Highlights of the ASCO Gastrointestinal Symposium 2014

More than 3,300 physicians, researchers, patient advocates and healthcare professionals from over 50 countries attended the American Society of Clinical Oncology (ASCO) Gastrointestinal Symposium in San Francisco (January 14-16, 2014), which featured the latest research on gastrointestinal cancer treatment and prevention. Results from several key studies presented at this meeting are outlined below.

RAINBOW (Abstract LBA7)

A second phase III trial of ramucirumab-based therapy for advanced gastric/gastroesophageal junction (GEJ) adenocarcinoma boosts support for the vascular endothelial grown factor receptor-2 (VEGFR-2) antagonist in the second-line setting. 

Whereas the REGARD trial showed that single-agent ramucirumab significantly improved overall survival (OS) from 3.8 to 5.2 months (p = 0.0473) when added to best supportive care, new data on combining ramucirumab with paclitaxel from the RAINBOW trial, the largest phase III trial of second-line therapy for advanced gastric/GEJ adenocarcinoma to date, increases this survival benefit even further to nearly 10 months.

The global, randomised, double-blind, placebo-controlled, RAINBOW trial of 665 patients with disease progression on first-line platinum/fluoropyrimidine-based combination therapy showed that the addition of ramucirumab to paclitaxel significantly prolonged the primary endpoint of OS from a median of 7.36 to 9.63 months (p = 0.0169). The difference between arms ultimately translated into a 19% reduction in the risk of death with inclusion of ramucirumab.

A consistent benefit from the addition of ramucirumab to paclitaxel was evident across all efficacy endpoints. Notably, pairing the monoclonal antibody with chemotherapy significantly prolonged median progression-free survival from 2.86 months to 4.40 months (p < 0.0001). Furthermore, inclusion of ramucirumab nearly doubled the overall response rate compared with paclitaxel alone (28% versus 16%; p = 0.0001) and significantly increased the disease control rate (80% versus 64%; p < 0.0001).

Ramucirumab was relatively well tolerated, although the benefits attained with the agent did come at the cost of a greater incidence of ≥ grade 3 neutropenia (40.7% versus 18.8%), leukopenia (17.4% versus 6.7%), hypertension (14.7% versus 2.7%), and fatigue (11.9% versus 5.5%). 

These adverse events did not, however, lead to increased treatment discontinuation in the ramucirumab arm, nor were rates of treatment-related deaths different between the two arms (4.0% with ramucirumab/paclitaxel versus 4.6% with paclitaxel alone). 

CAIRO3 (Abstract LBA388)

Final results from the phase III CAIRO3 trial provide guidance to clinicians who are uncertain about just how long a treatment holiday to give patients with metastatic colorectal cancer (mCRC) following bevacizumab-containing induction therapy. In a comparison of maintenance therapy with capecitabine and bevacizumab versus observation after six cycles of first-line capecitabine/oxaliplatin plus bevacizumab (CAPOX-B), the maintenance approach markedly improved progression-free survival (PFS) and yielded OS benefits in select patient subgroups. 

The 558 patients in this trial with stable disease or better after CAPOX-B were randomly assigned to undergo observation or to receive 625mg/m2 of oral capecitabine twice daily in combination with 7.5mg/kg of intravenous bevacizumab every three weeks on a continuous basis. At the time of first disease progression (PFS1), patients in both arms were retreated with CAPOX-B until second progression (PFS2) occurred.

Over the median follow-up period of 48 months, CAPOX-B was reintroduced in 60% of patients in the observation arm as compared with 47% in the maintenance arm.

The findings revealed a significant improvement in median PFS2, the primary study endpoint, with maintenance treatment versus observation (11.7 versus 8.5 months; p < 0.0001). Maintenance therapy also conferred a significant increase in median PFS1 as compared with observation (8.5 versus 4.1 months; p < 0.0001) and significantly improved the median time to second progression (13.9 versus 11.1 months; p < 0.0001). 

Importantly, the benefits observed with combination maintenance capecitabine and bevacizumab did not come at the expense of patient quality of life. 

The improvements in disease progression timing did not translate into a statistically significant OS benefit in the total study population (median OS: 21.6 months with maintenance therapy versus 18.1 months with observation; p = 0.22). However, this could have been affected by the greater proportion of patients with synchronous lesions in the maintenance arm versus the observation arm (79% versus 68%).

Subsequent multivariate analysis, evaluating the interactions between treatment and baseline covariates, found that maintenance therapy independently conferred a marked OS improvement in two groups of patients:

• Those with synchronous disease who underwent resection of the primary tumour (median OS: 25.0 months with maintenance therapy versus 18.0 months with observation; log-rank p < 0.0001)
• Those who attained a complete or partial response to induction therapy prior to randomisation (median OS: 24.1 months with maintenance therapy versus 18.8 months with observation; log-rank p < 0.0001).

Analysis of KRAS/NRAS mutations (Abstract LBA387)

More extensive genetic testing for RAS gene mutations beyond routine analysis of KRAS exon 2 may soon become a new standard of care to pinpoint which patients stand to benefit from anti-epidermal growth factor receptor (EGFR) therapy in the treatment of metastatic colorectal cancer (mCRC).

Whereas activating mutations in KRAS exon 2, present in approximately 40-50% of mCRC tumours, have long been known to blunt the response to EGFR-targeted agents in mCRC, new data point to a longer list of problematic mutations within the RAS gene family. Research was presented highlighting that activating mutations in exons 2, 3, and 4 of both KRAS and NRAS all predict a lack of response to panitumumab in the second-line setting, corroborating similar recent findings from the PRIME and PEAK trials regarding panitumumab efficacy in the first-line setting.

The data presented was obtained by a retrospective analysis of the randomised phase III 20050181 trial, which compared second-line treatment with FOLFIRI, both with and without panitumumab, in 1,186 patients with mCRC who were previously treated with oxaliplatin and bevacizumab. Banked tumour specimens previously classified as wild-type KRAS exon 2 were reanalysed to detect additional mutations in exons 3 and 4 of KRAS and exons 2, 3, and 4 of NRAS. The investigators successfully identified RAS mutational status in 85% of these 1,186 individuals.

Of note, 18% of patients with mCRC featuring wild-type KRAS exon 2 harboured additional RAS activating mutations that predict lack of response to panitumumab. In accord, improvements emerged when PFS and OS, the two co-primary study endpoints, were re-evaluated based on the new mutational profiles. Whereas the addition of panitumumab to FOLFIRI yielded an OS hazard ratio of 0.85 (p = 0.12) in the wild-type KRAS exon 2 population, the hazard ratio improved to 0.803 (p = 0.08) in patients with wild-type KRAS and NRAS. Similarly, the PFS hazard ratio improved from 0.73 (p = 0.004) among individuals with wild-type KRAS exon 2 population to 0.695 (p = 0.006) when restricted to the wild-type RAS group.

Two other important observations in the wild-type RAS subgroup of patients include:

• Improvement in the objective response rate in the second-line setting to 41% with the addition of panitumumab to FOLFIRI, as compared with 10% for FOLFIRI alone
• No additional toxicity or safety signals emerged in the wild-type RAS subgroup when combining panitumumab with FOLFIRI.

NSABP R-04 (Abstract 390)

New findings from a four-arm phase III clinical trial in patients with stage 2/3 rectal cancer indicate that combining pre-operative (neoadjuvant) radiation with either oral capecitabine or infusional 5-fluorouracil (5-FU) results in equivalent outcomes. The researchers also found that adding oxaliplatin to either of these regimens did not provide further benefit but increased overall treatment toxicity. Oral capecitabine is more convenient for patients.

In this four-arm clinical trial, 1,608 patients were randomly assigned to receive five weeks of radiation plus 5-FU (arm 1, 477 patients); 5-FU/oxaliplatin (arm 2, 329 patients); capecitabine (arm 3, 472 patients); or capecitabine/oxaliplatin (arm 4, 330 patients). Patients received one of these treatments for five weeks prior to surgical resection.

There were no significant differences between treatment arms in terms of local-regional control, disease-free survival, and OS. The three-year local-regional control rates ranged from 87.4 to 88.2%. In cases of ‘R0’ surgical resections (complete removal of all tumour; surgical margins are ‘clear’ or negative), 2-4% of stage 2 patients and 4-11% of stage 3 patients had a local recurrence within three years. In each of the treatment arms, about 80% of patients were still alive five years after surgery.

Infusional 5-FU and capecitabine had similar side effects. Patients who also received oxaliplatin, however, experienced significantly more diarrhoea and fatigue.

EVOLVE-1 (Abstract 172)

The treatment gap following sorafenib failure in advanced hepatocellular carcinoma (HCC) remains. New findings from the randomised phase III EVOLVE-1 trial show that the mTOR inhibitor everolimus confers no survival benefit when pitted against placebo in the salvage setting.

Andrew X Zhu, MD, PhD, discusses Abstract 172 EVOLVE-1: Phase 3 study of everolimus for advanced HCC that progressed during or after sorafenib. Photo by © ASCO/Todd Buchanan 2014(click to enlarge)

EVOLVE-1 was a large, global, randomised, placebo-controlled trial of 546 adult patients with Barcelona Clinic Liver Cancer stage B/C HCC and Child-Pugh class A liver function who had progressive disease following sorafenib treatment. More than 80% of patients in the trial discontinued prior sorafenib because of disease progression, whereas the balance (19%) stopped sorafenib because of intolerance. Patients were randomly assigned in a 2:1 ratio to receive everolimus (362 patients) or placebo (184 patients), both in combination with best supportive care.

At the end of the trial, the primary endpoint of overall survival (OS) was negative. Median OS was comparable between the two arms at 7.6 months with everolimus compared with 7.3 months with placebo (p = 0.675). The median time to progression also proved similar between arms at 3 months with everolimus versus 2.6 months with placebo.

A significant improvement in the disease control rate did emerge with everolimus as compared with placebo (p = 0.010). However, the overall response rate showed no difference between arms (2.2% with everolimus versus 1.6% with placebo).

The safety profile of everolimus in EVOLVE-1 matched that from other everolimus clinical trials in cancer, with no emergence of new safety signals. Of interest, grade 3/4 stomatitis (2.8%) and grade 3/4 non-infectious pneumonitis (1.6%) occurred infrequently with everolimus.

Although no virologic flares arose in patients infected with hepatitis C virus, hepatitis B virus (HBV) reactivation occurred more often with everolimus as compared with placebo (29 versus 10 patients). 

Tumour biomarker analyses are on going in EVOLVE-1, which may help to identify subgroups of patients with HCC who stand to benefit from everolimus treatment.