Changes in cancer drug benefits over time after insurance approval using ASCO and ESMO scores

There has been increasing development of anticancer drugs and growing demand for early access to these drugs, especially among patients who lack adequate therapies (Booth and Detsky 2019). This has led to a rise in the approval of these drugs, often based on the results of single-arm and phase 2 clinical trials (Kim and Prasad 2015; Prasad et al. 2015). However, basing reimbursement decisions on data from early-phase trials or interim data increases uncertainties (Raymakers et al. 2020). If a drug is granted conditional marketing approval, the drug faces either non-reimbursement or conditional reimbursement decision (Andersen et al. 2019). Even emerging evidence after reimbursement decision adds to the uncertainties, especially due to the escalating costs of oncology drugs (Mariotto et al. 2011; Howard et al. 2015; Saluja et al. 2018). This is because the economic evaluation of an oncology drug is based not only on efficacy, quality of life (QoL), and adverse events but also on the price of the drug. Under the circumstances, it has become increasingly challenging to make reimbursement and pricing decisions for anticancer drugs.

As new evidence on a drug emerges, the clinical value of the drug is likely to change over time. Considering the uncertainties surrounding novel therapies and the continuous emergence of new evidence, some researchers emphasized the need for developing a strategy to reassess the clinical benefits of anticancer drugs (Jenei et al. 2021). Besides, international societies in the pertinent field, such as the European Society for Medical Oncology (ESMO) and the American Society of Clinical Oncology (ASCO), proposed value assessment tools to evaluate the clinical benefits of anticancer drugs, which have already been utilized in several studies (Del Paggio et al. 2017; Saluja et al. 2018; Wong et al. 2020; Ha et al. 2022). The ESMO developed the Magnitude of Clinical Benefit Scale (ESMO-MCBS), which estimates the comparative clinical benefits of drugs based on survival and response outcomes, QoL, and toxicity (Cherny et al. 2015, 2017). Similarly, the ASCO’s value framework (ASCO-VF) is designed to measure the net health benefit (NHB) score, based on comparative clinical efficacy, toxicity, QoL, and palliative effects of drugs (Cherny et al. 2015, 2017).

Many countries including Korea have adopted health technology assessment (HTA) to aid reimbursement decisions. However, the process of re-evaluating the economic value of a drug based on continuously updated evidence can be time-consuming and resource-intensive. If the clinical value of a drug can be readily reassessed by incorporating new data using the ESMO-MCBS or ASCO-VF scores, these tools can serve as valuable screening tools for re-evaluation of the drug following reimbursement decision. Therefore, this study aims to estimate changes in the clinical benefits of anticancer drugs that were listed for reimbursement in Korea, based on the literature available at the time of the initial reimbursement decision (referred to herein as “primary publication”) and the paper published thereafter (“subsequent publication”). In addition, this study explores the potential for the ESMO-MCBS and ASCO-VF as screening tools for re-evaluation.

Approval and reimbursement decisions for anticancer drugs are often based on premature and insufficient evidence on clinical benefits. Over time, evidence gets updated, which changes the value of the drugs. In this current study, 61.6% of the included clinical trials updated the evidence following reimbursement appraisal. Moreover, the updates are made rather quickly. By analyzing publications issued after reimbursement decisions have already been made, we found that 62.5% of the subsequent publications were released within 1 year of reimbursement decisions. The updated evidence is shown to cover various outcomes, ranging from clinical efficacy and QoL to toxicity.

While efficacy is a crucial factor in assessing the value of a drug, this present study revealed that there were only minimal changes in average clinical efficacy from primary publication to subsequent one. The ESMO-MCBS and ASCO-VF NHB scores increased by only 0.04 and 0.43, respectively. Furthermore, the overall change in the value of anticancer drugs was substantially influenced by the bonus scores related to QoL and palliation measures.

The value of a drug demonstrated in subsequent publication varies considerably among drugs. The value increased for some drugs but decreased for others. Not only the direction but the magnitude of the change differed among drugs. Changes in the ESMO-MCBS scores ranged from − 2 to + 2, whereas changes in ASCO-MCBS scores varied more widely from − 15 to + 21.9. The value is likely to change, depending on the uncertainty associated with the initial evidence of the drug and the updated evidence.

There are discrepancies in the scoring system and factors considered between the ESMO-MCBS and the ASCO-VF NHB. For that reason, valuation results may differ for the same agent depending on the value framework used. For example, the ESMO-MCBS score for vemurafenib (therapy for melanoma) decreased from 4 at Time 1 to 2 at Time 2. This was a result of the change in the clinical benefit score according to long-term OS with crossover adjustment. On the other hand, the ASCO-VF NHB score for atezolizumab (therapy for non-small cell lung cancer) decreased from 49.7 at Time 1 to 34.7 at Time 2. This was because the tail of the curve for the bonus score was not assigned according to the long-term OS result. Despite the decrease in scores, when the other value framework was used for respective agent, changes in scores were not observed.

The proportion of clinical trials that exhibited a score change over time differed between the two frameworks used. In our analysis, only one-third of the included trials showed an ESMO-MCBS score change, which was similar to the findings of Thomson et al.’s (2021) study based on the European Medicine Agency’s approval data. In contrast, two-thirds (68%) of the trials showed an ASCO-VF NHB score change, which is similar to the findings of Delos Santos et al.’s (2021) study based on the Food and Drug Administration’s approval data. The difference between the two frameworks may have been attributable to a discrepancy in how variables were operationalized. In other words, the ESMO-MCBS scale is categorized into five levels, whereas the ASCO-VF NHB score for advanced disease is treated as a continuous variable. For that reason, the ESMO-MCBS score may have been less sensitive to changes in clinical benefit.

In our study, more than 50% of anticancer drugs were associated with high levels of uncertainty at the time of reimbursement appraisal. The mean preliminary scores for the drugs decreased after the updated evidence was made available. In some instances, a certain level of uncertainty persisted even in subsequent publications. For example, in three product/indication pairs, the primary publication was based on a non-randomized, phase 1 or 2 study, but new evidence was derived from a separate, randomized phase 3 study. However, trials following reimbursement decisions for all three cases used a crossover design, which made it difficult to measure OS gains with accuracy. In these cases, a type of uncertainty was not resolved but simply transferred into another type.

The process of conducting an economic evaluation is time-consuming and resource-intensive. The HTA of a drug for reimbursement purposes requires essential input parameters such as efficacy, QoL, and adverse events. The ASCO-VF and ESMO-MCBS tools also evaluate the value of a drug by considering multi-attribute outcomes. As decision-making in the face of uncertainty becomes more prevalent, the need for a re-evaluation process after reimbursement has also increased. As new evidence becomes available, evidence-based evaluation requires the incorporation of the updated literature. In these instances, value framework examined in this current study can serve as simple screening tools for re-evaluation. If the tools indicate a substantial decrease in the score for a drug after its reimbursement decision, there may be a need for a formal reassessment to maintain the reimbursement and pricing decisions of the drug. New evidence can either maintain, reinforce, or refute previous decisions. In South Korea, not all drugs need to be re-evaluated and the decision to re-evaluate reimbursement does not depend on the emergence of new evidence. However, for cancer drugs that are subject to the risk-sharing agreement (RSA) and exemption of economic evaluation, it is mandatory to re-evaluate reimbursement every 5 years until the contract ends.

While our study explored the potential of both frameworks as screening tools for re-evaluation after reimbursement, previous studies also investigated the relationship between the clinical benefits derived from the frameworks and reimbursement recommendations in Canada and Spain (Meyers et al. 2021; Nieto-Gómez et al. 2023). We focused on change in scores at the time of reimbursement and afterward, using both the ASCO and ESMO value frameworks. However, Meyers et al. and Nieto-Gómez et al. assessed the clinical benefit of cancer drugs that received a positive and negative recommendation for reimbursement using the ESMO-MCBS only. The previous studies showed significant differences in the ESMO-MCBS score between positive and negative decisions, but modest gains in OS. Anticancer drugs with a positive reimbursement decision had a higher proportion of RCTs and phase 3 trials, which is similar to our study. In the study by Nieto-Gómez et al., the ESMO-MCBS scores of 3 and 4 accounted for 68% of the reimbursed indications, which is comparable to the ESMO-MCBS mean score of 3.29 at the time of reimbursement observed in our study.

Despite the potential advantages of the tools, the ESMO-MCBS and ASCO-VF have several limitations that need to be considered. First, these tools do not measure clinical benefit against the best comparator used in real practice and there may be a discrepancy between the comparator considered in the reimbursement process and that used in clinical trials. Second, the ESMO-MCBS tends to focus on surrogate endpoints even though clinical benefit from surrogate endpoints may not necessarily result in final outcomes. On the other hand, the ASCO-VF tends to reflect the importance of final outcomes by attributing the greatest weight to OS. ASCO-VF cannot assess single-arm trials, which present a high level of uncertainty. For the effective application of the ESMO-MCBS and the ASCO-VF NHB to decision-making, further discussions are required as to how to reconcile discrepancies in the scoring methods and factors considered between the two tools.

Regardless of the discrepancies, the ASCO-VF and ESMO-MCBS have the potential as screening tools for re-evaluation after reimbursement decisions, which was explored in this present study. This is the first study to analyze post-reimbursement changes in value framework scores. Despite the strengths, caution should be taken in interpreting our findings due to the following limitations. First, we relied on the drugs that were listed in Korea. This choice may limit the generalizability of our findings. Nonetheless, it is worth noting that we included all anticancer drugs listed in Korea after the introduction of the positive listing system. Second, differences in the scoring methods and factors between frameworks can lead to different valuations of clinical benefits. Third, ESMO-MCBS: H-scale was published yet when the analysis was carried out in 2021. Future studies would be needed to include single-arm hematologic cancer using ESMO-MCBS:H. This limitation highlights the need to choose a framework carefully, depending on the purpose of valuation. Practical application of the framework should also consider country-specific factors, such as the reimbursement and drug pricing system and whether HTA has been implemented.

Subsequent publications of oncology drugs are often released after reimbursement decisions have been made. These publications provide updated evidence on the drugs. Updated data are frequently published with a year of initial reimbursement decision. Using the ASCO-VF and ESMO-MCBS frameworks, this study revealed that while some drugs demonstrated no change in the value of clinical benefits, others exhibited notable positive or negative changes. The two value frameworks can serve as preliminary screening tools for re-evaluating agents that exhibit high uncertainty or substantial changes in scores.

Below is the link to the electronic supplementary material.

NRY: investigation, visualization, formal analysis, project administration, writing of original draft, writing of review; YJN: investigation, visualization, formal analysis, project administration, writing of original draft, writing of review; JL investigation, visualization, formal analysis, project administration, writing of original draft, writing of review; IS: writing of original draft, writing of review, and editing; E-KL: writing of review and editing, conceptualization, methodology, and supervision; M-HP: writing of original draft, writing of review and editing, conceptualization, methodology, and supervision.

This study did not receive any specific grants from funding agencies in the public, commercial, or non-profit sectors.

Eui-Kyung Lee, Email: ekyung@skku.edu.

Mi-Hai Park, Email: Bestway00@skku.edu.