An assessment of mid-chemotherapy CT scanning as a predictor of a residual retroperitoneal mass in good prognosis metastatic teratoma of the testis

Abstract

4573 Background: There are 2 widely used approaches to treatment of stage 2 non-seminomatous germ cell tumours of the testis (NSGCT); in the UK, 3 or 4 cycles of chemotherapy (C) are generally followed by surgery in those (∼25%)with residual masses >1cm. In some US centres, intial surgery is followed by 2 cycles of C, with similarly good results (Williams et al, 1987). A critical view of US practice is that 2/3 of pts have unnecessary surgery. A US view of UK practice would be that 25% of pts have excessive C. We assessed whether a mid-C CT scan could predict which pts would require surgery, defined as those likely to have a residual post-C mass ≥ 1cm in maximum transverse diameter (MTD), and might thus be considered for early surgery after 2 cycles of C. Methods: Pts were recruited within MRC/EORTC TE20 (de Wit et al 2001), comparing 3 vs 4 x BEP and a 3 vs 5-day schedule. Pts had histologically proven seminoma (S) or NSGCT and were IGCCCG good prognosis. For this sub-study, pts had a pre-C retroperitoneal (RP) mass ≥ 1cm in MTD. Pts had CT scans at baseline, 4–5 weeks after the start of their 1st cycle of C and then 4 weeks post C to measure the MTD of the largest RP mass. Sensitivity and specificity were used via ROC curve analysis to assess the accuracy of predicting residual masses of <1cm by looking at various cut-offs in both absolute mid-C diameter (D) & pre- to mid-C shrinkage (SH). Results: 77 pts (76 testis 1o, 1 RP 1o). Using MTD at mid-C as a predictor of residual RP mass <1cm gave an optimal cutpoint of 2cm (sens=80%, spec=85%). Investigating SH from pre-C to mid-C as a predictor of residual RP mass <1cm gave an optimal SH of 70% (sens=57%, spec=82%). Combining SH and absolute D did not substantially improve on rules using D alone. Conclusions: This small study suggests a potential role for mid-C CT in selecting certain pts for early surgery. Our results suggest the best predictor of post-C residual mass requiring resection is one where pts with an RP mass >2cm by mid C are recommended to have their mass resected. This means that 80% of pts will be correctly predicted to have a residual RP mass <1cm, and should continue their C and 85% will be correctly predicted to have a residual RP mass ≥ 1 cm and thus require resection. No significant financial relationships to disclose.