Carcinoma of the rectum - Possible cellular predictors of metastatic potential and response to radiation therapy

BACKGROUND: Preoperative radiation therapy can markedly improve local control of rectal carcinoma. However, some tumors do not respond well to moderate doses of preoperative radiation and would be better served by more aggressive preoperative treatment (e.g., chemoradiotherapy). Cellular predictors of responsiveness to radiation can help to select lesions for more aggressive treatment. In addition, there is a need for cellular predictors of metastatic potential. This is particularly important in the setting of preoperative radiation-downstaging by preoperative treatment can obscure the true pathologic stage of a tumor and confound the usual selection criteria for postoperative chemotherapy. PURPOSE: This study was undertaken to determine if proliferating cell nuclear antigen (PCNA), p53, DNA ploidy, and S-phase fraction are associated with response to radiation and/or risk for distant metastatic disease and to determine if these cellular markers are best evaluated from preradiation biopsy specimen or the larger (but possibly altered) final surgical specimen. MATERIALS AND METHODS: Archival specimens from 23 cases of ultrasound T3 or T4 rectal carcinoma treated preoperatively with radiation therapy were reviewed. Eligible lesions had preradiation biopsy specimens of sufficient site for flow cytometric review of archival tissue. Factors considered included PCNA positivity, presence of mutant nuclear p53, more than 30 percent tumor cells in S-phase, and presence of aneuploidy. RESULTS: With a median follow-up of three years, overall freedom from relapse was 83 percent, with all but one failure being extrapelvic. PCNA positivity in the preradiation specimen was significantly (P = 0.025) associated with a greater risk of tumor recurrence. In addition, there was a trend to greater likelihood of ''probable downstaging'' (defined as surgical T stage less than preradiation ultrasound T stage) for lesions that were PCNA-negative or lesions with normal p53. Biomarkers measured in the postradiation surgical specimen were not associated with either freedom from relapse or response to radiation. Radiation treatment appeared to produce false-negatives in the final specimen. Thus, there were significantly more specimens converting from PCNA-positive to PCNA-negative after preoperative radiation than would be expected solely on thr basis of sampling errors (P = 0.004). Similar results were found for abnormal p53 findings (P = 0.02). CONCLUSIONS: Prospective studies of biomarkers should be based on pretreatment specimens if preoperative radiation is given. For carcinoma of the rectum, PCNA and p53 may be useful predictors of both metastatic potential and responsiveness to radiation.