Is chronic ulcerative stomatitis an entity? Clinical and immunological findings in 18 cases

ARTICLE

In 1990, we identified a new disease entity involving the oral mucosa which we termed chronic ulcerative stomatitis (CUS) [1, 2]. The clinical presentation is often diverse, bearing similarities to desquamative gingivitis, cicatricial pemphigoid or ulcerative lichen planus. This disorder has been found associated with unusual antinuclear antibodies which react preferentially with the lower layers of squamous epithelium and are referred to as stratified epithelium-specific antibodies (SES-ANA). The disease occurs almost exclusively in older women, the erosions being confined to the oral mucosa, the course is chronic with relapses, but the general condition of the patients is not affected. Most importantly, antimalarials have been found to produce clearing even in cases lasting as long as 20 years. At the same time, and independently, Parodi et al. [3] reported two similar cases, identified as erosive LP, one of them with concomitant cutaneous lesions of the LP type. Further studies have confirmed the original findings [4-7].

We report on a cohort of 18 patients found to be positive for SES-ANA, 15 of them had oral lesions of the CUS type, some have been followed-up for several years. We include also one case described in 1990, followed-up for 8 years. We provide new information on the clinical association of SES-ANA, their pathogenic significance, the course of the disease, and therapeutic modalities in addition to chloroquine.

Material and methods

Our cohort comprised 15 cases of CUS with typical clinical features and immunological findings.

In addition, we studied three cases associated with SES-ANA, but with no mucous membrane involvement. The antibodies were detected accidentally, in one patient the serum was studied for ANA because of suspected connective tissue disease (CTD), in the second patient because of a bullous eruption on the thighs after a burn injury, initially suspected of being BP. The third patient, a young woman investigated for ANA because of suspected CTD, with no oral involvement, had had stomatitis over 10 years earlier. At that time no immunological studies were performed.

Immunofluorescence studies were performed according to the methods described by Beutner et al. [8] using human and guinea pig esophagus as substrate, and HEp2 cells. In some patients, direct IF studies were also performed in the mucosa and in univolved skin. Histology was performed initially in all patients but due to a nonspecific pattern we later studied only cases with features of LP.

Because of the controversy concerning the relationship between CUS and LP, we used as controls 91 cases of LP of the skin and 25 cases of oral LP, studied immunologically and histologically.

Results

The main clinical features and immunological findings are presented in Table I. Of the 15 cases of ulcerative stomatitis (Fig. 1) associated with SES-ANA (Figs. 2 and 3), there were two males and three females younger than 40 years of age, but the majority of patients were older females.

Of special importance are the three cases of CUS demonstrating typical ulcerative stomatitis associated with LP of oral mucosa and the skin. Two of these women were in a younger age group. The histology of the cutaneous lesions was characteristic of LP. In another case, atypical lichenoid changes in the skin had no histological features of LP. In no typical case of LP were there immunological findings characteristic of CUS. In 52 out of 91 cases of cutaneous and 15 out of 25 oral LP, hyaline bodies were detected. Histology of LP was in general characteristic, that of erosive lesions ­ frequently difficult to interpret, whereas the histology of CUS was not characteristic (inflammatory infiltrates).

In IIF studies of LP, the anticytoplasmic and anti-membraneous antibodies (mainly around the basal cells) as reported by Lin et al. [9] were detected in a high proportion of cases. However, we found them to be non-specific, and they were not studied in detail. We can not exclude the fact that they are characteristic of cutaneous LP since they are present in a number of cases. No antibodies of this type were present in patients with CUS.

One patient with typical oral lesions of CUS associated with SES-ANA and erosive conjunctivitis (Fig. 4) was suspected of having cicatricial pemphigoid, but this was not confirmed either by IIF or DIF. Instead of linear immunoglobulin deposits, in vivo-fixed SES-ANA were detected in the conjunctival biopsy (Fig. 5). This, however, is by no means proof that this was a specific localisation of CUS to the conjunctiva since in vivo deposits were also found in the normal skin of this and other patients.

Of great interest is the case of typical CUS in a woman who, 4 years after the onset of mucosal changes, developed unusual cutaneous lesions (Fig. 6). The patient responded initially to chloroquine and for about 20 months this drug alone controlled the mucosal changes. Later, for the clearance of the mucosal lesions it was necessary to combine chloroquine with prednisone, however the disease tended to relapse. A severe relapse was accompanied by cutaneous involvement somewhat resembling erythema necrolyticum migrans. No glucagonoma was disclosed and high titers of IgA EmA were detected in the serum, which was not associated with any gastrointestinal symptoms. However, the duodenal biopsy showed flat mucosa with inflammatory infiltrates in the lamina propria, consistent with latent coeliac disease. The combined therapy with chloroquine and prednisone did not control either the mucosal or cutaneous changes. Only after introduction of a gluten-free diet (GFD), did the mucosal and skin lesions clear, with occasional slight relapses only in the skin. The relationship between CUS and latent coeliac disease is not known, however both cutaneous and mucosal lesions responded to the GFD. IgA EmA antibodies regressed under this diet, but SES-ANA persisted. This indicates that SES-ANA, a marker of CUS, may allow a retrospective recognition of past disease.

The patient reported in 1990, who showed a dramatic response to chloroquine, had several relapses during further follow-up, which were controlled with small doses of chloroquine combined with prednisone.

Discussion

We confirmed the finding in previous studies [1-7] that SES-ANA are quite unique antinuclear antibodies which are reactive exclusively with keratinocytes, and are non-reactive with HEp2 cells. We also confirmed that SES-ANA usually have high IF titers, as a rule higher on guinea pig esophagus than on monkey or human esophagus. Although the titers show fluctuations and are frequently lower on improvement, the antibodies do not disappear and there is no correlation with clinical condition.

According to Parodi these antibodies react in immunoblot with a 70 kD protein, as found with keratinocyte extract (personal communication). The study by Lee et al. ¹ showed that the sera of CUS patients bound a protein of approximately 72 kD present in keratinocytes, and were directed against the KET protein which is closely related to p53 tumor suppressor.

The newly detected gene KET is expressed exclusively in epithelia and thymus and thus may play a role in keratinocyte cell cycle control. This may explain the specificity of SES-ANA for keratinocytes.

Our study on a large cohort of 18 patients found to be positive for SES-ANA, has shown that in rare cases, SES-ANA may not be associated with CUS. In one case they were found in a patient who had had stomatitis many years earlier, two other patients did not remember having had any mucosal lesions in the past.

There was no correlation with clinical condition, and the titers were usually still high even after complete clearance of the mucosal lesions, and persisting for over 10 years. By direct immunofluorescence, they were found to be fixed in vivo in the uninvolved skin of the patients. Their pathogenic significance is not clear and merits further studies.

In a case with concomitant latent coeliac disease, the SES-ANA persisted after regression of the changes under GFD, whereas IgAEmA antibodies correlated with the cutaneous changes and disappeared after clearance of the skin.

We confirmed the complement fixation by SES-ANA (E.H. Beutner ­ personal communication), and we also found that the IF staining was in this reaction stronger, especially on human esophagus. However, the titers of SES-ANA are usually very high, and the IIF reaction is strong enough in the routine IF study. Complement fixation of SES-ANA does not provide proof for their pathogenicity, since SES-ANA not associated with CUS also activate complement.

We have shown that CUS might, in rare cases, occur in males and in women younger than 40 years of age, however we confirmed the former finding that the disease affects mainly older women. The concomitant lesions on the genital mucosa or conjunctiva found in 2 cases, are suggestive of a possibly more widespread involvement of mucous membranes. However, such an involvement is very rare. The differentiation between bullous and cicatricial pemphigoid is based on detection of in vivo bound and circulating SES-ANA.

The relationship between CUS and LP is demonstrated by the not infrequent association of these two diseases: in our cohort, LP of the skin and mucosa coexisted in 13% of CUS cases, and mucosal LP-like lesions in 20% of cases. LP- or LP-type changes are not infrequently associated with pemphigus neoplasticus (in 3 of our patients, with histology also characteristic of LP). However, IF findings are typical of PNP. The relationship of some CUS cases with LP should be stressed, however LP of the skin, LP pemphigoides, and LP with erosive oral changes were found to be immunologically different. It is possible that LP or LP-like lesions are GVHR, as seen in bone marrow allograft. This problem is not clear and must be studied further.

The course of CUS is variable, and the response to chloroquine is initially dramatic and favorable. The mechanism of action is however not known. It has been suggested that in autoimmune diseases chloroquine interferes with antigen processing and presentation, resulting in down-regulation of the immune response against autoantigen [10]. It was found to enhance apoptosis, to have an antiangiogenic effect due to apoptosis of the endothelial cells [11] and to inhibit the development of graft-versus-host disease [12]. In our large cohort, followed-up for several years, the antimalarials, initially very effective, proved in most cases to be not sufficient or the therapy could not be continued for a very prolonged period due to gastrointestinal and/or ocular side-effects. In such cases, combination of chloroquine with small doses of corticosteroids, or corticosteroids with sulfones, both in small doses, were found to control the disease. In spite of the remarkable response to chloroquine, which is the drug of choice, in most patients occasional relapses were not an infrequent finding and in such cases chloroquine as monotherapy did not prevent recurrences.

According to our new observations, we believe that the previous diagnostic criteria for CUS [1, 4] should be modified, and we propose the following major and minor criteria:

Major criteria

­ erosive or exfoliative oral lesions,

­ characteristic IIF and DIF findings as a basis for diagnosis.

Minor criteria

­ chronic course with relapses,

­ predominance of females in the older age groups,

­ response to chloroquine or chloroquine combined with small doses of corticosteroids.

CONCLUSION

REFERENCES

1. Jaremko WM, Beutner EH, Kumar V, Kipping H, Condry P, Zeid MY, Kauffman CL, Tatakis DN, Chorzelski TP. Chronic ulcerative stomatitis associated with a specific immunologic marker. J Am Acad Dermatol 1990; 22: 215-20.

2. Chorzelski TP. Chronic ulcerative stomatitis (CUS) Nowa jednostka chorobowa ze swoistym markerem immunologicznycm (SES-ANA). Przegl. Dermatol 1990; LXXVII (4): 229-34.

3. Parodi A, Cardo PP. Patients with erosive lichen planus may have antibodies directed to a nuclear antigen of epithelial cells. A study on the antigen nature. J Invest Dermatol 1990; 94: 689-93.

4. Beutner EH, Chorzelski TP, Parodi A, Schosster R, Guin J, Cardo PP, Maciejowska E, Valeski JE, Kumar V. Ten cases of chronic ulcerative stomatitis with stratified epithelium-specific antinuclear antibody. J Am Acad Dermatol 1991; 24: 781-2.

5. Church LF, Schosser RH. Chronic ulcerative stomatitis associated with stratified epithelial specific antinuclear antibodies. A case report of a newly described entity. Oral Surg Oral Med Oral Pathol 1992; 73: 579-82.

6. Levis JE, Beutner EH, Rostami R, Chorzelski TP. Chronic ulcerative stomatitis with stratified epithelium-specific antinuclear antibodies. Int J Dermatol 1996; 35: 272-51996.

7. Wörle B, Wollenberg A, Schaller M, Kunzelmann KH, Plewig G, Meurer M. Chronic ulcerative stomatitis. Br J Dermatol 1997; 137: 262-5.

8. Beutner EH, Chorzelski TP, Kumar V. Immunopathology of the skin J Wiley and Sons, New York, 1987.

9. Lin SCh, Sun A, Wu Y Ch, Chiang Ch P. Presence of anti-basal cell antibodies in oral lichen planus. J Am Acad Dermatol 1991; 26: 943-7.

10. Fox RI. Mechanism of action of hydroxychloroquine as an antirheumatic drug. Semin Arthritis Rheum 1993; 23 (2 suppl.): 82-91.

11. Potvin F, Petitclerc E, Marceau F, Poubelle PE. Mechanism of action of antimalarials in inflammation: induction of apoptosis in human endothelial cells. J Immunol 1997; 158: 1872-9.

12. Schultz KR, Gilman AL. The lysomotropic amines, chloroquine and hydroxychloroquine: a potentially novel therapy for graft-versus-host disease. Leuk-Lymphoma 1997; 24: 201-10.