Univariate and multivariate analyses were carried out to identify factors associated with the failure of 18-fluorodeoxyglucose positron emission tomography (FDG-PET) to correctly predict relapse-free survival in patients with nonseminomatous germ cell tumours. Ninety-three patients with negative postchemotherapy FDG-PET scan were analyzed in the retrospective study.
The FDG-PET result was validated by long-term follow-up and, in some patients, by resection of the residual tumour mass. The negative predictive value of FDG-PET was 81.7%.
Higher tumour marker levels and nodal stage at diagnosis, presence of residual mass, and teratoma or teratocarcinoma in the primary histology were associated with FDG-PET failure.