Dear Editor,
We read with great interest the article by Bergstra et al., which demonstrated that residence at lower latitudes is associated with younger age of onset of rheumatoid arthritis (RA).[1] We are impressed by the in-depth explanation of how the level of socioeconomic welfare in different countries can contribute the results.[1] This contribution provides new insights into disease development and highlights the need for improving public health policies in many developing countries.
Geographical distribution, including latitude and altitude, is an important factor in several diseases. Previous studies revealed an association between geographical distribution and age of onset and severity of diseases such as systemic sclerosis and systemic lupus erythematosus.[2, 3] Furthermore, regional climate differences in similar latitude can lead different immune status in autoimmune disease.[4] The climate is influenced not only latitude or altitude but also other factors, such as ocean currents, prevailing winds, and vegetation cover. In previous studies, there is a possible association between climate factors and disease activity.[5, 6] On the other hand, for the evaluation of country-level factors, I thought that some index is more appropriate for the association survey, such as the human development index (HDI), the multidimensional poverty index (MPI), the social progress index (SPI).[7, 8]
In this study, the locations of each healthcare facility are cl...
Dear Editor,
We read with great interest the article by Bergstra et al., which demonstrated that residence at lower latitudes is associated with younger age of onset of rheumatoid arthritis (RA).[1] We are impressed by the in-depth explanation of how the level of socioeconomic welfare in different countries can contribute the results.[1] This contribution provides new insights into disease development and highlights the need for improving public health policies in many developing countries.
Geographical distribution, including latitude and altitude, is an important factor in several diseases. Previous studies revealed an association between geographical distribution and age of onset and severity of diseases such as systemic sclerosis and systemic lupus erythematosus.[2, 3] Furthermore, regional climate differences in similar latitude can lead different immune status in autoimmune disease.[4] The climate is influenced not only latitude or altitude but also other factors, such as ocean currents, prevailing winds, and vegetation cover. In previous studies, there is a possible association between climate factors and disease activity.[5, 6] On the other hand, for the evaluation of country-level factors, I thought that some index is more appropriate for the association survey, such as the human development index (HDI), the multidimensional poverty index (MPI), the social progress index (SPI).[7, 8]
In this study, the locations of each healthcare facility are clearly identified, allowing us to obtain clear information of climate classification and country-level socioeconomic or development index. Due to this strength, we have the following two suggestions:
First, we thought that the possible reasonable explanation that negligible results within the same country are that the correlation is not only due to the latitudes but also other climate factors, thus we suggest conducting correlation analysis using not only latitude but also considering climate classification, such as Köppen climate classification, and specific factors such as temperature variations and humidity to investigate the correlation between these factors and the onset age of RA if data available. This comprehensive analysis will help to further elucidate potential discrepancies and improve the accuracy of the correlation.
Second, considering the multidimensional perspective on socioeconomic status, HDI may be better for presentation than gross domestic product (GDP). Previous studies demonstrated the association between malignancy mortality, fatigue severity in RA patients and HDI, and we believed that such similar index is more appropriate to represent the socio-economic status of the country.[9, 10] Thus, I suggest that HDI, MPI, SPI or other similar index is executed to achieve a more accurate assessment of factors at the country level.
In conclusion, this article leads to a great contribution that latitude is a remarkable associated factor in the onset of age in rheumatoid arthritis, this highlighting the strong association between epidemiology and geography. To solve potential discrepancies and urge improvement of accuracy, we provide the suggestion mentioned in the article.
References
1. Bergstra, S.A., et al., Country-level socioeconomic status relates geographical latitude to the onset of RA: a worldwide cross-sectional analysis in the METEOR registry. Ann Rheum Dis, 2023.
2. Qian, G., et al., Systemic lupus erythematosus patients in the low-latitude plateau of China: altitudinal influences. Lupus, 2014. 23(14): p. 1537-1545.
3. Walker, U.A., et al., Geographical variation of disease manifestations in systemic sclerosis: a report from the EULAR Scleroderma Trials and Research (EUSTAR) group database. Ann Rheum Dis, 2009. 68(6): p. 856-62.
4. Kondo, Y., et al., Effect of climatic environment on immunological features of rheumatoid arthritis. Sci Rep, 2023. 13(1): p. 1304.
5. Rozin, A., A. Balbir-Gurman, and D. Schapira, Seasonal distribution of relapse onset in rheumatoid arthritis and spondyloarthropathy: the possible effect of the solar factor. Clinical and experimental rheumatology, 2003. 21(2): p. 161-170.
6. Yang, J., et al., Seasonal distribution of systemic lupus erythematosus activity and its correlation with climate factors. Rheumatol Int, 2012. 32(8): p. 2393-9.
7. Behera, D.K., Measuring socio-economic progress in India: issues and challenges. Revista Galega de Economia, 2016. 25(2): p. 117-132.
8. Kar, M., New Indicators and Measurement Methods for Welfare in the Global Economy Era, in Redefining Global Economic Thinking for the Welfare of Society. 2022, IGI Global. p. 8-32.
9. Hifinger, M., et al., In rheumatoid arthritis, country of residence has an important influence on fatigue: results from the multinational COMORA study. Rheumatology (Oxford), 2016. 55(4): p. 735-44.
10. Razi, S., et al., The incidence and mortality of ovarian cancer and their relationship with the Human Development Index in Asia. Ecancermedicalscience, 2016. 10: p. 628.
Dear Editor,
We read with great interest the paper published by Perez-Garcia et al [1] on ‘Is methotrexate safe for men with an immune-mediated inflammatory disease and an active desire to become a father? Results of a prospective cohort study (iFAME-MTX)’. The authors bring original data to address an important and controversial topic regarding the use of MTX in men with immune-mediated inflammatory diseases (IMIDs) who desire to father a child. They prospectively evaluated the testicular toxicity profile of MTX through quantitative and qualitative study of the sperm from treated patients. The strength of this study lies in its prospective design and comprehensive evaluation of various fertility markers in men exposed to MTX such as semen parameters, reproductive hormone levels, sperm DNA fragmentation index, and the presence of MTX-polyglutamates (MTX-PG) in spermatozoa. The author included three distinct groups, namely MTX-starters, MTX-chronic, and healthy controls, to provide a comparative analysis and highlight the impact of both short-term and long-term MTX exposure on male fertility. They observe that MTX exposure in men with IMIDs does not significantly affect conventional semen parameters, reproductive endocrine markers, or sperm DNA fragmentation index. These results provide reassurance regarding the start or continuation of MTX therapy in men planning to father a child, which align with the recent recommendations from the American College of Rheumatology...
Dear Editor,
We read with great interest the paper published by Perez-Garcia et al [1] on ‘Is methotrexate safe for men with an immune-mediated inflammatory disease and an active desire to become a father? Results of a prospective cohort study (iFAME-MTX)’. The authors bring original data to address an important and controversial topic regarding the use of MTX in men with immune-mediated inflammatory diseases (IMIDs) who desire to father a child. They prospectively evaluated the testicular toxicity profile of MTX through quantitative and qualitative study of the sperm from treated patients. The strength of this study lies in its prospective design and comprehensive evaluation of various fertility markers in men exposed to MTX such as semen parameters, reproductive hormone levels, sperm DNA fragmentation index, and the presence of MTX-polyglutamates (MTX-PG) in spermatozoa. The author included three distinct groups, namely MTX-starters, MTX-chronic, and healthy controls, to provide a comparative analysis and highlight the impact of both short-term and long-term MTX exposure on male fertility. They observe that MTX exposure in men with IMIDs does not significantly affect conventional semen parameters, reproductive endocrine markers, or sperm DNA fragmentation index. These results provide reassurance regarding the start or continuation of MTX therapy in men planning to father a child, which align with the recent recommendations from the American College of Rheumatology and the British Society for Rheumatology [2]. While this study provides significant insights, it is important to acknowledge some limitations, and mitigate the authors conclusions.
First, the sample size, particularly in the MTX-chronic group, is small (n = 5), which may affect the generalizability of the findings and inadequately represent the diversity of IMID patients receiving chronic MTX therapy. Importantly, small groups lead to insufficient statistical powers and an inability to demonstrate statistical differences. As an example, median spermatozoa concentration is 37x106/mL of sperm in chronic MTX users, whereas it is 60x106/mL in healthy donors or 57x106/mL in IMIDs patients not treated. These numerical differences (38% and 35% decreases, respectively) may reach statistical significance with bigger groups. The study did not provide detailed information on the MTX dosage and treatment duration for the participants. Chronic MTX users may have been on different doses and treatment durations, leading to varying cumulative exposure levels. This information is crucial as the effects on fertility may depend on the total amount of MTX received over time. Without considering cumulative exposure, the conclusion may not accurately reflect the potential impact of long-term MTX use on fertility.
Second, the authors use biological surrogates (concentration of MTX-PG, spermatozoa quantity and morphology, DNA fragmentation index) to evaluate MTX impact on semen. The use of these biological surrogates, although reassuring, cannot replace the evaluation of clinically relevant endpoints that are live birth, birth defect, and developmental abnormality of the child. While MTX-PG is relatively low in spermatozoa, it is high in the seminal fluid, which could impact the newly fertilized egg. A recent meta-analysis graded the evidence on the impact paternal exposure of MTX from “low” to “very low” [3], and a small cohort study has reported birth defect after paternal exposure [4]. Therefore, although the current study brings forward important and reassuring data, we believe that the study abstract conclusion stating that MTX therapy can be safely started or continued in men with a wish to conceive is incompletely supported by the authors data and the current literature. The precautionary principle should be used while good quality clinical data are not available to keep our oath, “Primum non nocere”.
REFERENCES
1 Perez-Garcia LF, Röder E, Krijthe BP, et al. Is methotrexate safe for men with an immune-mediated inflammatory disease and an active desire to become a father? Results of a prospective cohort study (iFAME-MTX). Ann Rheum Dis Published Online First: 1 June 2023. doi:10.1136/ard-2023-224032
2 Russell MD, Dey M, Flint J, et al. British Society for Rheumatology guideline on prescribing drugs in pregnancy and breastfeeding: immunomodulatory anti-rheumatic drugs and corticosteroids. Rheumatology 2023;62:e48–88. doi:10.1093/rheumatology/keac551
3 Mouyis M, Flint JD, Giles IP. Safety of anti-rheumatic drugs in men trying to conceive: A systematic review and analysis of published evidence. Semin Arthritis Rheum 2019;48:911–20. doi:10.1016/j.semarthrit.2018.07.011
4 Saougou I, Markatseli TE, Papagoras C, et al. Fertility in male patients with seronegative spondyloarthropathies treated with infliximab. Joint Bone Spine 2013;80:34–7. doi:10.1016/j.jbspin.2012.03.004
Dear Editors,
I read with great interest the report by Bergstra et al. (2023) on a cross-sectional analysis of rheumatoid arthritis patients from the worldwide METEOR registry1. The study is commendable for its sample size and geographical distribution of rheumatoid patients from 17 countries and 93 hospitals. Latitude has been defined as a potential underlying associative factor of early onset of rheumatoid arthritis among individuals living near the equator, which could be explained with socioeconomic status and accessibility to social welfare.
I have a few questions regarding the outcome measures and patient distribution. First, from which 17 countries were patients included, and how many patients were presented from each country along with the corresponding hospitals' latitudes? Second, were the included hospitals and countries were stratified to lower, middle, and higher latitudes, considering the disparities in socioeconomic status among low- and middle-income (LCMI) countries? These LCMI countries often have wider socioeconomic gaps, creating a complex scenario where diagnostic and treatment modalities for rheumatoid arthritis, as well as healthcare awareness, may not be uniform across rural and urban settings. This could potentially impact patient representation in the study. Third, were the included patients from low and middle socioeconomic statuses primarily presented from metro cities? It is worth noting that a majority of the patients in thi...
Dear Editors,
I read with great interest the report by Bergstra et al. (2023) on a cross-sectional analysis of rheumatoid arthritis patients from the worldwide METEOR registry1. The study is commendable for its sample size and geographical distribution of rheumatoid patients from 17 countries and 93 hospitals. Latitude has been defined as a potential underlying associative factor of early onset of rheumatoid arthritis among individuals living near the equator, which could be explained with socioeconomic status and accessibility to social welfare.
I have a few questions regarding the outcome measures and patient distribution. First, from which 17 countries were patients included, and how many patients were presented from each country along with the corresponding hospitals' latitudes? Second, were the included hospitals and countries were stratified to lower, middle, and higher latitudes, considering the disparities in socioeconomic status among low- and middle-income (LCMI) countries? These LCMI countries often have wider socioeconomic gaps, creating a complex scenario where diagnostic and treatment modalities for rheumatoid arthritis, as well as healthcare awareness, may not be uniform across rural and urban settings. This could potentially impact patient representation in the study. Third, were the included patients from low and middle socioeconomic statuses primarily presented from metro cities? It is worth noting that a majority of the patients in this study are from India, with unequal distribution of hospitals with rheumatoid care facilities.
Fourth, how participating hospitals and included patients in India and countries nearing to equators were stratified by latitude beyond the Tropic of Cancer? Apart from latitude, air pollution and vitamin D deficiency are also considered potent risk factors for the onset of autoimmune diseases. The Italian cohort (n = 81,363) demonstrates that long-term exposure to air pollution increases the risk of autoimmune disease. Specifically, exposure to particulate matter of 10PM increases the risk of rheumatoid arthritis, while exposure to particles of size 2.5PM or less increases the risk of multiple autoimmune diseases2. The relationship between vitamin D deficiency and the prevalence of autoimmune diseases, including rheumatoid arthritis, remains debatable3. If vitamin D deficiency is not a surrogate marker and a plausible cause of rheumatoid arthritis, environmental triggers should be stratified by latitude.
It is important to note that latitude and socioeconomic status are not direct causes but rather indirect factors contributing to autoimmune diseases and rheumatoid arthritis. Considering the influence of air pollution, sunlight exposure, access to and awareness of healthcare services, demographics, age, sex, and genetic makeup, it is intriguing to consider latitude as a risk factor for rheumatoid arthritis and other autoimmune diseases1. Therefore, large-scale cohorts investigating risk factors for autoimmune diseases should be widely stratified.
References
1 Bergstra, S. A. et al. Country-level socioeconomic status relates geographical latitude to the onset of RA: a worldwide cross-sectional analysis in the METEOR registry. Ann Rheum Dis, doi:10.1136/ard-2023-224080 (2023).
2 Adami, G. et al. Association between long-term exposure to air pollution and immune-mediated diseases: a population-based cohort study. RMD Open 8, doi:10.1136/rmdopen-2021-002055 (2022).
3 Clasen, J. L., Cole, R., Aune, D., Sellon, E. & Heath, A. K. Vitamin D status and risk of rheumatoid arthritis: systematic review and meta-analysis. BMC Rheumatol 7, 3, doi:10.1186/s41927-023-00325-y (2023).
We would like to thank dr. Mahla for his interest in our paper and respond to the questions asked. The first question relates to the included patients, hospitals and countries. We were able to analyze 37,981 patients from 93 hospitals in 17 countries. Information on the included countries, and the number of patients and hospitals per country is provided in the original manuscript in table 1. Hospital latitudes are displayed in figure 1.
As we mentioned in the discussion section, one of the main limitations of our paper is that the number of patients per hospital/country strongly differs and it is impossible to determine to what extent these patients are representative of all patients with RA in a country. Especially in lower- and middle income economies, healthcare disparities may be significant and we were unable to include individual patients’ socioeconomic data . As an example, dr. Mahla mentions India, the country with the largest contribution of patients to our study. We performed a sensitivity analysis excluding India, and found very similar results.
Most countries were represented by only a limited number of hospitals. These were primarily located in cities. Whereas especially in lower income countries these hospitals serve a large area, this may have led to an underrepresentation of patients living in rural areas. We can only speculate on how this may have influenced our findings, which may not only depend on regional and personal socioeconomic circumst...
We would like to thank dr. Mahla for his interest in our paper and respond to the questions asked. The first question relates to the included patients, hospitals and countries. We were able to analyze 37,981 patients from 93 hospitals in 17 countries. Information on the included countries, and the number of patients and hospitals per country is provided in the original manuscript in table 1. Hospital latitudes are displayed in figure 1.
As we mentioned in the discussion section, one of the main limitations of our paper is that the number of patients per hospital/country strongly differs and it is impossible to determine to what extent these patients are representative of all patients with RA in a country. Especially in lower- and middle income economies, healthcare disparities may be significant and we were unable to include individual patients’ socioeconomic data . As an example, dr. Mahla mentions India, the country with the largest contribution of patients to our study. We performed a sensitivity analysis excluding India, and found very similar results.
Most countries were represented by only a limited number of hospitals. These were primarily located in cities. Whereas especially in lower income countries these hospitals serve a large area, this may have led to an underrepresentation of patients living in rural areas. We can only speculate on how this may have influenced our findings, which may not only depend on regional and personal socioeconomic circumstances and access to healthcare, but also on intangible factors that we were unable to measure.
Dr. Mahla gives a few examples: healthcare awareness, air pollution, vitamin D deficiency and genetic make-up. Unfortunately, reliable data on these factors were not available at either the patient or hospital level. Other factors such as demographics, age and sex were included in our model but did not satisfactorily explain the association between latitude and age at disease onset.
The final question relates to a stratification of latitude, for countries nearing the equator and for countries with a different socioeconomic status. We used a structural equation model, to assess the association between latitude and age at onset. In this model, we treated latitude as a continuous variable. Only some of the patient factors (ACPA, rheumatoid factor, ever smoking and sex) were included as dichotomous variables. All other variables, including indicators of country level socioeconomic status (e.g. GDP per capita) were treated as continuous variables. No stratified analyses were performed. We also refer to figures 2 and 3 in our original manuscript.
Only 8 hospitals are located in the southern hemisphere, at negative latitude. Since we observed no interaction between latitude and hemisphere (i.e. the association between latitude and age at onset was similar at the northern and southern hemisphere), all negative latitudes were artificially converted into positive latitudes.
We hope this response will contribute to further clarification of our results and conclusion.
The article says that "At the end of the inclusion consultation, rheumatologists initiated a first targeted therapy: 8.6% by bDMARDs etc.". That number is actually wrong, probably because a typing issue. The actual number would be 89.6%.
Recently, we have been very interested in the article: "Increased risk of osteoarthritis in patients with atopic disease" by Baker et al. [1]. In this article, the author analyzed the incidence of OA in patients with atopic disease in two databases, suggesting that the immune response to atopic illness contributes to the increased incidence of OA. This study provides a new direction for immunotherapy against OA.
Overall, the authors' research design is meticulous. In two different databases, the author divided the idiopathic disease exposure group and the non-exposed group. The author strictly stipulated the sequence of idiopathic diseases and OA to ensure the impact of idiopathic diseases on OA. Matching was performed for age, sex, race/ethnicity, education level, Charlson comorbidity score, follow-up time, and clinic visit frequency, effectively reducing these factors influence. Furthermore, OA risk was reduced after adjusting for BMI in the STARR cohort, confirming that BMI is an independent confounding factor. However, although the authors analyzed various limitations of this study in the Discussion section, there are still some unmentioned issues worth considering.
First, to reduce the influence of confounding factors, the author matched and adjusted multiple independent factors, including age, gender, and BMI. In addition, the impact of factors such as history of joint trauma and physical activity level were mentioned in the discussion sect...
Recently, we have been very interested in the article: "Increased risk of osteoarthritis in patients with atopic disease" by Baker et al. [1]. In this article, the author analyzed the incidence of OA in patients with atopic disease in two databases, suggesting that the immune response to atopic illness contributes to the increased incidence of OA. This study provides a new direction for immunotherapy against OA.
Overall, the authors' research design is meticulous. In two different databases, the author divided the idiopathic disease exposure group and the non-exposed group. The author strictly stipulated the sequence of idiopathic diseases and OA to ensure the impact of idiopathic diseases on OA. Matching was performed for age, sex, race/ethnicity, education level, Charlson comorbidity score, follow-up time, and clinic visit frequency, effectively reducing these factors influence. Furthermore, OA risk was reduced after adjusting for BMI in the STARR cohort, confirming that BMI is an independent confounding factor. However, although the authors analyzed various limitations of this study in the Discussion section, there are still some unmentioned issues worth considering.
First, to reduce the influence of confounding factors, the author matched and adjusted multiple independent factors, including age, gender, and BMI. In addition, the impact of factors such as history of joint trauma and physical activity level were mentioned in the discussion section. However, the pathogenesis of OA is complex. In addition to the factors mentioned in the article, the patient's smoking history, drinking history, and other living habits are essential factors affecting OA [2]. These confounding factors may affect the accuracy of the results. In another study, the author mentioned that in sedentary patients, the relationship between atopic dermatitis and osteoarthritis might be closer [3], which suggests that our occupation-related living habits also affect idiopathic diseases and the incidence of OA rate factor.
Second, recall bias is a problem that any retrospective study has to face. The author uses the International Classification of Disease (ICD) code 9 or 10 as the inclusion criteria for the disease. One of the problems here is the need for clinical experience in diagnosing the disease, and there may be differences in the results due to the diagnostic criteria of different ages and the diagnosis of other doctors.
Third, for idiopathic diseases, the authors discuss the contribution of idiopathic dermatitis and asthma to the increased incidence of OA. Idiopathic disease is an immune disease with elevated IgE levels. In addition to idiopathic dermatitis and asthma, joint diseases include hay fever [4]. The article does not discuss hay fever, which may include omitting objectives.
Also, we mentioned some suggestions. To exclude the influence of general lung diseases, the author compared the prevalence of OA in COPD patients and asthma patients, confirmed that COPD and other prevalent lung diseases do not affect the incidence of OA, and further confirmed that OA causes asthma through an immune response. We believe that the comparison between COPD patients and patients with non-idiopathic diseases can be increased, and it can be more directly explained that widespread lung diseases such as COPD do not affect the occurrence of OA. In this study, the impact of pre-existing idiopathic conditions on the event of OA was explored. In the future, whether patients with pre-existing OA will lead to the occurrence of idiopathic diseases is also a direction worthy of research.
We thank the authors again for their contributions and look forward to their responses.
Reference:
[1] M.C. Baker, K. Sheth, R. Lu, D. Lu, E.P. von Kaeppler, A. Bhat, D.T. Felson, W.H. Robinson, Increased risk of osteoarthritis in patients with atopic disease, Ann Rheum Dis 82(6) (2023) 866-872.
[2] J.M. Gwinnutt, M. Wieczorek, A. Balanescu, H.A. Bischoff-Ferrari, A. Boonen, G. Cavalli, S. de Souza, A. de Thurah, T.E. Dorner, R.H. Moe, P. Putrik, J. Rodriguez-Carrio, L. Silva-Fernandez, T. Stamm, K. Walker-Bone, J. Welling, M.I. Zlatkovic-Svenda, F. Guillemin, S.M.M. Verstappen, 2021 EULAR recommendations regarding lifestyle behaviours and work participation to prevent progression of rheumatic and musculoskeletal diseases, Ann Rheum Dis 82(1) (2023) 48-56.
[3] J. Liu, A. Martin, A. Thatiparthi, J.J. Wu, Association between atopic dermatitis and osteoarthritis among US adults in the 1999-2006 NHANES, J Eur Acad Dermatol Venereol 35(6) (2021) e375-e377.
[4] M. Rudwaleit, B. Andermann, R. Alten, H. Sorensen, J. Listing, A. Zink, J. Sieper, J. Braun, Atopic disorders in ankylosing spondylitis and rheumatoid arthritis, Ann Rheum Dis 61(11) (2002) 968-74.
We thank Dr Bitoun et al for their careful reading of our manuscript. We acknowledge that the small number of patients experiencing cancer progression in our study is a limitation as it reduced the number of covariates that could be analyzed. However, in individual unadjusted Cox models analyses we did not find that age, sex, cancer type, cancer stage, ICI type, ICI discontinuation, concomitant irAE or steroid duration were associated with progression free survival suggesting that inclusion of these covariates in a larger study might not have influenced our findings.
We do agree that arthritis disease activity is an important confounder of DMARD choice. However, in our cohort, arthritis grade at the time of presentation was similar across the groups (p=0.67). Unfortunately, other measures of disease activity such as DAS28 were not available in this retrospective study. In addition, to address potential bias arising from biologic DMARD-treated patients’ being treated earlier because of more severe arthritis, we included the time-dependent variable “time from ICI initiation to DMARD initiation” into our model (an analysis in which there was no censoring at the time of DMARD switch). With this covariate included, our results favoring methotrexate in comparison to a TNF inhibitor became if anything, stronger (HR 3.27, 95% CI 1.21-8.84, p=0.019). The two methods commonly used to account for immortal time bias are landmark models and time dependent covariates. We chos...
We thank Dr Bitoun et al for their careful reading of our manuscript. We acknowledge that the small number of patients experiencing cancer progression in our study is a limitation as it reduced the number of covariates that could be analyzed. However, in individual unadjusted Cox models analyses we did not find that age, sex, cancer type, cancer stage, ICI type, ICI discontinuation, concomitant irAE or steroid duration were associated with progression free survival suggesting that inclusion of these covariates in a larger study might not have influenced our findings.
We do agree that arthritis disease activity is an important confounder of DMARD choice. However, in our cohort, arthritis grade at the time of presentation was similar across the groups (p=0.67). Unfortunately, other measures of disease activity such as DAS28 were not available in this retrospective study. In addition, to address potential bias arising from biologic DMARD-treated patients’ being treated earlier because of more severe arthritis, we included the time-dependent variable “time from ICI initiation to DMARD initiation” into our model (an analysis in which there was no censoring at the time of DMARD switch). With this covariate included, our results favoring methotrexate in comparison to a TNF inhibitor became if anything, stronger (HR 3.27, 95% CI 1.21-8.84, p=0.019). The two methods commonly used to account for immortal time bias are landmark models and time dependent covariates. We chose to use time dependent covariates because landmark methods are dependent on the choice of landmark, which might be incorrect, and can be prone to bias.1
While clinical trial emulation could be very informative, we did not have a study population with similar enough pre-DMARD characteristics to accurately simulate a trial population. Of course, we also agree that a prospective clinical trial would ultimately best answer the important questions posed on how DMARDs may impact cancer progression while optimizing arthritis control and quality of life.
Finally, we do recognize that there are patients with very severe ICI arthritis who require fast acting treatment such as with a biologic DMARD, and we are not suggesting that effective therapy should be withheld in these cases. However, there may be other situations where there is clinical equipoise regarding DMARD choice, and in those situations our study suggests that methotrexate is a safer approach.
1Jones M, Fowler R. Immortal time bias in observational studies of time-to-event outcomes. J Crit Care. 2016 Dec; 36:195-199. PMID: 27546771.
We read with interest the article by Kristensen et al. based on primary ORAL Surveillance results (NCT02092467) enrolling rheumatoid arthritis (RA) patients aged ≥50 years and with ≥1 additional cardiovascular risk factor, who were treated with tofacitinib or tumor necrosis factor inhibitors [1,2]. According to the presence of age ≥65 years or smoking, two subpopulations (i.e., ’high-risk’ and ’low-risk’) were identified, among those receiving tofacitinib, with diverse relative event risk (i.e., malignancies, major cardiovascular events, myocardial infarction, venous thromboembolism, and all-cause death) [1].
One feature highlighted by Kristensen et al. is the increased risk of event in RA patients aged ≥ 65 years [1]. This elderly subpopulation comprises both elderly-onset RA patients, manifesting after the age of 60 years, and those patients diagnosed with RA early in life who naturally become members of this subset. The management of elderly RA patients may be challenging [3]. In fact, the typical high prevalence of comorbidities in these patients may be associated with a lower response to RA therapies and poor prognosis. Furthermore, the concomitant polypharmacy may make even more difficult the treatment of elderly patients due to increased likelihood of drug interactions, adverse events, and possible contraindications [3]. The increased frailty associated with older age may also complicate the management of such patients; this is a syndrome characterised by a de...
We read with interest the article by Kristensen et al. based on primary ORAL Surveillance results (NCT02092467) enrolling rheumatoid arthritis (RA) patients aged ≥50 years and with ≥1 additional cardiovascular risk factor, who were treated with tofacitinib or tumor necrosis factor inhibitors [1,2]. According to the presence of age ≥65 years or smoking, two subpopulations (i.e., ’high-risk’ and ’low-risk’) were identified, among those receiving tofacitinib, with diverse relative event risk (i.e., malignancies, major cardiovascular events, myocardial infarction, venous thromboembolism, and all-cause death) [1].
One feature highlighted by Kristensen et al. is the increased risk of event in RA patients aged ≥ 65 years [1]. This elderly subpopulation comprises both elderly-onset RA patients, manifesting after the age of 60 years, and those patients diagnosed with RA early in life who naturally become members of this subset. The management of elderly RA patients may be challenging [3]. In fact, the typical high prevalence of comorbidities in these patients may be associated with a lower response to RA therapies and poor prognosis. Furthermore, the concomitant polypharmacy may make even more difficult the treatment of elderly patients due to increased likelihood of drug interactions, adverse events, and possible contraindications [3]. The increased frailty associated with older age may also complicate the management of such patients; this is a syndrome characterised by a decrease of strength, reduced physiological function, and increased individual vulnerability [4]. In addition, elderly RA patients appear to have a diverse biological profile, at least partially, showing higher IL-6 and lower TNF levels in association with senescent T cells [5,6]. An impaired JAK/STAT pathway signalling has been also shown in immune cells of older patients [5,6]. Interestingly, a defective regulatory JAK/STAT cytokine signalling could be associated to a reduced response capacity of T cell subsets [5,6]. Thus, an immunological link could be suggested between rheumatoid inflammation in elderly patients and decreased immune cell responsiveness in JAK/STAT pathway. Although further studies are needed to clarify this issue, this immunological profile could partially explain an increased risk of the event in elderly RA patients treated with tofacitinib, described by Kristensen et al. [1].
Another feature highlighted by Kristensen et al. is the role of smoking habit in this context [1]. The exposure of lung to cigarette smoking is associated with increased oxidative stress, apoptosis, development of a proinflammatory state and autoantibody production in rheumatic diseases [7]. The effects of smoking habit may also stimulate the release of reactive oxygen species and reactive nitrogen species. Moreover, cigarette smoking may cause the release of some danger-associated molecular patterns, which in turn may lead to the activation of macrophages, neutrophils, and dendritic cells [7,8]. Furthermore, smoke exposure may enhance the proliferation of T lymphocytes [8]. It has been also reported that smokers may have a decreased NK cell-mediated lytic activity [8]. All these features could contribute to the increased risk of event of RA patients treated with tofacitinib, reported by Kristensen et al. [1].
All things considered, a better patient characterization, also considering the patient clinical characteristics, may be of crucial importance in tailoring RA treatment according to principles of precision medicine [9]. A more accurate patient profiling could indeed improve the clinical benefits simultaneously reducing the potential side effects of any treatment [9,10]. Therefore, a new approach may be suggested leading to a shift from the concept and practical challenge of ‘finding the patient for the trial’ to ‘finding the trial for the patient’ toward patient-centric approach [10].
In conclusion, all these observations about elderly patients and smoking habit may suggest further indications in accurately guiding a benefit/risk assessment about the treatment with tofacitinib in RA.
References:
1. Kristensen LE, Danese S, Yndestad A, et al. Identification of two tofacitinib subpopulations with different relative risk versus TNF inhibitors: an analysis of the open label, randomised controlled study ORAL Surveillance. Ann Rheum Dis 2023:ard-2022-223715. doi: 10.1136/ard-2022-223715. Epub ahead of print.
2. Ytterberg SR, Bhatt DL, Mikuls TR, et al. Cardiovascular and Cancer Risk with Tofacitinib in Rheumatoid Arthritis. N Engl J Med. 2022;386:316-326.
3. Serhal L, Lwin MN, Holroyd C, Edwards CJ. Rheumatoid arthritis in the elderly: Characteristics and treatment considerations. Autoimmun Rev. 2020;19:102528.
4. Khan KT, Hemati K, Donovan AL. Geriatric Physiology and the Frailty Syndrome. Anesthesiol Clin. 2019;37:453-474.
5. Samson LD, Engelfriet P, Verschuren WMM, et al. Impaired JAK-STAT pathway signaling in leukocytes of the frail elderly. Immun Ageing. 2022;19:5.
6. Shen-Orr SS, Furman D, Kidd BA, et al. Defective Signaling in the JAK-STAT Pathway Tracks with Chronic Inflammation and Cardiovascular Risk in Aging Humans. Cell Syst. 2016;3:374-384.e4.
7. Gwinnutt JM, Verstappen SM, Humphreys JH. The impact of lifestyle behaviours, physical activity and smoking on morbidity and mortality in patients with rheumatoid arthritis. Best Pract Res Clin Rheumatol. 2020;34:101562.
8. Arnson Y, Shoenfeld Y, Amital H. Effects of tobacco smoke on immunity, inflammation and autoimmunity. J Autoimmun. 2010;34:J258-65.
9. Giacomelli R, Afeltra A, Bartoloni E, et al. The growing role of precision medicine for the treatment of autoimmune diseases; results of a systematic review of literature and Experts' Consensus. Autoimmun Rev. 2021;20:102738.
10. Pitzalis C, Choy EHS, Buch MH. Transforming clinical trials in rheumatology: towards patient-centric precision medicine. Nat Rev Rheumatol. 2020;16:590-599.
Declarations:
-Ethics approval and consent to participate:
Not applicable.
-Consent for publication:
Not applicable, all the patients’ data are de-identified.
-Availability of data and material:
All data relevant to the study are included in the article.
-Competing interests:
The authors declare that they have no conflicts of interest for this work.
-Funding:
No funding.
-Authors' contributions:
All authors made substantial contributions to the conception or design of the work, the acquisition and interpretation of data. All authors contributed to the critical review and revision of the manuscript and approved the final version. All the authors agreed to be accountable for all aspects of the work.
We read with great interest the article by Smith et al [1], reinforcing the role of rituximab in the maintenance treatment of ANCA-associated vasculitis. This major study randomized the maintenance treatments of 170 GPA and MPA patients to rituximab (1000 mg every 4 months, up to month 20) and azathioprine (2 mg/kg/day). At the end of the month 24, relapse-free survival rate was 0.85 for the rituximab group (95% CI: 0.78–0.93) compared to 0.61 of the azathioprine group (95% CI: 0.51–0.73). Overall HR was 0.4 (95% CI: 0.27–0.61). There was no difference in rates of infection or hypogammaglobinemia, and rate of adverse events was found to be similar with previously published studies. Here we like to make two main points of interest.
First, after the last dose of rituximab at month 20, the relapse-free survival rate drops to 0.5 (95% CI: 0.40–0.63) in the study group, which is parallel to the azathioprine group (relapse-free survival rate: 0.22, 95% CI: 0.14–0.35) [1]. In the discussion, the authors comment on this by stating that ‘despite using a higher dose rituximab regimen, relapses still occurred’ and ‘together with increased risk of relapse after stopping rituximab and the associated safety risks illustrate the need for newer therapeutic agents for AAV.’ [1]. While we agree that newer therapeutic options are always welcome, data from this study underlines the efficacy of rituximab when it is used. An alternative of stopping rituximab therapy in two years may be th...
We read with great interest the article by Smith et al [1], reinforcing the role of rituximab in the maintenance treatment of ANCA-associated vasculitis. This major study randomized the maintenance treatments of 170 GPA and MPA patients to rituximab (1000 mg every 4 months, up to month 20) and azathioprine (2 mg/kg/day). At the end of the month 24, relapse-free survival rate was 0.85 for the rituximab group (95% CI: 0.78–0.93) compared to 0.61 of the azathioprine group (95% CI: 0.51–0.73). Overall HR was 0.4 (95% CI: 0.27–0.61). There was no difference in rates of infection or hypogammaglobinemia, and rate of adverse events was found to be similar with previously published studies. Here we like to make two main points of interest.
First, after the last dose of rituximab at month 20, the relapse-free survival rate drops to 0.5 (95% CI: 0.40–0.63) in the study group, which is parallel to the azathioprine group (relapse-free survival rate: 0.22, 95% CI: 0.14–0.35) [1]. In the discussion, the authors comment on this by stating that ‘despite using a higher dose rituximab regimen, relapses still occurred’ and ‘together with increased risk of relapse after stopping rituximab and the associated safety risks illustrate the need for newer therapeutic agents for AAV.’ [1]. While we agree that newer therapeutic options are always welcome, data from this study underlines the efficacy of rituximab when it is used. An alternative of stopping rituximab therapy in two years may be the continuation of rituximab at lower doses with increased intervals. Campaniello et al reported at EULAR 2022 congress that rituximab maintenance with “ultra low” (500–1000 mg per year) doses can be efficacious [2]. After remission induction, about %30 of the 83 ANCA-associated vasculitis patients received low dose rituximab therapy. Similar relapse rates were seen at month 24 (22.7% for the low dose group vs. 21.2% for the standard dose group). Infection rate was also lower for the low dose group (10.5% vs. 26.8%), albeit this did not reach statistical significance. In addition, the decision to continue rituximab beyond two years can be individualized, based on B-cell status as well as clinical response [3]
Second, we would like to inquire about the cumulative rituximab dose. The authors report that since minor relapses are included as part of the primary endpoint of relapse-free survival, this reflects the importance of minor relapses in cumulative treatment exposure. In addition, higher cumulative rituximab dose during maintenance (5000 mg) compared to MAINRITSAN study while 15% of rituximab group still having relapses is a sign of a patient subset with disease refractory to high dose rituximab [1]. The relationship of cumulative steroid and cyclophosphamide doses with adverse events is well known [4], however the same cannot be said of cumulative rituximab dose. Cumulative rituximab dose can be considered as an indirect measure of the length of rituximab treatment; in which case directly using the length of maintenance treatment with rituximab may be a more appropriate parameter to monitor. This is supported by the data presented in the study, as the cessation of rituximab at month 20 results in an increased rate of relapses.
References
1. Smith RM, Jones RB, Specks U, et al. Rituximab versus azathioprine for maintenance of remission for patients with ANCA-associated vasculitis and relapsing disease: an international randomised controlled trial. Ann Rheum Dis 2023;:ard-2022-223559. doi:10.1136/ard-2022-223559
2. Campaniello D, Treppo E, D’onofrio B, et al. Efficacy of Ultra-Low Dose Rituximab for Remission Maintenance Therapy in Anca-Associated Vasculitis. Ann Rheum Dis 2022;81:710–710. doi:10.1136/annrheumdis-2022-eular.4795
3. Arnold JE, Vital EMJ, Dass S, et al. P207 A proposal for individually tailored rituximab treatment based on clinical and B-cell biomarkers for remission maintenance in ANCA associated vasculitis. Rheumatology 2021;60:keab247.202. doi:10.1093/rheumatology/keab247.202
4. Hellmich B, Sanchez-Alamo B, Schirmer JH, et al. EULAR recommendations for the management of ANCA-associated vasculitis: 2022 update. Ann Rheum Dis Published Online First: 16 March 2023. doi:10.1136/ard-2022-223764
We read with interest the study recently published in ARD by Bass et al. This is the first retrospective study comparing methotrexate (MTX) to bDMARDs (TNF inhibitors (TNFi) and IL6 receptor inhibitors (IL6i)) for the treatment of immune checkpoint inhibitor (ICI)-associated arthritis. It concluded that bDMARDs were associated with more rapid arthritis control but also with a higher rate of cancer progression. Since it is the first study reporting such a comparison, it could become a reference and discourage rheumatologists to use bDMARDs for treating ICI-associated IA.
Even if the conclusions were cautious due to the retrospective nature of the study, a number of methodological weaknesses may challenge the confidence with those conclusions:
The number of events considered for comparing the progression free survival (PFS) between DMARDs is only 10, since 11 patients who progressed before DMARD introduction and 3 who were treated by different DMARDs were excluded from this analysis. With such a small number of events, regression models with many variables are at high risk of overfitting and sparse data bias[1].
The presence of immortal bias was identified, and there was an attempt to consider it. However, the way it was accounted for the analysis of PFS raises methodological concerns. Indeed, the index date to start follow-up for PFS was ICI initiation, meaning that patients were classified in each of the group (DMARDs vs bDMARDs), based on a criteria occu...
We read with interest the study recently published in ARD by Bass et al. This is the first retrospective study comparing methotrexate (MTX) to bDMARDs (TNF inhibitors (TNFi) and IL6 receptor inhibitors (IL6i)) for the treatment of immune checkpoint inhibitor (ICI)-associated arthritis. It concluded that bDMARDs were associated with more rapid arthritis control but also with a higher rate of cancer progression. Since it is the first study reporting such a comparison, it could become a reference and discourage rheumatologists to use bDMARDs for treating ICI-associated IA.
Even if the conclusions were cautious due to the retrospective nature of the study, a number of methodological weaknesses may challenge the confidence with those conclusions:
The number of events considered for comparing the progression free survival (PFS) between DMARDs is only 10, since 11 patients who progressed before DMARD introduction and 3 who were treated by different DMARDs were excluded from this analysis. With such a small number of events, regression models with many variables are at high risk of overfitting and sparse data bias[1].
The presence of immortal bias was identified, and there was an attempt to consider it. However, the way it was accounted for the analysis of PFS raises methodological concerns. Indeed, the index date to start follow-up for PFS was ICI initiation, meaning that patients were classified in each of the group (DMARDs vs bDMARDs), based on a criteria occurring after the start of follow-up. By construction, the event (progression or death) could not occur during the time between ICI start and introduction of the DMARD, and this time was actually different between groups. This may lead to so-called immortal time bias[2]. Interestingly, this immoral time was much longer in patients treated with MTX (486 days) than with TNFi (411 days) and IL6R (300 days) (p=0.02). This difference may be linked to the fact that the patients treated with bDMARDs were treated earlier because of a more severe arthritis. This is illustrated by the fact they received a higher maximum dose of glucocorticoid and a higher frequency of colitis for patients treated with TNFi. Thus, the unadjusted analysis may advantage the MTX group both because of a longer time from ICI initiation to DMARD initiation (immortal time) and a likely less severe IA. To account for the difference in time since ICI initiation, an adjusted analysis on this time was also performed. However, adjustment is not an adequate way to deal with immortal bias. Again, measuring the time between ICI initiation and DMARD initiation in each group implies that the DMARD chosen could affect that time, which is simply not possible. Therefore, these analyses cannot be interpreted.
A more relevant analysis would have been to analyse PFS since DMARD initiation, after matching or adjusting on the time from ICI initiation to DMARD initiation, and possibly other factors (however still limited by the small number of event). Alternatively, the study could rather be re-thought within the framework of target trial emulation[3,4]
Last, patients were treated in several sites having different habits in therapeutic strategies, with a clear centre effect in the use of drugs; for instance, 71% of patients treated at the MD Anderson received anti-IL-6R, but only 6% received MTX.
In conclusion, we appreciate the willingness of the authors to compare different therapeutic strategies for the treatment of ICI-associated IA. However, both immortal time and confounding biases complicated the interpretation of results. The adjustment on time from ICI initiation cannot correct this immortal bias, nor does it account for patients’ severity. Results might thus be interpreted with caution, and the statement that bDMARD is associated with higher risk of progression or death should not be interpreted as causality.
To solve this issue, we agree with the last sentence of the article, underlying that the best solution would be to run randomised controlled trials. In the absence of such a randomised trial, using large observational data to emulate those trials could also prove valuable. These studies should firstly compare steroids (considered as the gold standard for the treatment of ICI-associated IA) to one of these DMARDs individually and later compare one DMARD to another.
References :
1 Greenland S, Mansournia MA, Altman DG. Sparse data bias: a problem hiding in plain sight. BMJ 2016;352:i1981. doi:10.1136/bmj.i1981
2 Iudici M, Porcher R, Riveros C, et al. Time-dependent biases in observational studies of comparative effectiveness research in rheumatology. A methodological review. Ann Rheum Dis 2019;78:562–9. doi:10.1136/annrheumdis-2018-214544
3 Hernán MA, Robins JM. Using Big Data to Emulate a Target Trial When a Randomized Trial Is Not Available. Am J Epidemiol 2016;183:758–64. doi:10.1093/aje/kwv254
4 Matthews AA, Danaei G, Islam N, et al. Target trial emulation: applying principles of randomised trials to observational studies. BMJ 2022;378:e071108. doi:10.1136/bmj-2022-071108
Dear Editor,
Show MoreWe read with great interest the article by Bergstra et al., which demonstrated that residence at lower latitudes is associated with younger age of onset of rheumatoid arthritis (RA).[1] We are impressed by the in-depth explanation of how the level of socioeconomic welfare in different countries can contribute the results.[1] This contribution provides new insights into disease development and highlights the need for improving public health policies in many developing countries.
Geographical distribution, including latitude and altitude, is an important factor in several diseases. Previous studies revealed an association between geographical distribution and age of onset and severity of diseases such as systemic sclerosis and systemic lupus erythematosus.[2, 3] Furthermore, regional climate differences in similar latitude can lead different immune status in autoimmune disease.[4] The climate is influenced not only latitude or altitude but also other factors, such as ocean currents, prevailing winds, and vegetation cover. In previous studies, there is a possible association between climate factors and disease activity.[5, 6] On the other hand, for the evaluation of country-level factors, I thought that some index is more appropriate for the association survey, such as the human development index (HDI), the multidimensional poverty index (MPI), the social progress index (SPI).[7, 8]
In this study, the locations of each healthcare facility are cl...
Dear Editor,
Show MoreWe read with great interest the paper published by Perez-Garcia et al [1] on ‘Is methotrexate safe for men with an immune-mediated inflammatory disease and an active desire to become a father? Results of a prospective cohort study (iFAME-MTX)’. The authors bring original data to address an important and controversial topic regarding the use of MTX in men with immune-mediated inflammatory diseases (IMIDs) who desire to father a child. They prospectively evaluated the testicular toxicity profile of MTX through quantitative and qualitative study of the sperm from treated patients. The strength of this study lies in its prospective design and comprehensive evaluation of various fertility markers in men exposed to MTX such as semen parameters, reproductive hormone levels, sperm DNA fragmentation index, and the presence of MTX-polyglutamates (MTX-PG) in spermatozoa. The author included three distinct groups, namely MTX-starters, MTX-chronic, and healthy controls, to provide a comparative analysis and highlight the impact of both short-term and long-term MTX exposure on male fertility. They observe that MTX exposure in men with IMIDs does not significantly affect conventional semen parameters, reproductive endocrine markers, or sperm DNA fragmentation index. These results provide reassurance regarding the start or continuation of MTX therapy in men planning to father a child, which align with the recent recommendations from the American College of Rheumatology...
Dear Editors,
I read with great interest the report by Bergstra et al. (2023) on a cross-sectional analysis of rheumatoid arthritis patients from the worldwide METEOR registry1. The study is commendable for its sample size and geographical distribution of rheumatoid patients from 17 countries and 93 hospitals. Latitude has been defined as a potential underlying associative factor of early onset of rheumatoid arthritis among individuals living near the equator, which could be explained with socioeconomic status and accessibility to social welfare.
I have a few questions regarding the outcome measures and patient distribution. First, from which 17 countries were patients included, and how many patients were presented from each country along with the corresponding hospitals' latitudes? Second, were the included hospitals and countries were stratified to lower, middle, and higher latitudes, considering the disparities in socioeconomic status among low- and middle-income (LCMI) countries? These LCMI countries often have wider socioeconomic gaps, creating a complex scenario where diagnostic and treatment modalities for rheumatoid arthritis, as well as healthcare awareness, may not be uniform across rural and urban settings. This could potentially impact patient representation in the study. Third, were the included patients from low and middle socioeconomic statuses primarily presented from metro cities? It is worth noting that a majority of the patients in thi...
Show MoreWe would like to thank dr. Mahla for his interest in our paper and respond to the questions asked. The first question relates to the included patients, hospitals and countries. We were able to analyze 37,981 patients from 93 hospitals in 17 countries. Information on the included countries, and the number of patients and hospitals per country is provided in the original manuscript in table 1. Hospital latitudes are displayed in figure 1.
Show MoreAs we mentioned in the discussion section, one of the main limitations of our paper is that the number of patients per hospital/country strongly differs and it is impossible to determine to what extent these patients are representative of all patients with RA in a country. Especially in lower- and middle income economies, healthcare disparities may be significant and we were unable to include individual patients’ socioeconomic data . As an example, dr. Mahla mentions India, the country with the largest contribution of patients to our study. We performed a sensitivity analysis excluding India, and found very similar results.
Most countries were represented by only a limited number of hospitals. These were primarily located in cities. Whereas especially in lower income countries these hospitals serve a large area, this may have led to an underrepresentation of patients living in rural areas. We can only speculate on how this may have influenced our findings, which may not only depend on regional and personal socioeconomic circumst...
The article says that "At the end of the inclusion consultation, rheumatologists initiated a first targeted therapy: 8.6% by bDMARDs etc.". That number is actually wrong, probably because a typing issue. The actual number would be 89.6%.
Recently, we have been very interested in the article: "Increased risk of osteoarthritis in patients with atopic disease" by Baker et al. [1]. In this article, the author analyzed the incidence of OA in patients with atopic disease in two databases, suggesting that the immune response to atopic illness contributes to the increased incidence of OA. This study provides a new direction for immunotherapy against OA.
Show MoreOverall, the authors' research design is meticulous. In two different databases, the author divided the idiopathic disease exposure group and the non-exposed group. The author strictly stipulated the sequence of idiopathic diseases and OA to ensure the impact of idiopathic diseases on OA. Matching was performed for age, sex, race/ethnicity, education level, Charlson comorbidity score, follow-up time, and clinic visit frequency, effectively reducing these factors influence. Furthermore, OA risk was reduced after adjusting for BMI in the STARR cohort, confirming that BMI is an independent confounding factor. However, although the authors analyzed various limitations of this study in the Discussion section, there are still some unmentioned issues worth considering.
First, to reduce the influence of confounding factors, the author matched and adjusted multiple independent factors, including age, gender, and BMI. In addition, the impact of factors such as history of joint trauma and physical activity level were mentioned in the discussion sect...
We thank Dr Bitoun et al for their careful reading of our manuscript. We acknowledge that the small number of patients experiencing cancer progression in our study is a limitation as it reduced the number of covariates that could be analyzed. However, in individual unadjusted Cox models analyses we did not find that age, sex, cancer type, cancer stage, ICI type, ICI discontinuation, concomitant irAE or steroid duration were associated with progression free survival suggesting that inclusion of these covariates in a larger study might not have influenced our findings.
We do agree that arthritis disease activity is an important confounder of DMARD choice. However, in our cohort, arthritis grade at the time of presentation was similar across the groups (p=0.67). Unfortunately, other measures of disease activity such as DAS28 were not available in this retrospective study. In addition, to address potential bias arising from biologic DMARD-treated patients’ being treated earlier because of more severe arthritis, we included the time-dependent variable “time from ICI initiation to DMARD initiation” into our model (an analysis in which there was no censoring at the time of DMARD switch). With this covariate included, our results favoring methotrexate in comparison to a TNF inhibitor became if anything, stronger (HR 3.27, 95% CI 1.21-8.84, p=0.019). The two methods commonly used to account for immortal time bias are landmark models and time dependent covariates. We chos...
Show MoreWe read with interest the article by Kristensen et al. based on primary ORAL Surveillance results (NCT02092467) enrolling rheumatoid arthritis (RA) patients aged ≥50 years and with ≥1 additional cardiovascular risk factor, who were treated with tofacitinib or tumor necrosis factor inhibitors [1,2]. According to the presence of age ≥65 years or smoking, two subpopulations (i.e., ’high-risk’ and ’low-risk’) were identified, among those receiving tofacitinib, with diverse relative event risk (i.e., malignancies, major cardiovascular events, myocardial infarction, venous thromboembolism, and all-cause death) [1].
Show MoreOne feature highlighted by Kristensen et al. is the increased risk of event in RA patients aged ≥ 65 years [1]. This elderly subpopulation comprises both elderly-onset RA patients, manifesting after the age of 60 years, and those patients diagnosed with RA early in life who naturally become members of this subset. The management of elderly RA patients may be challenging [3]. In fact, the typical high prevalence of comorbidities in these patients may be associated with a lower response to RA therapies and poor prognosis. Furthermore, the concomitant polypharmacy may make even more difficult the treatment of elderly patients due to increased likelihood of drug interactions, adverse events, and possible contraindications [3]. The increased frailty associated with older age may also complicate the management of such patients; this is a syndrome characterised by a de...
We read with great interest the article by Smith et al [1], reinforcing the role of rituximab in the maintenance treatment of ANCA-associated vasculitis. This major study randomized the maintenance treatments of 170 GPA and MPA patients to rituximab (1000 mg every 4 months, up to month 20) and azathioprine (2 mg/kg/day). At the end of the month 24, relapse-free survival rate was 0.85 for the rituximab group (95% CI: 0.78–0.93) compared to 0.61 of the azathioprine group (95% CI: 0.51–0.73). Overall HR was 0.4 (95% CI: 0.27–0.61). There was no difference in rates of infection or hypogammaglobinemia, and rate of adverse events was found to be similar with previously published studies. Here we like to make two main points of interest.
Show MoreFirst, after the last dose of rituximab at month 20, the relapse-free survival rate drops to 0.5 (95% CI: 0.40–0.63) in the study group, which is parallel to the azathioprine group (relapse-free survival rate: 0.22, 95% CI: 0.14–0.35) [1]. In the discussion, the authors comment on this by stating that ‘despite using a higher dose rituximab regimen, relapses still occurred’ and ‘together with increased risk of relapse after stopping rituximab and the associated safety risks illustrate the need for newer therapeutic agents for AAV.’ [1]. While we agree that newer therapeutic options are always welcome, data from this study underlines the efficacy of rituximab when it is used. An alternative of stopping rituximab therapy in two years may be th...
We read with interest the study recently published in ARD by Bass et al. This is the first retrospective study comparing methotrexate (MTX) to bDMARDs (TNF inhibitors (TNFi) and IL6 receptor inhibitors (IL6i)) for the treatment of immune checkpoint inhibitor (ICI)-associated arthritis. It concluded that bDMARDs were associated with more rapid arthritis control but also with a higher rate of cancer progression. Since it is the first study reporting such a comparison, it could become a reference and discourage rheumatologists to use bDMARDs for treating ICI-associated IA.
Show MoreEven if the conclusions were cautious due to the retrospective nature of the study, a number of methodological weaknesses may challenge the confidence with those conclusions:
The number of events considered for comparing the progression free survival (PFS) between DMARDs is only 10, since 11 patients who progressed before DMARD introduction and 3 who were treated by different DMARDs were excluded from this analysis. With such a small number of events, regression models with many variables are at high risk of overfitting and sparse data bias[1].
The presence of immortal bias was identified, and there was an attempt to consider it. However, the way it was accounted for the analysis of PFS raises methodological concerns. Indeed, the index date to start follow-up for PFS was ICI initiation, meaning that patients were classified in each of the group (DMARDs vs bDMARDs), based on a criteria occu...
Pages