Prospective Cost-effectiveness Analysis of Cetuximab in Metastatic Colorectal Cancer: Evaluation of National Cancer Institute of Canada Clinical Trials Group CO.17 Trial

Evaluation of National Cancer Institute of Canada Clinical Trials Group CO.17 Trial

Nicole Mittmann; Heather-Jane Au; Dongsheng Tu; Christopher J. O'Callaghan; Pierre K. Isogai; Christos S. Karapetis; John R. Zalcberg; William K. Evans; Malcolm J. Moore; Jehan Siddiqui; Brian Findlay; Bruce Colwell; John Simes; Peter Gibbs; Matthew Links; Niall C. Tebbutt; Derek J. Jonker


J Natl Cancer Inst. 2009;101(17):1182-1192. 

In This Article


The NCIC CTG CO.17 trial is a landmark study that showed a statistically significant survival gain in patients with advanced metastatic colorectal cancer receiving cetuximab. In this economic analysis, cetuximab showed high incremental cost-effectiveness ratios when compared with best supportive care for all patients in the CO.17 trial. The incremental cost-effectiveness ratios were lower for patients with wild-type KRAS tumors, suggesting that providing cetuximab only to patients with wild-type KRAS may avoid the health-care system costs of using an expensive drug that would provide little or no clinical benefit to treat patients whose tumors do not express wild-type KRAS.

To our knowledge, this economic evaluation of the NCIC CTG CO.17 trial is the first large-scale analysis from an oncology cooperative group that has collected resource utilization and utility values prospectively in an international phase III clinical trial.[4] By using actual trial data, we avoided having to make assumptions about resource utilization that are often required in economic analyses that use models.

The mean incremental cost-effectiveness ratio of cetuximab compared with best supportive care for patients with wild-type KRAS tumors was lower and had narrower 95% confidence intervals than that for all CO.17 patients regardless of tumor KRAS mutation status, even though the incremental ratios remained high. The decrease in the incremental cost-effectiveness ratios for the subanalysis of patients with wild-type KRAS tumors was driven primarily by longer survival in the cetuximab group.

The clinical benefit from targeted therapies in oncology is limited to a subgroup of patients, but biomarkers for prospective selection of these patients have previously been lacking. The tumor mutation status of the KRAS oncogene has been investigated as a predictor of clinical outcomes.[7–13] The cost-effectiveness of cetuximab compared with best supportive care for the subset of the population with wild-type KRAS tumor status may actually be even more favorable (ie, the cost-effectiveness ratios may be lower) than that calculated in this analysis because the survival curves for patients with wild-type KRAS tumors and for those with mutant KRAS tumors had not yet converged at the end of the trial,[7] which indicates an even greater survival advantage for the wild-type KRAS tumor populations treated with cetuximab. However, the cost-effectiveness acceptability curves showed that there was a 0% probability of the threshold falling below $50 000 per life-year gained or below $50 000 per QALY gained for the wild-type KRAS tumor population.

With respect to survival, the increased clinical benefit of cetuximab compared with best supportive care alone in patients with wild-type KRAS tumors may encourage the routine clinical testing for KRAS mutations in tumors from patients with metastatic colorectal cancer. It is important to note that we did not include the cost of KRAS mutation testing (~$300–$600 per test) in this trial-based economic analysis because within this study, the cost of testing was applied to both cetuximab and best supportive care patients in the wild-type KRAS subset analysis. Thus, the incremental cost-effectiveness ratios were unlikely to have been affected by including the cost of KRAS mutation testing.

Economic evaluations should identify subgroups for which there is evidence for increased effectiveness of an intervention. A policy of treating the estimated 40% of patients whose tumors harbor mutated KRAS[7,16,17] with best supportive care alone would avoid the health-care costs of using cetuximab, an expensive drug from which this group of patients can expect little or no benefit. A recently published abstract from the US perspective reported a cost savings of $600 million associated with treating only the colorectal cancer patients whose metastatic tumors harbor wild-type KRAS with cetuximab.[31]

It is important for health-care payers to be informed about the financial impact and clinical benefits of a strategy of providing cetuximab only to patients with wild-type KRAS compared with providing best supportive care alone regardless of KRAS mutation status. This economic analysis evaluated the cost-effectiveness of the CO.17 trial population, which includes a subpopulation with wild-type KRAS tumors. This analysis is not an economic analysis of KRAS screening. An assessment of KRAS screening would require a separate analysis. We note that cetuximab is not yet reimbursed by the public payers in either of the two countries in which this clinical trial was conducted.

KRAS mutation analysis showed that treatment with cetuximab increases the survival advantage in a group of patients that can be selected as more likely to benefit from this treatment. Whether the incremental cost-effectiveness ratio continues to remain higher than what most payers are willing to pay speaks to the issue of what is considered an acceptable threshold and is currently being debated.

The health utility values obtained in the trial are similar to those reported in other studies. For example, in a survey of health-related quality of life among a community-based sample of individuals with colorectal carcinoma, Ramsey et al.[32] stratified colorectal patients by stages and time since diagnosis and reported that patients with stage IV colorectal cancer had a mean utility value of 0.84. This value is higher than any of the utility values in either treatment arm of CO.17. It is important to emphasize that this analysis used utilities that were measured with the HUI3 and were obtained directly from patients in the CO.17 trial. Thus, it is possible that the utility values used in this study may not reflect general population preferences for health states.[26] Considerable debate remains as to whose preference valuation, namely society's, patients', patients' families', and/or health-care professionals', should be considered in economic decision analyses.[33,34]

This study has five possible limitations. First, results of this analysis are based on clinical trial results and thus may not be generalizable to routine care. Second, the time horizon for the analysis was trial driven, and the maximum follow-up was between 18 and 19 months (1.58 years; the last point on the survival curve). However, given the advanced stage of the colorectal cancers and the short survival of patients on CO.17, it is unlikely that there would be any statistically significant survival gain beyond the trial duration for the overall population, but survival gain may be more important in the wild-type KRAS subset because these patients demonstrated better clinical outcomes than the rest of the population. Third, it is important to note that the CO.17 trial included patients who had chemorefractory disease, which is also the approved indication for cetuximab.[35] Fourth, the study population was a selected group of patients with prior chemotherapy who still maintained a reasonable performance status (ie, ECOG 0–2). Many heavily pretreated patients with advanced colorectal cancer have a poor performance status (ECOG 3–4), and the results of this study may not be generalizable to very ill patients. Fifth, the end of the trial occurred when the proportion alive was less than 10%. Therefore, modeling beyond the end of the trial may be uninformative, making this analysis essentially a lifetime cost–utility analysis. It is important to note that, as per CO.17 protocol, patients received cetuximab until they experienced disease progression or they no longer derived clinical benefit from therapy. Therefore, theoretically, patients could have received cetuximab in perpetuity. However, because less than 10% of the patients were alive at the end of the study time horizon, the decision to not model over a lifetime time horizon would not influence the economic outcome.

A determination of country-specific cost-effectiveness is important for health-care decision makers. The results of economic analyses based on multinational clinical trials may be subject to jurisdictional or geographical differences in patient population, clinical practice, and costs of resources.[36–38] Possible heterogeneity in between-country data may limit the geographical transferability of pharmacoeconomic data. A number of methods (simple and complex) are available to adjust for the multinational nature of resource and survival data.

The cost of cetuximab was a major cost driver in this economic analysis. Cetuximab costs approximately twice as much in the United States as it does in Europe.[19] Using a US cost ($6.73/mg) resulted in a doubling of the incremental ratios. Survival was also a major cost driver in this economic analysis, and limiting cetuximab therapy to those who had the greatest survival benefit (ie, those with wild-type KRAS tumors) improved the incremental cost-effectiveness ratio. However, not all costs were included in the economic evaluation; specifically, nonmedical and indirect costs were excluded. The incremental cost-effectiveness ratio was not sensitive to any cost variable except the cost of cetuximab.

The incremental cost-effectiveness ratios presented here represent the undiscounted ratios. In Canada, a 5% discount rate is used in economic analyses.[18] A discount rate was not applied to this analysis because the median survival of study subjects was less than 1 year and utility values were only collected over a 28-week period. Thus, we felt that discounting was unlikely to have a large impact because of the short survival of this patient group.

The economic burden of colorectal cancer is substantial.[39,40] The cost of new targeted therapies for this and other cancers has received much attention.[41,42] Norum[43] modeled irinotecan plus cetuximab vs no chemotherapy in patients with metastatic colorectal cancer in Norway and estimated incremental cost-effectiveness ratios ranging from €205 536 to €323 040 per life-year gained (€1 = US $1.59). Starling et al.[44] also analyzed cetuximab plus irinotecan vs active plus best supportive care in the United Kingdom and estimated incremental cost-effectiveness ratios of £42 975 per life-year gained and £57 608 per QALY (£1 = US $2). The incremental benefit for cetuximab in our study was smaller than either of these two studies. These studies, however, did not assess the effect of the tumor KRAS mutation status on the incremental cost-effectiveness.

Economic analyses are conducted to provide decision makers with information on the costs and clinical benefits of specific interventions. A number of jurisdictions, including Canada, the United Kingdom, Scotland, and Australia, require formal economic evaluations for drug reimbursement. It has been suggested that as the incremental cost-effectiveness ratio increases, the probability of reimbursement decreases.[45] Decisions regarding reimbursement are complex and involve many factors including affordability, burden of disease, improvement in survival, availability of alternatives, quality of life, and cost-effectiveness.[46]

Cost-effectiveness analysis has been increasingly used as a measure to inform treatment allocation in a shrinking health-care resource environment. The National Institute for Health and Clinical Excellence guidelines recommend considering factors in addition to cost-effectiveness for interventions that have incremental cost-effectiveness ratios greater than £30 000 per QALY (£1 = US $2).[45,47] In the United States, a benchmark threshold of approximately $50 000 per QALY exists. In the United Kingdom, a threshold of £30 000 per QALY has been applied.[48,49] A survey of health economists found a threshold of $60 000 per QALY to be acceptable.[50] Where thresholds have been suggested, they have not been focused on classes of drugs. However, a 2006 survey of medical oncologists inferred a $300 000-per-QALY threshold for oncology medications.[51] It is important to note that there is no empirical evidence to support a particular threshold value. Most jurisdictions have implicit thresholds,[52] but threshold values may not be formally declared if the setting of a threshold is value driven and may be politically sensitive.[45]

Our incremental cost-effectiveness ratio for cetuximab of $199 742 per life-year gained is considerably higher than the jurisdictional thresholds of which we are aware. However, restricting cetuximab to patients with wild-type KRAS tumors reduced the incremental cost-effectiveness ratio to $120 061, indicating that the use of cetuximab only in this group of patients would result in a more efficient use of health-care resources. Although the wild-type KRAS group had a greater survival gain with cetuximab vs best supportive care (3–4 months = 0.25–0.33 years) compared with the overall group (1.5 months = 0.13 years), the drug cost was also greater in the wild-type KRAS group because cetuximab was used for a longer time. However, the incremental cost-effectiveness ratio was still high for the wild-type KRAS group and would generally be considered unfavorable. To achieve a generally accepted level of cost-effectiveness, we hypothesize that the survival gain would need to be on the order of 6–8 months, even allowing for longer drug use. The cost-effectiveness acceptability curves revealed a 100% probability that cetuximab is not cost-effective compared with best supportive care above a threshold incremental cost-effectiveness ratio of $50 000 per QALY for both the entire population and the wild-type KRAS tumor subset.

Recent trials with cetuximab have shown statistically significant improvement in survival over best supportive care in metastatic colorectal cancer patients.[3–5] Additional trials have reported increased survival benefits in the wild-type KRAS tumor population.[6,7] Consequently, from a health-care system perspective, it would not be efficient to fund cetuximab treatment for all patients with advanced colorectal cancer.[7] Use of cetuximab may be restricted based on a patient's tumor KRAS status.[53]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.