ASCO Gastrointestinal Cancers Symposium 2024

Important new studies with the potential to impact clinical practice were presented at the 2024 ASCO Gastrointestinal Cancers Symposium held January 18 to 20 in San Francisco. NEJM Journal Watch Oncology and Hematology Associate Editor David H. Ilson, MD, PhD, was on hand and reports on some of them.

CheckMate 8HW

Researchers presented results from CheckMate 8HW, an industry-sponsored, international, open-label, phase 3 trial evaluating first-line treatments in patients with microsatellite instability (MSI)-high/DNA mismatch repair protein-deficient metastatic colorectal cancer. Patients were randomized to investigator's choice of chemotherapy (mFOLFOX6 or FOLFIRI, with or without bevacizumab or cetuximab), single-agent nivolumab (240 mg every 4 weeks followed by 480 mg monthly), or nivolumab combined with ipilimumab (1 mg/kg for 4 doses). An interim analysis was presented for the 303 patients treated with chemotherapy or ipilimumab/nivolumab.

MSI status was centrally confirmed in 84% of patients; 23% had RAS and BRAF wild-type cancers, 21% had KRAS- or NRAS-mutant cancer, and 25% had BRAF-mutant cancer; 13% had Lynch syndrome. The primary endpoint of progression-free survival (PFS) was markedly superior with ipilimumab/nivolumab compared with chemotherapy (median, not reached vs. 5.9 months; hazard ratio, 0.21; P<0.0001), as were 12-month and 24-month PFS (79% vs. 21% and 72% vs. 14%, respectively). All subgroups, including patients with RAS and BRAF mutations, benefited. No new safety signals emerged.

These results establish the combination of ipilimumab and nivolumab as a first-line treatment option in MSI-high metastatic colorectal cancer. Results from the single-agent nivolumab cohort are awaited.

Two trials evaluated the utility of circulating tumor DNA (ctDNA) in localized colorectal cancer.

BESPOKE CRC

The industry-sponsored, prospective, observational BESPOKE CRC trial evaluated use of a tumor-informed circulating tumor DNA (ctDNA) assay to potentially guide adjuvant therapy decisions in patients with stage II and III colon cancer. Two cohorts were enrolled, a minimal residual disease (MRD) cohort tested 2–12 weeks after surgery and prior to adjuvant chemotherapy (623 patients), and a surveillance cohort tested more than 2 weeks after adjuvant chemotherapy or more than 12 weeks after surgery if observed without treatment (655 patients).

In the MRD cohort, ctDNA positivity strongly correlated with inferior 2-year disease-free survival (DFS) compared with ctDNA negativity (30% vs. 92%; HR, 12.1; P<0.0001); the effect was greater in stage II (HR 18.1) than in stage III disease (HR 9.9). In MRD-positive patients, adjuvant chemotherapy improved 2-year DFS compared with observation (42% vs. 12%; HR, 3.06; P=0.0025), whereas in MRD-negative patients, adjuvant chemotherapy had no DFS effect compared with observation (94% and 90%; HR, 1.47). During surveillance, ctDNA negativity resulted in superior 2-year DFS compared with ctDNA positivity, either with adjuvant chemotherapy (98% vs. 23%; HR, 59.98) or without adjuvant chemotherapy (97% vs. 13%; HR, 80.10). Patients with sustained clearance of ctDNA during or after adjuvant treatment had superior DFS compared with those with transient or no clearance.

GALAXY

GALAXY, an industry-sponsored, prospective, observational trial from Japan, evaluated nearly 4000 patients with stage I–IV colorectal cancer using a tumor-informed ctDNA assay. In a MRD cohort of 2860 patients tested after surgery, patients with a negative ctDNA test had superior 2-year DFS compared with those with a positive test (86% vs, 29%; HR, 10.53; P<0.0001); results were similar in the subset of patients with stage II/III cancer (89% vs. 34%; HR, 12.05; P<0.0001). In ctDNA-positive patients receiving adjuvant chemotherapy, 2-year DFS was 90% in those with sustained ctDNA clearance as compared with 2% for those with transient or no clearance (P<0.0001). In a surveillance cohort of 1786 patients tested 4 weeks after adjuvant chemotherapy or after surgery alone without adjuvant treatment, ctDNA-negative patients had superior 2-year DFS compared with ctDNA-positive patients (95% vs. 6%; HR, 53.79; P<0.0001).

These two trials indicate the strong prognostic impact of a positive ctDNA test after surgical resection and improved DFS if ctDNA clears.

DYNAMIC-Rectal Study

The multicenter AGITG DYNAMIC-Rectal trial evaluated a tumor informed ctDNA assay performed at 4 and 7 weeks in patients with stage T3–4 N0/N+ rectal cancer after completion of neoadjuvant chemoradiotherapy and surgery. Patients were randomized either to physician's standard management dependent on surgical pathology or to management guided by ctDNA testing: ctDNA-positive patients were recommended to undergo 4 months of adjuvant chemotherapy; ctDNA-negative patients with N0 status were to be observed and those with N+ status could be observed or treated at the physician's discretion. Of the 230 patients, 20% had clinical N0 status before treatment and 30% had pathologic N0 status at surgery.

Adjuvant chemotherapy was given to 46% of the 155 patients in the ctDNA-informed group and to 77% of the 75 patients in the standard group; among those given adjuvant chemotherapy, a higher percentage received oxaliplatin plus fluorinated pyrimidine in the ctDNA-informed group (61%) than in the standard group (33%). Recurrence-free survival (RFS) at 2 and 3 years was similar in the ctDNA-informed and standard groups (84% and 84% at 2 years, 74% and 82% at 3 years; HR 1.38; P=0.28). CtDNA-negative patients, of whom 23% received chemotherapy, had superior RFS compared with ctDNA-positive patients, all of whom received chemotherapy (2-year RFS, 89% vs. 61%; 3-year RFS, 83% vs. 53%). Lung recurrence predominated in ctDNA-negative patients, liver recurrence in ctDNA-positive patients.

The results suggest that ctDNA testing can lead to reduced use of adjuvant chemotherapy in patients with rectal cancer who test negative, and that a negative test may fail to detect occult lung metastases.

COBRA NRG-GI005

Confounding data were presented from the phase 2/3 COBRA NRG-GI005 trial evaluating a non–tumor-informed ctDNA assay in patients with stage II colon cancer who were candidates for active surveillance. Patients were randomized to standard observation or ctDNA testing (Guardant Lunar assay); ctDNA-positive patients received 6 months of adjuvant chemotherapy with capecitabine/oxaliplatin or mFOLFOX6 and ctDNA-negative patients were observed. The primary endpoint of the first phase of the trial was ctDNA clearance at 6 months in the ctDNA-positive treated and observed cohorts.

After 635 of the planned 1408 patients were accrued, the trial was halted because ctDNA clearance was greater in the ctDNA-positive untreated observed group (3 of 7 patients, 43%) than in the ctDNA-positive chemotherapy-treated group (1 of 9 patients, 11%). The results raise concerns about the accuracy of the non–tumor-informed ctDNA assay used to select patients for observation or treatment.

PEC-CC

Investigators reported the prospective, multicenter PEC-CC study that evaluated peritoneal cytology findings in patients in Japan undergoing curative resection of stage II or III colorectal cancer. Of 1378 patients evaluated, 49% had stage II and 51% had stage III disease. A positive cytology was found in 2.5% of those with stage II and 5.2% of those with stage III disease. In patients with stage II disease, peritoneal recurrence occurred in 11.8% of cytology-positive compared with 1.5% of cytology-negative patients; patients with positive cytology had inferior 5-year RFS and overall survival (OS). In patients with stage III disease, peritoneal recurrence occurred in 10.8% of cytology-positive versus 2.5% of cytology-negative patients; there was no difference in RFS or OS between cytology-positive and -negative patients, because patients with stage III disease likely received adjuvant chemotherapy.

The authors conclude that positive peritoneal cytology in stage II resected colorectal cancer may be a novel high-risk factor to identify patients who may be candidates for adjuvant chemotherapy.

NETTER-2

In the industry-sponsored, phase 3 NETTER-2 trial, patients with newly diagnosed, well-differentiated grade 2 or 3 gastroenteropancreatic neuroendocrine tumors (NETs) were randomized to treatment with either high-dose octreotide-LAR (60 mg) every month or low-dose octreotide-LAR (30 mg) every month plus ¹¹⁷lutetium-dotatate every 8 weeks. Of the 226 patients, 54% had pancreatic primary tumors, 65% had grade 2 tumors, and the median Ki67 index was 17.

The primary endpoint of PFS was significantly improved with ¹¹⁷lutetium-dotatate/low-dose octreotide-LAR compared with high-dose octreotide-LAR (median, 22.8 vs. 8.5 months; HR, 0.276; P<0.0001). The response rate was also significantly higher with ¹¹⁷lutetium-dototate compared with high-dose octreotide-LAR (43% vs. 9%; odds ratio, 7.81; P<0.0001). There were no new safety signals.

¹¹⁷Lutetium-dototate will likely become a first-line treatment option for grade 2 or 3 gastroenteropancreatic NETs.

EMERALD-1

EMERALD-1, an industry-sponsored, randomized, placebo-controlled, phase 3 trial, evaluated transarterial chemoembolization (TACE) alone, with durvalumab, or with durvalumab and atezolizumab, in patients with unresectable hepatocellular cancer (HCC). Of the 616 patients, 52% were treated in Asia (non-Japan), 36% had hepatitis B, and 23% had hepatitis C.

PFS was significantly improved with the addition of durvalumab and atezolizumab to TACE compared with TACE alone (median, 15.0 vs. 8.2 months; HR, 0.77; P=0.032), but did not differ with the addition of durvalumab to TACE compared with TACE alone (median, 10.0 and 8.2 months; HR, 0.94; P=0.638). Response rates were higher with the addition of durvalumab and atezolizumab or durvalumab compared with TACE alone (43.6% and 41.0% vs. 29.6%). There were no new safety signals. OS results are pending.

These results indicate a potential response and PFS benefit with the addition of durvalumab and atezolizumab to TACE.

RAISE

Investigators reported results of the multicenter, phase 2 RAISE trial, evaluating the use of postoperative adjuvant radiation therapy in hepatocellular carcinoma after resection with a close surgical margin (<1 cm). Patients at six hospitals in China were randomized to observation or intensity-modulated radiation therapy to a dose of 50 Gy delivered in 25 fractions. Of the 148 patients, 87% had hepatitis B, 53% had cirrhosis, and 85% had a single tumor resected.

The primary endpoint of RFS at 2 years was improved with radiotherapy over observation (79% vs. 57%; HR, 0.55; P=0.043). Treatment-related grade 3/4 serious adverse events were rare. OS data are awaited.

Important updates were presented for two pivotal trials evaluating the addition of immune checkpoint inhibitors to perioperative chemotherapy in gastroesophageal adenocarcinoma.

KEYNOTE-585

The recently published industry-sponsored, randomized, phase 3 KEYNOTE-585 trial evaluated the addition of pembrolizumab to perioperative 5-FU plus cisplatin (CF) or 5-FU plus oxaliplatin plus docetaxel (FLOT) in resectable gastric or gastroesophageal junction (GEJ) adenocarcinoma. The trial indicated an improvement in rates of pathologic complete response (pCR) with the addition of pembrolizumab to chemotherapy, but no improvement in the coprimary endpoints of either event-free (EFS) or OS (NEJM JW Oncol Hematol Jan 22 2024 and Lancet Oncol 2024; 25:212).

In an updated analysis of the 203 patients receiving FLOT, the pCR rate was improved with versus without the addition of pembrolizumab (17% vs. 7%). However, similar to findings already reported for the CF cohort and the combined CF and FLOT cohorts, the addition of pembrolizumab to FLOT did not translate into a significant improvement in either EFS (HR, 0.79; 95% CI, 0.52–1.22) or OS (HR, 1.04; 95% CI, 0.66–1.66).

MATTERHORN

The industry-sponsored, global MATTERHORN trial evaluated the addition of durvalumab to perioperative FLOT chemotherapy in patients with resectable gastric or GEJ adenocarcinoma. Of 948 patients, 66% had clinical stage T3 disease, 69% had clinical stage N+ disease, 49% had PDL-1 status <5% by tumor area positive score, and 8% had MSI-high status. The pCR rate was improved with the addition of durvalumab (19% vs. 7%), with response enhancements seen in both microsatellite stable (OR, 2.98) and MSI-high subgroups (OR, 4.28). Results for EFS and OS are pending. There were no significant regional variations in pCR.

SKYSCRAPER-08

The industry-sponsored, randomized, double-blind, placebo-controlled SKYSCRAPER-08 trial evaluated the addition of the anti-PDL-1 agent atezolizumab and the anti-TIGIT antibody tiragolumab to chemotherapy (3-week cycles of paclitaxel plus cisplatin) compared with chemotherapy alone in 461 patients with metastatic esophageal squamous cancer.

The primary endpoints of PFS and OS were improved with the addition of atezolizumab and tiragolumab compared with chemotherapy alone (median PFS, 6.2 vs. 5.4 months; HR, 0.56; P<0.0001; median OS, 15.7 vs. 11.0 months; HR, 0.70; P=0.0024).

Because the control arm received chemotherapy alone and did not include immunotherapy, the contribution of the TIGIT inhibitor to atezolizumab plus chemotherapy is at best uncertain.