To the Editor:
We read with interest the article by Sharma, et al that evaluated a cohort of patients with symptoms of mucosal dryness undergoing diagnostic workup for the suspicion of primary Sjögren syndrome (pSS)1. The assessment also included a minor salivary gland (MSG) biopsy and patients were classified as pSS if fulfilling the histological [focus score (FS) ≥ 1], serological (anti-Ro based on the American College of Rheumatology/European League Against Rheumatism set or either anti-Ro or anti-La based on the American-European Consensus Group set), or both criteria2,3. Among the 229 subjects classified as having pSS, the authors observed a FS = 0 in 51 subjects (22%), between 0 and 1 in 11 subjects (5%), and ≥ 1 in 167 subjects (73%). According to this, patients were divided in 3 groups, and clinical and serological features were compared. In the FS ≥ 1 group, only patients with anti-Ro antibodies were selectively included to match those with FS = 0, all of whom have anti-Ro. We previously performed a similar exercise by retrospectively evaluating 383 unselected pSS patients who underwent MSG biopsy at the time of diagnosis and were scored with the FS4. Unlike Sharma, et al1, who grouped patients with FS ≥ 1, we considered patients with FS = 1 and those with FS > 1 separately and compared them with patients with FS = 0. Further, in the FS = 1 and FS > 1 groups, we included patients based only on the FS value, regardless of the serological status (either anti-Ro, anti-Ro and anti-La, or none). As shown in Table 1, seronegative patients represented almost 40% of the FS ≥ 1 cohort and when evaluated separately, they displayed some peculiar features worth discussing. For instance, despite a higher FS being associated with lower prevalence of sicca symptoms and a more severe clinical picture with extraglandular manifestations4,5,6, seronegative patients with FS ≥ 1 have an OR 4.6 (95% CI 1.43–14.8, P = 0.01) for xerostomia compared to those with either anti-Ro, or anti-Ro and anti-La antibodies (OR 0.22, 95% CI 0.07–0.7, P = 0.011 vs seronegative). Likewise, despite an overall higher prevalence of parotid gland swelling being reported in association with higher FS values4, seronegative patients with FS ≥ 1 have an OR 0.46 (95% CI 0.25–0.85, P = 0.013) for this manifestation as compared to those with either anti-Ro, or anti-Ro and anti-La antibodies (OR 2.17, 95% CI 1.18–4.0, P = 0.013 vs seronegative). Finally, as far as serological features are concerned, seronegative patients with FS ≥ 1 have an OR 0.065 (95% CI 0.031–0.136, P < 0.0001) and 0.38 (95% CI 0.20–0.73) for hypergammaglobulinemia and leukopenia, respectively, compared to those with either anti-Ro, or anti-Ro and anti-La antibodies (OR 15.3, 95% CI 7.4–32, P < 0.0001 and OR 2.64, 95% CI 1.37–5.1, P = 0.004, respectively). Of particular interest, no significant differences pertaining to the association with lymphoma were observed according to the serological status in patients with any FS value. It is important to mention that since our data result from the retrospective assessment of a cohort of established pSS with a mean follow-up duration of 5.8 years (SD 6.5 yrs), this allowed us to unmask associations between the FS and manifestations that may not be evident at disease onset but develop over time. Further, the inclusion of patients with any serological status in the FS ≥ 1 group allowed the observation that the positive association between severity of inflammatory infiltrate and some clinical and serological features such as parotid gland swelling, leukopenia, and hypergammaglobulinemia is not only absent but even opposite in seronegative patients (Table 2). Our results underscore the concept that when prospective data are not available, retrospective assessment of patients with overt disease, rather than cross-sectional evaluation of patients at the time of pSS diagnosis, may be more informative. In addition, when the histological status is one of the inclusion criteria, consecutive patients with any serological status should be included to prevent selection bias and encompass the wide spectrum of the disease.
Distribution of autoantibodies according to the focus score (FS) value.
Clinical and serological features of patients according to the focus score (FS) and autoantibody status. Patients with isolated anti-La are not displayed due to their low number.
Acknowledgment
Authors are grateful to the coinvestigators who coauthored our referenced research study (in alphabetical order): Chiara Baldini, Elena Bartoloni, Salvatore De Vita, Roberto Gerli, Roberto Giacomelli, Luca Quartuccio, Roberta Priori, and Guido Valesini.