Efficacy and Toxicity Associated with Cancer Drug Approvals in Switzerland: International Implications

Cancer is one of the leading health challenges of the 21st century. In response, regulatory agencies have increasingly implemented special pathways to expedite approval of new cancer therapies, often relying on surrogate endpoints to support authorisation.¹ Although these strategies may shorten time to approval, they can also increase uncertainty regarding clinical benefit, particularly when overall survival (OS) and quality-of-life data are limited at approval.²

Switzerland is among the European countries with the most reliable access to novel cancer therapies. Between 2020 and 2023, Swiss patients had access to 89% of oncology drugs approved by the European Medicines Agency, compared with a European average of 50%.³

But what is the clinical importance of these drugs? In this issue, Rahimzadeh and colleagues⁴ present a meta-analysis evaluating the evidence supporting oncology drugs approved by Swissmedic between 2001 and 2020, including both new active substances and indication extensions. Among 378 pivotal trials supporting cancer drug approvals, 241 were randomised and nearly 70% relied on surrogate measures, mainly progression-free survival (PFS), as the primary endpoint. Pooled hazard ratios for OS were 0.76 (95% CI 0.74–0.77) and 0.55 for PFS (95% CI 0.52–0.58). However, absolute clinical benefit was modest, with median gains of 2.4 months in OS and 3.5 months in PFS, highlighting the difference between favourable relative effects and modest absolute survival gains. Importantly, hazard ratios for OS could be estimated in 194 of 241 randomised trials (80%), whereas median OS gains were available in only 93 trials (39%). This discrepancy likely reflects that follow-up data were often immature at the time of analysis. These findings are particularly relevant because uncertainty at the time of approval can have important clinical and economic consequences, particularly in health systems where pricing and reimbursement are intended to be linked to clinical value.

Rahimzadeh et al.’s findings appear consistent with results reported internationally (Table 1). Across jurisdictions, many oncology drugs provide limited clinical benefit despite high prices. This persistent pattern raises a central question in modern oncology: how can we encourage development of and access to cancer drugs showing larger gains in meaningful patient outcomes?

Table 1.

Overall survival and progression-free survival supporting approvals of new cancer drugs across different jurisdictions.

Publication	Agency (Country)	Type of study	Period	Drugs n	Indications n	RCT%	OS HR (95% CI)	ΔOS m	PFS HR (95% CI)	ΔPFS m	QoL data	Safety data
Michaeli et al. J Clin Oncol. (2022)	FDA (US)	Meta-analysis	2003–2021	124	374	66	0.73 (0.72–0.75)	2.8	0.57 (0.54–0.60)	3.3	Not evaluated	Not evaluated
Ladanie et al. JAMA Netw Open. (2020)	FDA (US)	Meta-analysis	2000–2016	92	100	51	0.77 (0.73–0.81)	2.4	0.52 (0.47–0.57)	2.7	Not evaluated	Not evaluated
Fojo et al. JAMA Otolaryngol Head Neck Sur (2014)	FDA (US)	Narrative review	2002–2014	71	71	100	No pooled	2.1	No pooled	2.5	Not reported	Not reported
Davis et al. The BMJ (2017)	EMA (Europe)	Retrospective cohort	2009–2013	48	68	88	No pooled	2.7	No pooled	Not reported	Reported	Not reported
Mao et al. The Lancet Regional Health – Western (2025)	NMPA (China)	Retrospective cohort	2012–2021	56	92	81	0.68 (0.65–0.71)	Not reported	0.46 (0.42–0.52)	Not reported	Not reported	Reported
Zhang et al. JAMA Network Open (2022)	NMPA (China)	Systematic review Cross-sectional	2005–2020	78	141	82	No pooled	4.1	No pooled	Not reported	Not reported	Not reported

CI, Confidence Interval; Δ, delta; EMA, European Medicines Agency; FDA, Food and Drug Administration; HR: hazard ratio; IQR: interquartile range; m, months; NMPA, National Medical Products Administration; OS, overall survival; PFS, progression-free survival; RCT: randomised controlled trial.

Despite major advances in oncology, including genome-targeted therapies and immune checkpoint inhibitors, improvements in clinically meaningful outcomes have remained relatively modest. This situation reflects both therapeutic and regulatory limitations. In some tumour settings, clinical benefit often remains limited to small subsets of patients, with substantial heterogeneity across tumour types and drug classes.⁵ For example, as the authors report, in ovarian cancer, substantial improvements in PFS (HR 0.55) translated into only modest OS gains (HR 0.87), whereas immune checkpoint inhibitors showed more consistent improvements across both endpoints (OS HR 0.71; PFS HR 0.72).⁴

From a regulatory perspective, many approvals continue to rely on surrogate endpoints, immature survival data, and single-arm trials, contributing to an incomplete assessment of long-term clinical benefit.⁶ International regulatory collaborations may further reinforce the use of similar evidentiary standards across jurisdictions.⁷ Programmes such as Project Orbis promote coordinated international reviews and faster access to novel therapies.⁸

These uncertainties are particularly important because many oncology therapies are associated with substantial toxicity, meaning patients may be exposed to considerable harm for the prospect of modest clinical benefits. Few previous studies from other jurisdictions evaluated safety and efficacy together.² Their findings are concerning: approximately one in four patients experienced a serious adverse event, including life-threatening events, hospitalisation, or significant disability.⁴ Because patients enrolled in clinical trials are more selected and healthier than those treated in routine practice, toxicity rates may be even higher in real-world settings.⁹

The impact of Rahimzadeh et al.’s work also extends beyond Switzerland and other high-income countries, as many low- and middle-income countries rely on decisions from major regulatory agencies because of limited regulatory capacity.¹⁰

How can uncertainty regarding the clinical benefit of new cancer drugs be reduced? More robust evidence is needed at the time of approval when possible, but if it is not reported then, it must be collected in a timely manner during the post-approval period. Mature OS and quality-of-life data should subsequently inform regulatory and reimbursement decisions. In settings where OS is not feasible or requires prolonged follow-up, frameworks such as the European Society for Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS) or the Swiss OLUtool may help contextualise trial results by integrating efficacy, toxicity, and quality of life into broader assessments of clinical benefit.¹¹ Although conducted in Switzerland, the analyses by Rahimzadeh and colleagues likely reflect broader international trends and reinforce the need for more rigorous evidence standards before and after approval.

Contributors

AT conceptualised the manuscript, reviewed the literature, and wrote the original draft. All authors reviewed, edited, and approved the final manuscript.

Declaration of interests

Dr. Tibau reports receiving personal consulting fees from Alosa Health and a grant from Alfonso Martin Escudero Foundation. Dr Kesselheim reports serving as a consultant for Alosa Health and as an expert witness on behalf of a class of individual plaintiffs in a case against Gilead relating to its tenofovir-containing products, on behalf of a payor in a case against Teva regarding marketing of Copaxone, and on behalf of a group of state attorneys general and payors in a case against multiple generic manufacturers related to alleged price fixing. Dr Hwang reports serving as a member of the European Society for Medical Oncology (ESMO) Magnitude of Clinical Benefit Scale Working Group.