Over almost two decades, the prognostic impact of the decline of serum tumor markers after initiation of chemotherapy has been subject to considerable debate [8, 9]. However, only recently, it could be shown that the rate of decline of the serum tumor markers AFP and HCG according to their estimated serum half-lives can be used to further subdivide the groups of intermediate- and poor-risk patients. In a prospective intergroup phase III trial in the United States that randomized intermediate-risk and poor-risk patients to conventional-dose and high-dose first-line chemotherapy, it showed that those patients with a slow marker decline beyond the estimated marker half-life of AFP and HCG had inferior survival probabilities and profited more from treatment intensification compared to patients with marker declines according to their estimated marker half-lives [10]. This observation was recently confirmed in a randomized trial from the French Testicular Cancer Study Group that prospectively stratified patients according to their decline of serum AFP and HCG after an initial cycle of chemotherapy [11]. In this trial, patients with a slow marker decline had an inferior survival probabilities that could be significantly improved by early treatment intensification (see also Chap. 5). The German Testicular Cancer Study Group uses this information for an integrated, rational approach for first-line treatment of poor-risk patients (Fig. 8.1). An unresolved question, however, is the optimal calculation method of marker decline, which differed in the abovementioned trials and resulted in conflicting results depending on the algorithm used [12].

Prognostic factors have long been recognized to impact strongly on the results of first-salvage chemotherapy and have been the focus of several retrospective analyses [13–15]. In the first-salvage setting, the importance of prognostic factors is even greater as compared to initial diagnosis.

Although consistent prognostic variables have been identified, previous analyses suffered from substantial limitations. All have been too small to identify but a few variables reliably; some databases contained only incomplete information and were without source data verification; often external or internal validation of the results was missing; many analyses were based on outdated treatments that would no longer be considered standard today. Many of these obstacles have been overcome by the recent large collaboration of the International Prognostic Factor Study Group that included more than 1,500 patients in an analysis of prognostic factors for first-salvage treatment [15]. Seminoma histology was identified as a favorable prognostic factor. Adverse prognostic factors were (1) extragonadal primary tumors, in particular primary mediastinal non-seminomas; (2) less than complete remission or less than tumor-marker-negative partial remission to first-line treatment; (3) a progression-free interval of three months or less; (4) elevations of AFP at salvage, particularly if more than 1,000 ng/ml; (5) elevation of HCG at salvage to more than 1,000 U/l; and (6) the presence of liver, bone, or brain metastases (Table 8.2).