BSR Annual Conference 2014

LIVERPOOL, ENGLAND – Invasive temporal arterial biopsy could be replaced by cranial Doppler ultrasound for the diagnosis of giant cell arteritis, according to research presented at the British Society for Rheumatology annual conference.

Cranial Doppler ultrasound (CDUS) more accurately diagnosed giant cell arteritis (GCA) compared with a clinical diagnosis at 3 months, with 81% sensitivity and 98% specificity. Temporal arterial biopsy was associated with a sensitivity of 53%, although it was 100% specific.

The noninvasive test also had high positive and negative predictive values of 97% and 88%, respectively, compared with 100% and just 47% for temporary artery biopsy.

GCA is one of the most common types of vasculitis and is characterized by inflammation of the medium and large arteries, usually in the head and neck. Permanent vision loss and ischemic stroke are some of the more serious symptoms of GCA, which is also linked to the development of polymyalgia rheumatica in around 50% of cases. The latter can manifest around the same time as those of GCA or develop shortly after.

"The inflammation in the temporal artery, or indeed in any large vessel affected by GCA is segmental in nature," said researcher Adam Croft, MRCP, Ph.D., of the Queen Elizabeth Hospital, Birmingham, England.

"So of course that means that if you stick a biopsy needle in randomly, as happens in temporal artery biopsy, then there is a good chance that you’ll end up with normal histology even in an affected artery." Furthermore, it is important to bear in mind that some of the characteristics of the inflammation seen in GCA have been observed in other vasculitic conditions, the researcher said.

Catching the inflammation early and treating with high-dose steroids can be "sight saving," but until now it has been difficult to determine if all patients with suspected GCA really need such treatment, and there is a high chance of over diagnosis and treatment.

Dr. Croft and colleagues conducted a retrospective study of 87 individuals with suspected GCA who underwent CDUS between January 2005 and July 2013. The aim was to see if CDUS could improve upon temporal arterial biopsy when compared with the standard of a clinical diagnosis at 3 months defined using American College of Rheumatology (ACR) criteria for GCA.

At the 3-month follow-up, 36 patients had a clinical diagnosis of GCA and 51 did not. All patients had undergone CDUS and of these, 30 (34%) had a halo sign present. The halo sign is a characteristic imaging feature used to detect GCA on ultrasound, but may be present in healthy individuals and in those with other inflammatory vascular disorders.

Temporal artery biopsy was performed in just over a quarter of all study subjects, 13 (43%) of whom had a halo sign present and 11 (19%) who did not. Biopsies were positive in 7 (54%) of individuals with a halo sign present and in one (9%) of those without the sign.

Dr. Croft observed that CDUS was the strongest predictor for a diagnosis of GCA at 3 months, whereas clinical symptoms using the ACR criteria were insufficiently specific to accurately rule in or rule out a diagnosis.

CDUS may be particularly useful in patients at either end of the clinical spectrum, he added; that is where the pretest clinical probability is very high or very low, and perhaps biopsy can be avoided in these patients.

Data on the discontinuation of steroids following a negative CDUS result in the study were compared with historical findings on steroid withdrawal in patients who had a negative biopsy. Significantly more (87% vs. 22%, P less than .01) patients with a negative ultrasound than biopsy stopped this treatment as a result, "suggesting that the clinician confidence in a negative ultrasound result is much greater, and they were more actively discontinuing steroids in these patients," said Dr. Croft.

For ultrasound to take the place of biopsy, however, he noted that an experienced ultrasonographer would be needed be able to identify the often subtle GCA changes, and that there was a short time frame when ultrasound would be accurate.

"The sensitivity rapidly falls after 4-5 days," the researcher noted, due to the use of steroids. "So you have to have the infrastructure to get these scans done early."

Dr. Croft had no conflicts of interest.

LIVERPOOL, ENGLAND – Invasive temporal arterial biopsy could be replaced by cranial Doppler ultrasound for the diagnosis of giant cell arteritis, according to research presented at the British Society for Rheumatology annual conference.

Cranial Doppler ultrasound (CDUS) more accurately diagnosed giant cell arteritis (GCA) compared with a clinical diagnosis at 3 months, with 81% sensitivity and 98% specificity. Temporal arterial biopsy was associated with a sensitivity of 53%, although it was 100% specific.

The noninvasive test also had high positive and negative predictive values of 97% and 88%, respectively, compared with 100% and just 47% for temporary artery biopsy.

GCA is one of the most common types of vasculitis and is characterized by inflammation of the medium and large arteries, usually in the head and neck. Permanent vision loss and ischemic stroke are some of the more serious symptoms of GCA, which is also linked to the development of polymyalgia rheumatica in around 50% of cases. The latter can manifest around the same time as those of GCA or develop shortly after.

"The inflammation in the temporal artery, or indeed in any large vessel affected by GCA is segmental in nature," said researcher Adam Croft, MRCP, Ph.D., of the Queen Elizabeth Hospital, Birmingham, England.

"So of course that means that if you stick a biopsy needle in randomly, as happens in temporal artery biopsy, then there is a good chance that you’ll end up with normal histology even in an affected artery." Furthermore, it is important to bear in mind that some of the characteristics of the inflammation seen in GCA have been observed in other vasculitic conditions, the researcher said.

Catching the inflammation early and treating with high-dose steroids can be "sight saving," but until now it has been difficult to determine if all patients with suspected GCA really need such treatment, and there is a high chance of over diagnosis and treatment.

Dr. Croft and colleagues conducted a retrospective study of 87 individuals with suspected GCA who underwent CDUS between January 2005 and July 2013. The aim was to see if CDUS could improve upon temporal arterial biopsy when compared with the standard of a clinical diagnosis at 3 months defined using American College of Rheumatology (ACR) criteria for GCA.

At the 3-month follow-up, 36 patients had a clinical diagnosis of GCA and 51 did not. All patients had undergone CDUS and of these, 30 (34%) had a halo sign present. The halo sign is a characteristic imaging feature used to detect GCA on ultrasound, but may be present in healthy individuals and in those with other inflammatory vascular disorders.

Temporal artery biopsy was performed in just over a quarter of all study subjects, 13 (43%) of whom had a halo sign present and 11 (19%) who did not. Biopsies were positive in 7 (54%) of individuals with a halo sign present and in one (9%) of those without the sign.

Dr. Croft observed that CDUS was the strongest predictor for a diagnosis of GCA at 3 months, whereas clinical symptoms using the ACR criteria were insufficiently specific to accurately rule in or rule out a diagnosis.

CDUS may be particularly useful in patients at either end of the clinical spectrum, he added; that is where the pretest clinical probability is very high or very low, and perhaps biopsy can be avoided in these patients.

Data on the discontinuation of steroids following a negative CDUS result in the study were compared with historical findings on steroid withdrawal in patients who had a negative biopsy. Significantly more (87% vs. 22%, P less than .01) patients with a negative ultrasound than biopsy stopped this treatment as a result, "suggesting that the clinician confidence in a negative ultrasound result is much greater, and they were more actively discontinuing steroids in these patients," said Dr. Croft.

For ultrasound to take the place of biopsy, however, he noted that an experienced ultrasonographer would be needed be able to identify the often subtle GCA changes, and that there was a short time frame when ultrasound would be accurate.

"The sensitivity rapidly falls after 4-5 days," the researcher noted, due to the use of steroids. "So you have to have the infrastructure to get these scans done early."

Dr. Croft had no conflicts of interest.

LIVERPOOL, ENGLAND – Invasive temporal arterial biopsy could be replaced by cranial Doppler ultrasound for the diagnosis of giant cell arteritis, according to research presented at the British Society for Rheumatology annual conference.

Cranial Doppler ultrasound (CDUS) more accurately diagnosed giant cell arteritis (GCA) compared with a clinical diagnosis at 3 months, with 81% sensitivity and 98% specificity. Temporal arterial biopsy was associated with a sensitivity of 53%, although it was 100% specific.

The noninvasive test also had high positive and negative predictive values of 97% and 88%, respectively, compared with 100% and just 47% for temporary artery biopsy.

GCA is one of the most common types of vasculitis and is characterized by inflammation of the medium and large arteries, usually in the head and neck. Permanent vision loss and ischemic stroke are some of the more serious symptoms of GCA, which is also linked to the development of polymyalgia rheumatica in around 50% of cases. The latter can manifest around the same time as those of GCA or develop shortly after.

"The inflammation in the temporal artery, or indeed in any large vessel affected by GCA is segmental in nature," said researcher Adam Croft, MRCP, Ph.D., of the Queen Elizabeth Hospital, Birmingham, England.

"So of course that means that if you stick a biopsy needle in randomly, as happens in temporal artery biopsy, then there is a good chance that you’ll end up with normal histology even in an affected artery." Furthermore, it is important to bear in mind that some of the characteristics of the inflammation seen in GCA have been observed in other vasculitic conditions, the researcher said.

Catching the inflammation early and treating with high-dose steroids can be "sight saving," but until now it has been difficult to determine if all patients with suspected GCA really need such treatment, and there is a high chance of over diagnosis and treatment.

Dr. Croft and colleagues conducted a retrospective study of 87 individuals with suspected GCA who underwent CDUS between January 2005 and July 2013. The aim was to see if CDUS could improve upon temporal arterial biopsy when compared with the standard of a clinical diagnosis at 3 months defined using American College of Rheumatology (ACR) criteria for GCA.

At the 3-month follow-up, 36 patients had a clinical diagnosis of GCA and 51 did not. All patients had undergone CDUS and of these, 30 (34%) had a halo sign present. The halo sign is a characteristic imaging feature used to detect GCA on ultrasound, but may be present in healthy individuals and in those with other inflammatory vascular disorders.

Temporal artery biopsy was performed in just over a quarter of all study subjects, 13 (43%) of whom had a halo sign present and 11 (19%) who did not. Biopsies were positive in 7 (54%) of individuals with a halo sign present and in one (9%) of those without the sign.

Dr. Croft observed that CDUS was the strongest predictor for a diagnosis of GCA at 3 months, whereas clinical symptoms using the ACR criteria were insufficiently specific to accurately rule in or rule out a diagnosis.

CDUS may be particularly useful in patients at either end of the clinical spectrum, he added; that is where the pretest clinical probability is very high or very low, and perhaps biopsy can be avoided in these patients.

Data on the discontinuation of steroids following a negative CDUS result in the study were compared with historical findings on steroid withdrawal in patients who had a negative biopsy. Significantly more (87% vs. 22%, P less than .01) patients with a negative ultrasound than biopsy stopped this treatment as a result, "suggesting that the clinician confidence in a negative ultrasound result is much greater, and they were more actively discontinuing steroids in these patients," said Dr. Croft.

For ultrasound to take the place of biopsy, however, he noted that an experienced ultrasonographer would be needed be able to identify the often subtle GCA changes, and that there was a short time frame when ultrasound would be accurate.

"The sensitivity rapidly falls after 4-5 days," the researcher noted, due to the use of steroids. "So you have to have the infrastructure to get these scans done early."

Dr. Croft had no conflicts of interest.

LIVERPOOL, ENGLAND – Hydroxychloroquine was not associated with any adverse long-term complications in the children of women with systemic lupus erythematosus who used the drug during pregnancy or while breastfeeding in a multicenter, cross-sectional survey.

In fact the research, recently reported at the British Society for Rheumatology annual conference, showed that the disease-modifying antirheumatic drug (DMARD) reduced exposed children’s risk of admission to hospital for infection (14% vs. 24.8% for nonexposed children; P = .03) and the need for outpatient visits (12.6% vs. 23.1%; P = .0.03).

There was no significant increased risk for congenital anomalies (2.1% vs. 2.2%), including congenital heart block (2.0% vs. 2.9%), or developmental problems such as attention-deficit hyperactivity disorder (0% vs. 2.9% for ADHD). Mothers who used the DMARD were also no more at risk of developing eclampsia during their pregnancy than those who did not (11% vs. 9%; P = .75).

©Jupiterimages/thinkstockphotos.com

"Hydroxychloroquine should be continued during pregnancy and breastfeeding," advised study investigator Dr. Mary Gayed of the department of rheumatology at the University of Birmingham (England).

Dr. Gayed, who presented the study findings on behalf of the British Isles Lupus Assessment Group (BILAG), said that these were hopefully reassuring findings. "We know that immunosuppressive agents, including hydroxychloroquine, are used in SLE [systemic lupus erythematosus] during pregnancy to prevent disease flare and to ensure optimum outcomes for both mother and child," she said, explaining the rationale behind the research, "but there are few published data regarding the long-term outcome of these children."

Randomized, controlled study data have shown that women who continue treatment with hydroxychloroquine during their pregnancy exhibit no disease activity and require only low doses of steroids by the end of their pregnancy, Dr. Gayed observed. This has been corroborated by other studies that have shown that discontinuing the DMARD leads to higher disease activity and the need for higher steroid doses.

Lack of good disease control during pregnancy is linked to maternal complications such as high blood pressure, proteinuria, thrombocytopenia, and secondary antiphospholipid syndrome. "We also know that women who have active lupus during pregnancy have a reduced rate of live births, earlier deliveries, increased pregnancy losses, and increased small-for-gestational-age babies, which makes it key to control disease activity throughout pregnancy," said Dr. Gayed.

Dr. Gayed and associates have previously reported their findings on the long-term out comes of all DMARDs used to treat SLE in pregnancy and will release further data from their survey on the safety of azathioprine at the upcoming European League Against Rheumatism Congress next month in Paris. The current research question was to look at their data specifically with regard to hydroxychloroquine and how this affected maternal, neonatal, and childhood outcomes.

Their retrospective survey involved 200 women with SLE, defined using American College of Rheumatology criteria, who gave birth to a total of 287 children. Of these, 118 women used the DMARD during their pregnancy or while breastfeeding, with 149 children exposed to hydroxychloroquine during the neonatal or postnatal periods. The majority of women (n = 76) used the drug during pregnancy and breastfeeding, 42 during pregnancy only, and 4 only while they were breastfeeding. There were 102 women who did not use the drug during pregnancy or breastfeeding and who gave birth to 138 nonexposed children.

At the birth of their children, the mean age of mothers was 32 years, and they had a mean disease duration of 7.5 years overall, with a mean maternal disease duration of 6.5 years versus 8.6 years comparing the hydroxychloroquine-exposed with the nonexposed children. The majority (66%) of the women were white, 15% were South Asian, 10% Afro-Caribbean, 1% Chinese, 1% Hispanic, and the remainder of other or unstated ethnicity. There was no significant difference in the use of azathioprine, aspirin, or heparin, but mothers who used hydroxychloroquine were significantly more likely to use steroids than women who did not use the DMARD (66% vs. 51%; P = .02).

The median gestational age at delivery was 38 weeks in both hydroxychloroquine-exposed and nonexposed children. There was no statistical difference in birth weight between the groups, which was a mean of 2.8 kg in both exposed and nonexposed children. The median age of children exposed to hydroxychloroquine at enrollment was 2.1 years versus 4.6 years for nonexposed children.

The findings are limited by the fact that these are self-reported data, and it was not always possible to check medical records, Dr. Gayed said. Data on disease activity and prednisolone dose were also not available. That said, "this U.K. cross-sectional survey highlights that hydroxychloroquine is compatible with pregnancy," Dr. Gayed concluded.

Dr. Gayed and coauthors reported no conflicts of interest.

LIVERPOOL, ENGLAND – Hydroxychloroquine was not associated with any adverse long-term complications in the children of women with systemic lupus erythematosus who used the drug during pregnancy or while breastfeeding in a multicenter, cross-sectional survey.

In fact the research, recently reported at the British Society for Rheumatology annual conference, showed that the disease-modifying antirheumatic drug (DMARD) reduced exposed children’s risk of admission to hospital for infection (14% vs. 24.8% for nonexposed children; P = .03) and the need for outpatient visits (12.6% vs. 23.1%; P = .0.03).

There was no significant increased risk for congenital anomalies (2.1% vs. 2.2%), including congenital heart block (2.0% vs. 2.9%), or developmental problems such as attention-deficit hyperactivity disorder (0% vs. 2.9% for ADHD). Mothers who used the DMARD were also no more at risk of developing eclampsia during their pregnancy than those who did not (11% vs. 9%; P = .75).

©Jupiterimages/thinkstockphotos.com

"Hydroxychloroquine should be continued during pregnancy and breastfeeding," advised study investigator Dr. Mary Gayed of the department of rheumatology at the University of Birmingham (England).

Dr. Gayed, who presented the study findings on behalf of the British Isles Lupus Assessment Group (BILAG), said that these were hopefully reassuring findings. "We know that immunosuppressive agents, including hydroxychloroquine, are used in SLE [systemic lupus erythematosus] during pregnancy to prevent disease flare and to ensure optimum outcomes for both mother and child," she said, explaining the rationale behind the research, "but there are few published data regarding the long-term outcome of these children."

Randomized, controlled study data have shown that women who continue treatment with hydroxychloroquine during their pregnancy exhibit no disease activity and require only low doses of steroids by the end of their pregnancy, Dr. Gayed observed. This has been corroborated by other studies that have shown that discontinuing the DMARD leads to higher disease activity and the need for higher steroid doses.

Lack of good disease control during pregnancy is linked to maternal complications such as high blood pressure, proteinuria, thrombocytopenia, and secondary antiphospholipid syndrome. "We also know that women who have active lupus during pregnancy have a reduced rate of live births, earlier deliveries, increased pregnancy losses, and increased small-for-gestational-age babies, which makes it key to control disease activity throughout pregnancy," said Dr. Gayed.

Dr. Gayed and associates have previously reported their findings on the long-term out comes of all DMARDs used to treat SLE in pregnancy and will release further data from their survey on the safety of azathioprine at the upcoming European League Against Rheumatism Congress next month in Paris. The current research question was to look at their data specifically with regard to hydroxychloroquine and how this affected maternal, neonatal, and childhood outcomes.

Their retrospective survey involved 200 women with SLE, defined using American College of Rheumatology criteria, who gave birth to a total of 287 children. Of these, 118 women used the DMARD during their pregnancy or while breastfeeding, with 149 children exposed to hydroxychloroquine during the neonatal or postnatal periods. The majority of women (n = 76) used the drug during pregnancy and breastfeeding, 42 during pregnancy only, and 4 only while they were breastfeeding. There were 102 women who did not use the drug during pregnancy or breastfeeding and who gave birth to 138 nonexposed children.

At the birth of their children, the mean age of mothers was 32 years, and they had a mean disease duration of 7.5 years overall, with a mean maternal disease duration of 6.5 years versus 8.6 years comparing the hydroxychloroquine-exposed with the nonexposed children. The majority (66%) of the women were white, 15% were South Asian, 10% Afro-Caribbean, 1% Chinese, 1% Hispanic, and the remainder of other or unstated ethnicity. There was no significant difference in the use of azathioprine, aspirin, or heparin, but mothers who used hydroxychloroquine were significantly more likely to use steroids than women who did not use the DMARD (66% vs. 51%; P = .02).

The median gestational age at delivery was 38 weeks in both hydroxychloroquine-exposed and nonexposed children. There was no statistical difference in birth weight between the groups, which was a mean of 2.8 kg in both exposed and nonexposed children. The median age of children exposed to hydroxychloroquine at enrollment was 2.1 years versus 4.6 years for nonexposed children.

The findings are limited by the fact that these are self-reported data, and it was not always possible to check medical records, Dr. Gayed said. Data on disease activity and prednisolone dose were also not available. That said, "this U.K. cross-sectional survey highlights that hydroxychloroquine is compatible with pregnancy," Dr. Gayed concluded.

Dr. Gayed and coauthors reported no conflicts of interest.

LIVERPOOL, ENGLAND – Hydroxychloroquine was not associated with any adverse long-term complications in the children of women with systemic lupus erythematosus who used the drug during pregnancy or while breastfeeding in a multicenter, cross-sectional survey.

In fact the research, recently reported at the British Society for Rheumatology annual conference, showed that the disease-modifying antirheumatic drug (DMARD) reduced exposed children’s risk of admission to hospital for infection (14% vs. 24.8% for nonexposed children; P = .03) and the need for outpatient visits (12.6% vs. 23.1%; P = .0.03).

There was no significant increased risk for congenital anomalies (2.1% vs. 2.2%), including congenital heart block (2.0% vs. 2.9%), or developmental problems such as attention-deficit hyperactivity disorder (0% vs. 2.9% for ADHD). Mothers who used the DMARD were also no more at risk of developing eclampsia during their pregnancy than those who did not (11% vs. 9%; P = .75).

©Jupiterimages/thinkstockphotos.com

"Hydroxychloroquine should be continued during pregnancy and breastfeeding," advised study investigator Dr. Mary Gayed of the department of rheumatology at the University of Birmingham (England).

Dr. Gayed, who presented the study findings on behalf of the British Isles Lupus Assessment Group (BILAG), said that these were hopefully reassuring findings. "We know that immunosuppressive agents, including hydroxychloroquine, are used in SLE [systemic lupus erythematosus] during pregnancy to prevent disease flare and to ensure optimum outcomes for both mother and child," she said, explaining the rationale behind the research, "but there are few published data regarding the long-term outcome of these children."

Randomized, controlled study data have shown that women who continue treatment with hydroxychloroquine during their pregnancy exhibit no disease activity and require only low doses of steroids by the end of their pregnancy, Dr. Gayed observed. This has been corroborated by other studies that have shown that discontinuing the DMARD leads to higher disease activity and the need for higher steroid doses.

Lack of good disease control during pregnancy is linked to maternal complications such as high blood pressure, proteinuria, thrombocytopenia, and secondary antiphospholipid syndrome. "We also know that women who have active lupus during pregnancy have a reduced rate of live births, earlier deliveries, increased pregnancy losses, and increased small-for-gestational-age babies, which makes it key to control disease activity throughout pregnancy," said Dr. Gayed.

Dr. Gayed and associates have previously reported their findings on the long-term out comes of all DMARDs used to treat SLE in pregnancy and will release further data from their survey on the safety of azathioprine at the upcoming European League Against Rheumatism Congress next month in Paris. The current research question was to look at their data specifically with regard to hydroxychloroquine and how this affected maternal, neonatal, and childhood outcomes.

Their retrospective survey involved 200 women with SLE, defined using American College of Rheumatology criteria, who gave birth to a total of 287 children. Of these, 118 women used the DMARD during their pregnancy or while breastfeeding, with 149 children exposed to hydroxychloroquine during the neonatal or postnatal periods. The majority of women (n = 76) used the drug during pregnancy and breastfeeding, 42 during pregnancy only, and 4 only while they were breastfeeding. There were 102 women who did not use the drug during pregnancy or breastfeeding and who gave birth to 138 nonexposed children.

At the birth of their children, the mean age of mothers was 32 years, and they had a mean disease duration of 7.5 years overall, with a mean maternal disease duration of 6.5 years versus 8.6 years comparing the hydroxychloroquine-exposed with the nonexposed children. The majority (66%) of the women were white, 15% were South Asian, 10% Afro-Caribbean, 1% Chinese, 1% Hispanic, and the remainder of other or unstated ethnicity. There was no significant difference in the use of azathioprine, aspirin, or heparin, but mothers who used hydroxychloroquine were significantly more likely to use steroids than women who did not use the DMARD (66% vs. 51%; P = .02).

The median gestational age at delivery was 38 weeks in both hydroxychloroquine-exposed and nonexposed children. There was no statistical difference in birth weight between the groups, which was a mean of 2.8 kg in both exposed and nonexposed children. The median age of children exposed to hydroxychloroquine at enrollment was 2.1 years versus 4.6 years for nonexposed children.

The findings are limited by the fact that these are self-reported data, and it was not always possible to check medical records, Dr. Gayed said. Data on disease activity and prednisolone dose were also not available. That said, "this U.K. cross-sectional survey highlights that hydroxychloroquine is compatible with pregnancy," Dr. Gayed concluded.

Dr. Gayed and coauthors reported no conflicts of interest.

LIVERPOOL, ENGLAND – Further evidence that apremilast has multiple and sustained effects in psoriatic arthritis for at least 1 year were reported at the British Society for Rheumatology annual conference.

Data from two of the phase III studies that make up the PALACE (Psoriatic Arthritis Long-term Assessment of Clinical Efficacy) clinical trials program showed that the oral phosphodiesterase 4 inhibitor continued to suppress disease activity and improved pain and physical functioning associated with the skin and joint condition.

In the PALACE 1 (Psoriatic Arthritis Long-term Assessment of Clinical Efficacy 1) trial, 75 of 119 (63%) patients treated with apremilast 20 mg twice daily and 71 of 130 (55%) patients treated with apremilast 30 mg twice daily still had an American College of Rheumatology 20% (ACR20) response at 52 weeks.

Corresponding 52-week data from the PALACE 3 trial showed ACR20 responses in 56% of 116 patients and 62% of 127 patients for the two doses of apremilast, respectively. ACR50 and ACR70 were achieved by 25%-30% of patients, and 9%-10% achieved an ACR70.

Apremilast is marketed by Celgene as Otezla and gained approval from the Food and Drug Administration for the treatment of adults with active psoriatic arthritis in March. The FDA approval came with a caution that patients’ weight should be monitored regularly and that there had been increased reports of depression.

The long-term safety data from these two PALACE trials, however, showed no new safety concerns, according to the studies’ authors. The most common side effects seen were as reported previously and included diarrhea, nausea, and headache.

PALACE 1 and PALACE 3 were designed to assess the efficacy and safety of apremilast versus placebo in patients who had active psoriatic arthritis despite treatment with conventional nonbiologic disease-modifying antirheumatic drugs (nbDMARDs) with or without additional biologic treatment.

For inclusion in the trials, patients had to have disease duration of at least 6 months, meet Classification of Psoriatic Arthritis (CASPAR) criteria, and have three or more tender or swollen joints involved. Patients enrolled in PALACE 3 also had to have skin involvement, with at least plaque psoriasis of 2 cm in size or larger.

A total of 504 patients were enrolled into PALACE 1 (Ann. Rheum. Dis. 2013;72[Suppl. 3]:163) and 505 into PALACE 3 and initially randomized to twice-daily treatment with placebo or one of two doses of apremilast for 24 weeks. At this point, all patients in the placebo arm who had not already been rerandomized to active treatment were randomized to apremilast 20 mg or 30 mg. Patients originally randomized to active therapy continued with their treatment if they continued to respond.

"I think what comes out of these [data], and looking over time, is that the effect is sustained with all the caveats that go with that because, of course, people only stay on [the drug] if they are doing well," PALACE 3 investigator Dr. Christopher Edwards observed in an interview.

Dr. Edwards, who is a clinical rheumatologist at University Hospital Southampton, England, noted that the statistical power of long-term extension studies is perhaps a little less robust than the initial study phases, but that these long-term data from the PALACE trials do seem to show that the effects of treatment with apremilast are sustained and maybe even that the effect size improves slightly over time.

Physical function, assessed via the Health Assessment Questionnaire–Disability Index (HAQ-DI), was improved with apremilast treatment in both the PALACE 1 and PALACE 3 trials. In the latter trial, mean change in HAQ-DI scores at 16 weeks declined by 0.13 and 0.19, compared with baseline values in the apremilast 20-mg and 30-mg groups. In contrast, scores increased by 0.07 in the placebo group. Over the following year, HAQ-DI scores continued to improve, with the mean change in scores from baseline crossing the threshold of –0.30, which many think signifies a clinically meaningful change, Dr. Edwards explained.

HAQ-DI scores in PALACE 1 at 52 weeks were a respective –0.369 and –0.318 in the 20-mg and 30-mg groups, and at the 16-week assessment point, they had been –0.20 and –0.24, respectively, and –0.09 for placebo.

Furthermore, in the PALACE 3 trial, 29%-39% of patients treated with apremilast achieved a 75% improvement in skin involvement, assessed via the Psoriasis Area and Severity Index (PASI) at 1 year. PASI50 was achieved in 49%-54%.

Dr. Edwards observed that some of the additional data now available from PALACE 3 showed that there was a beneficial effect of apremilast on enthesitis and dactylitis – two unique features of psoriatic arthritis that are often not treated by the use of the nbDMARDS, such as methotrexate and sulfasalazine.

At 24 weeks, he noted that in other comparisons of patients treated with apremilast or placebo, active therapy yielded significant improvements in pain (assessed using a visual analog scale), improved Functional Assessment of Chronic Illness Therapy-Fatigue scores, a higher rate of achieving ‘good’ or ‘moderate’ European League Against Rheumatism responses, and greater modified Psoriatic Arthritis Response Criteria results.

Dr. Edwards has acted as a consultant to Celgene, the company that sponsored the studies. He has also received research support from Pfizer and honoraria from Roche, Abbott, and GlaxoSmithKline.

LIVERPOOL, ENGLAND – Further evidence that apremilast has multiple and sustained effects in psoriatic arthritis for at least 1 year were reported at the British Society for Rheumatology annual conference.

Data from two of the phase III studies that make up the PALACE (Psoriatic Arthritis Long-term Assessment of Clinical Efficacy) clinical trials program showed that the oral phosphodiesterase 4 inhibitor continued to suppress disease activity and improved pain and physical functioning associated with the skin and joint condition.

In the PALACE 1 (Psoriatic Arthritis Long-term Assessment of Clinical Efficacy 1) trial, 75 of 119 (63%) patients treated with apremilast 20 mg twice daily and 71 of 130 (55%) patients treated with apremilast 30 mg twice daily still had an American College of Rheumatology 20% (ACR20) response at 52 weeks.

Corresponding 52-week data from the PALACE 3 trial showed ACR20 responses in 56% of 116 patients and 62% of 127 patients for the two doses of apremilast, respectively. ACR50 and ACR70 were achieved by 25%-30% of patients, and 9%-10% achieved an ACR70.

Apremilast is marketed by Celgene as Otezla and gained approval from the Food and Drug Administration for the treatment of adults with active psoriatic arthritis in March. The FDA approval came with a caution that patients’ weight should be monitored regularly and that there had been increased reports of depression.

The long-term safety data from these two PALACE trials, however, showed no new safety concerns, according to the studies’ authors. The most common side effects seen were as reported previously and included diarrhea, nausea, and headache.

PALACE 1 and PALACE 3 were designed to assess the efficacy and safety of apremilast versus placebo in patients who had active psoriatic arthritis despite treatment with conventional nonbiologic disease-modifying antirheumatic drugs (nbDMARDs) with or without additional biologic treatment.

For inclusion in the trials, patients had to have disease duration of at least 6 months, meet Classification of Psoriatic Arthritis (CASPAR) criteria, and have three or more tender or swollen joints involved. Patients enrolled in PALACE 3 also had to have skin involvement, with at least plaque psoriasis of 2 cm in size or larger.

A total of 504 patients were enrolled into PALACE 1 (Ann. Rheum. Dis. 2013;72[Suppl. 3]:163) and 505 into PALACE 3 and initially randomized to twice-daily treatment with placebo or one of two doses of apremilast for 24 weeks. At this point, all patients in the placebo arm who had not already been rerandomized to active treatment were randomized to apremilast 20 mg or 30 mg. Patients originally randomized to active therapy continued with their treatment if they continued to respond.

"I think what comes out of these [data], and looking over time, is that the effect is sustained with all the caveats that go with that because, of course, people only stay on [the drug] if they are doing well," PALACE 3 investigator Dr. Christopher Edwards observed in an interview.

Dr. Edwards, who is a clinical rheumatologist at University Hospital Southampton, England, noted that the statistical power of long-term extension studies is perhaps a little less robust than the initial study phases, but that these long-term data from the PALACE trials do seem to show that the effects of treatment with apremilast are sustained and maybe even that the effect size improves slightly over time.

Physical function, assessed via the Health Assessment Questionnaire–Disability Index (HAQ-DI), was improved with apremilast treatment in both the PALACE 1 and PALACE 3 trials. In the latter trial, mean change in HAQ-DI scores at 16 weeks declined by 0.13 and 0.19, compared with baseline values in the apremilast 20-mg and 30-mg groups. In contrast, scores increased by 0.07 in the placebo group. Over the following year, HAQ-DI scores continued to improve, with the mean change in scores from baseline crossing the threshold of –0.30, which many think signifies a clinically meaningful change, Dr. Edwards explained.

HAQ-DI scores in PALACE 1 at 52 weeks were a respective –0.369 and –0.318 in the 20-mg and 30-mg groups, and at the 16-week assessment point, they had been –0.20 and –0.24, respectively, and –0.09 for placebo.

Furthermore, in the PALACE 3 trial, 29%-39% of patients treated with apremilast achieved a 75% improvement in skin involvement, assessed via the Psoriasis Area and Severity Index (PASI) at 1 year. PASI50 was achieved in 49%-54%.

Dr. Edwards observed that some of the additional data now available from PALACE 3 showed that there was a beneficial effect of apremilast on enthesitis and dactylitis – two unique features of psoriatic arthritis that are often not treated by the use of the nbDMARDS, such as methotrexate and sulfasalazine.

At 24 weeks, he noted that in other comparisons of patients treated with apremilast or placebo, active therapy yielded significant improvements in pain (assessed using a visual analog scale), improved Functional Assessment of Chronic Illness Therapy-Fatigue scores, a higher rate of achieving ‘good’ or ‘moderate’ European League Against Rheumatism responses, and greater modified Psoriatic Arthritis Response Criteria results.

Dr. Edwards has acted as a consultant to Celgene, the company that sponsored the studies. He has also received research support from Pfizer and honoraria from Roche, Abbott, and GlaxoSmithKline.

LIVERPOOL, ENGLAND – Further evidence that apremilast has multiple and sustained effects in psoriatic arthritis for at least 1 year were reported at the British Society for Rheumatology annual conference.

Data from two of the phase III studies that make up the PALACE (Psoriatic Arthritis Long-term Assessment of Clinical Efficacy) clinical trials program showed that the oral phosphodiesterase 4 inhibitor continued to suppress disease activity and improved pain and physical functioning associated with the skin and joint condition.

In the PALACE 1 (Psoriatic Arthritis Long-term Assessment of Clinical Efficacy 1) trial, 75 of 119 (63%) patients treated with apremilast 20 mg twice daily and 71 of 130 (55%) patients treated with apremilast 30 mg twice daily still had an American College of Rheumatology 20% (ACR20) response at 52 weeks.

Corresponding 52-week data from the PALACE 3 trial showed ACR20 responses in 56% of 116 patients and 62% of 127 patients for the two doses of apremilast, respectively. ACR50 and ACR70 were achieved by 25%-30% of patients, and 9%-10% achieved an ACR70.

Apremilast is marketed by Celgene as Otezla and gained approval from the Food and Drug Administration for the treatment of adults with active psoriatic arthritis in March. The FDA approval came with a caution that patients’ weight should be monitored regularly and that there had been increased reports of depression.

The long-term safety data from these two PALACE trials, however, showed no new safety concerns, according to the studies’ authors. The most common side effects seen were as reported previously and included diarrhea, nausea, and headache.

PALACE 1 and PALACE 3 were designed to assess the efficacy and safety of apremilast versus placebo in patients who had active psoriatic arthritis despite treatment with conventional nonbiologic disease-modifying antirheumatic drugs (nbDMARDs) with or without additional biologic treatment.

For inclusion in the trials, patients had to have disease duration of at least 6 months, meet Classification of Psoriatic Arthritis (CASPAR) criteria, and have three or more tender or swollen joints involved. Patients enrolled in PALACE 3 also had to have skin involvement, with at least plaque psoriasis of 2 cm in size or larger.

A total of 504 patients were enrolled into PALACE 1 (Ann. Rheum. Dis. 2013;72[Suppl. 3]:163) and 505 into PALACE 3 and initially randomized to twice-daily treatment with placebo or one of two doses of apremilast for 24 weeks. At this point, all patients in the placebo arm who had not already been rerandomized to active treatment were randomized to apremilast 20 mg or 30 mg. Patients originally randomized to active therapy continued with their treatment if they continued to respond.

"I think what comes out of these [data], and looking over time, is that the effect is sustained with all the caveats that go with that because, of course, people only stay on [the drug] if they are doing well," PALACE 3 investigator Dr. Christopher Edwards observed in an interview.

Dr. Edwards, who is a clinical rheumatologist at University Hospital Southampton, England, noted that the statistical power of long-term extension studies is perhaps a little less robust than the initial study phases, but that these long-term data from the PALACE trials do seem to show that the effects of treatment with apremilast are sustained and maybe even that the effect size improves slightly over time.

Physical function, assessed via the Health Assessment Questionnaire–Disability Index (HAQ-DI), was improved with apremilast treatment in both the PALACE 1 and PALACE 3 trials. In the latter trial, mean change in HAQ-DI scores at 16 weeks declined by 0.13 and 0.19, compared with baseline values in the apremilast 20-mg and 30-mg groups. In contrast, scores increased by 0.07 in the placebo group. Over the following year, HAQ-DI scores continued to improve, with the mean change in scores from baseline crossing the threshold of –0.30, which many think signifies a clinically meaningful change, Dr. Edwards explained.

HAQ-DI scores in PALACE 1 at 52 weeks were a respective –0.369 and –0.318 in the 20-mg and 30-mg groups, and at the 16-week assessment point, they had been –0.20 and –0.24, respectively, and –0.09 for placebo.

Furthermore, in the PALACE 3 trial, 29%-39% of patients treated with apremilast achieved a 75% improvement in skin involvement, assessed via the Psoriasis Area and Severity Index (PASI) at 1 year. PASI50 was achieved in 49%-54%.

Dr. Edwards observed that some of the additional data now available from PALACE 3 showed that there was a beneficial effect of apremilast on enthesitis and dactylitis – two unique features of psoriatic arthritis that are often not treated by the use of the nbDMARDS, such as methotrexate and sulfasalazine.

At 24 weeks, he noted that in other comparisons of patients treated with apremilast or placebo, active therapy yielded significant improvements in pain (assessed using a visual analog scale), improved Functional Assessment of Chronic Illness Therapy-Fatigue scores, a higher rate of achieving ‘good’ or ‘moderate’ European League Against Rheumatism responses, and greater modified Psoriatic Arthritis Response Criteria results.

Dr. Edwards has acted as a consultant to Celgene, the company that sponsored the studies. He has also received research support from Pfizer and honoraria from Roche, Abbott, and GlaxoSmithKline.

LIVERPOOL, ENGLAND – The risk of atypical femoral fractures with long-term bisphosphonate treatment may be relatively low overall, but clinicians need to be aware, experts said at the recent British Society for Rheumatology annual conference.

The topic was debated during an Osteoporosis Special Interest Group Session, with the motion put forth that the risk of atypical femoral fractures (AFFs) with bisphosphonates was a "ticking time bomb about to explode." While that was a rather extreme view to support, there is still cause for concern, according to Dr. Michael Stone, who examined the case in favor of the motion.

How common are atypical femoral fractures?

"The estimated prevalence is probably 2%-3% in patients taking bisphosphonates for 5 years or more," suggested Dr. Stone, who is director of the Bone Research Unit at Cardiff University and a consultant at the University Hospital Llandough in Cardiff, Wales. Importantly, the risk might be highest in patients who are taking glucocorticoids, so in that population of patients the use of bisphosphonates may need to be more cautious.

"We always suspected atypical fractures might happen [with long-term bisphosphonate use] and now we know that they do," he said, noting that there was a plausible mechanism of action with antiresorptive agents."

"It’s a type of fracture we’ve not really seen before, except in the context of hypophosphatasia, and particularly in the context of steroids, and the risk may well outweigh the benefits," Dr. Stone observed.

However, Dr. Emma Clark, consultant senior lecturer at the University of Bristol and a consultant at Southmead Hospital in Bristol, England, countered that the absolute prevalence of these unusual hip fractures was low. In one Swedish population study, for example, the absolute rate was 5 per 10,000 patient-years (N. Engl. J. Med. 2011;364:1728-37). While there are lots of data, none of them are particularly convincing that AFF are an explosion waiting to happen, she argued.

Recently, the American Society for Bone and Mineral Research (ASBMR) reported that the absolute risk of AFFs in patients treated with bisphosphonates may range from 3.2 to 50 cases per 100,000 person-years (J. Bone Miner. Res. 2014;29:1-23). However, long-term use may be associated with higher risk (around 100 per 100,000 person-years).

There is some evidence that the risk of AFF rises with treatment duration, perhaps as high as 100 cases per 100,000 patient years, and the ASBMR and the Medicines and Healthcare Products Regulatory Agency (MHRA) in England have both issued guidance on the long-term use of bisphosphonates in this context.

Atypical fractures are less common than osteoporotic fractures, Dr. Stone noted, and while confidence limits are wide, in the worst-case scenario, the risk might outweigh the benefits after about 7 years’ continuous use, according to the same Swedish population study cited above.

What are atypical hip fractures?

Part of the problem of determining the extent of the problem lies in defining exactly what constitutes an atypical fracture. These fractures are "strikingly different" from the compression fractures that are commonly seen in patients with osteoporosis, Dr. Stone said, with the notable feature that they tend to affect the lateral rather than medial cortex.

"They are quite often bilateral, and they often occur at the same level; it is quite extraordinary," he said.

There is no distinct diagnostic code for these, Dr. Clark pointed out, so data routinely collected on fractures in national registries in the United Kingdom cannot be used and X-rays are needed to really look at the association between bisphosphonate use and their development. She observed that patients who are taking bisphosphonates should have a higher risk of all fractures and not just AFF, compared with the general population, because that is why these patients are being given the drugs in the first place.

The ASBMR developed a working definition of AFF in 2010, which stated that to be defined as atypical these fractures must meet certain criteria. These included their location in the subtrochanteric region and femoral shaft, associated with minimal or no trauma, a fracture line that originates in the lateral cortex and is transverse or short oblique in orientation, affects only the lateral cortex, with no comminution, and a medial spike when the fracture is complete (J. Bone Miner. Res. 2010;25:2267-94).

The society has recently updated their definition (J. Bone Miner. Res. 2014;29:1-23) based on evidence available since 2010 that suggests that AFFs are "stress or insufficiency fractures." The authors of the ASBMR report note "the original case definition was revised to highlight radiographic features that distinguish AFFs from ordinary osteoporotic femoral diaphyseal fractures and to provide guidance on the importance of their transverse orientation. The requirement that fractures be noncomminuted was relaxed to include minimal comminution."

Dr. Clark observed that the ASBMR task force conclusion was that a causal relationship between bisphosphonates and atypical femoral fractures has not been established, but data continue to accumulate.

What should clinicians do if AFF is suspected?

"A significant number of atypical fractures are asymptomatic prior to completion," Dr. Stone warned. He noted that he had a relatively low threshold for scanning someone if he suspected AFF, but advised using "a reasonable amount of common sense" in that one might consider it in someone who is experiencing any new groin, thigh, or hip pain, or if there is worsening pain. In such cases, he might initially perform radiography, and then, if the level of clinical suspicion remains high, proceed to magnetic resonance imaging (MRI) or an isotope bone scan.

In terms of managing a patient with osteoporosis who develops AFF while taking bisphosphonates and is at high risk for other fractures, it is very difficult to know what to do. Stopping bisphosphonates might need to be considered because these drugs can impair healing of the fracture. Use of an alternative, such as strontium or teriparatide might be considered if the risk of fracture remains high.

In fact, according to MHRA guidance, stopping bisphosphonate therapy should be considered in patients suspected of AFF while they are evaluated, and should be based on an assessment of the benefits and risks of continuing treatment. The regulatory body also suggests that the need to continue therapy should be periodically evaluated, particularly after 5 or more years of use.

Drug manufacturer information provides guidance on the use of bisphosphonates and should be consulted regarding stopping treatment, Dr. Clark advised.

In an interview, Dr. Stone noted that patients on glucocorticoids should probably stop bisphosphonate treatment much sooner than the recommended 5 years, perhaps after 2-3 years of use. Conceding that that was controversial, because that’s not what the MHRA guidelines say, he noted that "for patients on glucocorticoids, where in the longer term you are going to be suppressing bone turnover and where the evidence of bisphosphonates reducing the risk of nonvertebral fracture is pretty much nonexistent, the concern should be higher in my opinion, and one needs to be cautious committing everybody without careful consideration to bisphosphonates beyond 2 or 3 years."

Dr. Clark observed in an interview: "I would also say I think we use bisphosphonates slightly differently now. We don’t consider them a treatment for people at low risk; we don’t give them to osteopenic, perimenopausal women who have not had a fracture; and we do not say it is a treatment for life anymore. We say it is a treatment for a fixed period of time at which point we should reassess."

Dr. Clark added: "I think it is something that we should all be aware of and we shouldn’t just use these medications without considering the risks." With regard to informing patients, she noted that she tells them that these sorts of fractures are rare, and the fracture-reducing benefits of bisphosphonates in high-risk patients with osteoporosis far outweigh their potential to cause these unusual fractures.

Atypical femoral fracture register planned

"We shouldn’t throw the baby out with the bath water," commented Dr. Eugene McCloskey, professor of adult bone disease at the University of Sheffield and a consultant at the Northern General Hospital in Sheffield, England.

"Osteoporotic fractures will remain a big problem, but we can’t ignore the fact that we are seeing patients with these atypical fractures, and I think they are really a heterogeneous group of patients that we need to categorize better," Dr. McCloskey added, noting that there were plans to set up a national AFF register in the United Kingdom.

Dr. Stone, Dr. Clark, and Dr. McCloskey have received research support, honoraria, or acted as consultants for several pharmaceutical companies involved in bone research, including Eli Lilly and Co., who provided an unrestricted educational grant to fund the Osteoporosis Special Interest Group Session.

LIVERPOOL, ENGLAND – The risk of atypical femoral fractures with long-term bisphosphonate treatment may be relatively low overall, but clinicians need to be aware, experts said at the recent British Society for Rheumatology annual conference.

The topic was debated during an Osteoporosis Special Interest Group Session, with the motion put forth that the risk of atypical femoral fractures (AFFs) with bisphosphonates was a "ticking time bomb about to explode." While that was a rather extreme view to support, there is still cause for concern, according to Dr. Michael Stone, who examined the case in favor of the motion.

How common are atypical femoral fractures?

"The estimated prevalence is probably 2%-3% in patients taking bisphosphonates for 5 years or more," suggested Dr. Stone, who is director of the Bone Research Unit at Cardiff University and a consultant at the University Hospital Llandough in Cardiff, Wales. Importantly, the risk might be highest in patients who are taking glucocorticoids, so in that population of patients the use of bisphosphonates may need to be more cautious.

"We always suspected atypical fractures might happen [with long-term bisphosphonate use] and now we know that they do," he said, noting that there was a plausible mechanism of action with antiresorptive agents."

"It’s a type of fracture we’ve not really seen before, except in the context of hypophosphatasia, and particularly in the context of steroids, and the risk may well outweigh the benefits," Dr. Stone observed.

However, Dr. Emma Clark, consultant senior lecturer at the University of Bristol and a consultant at Southmead Hospital in Bristol, England, countered that the absolute prevalence of these unusual hip fractures was low. In one Swedish population study, for example, the absolute rate was 5 per 10,000 patient-years (N. Engl. J. Med. 2011;364:1728-37). While there are lots of data, none of them are particularly convincing that AFF are an explosion waiting to happen, she argued.

Recently, the American Society for Bone and Mineral Research (ASBMR) reported that the absolute risk of AFFs in patients treated with bisphosphonates may range from 3.2 to 50 cases per 100,000 person-years (J. Bone Miner. Res. 2014;29:1-23). However, long-term use may be associated with higher risk (around 100 per 100,000 person-years).

There is some evidence that the risk of AFF rises with treatment duration, perhaps as high as 100 cases per 100,000 patient years, and the ASBMR and the Medicines and Healthcare Products Regulatory Agency (MHRA) in England have both issued guidance on the long-term use of bisphosphonates in this context.

Atypical fractures are less common than osteoporotic fractures, Dr. Stone noted, and while confidence limits are wide, in the worst-case scenario, the risk might outweigh the benefits after about 7 years’ continuous use, according to the same Swedish population study cited above.

What are atypical hip fractures?

Part of the problem of determining the extent of the problem lies in defining exactly what constitutes an atypical fracture. These fractures are "strikingly different" from the compression fractures that are commonly seen in patients with osteoporosis, Dr. Stone said, with the notable feature that they tend to affect the lateral rather than medial cortex.

"They are quite often bilateral, and they often occur at the same level; it is quite extraordinary," he said.

There is no distinct diagnostic code for these, Dr. Clark pointed out, so data routinely collected on fractures in national registries in the United Kingdom cannot be used and X-rays are needed to really look at the association between bisphosphonate use and their development. She observed that patients who are taking bisphosphonates should have a higher risk of all fractures and not just AFF, compared with the general population, because that is why these patients are being given the drugs in the first place.

The ASBMR developed a working definition of AFF in 2010, which stated that to be defined as atypical these fractures must meet certain criteria. These included their location in the subtrochanteric region and femoral shaft, associated with minimal or no trauma, a fracture line that originates in the lateral cortex and is transverse or short oblique in orientation, affects only the lateral cortex, with no comminution, and a medial spike when the fracture is complete (J. Bone Miner. Res. 2010;25:2267-94).

The society has recently updated their definition (J. Bone Miner. Res. 2014;29:1-23) based on evidence available since 2010 that suggests that AFFs are "stress or insufficiency fractures." The authors of the ASBMR report note "the original case definition was revised to highlight radiographic features that distinguish AFFs from ordinary osteoporotic femoral diaphyseal fractures and to provide guidance on the importance of their transverse orientation. The requirement that fractures be noncomminuted was relaxed to include minimal comminution."

Dr. Clark observed that the ASBMR task force conclusion was that a causal relationship between bisphosphonates and atypical femoral fractures has not been established, but data continue to accumulate.

What should clinicians do if AFF is suspected?

"A significant number of atypical fractures are asymptomatic prior to completion," Dr. Stone warned. He noted that he had a relatively low threshold for scanning someone if he suspected AFF, but advised using "a reasonable amount of common sense" in that one might consider it in someone who is experiencing any new groin, thigh, or hip pain, or if there is worsening pain. In such cases, he might initially perform radiography, and then, if the level of clinical suspicion remains high, proceed to magnetic resonance imaging (MRI) or an isotope bone scan.

In terms of managing a patient with osteoporosis who develops AFF while taking bisphosphonates and is at high risk for other fractures, it is very difficult to know what to do. Stopping bisphosphonates might need to be considered because these drugs can impair healing of the fracture. Use of an alternative, such as strontium or teriparatide might be considered if the risk of fracture remains high.

In fact, according to MHRA guidance, stopping bisphosphonate therapy should be considered in patients suspected of AFF while they are evaluated, and should be based on an assessment of the benefits and risks of continuing treatment. The regulatory body also suggests that the need to continue therapy should be periodically evaluated, particularly after 5 or more years of use.

Drug manufacturer information provides guidance on the use of bisphosphonates and should be consulted regarding stopping treatment, Dr. Clark advised.

In an interview, Dr. Stone noted that patients on glucocorticoids should probably stop bisphosphonate treatment much sooner than the recommended 5 years, perhaps after 2-3 years of use. Conceding that that was controversial, because that’s not what the MHRA guidelines say, he noted that "for patients on glucocorticoids, where in the longer term you are going to be suppressing bone turnover and where the evidence of bisphosphonates reducing the risk of nonvertebral fracture is pretty much nonexistent, the concern should be higher in my opinion, and one needs to be cautious committing everybody without careful consideration to bisphosphonates beyond 2 or 3 years."

Dr. Clark observed in an interview: "I would also say I think we use bisphosphonates slightly differently now. We don’t consider them a treatment for people at low risk; we don’t give them to osteopenic, perimenopausal women who have not had a fracture; and we do not say it is a treatment for life anymore. We say it is a treatment for a fixed period of time at which point we should reassess."

Dr. Clark added: "I think it is something that we should all be aware of and we shouldn’t just use these medications without considering the risks." With regard to informing patients, she noted that she tells them that these sorts of fractures are rare, and the fracture-reducing benefits of bisphosphonates in high-risk patients with osteoporosis far outweigh their potential to cause these unusual fractures.

Atypical femoral fracture register planned

"We shouldn’t throw the baby out with the bath water," commented Dr. Eugene McCloskey, professor of adult bone disease at the University of Sheffield and a consultant at the Northern General Hospital in Sheffield, England.

"Osteoporotic fractures will remain a big problem, but we can’t ignore the fact that we are seeing patients with these atypical fractures, and I think they are really a heterogeneous group of patients that we need to categorize better," Dr. McCloskey added, noting that there were plans to set up a national AFF register in the United Kingdom.

Dr. Stone, Dr. Clark, and Dr. McCloskey have received research support, honoraria, or acted as consultants for several pharmaceutical companies involved in bone research, including Eli Lilly and Co., who provided an unrestricted educational grant to fund the Osteoporosis Special Interest Group Session.

LIVERPOOL, ENGLAND – The risk of atypical femoral fractures with long-term bisphosphonate treatment may be relatively low overall, but clinicians need to be aware, experts said at the recent British Society for Rheumatology annual conference.

The topic was debated during an Osteoporosis Special Interest Group Session, with the motion put forth that the risk of atypical femoral fractures (AFFs) with bisphosphonates was a "ticking time bomb about to explode." While that was a rather extreme view to support, there is still cause for concern, according to Dr. Michael Stone, who examined the case in favor of the motion.

How common are atypical femoral fractures?

"The estimated prevalence is probably 2%-3% in patients taking bisphosphonates for 5 years or more," suggested Dr. Stone, who is director of the Bone Research Unit at Cardiff University and a consultant at the University Hospital Llandough in Cardiff, Wales. Importantly, the risk might be highest in patients who are taking glucocorticoids, so in that population of patients the use of bisphosphonates may need to be more cautious.

"We always suspected atypical fractures might happen [with long-term bisphosphonate use] and now we know that they do," he said, noting that there was a plausible mechanism of action with antiresorptive agents."

"It’s a type of fracture we’ve not really seen before, except in the context of hypophosphatasia, and particularly in the context of steroids, and the risk may well outweigh the benefits," Dr. Stone observed.

However, Dr. Emma Clark, consultant senior lecturer at the University of Bristol and a consultant at Southmead Hospital in Bristol, England, countered that the absolute prevalence of these unusual hip fractures was low. In one Swedish population study, for example, the absolute rate was 5 per 10,000 patient-years (N. Engl. J. Med. 2011;364:1728-37). While there are lots of data, none of them are particularly convincing that AFF are an explosion waiting to happen, she argued.

Recently, the American Society for Bone and Mineral Research (ASBMR) reported that the absolute risk of AFFs in patients treated with bisphosphonates may range from 3.2 to 50 cases per 100,000 person-years (J. Bone Miner. Res. 2014;29:1-23). However, long-term use may be associated with higher risk (around 100 per 100,000 person-years).

There is some evidence that the risk of AFF rises with treatment duration, perhaps as high as 100 cases per 100,000 patient years, and the ASBMR and the Medicines and Healthcare Products Regulatory Agency (MHRA) in England have both issued guidance on the long-term use of bisphosphonates in this context.

Atypical fractures are less common than osteoporotic fractures, Dr. Stone noted, and while confidence limits are wide, in the worst-case scenario, the risk might outweigh the benefits after about 7 years’ continuous use, according to the same Swedish population study cited above.

What are atypical hip fractures?

Part of the problem of determining the extent of the problem lies in defining exactly what constitutes an atypical fracture. These fractures are "strikingly different" from the compression fractures that are commonly seen in patients with osteoporosis, Dr. Stone said, with the notable feature that they tend to affect the lateral rather than medial cortex.

"They are quite often bilateral, and they often occur at the same level; it is quite extraordinary," he said.

There is no distinct diagnostic code for these, Dr. Clark pointed out, so data routinely collected on fractures in national registries in the United Kingdom cannot be used and X-rays are needed to really look at the association between bisphosphonate use and their development. She observed that patients who are taking bisphosphonates should have a higher risk of all fractures and not just AFF, compared with the general population, because that is why these patients are being given the drugs in the first place.

The ASBMR developed a working definition of AFF in 2010, which stated that to be defined as atypical these fractures must meet certain criteria. These included their location in the subtrochanteric region and femoral shaft, associated with minimal or no trauma, a fracture line that originates in the lateral cortex and is transverse or short oblique in orientation, affects only the lateral cortex, with no comminution, and a medial spike when the fracture is complete (J. Bone Miner. Res. 2010;25:2267-94).

The society has recently updated their definition (J. Bone Miner. Res. 2014;29:1-23) based on evidence available since 2010 that suggests that AFFs are "stress or insufficiency fractures." The authors of the ASBMR report note "the original case definition was revised to highlight radiographic features that distinguish AFFs from ordinary osteoporotic femoral diaphyseal fractures and to provide guidance on the importance of their transverse orientation. The requirement that fractures be noncomminuted was relaxed to include minimal comminution."

Dr. Clark observed that the ASBMR task force conclusion was that a causal relationship between bisphosphonates and atypical femoral fractures has not been established, but data continue to accumulate.

What should clinicians do if AFF is suspected?

"A significant number of atypical fractures are asymptomatic prior to completion," Dr. Stone warned. He noted that he had a relatively low threshold for scanning someone if he suspected AFF, but advised using "a reasonable amount of common sense" in that one might consider it in someone who is experiencing any new groin, thigh, or hip pain, or if there is worsening pain. In such cases, he might initially perform radiography, and then, if the level of clinical suspicion remains high, proceed to magnetic resonance imaging (MRI) or an isotope bone scan.

In terms of managing a patient with osteoporosis who develops AFF while taking bisphosphonates and is at high risk for other fractures, it is very difficult to know what to do. Stopping bisphosphonates might need to be considered because these drugs can impair healing of the fracture. Use of an alternative, such as strontium or teriparatide might be considered if the risk of fracture remains high.

In fact, according to MHRA guidance, stopping bisphosphonate therapy should be considered in patients suspected of AFF while they are evaluated, and should be based on an assessment of the benefits and risks of continuing treatment. The regulatory body also suggests that the need to continue therapy should be periodically evaluated, particularly after 5 or more years of use.

Drug manufacturer information provides guidance on the use of bisphosphonates and should be consulted regarding stopping treatment, Dr. Clark advised.

In an interview, Dr. Stone noted that patients on glucocorticoids should probably stop bisphosphonate treatment much sooner than the recommended 5 years, perhaps after 2-3 years of use. Conceding that that was controversial, because that’s not what the MHRA guidelines say, he noted that "for patients on glucocorticoids, where in the longer term you are going to be suppressing bone turnover and where the evidence of bisphosphonates reducing the risk of nonvertebral fracture is pretty much nonexistent, the concern should be higher in my opinion, and one needs to be cautious committing everybody without careful consideration to bisphosphonates beyond 2 or 3 years."

Dr. Clark observed in an interview: "I would also say I think we use bisphosphonates slightly differently now. We don’t consider them a treatment for people at low risk; we don’t give them to osteopenic, perimenopausal women who have not had a fracture; and we do not say it is a treatment for life anymore. We say it is a treatment for a fixed period of time at which point we should reassess."

Dr. Clark added: "I think it is something that we should all be aware of and we shouldn’t just use these medications without considering the risks." With regard to informing patients, she noted that she tells them that these sorts of fractures are rare, and the fracture-reducing benefits of bisphosphonates in high-risk patients with osteoporosis far outweigh their potential to cause these unusual fractures.

Atypical femoral fracture register planned

"We shouldn’t throw the baby out with the bath water," commented Dr. Eugene McCloskey, professor of adult bone disease at the University of Sheffield and a consultant at the Northern General Hospital in Sheffield, England.

"Osteoporotic fractures will remain a big problem, but we can’t ignore the fact that we are seeing patients with these atypical fractures, and I think they are really a heterogeneous group of patients that we need to categorize better," Dr. McCloskey added, noting that there were plans to set up a national AFF register in the United Kingdom.

Dr. Stone, Dr. Clark, and Dr. McCloskey have received research support, honoraria, or acted as consultants for several pharmaceutical companies involved in bone research, including Eli Lilly and Co., who provided an unrestricted educational grant to fund the Osteoporosis Special Interest Group Session.

LIVERPOOL, ENGLAND – Rituximab and mycophenolate mofetil could offer patients with rheumatoid arthritis and concomitant interstitial lung disease a survival advantage over other therapies, according to findings from a large, retrospective, multicenter study.

Data collated by the British Rheumatoid Interstitial Lung (BRILL) network found all-cause mortality in patients with rheumatoid arthritis (RA) and interstitial lung disease (ILD) treated with rituximab was 8% versus 31% for those treated with anti-tumor necrosis factor (TNF) alpha therapies (P = .03). Death from respiratory causes was also significantly lower in the rituximab-treated patients (4% vs. 15%, P = .04).

All-cause mortality was also significantly lower in patients with RA and ILD treated with mycophenolate mofetil (relative risk, 0.65) when compared with azathioprine (RR, 1.42) or other immunosuppressants, which included cyclophosphamide (RR, 1.65) and tacrolimus (RR, 1.74).

However, these data were retrospective, so prospective therapeutic trials are now needed, said Dr. Clive Kelly of Queen Elizabeth Hospital, Gateshead, England. Dr. Kelly, who was the driving force behind the set up of the BRILL network 2 years ago, stated that plans were already afoot to set up these trials with the intention of recruiting 200 patients with RA and ILD across the United Kingdom.

The aim of future studies would be to compare the use of rituximab versus anti-TNF therapies, with or without methotrexate, and mycophenolate mofetil versus azathioprine, with or without oral steroids, to see if the anticipated survival benefit associated with the newer agents is confirmed. Funding for the trials is yet to be secured, but it is hoped that recruitment can begin next year.

When people think of comorbidity in RA, they tend to think about cardiovascular disease first, Dr. Kelly noted at the British Society for Rheumatology annual conference. He observed that ILD has been associated with RA for well over 50 years and that respiratory disease, largely due to ILD and chronic lung damage, is now a "close second" behind heart disease as a cause of death in patients with RA.

Up to 40% of all cases of ILD are discovered in RA patients postmortem, he said, but with the increased use of high-resolution computed tomography (HRCT) in the 1990s, around a quarter of patients with RA were found to have changes on lung scanning.

Today, the prevalence of clinically significant lung disease is estimated to be around 5% in RA, with a lifetime risk thought to be around 7.7%. Patients with RA who develop ILD are around three times more likely to die than are those who do not, with a reported mean survival around 2.6 years from the diagnosis of lung disease until the advent of these recent studies.

The BRILL network consists of 16 centers located throughout the United Kingdom and was set up to establish a national database of all cases of ILD in RA patients that had been reported by these centers over a 25-year period, beginning in 1988. Patients were included in the database if clinical symptoms, such as shortness of breath or chest crackles, or pulmonary function tests had indicated ILD, and abnormal findings had been confirmed via HRCT.

For the current study, Dr. Kelly and coinvestigators from the BRILL Network looked at a subset of 188 patients with RA and ILD who had been diagnosed between 2000 and 2012. A control group of 188 patients with RA without ILD were included for comparison, matched for age, sex, and duration of RA.

Patients with ILD were more likely than those without to use biologics or immunosuppressants. Of the 188 patients with ILD, 57 were treated with either rituximab or anti-TNF drugs and 83 were treated with immunosuppressants.

The median age at which ILD was identified was 64 years, ranging from 42 to 87 years. The median duration of RA at ILD diagnosis was 9 years and the median duration of ILD was 4 years.

Dr. Kelly reported that RA preceded lung disease in the majority (83%) of cases, while 10% of patients developed ILD before the articular features of their disease became apparent. The onset of RA and ILD was "synchronous" in the remaining 7%.

HRCT results showed that the most common subtype of lung disease was interstitial pneumonia in 65% of cases, which carries the worst overall prognosis, he said. Nonspecific interstitial pneumonia, which is more responsive to early therapy, accounted for 24% of cases. The remaining patients had cryptogenic organizing pneumonia (6%) and or mixed subtypes (5%).

Survival over the 25 years improved, with 67% of RA-ILD patients dying from ILD in 1987-1993, compared with 30% in 2006-2012 (P less than .01). Nevertheless, these data are retrospective, Dr. Kelly observed, and trials are warranted to confirm the effects of specific treatments. He encouraged delegates at the meeting to participate in forthcoming BRILL network studies.

Dr. Kelly had no disclosures.

LIVERPOOL, ENGLAND – Rituximab and mycophenolate mofetil could offer patients with rheumatoid arthritis and concomitant interstitial lung disease a survival advantage over other therapies, according to findings from a large, retrospective, multicenter study.

Data collated by the British Rheumatoid Interstitial Lung (BRILL) network found all-cause mortality in patients with rheumatoid arthritis (RA) and interstitial lung disease (ILD) treated with rituximab was 8% versus 31% for those treated with anti-tumor necrosis factor (TNF) alpha therapies (P = .03). Death from respiratory causes was also significantly lower in the rituximab-treated patients (4% vs. 15%, P = .04).

All-cause mortality was also significantly lower in patients with RA and ILD treated with mycophenolate mofetil (relative risk, 0.65) when compared with azathioprine (RR, 1.42) or other immunosuppressants, which included cyclophosphamide (RR, 1.65) and tacrolimus (RR, 1.74).

However, these data were retrospective, so prospective therapeutic trials are now needed, said Dr. Clive Kelly of Queen Elizabeth Hospital, Gateshead, England. Dr. Kelly, who was the driving force behind the set up of the BRILL network 2 years ago, stated that plans were already afoot to set up these trials with the intention of recruiting 200 patients with RA and ILD across the United Kingdom.

The aim of future studies would be to compare the use of rituximab versus anti-TNF therapies, with or without methotrexate, and mycophenolate mofetil versus azathioprine, with or without oral steroids, to see if the anticipated survival benefit associated with the newer agents is confirmed. Funding for the trials is yet to be secured, but it is hoped that recruitment can begin next year.

When people think of comorbidity in RA, they tend to think about cardiovascular disease first, Dr. Kelly noted at the British Society for Rheumatology annual conference. He observed that ILD has been associated with RA for well over 50 years and that respiratory disease, largely due to ILD and chronic lung damage, is now a "close second" behind heart disease as a cause of death in patients with RA.

Up to 40% of all cases of ILD are discovered in RA patients postmortem, he said, but with the increased use of high-resolution computed tomography (HRCT) in the 1990s, around a quarter of patients with RA were found to have changes on lung scanning.

Today, the prevalence of clinically significant lung disease is estimated to be around 5% in RA, with a lifetime risk thought to be around 7.7%. Patients with RA who develop ILD are around three times more likely to die than are those who do not, with a reported mean survival around 2.6 years from the diagnosis of lung disease until the advent of these recent studies.

The BRILL network consists of 16 centers located throughout the United Kingdom and was set up to establish a national database of all cases of ILD in RA patients that had been reported by these centers over a 25-year period, beginning in 1988. Patients were included in the database if clinical symptoms, such as shortness of breath or chest crackles, or pulmonary function tests had indicated ILD, and abnormal findings had been confirmed via HRCT.

For the current study, Dr. Kelly and coinvestigators from the BRILL Network looked at a subset of 188 patients with RA and ILD who had been diagnosed between 2000 and 2012. A control group of 188 patients with RA without ILD were included for comparison, matched for age, sex, and duration of RA.

Patients with ILD were more likely than those without to use biologics or immunosuppressants. Of the 188 patients with ILD, 57 were treated with either rituximab or anti-TNF drugs and 83 were treated with immunosuppressants.

The median age at which ILD was identified was 64 years, ranging from 42 to 87 years. The median duration of RA at ILD diagnosis was 9 years and the median duration of ILD was 4 years.

Dr. Kelly reported that RA preceded lung disease in the majority (83%) of cases, while 10% of patients developed ILD before the articular features of their disease became apparent. The onset of RA and ILD was "synchronous" in the remaining 7%.

HRCT results showed that the most common subtype of lung disease was interstitial pneumonia in 65% of cases, which carries the worst overall prognosis, he said. Nonspecific interstitial pneumonia, which is more responsive to early therapy, accounted for 24% of cases. The remaining patients had cryptogenic organizing pneumonia (6%) and or mixed subtypes (5%).

Survival over the 25 years improved, with 67% of RA-ILD patients dying from ILD in 1987-1993, compared with 30% in 2006-2012 (P less than .01). Nevertheless, these data are retrospective, Dr. Kelly observed, and trials are warranted to confirm the effects of specific treatments. He encouraged delegates at the meeting to participate in forthcoming BRILL network studies.

Dr. Kelly had no disclosures.

LIVERPOOL, ENGLAND – Rituximab and mycophenolate mofetil could offer patients with rheumatoid arthritis and concomitant interstitial lung disease a survival advantage over other therapies, according to findings from a large, retrospective, multicenter study.

Data collated by the British Rheumatoid Interstitial Lung (BRILL) network found all-cause mortality in patients with rheumatoid arthritis (RA) and interstitial lung disease (ILD) treated with rituximab was 8% versus 31% for those treated with anti-tumor necrosis factor (TNF) alpha therapies (P = .03). Death from respiratory causes was also significantly lower in the rituximab-treated patients (4% vs. 15%, P = .04).

All-cause mortality was also significantly lower in patients with RA and ILD treated with mycophenolate mofetil (relative risk, 0.65) when compared with azathioprine (RR, 1.42) or other immunosuppressants, which included cyclophosphamide (RR, 1.65) and tacrolimus (RR, 1.74).

However, these data were retrospective, so prospective therapeutic trials are now needed, said Dr. Clive Kelly of Queen Elizabeth Hospital, Gateshead, England. Dr. Kelly, who was the driving force behind the set up of the BRILL network 2 years ago, stated that plans were already afoot to set up these trials with the intention of recruiting 200 patients with RA and ILD across the United Kingdom.

The aim of future studies would be to compare the use of rituximab versus anti-TNF therapies, with or without methotrexate, and mycophenolate mofetil versus azathioprine, with or without oral steroids, to see if the anticipated survival benefit associated with the newer agents is confirmed. Funding for the trials is yet to be secured, but it is hoped that recruitment can begin next year.

When people think of comorbidity in RA, they tend to think about cardiovascular disease first, Dr. Kelly noted at the British Society for Rheumatology annual conference. He observed that ILD has been associated with RA for well over 50 years and that respiratory disease, largely due to ILD and chronic lung damage, is now a "close second" behind heart disease as a cause of death in patients with RA.

Up to 40% of all cases of ILD are discovered in RA patients postmortem, he said, but with the increased use of high-resolution computed tomography (HRCT) in the 1990s, around a quarter of patients with RA were found to have changes on lung scanning.

Today, the prevalence of clinically significant lung disease is estimated to be around 5% in RA, with a lifetime risk thought to be around 7.7%. Patients with RA who develop ILD are around three times more likely to die than are those who do not, with a reported mean survival around 2.6 years from the diagnosis of lung disease until the advent of these recent studies.

The BRILL network consists of 16 centers located throughout the United Kingdom and was set up to establish a national database of all cases of ILD in RA patients that had been reported by these centers over a 25-year period, beginning in 1988. Patients were included in the database if clinical symptoms, such as shortness of breath or chest crackles, or pulmonary function tests had indicated ILD, and abnormal findings had been confirmed via HRCT.

For the current study, Dr. Kelly and coinvestigators from the BRILL Network looked at a subset of 188 patients with RA and ILD who had been diagnosed between 2000 and 2012. A control group of 188 patients with RA without ILD were included for comparison, matched for age, sex, and duration of RA.

Patients with ILD were more likely than those without to use biologics or immunosuppressants. Of the 188 patients with ILD, 57 were treated with either rituximab or anti-TNF drugs and 83 were treated with immunosuppressants.

The median age at which ILD was identified was 64 years, ranging from 42 to 87 years. The median duration of RA at ILD diagnosis was 9 years and the median duration of ILD was 4 years.

Dr. Kelly reported that RA preceded lung disease in the majority (83%) of cases, while 10% of patients developed ILD before the articular features of their disease became apparent. The onset of RA and ILD was "synchronous" in the remaining 7%.

HRCT results showed that the most common subtype of lung disease was interstitial pneumonia in 65% of cases, which carries the worst overall prognosis, he said. Nonspecific interstitial pneumonia, which is more responsive to early therapy, accounted for 24% of cases. The remaining patients had cryptogenic organizing pneumonia (6%) and or mixed subtypes (5%).

Survival over the 25 years improved, with 67% of RA-ILD patients dying from ILD in 1987-1993, compared with 30% in 2006-2012 (P less than .01). Nevertheless, these data are retrospective, Dr. Kelly observed, and trials are warranted to confirm the effects of specific treatments. He encouraged delegates at the meeting to participate in forthcoming BRILL network studies.

Dr. Kelly had no disclosures.

LIVERPOOL, ENGLAND – Biologic therapy has shown sustained benefits for the signs and symptoms of juvenile idiopathic arthritis as well as improved height over the course of up to 2 years of treatment, according to data from four studies presented at the British Society for Rheumatology annual conference.

The studies provided evidence indicating that:

• Treatment with etanercept over 2 years helped children with juvenile idiopathic arthritis (JIA) to gain height, although their overall vertical growth still lagged behind children in the general reference population.

• When etanercept was not used as a first-line biologic, it was most common to use adalimumab instead, followed by tocilizumab and infliximab.

• Adalimumab reduced the signs and symptoms of enthesitis-related arthritis (ERA) in children by week 12 of treatment and the effects were sustained at 1-year of follow up.

• The benefits of tocilizumab in reducing disease activity to a minimal level were preserved from 40 weeks to 2 years in most patients.

Etanercept and growth in JIA

"Etanercept therapy was associated with improvement in height z score over the first 2 years of therapy," and the effect was most apparent in children who were not receiving concomitant steroids, said Lianne Kearsley-Fleet, who reported data from the British Society for Paediatric and Adolescent Rheumatology (BSPAR Etanercept Cohort Study). The prospective, observational study was initiated in 2004 with the aim of recruiting all children with JIA who were starting treatment with etanercept. For comparison, a cohort of children newly starting methotrexate was also included.

The current analysis focused on the growth of children taking the biologic therapy because children with JIA are known to have restricted growth, which may be linked to inflammation and the use of corticosteroids. Of 658 children registered in the study, 191 had height data available at baseline and at 1 and 2 years’ follow-up.

Mean height and changes in height during biologic therapy over time were assessed by calculating z scores, and compared with values set by the World Health Organization for girls and boys of a similar age who did not have JIA.

The mean age of children included in the current analysis was 11 years, with a mean disease duration of 3.5 years. The majority (65%) were female, with concomitant oral steroids in 38% and methotrexate in 58%.

The mean height z score of the study population was –0.67 at baseline, –0.57 after 1 year, and –0.45 after 2 years’ etanercept therapy. "While it is still negative [compared with children of a similar sex and age], there is still significant improvement," said Ms. Kearsley-Fleet, a research assistant from the Arthritis Research UK Centre for Epidemiology at the University of Manchester, England.

"There was a lack of evidence for association between disease activity and improved growth," the researcher noted, adding that "anti-TNF inhibitors alone may be insufficient to increase growth."

Use of non–etanercept biologics

In a poster presentation at the meeting, Ms. Kearsley-Fleet and her colleagues at her institution also presented the findings from the Biologics for Children with Rheumatic Diseases Study. This ongoing, prospective, observational cohort study is looking at the use of biologics other than etanercept in the treatment of JIA.

Since the study started in 2010, and up until early April 2014, a total of 280 children with JIA had been recruited into the biologic cohort. The study also includes a parallel control cohort of 295 children being treated with methotrexate. Just under half (46%) of the children recruited into the biologic cohort started a non–etanercept biologic as their first biologic treatment, of which the majority (75%) received the medications off label. The main alternatives to etanercept being used were adalimumab in 38% of cases, followed by tocilizumab in 25% of children, and infliximab in 23%. Other drugs used as the first biologic was anakinra (13%), rituximab (less than 1%), and abatacept (less than 1%).

Adalimumab was also a popular subsequent biologic choice in 30% of cases, with 20% of patients using tocilizumab, and 25% infliximab as a subsequent biologic choice.

The research team reported that patients using a biologic for the first time were significantly younger and had shorter disease durations than did subsequent biologic users. The mean ages of first-, second-, and third-line users were 9, 12, and 11 years, respectively, and the mean disease durations were 2, 6, and 4 years. Subsequent biologic users also had higher active and limited joint counts than did first-time users.

Adalimumab reduces enthesitis-related arthritis symptoms

"Enthesitis-related arthritis is a category of juvenile idiopathic arthritis that primarily affects the peripheral joints and entheses, but can also affect the sacroiliac joints and spine," said Iain Sainsbury, Ph.D., associate director for RA in global medical affairs at AbbVie, in a presentation of data on the efficacy and safety of adalimumab in children with enthesitis-related arthritis (ERA) from the M11-328 study on behalf of the study’s authors.

Since adalimumab had been previously been shown to be an effective treatment for children aged 2-17 years with polyarticular JIA, the aim of the current study was to see if it could also be of benefit in patients with ERA.

The M11-328 study was a 12-week, multicenter, double-blind trial that enrolled 46 children aged 6-18 years. They were randomized 2:1 to receive adalimumab or placebo, with continued open-label treatment for up to 3 years.

The primary endpoint, the percent change in the number of active joints with arthritis from baseline to week 12, showed a clear benefit of treatment with the biologic. Indeed, adalimumab reduced the active joint count by 62.6% vs. a reduction of 11.6% with placebo (P = .039). The treatment response was maintained with continued adalimumab treatment for 1 year. Other signs and symptoms of ERA, such as pain and enthesitis, were also reduced by the active treatment.

Dr. Sainsbury said, "The safety profile in this patient population is consistent with what we see in all other patient populations," including children aged 2 years or older with polyarticular JIA.

CHERISH 2-year data show sustained tocilizumab benefits

Dr. Eileen Baildam of Alder Hey Children’s Hospital in Liverpool, England, reported 2-year follow-up data from the CHERISH phase III trial. This trial was conducted in 58 centers in 15 countries and consisted of three phases: an open-label, lead-in phase that involved 188 children who were treated with tocilizumab for 16 weeks; a double-blind phase in which 166 children were randomized to receive tocilizumab or placebo for 24 weeks; and an open-label extension involving 60 children who were treated with tocilizumab for a further 64 weeks, bringing the total duration of therapy to 104 weeks.

Data from the second phase of the trial, at 40 weeks’ follow-up have been previously reported and "showed a very significant improvement in the flare rate in those treated vs. those on placebo," Dr. Baildam said.

A total of 82% of the 188 children enrolled in the study completed 2 years’ treatment and exhibited improvements in JIA American College of Rheumatology (ACR) 50, 70, and 90 criteria. The percentage of children who were ACR70 responders at 40 weeks and at 104 weeks were 79.3% and 86.6%, respectively, and the percentage who were ACR90 responders were a respective 50% and 70.7%.

Dr. Baildam reported a continued benefit of the anti–interleukin-6 receptor therapy, with 73.5% of children showing minimal disease activity at 40 weeks and 58.5% showing minimal disease activity at 2 years. Benefit was also seen in children who had received prior biologics.

"Overall the safety profiles remained unchanged from those at week 40, there were no deaths, and no new or unexpected safety concerns," she concluded.

The Etanercept Cohort Study is supported by Pfizer; Ms. Kearsley-Fleet had no disclosures. The CHERISH trial was supported by Pfizer; Dr. Baildam has received speaker or advisory board fees from Roche and Pfizer. The adalimumab trial was supported by AbbVie; Dr. Sainsbury is an employee of AbbVie.

LIVERPOOL, ENGLAND – Biologic therapy has shown sustained benefits for the signs and symptoms of juvenile idiopathic arthritis as well as improved height over the course of up to 2 years of treatment, according to data from four studies presented at the British Society for Rheumatology annual conference.

The studies provided evidence indicating that:

• Treatment with etanercept over 2 years helped children with juvenile idiopathic arthritis (JIA) to gain height, although their overall vertical growth still lagged behind children in the general reference population.

• When etanercept was not used as a first-line biologic, it was most common to use adalimumab instead, followed by tocilizumab and infliximab.

• Adalimumab reduced the signs and symptoms of enthesitis-related arthritis (ERA) in children by week 12 of treatment and the effects were sustained at 1-year of follow up.

• The benefits of tocilizumab in reducing disease activity to a minimal level were preserved from 40 weeks to 2 years in most patients.

Etanercept and growth in JIA

"Etanercept therapy was associated with improvement in height z score over the first 2 years of therapy," and the effect was most apparent in children who were not receiving concomitant steroids, said Lianne Kearsley-Fleet, who reported data from the British Society for Paediatric and Adolescent Rheumatology (BSPAR Etanercept Cohort Study). The prospective, observational study was initiated in 2004 with the aim of recruiting all children with JIA who were starting treatment with etanercept. For comparison, a cohort of children newly starting methotrexate was also included.

The current analysis focused on the growth of children taking the biologic therapy because children with JIA are known to have restricted growth, which may be linked to inflammation and the use of corticosteroids. Of 658 children registered in the study, 191 had height data available at baseline and at 1 and 2 years’ follow-up.

Mean height and changes in height during biologic therapy over time were assessed by calculating z scores, and compared with values set by the World Health Organization for girls and boys of a similar age who did not have JIA.

The mean age of children included in the current analysis was 11 years, with a mean disease duration of 3.5 years. The majority (65%) were female, with concomitant oral steroids in 38% and methotrexate in 58%.

The mean height z score of the study population was –0.67 at baseline, –0.57 after 1 year, and –0.45 after 2 years’ etanercept therapy. "While it is still negative [compared with children of a similar sex and age], there is still significant improvement," said Ms. Kearsley-Fleet, a research assistant from the Arthritis Research UK Centre for Epidemiology at the University of Manchester, England.

"There was a lack of evidence for association between disease activity and improved growth," the researcher noted, adding that "anti-TNF inhibitors alone may be insufficient to increase growth."

Use of non–etanercept biologics

In a poster presentation at the meeting, Ms. Kearsley-Fleet and her colleagues at her institution also presented the findings from the Biologics for Children with Rheumatic Diseases Study. This ongoing, prospective, observational cohort study is looking at the use of biologics other than etanercept in the treatment of JIA.

Since the study started in 2010, and up until early April 2014, a total of 280 children with JIA had been recruited into the biologic cohort. The study also includes a parallel control cohort of 295 children being treated with methotrexate. Just under half (46%) of the children recruited into the biologic cohort started a non–etanercept biologic as their first biologic treatment, of which the majority (75%) received the medications off label. The main alternatives to etanercept being used were adalimumab in 38% of cases, followed by tocilizumab in 25% of children, and infliximab in 23%. Other drugs used as the first biologic was anakinra (13%), rituximab (less than 1%), and abatacept (less than 1%).

Adalimumab was also a popular subsequent biologic choice in 30% of cases, with 20% of patients using tocilizumab, and 25% infliximab as a subsequent biologic choice.

The research team reported that patients using a biologic for the first time were significantly younger and had shorter disease durations than did subsequent biologic users. The mean ages of first-, second-, and third-line users were 9, 12, and 11 years, respectively, and the mean disease durations were 2, 6, and 4 years. Subsequent biologic users also had higher active and limited joint counts than did first-time users.

Adalimumab reduces enthesitis-related arthritis symptoms

"Enthesitis-related arthritis is a category of juvenile idiopathic arthritis that primarily affects the peripheral joints and entheses, but can also affect the sacroiliac joints and spine," said Iain Sainsbury, Ph.D., associate director for RA in global medical affairs at AbbVie, in a presentation of data on the efficacy and safety of adalimumab in children with enthesitis-related arthritis (ERA) from the M11-328 study on behalf of the study’s authors.

Since adalimumab had been previously been shown to be an effective treatment for children aged 2-17 years with polyarticular JIA, the aim of the current study was to see if it could also be of benefit in patients with ERA.

The M11-328 study was a 12-week, multicenter, double-blind trial that enrolled 46 children aged 6-18 years. They were randomized 2:1 to receive adalimumab or placebo, with continued open-label treatment for up to 3 years.

The primary endpoint, the percent change in the number of active joints with arthritis from baseline to week 12, showed a clear benefit of treatment with the biologic. Indeed, adalimumab reduced the active joint count by 62.6% vs. a reduction of 11.6% with placebo (P = .039). The treatment response was maintained with continued adalimumab treatment for 1 year. Other signs and symptoms of ERA, such as pain and enthesitis, were also reduced by the active treatment.

Dr. Sainsbury said, "The safety profile in this patient population is consistent with what we see in all other patient populations," including children aged 2 years or older with polyarticular JIA.

CHERISH 2-year data show sustained tocilizumab benefits

Dr. Eileen Baildam of Alder Hey Children’s Hospital in Liverpool, England, reported 2-year follow-up data from the CHERISH phase III trial. This trial was conducted in 58 centers in 15 countries and consisted of three phases: an open-label, lead-in phase that involved 188 children who were treated with tocilizumab for 16 weeks; a double-blind phase in which 166 children were randomized to receive tocilizumab or placebo for 24 weeks; and an open-label extension involving 60 children who were treated with tocilizumab for a further 64 weeks, bringing the total duration of therapy to 104 weeks.

Data from the second phase of the trial, at 40 weeks’ follow-up have been previously reported and "showed a very significant improvement in the flare rate in those treated vs. those on placebo," Dr. Baildam said.

A total of 82% of the 188 children enrolled in the study completed 2 years’ treatment and exhibited improvements in JIA American College of Rheumatology (ACR) 50, 70, and 90 criteria. The percentage of children who were ACR70 responders at 40 weeks and at 104 weeks were 79.3% and 86.6%, respectively, and the percentage who were ACR90 responders were a respective 50% and 70.7%.

Dr. Baildam reported a continued benefit of the anti–interleukin-6 receptor therapy, with 73.5% of children showing minimal disease activity at 40 weeks and 58.5% showing minimal disease activity at 2 years. Benefit was also seen in children who had received prior biologics.

"Overall the safety profiles remained unchanged from those at week 40, there were no deaths, and no new or unexpected safety concerns," she concluded.

The Etanercept Cohort Study is supported by Pfizer; Ms. Kearsley-Fleet had no disclosures. The CHERISH trial was supported by Pfizer; Dr. Baildam has received speaker or advisory board fees from Roche and Pfizer. The adalimumab trial was supported by AbbVie; Dr. Sainsbury is an employee of AbbVie.

LIVERPOOL, ENGLAND – Biologic therapy has shown sustained benefits for the signs and symptoms of juvenile idiopathic arthritis as well as improved height over the course of up to 2 years of treatment, according to data from four studies presented at the British Society for Rheumatology annual conference.

The studies provided evidence indicating that:

• Treatment with etanercept over 2 years helped children with juvenile idiopathic arthritis (JIA) to gain height, although their overall vertical growth still lagged behind children in the general reference population.

• When etanercept was not used as a first-line biologic, it was most common to use adalimumab instead, followed by tocilizumab and infliximab.

• Adalimumab reduced the signs and symptoms of enthesitis-related arthritis (ERA) in children by week 12 of treatment and the effects were sustained at 1-year of follow up.

• The benefits of tocilizumab in reducing disease activity to a minimal level were preserved from 40 weeks to 2 years in most patients.

Etanercept and growth in JIA

"Etanercept therapy was associated with improvement in height z score over the first 2 years of therapy," and the effect was most apparent in children who were not receiving concomitant steroids, said Lianne Kearsley-Fleet, who reported data from the British Society for Paediatric and Adolescent Rheumatology (BSPAR Etanercept Cohort Study). The prospective, observational study was initiated in 2004 with the aim of recruiting all children with JIA who were starting treatment with etanercept. For comparison, a cohort of children newly starting methotrexate was also included.

The current analysis focused on the growth of children taking the biologic therapy because children with JIA are known to have restricted growth, which may be linked to inflammation and the use of corticosteroids. Of 658 children registered in the study, 191 had height data available at baseline and at 1 and 2 years’ follow-up.

Mean height and changes in height during biologic therapy over time were assessed by calculating z scores, and compared with values set by the World Health Organization for girls and boys of a similar age who did not have JIA.

The mean age of children included in the current analysis was 11 years, with a mean disease duration of 3.5 years. The majority (65%) were female, with concomitant oral steroids in 38% and methotrexate in 58%.

The mean height z score of the study population was –0.67 at baseline, –0.57 after 1 year, and –0.45 after 2 years’ etanercept therapy. "While it is still negative [compared with children of a similar sex and age], there is still significant improvement," said Ms. Kearsley-Fleet, a research assistant from the Arthritis Research UK Centre for Epidemiology at the University of Manchester, England.

"There was a lack of evidence for association between disease activity and improved growth," the researcher noted, adding that "anti-TNF inhibitors alone may be insufficient to increase growth."

Use of non–etanercept biologics

In a poster presentation at the meeting, Ms. Kearsley-Fleet and her colleagues at her institution also presented the findings from the Biologics for Children with Rheumatic Diseases Study. This ongoing, prospective, observational cohort study is looking at the use of biologics other than etanercept in the treatment of JIA.

Since the study started in 2010, and up until early April 2014, a total of 280 children with JIA had been recruited into the biologic cohort. The study also includes a parallel control cohort of 295 children being treated with methotrexate. Just under half (46%) of the children recruited into the biologic cohort started a non–etanercept biologic as their first biologic treatment, of which the majority (75%) received the medications off label. The main alternatives to etanercept being used were adalimumab in 38% of cases, followed by tocilizumab in 25% of children, and infliximab in 23%. Other drugs used as the first biologic was anakinra (13%), rituximab (less than 1%), and abatacept (less than 1%).

Adalimumab was also a popular subsequent biologic choice in 30% of cases, with 20% of patients using tocilizumab, and 25% infliximab as a subsequent biologic choice.

The research team reported that patients using a biologic for the first time were significantly younger and had shorter disease durations than did subsequent biologic users. The mean ages of first-, second-, and third-line users were 9, 12, and 11 years, respectively, and the mean disease durations were 2, 6, and 4 years. Subsequent biologic users also had higher active and limited joint counts than did first-time users.

Adalimumab reduces enthesitis-related arthritis symptoms

"Enthesitis-related arthritis is a category of juvenile idiopathic arthritis that primarily affects the peripheral joints and entheses, but can also affect the sacroiliac joints and spine," said Iain Sainsbury, Ph.D., associate director for RA in global medical affairs at AbbVie, in a presentation of data on the efficacy and safety of adalimumab in children with enthesitis-related arthritis (ERA) from the M11-328 study on behalf of the study’s authors.

Since adalimumab had been previously been shown to be an effective treatment for children aged 2-17 years with polyarticular JIA, the aim of the current study was to see if it could also be of benefit in patients with ERA.

The M11-328 study was a 12-week, multicenter, double-blind trial that enrolled 46 children aged 6-18 years. They were randomized 2:1 to receive adalimumab or placebo, with continued open-label treatment for up to 3 years.

The primary endpoint, the percent change in the number of active joints with arthritis from baseline to week 12, showed a clear benefit of treatment with the biologic. Indeed, adalimumab reduced the active joint count by 62.6% vs. a reduction of 11.6% with placebo (P = .039). The treatment response was maintained with continued adalimumab treatment for 1 year. Other signs and symptoms of ERA, such as pain and enthesitis, were also reduced by the active treatment.

Dr. Sainsbury said, "The safety profile in this patient population is consistent with what we see in all other patient populations," including children aged 2 years or older with polyarticular JIA.

CHERISH 2-year data show sustained tocilizumab benefits

Dr. Eileen Baildam of Alder Hey Children’s Hospital in Liverpool, England, reported 2-year follow-up data from the CHERISH phase III trial. This trial was conducted in 58 centers in 15 countries and consisted of three phases: an open-label, lead-in phase that involved 188 children who were treated with tocilizumab for 16 weeks; a double-blind phase in which 166 children were randomized to receive tocilizumab or placebo for 24 weeks; and an open-label extension involving 60 children who were treated with tocilizumab for a further 64 weeks, bringing the total duration of therapy to 104 weeks.

Data from the second phase of the trial, at 40 weeks’ follow-up have been previously reported and "showed a very significant improvement in the flare rate in those treated vs. those on placebo," Dr. Baildam said.

A total of 82% of the 188 children enrolled in the study completed 2 years’ treatment and exhibited improvements in JIA American College of Rheumatology (ACR) 50, 70, and 90 criteria. The percentage of children who were ACR70 responders at 40 weeks and at 104 weeks were 79.3% and 86.6%, respectively, and the percentage who were ACR90 responders were a respective 50% and 70.7%.

Dr. Baildam reported a continued benefit of the anti–interleukin-6 receptor therapy, with 73.5% of children showing minimal disease activity at 40 weeks and 58.5% showing minimal disease activity at 2 years. Benefit was also seen in children who had received prior biologics.

"Overall the safety profiles remained unchanged from those at week 40, there were no deaths, and no new or unexpected safety concerns," she concluded.

The Etanercept Cohort Study is supported by Pfizer; Ms. Kearsley-Fleet had no disclosures. The CHERISH trial was supported by Pfizer; Dr. Baildam has received speaker or advisory board fees from Roche and Pfizer. The adalimumab trial was supported by AbbVie; Dr. Sainsbury is an employee of AbbVie.

LIVERPOOL, ENGLAND – Anti–tumor necrosis factor treatment significantly decreased fatigue in patients with rheumatoid arthritis in an analysis of data from the British Society for Rheumatology Biologics Registers Rheumatoid Arthritis Register.

Of 6,835 patients included in the analysis, 38.8% reported high levels of fatigue at baseline, which was defined as a score of 12.5 or less on the vitality domain of the Short Form (SF)-36. Of these, 66% reported clinically relevant improvements of 10 U or more on the SF-36 vitality domain 6 months after starting anti–tumor necrosis factor (TNF) treatment.

"Randomized controlled trials have shown modest improvements in fatigue with anti-TNF therapy, but fatigue has not been the primary research question in these trials, and so that led us to ask what the effect would be in patients who had clinically relevant symptoms," said Katie Druce, who performed the research as part of her postgraduate studies at the University of Aberdeen (Scotland).

Data on 13,122 patients with RA who started treatment with etanercept, adalimumab, or infliximab between 2000 and 2008 were initially considered, but only those who had data on fatigue available at baseline and at 6-month assessments were included in the analysis.

At baseline, the mean age of participants was 57 years, and most (77%) were female and white (97%). A total of 31% were not working because of illness. The patients had mean scores of 6.57 on the 28-joint Disease Activity Score, 21.1 on the bodily pain and 24.8 on the vitality domains of the SF-36, and 2.03 on the mean Health Assessment Questionnaire Disability Index.

Fatigue improved at 6 months

Ms. Druce reported at the British Society for Rheumatology (BSR) annual conference that there was an absolute improvement in fatigue at 6 months of 22.5 U in the 2,652 patients who had reported a high level of fatigue at baseline. This exceeded the value of 10 U or more that was conservatively set as the minimum clinically important difference (MCID). Furthermore, in the patients who improved – that is going from a high level to a low level of fatigue – the difference was even greater, with improvement three times the MCID, at about 32 U.

Baseline predictors of improvement included being in good mental health, no history of hypertension or depression, and not being unemployed because of ill health. Other predictors included being female, rheumatoid factor positive, having a low level of disability, and not using steroids. Patients with multiple baseline predictors of improvement appeared to have the largest absolute improvement in fatigue at 6 months.

"Importantly, we saw no association with disease activity or measures of inflammation," Ms. Druce said, and the specific anti-TNF drug used was also not influential on the findings. She noted that baseline predictors of fatigue might be able to help stratify patients who need specific fatigue management.

Pain drives fatigue in RA

The improvements in RA-related fatigue seen with biologic therapy were driven by decreases in pain and not disease activity, Ms. Druce said, reporting the finding of a separate analysis elsewhere at the conference.

"Anecdotally, from both clinicians and patients, we know that they consider fatigue to be caused by disease activity, but when we look at the evidence from the literature, we see that there is an inconsistent link between fatigue and disease activity," she said. "Yet we know that fatigue improves after anti-TNF therapy commences. So this led us to look at which variables are driving change in fatigue that we see after anti-TNF therapies are commenced," she said.

Ms. Druce and her colleagues used structural equation modeling, which evaluates the effect of one variable on an outcome and the pathway by which it has that effect, to find that for every 1–standard deviation increase in change in disease activity, there was a 0.05 increase in change in fatigue. Although this does indicate some improvement, change in pain had a greater effect, with every 1–standard deviation increase in change in pain linked to a 0.31 increase in change in fatigue.

"What we found was that 82% of the effect of change in disease activity is mediated through other variables in [our] model," Ms. Druce explained. Indeed, it was found that almost two-thirds of the change in fatigue was mediated by pain (or pain and mental health), 27% through disability (or disability and mental health), and 9% directly through mental health.

"This led us to conclude that improvements in fatigue after commencing anti-TNF therapies are driven by improvements in pain rather than directly from disease activity, but that mental health and disability are also important, and, along with pain, represent potentially modifiable factors for us to target in routine treatment," Ms. Druce said.

How can fatigue be assessed?

Fatigue is an important facet of RA that should be included in patient assessment, said Sara Hewlett, Ph.D., professor of nursing at the University of the West of England in Bristol.

Speaking during a session organized by the BSR’s Specialist Interest Group in RA, she noted that there are three reasons why fatigue is important to assess: It is important to patients, it provides information beyond disease activity, and it responds to several interventions, including physical activity, psychological strategies, and biologic medications.

Dr. Hewlett noted that there have been many instruments used to try to assess fatigue in clinical trials, but none was specifically developed to assess fatigue in patients with RA, and many were not developed using recommended methodology. That was until the development and validation of the Bristol Rheumatoid Arthritis Fatigue Numerical Rating Scales (BRAF NRS).

On the BRAF NRS, patients circle a number from 0 to 10 to indicate the average level of fatigue that they have experienced, the impact that fatigue has had on them, and how well they have subsequently coped, in the past 7 days.

Ms. Druce said she had no financial disclosures. One of her coauthors had received honoraria and research funding from Pfizer and had been a consultant for Merck Sharp & Dohme. Ms. Hewlett said she had no disclosures other than helping to develop the BRAF NRS.

LIVERPOOL, ENGLAND – Anti–tumor necrosis factor treatment significantly decreased fatigue in patients with rheumatoid arthritis in an analysis of data from the British Society for Rheumatology Biologics Registers Rheumatoid Arthritis Register.

Of 6,835 patients included in the analysis, 38.8% reported high levels of fatigue at baseline, which was defined as a score of 12.5 or less on the vitality domain of the Short Form (SF)-36. Of these, 66% reported clinically relevant improvements of 10 U or more on the SF-36 vitality domain 6 months after starting anti–tumor necrosis factor (TNF) treatment.

"Randomized controlled trials have shown modest improvements in fatigue with anti-TNF therapy, but fatigue has not been the primary research question in these trials, and so that led us to ask what the effect would be in patients who had clinically relevant symptoms," said Katie Druce, who performed the research as part of her postgraduate studies at the University of Aberdeen (Scotland).

Data on 13,122 patients with RA who started treatment with etanercept, adalimumab, or infliximab between 2000 and 2008 were initially considered, but only those who had data on fatigue available at baseline and at 6-month assessments were included in the analysis.

At baseline, the mean age of participants was 57 years, and most (77%) were female and white (97%). A total of 31% were not working because of illness. The patients had mean scores of 6.57 on the 28-joint Disease Activity Score, 21.1 on the bodily pain and 24.8 on the vitality domains of the SF-36, and 2.03 on the mean Health Assessment Questionnaire Disability Index.

Fatigue improved at 6 months

Ms. Druce reported at the British Society for Rheumatology (BSR) annual conference that there was an absolute improvement in fatigue at 6 months of 22.5 U in the 2,652 patients who had reported a high level of fatigue at baseline. This exceeded the value of 10 U or more that was conservatively set as the minimum clinically important difference (MCID). Furthermore, in the patients who improved – that is going from a high level to a low level of fatigue – the difference was even greater, with improvement three times the MCID, at about 32 U.

Baseline predictors of improvement included being in good mental health, no history of hypertension or depression, and not being unemployed because of ill health. Other predictors included being female, rheumatoid factor positive, having a low level of disability, and not using steroids. Patients with multiple baseline predictors of improvement appeared to have the largest absolute improvement in fatigue at 6 months.

"Importantly, we saw no association with disease activity or measures of inflammation," Ms. Druce said, and the specific anti-TNF drug used was also not influential on the findings. She noted that baseline predictors of fatigue might be able to help stratify patients who need specific fatigue management.

Pain drives fatigue in RA

The improvements in RA-related fatigue seen with biologic therapy were driven by decreases in pain and not disease activity, Ms. Druce said, reporting the finding of a separate analysis elsewhere at the conference.

"Anecdotally, from both clinicians and patients, we know that they consider fatigue to be caused by disease activity, but when we look at the evidence from the literature, we see that there is an inconsistent link between fatigue and disease activity," she said. "Yet we know that fatigue improves after anti-TNF therapy commences. So this led us to look at which variables are driving change in fatigue that we see after anti-TNF therapies are commenced," she said.

Ms. Druce and her colleagues used structural equation modeling, which evaluates the effect of one variable on an outcome and the pathway by which it has that effect, to find that for every 1–standard deviation increase in change in disease activity, there was a 0.05 increase in change in fatigue. Although this does indicate some improvement, change in pain had a greater effect, with every 1–standard deviation increase in change in pain linked to a 0.31 increase in change in fatigue.

"What we found was that 82% of the effect of change in disease activity is mediated through other variables in [our] model," Ms. Druce explained. Indeed, it was found that almost two-thirds of the change in fatigue was mediated by pain (or pain and mental health), 27% through disability (or disability and mental health), and 9% directly through mental health.

"This led us to conclude that improvements in fatigue after commencing anti-TNF therapies are driven by improvements in pain rather than directly from disease activity, but that mental health and disability are also important, and, along with pain, represent potentially modifiable factors for us to target in routine treatment," Ms. Druce said.

How can fatigue be assessed?

Fatigue is an important facet of RA that should be included in patient assessment, said Sara Hewlett, Ph.D., professor of nursing at the University of the West of England in Bristol.

Speaking during a session organized by the BSR’s Specialist Interest Group in RA, she noted that there are three reasons why fatigue is important to assess: It is important to patients, it provides information beyond disease activity, and it responds to several interventions, including physical activity, psychological strategies, and biologic medications.

Dr. Hewlett noted that there have been many instruments used to try to assess fatigue in clinical trials, but none was specifically developed to assess fatigue in patients with RA, and many were not developed using recommended methodology. That was until the development and validation of the Bristol Rheumatoid Arthritis Fatigue Numerical Rating Scales (BRAF NRS).

On the BRAF NRS, patients circle a number from 0 to 10 to indicate the average level of fatigue that they have experienced, the impact that fatigue has had on them, and how well they have subsequently coped, in the past 7 days.

Ms. Druce said she had no financial disclosures. One of her coauthors had received honoraria and research funding from Pfizer and had been a consultant for Merck Sharp & Dohme. Ms. Hewlett said she had no disclosures other than helping to develop the BRAF NRS.

LIVERPOOL, ENGLAND – Anti–tumor necrosis factor treatment significantly decreased fatigue in patients with rheumatoid arthritis in an analysis of data from the British Society for Rheumatology Biologics Registers Rheumatoid Arthritis Register.

Of 6,835 patients included in the analysis, 38.8% reported high levels of fatigue at baseline, which was defined as a score of 12.5 or less on the vitality domain of the Short Form (SF)-36. Of these, 66% reported clinically relevant improvements of 10 U or more on the SF-36 vitality domain 6 months after starting anti–tumor necrosis factor (TNF) treatment.

"Randomized controlled trials have shown modest improvements in fatigue with anti-TNF therapy, but fatigue has not been the primary research question in these trials, and so that led us to ask what the effect would be in patients who had clinically relevant symptoms," said Katie Druce, who performed the research as part of her postgraduate studies at the University of Aberdeen (Scotland).

Data on 13,122 patients with RA who started treatment with etanercept, adalimumab, or infliximab between 2000 and 2008 were initially considered, but only those who had data on fatigue available at baseline and at 6-month assessments were included in the analysis.

At baseline, the mean age of participants was 57 years, and most (77%) were female and white (97%). A total of 31% were not working because of illness. The patients had mean scores of 6.57 on the 28-joint Disease Activity Score, 21.1 on the bodily pain and 24.8 on the vitality domains of the SF-36, and 2.03 on the mean Health Assessment Questionnaire Disability Index.

Fatigue improved at 6 months

Ms. Druce reported at the British Society for Rheumatology (BSR) annual conference that there was an absolute improvement in fatigue at 6 months of 22.5 U in the 2,652 patients who had reported a high level of fatigue at baseline. This exceeded the value of 10 U or more that was conservatively set as the minimum clinically important difference (MCID). Furthermore, in the patients who improved – that is going from a high level to a low level of fatigue – the difference was even greater, with improvement three times the MCID, at about 32 U.

Baseline predictors of improvement included being in good mental health, no history of hypertension or depression, and not being unemployed because of ill health. Other predictors included being female, rheumatoid factor positive, having a low level of disability, and not using steroids. Patients with multiple baseline predictors of improvement appeared to have the largest absolute improvement in fatigue at 6 months.

"Importantly, we saw no association with disease activity or measures of inflammation," Ms. Druce said, and the specific anti-TNF drug used was also not influential on the findings. She noted that baseline predictors of fatigue might be able to help stratify patients who need specific fatigue management.

Pain drives fatigue in RA

The improvements in RA-related fatigue seen with biologic therapy were driven by decreases in pain and not disease activity, Ms. Druce said, reporting the finding of a separate analysis elsewhere at the conference.

"Anecdotally, from both clinicians and patients, we know that they consider fatigue to be caused by disease activity, but when we look at the evidence from the literature, we see that there is an inconsistent link between fatigue and disease activity," she said. "Yet we know that fatigue improves after anti-TNF therapy commences. So this led us to look at which variables are driving change in fatigue that we see after anti-TNF therapies are commenced," she said.

Ms. Druce and her colleagues used structural equation modeling, which evaluates the effect of one variable on an outcome and the pathway by which it has that effect, to find that for every 1–standard deviation increase in change in disease activity, there was a 0.05 increase in change in fatigue. Although this does indicate some improvement, change in pain had a greater effect, with every 1–standard deviation increase in change in pain linked to a 0.31 increase in change in fatigue.

"What we found was that 82% of the effect of change in disease activity is mediated through other variables in [our] model," Ms. Druce explained. Indeed, it was found that almost two-thirds of the change in fatigue was mediated by pain (or pain and mental health), 27% through disability (or disability and mental health), and 9% directly through mental health.

"This led us to conclude that improvements in fatigue after commencing anti-TNF therapies are driven by improvements in pain rather than directly from disease activity, but that mental health and disability are also important, and, along with pain, represent potentially modifiable factors for us to target in routine treatment," Ms. Druce said.

How can fatigue be assessed?

Fatigue is an important facet of RA that should be included in patient assessment, said Sara Hewlett, Ph.D., professor of nursing at the University of the West of England in Bristol.

Speaking during a session organized by the BSR’s Specialist Interest Group in RA, she noted that there are three reasons why fatigue is important to assess: It is important to patients, it provides information beyond disease activity, and it responds to several interventions, including physical activity, psychological strategies, and biologic medications.

Dr. Hewlett noted that there have been many instruments used to try to assess fatigue in clinical trials, but none was specifically developed to assess fatigue in patients with RA, and many were not developed using recommended methodology. That was until the development and validation of the Bristol Rheumatoid Arthritis Fatigue Numerical Rating Scales (BRAF NRS).

On the BRAF NRS, patients circle a number from 0 to 10 to indicate the average level of fatigue that they have experienced, the impact that fatigue has had on them, and how well they have subsequently coped, in the past 7 days.

Ms. Druce said she had no financial disclosures. One of her coauthors had received honoraria and research funding from Pfizer and had been a consultant for Merck Sharp & Dohme. Ms. Hewlett said she had no disclosures other than helping to develop the BRAF NRS.

LIVERPOOL, ENGLAND – Reducing the once-weekly maintenance dose of etanercept from 50 mg to 25 mg was associated with worsening control of ankylosing spondylitis in an open-label, multicenter, pilot study.

Results of the ANSWER (Ankylosing Spondylitis with Etanercept Regimens) trial showed that halving the dose of the biologic almost halved the percentage of patients maintaining disease control at 6 months, from 92% to 52% of patients.

Although this was a small study, involving just 47 patients who were randomized to continue etanercept at a weekly dose of 50 mg (n = 24) or to drop down to 25 mg (n = 23) after competing 6 months’ treatment, the results suggest that other approaches are needed to see if the long-term dose of the drug can be reduced, commented study investigator Dr. Karl Gaffney of Norfolk and Norwich University Hospital NHS Foundation Trust, Norwich, England.

"A larger study is required to identify patients suitable for step-down," Dr. Gaffney said at the British Society for Rheumatology annual conference.

"We all know that anti-TNF [tumor necrosis factor] therapy is firmly established in clinical practice for patients with ankylosing spondylitis and this treatment has transformed the quality of life for many of our patients," Dr. Gaffney observed.

"The concept of dose reduction is very appealing from an economic perspective, but also in terms of the cumulative exposure and the risk of adverse effects," he added.

"In patients with rheumatoid arthritis, it’s been shown that lower doses may maintain the clinical response," Dr. Gaffney explained, citing the PRESERVE study published last year (Lancet 2013;381:918-29). However, there are limited data about whether reducing the dose of etanercept could also be successful in ankylosing spondylitis (AS) patients. This is why the ANSWER study was performed.

Patients with active AS who were not responding to conventional therapies and had not yet been treated with a biologic drug were consecutively recruited at two hospitals in England between September 2010 and September 2012. For inclusion, patients had to have sustained spinal disease and a Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) score of four or higher. The mean age of patients in the trial was 46.7 years, and the mean BASDAI at baseline was 6.8.

All recruited patients were treated with 50 mg etanercept, once weekly, for 6 months, and those with a "sufficient" clinical response were randomized to either continue this dose for a further 6 months or step down to 25 mg for continued maintenance treatment. A sufficient clinical response was defined as a 50% reduction in BASDAI or a fall of two or more units plus reduction in axial pain of 2 cm or more.

Three months into the maintenance period, patients who had been randomized to the step-down arm started to experience loss of disease control (60.9% vs. 91.7% of 50 mg–treated patients), with the gap widening by 6 months postrandomization and 39% fewer patients maintaining disease control (P = .003).

"There’s no doubt that the 50-mg group did better," Dr. Gaffney said. There were also statistically significant differences favoring the higher dose in a number of secondary outcome measures, which included other measurements of AS disease activity and quality of life. There were no significant differences in terms of adverse events between the two groups.

"However, 52% of the 25-mg arm did maintain their BASDAI response. So theoretically, there may be a subset of patients in whom we can offer this therapeutic option," Dr. Gaffney said. He noted that even in this small group of patients with short-term follow-up that the potential cost savings of being able to reduce the dose would be significant, at around £100,000 (about U.S. $170,000).

"We all have patients who do very well [on etanercept], and I think it would be very nice to be able to say, from 6 months, who we can attempt to step down," commented Dr. Nicola Goodson of University Hospital Aintree, Liverpool. Dr. Goodson chaired the session where the study findings were reported.

Although the ANSWER investigators did look for predictors of response, none were found to be significant. Patients continue to be monitored and the investigators may look at this again. Dr. Gaffney noted that all patients who had their 50-mg dose reinstated regained their clinical response.

He also highlighted during discussion: "A number of patients in the 50-mg group asked to be moved down to 25 mg at the end of the study because they wanted to explore the option. So I think this may well have some long-term consequences, especially in a disease area where we know that the treatment is more symptom modifying than radiographically modifying."

The study was funded by Pfizer. Dr. Gaffney has received research support and honoraria from Pfizer, Merck Sharp & Dohme, AbbVie, and UCB.

LIVERPOOL, ENGLAND – Reducing the once-weekly maintenance dose of etanercept from 50 mg to 25 mg was associated with worsening control of ankylosing spondylitis in an open-label, multicenter, pilot study.

Results of the ANSWER (Ankylosing Spondylitis with Etanercept Regimens) trial showed that halving the dose of the biologic almost halved the percentage of patients maintaining disease control at 6 months, from 92% to 52% of patients.

Although this was a small study, involving just 47 patients who were randomized to continue etanercept at a weekly dose of 50 mg (n = 24) or to drop down to 25 mg (n = 23) after competing 6 months’ treatment, the results suggest that other approaches are needed to see if the long-term dose of the drug can be reduced, commented study investigator Dr. Karl Gaffney of Norfolk and Norwich University Hospital NHS Foundation Trust, Norwich, England.

"A larger study is required to identify patients suitable for step-down," Dr. Gaffney said at the British Society for Rheumatology annual conference.

"We all know that anti-TNF [tumor necrosis factor] therapy is firmly established in clinical practice for patients with ankylosing spondylitis and this treatment has transformed the quality of life for many of our patients," Dr. Gaffney observed.

"The concept of dose reduction is very appealing from an economic perspective, but also in terms of the cumulative exposure and the risk of adverse effects," he added.

"In patients with rheumatoid arthritis, it’s been shown that lower doses may maintain the clinical response," Dr. Gaffney explained, citing the PRESERVE study published last year (Lancet 2013;381:918-29). However, there are limited data about whether reducing the dose of etanercept could also be successful in ankylosing spondylitis (AS) patients. This is why the ANSWER study was performed.

Patients with active AS who were not responding to conventional therapies and had not yet been treated with a biologic drug were consecutively recruited at two hospitals in England between September 2010 and September 2012. For inclusion, patients had to have sustained spinal disease and a Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) score of four or higher. The mean age of patients in the trial was 46.7 years, and the mean BASDAI at baseline was 6.8.

All recruited patients were treated with 50 mg etanercept, once weekly, for 6 months, and those with a "sufficient" clinical response were randomized to either continue this dose for a further 6 months or step down to 25 mg for continued maintenance treatment. A sufficient clinical response was defined as a 50% reduction in BASDAI or a fall of two or more units plus reduction in axial pain of 2 cm or more.

Three months into the maintenance period, patients who had been randomized to the step-down arm started to experience loss of disease control (60.9% vs. 91.7% of 50 mg–treated patients), with the gap widening by 6 months postrandomization and 39% fewer patients maintaining disease control (P = .003).

"There’s no doubt that the 50-mg group did better," Dr. Gaffney said. There were also statistically significant differences favoring the higher dose in a number of secondary outcome measures, which included other measurements of AS disease activity and quality of life. There were no significant differences in terms of adverse events between the two groups.

"However, 52% of the 25-mg arm did maintain their BASDAI response. So theoretically, there may be a subset of patients in whom we can offer this therapeutic option," Dr. Gaffney said. He noted that even in this small group of patients with short-term follow-up that the potential cost savings of being able to reduce the dose would be significant, at around £100,000 (about U.S. $170,000).

"We all have patients who do very well [on etanercept], and I think it would be very nice to be able to say, from 6 months, who we can attempt to step down," commented Dr. Nicola Goodson of University Hospital Aintree, Liverpool. Dr. Goodson chaired the session where the study findings were reported.

Although the ANSWER investigators did look for predictors of response, none were found to be significant. Patients continue to be monitored and the investigators may look at this again. Dr. Gaffney noted that all patients who had their 50-mg dose reinstated regained their clinical response.

He also highlighted during discussion: "A number of patients in the 50-mg group asked to be moved down to 25 mg at the end of the study because they wanted to explore the option. So I think this may well have some long-term consequences, especially in a disease area where we know that the treatment is more symptom modifying than radiographically modifying."

The study was funded by Pfizer. Dr. Gaffney has received research support and honoraria from Pfizer, Merck Sharp & Dohme, AbbVie, and UCB.

LIVERPOOL, ENGLAND – Reducing the once-weekly maintenance dose of etanercept from 50 mg to 25 mg was associated with worsening control of ankylosing spondylitis in an open-label, multicenter, pilot study.

Results of the ANSWER (Ankylosing Spondylitis with Etanercept Regimens) trial showed that halving the dose of the biologic almost halved the percentage of patients maintaining disease control at 6 months, from 92% to 52% of patients.

Although this was a small study, involving just 47 patients who were randomized to continue etanercept at a weekly dose of 50 mg (n = 24) or to drop down to 25 mg (n = 23) after competing 6 months’ treatment, the results suggest that other approaches are needed to see if the long-term dose of the drug can be reduced, commented study investigator Dr. Karl Gaffney of Norfolk and Norwich University Hospital NHS Foundation Trust, Norwich, England.

"A larger study is required to identify patients suitable for step-down," Dr. Gaffney said at the British Society for Rheumatology annual conference.

"We all know that anti-TNF [tumor necrosis factor] therapy is firmly established in clinical practice for patients with ankylosing spondylitis and this treatment has transformed the quality of life for many of our patients," Dr. Gaffney observed.

"The concept of dose reduction is very appealing from an economic perspective, but also in terms of the cumulative exposure and the risk of adverse effects," he added.

"In patients with rheumatoid arthritis, it’s been shown that lower doses may maintain the clinical response," Dr. Gaffney explained, citing the PRESERVE study published last year (Lancet 2013;381:918-29). However, there are limited data about whether reducing the dose of etanercept could also be successful in ankylosing spondylitis (AS) patients. This is why the ANSWER study was performed.

Patients with active AS who were not responding to conventional therapies and had not yet been treated with a biologic drug were consecutively recruited at two hospitals in England between September 2010 and September 2012. For inclusion, patients had to have sustained spinal disease and a Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) score of four or higher. The mean age of patients in the trial was 46.7 years, and the mean BASDAI at baseline was 6.8.

All recruited patients were treated with 50 mg etanercept, once weekly, for 6 months, and those with a "sufficient" clinical response were randomized to either continue this dose for a further 6 months or step down to 25 mg for continued maintenance treatment. A sufficient clinical response was defined as a 50% reduction in BASDAI or a fall of two or more units plus reduction in axial pain of 2 cm or more.

Three months into the maintenance period, patients who had been randomized to the step-down arm started to experience loss of disease control (60.9% vs. 91.7% of 50 mg–treated patients), with the gap widening by 6 months postrandomization and 39% fewer patients maintaining disease control (P = .003).

"There’s no doubt that the 50-mg group did better," Dr. Gaffney said. There were also statistically significant differences favoring the higher dose in a number of secondary outcome measures, which included other measurements of AS disease activity and quality of life. There were no significant differences in terms of adverse events between the two groups.

"However, 52% of the 25-mg arm did maintain their BASDAI response. So theoretically, there may be a subset of patients in whom we can offer this therapeutic option," Dr. Gaffney said. He noted that even in this small group of patients with short-term follow-up that the potential cost savings of being able to reduce the dose would be significant, at around £100,000 (about U.S. $170,000).

"We all have patients who do very well [on etanercept], and I think it would be very nice to be able to say, from 6 months, who we can attempt to step down," commented Dr. Nicola Goodson of University Hospital Aintree, Liverpool. Dr. Goodson chaired the session where the study findings were reported.

Although the ANSWER investigators did look for predictors of response, none were found to be significant. Patients continue to be monitored and the investigators may look at this again. Dr. Gaffney noted that all patients who had their 50-mg dose reinstated regained their clinical response.

He also highlighted during discussion: "A number of patients in the 50-mg group asked to be moved down to 25 mg at the end of the study because they wanted to explore the option. So I think this may well have some long-term consequences, especially in a disease area where we know that the treatment is more symptom modifying than radiographically modifying."

The study was funded by Pfizer. Dr. Gaffney has received research support and honoraria from Pfizer, Merck Sharp & Dohme, AbbVie, and UCB.

LIVERPOOL, ENGLAND – The risk of recurrent cancer in patients with rheumatoid arthritis did not increase with the use of biologic therapies, according to data just released from the British Society for Rheumatology Biologics Registers’ Rheumatoid Arthritis Register.

In fact, the risk of repeated cancer in patients with prior malignancies treated with biologic therapies was apparently decreased by around 50%, when compared with treatment with nonbiologic disease-modifying antirheumatic drugs (nbDMARDs).

The hazard ratios, adjusted for age and sex, for recurrent cancer were 0.55 for patients treated with drugs directed at tumor necrosis factor (TNF)-alpha and 0.47 for patients treated, off-label, with rituximab vs. nbDMARDs.

However, Dr. Luca Silva-Fernandez, who presented the findings at the British Society for Rheumatology annual conference, noted that patients treated with nbDMARDs were perhaps at higher risk of recurrent disease than were those who were treated with biologic agents at the start of their treatment, so the data do not imply that biologic agents are less likely to cause recurrent disease than do the older RA therapies.

"Our data suggest that patients with RA and prior malignancy selected to receive either an anti-TNF or rituximab therapy in the U.K. do not seem to have an increased risk of future incident malignancy," said Dr. Silva-Fernandez.

Patients treated with anti-TNFs or rituximab were more likely to have incident cancers, added Dr. Silva-Fernandez of the Arthritis Research UK Epidemiology Unit at the University of Manchester, England, where the BSRBR-RA Register is run.

Previous data on the use of biologic therapies in patients with prior cancer were published from the British Society for Rheumatology Biologics Registers (BSRBR) 4 years ago (Arthritis Care Res. 2010;62:755-63). These considered 293 patients with a prior malignancy identified from over 14,000 patients with RA enrolled in the register at the time. The present analysis included 425 patients who had cancer before being enrolled in the BSRBR-RA Register, which at the time of the updated analysis included around 19,000 patients.

In total, 425 new malignancies were detected in 159 of 3,787 (1.7%) patients who had been treated with an nbDMARD, 243 of 14,168 (8.9%) patients treated with an anti-TNF drug, and 23 of 257 (4.2%) patients who had received rituximab for the treatment of their RA.

There were several differences in baseline characteristics among the three groups. The mean age of patients in each treatment arm was 66.1, 62.7, and 67.3 years, respectively. A higher percentage (81%) of anti-TNF-treated patients were female when compared with the nbDMARD (74%) and rituximab (65%) arms. Biologic-treated patients tended to have a longer mean RA disease duration (anti-TNF 12 years and rituximab 14 years) than did the nbDMARD-treated patients (8 years), and higher disease severity scores at enrollment into the BSRBR-RA Register. The median time between the previous and current malignancy was 7.9 years for nbDMARDs, 11.5 years for anti-TNFs, and 5.4 years for rituximab.

The crude incident malignancy rates were 47/1,000 person-years in the nbDMARD cohort, 24/1,000 person-years in the anti-TNF-treated patients, and 25/1,000 person-years in the group treated with rituximab. Dr. Silva-Fernandez reported that the median follow-up was much shorter in rituximab-treated patients, at 3.9 years, than both the nbDMARD- (6.6 years) and anti-TNF–treated (6.9 years) groups.

After her presentation, Dr. Silva-Fernandez was asked to comment on the apparent "protective" effect of anti-TNFs on the recurrence of cancer. She replied that the nbDMARD and biologic groups were not really comparable, referring back to the differences in baseline characteristics, so such an association cannot be claimed.

With regards to a question on the use of rituximab in RA patients with a history of prior cancer, which was "off label" in this instance as it wasn’t used after anti-TNFs but "up-front," this might reflect a "channeling bias" on the part of the physicians, commented Dr. Kimmie Hyrich.

"These patients were those who had received rituximab as their first biologic for a number of reasons," said Dr. Hyrich, one of the principal investigators for the BSRBR-RA Register. "For many of them, it may have been a past cancer that made that decision," she added.

"I think, as physicians, we have a comfort in using rituximab in patients with past cancer, because it is not an absolute contraindication, so I think this is physician choice, and that’s probably why an off-license decision to treat with rituximab was made in these patients," said Dr. Hyrich, also from the University of Manchester, England.

Although patients with nonmelanoma skin cancer (NSMC) were excluded from the present analysis, Dr. Hyrich also noted in response to a question that the BSRBR RA Register team had previously reported on the rates of recurrence in patients with and without a prior history of this type of skin cancer.

Findings had shown that anti-TNFs did not appear to increase the risk of NSMC in patients without a prior history of skin cancer. In patients with a history of the disease, there was elevated risk of recurrence, regardless of whether patients received nbDMARDs or anti-TNFs, and it did not seem to occur more in the biologic-treated patients.

"I think the most striking finding, however, was regardless of treatment, all patients in the register, overall, have a marked increased risk of skin cancer, compared with general population, so I think RA itself and its treatment is probably the strongest risk factor," Dr. Hyrich said.

The BSRBR is funded by a grant from the British Society for Rheumatology (BSR), which receives funding from multiple drug companies. This income finances a separate contract between the BSR and the University of Manchester that provides and runs the BSRBR RA Register. Dr. Silva-Fernandez and Dr. Hyrich had no personal conflicts of interest.

LIVERPOOL, ENGLAND – The risk of recurrent cancer in patients with rheumatoid arthritis did not increase with the use of biologic therapies, according to data just released from the British Society for Rheumatology Biologics Registers’ Rheumatoid Arthritis Register.

In fact, the risk of repeated cancer in patients with prior malignancies treated with biologic therapies was apparently decreased by around 50%, when compared with treatment with nonbiologic disease-modifying antirheumatic drugs (nbDMARDs).

The hazard ratios, adjusted for age and sex, for recurrent cancer were 0.55 for patients treated with drugs directed at tumor necrosis factor (TNF)-alpha and 0.47 for patients treated, off-label, with rituximab vs. nbDMARDs.

However, Dr. Luca Silva-Fernandez, who presented the findings at the British Society for Rheumatology annual conference, noted that patients treated with nbDMARDs were perhaps at higher risk of recurrent disease than were those who were treated with biologic agents at the start of their treatment, so the data do not imply that biologic agents are less likely to cause recurrent disease than do the older RA therapies.

"Our data suggest that patients with RA and prior malignancy selected to receive either an anti-TNF or rituximab therapy in the U.K. do not seem to have an increased risk of future incident malignancy," said Dr. Silva-Fernandez.

Patients treated with anti-TNFs or rituximab were more likely to have incident cancers, added Dr. Silva-Fernandez of the Arthritis Research UK Epidemiology Unit at the University of Manchester, England, where the BSRBR-RA Register is run.

Previous data on the use of biologic therapies in patients with prior cancer were published from the British Society for Rheumatology Biologics Registers (BSRBR) 4 years ago (Arthritis Care Res. 2010;62:755-63). These considered 293 patients with a prior malignancy identified from over 14,000 patients with RA enrolled in the register at the time. The present analysis included 425 patients who had cancer before being enrolled in the BSRBR-RA Register, which at the time of the updated analysis included around 19,000 patients.

In total, 425 new malignancies were detected in 159 of 3,787 (1.7%) patients who had been treated with an nbDMARD, 243 of 14,168 (8.9%) patients treated with an anti-TNF drug, and 23 of 257 (4.2%) patients who had received rituximab for the treatment of their RA.

There were several differences in baseline characteristics among the three groups. The mean age of patients in each treatment arm was 66.1, 62.7, and 67.3 years, respectively. A higher percentage (81%) of anti-TNF-treated patients were female when compared with the nbDMARD (74%) and rituximab (65%) arms. Biologic-treated patients tended to have a longer mean RA disease duration (anti-TNF 12 years and rituximab 14 years) than did the nbDMARD-treated patients (8 years), and higher disease severity scores at enrollment into the BSRBR-RA Register. The median time between the previous and current malignancy was 7.9 years for nbDMARDs, 11.5 years for anti-TNFs, and 5.4 years for rituximab.

The crude incident malignancy rates were 47/1,000 person-years in the nbDMARD cohort, 24/1,000 person-years in the anti-TNF-treated patients, and 25/1,000 person-years in the group treated with rituximab. Dr. Silva-Fernandez reported that the median follow-up was much shorter in rituximab-treated patients, at 3.9 years, than both the nbDMARD- (6.6 years) and anti-TNF–treated (6.9 years) groups.

After her presentation, Dr. Silva-Fernandez was asked to comment on the apparent "protective" effect of anti-TNFs on the recurrence of cancer. She replied that the nbDMARD and biologic groups were not really comparable, referring back to the differences in baseline characteristics, so such an association cannot be claimed.

With regards to a question on the use of rituximab in RA patients with a history of prior cancer, which was "off label" in this instance as it wasn’t used after anti-TNFs but "up-front," this might reflect a "channeling bias" on the part of the physicians, commented Dr. Kimmie Hyrich.

"These patients were those who had received rituximab as their first biologic for a number of reasons," said Dr. Hyrich, one of the principal investigators for the BSRBR-RA Register. "For many of them, it may have been a past cancer that made that decision," she added.

"I think, as physicians, we have a comfort in using rituximab in patients with past cancer, because it is not an absolute contraindication, so I think this is physician choice, and that’s probably why an off-license decision to treat with rituximab was made in these patients," said Dr. Hyrich, also from the University of Manchester, England.

Although patients with nonmelanoma skin cancer (NSMC) were excluded from the present analysis, Dr. Hyrich also noted in response to a question that the BSRBR RA Register team had previously reported on the rates of recurrence in patients with and without a prior history of this type of skin cancer.

Findings had shown that anti-TNFs did not appear to increase the risk of NSMC in patients without a prior history of skin cancer. In patients with a history of the disease, there was elevated risk of recurrence, regardless of whether patients received nbDMARDs or anti-TNFs, and it did not seem to occur more in the biologic-treated patients.

"I think the most striking finding, however, was regardless of treatment, all patients in the register, overall, have a marked increased risk of skin cancer, compared with general population, so I think RA itself and its treatment is probably the strongest risk factor," Dr. Hyrich said.

The BSRBR is funded by a grant from the British Society for Rheumatology (BSR), which receives funding from multiple drug companies. This income finances a separate contract between the BSR and the University of Manchester that provides and runs the BSRBR RA Register. Dr. Silva-Fernandez and Dr. Hyrich had no personal conflicts of interest.

LIVERPOOL, ENGLAND – The risk of recurrent cancer in patients with rheumatoid arthritis did not increase with the use of biologic therapies, according to data just released from the British Society for Rheumatology Biologics Registers’ Rheumatoid Arthritis Register.

In fact, the risk of repeated cancer in patients with prior malignancies treated with biologic therapies was apparently decreased by around 50%, when compared with treatment with nonbiologic disease-modifying antirheumatic drugs (nbDMARDs).

The hazard ratios, adjusted for age and sex, for recurrent cancer were 0.55 for patients treated with drugs directed at tumor necrosis factor (TNF)-alpha and 0.47 for patients treated, off-label, with rituximab vs. nbDMARDs.

However, Dr. Luca Silva-Fernandez, who presented the findings at the British Society for Rheumatology annual conference, noted that patients treated with nbDMARDs were perhaps at higher risk of recurrent disease than were those who were treated with biologic agents at the start of their treatment, so the data do not imply that biologic agents are less likely to cause recurrent disease than do the older RA therapies.

"Our data suggest that patients with RA and prior malignancy selected to receive either an anti-TNF or rituximab therapy in the U.K. do not seem to have an increased risk of future incident malignancy," said Dr. Silva-Fernandez.

Patients treated with anti-TNFs or rituximab were more likely to have incident cancers, added Dr. Silva-Fernandez of the Arthritis Research UK Epidemiology Unit at the University of Manchester, England, where the BSRBR-RA Register is run.

Previous data on the use of biologic therapies in patients with prior cancer were published from the British Society for Rheumatology Biologics Registers (BSRBR) 4 years ago (Arthritis Care Res. 2010;62:755-63). These considered 293 patients with a prior malignancy identified from over 14,000 patients with RA enrolled in the register at the time. The present analysis included 425 patients who had cancer before being enrolled in the BSRBR-RA Register, which at the time of the updated analysis included around 19,000 patients.

In total, 425 new malignancies were detected in 159 of 3,787 (1.7%) patients who had been treated with an nbDMARD, 243 of 14,168 (8.9%) patients treated with an anti-TNF drug, and 23 of 257 (4.2%) patients who had received rituximab for the treatment of their RA.

There were several differences in baseline characteristics among the three groups. The mean age of patients in each treatment arm was 66.1, 62.7, and 67.3 years, respectively. A higher percentage (81%) of anti-TNF-treated patients were female when compared with the nbDMARD (74%) and rituximab (65%) arms. Biologic-treated patients tended to have a longer mean RA disease duration (anti-TNF 12 years and rituximab 14 years) than did the nbDMARD-treated patients (8 years), and higher disease severity scores at enrollment into the BSRBR-RA Register. The median time between the previous and current malignancy was 7.9 years for nbDMARDs, 11.5 years for anti-TNFs, and 5.4 years for rituximab.

The crude incident malignancy rates were 47/1,000 person-years in the nbDMARD cohort, 24/1,000 person-years in the anti-TNF-treated patients, and 25/1,000 person-years in the group treated with rituximab. Dr. Silva-Fernandez reported that the median follow-up was much shorter in rituximab-treated patients, at 3.9 years, than both the nbDMARD- (6.6 years) and anti-TNF–treated (6.9 years) groups.

After her presentation, Dr. Silva-Fernandez was asked to comment on the apparent "protective" effect of anti-TNFs on the recurrence of cancer. She replied that the nbDMARD and biologic groups were not really comparable, referring back to the differences in baseline characteristics, so such an association cannot be claimed.

With regards to a question on the use of rituximab in RA patients with a history of prior cancer, which was "off label" in this instance as it wasn’t used after anti-TNFs but "up-front," this might reflect a "channeling bias" on the part of the physicians, commented Dr. Kimmie Hyrich.

"These patients were those who had received rituximab as their first biologic for a number of reasons," said Dr. Hyrich, one of the principal investigators for the BSRBR-RA Register. "For many of them, it may have been a past cancer that made that decision," she added.

"I think, as physicians, we have a comfort in using rituximab in patients with past cancer, because it is not an absolute contraindication, so I think this is physician choice, and that’s probably why an off-license decision to treat with rituximab was made in these patients," said Dr. Hyrich, also from the University of Manchester, England.

Although patients with nonmelanoma skin cancer (NSMC) were excluded from the present analysis, Dr. Hyrich also noted in response to a question that the BSRBR RA Register team had previously reported on the rates of recurrence in patients with and without a prior history of this type of skin cancer.

Findings had shown that anti-TNFs did not appear to increase the risk of NSMC in patients without a prior history of skin cancer. In patients with a history of the disease, there was elevated risk of recurrence, regardless of whether patients received nbDMARDs or anti-TNFs, and it did not seem to occur more in the biologic-treated patients.

"I think the most striking finding, however, was regardless of treatment, all patients in the register, overall, have a marked increased risk of skin cancer, compared with general population, so I think RA itself and its treatment is probably the strongest risk factor," Dr. Hyrich said.

The BSRBR is funded by a grant from the British Society for Rheumatology (BSR), which receives funding from multiple drug companies. This income finances a separate contract between the BSR and the University of Manchester that provides and runs the BSRBR RA Register. Dr. Silva-Fernandez and Dr. Hyrich had no personal conflicts of interest.

LIVERPOOL, ENGLAND – A new analysis of data from the British Society for Rheumatology Biologics Registers’ Rheumatoid Arthritis Register confirms that anti–tumor necrosis factor treatment decreases patients’ risk for heart attack when compared with treatment with traditional, nonbiologic disease-modifying antirheumatic drugs, but a lower level of inflammation may not be the mechanism of action.

In the latest analysis, the risk of myocardial infarction was reduced by 40% if patients had been treated with an anti–tumor necrosis factor (anti-TNF) drug rather than a nonbiologic disease-modifying antirheumatic drug (nbDMARD) (adjusted hazard ratio, 0.6).

But this effect was not the result of a reduction in the severity of inflammation, said researcher Dr. Audrey Low of the Arthritis Research UK Epidemiology Unit at the University of Manchester, England, where the British Society for Rheumatology Biologics Registers’ Rheumatoid Arthritis (BSRBR-RA) Register is run. Nevertheless, the effect could still relate to attributes of these drugs themselves or better overall disease control, she and her coworkers reported at the British Society for Rheumatology annual conference.

"It is now well known that our patients with rheumatoid arthritis are at increased risk of cardiovascular disease," Dr. Low observed, adding that the risk of both cardiovascular morbidity and mortality is around 50% higher than in the general population. "In particular, the risk of MI [myocardial infarction] is increased," she noted.

Inflammation is thought to play an important role in the development of both atherosclerosis and rheumatoid arthritis, and thus dampening down the inflammatory response with anti-TNF drugs could potentially modify the elevated cardiovascular risk seen in RA patients.

Previously, Dr. Low reported that the risk of MI was lowered by 30% in patients treated with anti-TNFs versus nbDMARDs. The current study looked more deeply into the possible relationship, linking BSRBR-RA Register data to those in the Myocardial Ischaemia National Audit Project (MINAP), a large MI data set derived from all hospitalizations for heart attack in England and Wales. MINAP was established in 1999, 2 years before the BSRBR-RA Register, and it collects around 90,000 records of possible MI per year. As of 2012, MINAP consisted of more than 1 million records.

The aims of the current analysis, which included 14,258 patients with active RA, were to first look at the incidence of MI in patients treated with anti-TNFs (n = 11,200) versus nbDMARDs (n = 3,058), and then to see if the severity of MI was influenced by biologic treatment. The analysis included all patients who started treatment with one of three established anti-TNFs (etanercept, infliximab, and adalimumab) and were recruited into the BSRBR-RA Register between 2001 and 2008. Patients with prior cardiovascular disease were excluded.

To verify the occurrence of MI, data from the BSRBR-RA Register were linked to data from MINAP using patients’ forenames, surnames, date of birth, National Health Service number, postal code, and gender. Events occurring in the same 30-day time window in both data sets were considered the same event. MIs were then verified using criteria set by the American Heart Association and the European Society for Cardiology, with additional criteria of thrombolysis, angioplasty, and MI listed as the underlying cause of death on death certificates, based on the World Health Organization’s International Classification of Diseases, version 10.

A total of 252 first MIs were analyzed, of which 194 occurred in anti-TNF–treated patients at a median follow-up of 5.3 years and 58 occurred in those treated with nbDMARDs at a median follow-up of 3.5 years. The corresponding crude incidence rates were 3.5 and 5.6 per 1,000 patient-years.

"Looking at the MI severity, we only analyzed those MIs that had MINAP-associated data," which was just over half of the MIs (n = 143), Dr. Low said. While the absolute levels of three cardiac enzymes measured – peak troponin I, peak troponin T, and peak creatinine kinase – were higher in the anti-TNF cohort, compared with nbDMARDs, there were not enough events in each group to achieve statistical significance.

There were also no differences between the treatment groups in terms of other cardiac measures of severity, including the phenotype of MI (ST- vs. non-ST elevated), cardiac arrest, and length of hospitalization.

"So the second conclusion is that severity and post-MI mortality do not appear to be influenced by anti-TNF therapy," Dr. Low said.

BSR president Dr. Chris Deighton, who chaired the plenary session in which the findings were presented, noted that there seemed to be a high proportion of patients in the anti-TNF arm who were also treated with steroidal or nonsteroidal anti-inflammatory drugs (NSAIDs) at enrollment. Perhaps if the anti-TNFs were working well, then any reduction in the use of these other drugs over time might have had a confounding effect on the MI risk.

Dr. Low noted that the data on the use of steroids was quite crude and often the dose was not recorded, and even less information about NSAIDs was available. "This is something that we will have to work on, and it does remain to be a potential confounding factor in the analysis," she concluded.

The BSRBR is funded by a grant from the BSR, which receives funding from multiple drug companies. This income finances a separate contract between the BSR and the University of Manchester that provides and runs the BSRBR. Dr. Low had no personal conflicts of interest.

LIVERPOOL, ENGLAND – A new analysis of data from the British Society for Rheumatology Biologics Registers’ Rheumatoid Arthritis Register confirms that anti–tumor necrosis factor treatment decreases patients’ risk for heart attack when compared with treatment with traditional, nonbiologic disease-modifying antirheumatic drugs, but a lower level of inflammation may not be the mechanism of action.

In the latest analysis, the risk of myocardial infarction was reduced by 40% if patients had been treated with an anti–tumor necrosis factor (anti-TNF) drug rather than a nonbiologic disease-modifying antirheumatic drug (nbDMARD) (adjusted hazard ratio, 0.6).

But this effect was not the result of a reduction in the severity of inflammation, said researcher Dr. Audrey Low of the Arthritis Research UK Epidemiology Unit at the University of Manchester, England, where the British Society for Rheumatology Biologics Registers’ Rheumatoid Arthritis (BSRBR-RA) Register is run. Nevertheless, the effect could still relate to attributes of these drugs themselves or better overall disease control, she and her coworkers reported at the British Society for Rheumatology annual conference.

"It is now well known that our patients with rheumatoid arthritis are at increased risk of cardiovascular disease," Dr. Low observed, adding that the risk of both cardiovascular morbidity and mortality is around 50% higher than in the general population. "In particular, the risk of MI [myocardial infarction] is increased," she noted.

Inflammation is thought to play an important role in the development of both atherosclerosis and rheumatoid arthritis, and thus dampening down the inflammatory response with anti-TNF drugs could potentially modify the elevated cardiovascular risk seen in RA patients.

Previously, Dr. Low reported that the risk of MI was lowered by 30% in patients treated with anti-TNFs versus nbDMARDs. The current study looked more deeply into the possible relationship, linking BSRBR-RA Register data to those in the Myocardial Ischaemia National Audit Project (MINAP), a large MI data set derived from all hospitalizations for heart attack in England and Wales. MINAP was established in 1999, 2 years before the BSRBR-RA Register, and it collects around 90,000 records of possible MI per year. As of 2012, MINAP consisted of more than 1 million records.

The aims of the current analysis, which included 14,258 patients with active RA, were to first look at the incidence of MI in patients treated with anti-TNFs (n = 11,200) versus nbDMARDs (n = 3,058), and then to see if the severity of MI was influenced by biologic treatment. The analysis included all patients who started treatment with one of three established anti-TNFs (etanercept, infliximab, and adalimumab) and were recruited into the BSRBR-RA Register between 2001 and 2008. Patients with prior cardiovascular disease were excluded.

To verify the occurrence of MI, data from the BSRBR-RA Register were linked to data from MINAP using patients’ forenames, surnames, date of birth, National Health Service number, postal code, and gender. Events occurring in the same 30-day time window in both data sets were considered the same event. MIs were then verified using criteria set by the American Heart Association and the European Society for Cardiology, with additional criteria of thrombolysis, angioplasty, and MI listed as the underlying cause of death on death certificates, based on the World Health Organization’s International Classification of Diseases, version 10.

A total of 252 first MIs were analyzed, of which 194 occurred in anti-TNF–treated patients at a median follow-up of 5.3 years and 58 occurred in those treated with nbDMARDs at a median follow-up of 3.5 years. The corresponding crude incidence rates were 3.5 and 5.6 per 1,000 patient-years.

"Looking at the MI severity, we only analyzed those MIs that had MINAP-associated data," which was just over half of the MIs (n = 143), Dr. Low said. While the absolute levels of three cardiac enzymes measured – peak troponin I, peak troponin T, and peak creatinine kinase – were higher in the anti-TNF cohort, compared with nbDMARDs, there were not enough events in each group to achieve statistical significance.

There were also no differences between the treatment groups in terms of other cardiac measures of severity, including the phenotype of MI (ST- vs. non-ST elevated), cardiac arrest, and length of hospitalization.

"So the second conclusion is that severity and post-MI mortality do not appear to be influenced by anti-TNF therapy," Dr. Low said.

BSR president Dr. Chris Deighton, who chaired the plenary session in which the findings were presented, noted that there seemed to be a high proportion of patients in the anti-TNF arm who were also treated with steroidal or nonsteroidal anti-inflammatory drugs (NSAIDs) at enrollment. Perhaps if the anti-TNFs were working well, then any reduction in the use of these other drugs over time might have had a confounding effect on the MI risk.

Dr. Low noted that the data on the use of steroids was quite crude and often the dose was not recorded, and even less information about NSAIDs was available. "This is something that we will have to work on, and it does remain to be a potential confounding factor in the analysis," she concluded.

The BSRBR is funded by a grant from the BSR, which receives funding from multiple drug companies. This income finances a separate contract between the BSR and the University of Manchester that provides and runs the BSRBR. Dr. Low had no personal conflicts of interest.

LIVERPOOL, ENGLAND – A new analysis of data from the British Society for Rheumatology Biologics Registers’ Rheumatoid Arthritis Register confirms that anti–tumor necrosis factor treatment decreases patients’ risk for heart attack when compared with treatment with traditional, nonbiologic disease-modifying antirheumatic drugs, but a lower level of inflammation may not be the mechanism of action.

In the latest analysis, the risk of myocardial infarction was reduced by 40% if patients had been treated with an anti–tumor necrosis factor (anti-TNF) drug rather than a nonbiologic disease-modifying antirheumatic drug (nbDMARD) (adjusted hazard ratio, 0.6).

But this effect was not the result of a reduction in the severity of inflammation, said researcher Dr. Audrey Low of the Arthritis Research UK Epidemiology Unit at the University of Manchester, England, where the British Society for Rheumatology Biologics Registers’ Rheumatoid Arthritis (BSRBR-RA) Register is run. Nevertheless, the effect could still relate to attributes of these drugs themselves or better overall disease control, she and her coworkers reported at the British Society for Rheumatology annual conference.

"It is now well known that our patients with rheumatoid arthritis are at increased risk of cardiovascular disease," Dr. Low observed, adding that the risk of both cardiovascular morbidity and mortality is around 50% higher than in the general population. "In particular, the risk of MI [myocardial infarction] is increased," she noted.

Inflammation is thought to play an important role in the development of both atherosclerosis and rheumatoid arthritis, and thus dampening down the inflammatory response with anti-TNF drugs could potentially modify the elevated cardiovascular risk seen in RA patients.

Previously, Dr. Low reported that the risk of MI was lowered by 30% in patients treated with anti-TNFs versus nbDMARDs. The current study looked more deeply into the possible relationship, linking BSRBR-RA Register data to those in the Myocardial Ischaemia National Audit Project (MINAP), a large MI data set derived from all hospitalizations for heart attack in England and Wales. MINAP was established in 1999, 2 years before the BSRBR-RA Register, and it collects around 90,000 records of possible MI per year. As of 2012, MINAP consisted of more than 1 million records.

The aims of the current analysis, which included 14,258 patients with active RA, were to first look at the incidence of MI in patients treated with anti-TNFs (n = 11,200) versus nbDMARDs (n = 3,058), and then to see if the severity of MI was influenced by biologic treatment. The analysis included all patients who started treatment with one of three established anti-TNFs (etanercept, infliximab, and adalimumab) and were recruited into the BSRBR-RA Register between 2001 and 2008. Patients with prior cardiovascular disease were excluded.

To verify the occurrence of MI, data from the BSRBR-RA Register were linked to data from MINAP using patients’ forenames, surnames, date of birth, National Health Service number, postal code, and gender. Events occurring in the same 30-day time window in both data sets were considered the same event. MIs were then verified using criteria set by the American Heart Association and the European Society for Cardiology, with additional criteria of thrombolysis, angioplasty, and MI listed as the underlying cause of death on death certificates, based on the World Health Organization’s International Classification of Diseases, version 10.

A total of 252 first MIs were analyzed, of which 194 occurred in anti-TNF–treated patients at a median follow-up of 5.3 years and 58 occurred in those treated with nbDMARDs at a median follow-up of 3.5 years. The corresponding crude incidence rates were 3.5 and 5.6 per 1,000 patient-years.

"Looking at the MI severity, we only analyzed those MIs that had MINAP-associated data," which was just over half of the MIs (n = 143), Dr. Low said. While the absolute levels of three cardiac enzymes measured – peak troponin I, peak troponin T, and peak creatinine kinase – were higher in the anti-TNF cohort, compared with nbDMARDs, there were not enough events in each group to achieve statistical significance.

There were also no differences between the treatment groups in terms of other cardiac measures of severity, including the phenotype of MI (ST- vs. non-ST elevated), cardiac arrest, and length of hospitalization.

"So the second conclusion is that severity and post-MI mortality do not appear to be influenced by anti-TNF therapy," Dr. Low said.

BSR president Dr. Chris Deighton, who chaired the plenary session in which the findings were presented, noted that there seemed to be a high proportion of patients in the anti-TNF arm who were also treated with steroidal or nonsteroidal anti-inflammatory drugs (NSAIDs) at enrollment. Perhaps if the anti-TNFs were working well, then any reduction in the use of these other drugs over time might have had a confounding effect on the MI risk.

Dr. Low noted that the data on the use of steroids was quite crude and often the dose was not recorded, and even less information about NSAIDs was available. "This is something that we will have to work on, and it does remain to be a potential confounding factor in the analysis," she concluded.

The BSRBR is funded by a grant from the BSR, which receives funding from multiple drug companies. This income finances a separate contract between the BSR and the University of Manchester that provides and runs the BSRBR. Dr. Low had no personal conflicts of interest.