The incidence of myelodysplastic syndrome in patients receiving poly-ADP ribose polymerase inhibitors for treatment of solid tumors: A meta-analysis.

Abstract

3641 Background: Clinical trials have reported improved outcomes with PARPi (poly [adenosine diphosphate–ribose]-ADP polymerase inhibitor) therapy in ovarian, breast, pancreatic and lung cancers. There is concern that PARPi therapy may cause myelodysplastic syndrome (MDS). In this meta-analysis we seek to quantify the risk of MDS among patients treated with PARPi for solid tumor malignancies. Methods: We searched Medline, Embase, and Cochrane databases (up to January 6, 2020) to abstract randomized controlled trials that include a PARPi in the experimental arm in solid tumors. Combinations included PARPi versus (vs.) placebo, PARPi vs. cytotoxic treatment, and PARPi with cytotoxic treatment vs. cytotoxic treatment. We used to time-to-event curves to estimate person-time and calculated the incidence of MDS among all studies. We used random-effects Poisson regression models to estimate pooled incidence risk ratio (RR) for developing MDS. Results: We identified 14 studies, 10 in ovarian, 3 in breast, and 1 in pancreatic cancer patients. Of 5,646 patients, 62.3% received a PARPi alone or in combination with chemotherapy or bevacizumab, and 37.8% received treatment consisting of placebo alone or with chemotherapy or bevacizumab. PARPi were investigated as an upfront treatment in 2,827 patients, and as treatment for recurrence in 2,819 patients. The incidence of MDS was 6.73 cases vs. 3.85 per 1000 person-years in patients receiving PARPi as compared to control corresponding to a 3-year cumulative incidence of 2.0% and 1.1%. Accounting for intra-study clustering, PARPi use was associated with a 60% increase in risk (incidence RR 1.60, 95% Confidence Interval [CI] 0.89-2.87) of MDS compared to control. In the upfront setting, patients randomized to PARPi were twice as likely to develop MDS (RR 2.08, 95%, CI 1.39-3.64). Among patients treated for recurrence, the risk of MDS appeared to be similar among patient randomized to PARPi or control treatment (RR 1.13, 95% CI 0.35-3.64). In studies that compared PARPi in combination with other cytotoxic treatment vs. cytotoxic treatment alone, PARPi was associated with a large risk of MDS (RR 5.08, 95% CI 1.36-19.03). Conclusions: In pooled estimates from randomized controlled trials in solid tumors PARPi treatment appears to be associated with an increased incidence of MDS particularly in the upfront setting and when combined with cytotoxic treatment. Despite pooling 14 randomized trials our estimates remain imprecise due to the rarity of MDS.