Background: Early antibiotic administration reduces mortality in patients with severe sepsis. Administering antibiotics before blood cultures could potentially decrease time to treatment and improve outcomes, but the diagnostic yield of blood cultures drawn shortly after antibiotics is unknown.

There were several study limitations including: lack of sequential recruitment, lower than expected proportion of bacteremic patients, and variation in blood culture collection. Despite this, the magnitude of the findings are convincing and support current practice.

Bottom line: Continue to obtain blood cultures before antibiotics.

Citation: Cheng MP et al. Blood culture results before and after antimicrobial administration in patients with severe manifestations of sepsis. Ann Intern Med. 2019 Oct 15;171(8):547-54.

Dr. Waner is clinical instructor of medicine, hospital medicine, at the Rocky Mountain Veterans Affairs Regional Medical Center, Aurora, Colo.

Background: Early antibiotic administration reduces mortality in patients with severe sepsis. Administering antibiotics before blood cultures could potentially decrease time to treatment and improve outcomes, but the diagnostic yield of blood cultures drawn shortly after antibiotics is unknown.

There were several study limitations including: lack of sequential recruitment, lower than expected proportion of bacteremic patients, and variation in blood culture collection. Despite this, the magnitude of the findings are convincing and support current practice.

Bottom line: Continue to obtain blood cultures before antibiotics.

Citation: Cheng MP et al. Blood culture results before and after antimicrobial administration in patients with severe manifestations of sepsis. Ann Intern Med. 2019 Oct 15;171(8):547-54.

Dr. Waner is clinical instructor of medicine, hospital medicine, at the Rocky Mountain Veterans Affairs Regional Medical Center, Aurora, Colo.

Background: Early antibiotic administration reduces mortality in patients with severe sepsis. Administering antibiotics before blood cultures could potentially decrease time to treatment and improve outcomes, but the diagnostic yield of blood cultures drawn shortly after antibiotics is unknown.

There were several study limitations including: lack of sequential recruitment, lower than expected proportion of bacteremic patients, and variation in blood culture collection. Despite this, the magnitude of the findings are convincing and support current practice.

Bottom line: Continue to obtain blood cultures before antibiotics.

Citation: Cheng MP et al. Blood culture results before and after antimicrobial administration in patients with severe manifestations of sepsis. Ann Intern Med. 2019 Oct 15;171(8):547-54.

Dr. Waner is clinical instructor of medicine, hospital medicine, at the Rocky Mountain Veterans Affairs Regional Medical Center, Aurora, Colo.

Offering a combination of colonoscopy and fecal immunochemical testing (FIT), either in sequence or by choice, may significantly increase participation in colorectal cancer (CRC) screening, according to a prospective study involving more than 12,000 individuals in Poland.

Still, greater participation did not lead to significantly higher rates of advanced disease detection, reported lead author Nastazja Dagny Pilonis, MD, of the Maria Sklodowska-Curie National Research Institute of Oncology, Warsaw, and colleagues in Gastroenterology.

According to the investigators, screening programs that offer colonoscopy and FIT are more effective than those that offer colonoscopy alone, but an optimal combination protocol has yet to be established, and some parts of the world still rely upon a single diagnostic method.

“In Europe, CRC screening programs often implement only one screening modality: colonoscopy, sigmoidoscopy, or stool testing, depending on the health care provider,” the investigators wrote in Gastroenterology. They noted, however, that national guidelines in the United States recommend strategies that include more than one screening method. “‘One-size-fits-all’ approaches to CRC screening do not result in satisfactory participation” because of behavioral, cultural, and socioeconomic variation among individuals.

To improve understanding of the best ways to improve participation, the investigators conducted a prospective randomized trial, PICCOLINO, via the Polish Colonoscopy Screening Program. In total, 12,485 eligible individuals aged between 55 and 64 years received postal invitations to participate in CRC screening. Individuals were randomized in a 1:1:1 ratio into one of three mailing protocols, each of which involved an initial invitation, and, if needed, a second invitation that offered the following:

• Control group: colonoscopy, with nonresponders receiving the same invitation again
• Sequential group: colonoscopy, with nonresponders or refusers receiving a second invitation that offered FIT
• Choice group: choice between colonoscopy or FIT, with nonresponders receiving the same invitation again

The primary outcome was participation in screening within 18 weeks of enrollment. The secondary outcome was diagnostic yield for either advanced adenoma or CRC.

Out of the three groups, the control group had the lowest participation rate, at 17.5%, compared with 25.8% for the sequential group and 26.5% for the choice group. Multivariable logistic regression showed that individuals in the sequential and choice groups had 64% and 70% higher rates of participation, respectively. Across all groups, age of 60 years or older predicted 12% higher likelihood of participation; in contrast, location more than 40 kilometers from a testing center was associated with an 18% decrease in participation, compared with individuals who lived less than 20 kilometers away.

While the control and sequential groups had similar rates of colonoscopy participation, at 17.5% and 15.9%, respectively (P = .788), this rate was significantly lower, at 8.5%, in the choice group (P = .001). Conversely, the sequential group had a significantly lower rate of FITs than the choice group, at 9.9% versus 17.9%, respectively (P = .001). Among participants with a positive FIT, diagnostic work-up colonoscopies were performed in 70.0% of those in the sequential group and 73.3% in the choice group, “despite active call-recall efforts.”

Across all invited individuals, advanced disease detection rates were similar across groups, at 1.1% for both the control and the sequential group and 1.2% for the choice group. Among those who were actually screened, the control group had a slightly higher diagnostic yield for advanced neoplasia, at 6.5%, compared with 4.2% in the sequential group and 4.4% in the choice group; however, these differences were not statistically significant. In contrast, significantly more adenomas of any kind were detected in the control and sequential groups (5.6% for both) than the choice group (3.9%) (P < .001).

“Although the strategies which included FIT showed higher participation rates than the strategy of offering colonoscopy alone, these strategies did not result in increased detection rates of advanced neoplasia in the intention to screen analysis,” the investigators wrote. “An absolute increase in participation rates of 8%-10% seems insufficient to translate into higher advanced neoplasia detection at the population level.”

Dr. Pilonis and colleagues also suggested that the relatively low rate of diagnostic colonoscopy after positive FIT contributed to the suboptimal diagnostic yield.

“These rates are unsatisfactory taking into account significant call-recall efforts, but are within the range reported in other studies,” they wrote.

They also wrote that their study compared participation and detection between one-time colonoscopy and one-time screening strategies combining colonoscopy and FIT. In acknowledging this, they noted that these approaches have different screening intervals and uptake over time: “FIT has been shown to achieve higher participation rates than colonoscopy for one time screening, but its uptake over several rounds may not be superior to one time colonoscopy.” Furthermore, detection rates of the sequential or choice strategies for advanced disease may rise over time with further implementation, so the one-time screening may not be sufficient to reveal what could become significant differences.

The study was funded by the Polish Ministry of Health, the Polish Foundation of Gastroenterology, and the Centre of Postgraduate Medical Education in Warsaw. FITs, materials, and reagents were provided by Eiken Chemical. The investigators disclosed relationships with Boston Scientific, AbbVie, Olympus, and others.

Offering a combination of colonoscopy and fecal immunochemical testing (FIT), either in sequence or by choice, may significantly increase participation in colorectal cancer (CRC) screening, according to a prospective study involving more than 12,000 individuals in Poland.

Still, greater participation did not lead to significantly higher rates of advanced disease detection, reported lead author Nastazja Dagny Pilonis, MD, of the Maria Sklodowska-Curie National Research Institute of Oncology, Warsaw, and colleagues in Gastroenterology.

According to the investigators, screening programs that offer colonoscopy and FIT are more effective than those that offer colonoscopy alone, but an optimal combination protocol has yet to be established, and some parts of the world still rely upon a single diagnostic method.

“In Europe, CRC screening programs often implement only one screening modality: colonoscopy, sigmoidoscopy, or stool testing, depending on the health care provider,” the investigators wrote in Gastroenterology. They noted, however, that national guidelines in the United States recommend strategies that include more than one screening method. “‘One-size-fits-all’ approaches to CRC screening do not result in satisfactory participation” because of behavioral, cultural, and socioeconomic variation among individuals.

To improve understanding of the best ways to improve participation, the investigators conducted a prospective randomized trial, PICCOLINO, via the Polish Colonoscopy Screening Program. In total, 12,485 eligible individuals aged between 55 and 64 years received postal invitations to participate in CRC screening. Individuals were randomized in a 1:1:1 ratio into one of three mailing protocols, each of which involved an initial invitation, and, if needed, a second invitation that offered the following:

• Control group: colonoscopy, with nonresponders receiving the same invitation again
• Sequential group: colonoscopy, with nonresponders or refusers receiving a second invitation that offered FIT
• Choice group: choice between colonoscopy or FIT, with nonresponders receiving the same invitation again

The primary outcome was participation in screening within 18 weeks of enrollment. The secondary outcome was diagnostic yield for either advanced adenoma or CRC.

Out of the three groups, the control group had the lowest participation rate, at 17.5%, compared with 25.8% for the sequential group and 26.5% for the choice group. Multivariable logistic regression showed that individuals in the sequential and choice groups had 64% and 70% higher rates of participation, respectively. Across all groups, age of 60 years or older predicted 12% higher likelihood of participation; in contrast, location more than 40 kilometers from a testing center was associated with an 18% decrease in participation, compared with individuals who lived less than 20 kilometers away.

While the control and sequential groups had similar rates of colonoscopy participation, at 17.5% and 15.9%, respectively (P = .788), this rate was significantly lower, at 8.5%, in the choice group (P = .001). Conversely, the sequential group had a significantly lower rate of FITs than the choice group, at 9.9% versus 17.9%, respectively (P = .001). Among participants with a positive FIT, diagnostic work-up colonoscopies were performed in 70.0% of those in the sequential group and 73.3% in the choice group, “despite active call-recall efforts.”

Across all invited individuals, advanced disease detection rates were similar across groups, at 1.1% for both the control and the sequential group and 1.2% for the choice group. Among those who were actually screened, the control group had a slightly higher diagnostic yield for advanced neoplasia, at 6.5%, compared with 4.2% in the sequential group and 4.4% in the choice group; however, these differences were not statistically significant. In contrast, significantly more adenomas of any kind were detected in the control and sequential groups (5.6% for both) than the choice group (3.9%) (P < .001).

“Although the strategies which included FIT showed higher participation rates than the strategy of offering colonoscopy alone, these strategies did not result in increased detection rates of advanced neoplasia in the intention to screen analysis,” the investigators wrote. “An absolute increase in participation rates of 8%-10% seems insufficient to translate into higher advanced neoplasia detection at the population level.”

Dr. Pilonis and colleagues also suggested that the relatively low rate of diagnostic colonoscopy after positive FIT contributed to the suboptimal diagnostic yield.

“These rates are unsatisfactory taking into account significant call-recall efforts, but are within the range reported in other studies,” they wrote.

They also wrote that their study compared participation and detection between one-time colonoscopy and one-time screening strategies combining colonoscopy and FIT. In acknowledging this, they noted that these approaches have different screening intervals and uptake over time: “FIT has been shown to achieve higher participation rates than colonoscopy for one time screening, but its uptake over several rounds may not be superior to one time colonoscopy.” Furthermore, detection rates of the sequential or choice strategies for advanced disease may rise over time with further implementation, so the one-time screening may not be sufficient to reveal what could become significant differences.

The study was funded by the Polish Ministry of Health, the Polish Foundation of Gastroenterology, and the Centre of Postgraduate Medical Education in Warsaw. FITs, materials, and reagents were provided by Eiken Chemical. The investigators disclosed relationships with Boston Scientific, AbbVie, Olympus, and others.

Offering a combination of colonoscopy and fecal immunochemical testing (FIT), either in sequence or by choice, may significantly increase participation in colorectal cancer (CRC) screening, according to a prospective study involving more than 12,000 individuals in Poland.

Still, greater participation did not lead to significantly higher rates of advanced disease detection, reported lead author Nastazja Dagny Pilonis, MD, of the Maria Sklodowska-Curie National Research Institute of Oncology, Warsaw, and colleagues in Gastroenterology.

According to the investigators, screening programs that offer colonoscopy and FIT are more effective than those that offer colonoscopy alone, but an optimal combination protocol has yet to be established, and some parts of the world still rely upon a single diagnostic method.

“In Europe, CRC screening programs often implement only one screening modality: colonoscopy, sigmoidoscopy, or stool testing, depending on the health care provider,” the investigators wrote in Gastroenterology. They noted, however, that national guidelines in the United States recommend strategies that include more than one screening method. “‘One-size-fits-all’ approaches to CRC screening do not result in satisfactory participation” because of behavioral, cultural, and socioeconomic variation among individuals.

To improve understanding of the best ways to improve participation, the investigators conducted a prospective randomized trial, PICCOLINO, via the Polish Colonoscopy Screening Program. In total, 12,485 eligible individuals aged between 55 and 64 years received postal invitations to participate in CRC screening. Individuals were randomized in a 1:1:1 ratio into one of three mailing protocols, each of which involved an initial invitation, and, if needed, a second invitation that offered the following:

• Control group: colonoscopy, with nonresponders receiving the same invitation again
• Sequential group: colonoscopy, with nonresponders or refusers receiving a second invitation that offered FIT
• Choice group: choice between colonoscopy or FIT, with nonresponders receiving the same invitation again

The primary outcome was participation in screening within 18 weeks of enrollment. The secondary outcome was diagnostic yield for either advanced adenoma or CRC.

Out of the three groups, the control group had the lowest participation rate, at 17.5%, compared with 25.8% for the sequential group and 26.5% for the choice group. Multivariable logistic regression showed that individuals in the sequential and choice groups had 64% and 70% higher rates of participation, respectively. Across all groups, age of 60 years or older predicted 12% higher likelihood of participation; in contrast, location more than 40 kilometers from a testing center was associated with an 18% decrease in participation, compared with individuals who lived less than 20 kilometers away.

While the control and sequential groups had similar rates of colonoscopy participation, at 17.5% and 15.9%, respectively (P = .788), this rate was significantly lower, at 8.5%, in the choice group (P = .001). Conversely, the sequential group had a significantly lower rate of FITs than the choice group, at 9.9% versus 17.9%, respectively (P = .001). Among participants with a positive FIT, diagnostic work-up colonoscopies were performed in 70.0% of those in the sequential group and 73.3% in the choice group, “despite active call-recall efforts.”

Across all invited individuals, advanced disease detection rates were similar across groups, at 1.1% for both the control and the sequential group and 1.2% for the choice group. Among those who were actually screened, the control group had a slightly higher diagnostic yield for advanced neoplasia, at 6.5%, compared with 4.2% in the sequential group and 4.4% in the choice group; however, these differences were not statistically significant. In contrast, significantly more adenomas of any kind were detected in the control and sequential groups (5.6% for both) than the choice group (3.9%) (P < .001).

“Although the strategies which included FIT showed higher participation rates than the strategy of offering colonoscopy alone, these strategies did not result in increased detection rates of advanced neoplasia in the intention to screen analysis,” the investigators wrote. “An absolute increase in participation rates of 8%-10% seems insufficient to translate into higher advanced neoplasia detection at the population level.”

Dr. Pilonis and colleagues also suggested that the relatively low rate of diagnostic colonoscopy after positive FIT contributed to the suboptimal diagnostic yield.

“These rates are unsatisfactory taking into account significant call-recall efforts, but are within the range reported in other studies,” they wrote.

They also wrote that their study compared participation and detection between one-time colonoscopy and one-time screening strategies combining colonoscopy and FIT. In acknowledging this, they noted that these approaches have different screening intervals and uptake over time: “FIT has been shown to achieve higher participation rates than colonoscopy for one time screening, but its uptake over several rounds may not be superior to one time colonoscopy.” Furthermore, detection rates of the sequential or choice strategies for advanced disease may rise over time with further implementation, so the one-time screening may not be sufficient to reveal what could become significant differences.

The study was funded by the Polish Ministry of Health, the Polish Foundation of Gastroenterology, and the Centre of Postgraduate Medical Education in Warsaw. FITs, materials, and reagents were provided by Eiken Chemical. The investigators disclosed relationships with Boston Scientific, AbbVie, Olympus, and others.

People with rheumatic and musculoskeletal diseases (RMDs) who contract the SARS-CoV-2 virus appear more likely to die from COVID-19 if their rheumatologic condition is not being well controlled at the time of their infection.

New data from the COVID-19 Global Rheumatology Alliance (GRA) physician registry reported in Annals of the Rheumatic Diseases have found that the odds of dying from COVID-19 were 87% higher in individuals recorded as having moderate to high disease activity versus those reported to be in remission or having low disease activity.

“I think this really highlights the importance of continuing to appropriately, and actively, treat our patients, and the importance of controlling their disease,” Pedro Machado, MD, PhD, said in an interview. Dr. Machado, an associate professor in rheumatology and muscle diseases at University College London and a consultant rheumatologist at several U.K. hospitals, has been involved in the GRA physician registry from the start, and sits on the GRA steering committee.

Alongside higher disease activity, several other important factors were found to be associated with increased odds of dying from COVID-19 – older age, male gender, and the presence of one or more comorbidities, such as hypertension combined with cardiovascular disease or chronic lung disease.

These demographic and disease-based factors have been linked to an increased risk for COVID-19–related hospitalization before, both in people with RMDs and in the general population, but the latest GRA physician registry data now take that a step further, and link them also to an increased risk for death, together with several other factors more specific to RMDs.
 

Logging COVID-19 rheumatologic cases

Since the start of the global pandemic, the potential effects that SARS-CoV-2 infection might have on people with RMDs in particular has concerned the rheumatology community. The main worries being that, either because of the underlying RMD itself or to its treatment, there may be immunoregulatory deficits or other risk factors that would make individuals more susceptible to not only infection but also to developing more severe COVID-19 than the general population.

These concerns led to the rapid formation of the GRA and the COVID-19 GRA physician registry in March 2020 to collect and analyze data on adults with rheumatic disease and confirmed or presumptive COVID-19. Entries into the registry are made by or under the direction of rheumatologists, and this is a voluntary process.

“This population cannot ever be entirely representative of the population of patients with rheumatic diseases,” Dr. Machado acknowledged. There will be selection and other biases that affect the reported data. That said, it’s the largest database of reported COVID-19 cases in adult rheumatology patients across the world, with more than 9,000 cases so far included from multiple registries, including those based in Europe and North and South America. Data from one of these – the French RMD cohort – have also recently been published in Annals of the Rheumatic Diseases, showing much the same findings but on a national level.

Hospitalization was the focus of a previous report because “you need large sample sizes” to look at endpoints that occur less frequently. When the first analysis was done, there were around 600 cases from 40 countries in the registry with sufficient data that could be used. Now, with a greater number of recorded cases, factors influencing the risk for death could be examined.

Death rate and risk factors found

Data on 3,729 COVID-19 cases in people with RMDs were included in the current analysis, all recorded in the first few months of the registry being open and up until July 1, 2020. In all, 390 (10.5%) of people died. While this is “clearly higher” than reported in the general population in most countries, the analysis was not designed to calculate a precise estimate.

“It should not be taken as an estimate of the overall death rate among patients with rheumatic diseases and COVID-19,” Dr. Machado and coauthors have been keen to point out.

“Age is always the biggest risk factor,” Dr. Machado explained. “There’s always a gradient: the older the patient, the worse the outcome.”

Indeed, there was a threefold increased risk for death among those aged 66-75 years versus those who were 65 years or younger (odds ratio, 3.00), and a sixfold increased risk for patients older than 75, compared with the younger age group (OR, 6.18).

Having both hypertension and cardiovascular disease was associated with an OR of 1.89, and coexisting chronic lung disease also significantly increased the chances of dying from COVID-19 (OR, 1.68).

Being of male sex was associated with a 46% increased risk for death from COVID-19 versus being of female sex.

The risk for COVID-19 death also rose with the use of corticosteroids. Compared with no steroid use, there was a 69% increased risk for with death at doses of 10 mg or more prednisolone equivalent per day.

“The finding about moderate to high doses of steroids being associated with a worse outcome is consistent with the first report; it was the same for hospitalization,” Dr. Machado observed.

The general consensus on steroid use in the COVID-19 setting is that they should be continued as needed, but at the lowest possible dose, as outlined in provisional recommendations set out by the recently renamed European Alliance of Associations for Rheumatology.

The GRA physician registry findings provide further support for this, suggesting that disease control should be optimized with disease-modifying antirheumatic drugs, ideally without increasing the dose of steroids.

Surprise over sulfasalazine risk

“Taking all medications into account – such as methotrexate, leflunomide, hydroxychloroquine, [tumor necrosis factor] blockers, interleukin-6 blockers, and [Janus kinase] inhibitors – it is quite reassuring because we did not see an association with worse outcome with those drugs overall,” Dr. Machado said.

However, treatment with rituximab (OR, 4.0), sulfasalazine (OR, 3.6), and immunosuppressive agents such as azathioprine, cyclophosphamide, cyclosporine, mycophenolate, or tacrolimus (OR, 2.2), were associated with higher odds of dying from COVID-19 when compared with treatment with methotrexate alone.

The findings for rituximab and immunosuppressant use were perhaps not unexpected, but the possible association between sulfasalazine and COVID-19 death was “a bit intriguing,” Dr. Machado observed. “Sulfasalazine is believed to have low immunosuppressive effect.”

This warrants further investigation, but there are likely a range of confounding factors at play. One could be that people considered to be at higher risk may have been more often prescribed sulfasalazine because it was thought to be less immunosuppressive. Another might be because people taking sulfasalazine were more likely to be smokers, and they were also not advised to protect themselves from exposure to the virus (shielding) during the first wave of the pandemic, at least not in the United Kingdom.
 

Rituximab caution and vaccination

“Rituximab is a concern,” Dr. Machado acknowledged. “It is a concern that rheumatologists are now aware of and they are addressing, but then it’s a concern for a very specific subgroup of patients.”

While rheumatologists are, and will continue to prescribe it, there will be even more careful consideration over when, in whom, and how to use it during, and possibly even after, the pandemic.

“COVID is here to stay, it will become endemic, and it’s going to be part of our lives like the flu virus is,” Dr. Machado predicted.

Then there is the issue on vaccinating people against COVID-19, should those on rituximab still receive it? The answer is a yes, but, as with other vaccinations it’s all about the timing of when the vaccination is given.

Societies such as the British Society for Rheumatology have already begun to include guidance on this, recommending one of the available COVID-19 vaccines is given at least a month before the next or first dose of rituximab is due. As rituximab is given every few months, with doses sometimes spaced as much as 9 months or even a year apart, this should not be too much of a problem, but it is “better to have the vaccine first,” Dr. Machado said.
 

Has COVID-19 care improved in RMDs?

In separate research published in The Lancet Rheumatology, April Jorge, MD, of Massachusetts General Hospital and Harvard Medical School, both in Boston, and associates found that the risks of severe COVID-19 outcomes have improved over time, although they still “remain substantial.”

Dr. Jorge and colleagues looked at temporal trends in COVID-19 outcomes in patients with RMDs over the course of the first 6 months of the pandemic in 2020, using data from a large, multicenter, electronic health record network (TriNetX).

They formed two patient cohorts – a late (diagnosed from April 20 to July 20) and an early (diagnosed from January 20 to April 20) cohort – to see if outcomes had improved and discovered lower relative risks among patients in the late cohort for hospitalization (0.67), admission to the ICU (0.56), mechanical ventilation (0.39), acute kidney injury (0.66), renal replacement (0.53), and death (0.39).

“These results are encouraging,” but it’s difficult to match these different populations of patients, Dr. Machado said. “There are always factors that you cannot match for” and were not included in the U.S. analysis.

While there are important caveats in how the analysis was performed and thus in interpreting these data, they do “suggest that one of the reasons why outcomes have improved is because we have become better at treating these patients,” Dr. Machado added.

“Our treatment has improved, and our capacity to treat the complications has improved. We understand better how the disease behaves – we know that they can have thromboembolic complications that we can manage, and we are now able to manage ventilation issues better.”

Moreover, Dr. Machado said that, not only were clinicians more aware of what they should or should not do, there were treatments that were being used routinely or in some cases based on recent clinical trial results. “I think we are indeed treating these patients better.”

The COVID-19 GRA physician registry is financially supported by the American College of Rheumatology and EULAR. Dr. Machado had no relevant conflicts of interest.

People with rheumatic and musculoskeletal diseases (RMDs) who contract the SARS-CoV-2 virus appear more likely to die from COVID-19 if their rheumatologic condition is not being well controlled at the time of their infection.

New data from the COVID-19 Global Rheumatology Alliance (GRA) physician registry reported in Annals of the Rheumatic Diseases have found that the odds of dying from COVID-19 were 87% higher in individuals recorded as having moderate to high disease activity versus those reported to be in remission or having low disease activity.

“I think this really highlights the importance of continuing to appropriately, and actively, treat our patients, and the importance of controlling their disease,” Pedro Machado, MD, PhD, said in an interview. Dr. Machado, an associate professor in rheumatology and muscle diseases at University College London and a consultant rheumatologist at several U.K. hospitals, has been involved in the GRA physician registry from the start, and sits on the GRA steering committee.

Alongside higher disease activity, several other important factors were found to be associated with increased odds of dying from COVID-19 – older age, male gender, and the presence of one or more comorbidities, such as hypertension combined with cardiovascular disease or chronic lung disease.

These demographic and disease-based factors have been linked to an increased risk for COVID-19–related hospitalization before, both in people with RMDs and in the general population, but the latest GRA physician registry data now take that a step further, and link them also to an increased risk for death, together with several other factors more specific to RMDs.
 

Logging COVID-19 rheumatologic cases

Since the start of the global pandemic, the potential effects that SARS-CoV-2 infection might have on people with RMDs in particular has concerned the rheumatology community. The main worries being that, either because of the underlying RMD itself or to its treatment, there may be immunoregulatory deficits or other risk factors that would make individuals more susceptible to not only infection but also to developing more severe COVID-19 than the general population.

These concerns led to the rapid formation of the GRA and the COVID-19 GRA physician registry in March 2020 to collect and analyze data on adults with rheumatic disease and confirmed or presumptive COVID-19. Entries into the registry are made by or under the direction of rheumatologists, and this is a voluntary process.

“This population cannot ever be entirely representative of the population of patients with rheumatic diseases,” Dr. Machado acknowledged. There will be selection and other biases that affect the reported data. That said, it’s the largest database of reported COVID-19 cases in adult rheumatology patients across the world, with more than 9,000 cases so far included from multiple registries, including those based in Europe and North and South America. Data from one of these – the French RMD cohort – have also recently been published in Annals of the Rheumatic Diseases, showing much the same findings but on a national level.

Hospitalization was the focus of a previous report because “you need large sample sizes” to look at endpoints that occur less frequently. When the first analysis was done, there were around 600 cases from 40 countries in the registry with sufficient data that could be used. Now, with a greater number of recorded cases, factors influencing the risk for death could be examined.

Death rate and risk factors found

Data on 3,729 COVID-19 cases in people with RMDs were included in the current analysis, all recorded in the first few months of the registry being open and up until July 1, 2020. In all, 390 (10.5%) of people died. While this is “clearly higher” than reported in the general population in most countries, the analysis was not designed to calculate a precise estimate.

“It should not be taken as an estimate of the overall death rate among patients with rheumatic diseases and COVID-19,” Dr. Machado and coauthors have been keen to point out.

“Age is always the biggest risk factor,” Dr. Machado explained. “There’s always a gradient: the older the patient, the worse the outcome.”

Indeed, there was a threefold increased risk for death among those aged 66-75 years versus those who were 65 years or younger (odds ratio, 3.00), and a sixfold increased risk for patients older than 75, compared with the younger age group (OR, 6.18).

Having both hypertension and cardiovascular disease was associated with an OR of 1.89, and coexisting chronic lung disease also significantly increased the chances of dying from COVID-19 (OR, 1.68).

Being of male sex was associated with a 46% increased risk for death from COVID-19 versus being of female sex.

The risk for COVID-19 death also rose with the use of corticosteroids. Compared with no steroid use, there was a 69% increased risk for with death at doses of 10 mg or more prednisolone equivalent per day.

“The finding about moderate to high doses of steroids being associated with a worse outcome is consistent with the first report; it was the same for hospitalization,” Dr. Machado observed.

The general consensus on steroid use in the COVID-19 setting is that they should be continued as needed, but at the lowest possible dose, as outlined in provisional recommendations set out by the recently renamed European Alliance of Associations for Rheumatology.

The GRA physician registry findings provide further support for this, suggesting that disease control should be optimized with disease-modifying antirheumatic drugs, ideally without increasing the dose of steroids.

Surprise over sulfasalazine risk

“Taking all medications into account – such as methotrexate, leflunomide, hydroxychloroquine, [tumor necrosis factor] blockers, interleukin-6 blockers, and [Janus kinase] inhibitors – it is quite reassuring because we did not see an association with worse outcome with those drugs overall,” Dr. Machado said.

However, treatment with rituximab (OR, 4.0), sulfasalazine (OR, 3.6), and immunosuppressive agents such as azathioprine, cyclophosphamide, cyclosporine, mycophenolate, or tacrolimus (OR, 2.2), were associated with higher odds of dying from COVID-19 when compared with treatment with methotrexate alone.

The findings for rituximab and immunosuppressant use were perhaps not unexpected, but the possible association between sulfasalazine and COVID-19 death was “a bit intriguing,” Dr. Machado observed. “Sulfasalazine is believed to have low immunosuppressive effect.”

This warrants further investigation, but there are likely a range of confounding factors at play. One could be that people considered to be at higher risk may have been more often prescribed sulfasalazine because it was thought to be less immunosuppressive. Another might be because people taking sulfasalazine were more likely to be smokers, and they were also not advised to protect themselves from exposure to the virus (shielding) during the first wave of the pandemic, at least not in the United Kingdom.
 

Rituximab caution and vaccination

“Rituximab is a concern,” Dr. Machado acknowledged. “It is a concern that rheumatologists are now aware of and they are addressing, but then it’s a concern for a very specific subgroup of patients.”

While rheumatologists are, and will continue to prescribe it, there will be even more careful consideration over when, in whom, and how to use it during, and possibly even after, the pandemic.

“COVID is here to stay, it will become endemic, and it’s going to be part of our lives like the flu virus is,” Dr. Machado predicted.

Then there is the issue on vaccinating people against COVID-19, should those on rituximab still receive it? The answer is a yes, but, as with other vaccinations it’s all about the timing of when the vaccination is given.

Societies such as the British Society for Rheumatology have already begun to include guidance on this, recommending one of the available COVID-19 vaccines is given at least a month before the next or first dose of rituximab is due. As rituximab is given every few months, with doses sometimes spaced as much as 9 months or even a year apart, this should not be too much of a problem, but it is “better to have the vaccine first,” Dr. Machado said.
 

Has COVID-19 care improved in RMDs?

In separate research published in The Lancet Rheumatology, April Jorge, MD, of Massachusetts General Hospital and Harvard Medical School, both in Boston, and associates found that the risks of severe COVID-19 outcomes have improved over time, although they still “remain substantial.”

Dr. Jorge and colleagues looked at temporal trends in COVID-19 outcomes in patients with RMDs over the course of the first 6 months of the pandemic in 2020, using data from a large, multicenter, electronic health record network (TriNetX).

They formed two patient cohorts – a late (diagnosed from April 20 to July 20) and an early (diagnosed from January 20 to April 20) cohort – to see if outcomes had improved and discovered lower relative risks among patients in the late cohort for hospitalization (0.67), admission to the ICU (0.56), mechanical ventilation (0.39), acute kidney injury (0.66), renal replacement (0.53), and death (0.39).

“These results are encouraging,” but it’s difficult to match these different populations of patients, Dr. Machado said. “There are always factors that you cannot match for” and were not included in the U.S. analysis.

While there are important caveats in how the analysis was performed and thus in interpreting these data, they do “suggest that one of the reasons why outcomes have improved is because we have become better at treating these patients,” Dr. Machado added.

“Our treatment has improved, and our capacity to treat the complications has improved. We understand better how the disease behaves – we know that they can have thromboembolic complications that we can manage, and we are now able to manage ventilation issues better.”

Moreover, Dr. Machado said that, not only were clinicians more aware of what they should or should not do, there were treatments that were being used routinely or in some cases based on recent clinical trial results. “I think we are indeed treating these patients better.”

The COVID-19 GRA physician registry is financially supported by the American College of Rheumatology and EULAR. Dr. Machado had no relevant conflicts of interest.

People with rheumatic and musculoskeletal diseases (RMDs) who contract the SARS-CoV-2 virus appear more likely to die from COVID-19 if their rheumatologic condition is not being well controlled at the time of their infection.

New data from the COVID-19 Global Rheumatology Alliance (GRA) physician registry reported in Annals of the Rheumatic Diseases have found that the odds of dying from COVID-19 were 87% higher in individuals recorded as having moderate to high disease activity versus those reported to be in remission or having low disease activity.

“I think this really highlights the importance of continuing to appropriately, and actively, treat our patients, and the importance of controlling their disease,” Pedro Machado, MD, PhD, said in an interview. Dr. Machado, an associate professor in rheumatology and muscle diseases at University College London and a consultant rheumatologist at several U.K. hospitals, has been involved in the GRA physician registry from the start, and sits on the GRA steering committee.

Alongside higher disease activity, several other important factors were found to be associated with increased odds of dying from COVID-19 – older age, male gender, and the presence of one or more comorbidities, such as hypertension combined with cardiovascular disease or chronic lung disease.

These demographic and disease-based factors have been linked to an increased risk for COVID-19–related hospitalization before, both in people with RMDs and in the general population, but the latest GRA physician registry data now take that a step further, and link them also to an increased risk for death, together with several other factors more specific to RMDs.
 

Logging COVID-19 rheumatologic cases

Since the start of the global pandemic, the potential effects that SARS-CoV-2 infection might have on people with RMDs in particular has concerned the rheumatology community. The main worries being that, either because of the underlying RMD itself or to its treatment, there may be immunoregulatory deficits or other risk factors that would make individuals more susceptible to not only infection but also to developing more severe COVID-19 than the general population.

These concerns led to the rapid formation of the GRA and the COVID-19 GRA physician registry in March 2020 to collect and analyze data on adults with rheumatic disease and confirmed or presumptive COVID-19. Entries into the registry are made by or under the direction of rheumatologists, and this is a voluntary process.

“This population cannot ever be entirely representative of the population of patients with rheumatic diseases,” Dr. Machado acknowledged. There will be selection and other biases that affect the reported data. That said, it’s the largest database of reported COVID-19 cases in adult rheumatology patients across the world, with more than 9,000 cases so far included from multiple registries, including those based in Europe and North and South America. Data from one of these – the French RMD cohort – have also recently been published in Annals of the Rheumatic Diseases, showing much the same findings but on a national level.

Hospitalization was the focus of a previous report because “you need large sample sizes” to look at endpoints that occur less frequently. When the first analysis was done, there were around 600 cases from 40 countries in the registry with sufficient data that could be used. Now, with a greater number of recorded cases, factors influencing the risk for death could be examined.

Death rate and risk factors found

Data on 3,729 COVID-19 cases in people with RMDs were included in the current analysis, all recorded in the first few months of the registry being open and up until July 1, 2020. In all, 390 (10.5%) of people died. While this is “clearly higher” than reported in the general population in most countries, the analysis was not designed to calculate a precise estimate.

“It should not be taken as an estimate of the overall death rate among patients with rheumatic diseases and COVID-19,” Dr. Machado and coauthors have been keen to point out.

“Age is always the biggest risk factor,” Dr. Machado explained. “There’s always a gradient: the older the patient, the worse the outcome.”

Indeed, there was a threefold increased risk for death among those aged 66-75 years versus those who were 65 years or younger (odds ratio, 3.00), and a sixfold increased risk for patients older than 75, compared with the younger age group (OR, 6.18).

Having both hypertension and cardiovascular disease was associated with an OR of 1.89, and coexisting chronic lung disease also significantly increased the chances of dying from COVID-19 (OR, 1.68).

Being of male sex was associated with a 46% increased risk for death from COVID-19 versus being of female sex.

The risk for COVID-19 death also rose with the use of corticosteroids. Compared with no steroid use, there was a 69% increased risk for with death at doses of 10 mg or more prednisolone equivalent per day.

“The finding about moderate to high doses of steroids being associated with a worse outcome is consistent with the first report; it was the same for hospitalization,” Dr. Machado observed.

The general consensus on steroid use in the COVID-19 setting is that they should be continued as needed, but at the lowest possible dose, as outlined in provisional recommendations set out by the recently renamed European Alliance of Associations for Rheumatology.

The GRA physician registry findings provide further support for this, suggesting that disease control should be optimized with disease-modifying antirheumatic drugs, ideally without increasing the dose of steroids.

Surprise over sulfasalazine risk

“Taking all medications into account – such as methotrexate, leflunomide, hydroxychloroquine, [tumor necrosis factor] blockers, interleukin-6 blockers, and [Janus kinase] inhibitors – it is quite reassuring because we did not see an association with worse outcome with those drugs overall,” Dr. Machado said.

However, treatment with rituximab (OR, 4.0), sulfasalazine (OR, 3.6), and immunosuppressive agents such as azathioprine, cyclophosphamide, cyclosporine, mycophenolate, or tacrolimus (OR, 2.2), were associated with higher odds of dying from COVID-19 when compared with treatment with methotrexate alone.

The findings for rituximab and immunosuppressant use were perhaps not unexpected, but the possible association between sulfasalazine and COVID-19 death was “a bit intriguing,” Dr. Machado observed. “Sulfasalazine is believed to have low immunosuppressive effect.”

This warrants further investigation, but there are likely a range of confounding factors at play. One could be that people considered to be at higher risk may have been more often prescribed sulfasalazine because it was thought to be less immunosuppressive. Another might be because people taking sulfasalazine were more likely to be smokers, and they were also not advised to protect themselves from exposure to the virus (shielding) during the first wave of the pandemic, at least not in the United Kingdom.
 

Rituximab caution and vaccination

“Rituximab is a concern,” Dr. Machado acknowledged. “It is a concern that rheumatologists are now aware of and they are addressing, but then it’s a concern for a very specific subgroup of patients.”

While rheumatologists are, and will continue to prescribe it, there will be even more careful consideration over when, in whom, and how to use it during, and possibly even after, the pandemic.

“COVID is here to stay, it will become endemic, and it’s going to be part of our lives like the flu virus is,” Dr. Machado predicted.

Then there is the issue on vaccinating people against COVID-19, should those on rituximab still receive it? The answer is a yes, but, as with other vaccinations it’s all about the timing of when the vaccination is given.

Societies such as the British Society for Rheumatology have already begun to include guidance on this, recommending one of the available COVID-19 vaccines is given at least a month before the next or first dose of rituximab is due. As rituximab is given every few months, with doses sometimes spaced as much as 9 months or even a year apart, this should not be too much of a problem, but it is “better to have the vaccine first,” Dr. Machado said.
 

Has COVID-19 care improved in RMDs?

In separate research published in The Lancet Rheumatology, April Jorge, MD, of Massachusetts General Hospital and Harvard Medical School, both in Boston, and associates found that the risks of severe COVID-19 outcomes have improved over time, although they still “remain substantial.”

Dr. Jorge and colleagues looked at temporal trends in COVID-19 outcomes in patients with RMDs over the course of the first 6 months of the pandemic in 2020, using data from a large, multicenter, electronic health record network (TriNetX).

They formed two patient cohorts – a late (diagnosed from April 20 to July 20) and an early (diagnosed from January 20 to April 20) cohort – to see if outcomes had improved and discovered lower relative risks among patients in the late cohort for hospitalization (0.67), admission to the ICU (0.56), mechanical ventilation (0.39), acute kidney injury (0.66), renal replacement (0.53), and death (0.39).

“These results are encouraging,” but it’s difficult to match these different populations of patients, Dr. Machado said. “There are always factors that you cannot match for” and were not included in the U.S. analysis.

While there are important caveats in how the analysis was performed and thus in interpreting these data, they do “suggest that one of the reasons why outcomes have improved is because we have become better at treating these patients,” Dr. Machado added.

“Our treatment has improved, and our capacity to treat the complications has improved. We understand better how the disease behaves – we know that they can have thromboembolic complications that we can manage, and we are now able to manage ventilation issues better.”

Moreover, Dr. Machado said that, not only were clinicians more aware of what they should or should not do, there were treatments that were being used routinely or in some cases based on recent clinical trial results. “I think we are indeed treating these patients better.”

The COVID-19 GRA physician registry is financially supported by the American College of Rheumatology and EULAR. Dr. Machado had no relevant conflicts of interest.

Nanoparticle encapsulation may enable targeting of aberrant glucose metabolism in hepatocellular carcinoma (HCC), potentially amplifying the effects of existing therapies and overcoming resistance mechanisms, according to investigators.

In a preclinical trial involving cell lines, xenograft tumors, and mouse models, encapsulated 2-deoxy-D-glucose (2DG) nanoparticles enhanced the antineoplastic effects of sorafenib and checkpoint inhibitors and suppressed anti-programmed cell death protein 1 (PD1)–resistant tumors, reported lead author Kyo Sasaki, PhD, of Kyushu University in Fukuoka, Japan, and colleagues.

As a glycolysis inhibitor, 2DG acts against the Warburg effect, a cancer immune-resistance mechanism “in which a substantial amount of pyruvate is reduced to lactic acid instead of being directed into mitochondria,” the investigators wrote. Their report is in Cellular and Molecular Gastroenterology and Hepatology.

But this isn’t new information, and Dr. Sasaki and colleagues weren’t the first to address the Warburg effect with 2DG; two clinical trials reported signs of efficacy in patients with solid tumors, one in 2010 and the other in 2013.

“However, 2DG does not seem to have a significant effect on tumor growth at a dose that does not induce serious adverse effects,” wrote Dr. Sasaki and colleagues. “These results suggest a need to develop an efficient drug delivery system for 2DG.”

The investigators turned to nanoparticles, which accumulate in tumor tissue more than healthy tissue, thereby limiting off-target toxicity. Specifically, they encapsulated 2DG in nanoparticles of poly(lactic-co-glycolic acid) (PLGA), a Food and Drug Administration–approved biodegradable polymer.

After characterizing the physical properties of the encapsulated 2DG nanoparticles (2DG-PLGA-NPs), and observing tumor localization in nude mice with xenograft liver tumors, the investigators assessed cytotoxic effects.

Treatment resulted in “significant growth reduction” of not only xenograft liver tumors, but also xenograft renal, colon, and pancreatic tumors, “indicating the potential antitumor effects of this method against various tumors.” Furthermore, mice treated with encapsulated 2DG nanoparticles had significantly less weight loss compared with those receiving conventional 2DG, suggesting a reduction in 2DG-related adverse effects.

Additional experiments involving two immunocompetent mouse models with multiple large liver tumors added data to support to the relative efficacy of encapsulated versus nonencapsulated 2DG. Both mouse models had significant reductions in liver tumors when treated with 2DG-PLGA-NPs; in contrast, treatment with 2DG alone reduced tumor number in only one of the two mouse models and to a lesser degree than treatment with 2DG nanoparticles.

Further in vivo and ex vivo testing revealed that encapsulated 2DG nanoparticles exerted their cytotoxic effects via endoplasmic reticulum stress, oxidative stress, and inactivation of mTOR. Simultaneously, treatment was associated with CD8+ T-cell migration into tumor tissue via increased glucose uptake and IFN-gamma production in CD8+ T cells, reduced lactate production in tumors, and increased production of CXCL9/CXCL10/CXCL11 in both the tumors and CD8+ T cells.

According to the investigators, these findings suggested that 2DG-PLGA-NPs might upregulate PD-1 positive T cells in tumors, thereby enhancing the effects of a checkpoint inhibitor. Indeed, when syngeneic mice with anti-PD-1–resistant tumors were treated with encapsulated 2DG nanoparticles, the investigators observed significant reductions in tumor growth, compared with treatment using an isotype control, PLGA alone, or an anti-PD-1 antibody. And in nude mice with xenograft tumors, combination therapy with 2DG-PLGA-NPs and sorafenib significantly reduced tumor growth, compared with no treatment, 2DG, PLGA, or PLGA with sorafenib.

“2DG-PLGA-NPs amplified the antitumor effect of anti-PD1 or sorafenib, and showed an antitumor effect against anti-PD1–resistant tumors,” the investigators wrote.

Dr. Sasaki and colleagues also noted that encapsulated 2DG nanoparticles did not accumulate in nontumorous cirrhotic hepatocytes, which suggests that treatment would be safe for patients with chronic liver diseases.

“Another practical concern is the extent to which 2DG is effectively taken up by HCC cells,” the investigators wrote.

PET showed that the hepatic accumulation rate of F-2-fluoro-2-deoxyglucose (F-FDG), a radioactive tracer of 2DG, was 50% in well-differentiated HCC, and “much higher” in sorafenib-resistant HCC cells and poorly and moderately differentiated HCC cells.

“Thus, 2DG-PLGA-NPs are expected to be good therapeutic agent candidates for patients with advanced HCC,” the investigators concluded.

The investigators disclosed no conflicts of interest. Some authors received grants from the Japan Society for the Promotion of Science.

Nanoparticle encapsulation may enable targeting of aberrant glucose metabolism in hepatocellular carcinoma (HCC), potentially amplifying the effects of existing therapies and overcoming resistance mechanisms, according to investigators.

In a preclinical trial involving cell lines, xenograft tumors, and mouse models, encapsulated 2-deoxy-D-glucose (2DG) nanoparticles enhanced the antineoplastic effects of sorafenib and checkpoint inhibitors and suppressed anti-programmed cell death protein 1 (PD1)–resistant tumors, reported lead author Kyo Sasaki, PhD, of Kyushu University in Fukuoka, Japan, and colleagues.

As a glycolysis inhibitor, 2DG acts against the Warburg effect, a cancer immune-resistance mechanism “in which a substantial amount of pyruvate is reduced to lactic acid instead of being directed into mitochondria,” the investigators wrote. Their report is in Cellular and Molecular Gastroenterology and Hepatology.

But this isn’t new information, and Dr. Sasaki and colleagues weren’t the first to address the Warburg effect with 2DG; two clinical trials reported signs of efficacy in patients with solid tumors, one in 2010 and the other in 2013.

“However, 2DG does not seem to have a significant effect on tumor growth at a dose that does not induce serious adverse effects,” wrote Dr. Sasaki and colleagues. “These results suggest a need to develop an efficient drug delivery system for 2DG.”

The investigators turned to nanoparticles, which accumulate in tumor tissue more than healthy tissue, thereby limiting off-target toxicity. Specifically, they encapsulated 2DG in nanoparticles of poly(lactic-co-glycolic acid) (PLGA), a Food and Drug Administration–approved biodegradable polymer.

After characterizing the physical properties of the encapsulated 2DG nanoparticles (2DG-PLGA-NPs), and observing tumor localization in nude mice with xenograft liver tumors, the investigators assessed cytotoxic effects.

Treatment resulted in “significant growth reduction” of not only xenograft liver tumors, but also xenograft renal, colon, and pancreatic tumors, “indicating the potential antitumor effects of this method against various tumors.” Furthermore, mice treated with encapsulated 2DG nanoparticles had significantly less weight loss compared with those receiving conventional 2DG, suggesting a reduction in 2DG-related adverse effects.

Additional experiments involving two immunocompetent mouse models with multiple large liver tumors added data to support to the relative efficacy of encapsulated versus nonencapsulated 2DG. Both mouse models had significant reductions in liver tumors when treated with 2DG-PLGA-NPs; in contrast, treatment with 2DG alone reduced tumor number in only one of the two mouse models and to a lesser degree than treatment with 2DG nanoparticles.

Further in vivo and ex vivo testing revealed that encapsulated 2DG nanoparticles exerted their cytotoxic effects via endoplasmic reticulum stress, oxidative stress, and inactivation of mTOR. Simultaneously, treatment was associated with CD8+ T-cell migration into tumor tissue via increased glucose uptake and IFN-gamma production in CD8+ T cells, reduced lactate production in tumors, and increased production of CXCL9/CXCL10/CXCL11 in both the tumors and CD8+ T cells.

According to the investigators, these findings suggested that 2DG-PLGA-NPs might upregulate PD-1 positive T cells in tumors, thereby enhancing the effects of a checkpoint inhibitor. Indeed, when syngeneic mice with anti-PD-1–resistant tumors were treated with encapsulated 2DG nanoparticles, the investigators observed significant reductions in tumor growth, compared with treatment using an isotype control, PLGA alone, or an anti-PD-1 antibody. And in nude mice with xenograft tumors, combination therapy with 2DG-PLGA-NPs and sorafenib significantly reduced tumor growth, compared with no treatment, 2DG, PLGA, or PLGA with sorafenib.

“2DG-PLGA-NPs amplified the antitumor effect of anti-PD1 or sorafenib, and showed an antitumor effect against anti-PD1–resistant tumors,” the investigators wrote.

Dr. Sasaki and colleagues also noted that encapsulated 2DG nanoparticles did not accumulate in nontumorous cirrhotic hepatocytes, which suggests that treatment would be safe for patients with chronic liver diseases.

“Another practical concern is the extent to which 2DG is effectively taken up by HCC cells,” the investigators wrote.

PET showed that the hepatic accumulation rate of F-2-fluoro-2-deoxyglucose (F-FDG), a radioactive tracer of 2DG, was 50% in well-differentiated HCC, and “much higher” in sorafenib-resistant HCC cells and poorly and moderately differentiated HCC cells.

“Thus, 2DG-PLGA-NPs are expected to be good therapeutic agent candidates for patients with advanced HCC,” the investigators concluded.

The investigators disclosed no conflicts of interest. Some authors received grants from the Japan Society for the Promotion of Science.

Nanoparticle encapsulation may enable targeting of aberrant glucose metabolism in hepatocellular carcinoma (HCC), potentially amplifying the effects of existing therapies and overcoming resistance mechanisms, according to investigators.

In a preclinical trial involving cell lines, xenograft tumors, and mouse models, encapsulated 2-deoxy-D-glucose (2DG) nanoparticles enhanced the antineoplastic effects of sorafenib and checkpoint inhibitors and suppressed anti-programmed cell death protein 1 (PD1)–resistant tumors, reported lead author Kyo Sasaki, PhD, of Kyushu University in Fukuoka, Japan, and colleagues.

As a glycolysis inhibitor, 2DG acts against the Warburg effect, a cancer immune-resistance mechanism “in which a substantial amount of pyruvate is reduced to lactic acid instead of being directed into mitochondria,” the investigators wrote. Their report is in Cellular and Molecular Gastroenterology and Hepatology.

But this isn’t new information, and Dr. Sasaki and colleagues weren’t the first to address the Warburg effect with 2DG; two clinical trials reported signs of efficacy in patients with solid tumors, one in 2010 and the other in 2013.

“However, 2DG does not seem to have a significant effect on tumor growth at a dose that does not induce serious adverse effects,” wrote Dr. Sasaki and colleagues. “These results suggest a need to develop an efficient drug delivery system for 2DG.”

The investigators turned to nanoparticles, which accumulate in tumor tissue more than healthy tissue, thereby limiting off-target toxicity. Specifically, they encapsulated 2DG in nanoparticles of poly(lactic-co-glycolic acid) (PLGA), a Food and Drug Administration–approved biodegradable polymer.

After characterizing the physical properties of the encapsulated 2DG nanoparticles (2DG-PLGA-NPs), and observing tumor localization in nude mice with xenograft liver tumors, the investigators assessed cytotoxic effects.

Treatment resulted in “significant growth reduction” of not only xenograft liver tumors, but also xenograft renal, colon, and pancreatic tumors, “indicating the potential antitumor effects of this method against various tumors.” Furthermore, mice treated with encapsulated 2DG nanoparticles had significantly less weight loss compared with those receiving conventional 2DG, suggesting a reduction in 2DG-related adverse effects.

Additional experiments involving two immunocompetent mouse models with multiple large liver tumors added data to support to the relative efficacy of encapsulated versus nonencapsulated 2DG. Both mouse models had significant reductions in liver tumors when treated with 2DG-PLGA-NPs; in contrast, treatment with 2DG alone reduced tumor number in only one of the two mouse models and to a lesser degree than treatment with 2DG nanoparticles.

Further in vivo and ex vivo testing revealed that encapsulated 2DG nanoparticles exerted their cytotoxic effects via endoplasmic reticulum stress, oxidative stress, and inactivation of mTOR. Simultaneously, treatment was associated with CD8+ T-cell migration into tumor tissue via increased glucose uptake and IFN-gamma production in CD8+ T cells, reduced lactate production in tumors, and increased production of CXCL9/CXCL10/CXCL11 in both the tumors and CD8+ T cells.

According to the investigators, these findings suggested that 2DG-PLGA-NPs might upregulate PD-1 positive T cells in tumors, thereby enhancing the effects of a checkpoint inhibitor. Indeed, when syngeneic mice with anti-PD-1–resistant tumors were treated with encapsulated 2DG nanoparticles, the investigators observed significant reductions in tumor growth, compared with treatment using an isotype control, PLGA alone, or an anti-PD-1 antibody. And in nude mice with xenograft tumors, combination therapy with 2DG-PLGA-NPs and sorafenib significantly reduced tumor growth, compared with no treatment, 2DG, PLGA, or PLGA with sorafenib.

“2DG-PLGA-NPs amplified the antitumor effect of anti-PD1 or sorafenib, and showed an antitumor effect against anti-PD1–resistant tumors,” the investigators wrote.

Dr. Sasaki and colleagues also noted that encapsulated 2DG nanoparticles did not accumulate in nontumorous cirrhotic hepatocytes, which suggests that treatment would be safe for patients with chronic liver diseases.

“Another practical concern is the extent to which 2DG is effectively taken up by HCC cells,” the investigators wrote.

PET showed that the hepatic accumulation rate of F-2-fluoro-2-deoxyglucose (F-FDG), a radioactive tracer of 2DG, was 50% in well-differentiated HCC, and “much higher” in sorafenib-resistant HCC cells and poorly and moderately differentiated HCC cells.

“Thus, 2DG-PLGA-NPs are expected to be good therapeutic agent candidates for patients with advanced HCC,” the investigators concluded.

The investigators disclosed no conflicts of interest. Some authors received grants from the Japan Society for the Promotion of Science.

Key clinical point: Erenumab 70 mg was effective for the treatment of high-frequency episodic migraine (HFEM) or chronic migraine (CM) in a real-world setting.

Major finding: Erenumab was effective with a rapid progressive decrease in monthly migraine days (week 12: HEFM, 4.5 days; CM, 9.3 days). More than 50% of HEFM and 75% of CM patients responded to the treatment.

Study details: The data come from a real-life prospective cohort study of 372 patients affected by HEFM or CM with 3 or more prior preventive therapeutic failures.

Disclosures: This work was partially supported by the Italian Ministry of Health (Institutional Funding Ricerca Corrente). C Altamura, N Brunelli, A Fallacara, CM Costa, D Santangelo, A Salerno, B Mercuri, and A Carnevale declared no conflicts of interest. The remaining authors reported ties with various institutions and/or pharmaceutical companies.

Source: Barbanti P et al. Headache. 2020 Dec 18. doi: 10.1111/head.14032.

Key clinical point: Erenumab 70 mg was effective for the treatment of high-frequency episodic migraine (HFEM) or chronic migraine (CM) in a real-world setting.

Major finding: Erenumab was effective with a rapid progressive decrease in monthly migraine days (week 12: HEFM, 4.5 days; CM, 9.3 days). More than 50% of HEFM and 75% of CM patients responded to the treatment.

Study details: The data come from a real-life prospective cohort study of 372 patients affected by HEFM or CM with 3 or more prior preventive therapeutic failures.

Disclosures: This work was partially supported by the Italian Ministry of Health (Institutional Funding Ricerca Corrente). C Altamura, N Brunelli, A Fallacara, CM Costa, D Santangelo, A Salerno, B Mercuri, and A Carnevale declared no conflicts of interest. The remaining authors reported ties with various institutions and/or pharmaceutical companies.

Source: Barbanti P et al. Headache. 2020 Dec 18. doi: 10.1111/head.14032.

Key clinical point: Erenumab 70 mg was effective for the treatment of high-frequency episodic migraine (HFEM) or chronic migraine (CM) in a real-world setting.

Major finding: Erenumab was effective with a rapid progressive decrease in monthly migraine days (week 12: HEFM, 4.5 days; CM, 9.3 days). More than 50% of HEFM and 75% of CM patients responded to the treatment.

Study details: The data come from a real-life prospective cohort study of 372 patients affected by HEFM or CM with 3 or more prior preventive therapeutic failures.

Disclosures: This work was partially supported by the Italian Ministry of Health (Institutional Funding Ricerca Corrente). C Altamura, N Brunelli, A Fallacara, CM Costa, D Santangelo, A Salerno, B Mercuri, and A Carnevale declared no conflicts of interest. The remaining authors reported ties with various institutions and/or pharmaceutical companies.

Source: Barbanti P et al. Headache. 2020 Dec 18. doi: 10.1111/head.14032.

Key clinical point: Chronic migraine patients who achieved early reversal to episodic migraine are likely to have persistent reversion with long-term erenumab treatment.

Major finding: At 12 weeks, 54.1% (95% confidence interval [CI], 46.6%-61.6%) of patients showed reversal to episodic migraine. Continued treatment with erenumab showed a long-term persistent reversion in 96.8% (95% CI, 91.1%-99.3%) of patients at week 64.

Study details: Data on 181 migraine patients come from a post hoc analysis of a 12-week randomized double-blind trial and a 52-week open-label extension.

Disclosures: The study was funded by Amgen Inc. The authors reported ties with various institutions and/or pharmaceutical companies. F Zhang, GA Rippon, S Cheng, and DD Mikol are employed by and own stock in Amgen.

Source: Lipton RB et al. Cephalalgia. 2020 Dec 3. doi: 10.1177/0333102420973994.

Key clinical point: Chronic migraine patients who achieved early reversal to episodic migraine are likely to have persistent reversion with long-term erenumab treatment.

Major finding: At 12 weeks, 54.1% (95% confidence interval [CI], 46.6%-61.6%) of patients showed reversal to episodic migraine. Continued treatment with erenumab showed a long-term persistent reversion in 96.8% (95% CI, 91.1%-99.3%) of patients at week 64.

Study details: Data on 181 migraine patients come from a post hoc analysis of a 12-week randomized double-blind trial and a 52-week open-label extension.

Disclosures: The study was funded by Amgen Inc. The authors reported ties with various institutions and/or pharmaceutical companies. F Zhang, GA Rippon, S Cheng, and DD Mikol are employed by and own stock in Amgen.

Source: Lipton RB et al. Cephalalgia. 2020 Dec 3. doi: 10.1177/0333102420973994.

Key clinical point: Chronic migraine patients who achieved early reversal to episodic migraine are likely to have persistent reversion with long-term erenumab treatment.

Major finding: At 12 weeks, 54.1% (95% confidence interval [CI], 46.6%-61.6%) of patients showed reversal to episodic migraine. Continued treatment with erenumab showed a long-term persistent reversion in 96.8% (95% CI, 91.1%-99.3%) of patients at week 64.

Study details: Data on 181 migraine patients come from a post hoc analysis of a 12-week randomized double-blind trial and a 52-week open-label extension.

Disclosures: The study was funded by Amgen Inc. The authors reported ties with various institutions and/or pharmaceutical companies. F Zhang, GA Rippon, S Cheng, and DD Mikol are employed by and own stock in Amgen.

Source: Lipton RB et al. Cephalalgia. 2020 Dec 3. doi: 10.1177/0333102420973994.

Key clinical point: Rates of ipsilateral breast tumor recurrence (IBTR) in patients with early breast cancer treated with lumpectomy and intraoperative radiation therapy (IORT) are higher than that reported in TARGIT-A trial and with traditional whole-breast radiation therapy or other forms of partial breast irradiation.

Major finding: At 5 years, the absolute IBTR rate was 6.6% for the entire IORT-treated cohort. IBTR rate was 8% and 1.2% for patients who received primary IORT and unintentional IORT boost treatment, respectively. No recurrences were observed in the delayed IORT or intentional-boost IORT cohorts.

Study details: Findings from an updated analysis of TARGIT-R with a 5-year follow-up assessment in 667 patients with early-stage breast cancer who underwent lumpectomy and IORT.

Disclosures: The study did not receive any funding. The lead author did not report any conflicts of interest, but some co-investigators reported relationships with various pharmaceutical companies.

Source: Valente SA et al. Ann Surg Oncol. 2021 Jan 12. doi: 10.1245/s10434-020-09432-3.

Key clinical point: Rates of ipsilateral breast tumor recurrence (IBTR) in patients with early breast cancer treated with lumpectomy and intraoperative radiation therapy (IORT) are higher than that reported in TARGIT-A trial and with traditional whole-breast radiation therapy or other forms of partial breast irradiation.

Major finding: At 5 years, the absolute IBTR rate was 6.6% for the entire IORT-treated cohort. IBTR rate was 8% and 1.2% for patients who received primary IORT and unintentional IORT boost treatment, respectively. No recurrences were observed in the delayed IORT or intentional-boost IORT cohorts.

Study details: Findings from an updated analysis of TARGIT-R with a 5-year follow-up assessment in 667 patients with early-stage breast cancer who underwent lumpectomy and IORT.

Disclosures: The study did not receive any funding. The lead author did not report any conflicts of interest, but some co-investigators reported relationships with various pharmaceutical companies.

Source: Valente SA et al. Ann Surg Oncol. 2021 Jan 12. doi: 10.1245/s10434-020-09432-3.

Key clinical point: Rates of ipsilateral breast tumor recurrence (IBTR) in patients with early breast cancer treated with lumpectomy and intraoperative radiation therapy (IORT) are higher than that reported in TARGIT-A trial and with traditional whole-breast radiation therapy or other forms of partial breast irradiation.

Major finding: At 5 years, the absolute IBTR rate was 6.6% for the entire IORT-treated cohort. IBTR rate was 8% and 1.2% for patients who received primary IORT and unintentional IORT boost treatment, respectively. No recurrences were observed in the delayed IORT or intentional-boost IORT cohorts.

Study details: Findings from an updated analysis of TARGIT-R with a 5-year follow-up assessment in 667 patients with early-stage breast cancer who underwent lumpectomy and IORT.

Disclosures: The study did not receive any funding. The lead author did not report any conflicts of interest, but some co-investigators reported relationships with various pharmaceutical companies.

Source: Valente SA et al. Ann Surg Oncol. 2021 Jan 12. doi: 10.1245/s10434-020-09432-3.

Key clinical point: Low-dose capecitabine maintenance therapy for 1 year after standard adjuvant chemotherapy resulted in significantly better disease-free survival (DFS) but not overall survival (OS) in women with early triple-negative breast cancer (TNBC).

Major finding: Capecitabine vs. observation group had significantly higher estimated 5-year DFS (82.8% vs. 73%; hazard ratio [HR], 0.64; P = .03) and distant DFS (85.8% vs. 75.8%; HR, 0.60; P = .02) but not OS (85.5% vs. 81.3%; HR, 0.75; P = .22). Hand-foot syndrome (45.2%) was the most common adverse event in the capecitabine group, with 7.7% of patients experiencing grade 3 event.

Study details: Phase 3 SYSUCC-001 trial randomly allocated 434 women with early TNBC to receive either low-dose capecitabine maintenance (n=221) or undergo observation (n=213) within 4 weeks after completion of standard adjuvant chemotherapy.

Disclosures: This study was funded by Sun Yat-sen University, the National Natural Science Foundation of China, Natural Science Foundation of Guangdong Province, and F. Hoffmann-La Roche Ltd. The study investigators did not report any conflict of interest.

Source: Wang X et al. JAMA. 2021 Jan 5. doi: 10.1001/jama.2020.23370.

Key clinical point: Low-dose capecitabine maintenance therapy for 1 year after standard adjuvant chemotherapy resulted in significantly better disease-free survival (DFS) but not overall survival (OS) in women with early triple-negative breast cancer (TNBC).

Major finding: Capecitabine vs. observation group had significantly higher estimated 5-year DFS (82.8% vs. 73%; hazard ratio [HR], 0.64; P = .03) and distant DFS (85.8% vs. 75.8%; HR, 0.60; P = .02) but not OS (85.5% vs. 81.3%; HR, 0.75; P = .22). Hand-foot syndrome (45.2%) was the most common adverse event in the capecitabine group, with 7.7% of patients experiencing grade 3 event.

Study details: Phase 3 SYSUCC-001 trial randomly allocated 434 women with early TNBC to receive either low-dose capecitabine maintenance (n=221) or undergo observation (n=213) within 4 weeks after completion of standard adjuvant chemotherapy.

Disclosures: This study was funded by Sun Yat-sen University, the National Natural Science Foundation of China, Natural Science Foundation of Guangdong Province, and F. Hoffmann-La Roche Ltd. The study investigators did not report any conflict of interest.

Source: Wang X et al. JAMA. 2021 Jan 5. doi: 10.1001/jama.2020.23370.

Key clinical point: Low-dose capecitabine maintenance therapy for 1 year after standard adjuvant chemotherapy resulted in significantly better disease-free survival (DFS) but not overall survival (OS) in women with early triple-negative breast cancer (TNBC).

Major finding: Capecitabine vs. observation group had significantly higher estimated 5-year DFS (82.8% vs. 73%; hazard ratio [HR], 0.64; P = .03) and distant DFS (85.8% vs. 75.8%; HR, 0.60; P = .02) but not OS (85.5% vs. 81.3%; HR, 0.75; P = .22). Hand-foot syndrome (45.2%) was the most common adverse event in the capecitabine group, with 7.7% of patients experiencing grade 3 event.

Study details: Phase 3 SYSUCC-001 trial randomly allocated 434 women with early TNBC to receive either low-dose capecitabine maintenance (n=221) or undergo observation (n=213) within 4 weeks after completion of standard adjuvant chemotherapy.

Disclosures: This study was funded by Sun Yat-sen University, the National Natural Science Foundation of China, Natural Science Foundation of Guangdong Province, and F. Hoffmann-La Roche Ltd. The study investigators did not report any conflict of interest.

Source: Wang X et al. JAMA. 2021 Jan 5. doi: 10.1001/jama.2020.23370.

Key clinical point: In patients with human epidermal growth factor receptor 2 (HER2)-positive metastatic breast cancer (MBC) treated with pertuzumab, trastuzumab, and docetaxel, rash during treatment and diarrhea after docetaxel discontinuation are predictors of survival outcomes.

Major finding: Development of rash during docetaxel administration predicted progression-free survival (PFS; adjusted hazard ratio [aHR], 0.71; P = .006) and overall survival (OS; aHR, 0.66; P = .01). Development of rash after docetaxel interruption also predicted PFS (aHR, 0.55; P = .002) and OS (aHR, 0.52; P = .018). Diarrhea was prognostic only for PFS (aHR, 0.65; P = .011) after docetaxel discontinuation.

Study details: The data come from a retrospective analysis of CLEOPATRA trial involving 777 patients with de novo or recurrent advanced HER2-positive MBC treated with docetaxel plus trastuzumab with or without pertuzumab.

Disclosures: No specific funding source was identified. Some authors reported receiving travel grants, consultation fees, and honoraria from various pharmaceutical companies.

Source: Ferreira AR et al. Eur J Cancer. 2020 Dec 31. doi: 10.1016/j.ejca.2020.11.023.

Key clinical point: In patients with human epidermal growth factor receptor 2 (HER2)-positive metastatic breast cancer (MBC) treated with pertuzumab, trastuzumab, and docetaxel, rash during treatment and diarrhea after docetaxel discontinuation are predictors of survival outcomes.

Major finding: Development of rash during docetaxel administration predicted progression-free survival (PFS; adjusted hazard ratio [aHR], 0.71; P = .006) and overall survival (OS; aHR, 0.66; P = .01). Development of rash after docetaxel interruption also predicted PFS (aHR, 0.55; P = .002) and OS (aHR, 0.52; P = .018). Diarrhea was prognostic only for PFS (aHR, 0.65; P = .011) after docetaxel discontinuation.

Study details: The data come from a retrospective analysis of CLEOPATRA trial involving 777 patients with de novo or recurrent advanced HER2-positive MBC treated with docetaxel plus trastuzumab with or without pertuzumab.

Disclosures: No specific funding source was identified. Some authors reported receiving travel grants, consultation fees, and honoraria from various pharmaceutical companies.

Source: Ferreira AR et al. Eur J Cancer. 2020 Dec 31. doi: 10.1016/j.ejca.2020.11.023.

Key clinical point: In patients with human epidermal growth factor receptor 2 (HER2)-positive metastatic breast cancer (MBC) treated with pertuzumab, trastuzumab, and docetaxel, rash during treatment and diarrhea after docetaxel discontinuation are predictors of survival outcomes.

Major finding: Development of rash during docetaxel administration predicted progression-free survival (PFS; adjusted hazard ratio [aHR], 0.71; P = .006) and overall survival (OS; aHR, 0.66; P = .01). Development of rash after docetaxel interruption also predicted PFS (aHR, 0.55; P = .002) and OS (aHR, 0.52; P = .018). Diarrhea was prognostic only for PFS (aHR, 0.65; P = .011) after docetaxel discontinuation.

Study details: The data come from a retrospective analysis of CLEOPATRA trial involving 777 patients with de novo or recurrent advanced HER2-positive MBC treated with docetaxel plus trastuzumab with or without pertuzumab.

Disclosures: No specific funding source was identified. Some authors reported receiving travel grants, consultation fees, and honoraria from various pharmaceutical companies.

Source: Ferreira AR et al. Eur J Cancer. 2020 Dec 31. doi: 10.1016/j.ejca.2020.11.023.

Key clinical point: Hormone therapy (HT) use and benign breast disease (BBD) history are more strongly associated with a higher risk for symptom-detected breast cancer (sym-BC), whereas alcohol intake and obesity are associated with an increased screen-detected BC (scrn-BC) risk.

Major finding: Factors more strongly associated with higher sym-BC vs. scrn-BC risk were estrogen HT use (adjusted hazard ratio [aHR], 1.40; 95% confidence interval [CI], 1.15-1.71 vs. aHR, 0.95; 95% CI, 0.83-1.09), combined HT use (aHR, 2.07; 95% CI, 1.72-2.48 vs. aHR, 1.45; 95% CI, 1.27-1.65), and BBD history (aHR, 95% CI, 1.85; 1.64-2.08 vs. aHR, 1.43; 95% CI, 1.33-1.55). Factors more strongly associated with scrn-BC vs. sym-BC risk were alcohol intake (per drink aHR, 1.28; 95% CI, 1.18-1.39 vs. aHR, 1.07; 95% CI, 0.94-1.22) and body mass index ³30 kg/m² (aHR, 1.22; 95% CI, 1.01-1.48 vs. aHR, 0.76; 95% CI, 0.56-1.01).

Study details: The data come from an analysis of 77,206 women followed-up for a median of 14.8 years. BC was detected through screening or by symptoms in 2,711 and 1,281 women, respectively.

Disclosures: The study was supported by the American Cancer Society. The authors declared no conflicts of interest.

Source: Gaudet MM et al. Breast Cancer Res Treat. 2021 Jan 4. doi: 10.1007/s10549-020-06025-2.

Key clinical point: Hormone therapy (HT) use and benign breast disease (BBD) history are more strongly associated with a higher risk for symptom-detected breast cancer (sym-BC), whereas alcohol intake and obesity are associated with an increased screen-detected BC (scrn-BC) risk.

Major finding: Factors more strongly associated with higher sym-BC vs. scrn-BC risk were estrogen HT use (adjusted hazard ratio [aHR], 1.40; 95% confidence interval [CI], 1.15-1.71 vs. aHR, 0.95; 95% CI, 0.83-1.09), combined HT use (aHR, 2.07; 95% CI, 1.72-2.48 vs. aHR, 1.45; 95% CI, 1.27-1.65), and BBD history (aHR, 95% CI, 1.85; 1.64-2.08 vs. aHR, 1.43; 95% CI, 1.33-1.55). Factors more strongly associated with scrn-BC vs. sym-BC risk were alcohol intake (per drink aHR, 1.28; 95% CI, 1.18-1.39 vs. aHR, 1.07; 95% CI, 0.94-1.22) and body mass index ³30 kg/m² (aHR, 1.22; 95% CI, 1.01-1.48 vs. aHR, 0.76; 95% CI, 0.56-1.01).

Study details: The data come from an analysis of 77,206 women followed-up for a median of 14.8 years. BC was detected through screening or by symptoms in 2,711 and 1,281 women, respectively.

Disclosures: The study was supported by the American Cancer Society. The authors declared no conflicts of interest.

Source: Gaudet MM et al. Breast Cancer Res Treat. 2021 Jan 4. doi: 10.1007/s10549-020-06025-2.

Key clinical point: Hormone therapy (HT) use and benign breast disease (BBD) history are more strongly associated with a higher risk for symptom-detected breast cancer (sym-BC), whereas alcohol intake and obesity are associated with an increased screen-detected BC (scrn-BC) risk.

Major finding: Factors more strongly associated with higher sym-BC vs. scrn-BC risk were estrogen HT use (adjusted hazard ratio [aHR], 1.40; 95% confidence interval [CI], 1.15-1.71 vs. aHR, 0.95; 95% CI, 0.83-1.09), combined HT use (aHR, 2.07; 95% CI, 1.72-2.48 vs. aHR, 1.45; 95% CI, 1.27-1.65), and BBD history (aHR, 95% CI, 1.85; 1.64-2.08 vs. aHR, 1.43; 95% CI, 1.33-1.55). Factors more strongly associated with scrn-BC vs. sym-BC risk were alcohol intake (per drink aHR, 1.28; 95% CI, 1.18-1.39 vs. aHR, 1.07; 95% CI, 0.94-1.22) and body mass index ³30 kg/m² (aHR, 1.22; 95% CI, 1.01-1.48 vs. aHR, 0.76; 95% CI, 0.56-1.01).

Study details: The data come from an analysis of 77,206 women followed-up for a median of 14.8 years. BC was detected through screening or by symptoms in 2,711 and 1,281 women, respectively.

Disclosures: The study was supported by the American Cancer Society. The authors declared no conflicts of interest.

Source: Gaudet MM et al. Breast Cancer Res Treat. 2021 Jan 4. doi: 10.1007/s10549-020-06025-2.