Compared with what?

EMA approved tafasitamab with lenalidomide for a type of lymphoma in 2021. The main evidence came from 81 patients in a single arm study. There was no placebo and no active comparator. That evidence was enough for conditional approval because EMA judged that the treatment’s benefits outweighed its risks, and that earlier access for patients with limited options outweighed the uncertainty while further evidence was collected.[1]

Then NICE asked how the treatment compared with what the NHS already provided. There was no direct head-to-head comparison in a trial. The indirect evidence was uncertain, and the comparator survival estimates did not reflect how patients were doing in practice.[2]

NICE did not recommend the treatment for routine NHS use. It also rejected access through the Cancer Drugs Fund because collecting more data would not generate the missing comparative evidence. The cost-effectiveness estimates were also too high and uncertain, but the comparator problem could no longer be fixed by simply following the treated patients for longer.[2]

The evidence had answered the regulator’s question. It had not answered the payer’s.

We have written before that approval and access are different decisions. The comparator is often where those two paths first split.

Regulators and payers need different comparisons

A regulator asks whether a medicine works and whether its benefits outweigh its risks. A payer asks whether it offers enough additional value to fund instead of current care. A clinician asks whether they should prescribe it instead of the treatment they would otherwise choose.

Those questions can, and often do, require different comparators.

FDA’s June 2026 draft guidance recognises placebo, active treatment, no treatment, dose comparisons and external controls. An active comparator can be valuable, but a direct comparison with every treatment used in practice is not required for approval.[3]

EMA regards a randomised controlled trial as the standard. In exceptional circumstances, it can still accept a well-justified single arm study when the treatment effect can be isolated with sufficient certainty.[4]

The distinction is simple. A comparison can establish that a medicine works without establishing whether it is better than the treatment a payer funds or a clinician prescribes today.

Regulators and HTA bodies split most on comparators

This is not a theoretical difference between regulatory and market access teams. Tafuri and colleagues examined 31 parallel advice procedures involving regulators and health technology assessment (HTA) bodies. Full agreement was 77% for the patient population and 60% for endpoints. For the comparator, it was only 44%. Comparator questions also produced the highest disagreement, at 30%.[5]

A follow-up study looked at what companies then put into their development programmes. Seventeen of 21 studies used a comparator different from the one initially proposed. Comparator advice was also harder to implement than endpoint advice.[6]

The same pattern appears in completed assessments. Vreman and colleagues found that HTA bodies raised uncertainty about effects against relevant comparators for almost every indication they examined. Regulators raised it much less often.[7]

Different bodies were doing different jobs. The problem was leaving one of those jobs without the comparison it needed.

A randomised trial can still use the “wrong” comparator

Randomisation does not guarantee that the control group answers the payer’s question.

Padeliporfin was authorised in Europe after a randomised trial against active surveillance for low-risk prostate cancer.[8] But the company proposed it for a place in the NHS pathway where NICE considered radical therapies, such as surgery and radiotherapy, to be the relevant comparators.

The company provided no clinical-effectiveness analysis against those treatments. It also reported that a network meta-analysis was not possible because the available trials measured different outcomes. NICE could not establish whether padeliporfin offered a clinical benefit against the relevant alternatives and did not recommend it.[9]

There were other concerns, including limited unmet need and uncertainty about where the treatment would fit. The comparator was not the whole decision. It was still a gap the randomised trial could not fill.

Different medicines show different versions of the same problem.

Amivantamab received accelerated FDA approval from a non-randomised study.[10] NICE accepted an adjusted comparison with NHS care for decision making, but considered the size of benefit uncertain because of residual confounding and limitations in the real-world comparator evidence. An inappropriate EGFR inhibitor component was removed from the company’s updated comparator mix. NICE did not recommend routine use because its most plausible cost-effectiveness estimate was above the acceptable range. It also concluded that further data collection through the Cancer Drugs Fund would not resolve enough of the uncertainty.[11]

Elranatamab also received accelerated approval from a single arm study.[12] NICE needed comparisons with several treatments used at different points in the NHS pathway. The resulting unanchored comparisons, with no shared trial arm linking the treatments, were uncertain. NICE recommended managed access rather than routine use because longer follow-up and NHS data could address other uncertainties, including long-term survival and immunoglobulin use.[13]

Approval answered whether these medicines could work. The payer still needed to know what they should replace.

Standard care differs by market and changes over time

There is no universal standard of care. The relevant comparator may change by country, treatment line, biomarker, previous therapy or contraindication. A treatment used routinely in one market may be unavailable, unfunded or rarely prescribed in another.[7,14]

It can also move during development. A pivotal trial may take years. New competitors launch, clinical guidance changes and old treatments fall out of use.

Elranatamab illustrates both problems. NICE identified different comparators for different points in the multiple myeloma pathway. It also expected the use of one comparator to fall as newer medicines entered practice.[13]

Mapping one lead market once is not enough.

Ask payers and clinicians in every priority market: What would this treatment replace for this patient group today? Then keep asking as the programme develops.

Strong programmes compare directly where it matters most

The answer is not a trial arm for every treatment in every country. The strongest programmes choose one high-value direct comparison, then plan credible routes to the rest.

PARADIGM-HF compared sacubitril valsartan directly with enalapril, an established effective treatment for heart failure. NICE considered the trial high quality and its mortality and hospital-admission outcomes relevant. It accepted an ACE inhibitor class effect and used a network meta-analysis to support comparisons with angiotensin receptor blockers.[15,16]

ARISTOTLE compared apixaban directly with warfarin in 18,201 people with atrial fibrillation. Warfarin was the established treatment and the main direct comparison. A connected network allowed indirect comparisons with other anticoagulants, although NICE considered those results too uncertain to distinguish reliably between them. NICE recommended apixaban.[17,18]

FLAURA compared osimertinib with two active treatments, gefitinib and erlotinib, rather than placebo alone. NICE considered the trial broadly relevant to NHS practice and recommended osimertinib. Afatinib, then the most prescribed treatment in England, was not included. NICE examined other clinical evidence and an exploratory indirect comparison, rejected an assumption that all three comparators were equally effective, and carried the remaining uncertainty into its decision.[19,20]

None of these programmes answered every comparison directly. They answered an important one directly and used indirect evidence to address some of the rest, with uncertainty made explicit.

Choose the comparator with the greatest decision value

The pivotal comparator should do more than satisfy one meeting with one decision-maker. Choose the comparator that answers the most important questions across regulators, payers and clinicians, while remaining ethical, affordable and feasible within the development programme.

That requires an explicit trade-off:

Will regulators accept the control strategy?
Does it represent current care in priority markets?
Is it the treatment clinicians would otherwise prescribe?
Can the trial recruit and retain enough patients?
Will the comparison remain useful if the pathway changes?
Can the remaining comparators be reached through the existing evidence?

This is the logic of an Integrated Evidence Plan. Map the comparisons each decision-maker will need. Put the highest-value comparison into the trial, then identify which of the others can be built credibly from existing evidence.

The last question is easy to promise and expensive to get wrong.

Check the indirect comparison is possible first

When no head-to-head trial exists, comparator evidence usually comes from one or both of two places.

Patient-level data may be available from real-world sources, such as disease registries, electronic health records or insurance claims. If patient-level data cannot be accessed, published studies may provide aggregate data: averages and results for groups rather than individual patients.[21,26,27]

Run a full feasibility assessment before committing to either route. Start with the estimand, meaning the exact clinical question the comparison needs to answer. Then check whether the external data exist and align with the trial on time zero, eligibility criteria, baseline characteristics, outcome definitions, follow-up and what happens after treatment starts.[26,27]

Ideally, find that out before the pivotal protocol is fixed.

A manual evidence synthesis or systematic review can take months. Across 195 completed systematic reviews, the mean time from registration to publication was 67.3 weeks.[22]

Evidax TRACE uses an AI-augmented workflow to cut the evidence synthesis timeline from months to days. Experts remain in control and every decision is auditable. Both rigour and speed matter here.

Ask regulators, payers and clinicians before the protocol is fixed

Comparator choice should not pass from regulatory to market access after the protocol is locked.

Ask regulators which control strategy can establish effectiveness and support benefit-risk assessment. Ask payers what the medicine would replace in each intended population. Ask clinicians what they prescribe today and what evidence would make them switch.

Then put the answers in one evidence plan.

FDA recommends discussing the evidence strategy by the end of phase 2 at the latest.[3] The European parallel advice studies show why. Early discussion led companies to change proposed comparators and, in some cases, consider a third arm or redesign the trial.[5,6]

The right discussion does not just document disagreement. It changes the evidence while change is still possible.

Oncology, advanced therapy and orphan developers should prepare for JSC

The EU Joint Scientific Consultation, or JSC, gives developers advice during study planning on the evidence needed for a later Joint Clinical Assessment.[23]

Cancer medicines containing a new active substance and advanced therapy medicinal products (ATMPs) are already within the first phase of EU JCA. Other orphan medicinal products enter scope from 13 January 2028.[24]

For oncology and ATMP developers, prepare a JSC request for the next available window. For orphan developers whose pivotal programme can still change, prepare now too.

This is not a meeting that can simply be booked. Requests are accepted during set periods and capacity is limited. The EU planned only 8 to 12 medicinal product consultations during 2026.[25]

The pivotal comparator is one of, if not the, most important questions to take into that room.

Take home

Approval evidence may show that a medicine works without showing whether it should replace current care.
Ask payers and clinicians early: What would this treatment replace?
Map standard care across every intended market and refresh it as treatment pathways change.
Choose the direct comparator with the greatest decision value that the development programme can support.
Check that the literature and data can support an indirect comparison before relying on one.
Seek regulatory, payer and clinical advice before the pivotal protocol is fixed.
For oncology medicines, ATMPs and orphan medicines approaching JCA, prepare for Joint Scientific Consultation now.

References

European Medicines Agency. Minjuvi: EPAR. Marketing authorisation issued 26 August 2021. https://www.ema.europa.eu/en/medicines/human/EPAR/minjuvi
National Institute for Health and Care Excellence. Tafasitamab with lenalidomide for treating relapsed or refractory diffuse large B-cell lymphoma. Technology appraisal guidance TA883. Published 3 May 2023. https://www.nice.org.uk/guidance/ta883
US Food and Drug Administration. Demonstrating Substantial Evidence of Effectiveness for Human Drug and Biological Products. Revised draft guidance for industry. June 2026. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/demonstrating-substantial-evidence-effectiveness-human-drug-and-biological-products
European Medicines Agency. Reflection paper on establishing efficacy based on single-arm trials submitted as pivotal evidence in a marketing authorisation application. Adopted September 2024. https://www.ema.europa.eu/en/establishing-efficacy-based-single-arm-trials-submitted-pivotal-evidence-marketing-authorisation
Tafuri G, Pagnini M, Moseley J, et al. How aligned are the perspectives of EU regulators and HTA bodies? A comparative analysis of regulatory-HTA parallel scientific advice. British Journal of Clinical Pharmacology. 2016;82(4):965-973. https://doi.org/10.1111/bcp.13023
Tafuri G, Lucas I, Estevão S, et al. The impact of parallel regulatory-health technology assessment scientific advice on clinical development: assessing the uptake of regulatory and health technology assessment recommendations. British Journal of Clinical Pharmacology. 2018;84(5):1013-1019. https://doi.org/10.1111/bcp.13524
Vreman RA, Naci H, Goettsch WG, et al. Decision making under uncertainty: comparing regulatory and health technology assessment reviews of medicines in the United States and Europe. Clinical Pharmacology & Therapeutics. 2020;108(2):350-357. https://doi.org/10.1002/cpt.1835
European Medicines Agency. Tookad: EPAR. Marketing authorisation issued 10 November 2017. https://www.ema.europa.eu/en/medicines/human/EPAR/tookad
National Institute for Health and Care Excellence. Padeliporfin for untreated localised prostate cancer. Technology appraisal guidance TA546. Published 21 November 2018. https://www.nice.org.uk/guidance/ta546
US Food and Drug Administration. FDA grants accelerated approval to amivantamab-vmjw for metastatic non-small cell lung cancer. 21 May 2021. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-grants-accelerated-approval-amivantamab-vmjw-metastatic-non-small-cell-lung-cancer
National Institute for Health and Care Excellence. Amivantamab for treating EGFR exon 20 insertion mutation-positive advanced non-small-cell lung cancer after platinum-based chemotherapy. Technology appraisal guidance TA850. Published 14 December 2022. https://www.nice.org.uk/guidance/ta850
US Food and Drug Administration. FDA grants accelerated approval to elranatamab-bcmm for multiple myeloma. 14 August 2023. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-grants-accelerated-approval-elranatamab-bcmm-multiple-myeloma
National Institute for Health and Care Excellence. Elranatamab for treating relapsed and refractory multiple myeloma after 3 or more treatments. Technology appraisal guidance TA1023. Published 11 December 2024. https://www.nice.org.uk/guidance/ta1023
Wolters S, Jansman FGA, Postma MJ. Differences in evidentiary requirements between European Medicines Agency and European health technology assessment of oncology drugs: can alignment be enhanced? Value in Health. 2022;25(12):1958-1966. https://doi.org/10.1016/j.jval.2022.05.006
McMurray JJV, Packer M, Desai AS, et al. Angiotensin-neprilysin inhibition versus enalapril in heart failure. New England Journal of Medicine. 2014;371(11):993-1004. https://doi.org/10.1056/NEJMoa1409077
National Institute for Health and Care Excellence. Sacubitril valsartan for treating symptomatic chronic heart failure with reduced ejection fraction. Technology appraisal guidance TA388. Published 27 April 2016. https://www.nice.org.uk/guidance/ta388
Granger CB, Alexander JH, McMurray JJV, et al. Apixaban versus warfarin in patients with atrial fibrillation. New England Journal of Medicine. 2011;365(11):981-992. https://doi.org/10.1056/NEJMoa1107039
National Institute for Health and Care Excellence. Apixaban for preventing stroke and systemic embolism in people with non-valvular atrial fibrillation. Technology appraisal guidance TA275. Published 27 February 2013. https://www.nice.org.uk/guidance/ta275
Soria JC, Ohe Y, Vansteenkiste J, et al. Osimertinib in untreated EGFR-mutated advanced non-small-cell lung cancer. New England Journal of Medicine. 2018;378(2):113-125. https://doi.org/10.1056/NEJMoa1713137
National Institute for Health and Care Excellence. Osimertinib for untreated EGFR mutation-positive non-small-cell lung cancer. Technology appraisal guidance TA654. Published 14 October 2020. https://www.nice.org.uk/guidance/ta654
Patel D, Grimson F, Mihaylova E, et al. Use of external comparators for health technology assessment submissions based on single-arm trials. Value in Health. 2021;24(8):1118-1125. https://doi.org/10.1016/j.jval.2021.01.015
Borah R, Brown AW, Capers PL, Kaiser KA. Analysis of the time and workers needed to conduct systematic reviews of medical interventions using data from the PROSPERO registry. BMJ Open. 2017;7(2):e012545. https://doi.org/10.1136/bmjopen-2016-012545
European Commission. Joint Scientific Consultations. https://health.ec.europa.eu/health-technology-assessment/implementation-regulation-health-technology-assessment/joint-scientific-consultations_en
European Parliament and Council of the European Union. Regulation (EU) 2021/2282 of 15 December 2021 on health technology assessment. Official Journal of the European Union. 2021;L 458:1-32. https://eur-lex.europa.eu/eli/reg/2021/2282/oj/eng
Member State Coordination Group on Health Technology Assessment. Annual Work Programme 2026. https://health.ec.europa.eu/document/download/81db097f-a9cb-4404-837c-115b4289321d_en?filename=hta_hatcg_2026-awp_en.pdf
US Food and Drug Administration. Considerations for the Design and Conduct of Externally Controlled Trials for Drug and Biological Products. Draft guidance for industry. February 2023. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/considerations-design-and-conduct-externally-controlled-trials-drug-and-biological-products
National Institute for Health and Care Excellence. NICE real-world evidence framework. Corporate document ECD9. June 2022. https://www.nice.org.uk/corporate/ecd9