Purpose: The response of rectal cancers to neoadjuvant chemoradiotherapy is variable. Tumor hypoxia reduces the effectiveness of both radiation therapy and chemotherapy and is a well-known risk factor for tumor radioresistence. We hypothesized that imaging with the novel hypoxia-detecting agent, (60)Cu-diacetyl-bis (N(4)-methylthiosemicarbazone) ((60)Cu-ATSM), previously validated in cervical and lung cancers, would predict the response of rectal cancers to neoadjuvant chemoradiotherapy and prognosis.
Methods: Patients with locally invasive (T2-4) primary or node-positive rectal cancer located <12 cm from the anal verge were recruited for this pilot study. Pretreatment tumor size and stage were determined by endorectal ultrasonography, CT, and magnetic resonance imaging. Eleven patients also underwent clinical positron emission tomography with (18)F-fluorodeoxyglucose at the discretion of the treating clinician. The primary tumor was imaged by positron emission tomography with (60)Cu-ATSM, and accumulation of the tracer was measured semiquantitatively by determining the tumor-to-muscle activity ratio. Neoadjuvant chemoradiotherapy was then administered (within 2 weeks of (60)Cu-ATSM-positron emission tomography) and consisted of 45 Gy given in 25 fractions to the pelvis with continuous intravenous infusion of 5-fluorouracil (225 mg/m(2)/day). Proctectomy was performed six to eight weeks after neoadjuvant chemoradiotherapy and the tumor submitted to pathology for size measurement and staging. Tumor-to-muscle activity ratios were compared with tumor (18)F-fluorodeoxyglucose uptake, tumor response to neoadjuvant chemoradiotherapy, and with patient survival.
Results: Nineteen patients were enrolled in the study, two of whom were excluded from final analysis (1 death during neoadjuvant chemoradiotherapy and 1 tumor perforation during neoadjuvant chemoradiotherapy requiring emergent surgery). Of the 17 remaining patients, 14 had a reduction in tumor size and 13 were downstaged. The median tumor-to-muscle activity ratio of 2.6 discriminated those with worse prognosis from those with better prognosis. Both overall and progression-free survivals were worse with hypoxic tumors (tumor-to-muscle activity ratio >2.6) than with nonhypoxic tumors (tumor-to-muscle activity ratio <or=2.6; both P < 0.05). In addition, 2 of the 3 tumors with no change in size had tumor-to-muscle activity ratios >2.6 (positive predictive value 66 percent), whereas 6 of 14 with decreased size had tumor-to-muscle activity ratios >2.6 (negative predictive value 57 percent). Three of the 4 tumors not downstaged had tumor-to-muscle activity ratios >2.6 (positive predictive value 75 percent), whereas 5 of 13 downstaged tumors had tumor-to-muscle activity ratios >2.6 (negative predictive value 62 percent). The mean tumor-to-muscle activity ratio for downstaged tumors (2.2) was significantly lower than that of nondownstaged tumors (3.3) (P = 0.03). The difference in mean tumor-to-muscle activity ratio between downsized (2.3) and nondownsized (2.9) tumors did not reach statistical significance (P = 0.36). Tumor (18)F-fluorodeoxyglucose uptake (n = 11) did not correlate with (60)Cu-ATSM uptake (r = 0.4; P = 0.9) and there was no significant difference in mean tumor (18)F-fluorodeoxyglucose uptake between patients with hypoxic tumors and those with normoxic tumors (P = 0.3).
Conclusions: The results of this small pilot study suggest that (60)Cu-ATSM-PET may be predictive of survival and, possibly, tumor response to neoadjuvant chemoradiotherapy in patients with rectal cancer. A larger Phase II study is warranted to validate these results.