Purpose: There are several management options for patients with clinical stage I (CS1) nonseminomatous germ cell tumors (NSGCT); this study examined whether an 18fluorodeoxyglucose positron emission tomography (18FDG PET) scan could identify patients without occult metastatic disease for whom surveillance is an attractive option.
Methods: High-risk (lymphovascular invasion positive) patients with CS1 NSGCT underwent 18FDG PET scanning within 8 weeks of orchidectomy or marker normalization. PET-positive patients went off study; PET-negative patients were observed on a surveillance program. The primary outcome measure was the 2-year relapse-free rate (RFR) in patients with a negative PET scan (the negative predictive value). Assuming an RFR of 90% to exclude an RFR less than 80% with approximately 90% power, 100 PET-negative patients were required; 135 scanned patients were anticipated.
Results: Patients were registered between May 2002 and January 2005, when the trial was stopped by the independent data monitoring committee due to an unacceptably high relapse rate in the PET-negative patients. Of 116 registered patients, 111 underwent PET scans, and 88 (79%) were PET-negative (61% of preorchidectomy marker-negative patients v 88% of marker-positive patients; P = .002); 87 proceeded to surveillance, and one requested adjuvant chemotherapy. With a median follow-up of 12 months, 33 of 87 patients on surveillance relapsed (1-year RFR, 63%; 90% CI, 54% to 72%).
Conclusion: Though PET identified some patients with disease not detected by computed tomography scan, the relapse rate among PET negative patients remains high. The results show that 18FDG PET scanning is not sufficiently sensitive to identify patients at low risk of relapse in this setting.