Incidence of neurotoxicity and ototoxicity with cisplatin vs. non-cisplatin regimens: A meta-analysis with subgroup analyses based on renal eligibility criteria.

e15555 Background: Using glomerular filtration rate (GFR) to screen patients for trial eligibility may reduce the risk of cisplatin (CIS) associated nephrotoxicity compared to serum creatinine (SCr). The impact of using GFR or SCr on incidence of neurotoxicity and ototoxicity remains unknown. This meta-analysis compared incidence of neurotoxicity and ototoxicity among trials of solid tumors treated with CIS reporting nephrotoxicity when renal function was assessed using SCr or GFR. Methods: A PubMed literature search identified randomized trials comparing CIS to non-CIS containing chemotherapy regimens. Studies were included if performed from 1990-2010, used SCr or GFR as eligibility criteria, and reported incidence of WHO or NCI grade ≥3 nephrotoxic events. Separate analyses were performed on studies reporting grade ≥3 neurotoxic or ototoxic events. For neurotoxicity, studies comparing CIS to other neurotoxic drugs were excluded. Review articles, observational studies, phase 1 studies, non-randomized trials, studies without a comparator group, or studies not reported in English were excluded. Relative risk (RR) associated with CIS vs. non-CIS regimens was calculated and subgroup analyses were performed for studies using SCr, GFR, and either SCr or GFR for screening. Results: The literature search identified 2359 studies with 18 studies meeting all inclusion and exclusion criteria for neurotoxicity (N=3441 patients) and 9 for ototoxicity (N=2947). The RR for developing neurotoxicity was 1.27 (p=0.50) and ototoxicity was 2.32 (p=0.10) for CIS vs. non-CIS regimens. For neurotoxicity, the RR when SCr was used as eligibility criteria was 2.31 (p=0.11), 0.71 (p=0.54) when GFR was used, and 1.01 (p=0.99) when either was used. For ototoxicity, the RR for SCr was 2.33 (p=0.21) and 3.47 (p=0.20) for either. Studies using GFR reported zero ototoxic events. Conclusions: The risk of neurotoxicity and ototoxicity did not significantly change when SCr or GFR was used to screen patients. However, these analyses were performed on a subgroup of studies reporting nephrotoxicity and further research on all studies of neurotoxicity and ototoxicity is warranted.