We read with great interest the article by Brennan et al (1) that confirms the results we discovered approximately ten years ago through a similar study. These results were presented as abstracts at national and international meetings, including the XVIII ILAR Congress in 1993 (2–4). Unfortunately, when we submitted our data for publication it was repeatedly rejected by authoritative rheumatology journals because the reviewers considered the information “not relevant.” We would like to briefly summarize the data, which was presented in the past, to offer our contribution to this topic.
Similar to Brennan et al, we retrospectively reviewed 113 lower lip salivary gland biopsies (LL-SGB) and seroimmunologic profiles obtained from a group of patients (9 men, 104 women; ages 18–87 years; mean age 53 ± 12.3) reporting sicca symptoms and being evaluated for suspected Sjögren's syndrome. A seroimmunologic profile was considered evaluable when obtained at the same time as the lip biopsy and included total gammaglobulins, IgG, IgA, IgM, c-reactive protein (CRP), rheumatoid factor (RF), antinuclear antibodies (ANA), anti-Ro/SSA, anti-La/SSB, C3, and C4. On the basis of the focus score <1 or ≥ 1 and based on the European Community Study Group Classification Criteria (ECSGC) for Sjögren's syndrome (5), the patients were divided into two groups: group A had positive LL-SGB and group B had negative LL-SGB. The value of seroimmunologic parameters in predicting LL-SGB findings was evaluated by means of logistic multivariate regression analysis. Fifty-four patients in group A (88.5%) and 7 patients in group B (13.4%) satisfied the ECSGC criteria. Patients in group A showed a significant increase in total gammaglobulins, IgG, IgA, ESR, and CRP; a lower level of C4; and more frequent positivity of ANA, anti-Ro/SSA, anti-La/SSB, and RF compared with group B (Table 1). The logistic regression models constructed to evaluate the predictive value of various combinations of the seroimmunologic parameters revealed that only those including IgG levels demonstrated a good performance in predicting a focus score > 1 with a probability of wrong attribution ranging from 33.4% to 35.5%.
|Group A, n = 61 (focus score ≥1)||Group B, n = 52 (focus score <1)||P|
|Mean age, years||53.2 ± 13.0||52.9 ± 11.5|
|ESR, mm/1st hour||39.60 ± 25.66||17.90 ± 19.38||<0.001|
|γ-globulins, g/L||18.37 ± 7.25||12.44 ± 2.68||<0.001|
|CRP, mg/dl||1.11 ± 1.63||0.57 ± 0.33||<0.05|
|IgG, g/L||18.24 ± 6.71||12.40 ± 2.62||<0.001|
|IgA, g/L||3.16 ± 1.47||2.25 ± 0.94||<0.001|
|IgM, g/L||1.76 ± 1.02||1.84 ± 0.98||NS|
|C3, g/L||1.09 ± 0.26||1.03 ± 0.20||NS|
|C4, g/L||0.19 ± 0.07||0.21 ± 0.05||<0.05|
|ANA positive, %||78.6||48.0||<0.001|
|Anti-Ro/SSA positive, %||39.3||3.8||<0.001|
|Anti-La/SSB positive, %||29.5||–||<0.001|
|RF positive, %||62.3||21.7%||<0.001|
|ECSG, no (%)†||54 (88.5)||7 (13.4%)|
As Brennan et al suggest, we also have concluded that in the setting of highly selected patients with sicca symptoms being evaluated for a suspected Sjögren's syndrome, elevated IgG levels represented a simple and useful parameter that correlated with a positive LL-SGB (focus score > 1). We have suggested that this simple serologic parameter could be a potential predictor of a positive LL-SGB result in at least two-thirds of the cases.
We also agree with the conclusions of Brennan et al when they state that the value of IgG levels in predicting salivary lip focus score is relevant only in a selected population of patients with an increased prevalence of positive lip biopsy or primary Sjögren's syndrome (similar to the population we studied), and it is not generalizable in unselected populations of patients generically reporting sicca symptoms.
We appreciate that the study of Brennan et al rechallenges this topic and substantially confirms our “naive” results, and we are very pleased that this information has now been reconsidered to be “relevant” in clinical practice.
Finally, we would like to comment on Manthorpe's statement in his editorial in the same issue (6) where he said, “Will the lower lip biopsy procedure persist in the future? Certainly, yes, but with reduced frequency.” From an ethical, “patient-centered,” and scientific point of view, any reliable surrogate procedure that allows clinicians to avoid an invasive procedure such as the LL-SGB, is obviously welcome. However, in a more realistic perspective there are at least two conditions that will have to be satisfied before this will become true: first, the availability of evidence-based scientific data informing us that one simple, cheap, noninvasive test is specific and sensitive enough to substitute for lower lip biopsy; and second, a cost-benefit performance analysis must favor this option over the more invasive one. In the actual reimbursement system based on Diagnosis Related Group (DRG) (ICD-9-CM), the surgical procedure adopted to obtain an LL-SGB generates a “DRG point” that is 3-times higher than an common point DRG produced by a complicated connective tissue disease. Thus, the final result is that a patient with Sjögren's syndrome, in which an LL-SGB has been performed, is much more “remunerative” than a patient with complicated systemic lupus erythematosus.
In the current climate, one in which rheumatologists and other physicians work under remarkable financial pressure, we are not completely sure that the frequency of LL-SGB or other invasive diagnostic procedures will be destined to decrease in the near future. We think that rheumatologic institutions should begin a common effort devoted to attracting the attention of health authorities so that unnecessary, invasive patient procedures can be avoided. In this context, the evaluation of IgG levels as a reliable risk factor for positive LL-SGB results remains, if it is confirmed in non-selected patients, a promising and interesting issue although to date it is undoubted less “convenient” than a biopsy.