thanks for great summary!
ASCO 2013 Report — Gastrointestinal Cancer
ASCO 2013 Report — Gastrointestinal Cancer
- David H. Ilson, MD, PhDAU1252
Highlights of the latest research in pancreatic, gastric, and colorectal cancer
- David H. Ilson, MD, PhDAU1252
At this year's annual meeting of the American Society of Clinical Oncology (ASCO 2013), held May 31 through June 5 in Chicago, specialists discussed the latest findings in cancer research. The editors of NEJM Journal Watch Oncology and Hematology were in attendance to report highlights of the conference. Here, David Ilson, MD, reviews key presentations on new gastrointestinal cancer treatments.
A New Standard Therapy Option for Pancreatic Cancer
Von Hoff and colleagues reported results from the phase III, randomized IMPACT trial (Abstract 4005), which compared conventional gemcitabine therapy with nanoparticle albumin-bound (nab)-paclitaxel plus gemcitabine in 861 patients with advanced pancreatic cancer. Patients received either gemcitabine (1000 mg/m2) weekly for 7 weeks with 1 week off and then on days 1, 8, and 15 every 4 weeks or gemcitabine (1000 mg/m2) plus nab-paclitaxel (125 mg/m2) on days 1, 8, and 15 every 4 weeks. Most patients (93%) had Karnofsky performance status of 80 to 100, and 84% had metastatic disease to the liver. Combination therapy with gemcitabine and nab-paclitaxel was superior to gemcitabine in all endpoints, including overall survival (OS; 8.5 vs. 6.7 months; hazard ratio, 0.72; P=0.000015), progression-free survival (PFS; 5.5 vs. 3.7 months; HR, 0.69; P=0.000024), and response (23% vs. 7%; P=1.1 x 10-10). Grade 3 or 4 toxicities that were higher with nab-paclitaxel included neutropenia (38% vs. 27%), neuropathy (17% vs. 1%), fatigue (17% vs. 7%), and diarrhea (6% vs. 1%). Gemcitabine plus nab-paclitaxel represents a new standard-therapy option for advanced pancreatic cancer. Although superior to gemcitabine alone, this regimen did not achieve the survival benefit reported for FOLFIRINOX (oxaliplatin, irinotecan, leucovorin, and fluorouracil) versus gemcitabine (NEJM JW Oncol Hematol May 11 2011).
Lipid-Conjugated Gemcitabine Not Superior for Pancreatic Cancer
Poplin and colleagues reported data from the phase II, randomized, controlled LEAP study (Abstract 4007), which compared conventional gemcitabine therapy with CO-101 (a lipid-drug conjugate of gemcitabine) in 367 patients with advanced pancreatic cancer. Because the uptake of gemcitabine into cells may depend on expression of human equilibrative nucleoside transporter-1 (hENT1), tumors with low-hENT1 expression may be less responsive to gemcitabine. Thus, lipid conjugation may enhance cellular uptake of gemcitabine in such hENT1-low tumors. Patients received either gemcitabine (1000 mg/m2) weekly for 7 weeks with 1 week off and then on days 1, 8, and 15 every 4 weeks or CO-101; patients underwent tumor immunohistochemical testing for hENT1 expression. Outcomes were similar with gemcitabine versus CO-101 in all patients (HR, 1.072) and in those with low-hENT1–expressing tumors (HR, 0.994). In the gemcitabine arm, outcome also did not differ as a function of high- versus low-hENT1 expression (HR, 1.147). In metastatic pancreatic cancer, hENT1 expression was neither predictive nor prognostic regarding outcome to gemcitabine treatment, and a lipid-conjugated gemcitabine was not superior in patients regardless of hENT1 expression.
Low-hENT1 Level Predicts Gemcitabine Benefit in Pancreatic Cancer
Contrary to results of the LEAP study (see above), findings by Neoptolemos and colleagues (Abstract 4006) indicated a potential prognostic and predictive benefit of hENT1 expression regarding adjuvant treatment with gemcitabine or 5-fluorouracil (5-FU) after pancreatic cancer resection. HENT1 expression was studied in 352 patients undergoing resection for pancreatic cancer: 176 from the ESPAC 1 trial of adjuvant 5-FU with or without radiotherapy versus no treatment (N Engl J Med 2004; 350:1200) and 176 from the ESPAC 3 trial of adjuvant 5-FU versus adjuvant gemcitabine (JAMA 2012; 308:147). Low- versus high-hENT1 expression was associated with shorter median survival in gemcitabine-treated patients (17.1 vs. 26.2 months; P=0.002) but not with a significant survival difference in 5-FU–treated patients (21.9 and 25.6 months; P=0.362). Multivariate analysis indicated that hENT1 expression was a predictive marker for adjuvant gemcitabine but not for 5-FU. These provocative data, in contrast to the negative predictive utility of hENT1 expression in metastatic pancreatic cancer, require further validation and potential testing in prospective trials.
S-1 Superior to Gemcitabine in Resected Pancreatic Cancer
Fukotomi and colleagues presented data from the phase III, randomized JASPAC 01 trial (Abstract 4008), which compared adjuvant therapy with gemcitabine versus S-1 in 385 patients with resected pancreatic cancer, most of whom had R0 resection (86%–88%) and node-positive disease (62%–64%). Patients received gemcitabine (1000 mg/m2) weekly for 3 weeks on and 1 week off per month for 6 months or S-1 (80–100 mg/m2/day) for 4 weeks on and 2 weeks off for 6 months. Relapse-free survival at 2 years was significantly better with S-1 versus gemcitabine (49% vs. 29%; HR, 0.56; P<0.0001), as was OS at 2 years (70% vs. 53%; HR, 0.56; P<0.0001). S-1 had a lesser degree of hematologic toxicity. The superiority of S-1 over gemcitabine was surprising, given that the ESPAC 3 data indicated parity for adjuvant 5-FU versus gemcitabine. Further study and validation are required.
Adding Radiotherapy to Chemotherapy in Pancreatic Cancer
Hammel and colleagues presented results of the phase III LAP 07 trial (Abstract 4003), which compared gemcitabine versus gemcitabine followed by capecitabine and radiotherapy in 722 patients with locally advanced, unresectable pancreatic cancer. Patients were randomized initially to receive standard gemcitabine chemotherapy versus gemcitabine plus erlotinib. The 420 patients who maintained either stable disease or a response at 4 months were then randomized to continue chemotherapy or switch to capecitabine plus concurrent 5040 cGy of radiotherapy. OS was similar in patients assigned to chemotherapy versus chemoradiotherapy (16.4 and 15.2 months, respectively; HR 1.03; P=0.83). There was also no difference in OS in all patients assigned to gemcitabine or gemcitabine plus erlotinib (13.6 and 11.9 months, HR 1.19; P=0.093); PFS was also similar (10.7 vs. 9.3 months; P=0.1526). The results from this study address two important issues: Adding radiotherapy to chemotherapy in locally advanced and unresectable pancreatic cancer is not superior to continuing chemotherapy alone, and the addition of erlotinib to chemotherapy in these patients also failed to improve outcome.
Lapatinib plus Chemotherapy for HER2-Positive Gastric Cancer
Hecht and colleagues presented data from the phase III, randomized, double-blind LOGIC trial (Abstract LBA4001), which compared treatment with capecitabine (1700 mg/m2) for 14 days every 21 days and oxaliplatin (130 mg/m2) on day 1 with or without lapatinib (1250 mg daily) for human epidermal growth factor receptor-2 (HER2)-positive gastric cancer. The trial involved 545 patients with esophageal and gastroesophageal junction cancer (11%–14%) and distal gastric cancer (86%–89%); 40% of patients were from Asia. OS trended toward improvement with lapatinib recipients versus nonrecipients (12.2 and 10.5 months; HR, 0.91; P=0.35); no difference in PFS was observed (6.0 and 5.4 months; HR, 0.86; P=0.10). Response was increased from 40% to 53% with lapatinib. The subset of patients treated in Asia achieved a survival benefit with lapatinib (10.9 vs. 16.5 months; HR, 0.68). Serious adverse events were higher with lapatinib (27% vs. 19%). These findings, which indicate a potential benefit from lapatinib in Asian gastric cancer patients, contrast with results from the previously reported TOGA trial (NEJM JW Oncol Hematol Sep 14 2010), in which all HER2-positive patients globally benefited from the addition of trastuzumab to chemotherapy.
Minimal vs. Intensive Screening After Colorectal Cancer Resection
Mant and colleagues reported results from the randomized, controlled FACS trial (Abstract 3500), which evaluated the utility of screening strategies during follow-up after surgery for 1202 patients with stage I–III colorectal cancer. Minimal testing (a single follow-up computed tomography [CT] scan at 12–18 months) was compared with more-intensive testing: one CT scan plus carcinoembryonic antigen [CEA] testing ever 3 months for 2 years and every 6 months for the next 3 years, with or without intensive CT scan follow-up every 6 months for 2 years, then annually for the next 3 years. Detection of potentially resectable recurrent disease was superior for CEA screening plus one CT scan compared with minimal testing (6.7% vs. 2.3%; odds ratio 2.7), but the addition of intensive CT scan screening with or without CEA screening did not increase detection rates (5.8% and 8.0%). No strategy led to any relative survival improvements. The authors concluded that a reasonable screening strategy includes CEA screening and a single follow-up CT scan at 12–18 months after resection of stage I–III colorectal cancer.
Calcium and Magnesium Ineffective with FOLFOX for Colon Cancer
Loprinzi and colleagues reported results of the phase III, randomized, placebo-controlled, double-blind N08CB trial (Abstract 3501), which evaluated the addition of calcium and magnesium infusion to FOLFOX (fluorouracil, leucovorin, and oxaliplatin) chemotherapy as a strategy to reduce neurologic toxicity. A total of 353 patients with colon cancer undergoing adjuvant FOLFOX infusion were randomized to one of three treatment arms: FOLFOX alone, calcium (1 g) and magnesium (1 g) administered before FOLFOX, or calcium and magnesium administered before and after FOLFOX. There was no difference in either acute or chronic neurologic toxicities for either schedule of calcium and magnesium infusion compared with chemotherapy alone, and no difference in administered doses of oxaliplatin. These important data indicate that the practice of calcium and magnesium infusion for oxaliplatin should be discontinued.
Maintenance Chemotherapy for Metastatic Colorectal Cancer
Two studies evaluated maintenance chemotherapy strategies for patients with metastatic colorectal cancer. In the phase III, randomized CAIRO 3 trial (Abstract 3502), Koopman and colleagues studied 558 previously untreated patients who initially received 6 cycles of capecitabine (1000 mg/m2) twice daily for 14 days every 21 days and oxaliplatin (130 mg/m2) plus bevacizumab (7.5 mg/kg) on day 1. Patients with response or stable disease were randomized to observation or continuous capecitabine (625 mg/m2 twice daily) plus bevacizumab (7.5 mg/mg every 3 weeks). The combination of capecitabine, oxaliplatin, and bevacizumab was resumed at disease progression. At median follow-up of 33 months, PFS was significantly longer with active maintenance therapy versus observation (7.4 vs. 4.1 months; HR 0.44; P<0.0001). PFS after subsequent resumption of full therapy was not different between the observation and maintenance groups (10.4 months), but OS favored maintenance therapy over observation (21.7 vs. 17.9 months; HR 0.77; P=0.02). These improvements in PFS and OS suggest that maintenance therapy with capecitabine and bevacizumab may offer a modest survival benefit versus observation in metastatic colorectal cancer.
In the phase III, randomized, open-label, multicenter, noninferiority, SAKK 41/06 trial (Abstract 3503), Koeberle and colleagues studied 262 Swiss patients with metastatic colorectal cancer who achieved stable disease or response to initial chemotherapy plus bevacizumab. Patients were then randomized to a maintenance strategy of observation alone versus continuation of bevacizumab (7.5 mg/kg every 3 weeks). At median follow-up of 28.6 months, bevacizumab versus observation led to modestly longer time to progression (17.9 vs. 12.6 weeks; HR, 0.72) as well as minimally longer PFS (9.5 vs. 8.5 months; HR, 0.73) and OS (24.9 vs. 22.8 months; HR, 0.87). The results failed to demonstrate noninferiority for observation. The limited difference in treatment arms and cost-analysis data argue against the use of bevacizumab alone as a maintenance strategy in metastatic colorectal cancer. Data from the two trials support the use of single-agent fluoropyrimidine chemotherapy plus bevacizumab as maintenance therapy compared to observation alone or bevacizumab alone.
Cetuximab for KRAS Wild-Type Metastatic Colorectal Cancer
Two trials evaluated the use of cetuximab for patients with KRAS wild-type metastatic colorectal cancer. Primrose and colleagues reported results from the new randomized, EPOC trial (Abstract 3504), which compared perioperative chemotherapy alone with or without cetuximab in 271 patients with resectable liver metastases from colorectal cancer. Patients were treated with preoperative FOLFOX or FOLFIRI (fluorouracil, leucovorin, and irinotecan) with or without cetuximab before and after hepatic resection. The trial was closed early given the significant inferiority in PFS in the cetuximab arm (14.1 vs. 20.5 months; HR, 1.49) and no difference in rates of hepatic resection with cetuximab versus chemotherapy alone (90% and 87%). OS data, although not mature, also indicated a trend for inferior survival in the cetuximab arm (HR, 1.48). Given that these and other recent data regarding the adjuvant use of cetuximab in colon cancer indicate potentially worse outcomes with cetuximab, this agent should not be used as part of chemotherapy in the setting of hepatic resection.
Heinemann and colleagues reported results from the randomized, AIO KRK-0306 (FIRE-3) trial (Abstract LBA3506), which compared FOLFIRI with either bevacizumab or cetuximab as first-line chemotherapy in 592 German patients with KRAS wild-type metastatic colorectal cancer. Response rate (the primary endpoint) did not differ with bevacizumab versus cetuximab (58% and 62%; P=0.183), nor did PFS (10.3 vs 10.0 months; HR 1.06; P=0.547). However, a survival difference emerged beyond 24 months of follow-up favoring cetuximab (28.7 vs. 25 months; HR, 0.77; P=0.017). The authors could not explain the absence of a difference in response or PFS for the two treatment arms leading to a late separation in survival curves. Further follow-up on this trial is awaited.
FOLFIRI vs. FOLFOXIRI for Metastatic Colorectal Cancer
Falcone and colleagues updated data from the phase III, randomized, TRIBE trial (Abstract 3505), which compared treatment with FOLFIRI plus bevacizumab versus FOLFOXIRI plus bevacizumab in 508 Italian patients with metastatic colorectal cancer. Response was superior in the FOLFIXIRI arm versus the FOLFIRI arm (65% vs. 53%; P=0.006), as was both PFS (12.1 vs. 9.7 months; HR, 0.77; P=0.0012) and OS (31.0 vs. 25.8 months; HR, 0.79; P=0.0054). However, rates of resection of metastatic disease did not differ between FOLFOXIRI and FOLFIRI (15% and 12%; P=0.327). Given the failure to improve rates of resection, it remains to be established which patients are most likely to benefit from the more intensive FOLFOXIRI plus bevacizumab.
Editor Disclosures at Time of Publication
Disclosures for David H. Ilson, MD, PhD at time of publication Consultant / Advisory board Clovis; Eli-Lilly; ImClone Speaker's bureau Genentech Grant / research support Bayer Editorial boards HemOnc Today; Journal of Clinical Oncology