Need for consensus guidelines to standardise the assessment of germinal centres and other histopathological parameters in salivary gland tissue of patients with primary Sjögren's syndrome.

We have read with great interest the letter to the editor by van Roon et al commenting on our paper ‘Towards personalised treatment in primary Sjögren’s syndrome: baseline parotid histopathology predicts responsiveness to rituximab treatment’. The authors argue that there is a need for standardisation of the histopathological characteristics of salivary gland tissue of patients with primary Sjögren’s syndrome (pSS), in general, and of the presence of germinal centres (GCs), in particular. We fully agree with van Roon et al and other authors about the need for consensus guidelines to standardise the histopathological evaluation of salivary gland biopsies in patients with pSS. A standardised scoring system may facilitate prognostication and stratification of patients with pSS and is needed for a valid evaluation of various clinical trials. In particular, histological definition of GCs in salivary gland tissue is warranted, since these structures can be difficult to detect in diagnostic H&E-stained tissue sections. Detection of GCs in the periductal lymphoid infiltrates of the salivary glands is clinically relevant, because the presence of these structures is associated with more severe disease. Importantly, the presence of GCs in minor salivary gland biopsies has been postulated to be a predictor of patients who are at risk of lymphoma development. 6 It has to be mentioned, however, that recently, we were not able to confirm these findings for a larger number of mucosa-associated lymphoid tissue (MALT) lymphomas in parotid glands of patients with pSS (Haacke et al, unpublished data). In our study, we defined GCs in H&E-stained sections as lighter areas within the lymphoid infiltrate composed of both lymphoid cells (centrocytes, centroblasts) and cells with a nonlymphoid nature (macrophages and follicular dendritic cells (FDCs)) (figure 1A). Furthermore, the GCs were scored independently by two experienced pathologists. For the inexperienced eye, GCs may be overlooked, because of their small size, or lighter areas within the infiltrate may erroneously be scored as GCs, while in fact they represent lymphoepithelial lesions. For proper and easy detection of GCs, also by less-trained persons, additional immunohistochemical staining might be helpful. Therefore, we propose to stain for B-cell lymphoma 6 (Bcl-6) to define and identify GCs. Bcl-6 is a transcription factor expressed at high levels by GC B-cells. Like GCs in peripheral lymphoid organs, GCs in salivary glands of patients with pSS are also consistently positive for Bcl-6. As shown in figure 1B, staining for Bcl-6 allows the easy and unequivocal detection and scoring of GCs in salivary gland biopsies, both in minor and major (parotid) salivary glands. Implementation of Bcl-6 staining is relatively easy, since it is routinely used in pathology laboratories worldwide for the diagnosis of lymphomas. Other markers, as proposed by Fisher et al and van Roon et al, are less specific and less suitable to detect GCs in routine diagnostics. For example, activation-induced deaminase, an enzyme essential for the function of GCs B-cells, is expressed only by a minority of GCs B-cells in minor salivary glands of patients with pSS, which may make GCs harder to detect. The long isoform of CD21 (CD21L) has also been suggested for detection of GCs. CD21L is expressed by FDCs. However, although FDCs are a prerequisite for GC function and development, the presence of these cells does not necessarily imply that GCs are present. Indeed, ectopic lymphoid infiltrates in salivary gland tissue of patients with pSS can contain FDC networks in the absence of GCs. 9 Staining for the CD21L may therefore result in an overestimation of the number of GCs present in salivary gland tissue. In our study we observed a relative high proportion of the parotid salivary gland biopsies presented with GCs at baseline— 67% and 68% of patients in the placebo-treated and rituximabtreated groups, respectively. These are relatively high percentages compared with the general pSS population, in which approximately one-quarter of the minor salivary gland biopsies exhibit GCs. The reason for this high baseline characteristic can be attributed to the inclusion criteria of our study. In our study, the patients with pSS were all positive for anti-SSA antibodies and