To the Editor:
We have read with great interest in the article by Wang, et al on the higher incidence of non-Hodgkin lymphoma (NHL) in patients with primary Sjögren syndrome (pSS) and the higher incidence of pSS in patients with NHL1. Thank you for the discovery of the bilateral relationship between pSS and NHL in this nationwide population-based study1. However, we have some questions and findings that we would like to discuss.
First, although the pSS cohort excluded patients diagnosed with other specific autoimmune diseases before the index date, it was hard to identify if there were patients who had other autoimmune diseases diagnosed during the study period, from the date of application date for a catastrophic illness certificate (CIC) of pSS in patients with pSS until NHL was diagnosed. In addition, there was a possibility of underdiagnosis of other specific autoimmune diseases before the index date that may influence the incidence of NHL in patients with pSS, since other specific autoimmune diseases (e.g., rheumatoid arthritis, systemic lupus erythematosus) are consistently reported for increasing NHL risks, and it is common for pSS to be combined with other autoimmune diseases2. Except for the confounders mentioned above, obesity and immunocompromised patients (e.g., HIV/AIDS and organ transplantation) were also potential risk factors of NHL3. All prominent confounders including those that Wang, et al indicated in the discussion1 such as smoking habit4, family history5, specific virus or bacterial infections, and other confounders referred to in this letter that may influence incidence rate of pSS and NHL3,6 should all be adjusted to make the data of this study more accurate and reliable.
Second, there may be a delay in the CIC application, and the interval between the diagnostic date and the CIC application date should be taken into consideration. Patients may have had the diagnosis of pSS long before they received the CIC.
Third, in both the pSS and NHL cohorts, the treatments for the diseases were not taken into consideration. The treatments for pSS (e.g., some disease-modifying antirheumatic drugs, nonsteroidal antiinflammatory drugs, systemic corticosteroids) may potentially increase the incidence of NHL in patients with pSS and vice versa7; the treatments for NHL may also affect the incidence of pSS in patients with NHL. For example, rituximab can be administered for both NHL and pSS8. Further, assessment of disease severity and treatments may better predict the relationship between NHL and pSS8.
If the questions mentioned in correspondence could be further clarified, this study could become more representative and complete.
Footnotes
Y.T. Liang and P.Y. Leong contributed equally to this article.
The authors have declared no conflicts of interests.
- Copyright © 2021 by the Journal of Rheumatology