CARCINOGENIC POTENTIAL OF SILICONE BREAST IMPLANTS - A CONNECTICUT STATEWIDE STUDY

To clarify the carcinogenic potential of silicone breast implants, 680 implant procedures performed on women in Connecticut with no prier hi story of cancer were correlated with the subsequent development of pri mary breast and nonbreast cancers. Neoplastic events after the placeme nt of silicone breast implants during the 13-year interval from Octobe r 1, 1980, through September 30, 1993, were quantified using a retrosp ective, linked-registry method. ICD-9-CM discharge codes contained in the Uniformed Hospital Discharge Data Sets (UHDDS) from 34 hospitals a cross Connecticut were linked to procedure codes for unilateral and bi lateral implants, and to medical histories for new malignancies after the implant procedures. Data were cross-linked to the Connecticut Tumo r Registry to confirm the clinical history of each cancer. The rates o f breast and nonbreast cancers in patients with silicone br east impla nts were compared with those of a control population drawn from the UH DDS of 1022 women undergoing sterilization by tubal ligation between 1 981 to 1985. Ages (mean +/- SD) were similar in the implant group (34 +/- 10 years) and in the sterilization group (32 +/- 6 years). The mea n follow-up in the implant group (4.6 years) was also similar to that of the control group (5.4 years). Compared with the control group, the implant group demonstrated lower rates of breast cancer (0.59 versus 0.88 percent, p = 0.35) and nonbreast cancer (0.59 versus 2.7 percent, p = 0.001). Correspondingly, the implant group had a lower relative r isk of breast cancel(relative risk = 0.67, 95 percent, confidence inte rval = 0.20 to 2.17) and nonbreast cancer (relative risk = 0.21, 95 pe rcent, confidence interval = 0.07 to 0.60). Based on these data, it wa s concluded that silicone breast implants are not carcinogenic, becaus e they are not associated with increased rates of either breast or non breast cancers. The validity and threats to the conclusions are discus sed, and the results are placed into contest with similar findings fro m other studies.