The sera of 127 non-diabetic children after mumps-infection were investigated for the presence of islet cell antibodies and islet cell surface antibodies. The study also included 4 children who developed diabetes mellitus shortly after an active mumps vaccination. 21 of the non-diabetic children and four of the vaccinated children exhibited islet cell cytoplasmic antibodies. Islet cell surface antibodies were observed more frequently, namely in 43 out of 68 patients studied after mumps infection and in 32 out of 44 patients studied after different viral diseases. With one exception, none of the mumps-infected children and none of the other viral infected patients developed diabetes mellitus.

Cytoplasmic islet-cell antibodies, insulin antibodies, islet-cell surface antibodies and islet-cell specific cytotoxicity were determined in serum of the following groups: 131 patients with type I diabetes, 19 with type II diabetes, 29 with mumps, 29 with enterovirus infections, 18 with measles and 28 healthy controls. Cytoplasmic islet-cell antibodies were found predominantly in type I diabetics. Islet-cell surface antibodies, on the other hand, were relatively frequently (60-80%) present in sera of both diabetics and patients with various virus infections. Islet-cell specific cytotoxicity in vitro was found not only in sera of diabetics, but also of patients with mumps or enterovirus infections. Sera of five patients with measles, however, had cytotoxic reactions comparable to those of the controls. These results suggest that cytotoxic antibody reactions against islet cells in vitro occur also in sera of non-diabetic patients. Under certain circumstances, infections which induce such immune reactions may be of significance in the pathogenesis of diabetes.

Serum samples of patients suffering from diabetes mellitus were tested for complement-fixing and non complement-fixing islet cell antibodies, antinuclear antibodies and circulating immune complexes. There was no correlation between circulating immune complexes or antinuclear antibodies and secondary diabetic complications. A close relationship was found between the ICA titer and complement fixation of ICA. The incidence of ICA at the onset of the disease was higher in the patients under the age of 10 (85%) and decreased with increasing age up to 45% in patients with onset above age 20. In five patients being positive and four patients being negative for ICA at onset of disease, changes and fluctuations in antibody titers were observed over 38 months. Since manifestation of diabetes mellitus is believed to be an endpoint of a long lasting autoimmune process, our observations indicate that the autoimmune phenomena are merely indicators of ongoing autoimmune reactions not necessarily reflecting the state of autoaggression or islet cell destruction.

Cytoplasmic islet-cell antibodies, insulin antibodies, islet-cell surface antibodies and islet-cell specific cytotoxicity were determined in serum of the following groups: 131 patients with type I diabetes, 19 with type II diabetes, 29 with mumps, 29 with enterovirus infections, 18 with measles and 28 healthy controls. Cytoplasmic islet-cell antibodies were found predominantly in type I diabetics. Islet-cell surface antibodies, on the other hand, were relatively frequently (60-80%) present in sera of both diabetics and patients with various virus infections. Islet-cell specific cytotoxicity in vitro was found not only in sera of diabetics, but also of patients with mumps or enterovirus infections. Sera of five patients with measles, however, had cytotoxic reactions comparable to those of the controls. These results suggest that cytotoxic antibody reactions against islet cells in vitro occur also in sera of non-diabetic patients. Under certain circumstances, infections which induce such immune reactions may be of significance in the pathogenesis of diabetes.

Islet cell antibodies were investigated in 127 non-diabetic children after mumps infection and in four out of seven children who developed diabetes mellitus shortly after active mumps vaccination. Twenty-one of the children who had mumps and all four vaccinated children who were tested had islet cell cytoplasmic antibodies. In contrast, islet cell surface antibodies were detected in 43 out of 68 patients with mumps infection and in 32 out of 44 patients with other viral diseases. All but one mumps-infected child and all the other viral infected patients investigated did not develop diabetes mellitus. The mumps-infected ICA positive children did not show those HLA-frequencies associated with Type 1 diabetes.

To connect mumps and diabetes mellitus in children is an old problem in medical literature. The typical occurrence of ICA at the onset of diabetes in children, as well as the incidence of ICA approximately 3 weeks after mumps infection support the hypothesis of a direct relationship between virus infection and diabetes. But the mumps infection alone is not the key factor. Mumps vaccination may not provide protection against diabetes mellitus, it may even provoke it. (Genetic determination, expressed by the HLA-phenotype in all the patients reported, does not allow a differentiation.)

To investigate whether the development of islet-cell antibodies (ICA) in the course of mumps infection is associated with a "diabetes-like" immunogenetic condition, 45 children with mumps complications as well as 56 children with insulin-dependent diabetes mellitus (IDDM) were typed for HLA ABC and DR antigens. ICA were detected in 14 out of 35 mumps patients. In the IDDM group, significant deviations from antigen frequencies of normal controls were observed for HLA Bw39, DR2, DR3, and DR4. In contrast, in ICA positive mumps patients, the frequency of these antigens was normal, but Aw24 was significantly increased. Thus, no immunogenetic similarities of both groups of patients could be detected.

The detection of islet cell antibodies lead to an increasing interest in autoimmune mechanisms in Typ I diabetes mellitus. Other phenomena such as insulitis in juvenile diabetics and in experimental animals, cellular immune reactions and concomitant antibodies against other endocrine organs, antinuclear antibodies and circulating immune complexes supported these suggestions. HLA-association and viral infections could be predisposing and inducing factors, although there are no clear correlations to any viral infection in a larger number of patients so far. For clinical and therapeutic purposes there are not enough sufficient criteria to demonstrate a pathologic autoimmune process in patients before developing diabetes. Up to now there is no realistic possibility and justification for starting an early immunosuppressive therapy.

In 1972, Wolcott and Rallison described three siblings with a combination of infancy-onset diabetes mellitus and multiple epiphyseal dysplasia. We have observed a brother and sister with the same disorder. The chondro-osseous lesions are those of a spondylo-epiphyseal dysplasia. The diabetes mellitus is relatively mild. Histologic and electron microscopic studies of chondro-osseous tissue show findings similar to those in other epiphyseal and spondylo-epiphyseal dysplasias. In addition, however, atypical collagen-like fibres are found inside and outside chondrocytes. Collagen production seems to be normal in cultured fibroblasts. From the available data it appears that the association of characteristic chondro-osseous and endocrine abnormalities is non-random and that the lesions are independent manifestations of a pleiotropic gene. We propose to call this disorder the Wolcott-Rallison Syndrome.

Mumps infection was the first and still is the most frequently mentioned virus infection in connection with the development of type I diabetes mellitus. In 21 mumps-infected children islet cell antibodies were demonstrated during 1979 to 1981. These positive results could not be obtained in the subsequent years. Islet cell surface antibodies, however were detectable in a high percentage of patients infected with mumps as well as with other viruses. All but one of those patients showing anti-islet cell immune activity did not develop diabetes mellitus. Their HLA constellation was not comparable with those of type I diabetic patients. In contrast the HLA constellation of 7 children who developed diabetes mellitus shortly after mumps vaccination did include DR4 in each case as well as DR3 in 3 cases.