Poly (ADP-Ribose) Polymerase Inhibition for Chemotherapy-Induced Peripheral Neuropathy: A Meta-Analysis of Placebo-Controlled Trials

Introduction

Pain, numbness, and tingling in the hands and feet characterize chemotherapy-induced peripheral neuropathy. This set of symptoms occurs in 20%–60% of cancer patients who receive agents such as paclitaxel and oxaliplatin as well as others. ¹ Symptoms can last for years. Although distressing in their own right, these symptoms also negatively impact functionality, thus predisposing patients to falls and other causes of injury. ²

Although interventions for neuropathy can provide modest palliation, no intervention provides major durable symptom relief. Interestingly, Ta et al. studied poly (ADP-ribose) polymerase (PARP) inhibition for the palliation of chemotherapy-induced peripheral neuropathy. ³ These investigators focused on veliparib, one of several PARP inhibitors. They amplified nerve-related pain in animals exposed to neuropathy-inducing chemotherapy and observed that veliparib had a favorable modulating effect on pain thresholds. Similarly, in many other preclinical settings, others have reported that PARP inhibitors diminish neuroinflammation with corresponding reductions in neuropathic pain.^4–21 In one of the few cancer clinical studies on this topic, Balko et al. re-examined a 34-patient pilot trial and reported that, among patients who had previously received paclitaxel, 21% described an improvement in pain after exposure to veliparib. ²²

The foregoing evidence, in concert with the increased clinical use of PARP inhibitors in the treatment of ovarian and other cancers, prompted us to advance and test the hypothesis that PARP inhibition reduces the risk of developing chemotherapy-induced peripheral neuropathy. This meta-analysis was intended to generate further clinical data on the potential role of PARP inhibition in the prevention or palliation of chemotherapy-induced peripheral neuropathy.

Methods

The primary goal of this meta-analysis was to assess the purported relationship between PARP inhibition and prevention/palliation of peripheral neuropathy in a clinical setting. This meta-analysis closely adhered to the PRISMA criteria for the conduct and reporting of meta-analyses. ²³

PubMed was used as the primary search engine because of its comprehensiveness; the following search terms were used: olaparib, veliparib, niraparib, PARP inhibitor, ovarian cancer, fallopian cancer, peritoneal cancer, breast cancer, cancer, malignancy, BRCA, paclitaxel, palliative, neuropathy, and pain. Other articles were retrieved by review of references from previously acquired articles and by discussions with colleagues. Articles were deemed eligible for inclusion if they (1) were reported as a full publication, (2) tested a PARP inhibitor clinically, (3) incorporated a blinded placebo counterpart to the PARP inhibitor into the trial design, and (4) reported on neuropathy. This study focused on only placebo-controlled trials because of the established role of such inert substances in capturing drug effects. ²⁴ Articles were initially assessed for inclusion by abstract review. To determine final eligibility, two investigators (R.B. and A.J.) reviewed all articles in full.

Each trial, as described in these articles, was weighted by its respective sample size. Relative risks for all grades of neuropathy in arms assigned to a PARP inhibitor versus placebo were calculated for each trial. Trial heterogeneity was assessed by means of I² test as well as by the investigators' qualitative assessment of the trial design.

Results

Overview

A total of five clinical trials, inclusive of 843 patients, met this study's eligibility criteria (Fig. 1).

OPEN IN VIEWER

FIG. 1.

Consort diagram: five clinical trials formed the basis for the analyses presented in this study.

These five trials tested the PARP inhibitors, veliparib (2) and olaparib (3). Four trials tested PARP inhibition in conjunction with paclitaxel-based concurrent chemotherapy, and one tested PARP inhibition alone. Cancer types included breast cancer (1), gastric cancer (2), nonsmall cell lung cancer (1), and ovarian cancer (1). Only one trial examined PARP inhibition as a long-term maintenance therapy for ovarian cancer patients with adverse events assessed at two years and beyond. ²⁵

PARP inhibition and neuropathy

The pooled overall relative risk of developing neuropathy with PARP inhibition was 1.06 (95% confidence interval: 1–1.4) (Fig. 2), a finding that suggests PARP inhibition does not protect against neuropathy. Subgroup analyses that entailed removal of trials that seemed qualitatively heterogeneous from the meta-analysis did not alter conclusions.

OPEN IN VIEWER

FIG. 2.

Forest plot: In examining these trials in aggregate, PARP inhibition did not exert a protective effect for chemotherapy-induced peripheral neuropathy. This plot shows relative risks along with 95% confidence intervals.

Of note, substantial interstudy heterogeneity was observed based on the foregoing qualitative data and the I² of 73.1%

Discussion

This meta-analysis yielded disappointing results that do not support the hypothesis that PARP inhibition reduces the risk of chemotherapy-induced peripheral neuropathy. The findings reported in this study suggest that prescribing a PARP inhibitor in conjunction with a chemotherapy agent such as paclitaxel does not attenuate neuropathy risk.

How might one reconcile the foregoing conclusion with the much earlier results that prompted this study? First, we assumed that the purported salutary effects of PARP inhibitors comprise a class effect as opposed to a specific effect from an individual drug. Previous studies that pointed to an inverse relationship between neuropathy and PARP inhibition had focused on veliparib. ³ The fact that we opted to study PARP inhibitors as a class may have attenuated any purported palliative effect specific to veliparib. Second, this study does not address the possibility that PARP inhibitors are good at palliating—as opposed to preventing—chemotherapy-induced peripheral neuropathy. Once again, many of the studies, as referenced earlier, suggest that PARP inhibition palliates existing neuropathy. Of note, the vast majority of clinical trials that had tested PARP inhibitors in patients with baseline chemotherapy-induced peripheral neuropathy did not report on the latter, presumably because of assumptions that this baseline symptomatology would not change.^25–33 These circumstances forced us to focus this meta-analysis on trials that tested PARP inhibitors that were administered concurrently with chemotherapy, thereby enabling us to comment only on the prevention—as opposed to palliation—of neuropathy.

Finally, this study has strengths and limitations. With respect to the latter, these five trials included patients who had different cancer types and who had been prescribed different types of chemotherapy. Hence, heterogeneity between trial, as further evidenced by the I² of >70%, is not surprising but can be viewed as a limitation. However, the meta-analytic approach, employed here, can also be viewed as a strength. PARP inhibitors give rise to serious adverse events, such as myelosuppression and fatigue.^33,34 To use available data to address a clinically important question spares patients this toxicity. At the same time, as a result of this meta-analysis, health care providers are better equipped to educate cancer patients on risks and benefits of PARP inhibitors. Based on this study, health care providers are unable to speak of whether PARP inhibitors palliate chemotherapy-induced peripheral neuropathy, but, clearly, based on the results reported here, they are able to explain to patients that PARP inhibitors do not appear to prevent these symptoms, despite previous studies that might have suggested otherwise.