Treatment of Relapsed/Refractory Germ Cell Tumours: An Equipoise Between Conventional and High Dose Therapy

Abstract

Patients with relapsed germ cell tumors (GCT) are potentially curable despite the failure of initial cisplatin based therapy. The recent identification of robust prognostic markers has helped clarify this area of uncertainty. However, unlike the initial treatment of metastatic disease where cisplatin based combination therapy is standard of care, there is no consensus on treatment in the relapsed setting. Instead there is a genuine equipoise between high dose therapy (HDCT) and conventional dose therapy (CDCT) in this relapsed setting. The randomised data fails to support the use of HDCT although retrospective analysis consistently shows it may be superior, especially if tandem or triple high dose therapy is used. This has resulted in different approaches worldwide with some stating HDCT should not be used outside of a trial and others suggesting it is standard of care. Re-challenging relapsed patients with conventional dose cisplatin based therapy would seem to be counterintuitive, but the addition of drugs such as ifosamide and paclitaxel results in genuine activated in selected patients with relapsed disease. This appears particularly relevant as HDCT is associated with significant morbidity and mortality, which was highlighted in a randomised trial. However large single institution data suggests this treatment related toxicity may not be such as concern and treatment related mortality can be as low as only 3 % in selected centres. Overall there is a need to answer this question of CDCT v.s. HSCT definitively. For this reason a prospective global randomised trial (TIGER trial) comparing conventional dose therapy (paclitaxel, ifosamide and cisplatin [TIP]) and triple high dose therapy (paclitaxel and ifosamide followed by high dose carboplatin and etoposide [TI CE]) is planned for 2012 (TIGER trial). Due to the rarity of this disease and complexities associated with a global trial, this study faces many challenges. Nevertheless it is hoped that it will finally address this area major of uncertainty in GCTs.