Highlights of the Gastrointestinal Cancers Symposium

July 2021 ASCO GI 2021 Multiple oncologists
Human body digestive system anatomy. 3d illustration

Authors: 1 Willem Lybaert MD, 2 Anne Demols MD PhD, 3 Isabel Dero MD, 4 Ivan Borbath MD PhD, 5 Joëlle Collignon MD

1 Medical Oncology, AZ Nikolaas – AZ Lokeren – UZA, Sint-Niklaas – Lokeren – Antwerpen.

2 Department of Gastro-enterology and GI oncology, Hôpital Erasme, Brussels.

3 Department of Gastro-enterology and GI oncology, AZ Monica, Antwerpen.

4 Department of Gastro-enterology/Hepatology and GI oncology, UCLouvain, Brussels.

5 Medical Oncology, CHU-Sart-Tilman, Liège.

The annual Gastrointestinal Cancers Symposium was held from 15 till 17 January 2021 in a virtual format. During this meeting, evolving immunotherapy options were shown in the upper digestive tractus, already presented at ESMO 2020. Besides immunotherapy, also the total neoadjuvant treatment approach (TNT) in rectal cancer reached sufficient attention, changing our current standard of practice in 2021. Targeted therapies for different GI tumour locations showed promising results, paving the way for more personalized medicine. In this report, the most important headlines will be discussed, with comments on the clinical relevance of the different studies.

Esophageal, gastroesophageal (GEJ) and gastric cancer

2020 was the year of advances in the treatment of locally advanced/metastatic esophageal and gastric cancer with mainly news in targeted treatment and immunotherapy. At ASCO GI this year, we had updates and new trials in the same setting.

For advanced metastatic esophageal cancer, abstracts in the poster session showed data with long-term follow-up for nivolumab in squamous cell carcinoma.

First, ATTRACTION-1/ONO-4538-07, a phase II monotherapy of nivolumab 3 mg/kg Q2W in esophageal squamous cell cancer (ESCC) intolerant or refractory to fluoropyrimidine-, platinum- and taxane-based chemotherapy with a minimum of five years of follow-up: 65 patients were enrolled with ECOG PS 0/1 and 2 prior lines or more (≤ 2: 32%, 3: 37%, ≥ 4: 31%) and received 3 mg/kg nivolumab intravenously every 2 weeks in 6-week cycles until disease progression or unacceptable toxicity. The primary endpoint was centrally-assessed objective response rate (ORR), defined as the proportion of patients whose best overall response was either a complete or partial response. Secondary endpoints included overall survival (OS), investigator-assessed ORR, progression-free survival (PFS), change in tumour burden, time to response, time to disease progression and duration of response. Long-term follow-up confirms durable efficacy with 10,8 months mean OS and the estimated 5-year OS rate was 6.3%. The median PFS was 1.5 months and the estimated 5-year PFS rate was 6.8%. No new toxicities were identified, and long survivors tend to show more complete response (CR) to nivolumab.

Second trial, ATTRACTION-3: a phase 3 study of nivolumab 3 mg/kg Q2W vs. docetaxel 75 mg/m² IV Q3W or paclitaxel 100 mg/mg² Q1W for 6 weeks and then one week off in patients with advanced ESCC refractory or intolerant to one prior line of fluoropyrimidine and platinum chemotherapy: 419 patients were randomized with ECOG PS 0/1. The primary endpoint was OS. With 3-year follow-up, nivolumab confirmed a clinically meaningful improvement in long-term OS (median OS: 10.9 months in the nivolumab group vs. 8.5 months in the chemotherapy group). There were no new safety signals and no major late onset treatment related adverse events were observed. Now since December 2020, nivolumab is reimbursed in second-line for ESCC.

We can now follow guidelines for metastatic ESCC with fluoropyrimidine+platinum-based chemotherapy in first-line, nivolumab in second-line and taxane- or irinotecan-based chemotherapy in third-line.

But, immunotherapy has also shown benefit in first-line metastatic esophagogastric cancer and even in the adjuvant setting.

Data were presented on health-related quality of life (HRQoL) of pembrolizumab + chemotherapy vs. chemotherapy as first-line therapy in patients with advanced esophageal cancer: the phase III KEYNOTE-590 study. Dual primary endpoints were OS and PFS; secondary endpoints were ORR and PROs (change from baseline to week 18 in QLQ-C30 GHS/QoL, and QLQ-OES18 pain, reflux and dysphagia scores); exploratory endpoint was change from baseline to week 18 in EQ-5D-5L. In the primary analysis, the addition of pembrolizumab to platinum-based chemotherapy significantly increases ORR, PFS and OS (median OS 12,4 months vs. 9,8 months) in patients with advanced esophageal carcinoma over chemotherapy with a manageable safety profile. The benefit was seen in all patients and independent of CPS-status, although squamous histology and CPS ≥ 10 may get a maximum benefit (median 13,9 months vs. 8,8 months). All HRQoL measures remained stable and similar over 18 weeks in the pembrolizumab + CT and placebo + CT treatment groups, with no difference between groups. The addition of pembrolizumab to CT maintained HRQoL measures relative to baseline and did not worsen HRQoL compared with CT alone.

At ESMO 2020, we had results from CheckMate 577, a randomized, double-blind phase III study of nivolumab (NIVO) vs. placebo (PBO) as adjuvant treatment in patients with resected esophageal or gastroesophageal junction cancer (EC/GEJC). Primary endpoints were DFS and secondary endpoints, OS and OS at years 1, 2 and 3. 794 patients were enrolled with 532 in nivolumab arm and 262 patients in the placebo arm. Nivolumab provided superior DFS with a 31% reduction in the risk of recurrence or death and a doubling in median DFS (22,4 months vs. 11 months). At ASCO GI 2021, Eric Van Cutsem presented PRO exploratory endpoints: FACT-E, EQ-5D-3L, ECS and FACT-G7. Results from Checkmate 577 demonstrated that patients who were treated with nivolumab and placebo showed trends for improvement and maintenance of HRQoL from baseline. There was no significant difference in time to first deterioration of HRQoL between nivolumab and placebo. Patients treated with nivolumab did not experience a reduction in HRQoL, which is good news to become a new standard.

An interesting association of pembrolizumab and lenvatinib showed promising antitumour activity in several cancers in early-phase trials and is FDA approved for patients with previously treated advanced endometrial cancers not MSI-H/dMMR. Chung presented the results from the gastric cohort from LEAP-005: a phase II multicohort study of lenvatinib + pembrolizumab in patients with previously treated selected solid tumors. Patients with at least 2 priors lines received lenvatinib, an oral multikinase inhibitor targeting VEGFR1-3, FGFR1-4, PDGFRα and the oncogenes RET and KIT given at 20 mg once daily + pembrolizumab 200 mg Q3W for up to 35 cycles of pembrolizumab (approximately 2 years) or until confirmed disease progression, unacceptable toxicity, or withdrawal of consent. 31 patients were enrolled; ORR was 10%, median DOR was not reached (range, 2.1+ to 2.3+ months). DCR was 48%. Median PFS was 2.5 months. Median OS was 5.9 months. The safety profile was manageable and based on these data, enrollment in the gastric cancer cohort has been expanded to 100 patients.

Besides immunotherapy, ASCO GI 2021 showed also interesting data in targeted therapy.

In advanced-line adenocarcinoma of gastric cancers/gastroesophageal junction (GC/GEJC), the TAGS study showed an improved OS with trifluridine/tipiracil (FTD/TPI) + best supportive care (BSC) vs. placebo + BSC in heavily pretreated metastatic GC/GEJC patients with a 31% risk reduction in death in third- and later-lines. The primary endpoint, OS, was met and yet published. Median OS was 5·7 months in the trifluridine/tipiracil group and 3∙6 months in the placebo group. At ASCO GI, Tabernero presented TAGS pre-planned subgroup analysis: efficacy, safety and QoL of trifluridine/tipiracil in third-line or beyond third-line (4L+). Patients were divided into two groups: patients who had received FTD/TPI or placebo after two lines of previous systemic therapy (3L) (n = 126 vs. 64), and patients who had received FTD/TPI or placebo (n = 211 vs. 106) after three or more lines of previous systemic therapy (4L+). The 2 groups were well balanced except that FTD/TPI was more frequently used in fourth-line and later-line in Japan and USA than in Europe. This subgroup analysis confirms the efficacy of FTD/TPI vs. placebo for third- and later-line treatment of GC/GEJC. The mOS for FTD/ TPI vs. placebo in 3L was 6.8 vs. 3.2 months (HR, 0.67; P=0.0318), whereas, in 4L+, it was 5.2 vs. 3.7 months (HR, 0.72; P=0.0192). The results suggest a superior survival benefit with FTD/TPI in patients treated in the third-line setting than in later-lines. And QoL in third- and later-lines is consistent with the overall TAGS study population. Trifluridine/tipiracil is now reimbursed in Belgium and can be used in third- and later-line. Irinotecan-based chemotherapy is also an option to be discussed before or after trifluridine/tipiracil.

For targeted therapy in first-line, since 2010, the only positive trial was ToGA showing benefit in OS for trastuzumab added to platinum-based chemotherapy in HER2-positive GC/GEJC. Median overall survival was 13·8 months in those assigned to trastuzumab + chemotherapy compared with 11,1 months in those assigned to chemotherapy alone.

At ASCO GI 2021 was presented the update of FIGHT trial: a randomized double-blind placebo-controlled phase 2 study of bemarituzumab combined with modified FOLFOX6 in first-line treatment of advanced gastric/gastroesophageal junction adenocarcinoma. Bemarituzumab is a first-in-class IgG1 targeted antibody specific for FGFR2b receptor overexpressed in 3-61% of GC depending on tumour stage and assay. A phase 1 study of bema monotherapy in solid tumors confirmed objective response rate (ORR) of 18% in patients with refractory FGFR2b+ GC. The primary endpoint was investigator-assessed PFS, and key secondary endpoints include OS, ORR and frequency of adverse events. Of 910 1L GC pts whose tumors were evaluated, 275 (30%) were FGFR2b+. 155 pts were randomized, 77 to bema+mFOLFOX6 and 78 to placebo+mFOLFOX6. Improved efficacy was observed in all 3 endpoints (PFS, OS, ORR):  improvement in median PFS of 9.5 mo (bema) vs. 7.4 mo (pbo) (HR 0.68; p=0.07), median OS not reached in the bema arm vs. 12.9 mo for pbo (HR 0.58, p=0.03). Among patients with measurable disease, ORR improved from 40% (pbo) to 53% (bema). An important finding is also increasing efficacy in all 3 endpoints with higher levels of FGFR2b overexpression: if IHC 2+/3+ ≥ 10%, median PFS was 14,1 mo for bema cohort vs. 7,3 mo for the placebo cohort (HR 0,44). Stomatitis was higher in the bema arm (31.6% vs. 13.0%) and corneal AEs were more common in the bema arm (67% vs. 10%), but grade 3 and higher adverse events, serious adverse events and deaths were comparable across arms. These results support phase III in GC and also testing the drug in others FGFR2b+ tumour types.

Biliary tract cancer

After this ASCO GI 2021, there is no change in clinical practice of the first-line treatment of advanced biliary cancers (ABC); the combination of gemcitabine+cisplatin (ABC-02, Valle et al. N Engl J Med 2010) remains the standard of care whatever the molecular/genetic profile of the biliary tumour is. However, efficacy data of the phase I/II trial combining silmitasertib (an oral small molecule inhibitor of casein kinase 2 that exhibited preclinical antitumour activity and strong synergism with gemcitabine+cisplatin) with gemcitabine+cisplatin showing ORR of 32.1% and mPFS of 11.1 months (as compared to 8 months mPFS in the gemcitabine+cisplatin arm of the ABC-02 trial) (MJ Borad, abstract 312), are promising and a randomized phase 3 is already planned.

Molecular/genetic characterization of biliary tract cancers (BTC) is of great importance in the area of precision medicine and this point was highlighted in many presentations during this congress. Emergence of molecular targeted therapies is of great interest in BTC, especially for fusions/alterations of FGFR2 (almost exclusively in intrahepatic cholangiocarcinoma), IDH1 mutations (intrahepatic cholangiocarcinoma), HER2 amplification or ERBB2 mutations (especially in gallbladder and extrahepatic cholangiocarcinoma), BRAF mutations and NTRK fusions (rare, less than 1%).

It is already demonstrated that FGFR2 fusions/alterations confer a better prognosis for the patients; this was also reported in abstract 303 (GK Abou-Alpha), a retrospective analysis of 132 intrahepatic cholangiocarcinoma patients (including 15 tumours with FGFR2 fusions/rearrangements) receiving chemotherapy, demonstrating a numerically longer mOS in these patients as compared to patients with no alteration. On the contrary, IDH1 mutations have no prognostic impact.

Even if FOLFOX is the standard of care in second-line therapy for advanced biliary tract cancers (ABC-06 trial, Lamarca et al., 2019), some targeted therapies have been studied in ≥ second-line treatment for ABC.

For FGFR2 fusions/alterations, several anti-FGFR inhibitors have already been tested in second- and more-lines. In April 2020, pemigatinib has been FDA approved in second-line therapy, after the release of the FIGHT 202 phase 2 trial results (Abou-Alpha et al. Lancet Oncol 2020). In this trial, patients harbouring FGFR2 fusions or rearrangements have an ORR of 35.5% and a mPFS of 6.9 months. At this ASCO GI, the results of the exploratory endpoint QoL have been presented (J Valle, abstract 276), using EORTC QLQ-C30 and specific QLQ-BIL21 questionnaires. In patients with CR/PR or SD, pemigatinib was associated with stable overall health status, decrease of pain and anxiety, despite increase of perceived treatment side effects (related to treatment length).

These efficacy results of anti-FGFR inhibitors were confirmed by the presentation of the final results of the open-label phase 2 study of infigratinib in 108 patients with previously treated FGFR2 fusions/rearrangements tumours (M Javle, abstract 265). Confirmed ORR was 23.1% (95% CI 15.6-32.2%), SD was 61.1% and median PFS was 7.3 months (95% CI 5.6-7.6 months). Toxicity was manageable and safety profile was consistent with anti-FGFR agents (ie. hyperphosphatemia, eye disorders and fatigue).

Based on these efficacy data in second- and more-lines, randomized phase 3 trials are ongoing in first-line therapy comparing the standard chemotherapy gemcitabine+cisplatin to pemigatinib (FIGHT 302 study) or to infigratinib (PROOF trial).

For IDH-1 mutations, ivosidenib (an oral small molecule inhibitor of mutant IDH1) has been tested in ≥2 line setting in the ClarIDHy randomized double-blinded phase 3 trial. 187 patients harbouring mutations of IDH-1 were randomized 2:1 between placebo and ivosidenib, and crossover was permitted after progression on placebo (and occured in 70% of the patients in the placebo arm). Primary endpoint was met with a significant increase of mPFS from 1.4 months to 2.7 months (p<0.0001). Ivosidenib also resulted in a numerous improvement in secondary OS despite a high crossover rate and this was confirmed after RPSFT adjustment for crossover (HR=0.49, one-sided p<0.0001) (AX Zhu, abstract 266). Toxicity profile was manageable and ivosidenib preserved QLQ-C30 physical functioning and was favored on the QLQ-C30 pain subscale (AX Zhu, abstract 266). Unfortunately, despite this positive trial, ivosidenib is not yet accessible in Belgium.

Some phase I or II trials have been presented showing promising activity of anti-HER2 therapy in later lines of treatment. In the abstract 299 (F Meric-Bernstam et al., phase 1 study), zanidatamab, a bispecific HER2-targeted antibody, was well tolerated with promising and durable antitumour activity (confirmed ORR of 47%, median duration of response 6.6 months) in patients with HER2-overexpressing BTC after progression on standard of care. Based on these data, zanidatamab is now being evaluated in an ongoing global phase 2b study. In the biliary cohort of the phase 2 SUMMIT basket trial (JJ Harding, abstract 320), somatic ERBB2 BTC mutant, previously treated patients (n=25) were treated with neratinib, an oral irreversible tyrosine kinase inhibitor of ERBB1, ERBB2 and ERBB4. Neratinib was safe and tolerable, and showed antitumour activity (ORR 16%). Future combination strategies are in development.

Pancreatic cancer

For resectable and borderline resectable pancreatic cancer, the question of radiation in neoadjuvant/adjuvant therapy is on the stage. A lot of questions remain open including patient selection (borderline only?), radiation type (SBRT vs. conventional), sequence of radiation (neoadjuvant vs. adjuvant), and to date there are no prospective phase 3 data assessing this question.

In the Alliance A021501 trial (MHG Katz, abstract 377), borderline resectable pancreatic tumours were randomized between neoadjuvant mFOLFIRINOX + adjuvant mFOLFOX6 vs. neoadjuvant mFOLFIRINOX + 5 days hypofractionated radiation or SBRT + adjuvant mFOLFOX6. In this trial, an interim analysis was planned after evaluation of R0 resection of the first 30 patients in each arm. Because the R0 resection rate was only 33% (10/30 patients) in the radiation arm, inclusions were stopped for futility in this arm due to the prespecified closure of either arm if R0 resection rate was ≤ 37%. Inclusions were continued in the mFOLFIRINOX arm for a total of 70 patients (vs. 56 patients in the radiation arm). Resection rates were 49% and 35% in the mFOLFIRINOX and radiation arms, respectively. Among patients who underwent surgery, 18-mo OS rate (primary endpoint) was 93.1% (95% CI: 84.3-100) in the mFOLFIRINOX arm and 78.9% (95% CI: 62.6-99.6) in the radiation arm. After a median follow-up of 27 and 31 months, addition of radiation did not improve OS, mOS was 31 months in the mFOLFIRINOX arm and 17.1 months in the radiation arm. Modified FOLFIRINOX was declared efficacious using the prespecified final efficacy decision rule (improvement in the 18-months overall survival rate of 13% over a historical rate of 50%).

In the metastatic setting, there are no practice changing-reported trials. In front-line therapy, adding ramucirumab to mFOLFIRINOX in a randomized double-blinded, placebo-controlled phase 2 trial (WL Shaib, abstract 413) was well tolerated, but did not improve mPFS (primary endpoint; 5.6 months for the ramucirumab+mFOLFIRINOX vs. 6.7 months for placebo+mFOLFIRINOX), RR or OS.

For gBRCA mutants, the final OS results of the POLO phase 3 trial have been presented (T Golan, abstract 378). In this trial, metastatic germline BRCA-mutated pancreatic cancers whose disease had not progressed on first-line platinum-based chemotherapy, were treated with olaparib maintenance. Primary endpoint was met and patients had significantly longer mPFS with maintenance olaparib than with placebo (HR=0.53; 95% CI, 0.35-0.82; p=0.004). Although HR for OS (0.83) was in favor of maintenance olaparib vs. placebo and that, at data cutoff analysis 73.5% of olaparib patients were still alive (as compared to 41.2%), there was no statistically significant difference between olaparib mOS (19.0 months) and placebo mOS (19.2 months). However, primary mPFS and other key secondary endpoints support a clinical benefit of maintenance olaparib: median time to first subsequent cancer treatment was significantly longer in the olaparib group (9.0 months vs 5.4 months, p<0.0001, HR=0.44) and median time to second progression (investigator-assessed) was clearly longer (16.9 months vs. 9.3 months, NS). These data support the use of olaparib as maintenance therapy in this setting.

Hepatocellular carcinoma (HCC)

The year 2020 was a landmark for the treatment of HCC because of the results of the IMbrave150 study, a phase 3 randomised controlled trial that demonstrated the superiority of the combination atezolizumab+bevacizumab over sorafenib in advanced (BCLC C) HCC. At this year’s ASCO-GI meeting, Dr. Finn reported the post-hoc, descriptive OS analysis conducted with 12 months of additional follow-up from the primary analysis (abstract 267). At the clinical cut-off date of Aug. 31, 2020, median follow-up was 15.6 months and 280 OS events were observed. Median OS was 19.2 months with atezo + bev vs. 13.4 months with sor (HR, 0.66 [95% CI, 0.52, 0.85]; P=0.0009). Survival at 18 mo was 52% with atezo+bev and 40% with sor. Survival benefit with atezo+bev over sor was generally consistent across subgroups and with the primary analysis. The updated objective response rate (ORR 29.8% per RECIST 1.1) with atezo+bev was in line with the primary analysis, with more pts achieving complete response (CR 7.7%) than previously reported. Safety was aligned with the primary analysis, with no new signals identified. Dr. Finn concluded thatIMbrave150 showed consistent clinically meaningful treatment benefit and safety with 12 mo of additional follow-up. The combination provides the longest survival seen in a front-line phase III study in advanced HCC, confirming atezo+bev as a standard of care for previously untreated, unresectable HCC.

Another systemic immunotherapy treatment currently assessed, is the combination of anti-CTLA-4 and anti-PD-1. Dr. El-Khoueiry presented the long-term follow-up results from the CheckMate 040 nivo/ipi cohort, with a minimum follow-up of 44 months, in patients with advanced HCC previously treated with sorafenib. They were randomized to three arms – arm A: nivo 1, ipi 3 mg/kg every three weeks times four followed by nivo maintenance; arm B: nivo 3, ipi 1 mg/kg every three weeks times four followed by nivo maintenance; arm C, nivo 3 every two weeks, ipi 1 mg/kg Q6 weeks continuously. The key endpoints of the study were safety and objective response rate. The majority of the patients were BCLC C. More than 80% of patients had extrahepatic spread. More than 20% of patients across arms had two or more prior lines of therapy. Response outcomes at 44 months of follow-up were consistent with the primary analysis. Response rate continued to be at 32% in arm A with 8% complete responses, higher than in arms B and C. Of note, very durable responses were achieved across treatment arms with the duration of response approaching four years in some cases. The 36-month overall survival rates were 42%, 26%, and 30%, respectively. Immune-mediated adverse events were reported more frequently in arm A than arms B and C; most were reversible. So, authors concluded that the nivo 1, ipi 3 combination (arm A) continues to demonstrate clinically meaningful responses and long-term survival with OS rates at 36 months reaching 42% in that arm. The safety profile was manageable with longer follow-up with no new safety signals. According to authors, these results further support the use of this combination as second-line treatment for patients with advanced HCC who previously received sorafenib. Currently, the CHECKMATE-9DW is recruiting patients in first-line advanced HCC, comparing arm A to sorafenib/lenvatinib.

Our colleague Jean-Luc Van Laethem presented the phase II KEYNOTE-224 study (abstract 297), assessing the efficacy of pembrolizumab monotherapy for previously untreated advanced HCC. The study enrolled 51 pts, of which many came from Belgian centers. The median time from the first dose to data cutoff (July 31, 2020) was 21 (range 17-23) mo. The median age of pts was 68 (range 41-91) years, one pt was HBV+, 80% had alcohol use, 8% were HCV+, 18% had vascular invasion, 35% had extrahepatic disease, 33% had BCLC stage B disease and 67% had BCLC stage C HCC. ORR was 16% (95% CI, 7-29) and was similar across most subgroups. Median DOR was not reached (range 3-20+ mo); 70% were estimated to have response duration ≥12 mo. Best overall responses were 8 (16%) PR, 21 (41%) SD and 17 (33%) PD. The median PFS was 4 (95% CI, 2-6) mo, and median OS was 17 (95% CI, 8-NA) mo. PFS rate at 18 mo was 16%, and OS rate at 18 mo was 46%. Immune-mediated AEs and infusion reactions occurred in 11 (22%) pts. So, in pts with advanced HCC and no prior systemic therapy, pembrolizumab monotherapy can provide durable antitumour activity and promising overall survival in advanced HCC. Hopefully, these findings will support evaluation of pembrolizumab-based regimens for the treatment of HCC in the frontline setting and provide evidence of the efficacy of CPI as monotherapy in a subset of patients.

In abstract 335, Dr. Yarchian and colleagues presented their study of neoadjuvant cabozantinib and nivolumab in patients with borderline resectable or local advanced hepatocellular carcinoma. The characteristics of patients enrolled in the study included a high rate of multinodular disease, portal vein invasion, infiltrated disease and very large tumours, including tumours over 10 cm. Of the 15 patients enrolled, 12 underwent a successful margin negative resection, and 5 of 12 actually had a major or complete pathologic response at the time of surgery as defined by greater than 90% tumour necrosis on the surgical resection specimen. All patients with an elevated AFP at baseline achieved AFP drop of 30% or greater over the course of neoadjuvant therapy. The patients who had major pathologic responses, had a disease-free survival of greater than 200 days after surgery, whereas patients who did not achieve major pathologic responses, had a high rate of disease recurrence quite early on. A deeper analysis of the tumour immune microenvironment of responders versus non-responders using imaging mass cytometry allowed generating a simplified network of cell types. B-cells were spatially distinct in the responder tumour immune microenvironment than non-responders; the neighboring cells were other B-cells consistent with tumour lymphoid aggregates. This study is the first to use systemic therapy to attempt to downstage patients outside of resection criteria. It shows the feasibility of this approach, and the importance of B-cell infiltration in responding tumour tissue.

In the same line, abstract 300, presented by Dr. Lewandowski, reports on the use of radioembolization with Yttrium-90 microspheres as neoadjuvant therapy to transplantation or resection in HCC. This was a retrospective, single-arm, multicenter study on 162 consecutive patients treated between January 2014 and December 2017. Their key eligibility criteria included unresectable solitary lesions up to 8 cm, no vascular invasion or extrahepatic disease. Patients had Child-Pugh A, BCLCA or C (ECOG 0-1 for the latter). Radioembolization was performed either in a selective lobar or a mixture of selective and lobar therapies. The endpoint was to compare the overall response rate (OR), duration of response and overall survival (OS) in patients treated with Y-90 as neoadjuvant to surgery vs. those who did not go on to surgery. Overall, OR was 72.2%, among which 76.1% of patients experienced duration of response > 6 months. The 3-year OS of all patients was 86.6%. Y-90 treatment served as neoadjuvant therapy prior to surgery in 45 patients (27.8%): 34 of those patients went on to transplantation, and 11 on to surgical resection. For this neoadjuvant arm, OR was 80%, with a duration response > 6 months of 30.6%. The 3-year overall survival was 92.8%. For the 72.2% of patients that were not transplanted or resected, OR was 91.5%, with a duration of response > 6 months of 72.9% and a 3-year overall survival of 83.5%. Importantly, the histopathology from the explants showed that 66% of patients achieved a complete pathologic necrosis; while 22.2% had extensive pathologic necrosis, meaning 50% to 90% necrosis. Authors conclude that in their hands, treatment of solitary HCC with Y-90 glass microspheres provides strong OR, duration of response and overall survival both as neoadjuvant therapy to transplantation or resection and as treatment in non-surgical candidates. However, authors do not report on surgical complications related to treatment, nor long-term OS of disease-free survival. Nevertheless, radioembolization appears very effective as neoadjuvant therapy before transplantation.

Colorectal cancer (CRC)

At this ASCO GI, the final results of the KEYNOTE-177, a phase 3 randomized study of pembrolizumab vs. chemotherapy for MSI-high advanced colorectal cancer, were presented. The study design was a 1:1 randomisation in MSI-high treatment naive patients between pembrolizumab 200 mg every 3 weeks and chemotherapy at the investigator’s choice, with an optional crossover from the chemotherapy arm to pembrolizumab when verified PD by RECIST, central review. 307 patients were included. There were dual primary endpoints, PFS and OS, and secondary endpoints ORR and safety. Superiority of pembrolizumab was shown for PFS at 24 months follow-up with a PFS of 16.5 months vs. 8.2 months. Also the median duration of respons (DOR) was 10.6 months in the chemotherapy arm, where the median DOR has not been reached at a median study follow-up of 32.4 months. Notify that the median time to response was equal, 2.2 months, in both groups. The PFS2, time from randomization to progression on the next line of therapy or any cause of death, were shown. With an effective crossover rate of 59% of patients in the chemotherapy arm, the median PFS2 was 23.5 months in this group; however the median PFS2 has not been reached yet in the pembrolizumab arm. The adverse event rate was also much better in the pembrolizumab arm, with 22% grade 3 AE vs. 66%. The auto-immune related AE in patients who received pembrolizumab in first-line were very manageable at 3% of colitis and 3% of hepatitis. In conclusion, pembrolizumab provided clinically and statistically meaningful benefit in PFS and PFS2 and responses were durable with a manageable safety profile. We await keenly the OS data. Data support pembrolizumab as the new standard of care in first-line patients with MSI-high mCRC.

Updated results of the final analysis on OS of the TASCO1 trial were presented, a randomized phase II study evaluating trifluridine/tipiracil + bevacizumab in first-line unresectable mCRC patients who are not eligible for intensive therapy. The criterium ‘non-eligible’ was accorded to the investigator judgement (elderly patients, low tumour burden, severe co-morbidity). The study design was a 1:1 randomization to either trifluridine/tipiracil + bevacizumab (TT-B) or capecitabine + bevacizumab (C-B), standard used dosing. Primary endpoint was PFS, secondary endpoints were OS, ORR, DCR, QoL and safety. In this well-balanced population there was a median PFS of 9.23 months in the TT-B group vs. 7.82 months in the C-B group. A delta of 4.64 months was shown in the OS data, 22.31 months in the TT-B group vs. 17.67 months in the C-B group. There was no intention to make a statistical analysis in this trial. Further efficacy data will be obtained from the ongoing phase III trial, SOLSTICE study NCT 03869892.

Circulating tumour DNA (ctDNA)

The use of circulating tumour DNA (ctDNA) analysis as biomarker in our daily practice becomes more clear and defined. Molecular residual disase (MRD) detection rate postoperative is depending on the stage at time of diagnosis: 10% stage II, 25% stage III and >50% in stage IV. CtDNA has a predictive value >95%  and a specificity >95% for recurrence.

Although a high proportion of patients are treated with curative intent, there are relapse rates around 20-30%. If we could identify molecular residual disease (MRD), we could better stratify patients into low-risk or high-risk patients for recurrence. The latter was the main hypothesis in the trial that was presented, also to assess the post-therapy risk in ctDNA positive patients as well as to determine the lead time for ctDNA detection compared to CT recurrence.

260 patients with stage I,II and III CRC were monitored for 3 years with blood samples post-surgery, then every 3 months and by CT-scan performed at 12 and 36 months. Based on the post-surgery blood samples, they found a high risk of recurrence in the ctDNA positive patients and a low risk in the ctDNA negative group. The same conclusions could be drawn from the post-adjuvant treatment blood samples, namely a high relapse risk in the ctDNA positive group. Furthermore, on longitudinal samples they saw a shift of ctDNA negative patients becoming ctDNA positive, indicating that serial sampling has a benefit over a single time point sampling. And finally, they found that ctDNA had a median lead time of around 8 months compared to detection of recurrence on CT-scans.

When comparing ctDNA with the already established biomarker CEA, not in single time point measurements but in longitudinal samples, all the predictive power is in the ctDNA samples, so that ctDNA is outperforming CEA in recurrence-free survival prediction.

In order to get ctDNA implemented in the clinic, we have to test it in specific clinical settings using randomized trials. This might also lead to a non-radiological monitoring in the ctDNA negative group and a more intense radiological surveillance in the ctDNA positive group. Therefore, different trials are ongoing in stage II and stage III CRC patients. We can only look forward to this new and better biomarker for our clinical practice.


In this topic of the congress, a nice overview was shown on who needs a total neoadjuvant treatment (TNT), based on the PRODIGE 23 and the RAPIDO TRIAL, both presented at last ASCO. Furthermore, they try to define patients in whom organ preservation could be an option, based on the OPRA trial and how to deal with the elderly patients with locally advanced rectal cancer (NACRE trial).

In the PRODIGE 23 trial, 461 patients with T3-T4 advanced rectal cancers were randomized 1:1, well balanced population, into a group (CRT arm) receiving 5 weeks of radiation 50.4 Gy in combination with capecitabine, followed by TME after 7 weeks, followed by 6 months of adjuvant chemotherapy; another group (TNT arm) receiving 6 cycles of neoadjuvant mFOLFIRINOX followed by 5 weeks of radiation 50.4Gy in combination with capecitabine, followed by TME after 7 weeks, followed by 3 months of adjuvant treatment. The investigators could conclude that mFOLFIRINOX was a safe regimen and that in the TNT arm, upfront chemotherapy with mFOLFIRINOX

  • significantly increased the probability of pathological complete response
  • significantly increased the probability of curative intent surgery
  • improved DFS and metastasis-free survival

Therefore, TNT with mFOLFIRINOX should be considered as a new option of care for the initial management of patients with T3-T4 rectal cancers. This was implemented in the updated NCCN guidelines version 6.2020.

In the RAPIDO trial, 920 patients with cT4a/b, extramural vascular invasion, cN2, involved mesorectal fascia, were 1:1 randomly assigned to the standard arm with capecitabine-based chemoradiotherapy (50.4. Gy) followed by TME and optional adjuvant chemotherapy or to the experimental arm with short course radiotherapy (5x5Gy), followed by delayed TME with in the waiting period 18 weeks of chemotherapy (CAPOX or FOLFOX).

The primary endpoint was 3y-DrTF (disease-related treatment failure). They concluded to:

  • a significantly improved number of pCR (14% standard arm vs. 28% in the TNT arm)
  • 7% lower disease-related treatment failure (30.4% standard vs. 23.% TNT)
  • 7% lower distant metastasis rate (26.8% standard vs. 20% TNT)
  • 3-year OS 89% in both groups
  • no differences in surgery, postoperative complications and QoL

This treatment could be considered as a new standard of care, also recommended in the NCCN guidelines version 6.2020.

The OPRA trial, also presented at ASCO last year, could provide us with criteria that we can use for patient selection, for choosing the optimal regimen in a non-operative management (NOM). In this trial, patients with stage II or III distal rectal cancer were randomized to either induction FOLFOX or CAPOX followed by chemoradiation, or chemoradiation followed by consolidation FOLFOX or CAPOX. After restaging, patients that had a significant clinical response underwent NOM, while those that did not have a significant clinical response underwent TME.

  • DFS and distant metastases-free survival were not statistically different between the two groups
  • The rates of organ preservation were significantly higher in the consolidation arm (60% vs. 40% in the induction arm)
  • NOM was feasible in over half of the patients in the consolidation arm

Recently published ASTRO guidelines provide us with practical guidance on the role of NOM. They state that NOM is conditionally recommended if a clinical complete response is achieved in patients who would have a permanent colostomy or inadequate bowel function with TME and in patients who decline TME and agree to close follow-up.

The ASTRO guidelines recommend a regimen of treatment, but this recommendation however has the data of the GERMAN trial and the OPRA trial outside the review window. The ASTRO guidelines also help us for a practical response assessment.


SINCE OUR PATIENT POPULATION IS AGING, and over 40% of patients are over 75 years old, with more severe toxicities, SOME CRITICAL EVALUATION IN THIS GROUP OF PATIENTS IS WARRANTED to find the best compromise between cure of the cancer and QoL.

The NACRE PRODIGE 42 GERICO 12 is a prospective trial were 103 patients of 75 years or more were randomized 1:1 to the standard of care arm (radiochemotherapy plus capecitabine followed by delayed surgery) or to the experimental arm (short course radiation 5x5Gy followed by delayed surgery). Two primary objectives in this trial: first a non-inferiority R0 resection and second a superiority for preservation of autonomy.

The median age was 80 years, one third of the patients had a performance status 0 and there was no difference of tumour characteristics between groups. 24% of patients in the standard arm had severe toxicities compared to 12% in the experimental arm. There was no difference in surgical procedures between the groups and there was no statistical difference in the ypT stage and in the N stage; although they saw a greater downstaging in the standard arm.

R0 resection rate was 90% in the standard arm, 86% in the experimental arm, statistically significant for non-inferiority. There was also a statistical advantage for OS in the experimental arm. They further conclude that the short course radiotherapy was associated with a better tolerance and a better compliance. Due to the limited number of patients, these results however need to be confirmed.

Neuroendocrine tumors (NETs)

The current treatment options for NEN (= NET + NEC) are:

  • Surgery
  • Somatostatin-analogues
  • Everolimus
  • Sunitinib/surufatinib
  • PRRT
  • Chemotherapy: cisplatin/carboplatin+etoposide, CAPTEM, FOLFOX, FOLFIRI
  • Liver directed therapy: bland embolisation, chemoembolisation and SIRT
  • Clinical studies
  • Immunotherapy: atezolizumab and durvalumab in first-line SCLC, in the other NENs unclear position: possibly when lung origin and in GEP NECs…

Rocio Garcia-Carbonero discussed at ASCO GI the results of the phase II/III AXINET trial.

Angiogenesis plays an important role in NET development and progression. Sunitinib is reimbursed in the treatment of pancreatic NETs.

Axitinib is a potent and selective VEGFR-1,2,3 inhibitor, with proven activity against several vascular-dependent solid tumors. The aim of this randomized, double-blind phase II/III study was to assess the efficacy of axitinib in patients with advanced G1-2 extra-pancreatic NETs. Eligible pts were randomized (1:1) to receive octreotide LAR (30 mg IM q4w) with axitinib (5 mg BID) or placebo BID until disease progression or unacceptable toxicity. Pts were stratified by time from diagnosis to study entry ( > or < 12m), primary tumour site (GI tract vs. non-GI) and Ki-67 index (< 5% vs. > 5%). Prior therapy with SSA, IFN and up to 2 lines of systemic treatment was allowed, but not prior VEGF- or VEGFR-targeted drugs. Clinical and/or radiological disease progression within 12 months prior to study entry was required. The primary endpoint was progression-free survival (PFS). Secondary endpoints included overall survival (OS), time to progression, overall response rate (ORR), duration of response, biochemical response and safety. 256 pts were randomized (106 in the phase II part and 150 additional pts in the phase III part), 126 to axitinib and 130 to placebo. The main characteristics of the study population were: median age 61 years (range: 21-85), 52% male, PS 0-1 (64-35%), G1-2 (29%-71%), primary tumour site GI (40%) – lung (17%) – other (32%). Prior therapies included: SSA (46%), everolimus (13%), chemotherapy (13%), TACE (5%) and PRRT (2%). ORR was significantly higher in axitinib- vs. placebo-treated patients (17.5% vs. 3.8%, p=0.0004). PFS per investigator assessment also favored axitinib vs. placebo-treated patients, although the difference did not reach statistical significance (median PFS 17.2 vs. 12.3 months, respectively, HR 0.816, p=0.169). Grade 3-4 treatment-related AEs occurred more frequently in the axitinib vs. placebo arm (52% vs. 13.8%) and included hypertension (21% vs. 6%), cardiac disorders (3.2% vs. 0.7%), diarrhoea (13% vs. 1.5 %), asthenia (9% vs. 3%) and nausea&vomiting (2% vs. 0.7%). There were 3 treatment-related deaths, 1 in the axitinib arm (cardiac failure) and 2 in the placebo arm (myocardial infarction and hepatorenal syndrome). Although the study failed to demonstrate a significant PFS benefit per investigator assessment, axitinib in combination with octreotide LAR demonstrated activity and had a tolerable safety profile in patients with advanced G1-2 extra-pancreatic NETs.

At this moment thus, axitinib is not taking in a position in second-line/third-line treatment of extra-pancreatic NETs (in contrast to surufatinib), PRRT remains first choice treatment here.

Acknowledgement: Special word of thanks to Bayer for delivering contributing articles.


1. Kato K et al. Nivolumab in advanced esophageal squamous cell carcinoma (ATTRACTION-1/ONO-4538-07): Minimum of five-year follow-up. Poster presented at ASCO GI 2021.

2. Chin K et al. Three-year follow-up of ATTRACTION-3: A phase III study of nivolumab (Nivo) in patients with advanced esophageal squamous cell carcinoma (ESCC) that is refractory or intolerant to previous chemotherapy. Poster presented at ASCO GI 2021.

3. Mansoor W et al. Health-related quality of life (HRQoL) of pembrolizumab plus chemotherapy versus chemotherapy as first-line therapy in patients with advanced esophageal cancer: The phase III KEYNOTE-590 study. Abstract presented at ASCO GI 2021.

4. Van CutsemE et al. Checkmate 577: Health-related quality of life (HRQoL) in a randomized, double-blind phase III study of nivolumab (NIVO) versus placebo (PBO) as adjuvant treatment in patients (pts) with resected esophageal or gastroesophageal junction cancer (EC/GEJC). Abstract presented at ASCO GI 2021.

5. Chung HC et al. LEAP-005: A phase II multicohort study of lenvatinib plus pembrolizumab in patients with previously treated selected solid tumors: Results from the gastric cancer cohort. Poster presented at ASCO GI 2021.

6. Shitara K et al. Trifluridine/tipiracil versus placebo in patients with heavily pretreated metastatic gastric cancer (TAGS): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Oncology 2018;19 (11):1437-1448.

7. Tabernero J et al. Trifluridine/tipiracil outcomes in third- or later lines versus placebo in metastatic gastric cancer treatment: An exploratory subgroup analyses from the TAGS study. Poster presented at ASCO GI2021.

8. Bang YJ et al. Trastuzumab in combination with chemotherapy versus chemotherapy alone for treatment of HER2-positive advanced gastric or gastro-oesophageal junction cancer (ToGA): a phase 3, open-label, randomised controlled trial. Lancet2010; 376: 687–97.

9. Wainberg ZA et al. Randomized double-blind placebo-controlled phase 2 study of bemarituzumab combined with modified FOLFOX6 (mFOLFOX6) in first-line (1L) treatment of advanced gastric/gastroesophageal junction adenocarcinoma (FIGHT). Abstract orally presented at ASCO GI 2021.

10. Borad MJ et al. Silmitasertib (CX-4945) in combination with gemcitabine and cisplatin as first-line treatment for patients with locally advanced or metastatic cholangiocarcinoma: a phase Ib/II study. Abstract presented at ASCO GI 2021.

11. Abou-Alfa GK et al. Effects of FGFR2 alterations on survival in patients receiving systemic chemotherapy for intrahepatic cholangiocarcinoma. Abstract presented at ASCO GI 2021.

12. Valle JW et al. Longitudinal evaluation of quality of life (QoL) in patients (pts) with FGFR2-driven cholangiocarcinoma (CCA) treated with pemigatinib. Abstract presented at ASCO GI 2021.

13. Javle MM et al. Final results from a phase II study of infigratinib (BGJ398), an FGFR-selective tyrosine kinase inhibitor, in patients with previously treated advanced cholangiocarcinoma harboring an FGFR2 gene fusion or rearrangement. Abstract presented at ASCO GI 2021.

14. Zhu AX et al. Final results from ClarIDHy, a global, phase III, randomized, double-blind study of ivosidenib (IVO) versus placebo (PBO) in patients (pts) with previously treated cholangiocarcinoma (CCA) and an isocitrate dehydrogenase 1 (IDH1) mutation. Abstract presented at ASCO GI 2021.

15. Meric-Bernstam F et al. Zanidatamab (ZW25) in HER2-positive biliary tract cancers (BTCs): results from a phase I study. Abstract presented at ASCO GI 2021.

16. Harding JJ et al. Targeting HER2 (ERBB2) mutation-positive advanced biliary tract cancers with neratinib: results from the phase II SUMMIT ‘basket’ trial. Abstract presented at ASCO GI 2021.

17. Katz MHG et al. Alliance A021501: Preoperative mFOLFIRINOX or mFOLFIRINOX plus hypofractionated radiation therapy (RT) for borderline resectable (BR) adenocarcinoma of the pancreas. Abstract presented at ASCO GI 2021.

18. Shaib WL et al. Phase II randomized, double-blind study of mFOLFIRINOX plus ramucirumab versus mFOLFIRINOX plus placebo in advanced pancreatic cancer patients (HCRN GI14-198). Abstract presented at ASCO GI 2021.

19. Golan T et al. Overall survival from the phase 3 POLO trial: maintenance olaparib for germline BRCA-mutated metastatic pancreatic cancer. Abstract presented at ASCO GI 2021.

20. Finn RS et al. Atezolizumab plus Bevacizumab in Unresectable Hepatocellular Carcinoma. N Engl J Med. 2020;382(20):1894-905.

21. Finn RS et al. IMbrave150: Updated overall survival (OS) data from a global, randomized, open-label phase III study of atezolizumab (atezo) + bevacizumab (bev) versus sorafenib (sor) in patients (pts) with unresectable hepatocellular carcinoma (HCC); Gastrointestinal Cancers Symposium 2021: American Society of Clinical Oncology.

22. El-Khoueiry AB et al. Nivolumab (NIVO) plus ipilimumab (IPI) combination therapy in patients (Pts) with advanced hepatocellular carcinoma (aHCC): Long-term results from CheckMate 040; Gastrointestinal Cancers Symposium 2021: American Society of Clinical Oncology.

23. Laethem J-LV et al. Pembrolizumab (pembro) monotherapy for previously untreated advanced hepatocellular carcinoma (HCC): Phase II KEYNOTE-224 study; Gastrointestinal Cancers Symposium 2021: American Society of Clinical Oncology.

24. Yarchoan M et al. Feasibility and efficacy of neoadjuvant cabozantinib and nivolumab in patients with borderline resectable or locally advanced hepatocellular carcinoma. (HCC); Gastrointestinal Cancers Symposium 2021: American Society of Clinical Oncology.

25. Lewandowski R et al. Use of yttrium-90 (Y90) glass microspheres (TheraSphere) as neoadjuvant to transplantation/resection in hepatocellular carcinoma: Analyses from the LEGACY study; Gastrointestinal Cancers Symposium 2021: American Society of Clinical Oncology.

26. Shiu KK et al. KEYNOTE-177: Phase III randomized study of pembrolizumab versus chemotherapy for microsatellite instability-high advanced colorectal cancer. Abstract presented at ASCO GI 2021.

27. Van Cutsem E et al. Trifluridine/tipiracil plus bevacizumab in patients with untreated metastatic colorectal cancer ineligible for intensive therapy: the randomized TASCO1 study. Ann Oncol. 2020;31(9):1160-1168.

28. Henriksen T et al. Circulating tumor DNA analysis for assessment of recurrence risk, benefit of adjuvant therapy, and early relapse detection after treatment in colorectal cancer patients. Abstract presented at ASCO GI 2021.

29. Dasari A et al. Circulating Tumor DNA. Oral presentation at ASCO GI 2021.

30. Fernandez-Martos C et al. Who Needs Total Neoadjuvant Therapy? Oral presentation at ASCO GI 2021.

31. Conroy T et al. Total neoadjuvant therapy with mFOLFIRINOX versus preoperative chemoradiation in patients with locally advanced rectal cancer: Final results of PRODIGE 23 phase III trial, a UNICANCER GI trial. Oral presentation at ASCO 2020.

32. Bahadour et al. Short-course radiotherapy followed by chemotherapy before total mesorectal excision (TME) versus preoperative chemoradiotherapy, TME, and optional adjuvant chemotherapy in locally advanced rectal cancer (RAPIDO): a randomised, open-label, phase 3 trial. Lancet Oncology 2021 Jan;22(1):29-42.

33. Garcia-Aguilar J et al. Preliminary results of the organ preservation of rectal adenocarcinoma (OPRA) trial. J Clin Oncol 38,2020(suppl:abstract 4008).

34. Wo JY et al. Radiation Therapy for Rectal Cancer: Executive Summary of an ASTRO Clinical Practice Guideline. Pract Radiat Onc 2021.

35. Francois E et al. NACRE: A randomized study comparing short course radiotherapy with radiochemotherapy for locally advanced rectal cancers in the elderly-Preliminary results. Abstract presented at ASCO GI 2021.

36. Garcia-Carbonero R et al. A phase II/III randomized double-blind study of octreotide acetate with axitinib versus octreotide acetate with placebo in patients with advanced G1-G2 NETs of non-pancreatic origin (AXINET trial-GETNE-1107). Abstract orally presented at ASCO GI 2021.