DNA repair biomarkers to guide usage of combined PARP inhibitors and chemotherapy: A meta-analysis and systematic review

Zoe Phan,Caroline E Ford,C Elizabeth Caldon

DOI: https://doi.org/10.1016/j.phrs.2023.106927

Abstract:Purpose: The addition of PARP inhibitors to chemotherapy has been assessed in > 80 clinical trials across multiple malignancies, on the premise that PARP inhibitors will increase chemotherapy effectiveness regardless of whether cancers have underlying disruption of DNA repair pathways. Consequently, the majority of combination therapy trials have been performed on patients without biomarker selection, despite the use of homologous recombination deficiency to dictate use of PARP inhibitors in the maintenance setting. An unresolved question is whether biomarkers are needed to identify patients who respond to combination PARP inhibitors and chemotherapy. Methods: A systematic literature review identified studies using PARP inhibitors in combination with chemotherapy versus chemotherapy alone, where the study included a biomarker of DNA repair function (BRCA1, BRCA2, homologous recombination deficiency test, ATM, ERCC1, SLFN11). Hazard ratios (HR) were pooled in a meta-analysis using generic inverse-variance, and fixed or random effects modelling. Subgroup analyses were conducted on biomarker selection and type of malignancy. Results: Nine studies comprising 2547 patients met the inclusion criteria. Progression-free survival (PFS) was significantly better in patients with a DNA repair biomarker (HR: 0.57, 95% CI: 0.48-0.68, p < 0.00001), but there was no benefit in patients who lacked a biomarker (HR: 0.94, 95% CI: 0.82-1.08, p = 0.38). Subgroup analysis showed that BRCA status and SLFN11 biomarkers could predict benefit, and biomarker-driven benefit occurred in ovarian, breast and small cell lung cancers. The addition of PARP inhibitors to chemotherapy was associated with increased grade 3/4 side effects, and particularly neutropenia. Conclusions: Combination therapy only improves PFS in patients with identifiable DNA repair biomarkers. This indicates that PARP inhibitors do not sensitise patients to chemotherapy treatment, except where their cancer has a homologous recombination defect, or an alternative biomarker of altered DNA repair. While effective in patients with DNA repair biomarkers, there is a risk of high-grade haematological side-effects with the use of combination therapy. Thus, the benefit in PFS from combination therapy must be weighed against potential adverse effects, as individual arms of treatment can also confer benefit.