AIMS: To examine a large series of patients in whom both red cell autoantibodies and carcinoma are present; and to determine whether this rare occurrence is a true association or a chance event. METHODS: The laboratory records of 160 patients (76 men, 84 women; mean age 68 years) with erythrocyte autoantibodies and confirmed carcinoma were examined for site of tumour origin and clinical and immunohematological findings. To test whether the concomitant occurrence of autoantibodies and carcinoma was fortuitous, data on total population and carcinoma incidence were included in a chi 2 analysis. RESULTS: The association was significant (chi 2 = 97.5, p < 0.0005); erythrocyte autoantibodies and carcinoma were found together 12-13 times more often than expected from their relative frequencies. Autoantibodies occurred with a variety of carcinomas, particularly those of breast, lung, colon, rectum, and prostate; this largely reflected tumour incidence. Adenocarcinoma, squamous, anaplastic, and transitional cell types were all represented. Warm, cold, and mixed autoantibodies were not associated with particular tumour sites or histology. Eighty six patients had haemolysis of varying severity, 37 had metastatic disease, and 28 died within a few months of presentation. CONCLUSIONS: The presence of erythrocyte autoantibodies and carcinoma in the same patient is a true association and probably reflects a fundamental disturbance in immune homeostasis. It tends to occur with a large tumour mass and metastatic disease, and generally indicates a poor prognosis.
AIMS: To examine a large series of patients in whom both red cell autoantibodies and carcinoma are present; and to determine whether this rare occurrence is a true association or a chance event. METHODS: The laboratory records of 160 patients (76 men, 84 women; mean age 68 years) with erythrocyte autoantibodies and confirmed carcinoma were examined for site of tumour origin and clinical and immunohematological findings. To test whether the concomitant occurrence of autoantibodies and carcinoma was fortuitous, data on total population and carcinoma incidence were included in a chi 2 analysis. RESULTS: The association was significant (chi 2 = 97.5, p < 0.0005); erythrocyte autoantibodies and carcinoma were found together 12-13 times more often than expected from their relative frequencies. Autoantibodies occurred with a variety of carcinomas, particularly those of breast, lung, colon, rectum, and prostate; this largely reflected tumour incidence. Adenocarcinoma, squamous, anaplastic, and transitional cell types were all represented. Warm, cold, and mixed autoantibodies were not associated with particular tumour sites or histology. Eighty six patients had haemolysis of varying severity, 37 had metastatic disease, and 28 died within a few months of presentation. CONCLUSIONS: The presence of erythrocyte autoantibodies and carcinoma in the same patient is a true association and probably reflects a fundamental disturbance in immune homeostasis. It tends to occur with a large tumour mass and metastatic disease, and generally indicates a poor prognosis.