Roundup of Abstracts on New Therapies for Gastrointestinal Cancers From ASCO 2023

Get Permission

The 2023 ASCO Annual Meeting featured thousands of abstracts. In addition to our in-depth coverage of pivotal research from the meeting, The ASCO Post highlights the following studies of novel therapies for gastrointestinal cancers.

Addition of Anti-TIGIT Agent in Unresectable Liver Cancer

In the phase Ib/II MORPHEUS-Liver study, the addition of the novel agent tiragolumab to standard-of-care atezolizumab plus bevacizumab improved responses as a first-line treatment in patients with unresectable, locally advanced or metastatic hepatocellular carcinoma.1 The findings point to anti-TIGIT treatment as a potential enhancer of checkpoint inhibitors, according to Richard S. Finn, MD, of the University of California Los Angeles and the Jonsson Comprehensive Cancer Center.

Richard S. Finn, MD

Richard S. Finn, MD

“Tiragolumab plus atezolizumab and bevacizumab may represent a novel first-line treatment option for untreated hepatocellular carcinoma,” Dr. Finn said.

TIGIT is a co-inhibitory receptor and immune checkpoint expressed on activated T cells, natural killer cells, and regulatory T cells, and it has been implicated in liver cancer and other tumors. Tiragolumab is an anti-TIGIT monoclonal antibody that may augment antitumor responses when combined with other immunotherapies, such as inhibitors of PD-L1.

The triplet was evaluated in the open-label multicenter randomized umbrella MORPHEUS-Liver study, whose 58 participants had previously untreated advanced liver cancers. MORPHEUS comprises multiple early-phase trials designed to identify early signals of efficacy and safety of cancer immunotherapy using atezolizumab/bevacizumab as a backbone. There were 18 patients randomly assigned to receive the control doublet and 40, to receive triplet therapy. Most baseline characteristics were balanced, except in the control arm, there were more patients with a good performance status and fewer patients with advanced features; the experimental arm had fewer patients with alfa-fetoprotein levels ≥ 400 μg/L.

The triplet of tiragolumab, atezolizumab, and bevacizumab boosted response rates, which were 11.1% with atezolizumab plus bevacizumab and 42.5% with the addition of the anti-TIGIT compound. At a median follow-up of 14.3 months for the triplet and 12.1 months for standard therapy, investigator-assessed median progression-free survival was 11.1 months vs 4.2 months (hazard ratio [HR] = 0.42), respectively.

Grade 3 or 4 adverse events were recorded in 50.0% of the control arm and 40.0% of the triplet arm; 16.7% and 15.0%, respectively, were grade 5, and those leading to withdrawal occurred in 5.6% and 12.5%, respectively. All-grade immune-mediated rash increased in frequency with the addition of tiragolumab (47.5% vs 33.3%), but there were no instances of grade 3 or 4 immune-mediated rash in either arm, Dr. Finn reported.

“Recognizing the control arm is somewhat small and might not represent what we anticipate from atezolizumab plus bevacizumab, there were two statistical analyses leveraging the IMbrave150 data set2 [which confirmed the efficacy of atezolizumab plus bevacizumab] to create hybrid control arms,” Dr. Finn explained. The Bayesian dynamic modeling analysis calculated hazard ratios of 0.49 in a conservative model and 0.58 in an aggressive model; a model that was based on optimal full matching calculated the hazard ratio to be 0.72. “These two complementary hybrid control analyses support a treatment benefit for the tiragolumab-containing regimen,” he said.

The results from MORPHEUS-Liver support the planned double-blind randomized phase III IMbrave152/SKYSCRAPER-14 study. This trial is expected to begin soon and will further investigate this regimen, Dr. Finn said.

Neoadjuvant Therapy for Colon Cancer Yields Mixed Results

In the randomized phase III NeoCol trial, neoadjuvant chemotherapy was not superior to standard upfront surgery and adjuvant therapy in patients with colon cancer, but it did offer some relative benefits, according to Lars Henrik Jensen, MD, PhD, of the Danish Colorectal Cancer Center South, University Hospital of Southern Denmark, Vejle.3

Lars Henrik Jensen, MD, PhD

Lars Henrik Jensen, MD, PhD

In both arms, the 5-year disease-free survival rate exceeded 75% (P = .94), and the overall survival rated exceeded 85% (P = .95). “But the neoadjuvant approach may have more favorable outcomes in terms of the number of chemotherapy cycles, risk of chemotherapy toxicity, and some surgical complications, as well as for downsizing and downstaging,” he noted.

As Dr. Jensen pointed out, patients with locally advanced colon cancer present a therapeutic challenge. The question is whether survival can be improved and toxicity minimized by optimizing the timing of surgical and systemic treatments. And for this objective, the neoadjuvant approach has potential benefits, he added.

NeoCol randomly assigned 250 patients with locally advanced colon cancer to upfront surgery or neoadjuvant chemotherapy with either 3 cycles of oxaliplatin, capecitabine every 3 weeks (CAPOX) or 4 cycles of oxaliplatin, 5FU every 2 weeks (FOLFOX). A total of 73% of the tumors were classified as T3, whereas 26% were classified as T4 on baseline imaging. The upfront surgery arm received a mean of 5.9 postoperative chemotherapy cycles; the neoadjuvant therapy arm received 2.7 preoperative cycles and 4.1 postoperative cycles. Surgery was performed a median of 7 days and 74 days, respectively, from randomization and was laparoscopic for 68% and 75%, respectively.

“Surgery proved to be a safe procedure after neoadjuvant chemotherapy. The absolute numbers are small, but the relative risk of ileus or anastomotic leakage favored neoadjuvant chemotherapy,” reported Dr. Jensen.

For upfront surgery vs neoadjuvant surgery, respectively, ileus was seen in 8% and 3% and anastomotic ­leakage in 8% and 2%. Tumors were downstaged with neoadjuvant ­treatment, with 3% of patients achieving complete responses and 7% achieving in situ status or better; no patient in this arm had lymph node spread, and few patients had vascular invasion. Significantly fewer patients in this arm needed adjuvant therapy.

Of note, sensory neuropathy grade 3 or 4 was more common with upfront surgery (11% vs 7%) during treatment and during follow-up (5% vs 2%).

T-DXd in HER2-Positive Colorectal Cancer

According to the primary analysis of the multicenter phase II DESTINY-CRC02 trial, fam-trastuzumab deruxtecan-nxki (T-DXd) showed antitumor activity in patients with HER2-positive metastatic colorectal cancer at both 5.4 mg/kg and 6.4 mg/kg. Antitumor efficacy was observed irrespective of RAS mutation status or previous treatment with an anti-HER2 agent, as reported at the meeting by Kanwal Pratap Singh Raghav, MBBS, MD, of The University of Texas MD Anderson Cancer Center, Houston.4

Kanwal Pratap Singh Raghav, MBBS, MD

Kanwal Pratap Singh Raghav, MBBS, MD

The study assessed the efficacy and safety of these two doses of T-DXd in 82 patients with HER2-positive unresectable, recurrent or metastatic colorectal cancer. Patients’ tumors were RAS wild-type or mutated and BRAF wild-type, and patients had received a median of three to four prior lines of therapy, including prior anti-HER2 therapies if indicated. The primary endpoint was confirmed objective response rate by blinded independent central review.

Confirmed responses were seen in 37.8% of patients receiving 5.4 mg/kg (vs 27.5% receiving 6.4 mg/kg); in 46.9% of patients with immunohistochemistry (IHC) 3+ HER2 status (vs 5.6% with IHC 2+/in situ hybridization–positive status); in 39.7% of patients with wild-type RAS (vs 28.6% with RAS mutations); and in 41.2% of patients who received prior anti-HER2 therapy, Dr. Raghav reported.

Overall, safety was consistent with the known safety profile of T-DXd and favored the 5.4-mg/kg dose. All-grade interstitial lung disease and pneumonitis rates were lower at 5.4 mg/kg (8.4% vs 12.8% with 6.4 mg/kg).

“These promising results support T-DXd at 5.4 mg/kg as the optimal dose, as a single agent, in this patient population, because of its positive benefit/risk profile,” Dr. Raghav concluded.

PRODIGE 23 Update: Neoadjuvant Chemotherapy for Rectal Cancer

The 7-year results of the phase III PRODIGE 23 trial have shown that overall survival in locally advanced rectal cancer can be improved with neoadjuvant chemotherapy using modified ­FOLFIRINOX (leucovorin, fluorouracil, irinotecan, oxaliplatin) followed by chemoradiation, surgery, and adjuvant chemotherapy, as compared with standard chemoradiation, surgery, and adjuvant chemotherapy.5

Thierry Conroy, MD, of the Institut de Cancérologie de Lorraine in Vandoeuvre-Lès-Nancy, France, reported the 7-year follow-up of the PRODIGE 23 patients. PRODIGE 23 previously showed,6 as did the RAPIDO trial,7 that this total neoadjuvant approach (TNT) could result in higher metastasis-free survival, but the studies did not show an improvement in overall survival with 44 months’ and 60 months’ follow-up, respectively. Dr. Conroy reported the primary and secondary endpoints of PRODIGE 23 with mature 7-year follow-up.

Thierry Conroy, MD

Thierry Conroy, MD

The study enrolled 461 patients with cT3 or cT4, M0 rectal adenocarcinomas < 15 cm from the anal verge. They were randomly assigned to the standard of care (radiotherapy plus capecitabine, followed by total mesorectal excision, and then by adjuvant chemotherapy with modified FOLFOX6 [fluorouracil, leucovorin, oxaliplatin] or capecitabine) or neoadjuvant therapy with modified FOLFIRINOX (followed by radiotherapy plus capecitabine, then by total mesorectal excision, and then by adjuvant modified FOLFOX6 or capecitabine).

“TNT with modified FOLFIRINOX should now be considered as one of the best options in patients with locally advanced rectal cancers,” Dr. Conroy said. “This strategy is now being investigated in rectal cancer for tailored management, according to response to induction modified FOLFIRINOX.”

Cumulative incidences for the key outcomes for TNT vs standard therapy at 7 years included: rectal cancer recurrences (5.3% vs 8.1%), metastatic disease (26.4% vs 34.6%), and metastases as a first event (16.4% vs 24.3%). The 7-year disease-free survival rate was 67.6% vs 62.5%; the cancer-specific survival rate was 84.9% vs 79.6%; and the overall survival rate was 81.9% vs 76.1%.

TNT with modified FOLFIRINOX conveyed a restricted mean survival time (average time to event over the 7-year follow-up) of 5.7 months for disease-free survival (P = .048); 7.1 months for metastasis-free survival (P = .011); 3.8 months for cancer-specific survival (P = .051); and 4.3 months for overall survival (P = .033).

Second cancers and deaths from other causes were included in the definition of the disease-free survival. Unrelated second cancers were more frequently detected during TNT (6.1% vs 2.2%). Moreover, there was no reduction in survival after distant failure in the TNT arm—a phenomenon known as adjuvant therapy–related shortening of survival, according to Dr. Conroy.

“The disease-free and metastasis-free survival benefits are durable, there has been no increase in local recurrences, the quality of life is similar [to previous reports] or improved, and the safety profile is unchanged,” he added. PRODIGE 23 is the first randomized trial to demonstrate an overall survival benefit since the Swedish Rectal Cancer Trial on preoperative radiotherapy in 1997.8

NAPOLI-3 Update on NALIRIFOX in Pancreatic Cancer

In an updated analysis of the global phase III NAPOLI-3 trial, the NALIRIFOX regimen (liposomal irinotecan, fluorouracil, leucovorin, oxaliplatin) maintained its improvements in median overall survival and progression-free survival at 12 and 18 months, relative to gemcitabine and nab-paclitaxel.9 The significant results shown in the earlier analysis9 “were supported by outcomes at 12 and 18 months for overall survival and progression-free survival, and these outcomes were also identified in the key prespecified subgroups,” according to Eileen M. O’Reilly, MD, of Memorial Sloan Kettering Cancer Center, New York.

Eileen M. O’Reilly, MD

Eileen M. O’Reilly, MD

“These results support NALIRIFOX as a new reference regimen in front-line untreated pancreas cancer,” Dr. O’Reilly said, noting that except for maintenance studies, the last positive phase III front-line study in pancreatic cancer occurred a decade ago.

The NAPOLI-3 trial evaluated NALIRIFOX in 770 patients with previously untreated metastatic pancreatic ductal adenocarcinoma, 80% of whom had liver metastases. Significantly improved overall survival and progression-free survival were reported with NALIRIFOX over gemcitabine plus nab-paclitaxel, as reported at the 2023 ASCO GI Cancers Symposium.10 At a median follow-up of 16.1 months, median overall survival was 11.1 months with ­NALIRIFOX compared with 9.2 months with gemcitabine plus nab-paclitaxel (hazard ratio = 0.83; P = .04), and median progression-free survival was 7.4 months vs 4.6 months (hazard ratio = 0.69; P < .0001), respectively.

In the updated analysis, overall survival in the intention-to-treat population was, for NALIRIFOX vs standard therapy, 45.6% vs 39.5% at 12 months and 26.2% vs 19.3% at 18 months. Progression-free survival was 27.4% vs 13.9% at 12 months and 11.4% and 3.6% at 18 months, respectively. Analysis of all the major subgroups—such as performance status, liver metastases, and age—consistently favored NALIRIFOX, Dr. O’Reilly reported.

Treatment-related adverse events grade ≥ 3 were frequent in both arms, 71% with NALIRIFOX and 68% with gemcitabine plus nab-paclitaxel. Serious treatment-related adverse events were noted in 26% and 19%, respectively, and those leading to death were seen in about 2% per arm. “As you distill these data, one thing to keep in mind is the median duration of treatment was about one-third longer with NALIRIFOX, 24 weeks vs 18 weeks,” she pointed out.

“Toxicity was as you might expect for the components of each regimen,” she noted. Incidence of serious adverse events slightly favored NALIRIFOX, 51% vs 54%. There were more hematologic adverse events with gemcitabine plus nab-paclitaxel and more gastrointestinal toxicity with NALIRIFOX. “Peripheral neuropathy was relatively similar, possibly less on the NALIRIFOX arm because of the lower dose of oxaliplatin used in the study,” Dr. O’Reilly noted.

Neoadjuvant Approach Falls Short in Pancreatic Cancer in NORPACT-1

In the multicenter randomized phase II NORPACT-1 trial of patients with resectable cancer of the pancreatic head, a short course of neoadjuvant FOLFIRINOX was not associated with improved overall survival compared with upfront surgery and adjuvant chemotherapy, according to findings reported by Knut Jørgen Labori, MD, of Oslo University Hospital and Institute of Clinical Medicine.11

Knut Jørgen Labori, MD

Knut Jørgen Labori, MD

The study explored outcomes after four cycles of neoadjuvant FOLFIRINOX followed by restaging and surgery, followed by adjuvant modified FOLFIRINOX for eight cycles vs upfront surgery followed by adjuvant modified FOLFIRINOX for 12 cycles.

Of 140 patients enrolled, 77 patients were randomly assigned to the neoadjuvant arm, but just 64 received it, largely because of an inability to biopsy the tumor or establish biliary drainage; 18% of patients were found to have metastases before surgery. Neoadjuvant therapy was completed by 40 patients (52%), and surgical resection was completed for 63 patients (82%) in the original cohort.

Of the 63 patients randomly assigned to upfront surgery, 56 (89%) ultimately underwent resection. As for adjuvant therapy, 39 patients (51%) on the neoadjuvant arm completed all cycles; for 73% of this arm, treatment was gemcitabine-based. In the upfront surgery arm, 32 patients (51%) completed adjuvant chemotherapy, and for 58%, this was gemcitabine-based. Therefore, more than 50% of patients received gemcitabine-based adjuvant therapy and not adjuvant FOLFIRINOX, according to the study design.

Neoadjuvant therapy did not improve overall survival. In fact, a numerical advantage was observed for upfront surgery, which yielded a median overall survival of 38.5 months as compared with 25.1 months with the neoadjuvant approach (hazard ratio = 1.52; P = .096) in the intention-to-treat analysis. The per-protocol analysis reached a similar conclusion (hazard ratio = 1.46; P = .158), Dr. Labori reported. 

DISCLOSURE: Dr. Finn has served as a consultant or advisor to AstraZeneca, Bayer, Bristol Myers Squibb, CStone Pharmaceuticals, Eisai, Exelixis, Genentech/Roche, Hengrui Therapeutics, Eli Lilly, Merck, and Pfizer. Dr. Raghav reported financial relationships with Bayer, Daiichi Sankyo/Astra Zeneca, Eisai, Merck, and Seagen. Dr. O’Reilly reported financial relationships with Boehringer Ingelheim, BioNTech, Ipsen, Merck, Novartis, AstraZeneca, BioSapien, Astellas, Thetis, Autem, BMS, Tempus, Fibrogen, Merus, Agios (spouse), Genentech-Roche (spouse), Eisai (spouse), and Servier (spouse). Dr. Jensen, Dr. Conroy, and Dr. Labori reported no conflicts of interest.


1. Finn RS, Ryoo BY, Hsu CH, et al: Results from the MORPHEUS-Liver study: Phase Ib/II randomized evaluation of tiragolumab in combination with atezolizumab and bevacizumab in patients with unresectable, locally advanced or metastatic hepatocellular carcinoma. 2023 ASCO Annual Meeting. Abstract 4010. Presented June 4, 2023.

2. Finn RS, Qin S, Ikeda M, et al; IMbrave150 Investigators: Atezolizumab plus bevacizumab in unresectable hepatocellular carcinoma. N Engl J Med 382:1894-1905, 2020.

3. Jensen LH, Kjaer ML, Larsen FO, et al: Phase III randomized clinical trial comparing the efficacy of neoadjuvant chemotherapy and standard treatment in patients with locally advanced colon cancer: The NeoCol trial. 2023 ASCO Annual Meeting. Abstract LBA3503. Presented June 4, 2023.

4. Raghav KPS, Siena S, Takashima A, et al: T-DXd in patients with HER2-overexpressing/amplified metastatic colorectal cancer: Primary results from the multicenter, randomized, phase 2 DESTINY-CRC02 study. 2023 ASCO Annual Meeting. Abstract 3501. Presented June 4, 2023.

5. Conroy T, Etienne PL, Rio E, et al: Total neoadjuvant therapy with mFOLFIRINOX versus preoperative chemoradiation in patients with locally advanced rectal cancer: 7-year results of PRODIGE 23 phase III trial, a UNICANCER GI trial. 2023 ASCO Annual Meeting. Abstract LBA3504. Presented June 4, 2023.

6. Conroy T, Bosset JF, Etienne PL, et al: Neoadjuvant chemotherapy with FOLFIRINOX and preoperative chemoradiotherapy for patients with locally advanced rectal cancer (UNICANCER-PRODIGE 23): A multicentre, randomised, open-label, phase 3 trial. Lancet Oncol 22:702-715, 2021.

7. Dijkstra EA, Nilsson PJ, Hospers GAP, et al: Locoregional failure during and after short-course radiotherapy followed by chemotherapy and surgery compared to long-course chemoradiotherapy and surgery: A five-year follow-up of the RAPIDO trial. Ann Surg. January 20, 2023 (early release online).

8. Swedish Rectal Cancer Trial; Cedermark B, Dahlberg M, Glimelius B, et al: Improved survival with preoperative radiotherapy in resectable rectal cancer. N Engl J Med 336:980-987, 1997.

9. O’Reilly EM, Melisi D, Macarulla T, et al: Liposomal irinotecan + 5-fluorouracil/leucovorin + oxaliplatin (NALIRIFOX) versus nab-paclitaxel + gemcitabine in treatment-naive patients with metastatic pancreatic ductal adenocarcinoma: 12- and 18-month survival rates from the phase 3 NAPOLI 3 trial. 2023 ASCO Annual Meeting. Abstract 4006. Presented June 4, 2023.

10. Wainberg ZA, Melisi D, Mararulla T, et al: NAPOLI-3: A randomized, open-label phase 3 study of liposomal irinotecan + 5-fluorouracil/leucovorin + oxaliplatin (NALIRIFOX) versus nab-paclitaxel + gemcitabine in treatment-naive patients with metastatic pancreatic ductal adenocarcinoma. 2023 ASCO GI Cancers Symposium. Abstract LBA661. Presented January 20, 2023.

11. Labori KJ, Bratlie SO, Biörserud C, et al: Short-course neoadjuvant FOLFIRINOX versus upfront surgery for resectable pancreatic head cancer: A multicenter randomized phase II trial (NORPACT-1). 2023 ASCO Annual Meeting. Abstract LBA4005. Presented June 2, 2023.