Targeting PARP for Chemoradiosensitization: Opportunities, Challenges, and the Road Ahead

Introduction

In many cancer patients, the dose of radiation therapy (RT) that can be safely administered is insufficient to achieve high rates of local tumor control and cure. In others, damage to normal tissues is a concern even at moderate doses. In these settings, RT - or chemoradiation (CRT) - ideally would be combined with novel targeted drugs that can enhance the tumoricidal effects of standard therapy but without significantly increased normal tissue toxicity [1]. Over the past decade, major advances in precision medicine have supplied the field of radiation oncology with countless opportunities to enhance the anti-tumor effects of CRT [2]. However, a large body of preclinical research and clinical investigations on molecular targeted drugs has not yet translated into any meaningful number of combinations of RT or CRT with targeted radiosensitizers that are approved by the United States Food and Drug Administration (FDA) [3]. In fact, to date the epidermal growth factors receptor-directed monoclonal antibody cetuximab remains the only FDA-approved targeted agent for concurrent administration with RT (in head and neck (H&N) cancers). There exist considerable challenges to clinical translation of combining targeted drugs with CRT or RT that the field has only recently begun to fully appreciate [4, 5].

Since ionizing radiation exerts its tumoricidal effects predominantly by damaging DNA, the DNA damage response (DDR) provides a range of promising therapeutic targets. Poly(ADP-ribose) polymerase (PARP) is the primary sensor and signaler of single-strand breaks induced by DNA damaging cytotoxic agents. Pharmacological inhibition of PARP is associated with increased sensitivity to many such agents that are in routine clinical use. The radiosensitizing effects of PARP inhibitors have been appreciated for decades and moderate potentiation of ionizing radiation has been demonstrated in a broad spectrum of cancer cell lines in vitro and xenograft models of cancer in vivo [6].

The first indication that PARP inhibitors might exert differential effects on tumors and normal tissues arose from in vitro reports that radiosensitization was only observed in actively proliferating cells [7, 8]. Consistent with this, recent in vivo studies have shown that PARP inhibition exacerbates the acute radiation toxicity observed in rapidly proliferating normal tissues such as the skin and the esophagus [9]. Of note, the most marked effects were observed in the context of the highly potent compound talazoparib, which exhibits marked PARP trapping activity. Attempts to study effects on the lungs were initially confounded by coincidental irradiation of the esophagus, but recent advances in preclinical irradiation technologies enabled Jiang and colleagues to show that the clinical PARP inhibitor olaparib had no impact on acute or delayed lung toxicity [10]. This important study used esophageal-sparing hemi-thoracic irradiation in a urethane induced model of lung cancer to demonstrate convincingly that olaparib enhances the therapeutic ratio by enhancing tumor control without exacerbating lung toxicity.

Overall, the growing body of preclinical evidence supports the use of RT/PARP inhibitor combinations in clinical scenarios where tumors are rapidly proliferating and adjacent normal tissues are more quiescent. It also predicts that PARP inhibition would exacerbate RT-induced acute toxicity in highly proliferative normal tissues. In this article, we examine the validity of these predictions by reviewing emerging data from a series of recent early-phase clinical trials evaluating combinations of CRT or RT and PARP inhibitors in different tumor types.

De Haan et al., Phase I and Pharmacologic Study of Olaparib in Combination with High-dose Radiotherapy with and without Concurrent Cisplatin for Non-Small Cell Lung Cancer. Clin Cancer Res 2021 [11]

Summary:

This was a single-institution phase I trial adding olaparib to concurrent or sequential CRT in inoperable locally advanced non-small cell lung carcinoma (NSCLC) or to RT alone in oligometastatic disease. Concurrent CRT consisted of intensity-modulated radiation therapy to 66 Gy at 2.75 Gy per fraction with daily cisplatin at 6 mg/m². A time-to-event continual reassessment method (TITE-CRM) with a 1-year dose-limiting toxicity (DLT) period was used. Ten patients were enrolled at the starting dose and de-escalation dose levels 25 mg twice daily (BID) and 25 mg once daily (QD) olaparib + concurrent CRT. The estimated DLT likelihood at the 25 mg QD level was 31% due to hematologic and late esophageal DLTs; this exceeded the prespecified maximum tolerated dose (MTD) definition (dose level at which no more than 15% of patients experience DLT). Nine of 10 patients (90%) in the concurrent CRT group developed grade (G) 2/3 dysphagia and 78% did so in the sequential CRT/RT group (14/18). After initial recovery of dysphagia, a deterioration to G2/3 was observed at 3–4 months posttreatment in 8 patients. Esophageal endoscopy revealed 6 esophageal ulcers and 2 esophageal stenoses in 7 patients. Patients who developed late esophageal adverse events (AEs) had significantly higher esophageal metabolic activity 6 weeks posttreatment on fluorodeoxyglucose (FDG)-positron emission tomography (PET). Incidence and severity of late esophageal AEs was lower in patients not receiving concurrent CRT and with QD olaparib dose. Grade 2 or higher pneumonitis incidence was 7% at 1 year. However, at a latency of 1.0 – 2.8 years, severe pulmonary AEs (bronchial strictures, pulmonary fibrosis, and fatal hemorrhage) were observed, resulting in 18% G5 events (5/28). Two-year local control for all patients was 84% (95% CL, 58–95%) with a median follow-up of 14 months. In pharmacodynamic (PD) and pharmacokinetic (PK) studies, olaparib at 25 mg QD reduced baseline poly ADP-ribose (PAR) levels by more than 95% and also abolished RT-induced PARylation in peripheral blood mononuclear cells (PBMCs) at the time of radiation. Pharmacodynamic efficacy of PARP inhibition in tumor cells was observed to a similar extent as in PBMCs.

Commentary:

The authors of this study are applauded for providing in great detail the toxicity profiles associated with the addition of olaparib to CRT/RT as well as carefully acquired PD/PK data. While hematologic toxicity was expected, the occurrence of severe late esophageal and pulmonary AEs was not. Because radiosensitization by PARP inhibition is DNA replication-dependent, increased acute toxicity in early-responding normal tissues, such as esophageal mucosa, is logical. In our opinion, the observed ‘late’ esophageal toxicity is likely to represent a ‘consequential late effect’ since patients who developed late toxicity tended to have more severe acute dysphagia that persisted for longer without full recovery. This hypothesis is supported by the finding that these patients exhibited higher esophageal metabolic activity on FDG-PET post-treatment. The application of more stringent radiation dose-constraints for the esophagus has reduced the risk of acute esophagitis [12], and therefore it may also reduce the incidence of late complications. Similarly, optimization of RT to reduce lung dose may decrease the incidence of pulmonary complications [13]. It is unclear to what extent the use of hypofractionated RT (2.75 Gy/day) and concomitant daily cisplatin impacted RT toxicity, but it seems likely that standard fractionation schedules of 1.8–2 Gy are more suitable to the evaluation of these radiosensitizers with (C)RT. Data on treatment volume and involvement of blood vessels and airways are not reported and would have helped the interpretation of the toxicity results.

There are important conclusions to be drawn from this study: 1) The addition of targeted radiosensitizers to standard CRT/RT is not without risks, highlighting the importance of continually improving RT techniques to limit the potential toxicity of novel treatment combinations. 2) Not only careful short-term follow-up but also long-term follow-up is critical to a full understanding of the potential late effects of such combinations. 3) PD/PK data indicate that effective radiosensitization can be achieved by much lower doses of olaparib than those used in monotherapy or in combination with chemotherapy, which has important implications for current and future trials in this setting.

Karam et al, Final Report of a Phase I Trial of Olaparib with Cetuximab and Radiation for Heavy Smoker Patients with Locally Advanced Head and Neck Cancer. Clinical Cancer Research, 2018 [14]

Summary:

This phase I study used TITE-CRM to evaluate escalating doses of olaparib in combination with RT (69.3 Gy in 33 fractions) and cetuximab (loading dose 400 mg/m² one week pre-RT then weekly 250 mg/m² during RT) in a total of 17 patients with locally advanced H&N cancer and a smoking history of at least 10 pack years. Olaparib was commenced 10 days prior to RT with a starting dose of 50 mg BID. Olaparib dose escalation commenced as indicated by the TITE-CRM model, with two patients receiving 200 mg BID and four patients receiving 100 mg BID. However, high rates of G3–4 dermatitis, mucositis, and dysphagia necessitated dose de-escalation and while the MTD of olaparib in this combination was determined to be 50 mg BID, the recommended phase II dose (RP2D) was defined as 25 mg BID.

Commentary:

This study supports the findings of the NKI study testing olaparib in NSCLC (de Haan et al. [11]) by demonstrating exacerbation of acute radiation toxicity by relatively low doses of olaparib. Interpretation of the data are hampered by the high levels of acute toxicity expected with standard of care RT-cetuximab, especially in patients with a smoking history, and by the limitations of the toxicity scoring system. It would be interesting to see if olaparib could be combined more effectively with RT for H&N cancer patients in the absence of an additional systemic agent; such a trial would be difficult to perform given the current guidelines that specify RT-cisplatin or RT-cetuximab as standard of care.

Argiris et al., A Dose-finding Study Followed by a Phase II Randomized, Placebo-controlled Trial of Chemoradiotherapy With or Without Veliparib in Stage III Non-small-cell Lung Cancer: SWOG 1206 (8811), Clin Lung Cancer 2021 [15]

Summary:

SWOG 1206 was a multi-institutional combined phase I MTD finding (3+3 design) and randomized phase II trial testing the addition of veliparib to concurrent CRT and consolidation chemotherapy for treatment of unresectable stage III NSCLC. CRT consisted of standard 60 Gy at 2 Gy per fraction with concurrent weekly carboplatin and paclitaxel. During the phase I part, 21 patients were enrolled and the MTD/RP2D was established as 120 mg BID. In the phase II part, 31 patients were analyzable, 18 of whom were randomized to veliparib + CRT (during thoracic RT and consolidation phase) and 13 received placebo + CRT. Grade 3–5 AEs were observed in 66% of patients receiving veliparib + CRT patients in the phase I/II study parts. The most common AEs were hematologic as well as esophagitis. Grade 3–5 esophagitis, dysphagia, esophageal pain, or/and perforation was reported in 16% (6/38) of phase I/II patients treated with veliparib + CRT compared to 8% (1/13) in the placebo + CRT cohort. In addition, during consolidation veliparib was associated with 11.5% esophagus-related G3–5 AEs vs none in the placebo group. Late G2+ AEs were not detailed although 1 patient who received veliparib was noted to have developed a G4 esophageal fistula 15 months after consolidation. With a median follow-up of 26.9 months in the phase II cohort, veliparib and placebo were associated with median progression-free survival (PFS) times of 9.3 and 9.9 months, respectively. Treatment efficacy could not be fully evaluated owing to early study closure as a result of the publication of the PACIFIC trial[16].

Commentary:

Veliparib is less potent than the other PARP inhibitors in clinical use and exhibits markedly less PARP trapping activity than olaparib or talazoparib for example. As such, the observation that it was associated with less esophageal (and pulmonary) toxicity than the combination of olaparib with RT and CRT in the study by de Haan et al. [11] is perhaps not surprising. However, comparisons are limited by differences in study design and CRT backbone and there does appear to be a signal of increased acute esophageal toxicity with the addition of veliparib, mirroring the findings of de Haan et al. [11] Treatment efficacy is difficult to evaluate due to small patient numbers and early study closure.

Kozono et al., Veliparib in combination with carboplatin/paclitaxel-based chemoradiotherapy in patients with stage III non-small cell lung cancer. Lung Cancer 2021 [17]

Summary:

This was a multi-institutional phase I study to assess the safety and efficacy of adding veliparib to CRT and consolidation chemotherapy in unresectable stage III NSCLC. CRT consisted of standard 60 Gy at 2 Gy per fraction with weekly carboplatin and paclitaxel. The study followed a 3+3 design with six dosing cohorts from 60 mg to 240 mg BID veliparib with CRT. Forty-eight patients were enrolled with 12 patients in Cohort 5 (RP2D). The MTD/RP2D of veliparib was 240 mg BID with CRT. Veliparib-related AEs were reported in 88% of patients, with 56% G3/4 and 23% classified as serious. The most common side effects were bone marrow suppression nausea, and esophagitis. All-grade and G3/4 esophagitis were observed in 67% and 8% of patients, respectively. Sepsis was the only veliparib-related death reported (n=1). With a median follow-up of 26.5 months, median PFS was 19.6 months. Out of 16/48 patients with progression, 4/16 recurred in the irradiated lung or nodes.

Commentary:

This phase I study tested a higher dose of veliparib than SWOG 1206; this likely explains the relatively high rate of observed veliparib-associated AEs and draws attention to the importance of generating standardized DLT criteria in phase I studies of RT/drug combinations. Whether the rate of radiation esophagitis was increased by veliparib in this study is difficult to appreciate. The observed 8% rate of G3+ esophagitis is similar to that associated with standard CRT[18], and neither the isolated rate of G2 acute esophagitis nor late AEs were detailed. The favorable PFS and low number of local recurrences reported are consistent with the notion that this combination provided local treatment efficacy.

Sim et al., A randomized phase II trial of veliparib, radiotherapy, and temozolomide in patients with unmethylated MGMT glioblastoma: the VERTU study. Neuro Oncol 2021 [19]

Summary:

The VERTU study was a randomized phase II trial in which 125 patients with newly diagnosed O6-methylguanine-DNA-methyltransferase (MGMT) unmethylated glioblastoma were randomized (2:1) to receive RT (60 Gy in 30 fractions) with concomitant veliparib 200 mg BID or placebo, followed by standard dose adjuvant temozolomide (TMZ) with veliparib 40 mg BID, days 1–7, or placebo. Grade 3–4 toxicity was observed in 55% of patients in both arms with higher rates of G3–4 thrombocytopenia, neutropenia, seizure and fatigue in the veliparib arm. Dose reductions were not required in the concomitant phase, which was well tolerated, but were frequent in the adjuvant phase of the experimental arm, with mean veliparib dosing being 60% and mean TMZ dosing 64% compared with 85% in the control arm. The addition of veliparib did not improve PFS or overall survival (OS), with median OS being 12.7 months in the experimental arm and 12.8 months in the control arm. Similarly, no effect on health-related quality of life (HRQL) was detected.

Commentary:

The safety and toxicity data from this study are encouraging, with veliparib deliverable at full single agent dose in combination with radical partial brain RT. These observations are consistent with the preclinical data that showed the radiosensitizing effects of PARP inhibitors to be observed only in replicating cells (which are not present in the normal brain). It is not surprising that dose reductions of veliparib and/or TMZ were required in the adjuvant phase, given previous data showing exacerbation of TMZ-related myelosuppression by various PARP inhibitors. The lack of an efficacy signal was disappointing; this may reflect the relatively low potency of veliparib as a PARP inhibitor and radiosensitizer, and may be linked to the lack of PARP trapping activity of this compound.

Discussion

Toxicity

Collectively, the five recent early-phase trials (olaparib n=2, veliparib n=3) reviewed here confirm the hematologic toxicity associated with the integration of PARP inhibitors with standard (C)RT. They also illustrate the potential for increased acute toxicity in proliferating tissues, such as esophageal and oropharyngeal mucosa, that are vulnerable to the replication-targeted effects of PARP inhibitors (Fig. 1) [3]. This is an important consideration for trials incorporating PARP inhibitors with CRT in H&N or gastrointestinal cancers. Olaparib combined with hypofractionated CRT also reported unexpected late occurring esophageal and pulmonary toxicity, which at least for the former we believe to represent consequential effects of severe acute toxicity. These findings highlight the need for continued refinement of RT techniques to facilitate the integration of targeted radiosensitizers into standard-of-care, for example through methods that promote esophageal mucosal regeneration during CRT [12]. The data also reinforce the importance of careful and detailed assessment and reporting of acute and late RT complications, especially those occurring beyond the first year of follow-up. This will require a change in both the focus and design of early-phase clinical trials, which have traditionally emphasized reporting of acute DLTs. Novel statistical methodologies such as the TITE-CRM design provide opportunities to achieve more balanced and informative assessment of toxicities.

Fig. 1.

Potential challenges to adding PARP inhibitors to standard-of-care therapy. ICI, immune checkpoint inhibitors. Modified from ref.[3].

Efficacy and Biomarkers

Despite the reproducible radiosensitizing effects of PARP inhibitors seen in laboratory studies, the trials discussed here have failed to demonstrate striking signals of clinical efficacy. There may be numerous reasons for this, including trial design, sample size, and the use of a weakly PARP trapping inhibitor (veliparib). In addition, it is common practice in preclinical studies to combine small molecule inhibitors with RT but not CRT [4]. Conceivably, as the combination of platinum +/− taxane chemotherapeutics already targets proliferating cells for radiosensitization, this may not be further enhanced by PARP inhibition (Fig. 1). Another possibility is that there is inter-tumoral heterogeneity in the radiosensitizing effects of PARP inhibitors and predictive biomarkers will be needed to identify patients most likely to benefit from treatment intensification [5]. These may be biomarkers mechanistically related to the effects of PARP inhibitors, for example proliferation indices or DDR factors indictive of a dependency on PARP-mediated DNA repair. Alternatively, biomarkers of radioresistance, such as mutant KEAP1 or/and KRAS in NSCLC, may be useful to select patients in whom the additional toxicity resulting from PARP inhibition can be justified in an attempt to reduce the high local recurrence rate associated with CRT in these tumors [20].

Conclusions

In order to realize the promise of efficacious and safe radiosensitization by combining targeted drugs and (C)RT, there is an unmet need to develop efficient and scientifically rich clinical studies. One example is the CONCORDE trial, a multidisciplinary collaboration between academia and the pharmaceutical industry testing multiple DDR inhibitors within a Bayesian adaptive model-based framework [21]. In addition, consideration should be given to the combination of DDR inhibitors and (C)RT with immune checkpoint inhibitors (ICI), given their increasing role in the standard management of many disease sites and the potential for synergistic interactions between these three therapeutic classes. More preclinical studies are also urgently needed, in particular a) to deepen our understanding of the interactions of DDR inhibitors with systemic therapy (chemotherapy as well as ICI), and b) to reveal predictive biomarkers that may identify patient subsets most like to benefit from novel combination therapies. Taken together, the clinical trials discussed here illustrate the opportunities and challenges associated with integrating targeted agents into standard-of-care therapies. Valuable lessons learned from these trials illuminate the road ahead where considerable enthusiasm still exists for combining PARP inhibitors with (C)RT +/− ICI.