Sequential or Combination Therapy in Untreated Metastatic CRC?

The XELAVIRI trial failed to demonstrate the noninferiority of a sequential escalation treatment strategy vs initial combination therapy in mCRC.

The XELAVIRI trial failed to demonstrate the noninferiority of a sequential escalation treatment strategy with fluoropyrimidines, irinotecan, and bevacizumab compared with initial combination therapy in patients with untreated metastatic colorectal cancer.

However, researchers led by Dominik Paul Modest, MD, of Comprehensive Cancer Center, University of Munich, Germany, wrote that although noninferiority was not demonstrated, it could also not be ruled out.

“A critical aspect of the current study consists of the observation that sequential escalation of treatment – that is, the addition of irinotecan in arm A was only performed in a minority (37.7%) of patients,” Modest and colleagues wrote in The Journal of Clinical Oncology. “Given the median treatment duration of 7.4 months observed in the sequential treatment arm, many patients seemingly discontinued treatment without progression and therefore did not rely on the salvage escalation within the foreseen time interval of first-line therapy.”

The XELAVIRI trial tested the noninferiority of initial treatment with fluoropyrimidines plus bevacizumab followed by irinotecan at first progression (arm A) with fluoropyrimidines plus irinotecan plus bevacizumab (arm B) in 421 patients with untreated metastatic colorectal cancer. Given a 90% confidence interval (CI), a power of 70% and a one-sided α of .05, the margin for noninferiority was set at 0.8.


The median duration of treatment was 7.4 months in arm A and 6.7 months in arm B. Only about one-third of patients (37.7%) in the sequential treatment arm had treatment escalation. Reasons for no escalation included no progression on treatment (68.3%), physician decision (8.3%), patient wish (9.1%), adverse events (4.5%), death (0.8%), protocol violation (4.5%), and unknown reason (3.8%). About two-third of patients (63.2%) received irinotecan at some point during the course of treatment in arm A compared with 100% of patients in arm B.

Noninferiority was not shown between the two arms (hazard ratio [HR], 0.86; 90% CI, 0.73–1.02).

The researchers wrote that these efficacy data should be interpreted “in the context of competing treatment concepts of metastatic colorectal cancer.”

“In the pre-antibody era, studies demonstrated that sequential therapy is not worse than initial combination treatment,” the researchers wrote. “In contrast, in the antibody era, an association of intensification of cytotoxic therapy in first-line treatment with response parameters that, in turn, correlate with prolonged survival was observed in multiple studies predominantly evaluating RAS wild-type, and to a lesser extent, RAS mutant metastatic colorectal cancer.”

The XELAVIRI study did show that those patients with RAS/BRAF wild-type benefitted from combination chemotherapy (HR, 0.61; 90% CI, 0.46–0.82; P = .005) but those with RAS-mutant tumors did not (HR, 1.09; 90% CI, 0.81–1.46; P = .58).

“Our study suggests that sequential escalation of treatment may have a negative effect on outcome in patients with RAS/BRAF wild-type tumors,” the researchers wrote. “In contrast, in this molecularly favorable subgroup, more intensive initial therapy seems to be associated with a clinically relevant improvement in outcome.”

Commenting on the results, Michel Ducreux, MD, PhD, of Gustave Roussy, France, said it was an "interesting study showing that sequential treatment should be used only in wild-type RAS tumors."