Linkage analysis of malignancy-associated sarcoidosis

Abstract

Study objective: To determine whether a reported association between sarcoidosis and malignancy can be supported.

Design: In this retrospective survey in a general community setting, we crossmatched the Kaiser Permanente Northwest Region (KPNW) Tumor Registry comprising 3 x 10(4) cases observed over 32 years against a sarcoidosis registry of 243 cases observed over 24 years. We used linkage criteria, eg, previously reported associated malignancies, late age onset of sarcoidosis, and close temporal proximity, to identify joint occurrences suggesting an etiologic relationship. We assessed the feasibility of detecting an association of sarcoidosis and Hodgkins disease (HD) by computing the required sample size based on incidence estimates of sarcoidosis and HD in this population. Medical records of 241 persons identified by the Tumor Registry as having HD were reviewed.

Setting: KPNW, a health maintenance organization.

Population: We studied 6.8 x 10(6) KPNW member-years from 1960 to 1992 at risk for malignancy; 5.6 x 10(6) member-years from 1971 to 1992 at risk for sarcoidosis.

Results: Six of the 11 (55%; 95% confidence interval, 26 to 84) correctly classified sarcoidosis and malignancy (S&M) cases met two or more linkage criteria. A seventh case, incorrectly classified as sarcoidosis, exhibited a necrotizing systemic granulomatous process, believed to represent a response to a fatal lymphoproliferative disorder. The mean age (40 years) of the 11 patients with S&M was 9 years higher than the mean age of the patients with sarcoidosis in the population from which they were drawn. The stage of sarcoidosis at diagnosis among the 11 S&M cases was exclusively 0 and I and thus divergent from expected values in this population. We found no instances of sarcoidosis accompanying HD.

Conclusion: Linkage analysis provides evidence that S&M may be etiologically related in at least a quarter of cases in which both are present.