Immunogenetic markers and immune response in patients with recurrent oral ulceration

Int. J. Oral Surg, 1983: 12: 23-30 (Key words: ulcer. aphthous: immunology; HLA; hypersensitivity, morons membrane. ural)

Immunogenetic markers and immune response in patients with recurrent oral ulceration MARIA MALMSTROM, OSMO P. SALO AND FREJ FYHRQUIST Department of Oral Surgery, Institute ofDentistry, University ofHelsinki; Department ofDermatology, and the Fourth Department of Medicine, Helsinki University Central Hospital. Finland

ABSTRACT - 20 patients, aged 20 to 72 years (mean 36.5 years), 14with recurrent (RAS) and 6 with recurrent cicatrizing (RCAS) aphthous stomatitis were studied. 3 patients (15%)had the HLA locus All antigen, whose frequency in the Finnish population is 8~~. 5 patients (25%) had B12, which occurs in 15% of the normal population. Results of routine serological tests were normal. All had normal serum levelsof IgG, IgM, IgA and complements C3 and C4. 4 patients, 2 with RAS and 2 with RCAS, had raised serum IgE. Precipitating antibodies against milk protein were detected in 2 patients and against gluten in I. In 4 patients, tests for immediate allergy were positive. 5 patients had antibodies to double-stranded DNA Delayed hypersensitivity reactions werenormal, and the PHA stimulation oflymphocytes elicited normal T'-cellresponses in all patients except one with RCAS. In this patient, there was a striking parallellism between an increase in PHA-reactive lymphocytes and clinical improvement. The serum of this patient contained a binder for '''I-labelled PHA, a binder not consistently detected in the other patients with ROU. Lymphocyte dysfunction may playa role in ROU. Of the 16 biopsy specimens of aphthous tissue studied by direct immunofluorescence for IgG, IgM, IgA, fibrinogen and C3, 15 specimens contained deposits of C3 in and along mucosal vessels, whereas among the 15 controls only 1 specimen of erosive lichen planus showed deposits of C3 along capillary walls. Immune complexesprecipitating in capillary wallsappear to be a common feature of ROU.

(Received for publication 26 October 198/. accepted 6 July 1982)

The etiology of recurrent oral ulceration (ROD) is still unknown. Attempts have been made to determine whether this disease might be either genetically determined or caused by a defective immune response. The pathogenesis has been

studied chiefly by examination of diseased tissue. SIRCUS et af.2 6 reported that among their patients with ROD, 46% had a family history of the disease. Since then, a number of

24

MALMSTRbM, SALO AND FYHRQUIST

investigators"·14.16,22.25 have sought a genetic explanation for recurrent aphthous stomatitis (RAS), the most common form of recurrent oral ulceration, but results to date have not been convincing. Among the investigators who, since the mid1960s, have studied immunogenetic markers and the immune response in ROD are CHALLACOMBE et al.", DOLBy6, DONATSKy9 and LEHNER'6-19. CHALLACOMBE et al:" observed that among their patients with ROD, the frequencies of the HLA antigens A2, B12 and Aw29 were significantly increased. DOLBy6 and PLATZ et al?", on the other hand, reported no deviations in the incidence of HLA-A or -B antigens in their patients with RAS. The pathogenesis in ROD has been postulated to be autoimmune. LEIiNER'6.17 detected lymphocytes sensitized to saline extracts of foetal as well as adult oral mucosa in patients with RAS. His patients also had raised levels of circulating antibodies to oral epithelial cells. Possible causal connections between these observations and RAS have not been identified. Microorganisms and food allergens have been proposed" as being exogenous aetiological factors in ROD. The search for viral causes for ROD has been unsuccessful". DONATSKY & BENDIXEN? demonstrated cellular hypersensitivity to Streptococcus sanguis A2 in ROD but were unable to relate this finding to the pathogenesis of the disease. Based on their studies of patients with jejunal abnormalities and RAS, FERGUSON et al.!" have suggested that gluten might be an aetiological factor in RAS. Histological examination of the ulcerous tissue in ROD has yielded no conclusive clarification ofeither aetiology or pathogenesis. We have examined a selected series of patients with ROD for possible immunogenetic markers and defects in their immune response.

Material and methods The series comprised 20 patients, 8 men and 12 women, aged 20 to 72 years (mean 36.5 years), treated

Table I. Patient information for 20 patients with recurrent oral ulcerations, either recurrent aphthous stomatitis (RAS) or recurrent cicatrizing aphthous stomatitis (RCAS)

RAS

No. of patients

Age range (mean)

Sex M/F

14

20-72 (33.2) 26-63 (44.2)

5/9

RCAS

6

Total

20

20-72 (36.5)

3/3

8/12

for ROU at the Department of Oral Surgery, the Institute of Dentistry, University of Helsinki, and at the Department of Dermatology, Helsinki University Central Hospital. The patients were grouped according to the WHO classification": 14 had small recurrent aphthae (RAS) and 6 had large cicatrizing aphthae (RCAS) (Table 1). The history of disease ranged from 1 to 20 years. No patient had herpetiform aphthae or Behcet's syndrome. The immunogenetic markers of the HLA-A and-B loci were determined for all patients at the Tissue Typing Laboratory of the Finnish Red Cross Transfusion Service, Helsinki. The haplotypes and genotypes of 5 consanguineous relatives of one of the patients with RAS were also determined; all 5 relatives had a history of aphthae. Routine serological and immunological tests (Table 2) were performed with standard methods at the Fourth Department of Medicine, Helsinki University Central Hospital. The following special tests (for lymphocyte function) were performed on one patient (VT) with RCAS. Whole blood was cultured with crystalline kidney bean leucoagglutinin (Pharmacia Fine Chemicals, Uppsala, Sweden) to follow the changes in lymphocyte function during the clinical course. After cytocentrifugation, sedimented cells were stained and counted for transformed cells. The degree of lymphocyte stimulation was also measured quantitatively by incorporating mI_ desoxyuridine into lymphocytes purified by Isopaque-Ficoll centrifugation. Before being suspended, the purified lymphocytes were washed 3 times in culture medium (Eagle's MEM) containing 10% autologous or homologous serum. Cultures were incubated in an atmosphere of 5% CO 2 in air for 4

IMMUNOLOGY AND APHTHOUS ULCERATIONS

25

Table 2. Serological and immunological tests performed with routine methods on 20 patients with recurrent oral ulcerations ESR Hb Hematocrit Er MCH Leuk - differentiation Tromb MCV Antistreptolysin (AST) Antistaphylysin (ASTA) CRP Waaler-Rose Latex fixation Cold agglutinin Coombs, direct Immunoglobulins -

IgG IgA IgM IgE

Complement - C3 - C4 Nuclear antibodies - total - IgG

- IgM - DNA, double-stranded Organ-specific antibodies Viral antibodies Serum protein electrophoresis PHA stimulation Delayed hypersensitivity (skin tests with: - tuberculin - trichophytin - oidiomycin) Precipitating antibodies against - milk protein - gluten HLA-typing at A and B loci

days and contained I x la' lymphocytes in 2 ml of medium.

Macrophage migration inhibit ion factor (MIF) was measured according to the method of SeDORG & BENDIXEN" with crystalline leucoagglutinin used antigen. Quantitation of T and B lymphocytes was performed with the rosette technique". "'I-labelled leucoagglutinin was prepared with the chloramin T method" . The "'l-labelled material was purified on a Sephadex G·IOO column eluted with 0.15 M phosphate buffer, pH 7.5, containing ISO mrnol NaC!. Purified label was stored at -20°C for less than 2 weeks. Binding of "'I-Ieucoagglutinin to serum factors was studied with gel filtration . Patient or control serum, 0.2 ml, was incubated with 0.2 ml of I1~I-leucoagglutinin, 100,000 counrs/luu s for 15 min at room temperature. The mixture was then eluted on a Sepbarose 4 B (Pharmacia Fine Chemicals, Uppsala, Sweden)column, 70x I em, elution rate 0.5 ml/min, pressure 15 em H,o, at +4°C. Biopsies of the ulcerative tissue weretaken from 16 patients, 10 with RAS and 6 with RCAS. As controls were used biopsies from 15 patienls with oral nonaphthous ulcers, 6 with erosive lichen ruber planus and 9 with non-specific ulcerations. Specimens were studied with routine histologicaltechniques, as wellas by direct immunofluorescence microscopy with

FITC-labelled antisera for IgG, IgM and IgA, fibrinogen, and complement component C3, at the Department of Dermatology, Helsinki University Central Hospital.

Results HLA ant igen. Determination for A-locus antigens showed that 3 patients, 2 with RAS and I with RCAS, had All. The frequency of All in the normal Finnish population is 8%; in our series it was 15% . No other A-locus antigens occurred with greater than normal frequency". Of the 20 patients, 5 or 25% (4 with RAS and I with RCAS) had the antigen B12, which occurs in 15% ofthe Finnish population". The antigen B13, present in 6% of the normal population, was detected in 3 patients with RAS and I with RCAS, or in 20%. I RAS patient had (the antigen) B40, a frequency of 5% versus 15% in the general populatiorr" (Table 3). Of the 6 persons related to I RAS patient whose baplotypes and genotypes were determined, the propositus and 3 relatives had B12.

26

MALMSTROM, SALO AND FYHRQUIST

Table 3. B-locus antigens among 14 patients with recurrent aphthous stomatitis (RAS) and 6 patients with recurrent cicatrizing aphthous stomatitis (RCAS) B5 RAS RCAS

B8

BI2

B13

BI5

3

4 I

3 I

4

3

5 25%

4

20%

4 20%

2

Total

2 10%

15~;

Serology. In no patient did results of routine serological tests differ from reference values. Serum concentration of IgG, IgM and IgA, as well as C3 and C4 were also within normal limits. 4 patients, 2 each with RAS and RCAS, had high concentrations of total IgE (Table 4). None had raised antibody titres against herpes simplex or other viruses. 2 patients, one each with RAS and RCAS, had precipitating antibodies against milk protein, and one, with RCAS, against gluten, which is not seen in a normal population. 4 patients with RAS and I with RCAS had antibodies to double-stranded DNA (binding capacity more than 4 mg DNA/I). PHA-stimulation oflymphocytes also elicited a normal reaction in all patients except one (VT). Hypersensitivity. Skin tests for immediate hypersensitivity to 20 common allergens gave at least I positive reaction in 4 patients with RAS which is seen only in atopia. The delayed hypersensitivity reaction was normal in all except one patient (VT).

B40

I 5%

Patient VT. Cutaneous tests for delayed hypersensitivity revealed an exaggerated reaction both to tuberculine and to crystalline leucoagglutinin applied intradermally; 0.1 TV of Mycobacterium tuberculosis (Purified Protein Derivative, Parke Davis) gave a 20 x 20 mm reaction, and 1.0TV gave a 60 x 65 mm intense reaction with vesicles and fever at 39°C. The intradermal injection of 5 flg of crystalline Ieucoagglutinin? yielded a similar reaction. After 24 h, there was a 112x 95 mm and after 48 han 80 x 75 mm intense reaction with vesicles and fever at 39.6°C. The result in 4 healthy controls given the same dose of leucoagglutinin was 33 x 27 mm after 24 hand 22 x 25 mm after 48 h (mean values), with no fever or vesicles. The numbers of T and B lymphocytes in peripheral blood were normal. PHA-stimulated blast transformation was suppressed (Fig. l) and conspicuously related to the clinical course of RAS. An absence or a low %of PHA-reactive lymphocytes was associated with an increased number of aphthae and an

Table 4. Serum concentrations of immunoglobulins IgG, IgM, IgA and IgE and complements C3 and C4 among 20 patients with recurrent oral ulceration Normal range

IgM IgE

8.0-18.0 0.9- 4.5 0.6- 2.4 <110

mg/rnl mg/rnl mg/ml U/ml

13.08 mg/ml ± 3.21* 2.38 mg/ml± 0.72 1.58 mg/ml ± 1.10 68.65 U/ml±55.53

C3 C4

0.7- I.7 mg/ml 0.2- 0.5 mg/ml

0.8 mg/ml j, 0.12 0.37 mg/ml± 0.13

IgG

19A

*

±S.D.

Mean values for the patients

IMMUNOLOGY AND APHTHOUS ULCERATIONS

27

Ij. ;;.

Fig. 1. % PHA-stimulated blast transformation of lymphocytes (0) and periods of clinical improvement of the RAS (rectangles marked "I") during a 12month follow-up period. Clinical improvement denotes disappearance of aphthous lesions and subjective symptoms.

Fig. 3. Deposit of complement component C3 in a

mucosal vessel in aphthous tissue.

exacerbation of symptoms. When tested with incorporation of I25I-desoxyuridine, the patient's lymphocytes wereclearly hyporeactive to PHA doses ranging from 0.5 to 10pg/ml (Fig. 2). Washing the lymphocytes did not improve their PHA responsiveness.

15 ('1)0 ~

)C

2

a. u1Q l&I

Z Q

'"

::J )0-

>< 5

oVI

l&I

Q

I

!:!

Studies with gel filtration provided evidence that the patient's serum, but not the pooled control serum, contained a macromolecular factor capable of binding "'I-Iabelled PHA (leucoagglutinin). Serum from the patient did not inhibit PHA-induced blast transformation of lymphocytes from normal subjects. Nor did high doses of PHA increase incorporation of I25I-desoxyuridine into patient lymphocytes . The observed binding factor moved in the CX zmacroglobuli n fraction on paper electrophoresis. Its molecular weight wasabout 800,000 as estimated from Sepharose gelliltration studies. Histology. Routine histological examination of the aphthous tissue from 16pa tients (10 with RAS and 6 with RCAS) shed no new light on the established histological picture . Examination by direct immunofluorescence revelaed no IgG, IgM or IgA and no fibrinogen. However, 15 of the 16 biopsies showed deposits of C3 along capillary walls (Fig. 3). In the control group, only one biopsy from a patient with erosive lichen planus showed deposits of complement C3 along capillary walls.

O ..........-.::::r---,__--,..__....,

o

1 10 20 LEUCOAGGLUTININ P9/2 ml Fig . 2. I25I-desoxyuridine incorporation into patient (0 .) and control (.6 4) lymphocytes at different concentrations of PHA (Ieucoagglutinin), in the o 4) presence of autologous Co 6) or homologous C serum . Mean±S.E.M. of 3 experiments .

Discussion Among their patients with ROU, DOLBY' and PLATZ et a/,2" detected no deviations from the normal frequencies of HLA-A and -B loci anti gens. By contrast, when CHA L LACOMD B et at:', typed a series of 100 patients with ROU,

28

MALMSTR6M, SALO AND FYHRQUIST

43% had the antigen B12,compared with 22% in their control group, a deviation of21%. In our series, the corresponding deviation was 10%. In the series of CHALLACOMBE et al,', 33% of the patients and II % of the controls had the haplotype HLA-A2, B12. None of our patients had this haplotype. That HLA-BI2 might be associated with ROD is of diagnostic aid, since ROD can be confused with many ulcerative diseases, such as the early stages of Behcet's syndrome, known to be associated with B5'o. 2 of our patients with RCAS had B5, and Behcet's syndrome appeared to be developing in one of them. The combination B8, Bl5 is associated with SLE13, although Dvlocus antigens seem to have an even stronger association with this disease. 2 of our patients with RCAS had B8 and 4 with RAS had B15, but none had the combination (Table 3).5 ofour 20 patients had antibodies to doublestranded DNA, antibodies present in 60-85% of patients with SLE but absent in healthy persons. These results suggest a relationship between SLE and ROD. Only a few patients had abnormal immune responses. 4 were hypersensitive to food allergens, and 2 ofthese had high serum total IgE. In only I patient with RAS did the theory of FERGUSON et aI. 10 about the relationship between RAS and hypersensitivity to gluten find support. BRODY & SILVERMAN3 reported decreased serum concentrations of IgA in RAS, whereas LEHNER 19 reported raised levelsofboth IgA and IgG in RCAS. In our series, IgG, IgM and IgA were all within normal limits. That complement components C3 and C4 were within normal limits agrees with the observations of the aforementioned authors':". One patient (VT) had an impaired response to PHA-stimulation in vitro not previously shown in ROD. The serum factor that bound 1'5I-Iabelled PHA (leucoagglutinin) had the characteristics of ct,-macroglobulin. This agrees with a recent report! that PHA inhibitors in sera from patients with Hodgkin's disease may be related to
That washing of lymphocytes did not restore PHA reactivity suggests that serum inhibitors were of minor importance. Saturating amounts of PHA did not restore PHA reactivity to normal levels. Hence, suppressed PHA reactivity could not be attributed to the observed PHA binder, an observation further corroborated by the fact that the patient's serum did not suppress PHA reactivity oflymphocytes from healthy controls. The most intriguing feature of this case was the relation of suppressed PHA reactivity to an exacerbation of the clinical course of RCAS. This relation remains unexplained at present. Exacerbated RCAS may result in an increased production of serum PHA inhibitor, which would be reflected in an in vitro suppression of PHA. Whether this implies impaired T-cell function in vivo remains uncertain. NEWBLE et al.'3 detected serum inhibitors to PHA stimulation in the acute stage of viral hepatitis. They did not characterize the PHA binder in hepatitis serum, but they suggested that impaired PHA response in the acute stage of viral hepatitis was due to the virus. We found no evidence of viral infection in our patient, except one transient rise in serum antibody to herpes simplex. This patient was negative for Au antigen. Experiments performed by others on PHA inhibitors present in normal blood and in liver cirrhosis" showed that washing the lymphocytes restored normal PHA reactivity. This did not occur with our patient. Hence, the observed defect seems to reside in this patient's lymphocytes, although it could also be caused by serum factors or by factors adhering to the lymphocyte surface that are not easily removable by gentle washing. Neither did addition of normal serum to the incubation medium restore normal PHA reactivity. The exaggerated cutaneous reaction of this patient to both tuberculine and PHA is in apparent contradiction to the suppressed PHA reactivity in vitro. However, as a potent lymphocyte stimulator, PHA may be able to elicit cutaneous manifestations even when very

IMMUNOLOGY AND APHTHOUS ULCERATIONS

few active lymphocytes are present. Release of the migration inhibitory factor (MIF), found to be normal in our patient, and of the skin reactive factor may suffice. Failure of in vitro blastogenesis with PHA does not mean that such factors (lymphokines) are not released, since they may be produced independently of proliferation in vivo5 • The intense cutaneous effect of PHA and tuberculine in our patient might also be attributable to the cytotoxic effects of stimulated lymphocytes". Moreover, a similar reaction may take place in the oral mucosa, causing lymphocytes, locally stimulated by oral antigens, to release cytotoxic substance(s) and to act as killer cells. Among their patients with ROD, DONATSKY & DABELSTENs detected deposits of both IgG and C3 distributed in aphthous tissue. We found no deposits oflgG, and the deposits ofC3 detected were situated in and along the mucosal vessels. Our observations agree in part with those ofLUDERSCHMlDT et al," who, in addition to complement components, also detected either IgG or IgM in the aphthous tissue of some of their patients . Complement activation in the vessel wall, mediated either via immune complexes or via some other mechanism , can lead to vasculitis and ulceration in the mouth. Acknowledgement - We want to thank T. Weber, Minerva Institute, Kauniainen, Finland for performing the PHA stimulation.

References I. AMLOT, P. L. & UNGER, A.: The nature of phytohaemagglutinin (PHA) inhibition by serum from patients with Hodgkin's disease. European Immunology Meeting, Amsterdam, September 17-19 , 1975 (abstract). 2. BLAESE, R. M., WElDEN, P., OPPENHEIM, J. J. & WAlDMAN, T. A.: Phytohemagglutinin as a skin test for the evaluation of cellular immune competence in man. J. Lab . cu« Med. 1973: 81: 538-548. 3. BRODY, H . A. & SILVERMAN, S.: Studies on recurrent oral aphthae (I). Clin ical and laboratory comparisons. Oral Surg, 1969: 27: 27-34. 4. CHALLACOMBE, S. J., BATCHELOR, J. R. ,

29

KENNEDY, 1. A. & LEHNER, T.: HLA antigens in of recurrent oral ulceration. Archives Dermatology 1977: 113: 1717-1719. 5. DAVID, J . R. : Mediators produced by sensitized lymphocytes. Fed. Proc. 1971: 30 : /730-1735. 6. DOLBY, A. E.: HLA in recurrent aphthous stomatitis. The HLA & Disease Registry, Copenhagen ]975 . 7. DONATSKY, O. & BENDIXEN, G. : In vitro demonstration of cellular hypersentivity to streptococcus 2A in recurrent aphthous stomatitis by means of the leucocyte migration test. Acta Allergol. 1972: 27: 137-144. 8. DONATSKY, O. DABELSTEN, E.: Deposits of immunoglobulin G and complement C3 in recurrent aphthous ulcerations. Scand. J. Dent. Res. 1977: 85: 419-425. 9. DONATSKY, 0.: Recurrent aphthous stomatitis. Immunological aspects. An academic dissertation. Copenhagen 1978, pp. 15-17, 18-24. 10. FERGUSON, R., BASU, M. K. & ASQUlTII, P. & CooKB, W. T.: Jejunal mucosal abnormalities in patients with recurrent aphthous ulceration. Br . Med . J. 1976: 1: II-B . II. FRANCIS, T. C.: Recurrent aphthous stomatitis and Behcet's disease. A review. Oral Surg. 1970: 30 : 476-487. 12. HOLM, G. & PERLMANN, R .: Cytotoxic potential of stimulated human lymphocytes. J. Exp , Med. 1967: 125 : 721-736. 13. Hsu, C. C. S. & LEBVY, C. M.: Inhibition ofPH Astimulated lymphocyte transformation by plasma from patients with advanced alcoholic cirrhosis. Clln. Exp. Immunol. 1971: 8: 749-760. 14. HUNTER, W. M. & GREENWOOD, F. C.: Preparation ofiodine-131 labeled human growth hormone of high specific activity. Nature 1962: 194 : 495-496. 15. JONDAL, M ., HOLM, G. & WIGZELL, H.: Surface markers of human T and B lymphocytes. A large population of'lymphocytes forming non-immune rosettes with sheep red blood cells. J. Exp, Med. 1972: 136: 207-215. 16. LEHNER, T.: Recurrent aphthous ulceration and autoimmunity. Guy's Hasp, Gaz, 1965: 79 : 179182. 17. LEHNER, T. : Pathology of recurrent oral ulcer ation and oral ulceration in Behcet's syndrome: light, electron and fluoresence microscopy. J. Pathol. 1969: 97: 481-494. 18. LEHNER, T. : Characterization of mucosal antibodies in recurrent aphthous ulceration and Behcet's syndrome. Arch. Oral Bioi. 1969: 14 : 843-853 . 19. LEHNER, T.: Immunoglobulin estimation of blood and saliva in human recurrent oral

30

MALMSTROM, SALO AND FYHRQUIST

ulceration. Arch. Oral Bioi. 1969 : 14 : 351-364. 20. LOBITZ, JR., W. C., CIVATIE, J., THIVOLET, J., BBTUEL, H . & THORSBY, E.: Dermatology. In: DAUSSET, J. & SVEJGAARD, A. (eds.) : HLA and Disease. MunksgaardfWilkins 1977, pp . 126-148. 21. LunERSCHMIDT, CR., WOLFF, H . H. & SCHERER, R .: Aphten : histologische, immunfluoreszensund immunelektromikroskoplsche Studie zur Pathogenese. Der Hautarzt 1981 : 32: 364-369. 22. MILLER, M. F., GARFUNKEL, A . A., RAM, C. & SHIP, I. I. : Inheritance patterns in recurrent aphthous ulcers. Twin and pedigree data. Oral Surg, 1977: 43: 886-891. 23. NBWBLE, D. I., HOLMES, K . T., WANOEL, A. G. & FORBES, I. J.: Immune reactions in acute viral hepatitis. Clin. Exp, Immunol. 1975 : 20: 17-28. 24. PLATZ, P., RYDER, L. P. & DONATSKY, 0.: No deviations ofHLA-A and -B antigens in patients with recurrent aphthous stomatitis. Tissue Antigens 1976 : 8 : 279-280. 25. SHIP, I. I.: Inheritance of aphthous ulcers of the mouth. J. Dent. Res. 1965: 44 : 837-844. 26. SIRCUS, W., CHuRCH, R. & KELLEHER, J. : Recurrent aphthous ulceration of the mouth. A

study of th e natural history, aetiology and treatment. Quart . J. Med. 1957: 26: 235-249. 27. S0BORO, M. & BENDIXEN, G.: Human lymphocyte migration as a parameter of hypersensitivity. Acta Med. Scand. 1967 : 181: 247-256. 28. VARPELA, E., TIILIKAINEN, A. , VARPELA, M. & TuKJAINEN, P.: High prevalences of HLA-BI5 and HLA DW6 in patients with cryptogenetic fibrosing alveolitis. Tissue Antigens 1979 : 14: 6871. 29. World Health Organization: Application of the international classification of diseases to dentistry and stomatology. ICD-DA, Geneva 1973 .

Address: Maria Malmstrom Department of Oral Surgery Institute of Dentistry University of Helsintd Mannerhelmlntie 172 00280 H elsinki 28 Finland