Valuing the benefits of new anticancer drugs

Improvements in survival and cancer-related symptoms must be weighed up against treatment-related adverse effects and financial burden

Annual expenditure on anticancer drugs by the Pharmaceutical Benefits Scheme (PBS) has risen from $65 million in 1999–00 to $466 million in 2011–12 — an average annual increase of 19%.1 This is more than double Australia’s average annual increase in health expenditure of 8% over the same period.2 New techniques in radiation oncology and surgical oncology have also increased expenditure on cancer. It has never been more important to assess the value of new cancer treatments, weighing up the balance between benefits, harms and costs. In oncology, clinical benefit is defined by improvements in survival and cancer-related symptoms, and must be traded off against treatment-related adverse effects and financial burden. In Australia, evaluation of the trade-offs for new anticancer drugs has been predominantly a matter of health care policy, and determined by the Pharmaceutical Benefits Advisory Committee (PBAC).

People affected by cancer also face decisions about paying for unfunded treatments and need to understand the degree of benefit a treatment may offer. This is not straightforward, as illustrated by an American study published in 2012, which reported that 69% of patients with metastatic lung cancer and 81% of patients with metastatic colorectal cancer did not understand that their treatment was “not at all likely to cure” their cancer.3 It is even more difficult to balance modest benefits in survival against toxicity and cost, especially when new treatments are portrayed in the media as major breakthroughs.

The world’s two leading medical oncology organisations, the American Society of Clinical Oncology (ASCO) and the European Society for Medical Oncology (ESMO), have recently formulated the first objective, validated and reproducible frameworks to assess the value of anticancer drugs.4,5 These frameworks aim to improve transparency and provide independent measures of the degree of benefit a treatment offers. Both frameworks use data from individual trials to assess the value of a new treatment in comparison with the standard of care used in the trial. Important similarities between the frameworks include separate scoring systems for treatments given with curative intent versus palliative intent; higher value given to improvements in overall survival than to improvements in surrogate endpoints such as progression-free survival (PFS) or response rate; and upward adjustments for improvements in quality of life or reductions in toxicity. In this article, we summarise and compare the frameworks using two examples, and discuss their relevance to clinical practice and health care policy.

The ASCO has developed two instruments — one for potentially curable disease (adjuvant setting) and another for advanced disease — that yield a net health benefit (NHB) score for a new treatment in comparison with the standard of care, using published trial data. In advanced disease, the maximum NHB score is 130, indicating substantial improvements in overall survival, toxicity, cancer-related symptoms and treatment-free interval. The primary outcome contributes a maximum of 80 points to NHB, with higher scores for bigger improvements and benefits in overall survival rather than PFS or objective tumour response rate. The adjustment for toxicity ranges from − 20 for much more toxic to + 20 for much less toxic. Significant improvements in cancer-related symptoms contribute an additional 10 points. An improvement in the treatment-free interval (from completion of the new treatment to initiation of the next treatment) contributes a maximum of 20 points. The total NHB score is judged against the cost of acquiring the drug and patient copayments. The ASCO framework does not define benchmarks for the NHB indicating high value, or a level that justifies funding, leaving individuals and policy makers to draw their own conclusions.

The ASCO framework is summarised with two worked examples in the Box. An evaluation of erlotinib versus chemotherapy as first-line treatment for metastatic non-small cell lung cancer with a driver mutation of the epidermal growth factor receptor (EGFR-mutated)6 yielded an NHB score of 44/130 using the ASCO framework, indicating moderate benefit. In contrast, an evaluation of bevacizumab used in addition to chemotherapy in unselected patients with metastatic non-small cell lung cancer7 yielded an NHB score of 16/130, indicating modest benefit. The ASCO framework can also be used to compare the prices of two treatment strategies; however, in Australia, determining the additional costs of the new treatments is complex. Erlotinib is listed on the PBS for the indication above and, in 2016, costs patients $38.30 per month ($6.20 per month for concession-card holders), with an approximate price paid by the PBS of $1500/month. Bevacizumab is not listed on the PBS for the indication above, and would cost patients about $8000 per month.

The ESMO instruments categorise the benefits of anticancer drugs compared with current standard of care. The instruments yield an alphabetic grade (A, B or C) for treatments that are potentially curative, and a numerical grade (5, 4, 3, 2 or 1) for treatments unlikely to be curative, both ranked in diminishing order of benefit. The grades are assigned based on the observed benefits in survival or PFS in both relative terms (hazard ratios [HRs]) and absolute terms (numbers of months). This ensures an emphasis on benefits that are clinically important rather than just statistically significant. The ESMO framework evaluates the survival benefit first, and then adjusts the benefit grade for toxicity and quality of life. Financial cost was deliberately excluded from the ESMO framework because of differences between countries in costs and health service delivery. ESMO proposes that each health care jurisdiction make its own independent judgements about value for money and willingness to pay after considering the grade of clinical benefit.

The ESMO framework is summarised using the same two worked examples in the Box. Erlotinib compared with standard chemotherapy as first-line treatment for metastatic EGFR-mutated non-small cell lung cancer6 was assigned a grade 4 benefit using the ESMO framework, the highest possible grade for a non-curative treatment with PFS as the primary endpoint. Erlotinib received a high clinical value rating as it substantially reduced the rate of progression (HR, 0.37), improved median PFS by 4.5 months, and had less toxicity than standard-of-care chemotherapy. In contrast, bevacizumab in combination with chemotherapy for unselected metastatic non-small cell lung cancer7 was assigned a grade of 2, because the survival benefits were judged modest (HR for death, 0.79, with an absolute improvement in median survival of 2 months), and were not associated with improvements in toxicity or quality of life.

The ASCO and ESMO frameworks led to similar conclusions in the two examples, yielding higher benefit ratings for erlotinib in EGFR-mutated non-small cell lung cancer than for bevacizumab in unselected non-small cell lung cancer. The conclusions are consistent with the PBAC decision to fund erlotinib, but not bevacizumab, for the indications specified.

The simultaneous development of these frameworks highlights the global recognition that health care costs are increasing rapidly while resources are limited, and the need to establish rigorous assessments of the value of new treatments. This is important regardless of whether it is a jurisdiction or an individual deciding whether to pay for a new expensive anticancer drug.

In Australia, the Therapeutic Goods Administration (TGA) approves the marketing of new drugs based on safety and efficacy, but it is the PBAC that makes recommendations to the Health Minister, based on clinical and economic considerations, about which drugs should be PBS listed and hence funded by the government. In the United States, the Food and Drug Administration is the sole agency responsible for the approval of new drugs for marketing, but it is not permitted to consider cost or cost-effectiveness in its deliberations. The ASCO and ESMO frameworks allow policy makers and health care funders from different health care systems to consider the value of a new treatment together with local considerations such as costs, disease prevalence and competing priorities. These frameworks provide an independent, objective and structured approach that could increase the transparency of decisions made by the PBAC and similar agencies.

Clinicians should find the frameworks useful when assessing and explaining the value of new anticancer drugs, because they provide standardised metrics for expressing benefit. This should allow clearer communication with patients, particularly in situations where new treatments are unfunded, expensive and have modest benefits. The ESMO framework is probably easier to use and expresses value in simple grades. The ASCO framework is more complex, and expresses value with a numerical score that requires more explanation and interpretation. The ASCO plans further development of its framework to facilitate doctor–patient communication and shared decision making.4 An important enhancement would be the capacity for individuals and groups to assign weights to the factors that are most important to them. For example, some prioritise survival benefits regardless of cost or toxicity, whereas others prioritise quality of life over survival time. The incorporation of such weights would help communities, patients and doctors make decisions that better reflect their preferences.

The costs of health care are burgeoning and this approach warrants consideration beyond oncology. All treatments require rigorous assessments of efficacy, safety and cost-effectiveness. These frameworks provide important first steps towards standard definitions, metrics and criteria for assessing the value of new anticancer drugs.