Neoadjuvant Chemoradiation Plus Veliparib or Pembrolizumab Fails to Improve NAR Score in Locally Advanced Rectal Cancer

Both veliparib plus total neoadjuvant therapy and pembrolizumab (Keytruda) plus total neoadjuvant therapy failed to improve short-term outcomes in unselected patients with locally advanced rectal cancer, according to long-term findings from the phase 2 NRG-GI002 trial (NCT02921256).

Pembrolizumab with total neoadjuvant therapy did demonstrate a benefit in 3-year overall survival (OS) rate at 95% compared with 87% with total neoadjuvant therapy alone (HR, 0.35; 95% CI, 0.12-1.00; P = .04). However, other outcomes were not statistically significantly improved by the addition of pembrolizumab.

Presenter Thomas J. George, MD, FACP, noted during the 2023 Gastrointestinal Cancers Symposium that the reason for these differences in outcome results was not entirely understood.

“NRG-GI002 helps to provide further total neoadjuvant therapy outcome data from multi-institutional national clinical trials that can benchmark and be used for future locally advanced rectal cancer outcomes,” said George, a professor in the Department of Medicine in the Division of Hematology & Oncology, and director of the Gastrointestinal Oncology Program at the University of Florida.

Veliparib Experimental Arm

In the veliparib experimental arm (n = 90), the majority of patients were male (67.8%), younger than 60 years (64.5%), not Hispanic or Latino (94.4%), had distal location (60%), bulky disease (54.4%), cN2 stage disease (45.6%), cT3 stage disease (70.0%), and were candidates for sphincter-sparing surgery (61.1%). George noted that these characteristics matched a national trend for early-onset rectal cancer development.

Similarly in the control arm (n = 88), most patients were male (67.0%), younger than 60 years (68.2%), not Hispanic or Latino (95.5%), had distal location (69.3%), bulky disease (62.5%), cN2 stage disease (40.9%), and cT3 stage disease (68.2%). However, in this arm, most patients were not candidates for sphincter-sparing surgery (61.4%).

At a median follow-up of 3.5 years, the mean neoadjuvant rectal (NAR) score in the veliparib arm was 13.3 (95% CI, 10.1-16.5) compared with 12.5 (95% CI, 9.7-15.2) in the control arm (mean difference, –0.8; 95% CI, –5.0 to 3.3; P = .81). The 3-year DFS rate was 60% with veliparib compared with 67% with total neoadjuvant therapy alone (HR, 1.36; 95% CI, 0.83-2.25; P = .23), and the 3-year OS rate was 85% vs 92%, respectively (HR, 2.13; 95% CI, 0.86-5.29; P = .10).

In terms of secondary end points, the pCR rate with veliparib was 33.8% compared with 21.6% for the control arm (P = .14), and the clinical complete response (cCR) rates were 33.3% and 28.2%, respectively (P = .60). R0 resection rates were 83.1% in the veliparib arm and 85.1% in the control arm (P = .82), and the rate of patients who received surgery was 59.3% vs 52.5%, respectively (P = .43).

Pembrolizumab Experimental Arm

In the pembrolizumab experimental arm (n = 90), the majority of patients were male (66.7%), younger than 60 years (64.5%), not Hispanic or Latino (86.7%), had distal location (72.2%), bulky disease (66.7%), cN1 or cN2 stage disease (38.9% each), cT3 stage disease (74.4%), and were candidates for sphincter-sparing surgery (54.4%).

Characteristics were similar in the control arm (n = 95) where most patients were male (69.5%), younger than 60 years (61.1%), not Hispanic or Latino (87.4%), had distal location (71.6%), bulky disease (57.9%), cN2 stage disease (38.9%), cT3 stage disease (71.6%), and were candidates for sphincter-sparing surgery (53.7%).

The mean NAR score in the pembrolizumab arm was 11.5 (95% CI, 8.5-14.5) compared with 14.4 (95% CI, 11.1-17.7) in the control arm (mean difference, 2.9; 95% CI, –1.6 to 7.3; P = .21). The 3-year DFS rate was matched in both arms at 64% (HR, 0.95; 95% CI, 0.58-1.55; P = .82).

The pCR rate was 31.9% with veliparib compared with 29.4% with total neoadjuvant therapy alone (P = .75), and the cCR rates were 13.9% and 13.6%, respectively (P = .95). The R0 resection rate was 94.0% in the pembrolizumab arm and 89.4% in the control arm (P = .36), and the rate of patients who went on to surgery was 59.4% vs 71% (P = .15).

Study Design

Total neoadjuvant therapy is an accepted standard of care for patients with locally advanced rectal cancer to ensure that high-risk patients are able to receive systemic therapy, as previously 25% to 70% did not. However, responses are still fairly limited with this approach.

The NRG-GI002 trial consisted of nested, randomized phase 2 experimental arms within a master clinical trial protocol. “NRG-GI002 was deliberately designed to rapidly seek activity-seeking signals for new agents in this total neoadjuvant paradigm,” George said.

Eligible patients had very high-risk stage II and III rectal cancer, including those with a distal location, bulky, cN2 disease, and/or not a candidate for sphincter-sparing surgery at the time of presentation. However, patients were not selected for tumor markers such as mismatch repair deficiency.

They were randomly assigned to either the control arm or one of the experimental arms and were stratified by cT and cN stage. In the control arm, patients received FOLFOX (leucovorin, fluorouracil, and oxaliplatin) chemotherapy for 8 cycles followed 3 to 4 weeks later by long-course radiation therapy (50.4 Gy) with concurrent capecitabine (825 mg/m2 orally twice daily) and then followed 8 to 12 weeks later by total mesorectal surgical resection. The first experimental arm added veliparib (400 mg twice daily) to the chemoradiotherapy regimen and the second experimental arm added pembrolizumab (200 mg intravenously every 3 weeks for 6 doses). In both experimental arms, the total neoadjuvant therapy was unchanged.

George noted that veliparib was added with the idea that a PARP inhibitor could enhance radiation-induced synthetic lethality, and pembrolizumab was tested based on the hypothesis that radiation could enhance anti-tumor immunogenicity with upregulation of immune markers.

The primary end point was NAR score, to quantify downstaging of the tumor, with a target score reduction of 4.70 points. Key secondary end points included pathologic complete response (pCR) rates, cCR rates, OS, disease-free survival (DFS), and toxicity.

George noted that analysis of tumor and circulating biomarkers is ongoing to identify potential subgroups of patients that may benefit from these added treatments.

Reference

George TJ, Yothers G, Rahma OE, et al. Long-term results from NRG-GI002: A phase II clinical trial platform using total neoadjuvant therapy (TNT) in locally advanced rectal cancer (LARC). J Clin Oncol. 2023;41(suppl 4):7. doi:10.1200/JCO.2023.41.3_suppl.7