Theresa Wald ^1,2* Gunnar Wichmann ^1,2 Veit Zebralla ^1,2 Matthaeus Stoehr ^1,2 Markus Pirlich ^1,2 Susanne Wiegand ³ Viktor Kunz ^1,2 Andreas Dietz ^1,2

• ¹ University Hospital Leipzig, Leipzig, Germany
• ² Comprehensive Cancer Center Central Germany, Leipzig, Lower Saxony, Germany
• ³ University Medical Center Schleswig-Holstein, Kiel, Schleswig-Holstein, Germany

Introduction The larynx-organ preservation (LOP) trial DeLOS-II enrolled n = 173 patients with advanced laryngeal/hypopharyngeal squamous cell carcinoma (LHSCC) amenable (only curatively resectable) through total laryngectomy (TL) to receive induction-chemotherapy (IC) with TPF (docetaxel [T], cisplatin [P], and 5-fluoruracil [F]) (arm A, 85 patients) or additional cetuximab (E) weekly (arm B, 88 patients). Responders with endoscopic estimated surface shrinkage (ETSS) ≥ 30% after 1 cycle IC (IC-1) received further 2 cycles IC followed by radiotherapy (RT), whereas TL was recommended for non-responders. Arm B failed to show superior 24-months laryngectomy-free (LFS) and overall survival (OS), the protocol-specified primary and secondary endpoints. Ten years after last-patient-out we were interested in long-term outcome of our clinic’s DeLOS-II patients. Methods Our cohort of 52 DeLOS-II patients accrued between 2007-2012 included 27/25 patients randomized to arms A/B. F was omitted because of severe toxicity with amendment 2 of DeLOS-II protocol leading to 21 and 31 patients receiving TPF or TP IC-backbone, respectively. Follow-up data was collected using electronic health records and information from the German Centre for Cancer Registry Data to evaluate long-term LFS and OS in treatment groups. Results According to ETSS ≥ 30%, 42 patients (80.8%; 21/21 corresponding to 77.8%/84.0% in arms A/B) were responders to IC-1 and underwent the LOP attempt. Recommending early TL to non-responders (ETSS < 30%), 8 patients (5/3 in A/B) underwent early TL. At 125 months, 22 (8/14) patients were alive, 17 (6/11) with functioning and 5 (2/3) without larynx. Arm B had superior OS (p = 0.023). Disease-specific (DSS) and tumor-specific survival were not different, whereas non-cancer-related survival (NCRS) was impaired in arm A (p = 0.018). Receiving TP or TPF IC did not significantly influence survival. Pairwise comparing OS of patients receiving TP, TPF, TPE, and TPFE revealed a benefit from cetuximab in TPE vs. TP (p = 0.020). Conclusion While the per-protocol DeLOS-II results earlier reported comparable 24-months LFS and OS in arms A and B, our sub-cohort’s long-term follow-up data demonstrate superior 125-months outcome in arm B.