User login
An uncommon but potentially deadly inflammatory lung disease is emerging among children with systemic juvenile idiopathic arthritis, and its history appears to coincide with the rise of powerful biologics as first-line therapy for children with the disease.
Most confirmed cases of systemic juvenile idiopathic arthritis with lung disease (sJIA-LD) are in the United States. But it’s popping up in other places that have adopted early biologic treatment for sJIA – including Canada, South America, Europe, and the Middle East.
The respiratory symptoms are relatively subtle, so by the time of lung disease detection, the amount of affected lung can be extensive, said Elizabeth Mellins, MD, a Stanford (Calif.) University researcher who, along with first author Vivian Saper, MD, recently published the largest case series comprising reports from 37 institutions (Ann Rheum Dis. 2019 Sep 27. doi: 10.1136/annrheumdis-2019-216040). By the end of follow-up, 22 of the 61 children in her cohort had died, including all 12 patients who demonstrated excessively high neutrophil levels in bronchoalveolar lavage samples.
Another recent report, authored by Grant Schulert, MD, PhD, and colleagues of the Cincinnati Children’s Hospital Medical Center, described 18 patients, 9 of whom were also included in the Stanford cohort (Arthritis Rheumatol. 2019 Aug 5. doi: 10.1002/art.41073).
Both investigators have now identified new patients.
“We are aware of 60 additional cases beyond what were included in our series,” Dr. Mellins said in an interview, bringing her entire cohort to 121. Dr. Schulert also continues to expand his group, detailing nine new cases at a recent private meeting.
“We are up to 27 now,” he said. “The features of these new patients are all very similar: The children are very young, all have had macrophage activation syndrome in the past and very-difficult-to-control JIA. Reactions to tocilizumab [Actemra] were also not uncommon in this group.”
Dr. Mellins also saw this association with allergic-type tocilizumab reactions, severe delayed hypersensitivity reactions to anakinra (Kineret) or canakinumab (Ilaris). Although serious lung disease in sJIA patients is not unheard of, this phenotype was virtually unknown until about a decade ago. Both investigators said that it’s been rising steadily since 2010 – just about the time that powerful cytokine-inhibiting biologics were changing these patients’ world for the better. After decades of relying almost solely on steroids and methotrexate, with rather poor results and significant long-term side effects, children were not only improving, but thriving. Gone was the life-changing glucocorticoid-related growth inhibition. Biologics could halt fevers, rash, and joint destruction in their tracks.
“For the first time in history, these kids could look forward to a more or less normal life,” Dr. Schulert said.
But the emergence of this particular type of lung disease could throw a pall over that success story, he said. If sJIA-LD is temporally associated with increasing reliance on long-term interleukin-1/IL-6 inhibition in children with early-onset disease, could these drugs actually be the causative agent? The picture remains unclear.
Some of the 18 in his initial series have improved, while 36% of those in the Stanford series died. Most who do recover stay on their IL-1 or IL-6 blocking therapy with good disease control without further lung problems. Both investigators found compelling genetic hints, but nothing conclusive. Children with trisomy 21 appear especially vulnerable. Most patients are very young – around 2 years old – but others are school aged. Some had a history of macrophage activation syndrome. Some had hard-to-control disease and some were clinically well controlled when the lung disease presented.
There are simply no answers yet.
With so many potential links, all unproven, clinicians may question the wisdom of embarking on long-term biologic therapy for their children with sJIA. Peter Nigrovic, MD, of Boston Children’s Hospital, addressed this in an accompanying editorial (Arthritis Rheumatol. 2019 Aug 7. doi: 10.1002/art.41071).
“My take on this is that it’s a very worrisome trend,” he said in an interview. “We’ve been going full bore toward early biologic therapy in sJIA and at the same time we are seeing more of this lung disease. Is it guilt by association? Or is there something more? The challenge for us is not to jump too soon to that conclusion.”
Although the association is there, he said, association does not equal causation. And there’s no doubt that biologics have vastly improved the lives of sJIA patients. “The drugs might be causal, and I worry about that and think we need to study it. But we absolutely need stronger evidence before we change practice.”
“This is a new manifestation of the disease, and it’s coming at the same time we are changing the treatment paradigm,” Dr. Nigrovic continued. “It could be because of interleukin-1 or interleukin-6 blockade. There is biological plausibility for such a link. It could also be related to the fact that we are using less steroids and methotrexate, which might have been preventing this. The appearance of sJIA lung disease could also be that a distinct secular trend unrelated to treatment, just as we saw amyloid come and go in this population in Europe. These other therapies were actually preventing this. We just don’t know.”
Clinical characteristics
Children presented with similar symptoms. Respiratory symptoms are usually subtle and mild. These can include tachypnea, hypoxia (43% in the Stanford series), and pulmonary hypertension (30% in the Stanford series).
Digital clubbing, often with erythema, was a common finding. Some children showed pruritic, nonevanescent rashes. Eosinophilia occurred in 37% of the Stanford series and severe abdominal pain in 16%, although Dr. Mellins noted that belly pain may be underestimated, as it was only volunteered, not queried, information.
“There are some red flags that should raise suspicion even without obvious respiratory symptoms,” Dr. Mellins said. These include lymphopenia, unexplained abdominal pain, eosinophilia, an unusual rash, and finger clubbing with or without erythema.
Findings on imaging were consistent in both series. Several key clinic features emerged: pleural thickening, septal thickening, bronchial wall or peribronchovascular thickening, “tree-in-bud” opacities, “ground-glass” opacities, peripheral consolidation, and lymphadenopathy.
“The imaging findings correspond to two things,” Dr. Schulert said. “The first is inflammation in the interstitium, which is evidence of chronic and ongoing inflammation. The other thing is that the alveoli are filled with a lipoproteinaceous material which is actually surfactant that’s not being normally recycled by the lung macrophages. You can see these features in other conditions where there’s a problem with lung macrophages, like pulmonary alveolar proteinosis, genetic and autoimmune disorders, infections, or inhalants.”
Pathology showed alveolar filling – a location in the lung that hides usual symptoms until the lung disease is advanced. Prior drug reactions were common. Tocilizumab anaphylaxis occurred in close to 40% of the Stanford series – a surprising finding given the 0.6% reaction incidence in the drug’s sJIA trials. Dr. Schulert saw a similar story.
“In our cohort we also observed a striking number of adverse events to cytokine-targeted biologics exposure,” Dr. Schulert said. “Most of these reactions were to tocilizumab, and were described variously from pain and feeling unwell, to difficulty breathing, to anaphylaxis.”
In a risk analysis, Dr. Schulert determined that adverse events to cytokine-targeting biologics increased the likelihood of lung disease more than 13 times (odds ratio, 13.6).
“We also identified a statistically significant association with history of macrophage activation syndrome when compared to controls (OR, 14.5),” Dr. Schulert and associates wrote.
Genetics
Both the Cincinnati and Stanford teams conducted genetic analyses on some of their patients.
Among eight lung biopsy samples, Dr. Schulert found 37 differentially expressed genes: 36 with increased expression and 1 with decreased expression. Many of the up-regulated genes are involved in interferon-gamma response. Two (CXCL10 and CXCL9) are interferon-induced chemokines associated with macrophage activation syndrome. The down-regulated gene, PADI4, modulates immune response in lupus, and has been associated with the risk of interstitial lung disease in RA.
Dr. Mellins and her team analyzed whole-exome sequencing data from 20 patients and found some rare protein-altering gene variants in genes related to pulmonary alveolar proteinosis, all of which were heterozygous and shared with a healthy parent. But none of them could be directly tied to the disorder.
Another genetic puzzle demands attention, she said. About 10% of the children had trisomy 21 – a stark contrast to the typical 0.2% prevalence among a control group of sJIA patients without any known lung disease in the Childhood Arthritis and Rheumatology Research Alliance Registry cohort, similar to the background population rate. There were suggestions of more aggressive lung disease in all six of these children. Four presented with hypoxia, and two showed advanced interstitial fibrosis. Children with trisomy 21 also seemed more susceptible to infections; 83% had a viral or fungal lung infection at diagnosis, compared with 29% of those without trisomy 21.
Prior exposure to cytokine inhibitors
Parenchymal lung disease and pulmonary hypertension complicating sJIA was first highlighted in a series of 25 cases reported by Kimura et al. in 2013. These authors raised the question of the possible relationship of this and the increasing use of anti–IL-1 and anti–IL-6 biologics in sJIA treatment.
Following this lead, Dr. Mellins started looking into this new clinical entity in 2015. By then, she was identifying some past cases by autopsy records and current cases by clinical presentation. She saw a dramatic shift over time. From 2002 to 2011, she identified four cases, half of which had been exposed to IL-1/IL-6 inhibitors. From 2012 to 2014, eight new cases came to light, and seven had been exposed to those drugs. The crescendo continued from 2015 to 2017. During those years, Dr. Mellins and associates identified 10 new patients, 7 of whom had taken interleukin-inhibiting biologics. The mean time from initial drug exposure to diagnosis was a little more than 1 year.
An adjusted analysis comparing sJIA-LD patients and sJIA patients without lung disease didn’t find any significant difference in drug exposure. However, children with lung disease were more likely to have taken anakinra before the symptoms developed. Additionally, the symptoms of clubbing, abdominal pain, eosinophilia, hyperenhancing lymph nodes, and pulmonary alveolar proteinosis were much more common in children who’d taken the drugs.
The authors pointed out that this association does not prove causality and is confounded by the concomitant reduction in glucocorticoids with IL-1/IL-6 inhibitor use. And the vast majority of children with sJIA take cytokine inhibitors with no problems.
“Possibly, drug exposure may promote lung disease in a subset of children with sJIA, among the substantially larger group of patients who derive striking benefit from these drugs,” Dr. Mellins said, “Importantly, our results argue strongly for more investigation into a possible connection.”
Survival
After a mean follow-up of 1.7 years, the Stanford group saw high mortality. The 5-year survival rate translated to a mortality incidence of 159 deaths per 1,000 person-years, compared with 3.9 per 1,000 person-years in a historical cohort of sJIA patients who required biologic therapy.
Diffuse lung disease was the cause of 12 deaths; 5 of these patients also had macrophage activation syndrome at the time of death. Factors significantly associated with shortened survival included male sex, hypoxia at presentation, and neutrophilic bronchoalveolar lavage with more than 10 times the normal count. In an adjusted analysis, all of these variables fell out. However, none of the children with excessively high neutrophilic bronchoalveolar lavage survived.
Does it affect adults?
Could adults be experiencing the same disorder? There is some evidence to support it: The Food and Drug Administration adverse event website shows alveolar disease or pulmonary hypertension in 39 adults who have been exposed to IL-1 or IL-6 inhibition. Of these, 23 had RA, 11 adult-onset Still’s disease, and 5 unclassified rheumatic disorders.
The research groups were supported by grants from the sJIA Foundation, the Lucile Packard Foundation for Children’s Health, Stanford graduate fellowships, the Life Sciences Research Foundation, the Bill & Melinda Gates Foundation, Cincinnati Children’s Research Foundation, the Childhood Arthritis and Rheumatology Research Alliance, the Arthritis Foundation, and the National Institutes of Health. Many authors on both papers reported financial ties to Genentech, which markets tocilizumab, and other pharmaceutical companies*. Dr. Nigrovic reported receiving consulting fees and research support from Novartis and other companies.
SOURCES: Saper V et al. Ann Rheum Dis. 2019 Sep 27. doi: 10.1136/annrheumdis-2019-216040; Schulert GS et al. Arthritis Rheumatol. 2019 Aug 5. doi: 10.1002/art.41073; Nigrovic PA. Arthritis Rheumatol. 2019 Aug 7. doi: 10.1002/art.41071.
*Correction, 10/12/19: An earlier version of this article misstated the manufacturer of Actemra (tocilizumab).
This article was updated 10/14/19.
An uncommon but potentially deadly inflammatory lung disease is emerging among children with systemic juvenile idiopathic arthritis, and its history appears to coincide with the rise of powerful biologics as first-line therapy for children with the disease.
Most confirmed cases of systemic juvenile idiopathic arthritis with lung disease (sJIA-LD) are in the United States. But it’s popping up in other places that have adopted early biologic treatment for sJIA – including Canada, South America, Europe, and the Middle East.
The respiratory symptoms are relatively subtle, so by the time of lung disease detection, the amount of affected lung can be extensive, said Elizabeth Mellins, MD, a Stanford (Calif.) University researcher who, along with first author Vivian Saper, MD, recently published the largest case series comprising reports from 37 institutions (Ann Rheum Dis. 2019 Sep 27. doi: 10.1136/annrheumdis-2019-216040). By the end of follow-up, 22 of the 61 children in her cohort had died, including all 12 patients who demonstrated excessively high neutrophil levels in bronchoalveolar lavage samples.
Another recent report, authored by Grant Schulert, MD, PhD, and colleagues of the Cincinnati Children’s Hospital Medical Center, described 18 patients, 9 of whom were also included in the Stanford cohort (Arthritis Rheumatol. 2019 Aug 5. doi: 10.1002/art.41073).
Both investigators have now identified new patients.
“We are aware of 60 additional cases beyond what were included in our series,” Dr. Mellins said in an interview, bringing her entire cohort to 121. Dr. Schulert also continues to expand his group, detailing nine new cases at a recent private meeting.
“We are up to 27 now,” he said. “The features of these new patients are all very similar: The children are very young, all have had macrophage activation syndrome in the past and very-difficult-to-control JIA. Reactions to tocilizumab [Actemra] were also not uncommon in this group.”
Dr. Mellins also saw this association with allergic-type tocilizumab reactions, severe delayed hypersensitivity reactions to anakinra (Kineret) or canakinumab (Ilaris). Although serious lung disease in sJIA patients is not unheard of, this phenotype was virtually unknown until about a decade ago. Both investigators said that it’s been rising steadily since 2010 – just about the time that powerful cytokine-inhibiting biologics were changing these patients’ world for the better. After decades of relying almost solely on steroids and methotrexate, with rather poor results and significant long-term side effects, children were not only improving, but thriving. Gone was the life-changing glucocorticoid-related growth inhibition. Biologics could halt fevers, rash, and joint destruction in their tracks.
“For the first time in history, these kids could look forward to a more or less normal life,” Dr. Schulert said.
But the emergence of this particular type of lung disease could throw a pall over that success story, he said. If sJIA-LD is temporally associated with increasing reliance on long-term interleukin-1/IL-6 inhibition in children with early-onset disease, could these drugs actually be the causative agent? The picture remains unclear.
Some of the 18 in his initial series have improved, while 36% of those in the Stanford series died. Most who do recover stay on their IL-1 or IL-6 blocking therapy with good disease control without further lung problems. Both investigators found compelling genetic hints, but nothing conclusive. Children with trisomy 21 appear especially vulnerable. Most patients are very young – around 2 years old – but others are school aged. Some had a history of macrophage activation syndrome. Some had hard-to-control disease and some were clinically well controlled when the lung disease presented.
There are simply no answers yet.
With so many potential links, all unproven, clinicians may question the wisdom of embarking on long-term biologic therapy for their children with sJIA. Peter Nigrovic, MD, of Boston Children’s Hospital, addressed this in an accompanying editorial (Arthritis Rheumatol. 2019 Aug 7. doi: 10.1002/art.41071).
“My take on this is that it’s a very worrisome trend,” he said in an interview. “We’ve been going full bore toward early biologic therapy in sJIA and at the same time we are seeing more of this lung disease. Is it guilt by association? Or is there something more? The challenge for us is not to jump too soon to that conclusion.”
Although the association is there, he said, association does not equal causation. And there’s no doubt that biologics have vastly improved the lives of sJIA patients. “The drugs might be causal, and I worry about that and think we need to study it. But we absolutely need stronger evidence before we change practice.”
“This is a new manifestation of the disease, and it’s coming at the same time we are changing the treatment paradigm,” Dr. Nigrovic continued. “It could be because of interleukin-1 or interleukin-6 blockade. There is biological plausibility for such a link. It could also be related to the fact that we are using less steroids and methotrexate, which might have been preventing this. The appearance of sJIA lung disease could also be that a distinct secular trend unrelated to treatment, just as we saw amyloid come and go in this population in Europe. These other therapies were actually preventing this. We just don’t know.”
Clinical characteristics
Children presented with similar symptoms. Respiratory symptoms are usually subtle and mild. These can include tachypnea, hypoxia (43% in the Stanford series), and pulmonary hypertension (30% in the Stanford series).
Digital clubbing, often with erythema, was a common finding. Some children showed pruritic, nonevanescent rashes. Eosinophilia occurred in 37% of the Stanford series and severe abdominal pain in 16%, although Dr. Mellins noted that belly pain may be underestimated, as it was only volunteered, not queried, information.
“There are some red flags that should raise suspicion even without obvious respiratory symptoms,” Dr. Mellins said. These include lymphopenia, unexplained abdominal pain, eosinophilia, an unusual rash, and finger clubbing with or without erythema.
Findings on imaging were consistent in both series. Several key clinic features emerged: pleural thickening, septal thickening, bronchial wall or peribronchovascular thickening, “tree-in-bud” opacities, “ground-glass” opacities, peripheral consolidation, and lymphadenopathy.
“The imaging findings correspond to two things,” Dr. Schulert said. “The first is inflammation in the interstitium, which is evidence of chronic and ongoing inflammation. The other thing is that the alveoli are filled with a lipoproteinaceous material which is actually surfactant that’s not being normally recycled by the lung macrophages. You can see these features in other conditions where there’s a problem with lung macrophages, like pulmonary alveolar proteinosis, genetic and autoimmune disorders, infections, or inhalants.”
Pathology showed alveolar filling – a location in the lung that hides usual symptoms until the lung disease is advanced. Prior drug reactions were common. Tocilizumab anaphylaxis occurred in close to 40% of the Stanford series – a surprising finding given the 0.6% reaction incidence in the drug’s sJIA trials. Dr. Schulert saw a similar story.
“In our cohort we also observed a striking number of adverse events to cytokine-targeted biologics exposure,” Dr. Schulert said. “Most of these reactions were to tocilizumab, and were described variously from pain and feeling unwell, to difficulty breathing, to anaphylaxis.”
In a risk analysis, Dr. Schulert determined that adverse events to cytokine-targeting biologics increased the likelihood of lung disease more than 13 times (odds ratio, 13.6).
“We also identified a statistically significant association with history of macrophage activation syndrome when compared to controls (OR, 14.5),” Dr. Schulert and associates wrote.
Genetics
Both the Cincinnati and Stanford teams conducted genetic analyses on some of their patients.
Among eight lung biopsy samples, Dr. Schulert found 37 differentially expressed genes: 36 with increased expression and 1 with decreased expression. Many of the up-regulated genes are involved in interferon-gamma response. Two (CXCL10 and CXCL9) are interferon-induced chemokines associated with macrophage activation syndrome. The down-regulated gene, PADI4, modulates immune response in lupus, and has been associated with the risk of interstitial lung disease in RA.
Dr. Mellins and her team analyzed whole-exome sequencing data from 20 patients and found some rare protein-altering gene variants in genes related to pulmonary alveolar proteinosis, all of which were heterozygous and shared with a healthy parent. But none of them could be directly tied to the disorder.
Another genetic puzzle demands attention, she said. About 10% of the children had trisomy 21 – a stark contrast to the typical 0.2% prevalence among a control group of sJIA patients without any known lung disease in the Childhood Arthritis and Rheumatology Research Alliance Registry cohort, similar to the background population rate. There were suggestions of more aggressive lung disease in all six of these children. Four presented with hypoxia, and two showed advanced interstitial fibrosis. Children with trisomy 21 also seemed more susceptible to infections; 83% had a viral or fungal lung infection at diagnosis, compared with 29% of those without trisomy 21.
Prior exposure to cytokine inhibitors
Parenchymal lung disease and pulmonary hypertension complicating sJIA was first highlighted in a series of 25 cases reported by Kimura et al. in 2013. These authors raised the question of the possible relationship of this and the increasing use of anti–IL-1 and anti–IL-6 biologics in sJIA treatment.
Following this lead, Dr. Mellins started looking into this new clinical entity in 2015. By then, she was identifying some past cases by autopsy records and current cases by clinical presentation. She saw a dramatic shift over time. From 2002 to 2011, she identified four cases, half of which had been exposed to IL-1/IL-6 inhibitors. From 2012 to 2014, eight new cases came to light, and seven had been exposed to those drugs. The crescendo continued from 2015 to 2017. During those years, Dr. Mellins and associates identified 10 new patients, 7 of whom had taken interleukin-inhibiting biologics. The mean time from initial drug exposure to diagnosis was a little more than 1 year.
An adjusted analysis comparing sJIA-LD patients and sJIA patients without lung disease didn’t find any significant difference in drug exposure. However, children with lung disease were more likely to have taken anakinra before the symptoms developed. Additionally, the symptoms of clubbing, abdominal pain, eosinophilia, hyperenhancing lymph nodes, and pulmonary alveolar proteinosis were much more common in children who’d taken the drugs.
The authors pointed out that this association does not prove causality and is confounded by the concomitant reduction in glucocorticoids with IL-1/IL-6 inhibitor use. And the vast majority of children with sJIA take cytokine inhibitors with no problems.
“Possibly, drug exposure may promote lung disease in a subset of children with sJIA, among the substantially larger group of patients who derive striking benefit from these drugs,” Dr. Mellins said, “Importantly, our results argue strongly for more investigation into a possible connection.”
Survival
After a mean follow-up of 1.7 years, the Stanford group saw high mortality. The 5-year survival rate translated to a mortality incidence of 159 deaths per 1,000 person-years, compared with 3.9 per 1,000 person-years in a historical cohort of sJIA patients who required biologic therapy.
Diffuse lung disease was the cause of 12 deaths; 5 of these patients also had macrophage activation syndrome at the time of death. Factors significantly associated with shortened survival included male sex, hypoxia at presentation, and neutrophilic bronchoalveolar lavage with more than 10 times the normal count. In an adjusted analysis, all of these variables fell out. However, none of the children with excessively high neutrophilic bronchoalveolar lavage survived.
Does it affect adults?
Could adults be experiencing the same disorder? There is some evidence to support it: The Food and Drug Administration adverse event website shows alveolar disease or pulmonary hypertension in 39 adults who have been exposed to IL-1 or IL-6 inhibition. Of these, 23 had RA, 11 adult-onset Still’s disease, and 5 unclassified rheumatic disorders.
The research groups were supported by grants from the sJIA Foundation, the Lucile Packard Foundation for Children’s Health, Stanford graduate fellowships, the Life Sciences Research Foundation, the Bill & Melinda Gates Foundation, Cincinnati Children’s Research Foundation, the Childhood Arthritis and Rheumatology Research Alliance, the Arthritis Foundation, and the National Institutes of Health. Many authors on both papers reported financial ties to Genentech, which markets tocilizumab, and other pharmaceutical companies*. Dr. Nigrovic reported receiving consulting fees and research support from Novartis and other companies.
SOURCES: Saper V et al. Ann Rheum Dis. 2019 Sep 27. doi: 10.1136/annrheumdis-2019-216040; Schulert GS et al. Arthritis Rheumatol. 2019 Aug 5. doi: 10.1002/art.41073; Nigrovic PA. Arthritis Rheumatol. 2019 Aug 7. doi: 10.1002/art.41071.
*Correction, 10/12/19: An earlier version of this article misstated the manufacturer of Actemra (tocilizumab).
This article was updated 10/14/19.
An uncommon but potentially deadly inflammatory lung disease is emerging among children with systemic juvenile idiopathic arthritis, and its history appears to coincide with the rise of powerful biologics as first-line therapy for children with the disease.
Most confirmed cases of systemic juvenile idiopathic arthritis with lung disease (sJIA-LD) are in the United States. But it’s popping up in other places that have adopted early biologic treatment for sJIA – including Canada, South America, Europe, and the Middle East.
The respiratory symptoms are relatively subtle, so by the time of lung disease detection, the amount of affected lung can be extensive, said Elizabeth Mellins, MD, a Stanford (Calif.) University researcher who, along with first author Vivian Saper, MD, recently published the largest case series comprising reports from 37 institutions (Ann Rheum Dis. 2019 Sep 27. doi: 10.1136/annrheumdis-2019-216040). By the end of follow-up, 22 of the 61 children in her cohort had died, including all 12 patients who demonstrated excessively high neutrophil levels in bronchoalveolar lavage samples.
Another recent report, authored by Grant Schulert, MD, PhD, and colleagues of the Cincinnati Children’s Hospital Medical Center, described 18 patients, 9 of whom were also included in the Stanford cohort (Arthritis Rheumatol. 2019 Aug 5. doi: 10.1002/art.41073).
Both investigators have now identified new patients.
“We are aware of 60 additional cases beyond what were included in our series,” Dr. Mellins said in an interview, bringing her entire cohort to 121. Dr. Schulert also continues to expand his group, detailing nine new cases at a recent private meeting.
“We are up to 27 now,” he said. “The features of these new patients are all very similar: The children are very young, all have had macrophage activation syndrome in the past and very-difficult-to-control JIA. Reactions to tocilizumab [Actemra] were also not uncommon in this group.”
Dr. Mellins also saw this association with allergic-type tocilizumab reactions, severe delayed hypersensitivity reactions to anakinra (Kineret) or canakinumab (Ilaris). Although serious lung disease in sJIA patients is not unheard of, this phenotype was virtually unknown until about a decade ago. Both investigators said that it’s been rising steadily since 2010 – just about the time that powerful cytokine-inhibiting biologics were changing these patients’ world for the better. After decades of relying almost solely on steroids and methotrexate, with rather poor results and significant long-term side effects, children were not only improving, but thriving. Gone was the life-changing glucocorticoid-related growth inhibition. Biologics could halt fevers, rash, and joint destruction in their tracks.
“For the first time in history, these kids could look forward to a more or less normal life,” Dr. Schulert said.
But the emergence of this particular type of lung disease could throw a pall over that success story, he said. If sJIA-LD is temporally associated with increasing reliance on long-term interleukin-1/IL-6 inhibition in children with early-onset disease, could these drugs actually be the causative agent? The picture remains unclear.
Some of the 18 in his initial series have improved, while 36% of those in the Stanford series died. Most who do recover stay on their IL-1 or IL-6 blocking therapy with good disease control without further lung problems. Both investigators found compelling genetic hints, but nothing conclusive. Children with trisomy 21 appear especially vulnerable. Most patients are very young – around 2 years old – but others are school aged. Some had a history of macrophage activation syndrome. Some had hard-to-control disease and some were clinically well controlled when the lung disease presented.
There are simply no answers yet.
With so many potential links, all unproven, clinicians may question the wisdom of embarking on long-term biologic therapy for their children with sJIA. Peter Nigrovic, MD, of Boston Children’s Hospital, addressed this in an accompanying editorial (Arthritis Rheumatol. 2019 Aug 7. doi: 10.1002/art.41071).
“My take on this is that it’s a very worrisome trend,” he said in an interview. “We’ve been going full bore toward early biologic therapy in sJIA and at the same time we are seeing more of this lung disease. Is it guilt by association? Or is there something more? The challenge for us is not to jump too soon to that conclusion.”
Although the association is there, he said, association does not equal causation. And there’s no doubt that biologics have vastly improved the lives of sJIA patients. “The drugs might be causal, and I worry about that and think we need to study it. But we absolutely need stronger evidence before we change practice.”
“This is a new manifestation of the disease, and it’s coming at the same time we are changing the treatment paradigm,” Dr. Nigrovic continued. “It could be because of interleukin-1 or interleukin-6 blockade. There is biological plausibility for such a link. It could also be related to the fact that we are using less steroids and methotrexate, which might have been preventing this. The appearance of sJIA lung disease could also be that a distinct secular trend unrelated to treatment, just as we saw amyloid come and go in this population in Europe. These other therapies were actually preventing this. We just don’t know.”
Clinical characteristics
Children presented with similar symptoms. Respiratory symptoms are usually subtle and mild. These can include tachypnea, hypoxia (43% in the Stanford series), and pulmonary hypertension (30% in the Stanford series).
Digital clubbing, often with erythema, was a common finding. Some children showed pruritic, nonevanescent rashes. Eosinophilia occurred in 37% of the Stanford series and severe abdominal pain in 16%, although Dr. Mellins noted that belly pain may be underestimated, as it was only volunteered, not queried, information.
“There are some red flags that should raise suspicion even without obvious respiratory symptoms,” Dr. Mellins said. These include lymphopenia, unexplained abdominal pain, eosinophilia, an unusual rash, and finger clubbing with or without erythema.
Findings on imaging were consistent in both series. Several key clinic features emerged: pleural thickening, septal thickening, bronchial wall or peribronchovascular thickening, “tree-in-bud” opacities, “ground-glass” opacities, peripheral consolidation, and lymphadenopathy.
“The imaging findings correspond to two things,” Dr. Schulert said. “The first is inflammation in the interstitium, which is evidence of chronic and ongoing inflammation. The other thing is that the alveoli are filled with a lipoproteinaceous material which is actually surfactant that’s not being normally recycled by the lung macrophages. You can see these features in other conditions where there’s a problem with lung macrophages, like pulmonary alveolar proteinosis, genetic and autoimmune disorders, infections, or inhalants.”
Pathology showed alveolar filling – a location in the lung that hides usual symptoms until the lung disease is advanced. Prior drug reactions were common. Tocilizumab anaphylaxis occurred in close to 40% of the Stanford series – a surprising finding given the 0.6% reaction incidence in the drug’s sJIA trials. Dr. Schulert saw a similar story.
“In our cohort we also observed a striking number of adverse events to cytokine-targeted biologics exposure,” Dr. Schulert said. “Most of these reactions were to tocilizumab, and were described variously from pain and feeling unwell, to difficulty breathing, to anaphylaxis.”
In a risk analysis, Dr. Schulert determined that adverse events to cytokine-targeting biologics increased the likelihood of lung disease more than 13 times (odds ratio, 13.6).
“We also identified a statistically significant association with history of macrophage activation syndrome when compared to controls (OR, 14.5),” Dr. Schulert and associates wrote.
Genetics
Both the Cincinnati and Stanford teams conducted genetic analyses on some of their patients.
Among eight lung biopsy samples, Dr. Schulert found 37 differentially expressed genes: 36 with increased expression and 1 with decreased expression. Many of the up-regulated genes are involved in interferon-gamma response. Two (CXCL10 and CXCL9) are interferon-induced chemokines associated with macrophage activation syndrome. The down-regulated gene, PADI4, modulates immune response in lupus, and has been associated with the risk of interstitial lung disease in RA.
Dr. Mellins and her team analyzed whole-exome sequencing data from 20 patients and found some rare protein-altering gene variants in genes related to pulmonary alveolar proteinosis, all of which were heterozygous and shared with a healthy parent. But none of them could be directly tied to the disorder.
Another genetic puzzle demands attention, she said. About 10% of the children had trisomy 21 – a stark contrast to the typical 0.2% prevalence among a control group of sJIA patients without any known lung disease in the Childhood Arthritis and Rheumatology Research Alliance Registry cohort, similar to the background population rate. There were suggestions of more aggressive lung disease in all six of these children. Four presented with hypoxia, and two showed advanced interstitial fibrosis. Children with trisomy 21 also seemed more susceptible to infections; 83% had a viral or fungal lung infection at diagnosis, compared with 29% of those without trisomy 21.
Prior exposure to cytokine inhibitors
Parenchymal lung disease and pulmonary hypertension complicating sJIA was first highlighted in a series of 25 cases reported by Kimura et al. in 2013. These authors raised the question of the possible relationship of this and the increasing use of anti–IL-1 and anti–IL-6 biologics in sJIA treatment.
Following this lead, Dr. Mellins started looking into this new clinical entity in 2015. By then, she was identifying some past cases by autopsy records and current cases by clinical presentation. She saw a dramatic shift over time. From 2002 to 2011, she identified four cases, half of which had been exposed to IL-1/IL-6 inhibitors. From 2012 to 2014, eight new cases came to light, and seven had been exposed to those drugs. The crescendo continued from 2015 to 2017. During those years, Dr. Mellins and associates identified 10 new patients, 7 of whom had taken interleukin-inhibiting biologics. The mean time from initial drug exposure to diagnosis was a little more than 1 year.
An adjusted analysis comparing sJIA-LD patients and sJIA patients without lung disease didn’t find any significant difference in drug exposure. However, children with lung disease were more likely to have taken anakinra before the symptoms developed. Additionally, the symptoms of clubbing, abdominal pain, eosinophilia, hyperenhancing lymph nodes, and pulmonary alveolar proteinosis were much more common in children who’d taken the drugs.
The authors pointed out that this association does not prove causality and is confounded by the concomitant reduction in glucocorticoids with IL-1/IL-6 inhibitor use. And the vast majority of children with sJIA take cytokine inhibitors with no problems.
“Possibly, drug exposure may promote lung disease in a subset of children with sJIA, among the substantially larger group of patients who derive striking benefit from these drugs,” Dr. Mellins said, “Importantly, our results argue strongly for more investigation into a possible connection.”
Survival
After a mean follow-up of 1.7 years, the Stanford group saw high mortality. The 5-year survival rate translated to a mortality incidence of 159 deaths per 1,000 person-years, compared with 3.9 per 1,000 person-years in a historical cohort of sJIA patients who required biologic therapy.
Diffuse lung disease was the cause of 12 deaths; 5 of these patients also had macrophage activation syndrome at the time of death. Factors significantly associated with shortened survival included male sex, hypoxia at presentation, and neutrophilic bronchoalveolar lavage with more than 10 times the normal count. In an adjusted analysis, all of these variables fell out. However, none of the children with excessively high neutrophilic bronchoalveolar lavage survived.
Does it affect adults?
Could adults be experiencing the same disorder? There is some evidence to support it: The Food and Drug Administration adverse event website shows alveolar disease or pulmonary hypertension in 39 adults who have been exposed to IL-1 or IL-6 inhibition. Of these, 23 had RA, 11 adult-onset Still’s disease, and 5 unclassified rheumatic disorders.
The research groups were supported by grants from the sJIA Foundation, the Lucile Packard Foundation for Children’s Health, Stanford graduate fellowships, the Life Sciences Research Foundation, the Bill & Melinda Gates Foundation, Cincinnati Children’s Research Foundation, the Childhood Arthritis and Rheumatology Research Alliance, the Arthritis Foundation, and the National Institutes of Health. Many authors on both papers reported financial ties to Genentech, which markets tocilizumab, and other pharmaceutical companies*. Dr. Nigrovic reported receiving consulting fees and research support from Novartis and other companies.
SOURCES: Saper V et al. Ann Rheum Dis. 2019 Sep 27. doi: 10.1136/annrheumdis-2019-216040; Schulert GS et al. Arthritis Rheumatol. 2019 Aug 5. doi: 10.1002/art.41073; Nigrovic PA. Arthritis Rheumatol. 2019 Aug 7. doi: 10.1002/art.41071.
*Correction, 10/12/19: An earlier version of this article misstated the manufacturer of Actemra (tocilizumab).
This article was updated 10/14/19.
