PMID 26132941 reports two randomized, double-blind, placebo-controlled trials of a defined afamelanotide implant in adults with erythropoietic protoporphyria, or EPP. The paper's U.S. trial, NCT01605136, reported median pain-free direct-sunlight exposure of 69.4 versus 40.8 hours over six months, P=0.04. The European trial, NCT00979745, reported 6.0 versus 0.8 hours over nine months, P=0.005. FDA's broader U.S. analysis produced different medians and P=0.055; another missing-diary sensitivity analysis gave P=0.091. The findings contributed to approval of the finished Scenesse implant for a specific adult EPP indication, not approval or validation of “Melanotan I” powder or another material.[1–8]
Study at a glance
| Field | Trial record |
|---|---|
| Research question | Did the sponsor's afamelanotide implant increase pain-free direct-light exposure compared with a vehicle implant in adults with confirmed EPP? |
| Design | Two multicenter, randomized, parallel-group, double-blind, placebo-controlled trials |
| Population | Paper: 74 European participants and 94 U.S. participants; one U.S. placebo participant was withdrawn before exposure |
| Sites | Eight European and seven U.S. porphyria centers |
| Test article | Defined bioresorbable implant containing 16 mg of afamelanotide |
| Comparator | Identical-appearing implant containing the polymer vehicle without afamelanotide |
| Follow-up | Nine-month European and six-month U.S. efficacy periods; the U.S. final safety visit was at day 360 |
| Main endpoint | Diary-recorded hours of sunlight exposure without pain under trial-specific daytime and exposure definitions |
| Registration | U.S. CUV039: NCT01605136; European CUV029: NCT00979745 |
| Publication | PMID 26132941; PMCID PMC4780255; DOI 10.1056/NEJMoa1411481 |
| Sponsor | Clinuvel Pharmaceuticals; the paper also reports NIH funding |
| Current product context | Scenesse, NDA 210797, approved October 8, 2019; current DailyMed update May 11, 2026, with prescribing information revised August 2024 |
The finished form and strength identify the study article; they are not preparation, implantation, or treatment instructions.
Two trials with different endpoints and histories
The European study preceded the U.S. study. Europe used a nine-month period, five planned implants, and a 10 a.m.–3 p.m. window. It counted outdoor time on pain-free days classified as having been spent mostly in direct sun, excluding mixed sun-and-shade days. The U.S. trial used three planned implants over six months, recruited around brighter months, and separately recorded direct sunlight and shade between 10 a.m. and 6 p.m. These are not interchangeable exposure measurements.[1][5][6]
FDA also documents an important endpoint history. The European protocol originally emphasized maximum and total severity of phototoxic reactions. After a different U.S. study, CUV030, failed its original endpoints, the sponsor developed a pain-free sunlight endpoint from that study's data. The European CUV029 analysis plan was changed while CUV029 remained blinded, before its unblinding; the later CUV039 protocol used the new endpoint prospectively. That is different from saying the same primary endpoint was fixed from the beginning in both trials.[6]
Participants were adults with biochemically confirmed EPP. The paper excluded clinically significant hepatic or other organ dysfunction, skin cancer or premalignant lesions, other photodermatoses, and pregnancy. The populations were predominantly White and do not represent all people with porphyria or significant liver involvement. Computer-generated site-specific allocation and matching implants supported masking, but the paper acknowledges that pigmentation partly revealed assignment.[1]
Primary results and what the numbers mean
| Analysis | Active median | Placebo median | Comparison |
|---|---|---|---|
| Paper / U.S. diary analysis, 46 active and 43 placebo | 69.4 hours | 40.8 hours | P=0.04 in paper; FDA reports applicant P=0.044 |
| FDA U.S. analysis, 48 active and 45 placebo | 64.1 hours | 40.5 hours | P=0.055 |
| Paper / European analysis, 38 active and 36 placebo | 6.0 hours | 0.8 hours | P=0.005 |
| FDA / current label European summary, same 38 and 36 | 6.0 hours | 0.75 hours | FDA P=0.005 |
The current label's U.S. medians are 64.1 and 40.5 hours, not the paper's 69.4 and 40.8. The European placebo median is displayed with greater precision in the FDA/registry sources. Keep each source's population and precision visible.[1][3–6]
The raw differences between the paper's medians are 28.6 U.S. hours and 5.2 European hours. They are descriptive subtraction, not a separately established causal effect or a basis for comparing the trials. FDA notes that the highly skewed distributions have different shapes, so the rank-based test should not simply be described as a test of median differences.[6]
FDA reports a U.S. Hodges–Lehmann estimate of 24.0 hours, 95% CI 0.3 to 50.3, in the 89-person diary analysis. This is an estimate based on pairwise between-group differences, not subtraction of the displayed medians. FDA says it was added in the study report rather than specified in the protocol or analysis plan. The confidence-interval and P-value guide explains why the estimate, interval, test and analysis population all matter.[6]
FDA's missing-data analyses
The U.S. study randomized 94 people: 48 active and 46 placebo. One placebo participant with a pre-existing eye finding was withdrawn before receiving or being dispensed an implant. FDA excluded that person from efficacy and safety analyses. Its tables labeled “all randomized” therefore use the 93-person randomized-and-dispensed population, not all 94 initially randomized people.[6]
The applicant excluded four more people, two in each arm, who returned no diaries, leaving 89 in the paper's primary diary analysis. FDA's first sensitivity analysis included those four and assigned them zero pain-free sunlight hours, while retaining reported observations for people with at least one diary. It yielded P=0.055 and a Hodges–Lehmann estimate of 23.5 hours, 95% CI 0.0 to 48.0.[6]
A second FDA analysis also accounted for some people missing one or two whole diaries. For those who completed treatment, it scaled observed exposure to the missing diary periods; for relevant treatment discontinuations, it assumed no additional pain-free sunlight exposure after stopping. This gave P=0.091 and an estimate of 22.8 hours, 95% CI −1.8 to 47.0. These analyses addressed missing whole diaries, not every missing individual day.[6]
The results retain a similar direction, but their classification around P=0.05 changes with population and assumptions. FDA nonetheless concluded that the total evidence supported the specific finished-product indication, with CUV039 pivotal and CUV029 supportive. CUV030's post hoc sunlight endpoint was not independent confirmatory efficacy evidence.[6][7] See trial estimands and missing data for why assumptions cannot be omitted from a result summary.
European diary and analysis limitations
The European diary accidentally omitted printed boxes for 2–3 p.m., part of the intended primary window. Participants were asked to add responses by hand, but FDA found that hour underreported. A 10 a.m.–2 p.m. sensitivity analysis remained favorable, P=0.006; this does not restore the missing hour or make the diary identical to the U.S. instrument.[6]
All 74 treated European participants returned at least one diary, but that does not mean complete daily data. FDA also describes differences between planned and implemented handling of missing pain scores. Its review considered these data-quality issues when interpreting supportive evidence. The two studies should not be pooled into one universal number of “extra safe sunlight hours.”[6]
Secondary outcomes and multiplicity
The paper reports 77 versus 146 European phototoxic reactions and P=0.04. The U.S. per-person reaction comparison was not statistically significant, P=0.60; Table 2 gives a median of one reaction in each arm. The paper's narrative also prints “46 and 43 reactions,” but these numbers match its U.S. analysis-population counts and do not agree with the table's per-person summaries. They are not presented here as verified total event counts.[1]
The disease-specific EPP quality-of-life questionnaire favored active treatment at several visits. For example, U.S. mean changes at day 180 were 51.1 versus 36.8 points, nominal P=0.02. The European final-visit comparison was not statistically significant, P=0.06, and the general Dermatology Life Quality Index did not show the same improvement pattern. FDA considered the submitted evidence insufficient to establish that the quality-of-life instruments were fit for the proposed efficacy assessment and did not use those results in its overall efficacy conclusion.[1][6]
The paper reports photoprovocation in 21 U.S. participants, whereas the current registry table reports 20, ten per arm. Some reported values also differ, so the records should not be silently merged. FDA notes that the small subsets were not powered for these comparisons, had exposure ceilings, and were not part of a multiplicity-controlled testing hierarchy.[1][3][6]
FDA found no proposed multiplicity adjustment for the secondary endpoint families in either trial. Their P values are nominal; favorable secondary findings are not a collection of independently confirmed benefits.[6]
Adverse events and postmarketing information
| Paper-reported safety measure | European active | European placebo | U.S. active | U.S. placebo |
|---|---|---|---|---|
| Any adverse event | 34/38 | 32/36 | 45/48 | 39/45 |
| Serious adverse events reported | 1 | 0 | 3 | 2 |
The paper reports four serious events in active groups and two in placebo groups, judged unrelated by investigators, and no deaths in these two trials. Relatedness is an assessment, not proof of no risk. The wider FDA development-program review includes additional studies and a later death, so the paper's no-death statement should not be generalized to that whole program.[1][6]
Common reported events included headache, nausea, nasopharyngitis and back pain. Table 5 records implant-site discoloration in 4/38 European and 9/48 U.S. active participants, versus none in either placebo arm. The paper's broader narrative pigmentation description is not substituted for those specifically coded event counts.[1]
The current European registry's adverse-event module does not reproduce the paper's safety table: it lists zero serious events and smaller nonserious-event totals. The available sources do not establish identical reporting scope, and these entries are not used to erase the paper's reported serious event. The U.S. registry safety module specifies day 0–180, distinct from the later safety-follow-up visit.[3][4]
Current labeling adds postmarketing reports of serious hypersensitivity, including anaphylaxis, and skin-pigmentation monitoring warnings. Its pooled trial adverse-reaction table covers three studies and 244 treated participants through month six, not just the two studies in this paper. Voluntary postmarketing reports cannot establish a reliable incidence by themselves.[5] See adverse events and tolerability for these reporting distinctions.
Enrollment, treatment and completion are different counts
| U.S. population | Active | Placebo |
|---|---|---|
| Initially randomized | 48 | 46 |
| Dispensed / treated | 48 | 45 |
| Primary diary analysis | 46 | 43 |
| Received all three implants, FDA exposure table | 46 | 42 |
| Completed six-month study period, registry | 45 | 42 |
| Completed later safety follow-up, registry | 44 | 40 |
The original paper's combined completion-and-implant row gives 46 and 42; the separate registry completion counts are 45 and 42. Completing every planned implant and completing every visit are not identical events. The four U.S. placebo discontinuations in the paper include the pre-dose withdrawal, leaving three additional treated-placebo discontinuations.[1][3][6]
In Europe, 34 participants in each group completed the study, but FDA's exposure table shows 34 active and 35 placebo participants received all five implants; one placebo participant left after the last implant but before the final visit. FDA's European disposition prose says 76 randomized and 74 treated, while its following table, the paper and the current registry report 38/36, totaling 74. That source discrepancy remains unresolved here; no additional arm assignments or exclusions are invented.[1][4][6]
Sponsor role, registry status and access limits
Clinuvel designed and sponsored the trials with investigator input, supplied implants, monitored the studies and managed the recorded data. Investigators supplied site data; authors received the locked database to check data and analyses. CPR Pharma Services performed the statistical analyses. The first two authors drafted the manuscript with the last two authors' assistance, and the paper says no nonauthor contributed to writing.[1]
The paper reports Clinuvel support, NIH grant R15 HL117199 and career award K23 DK095946. Author disclosures include company consulting, grants, travel, stock or patent interests; not every disclosed relationship concerns Clinuvel. Sponsor involvement is important provenance, not proof that results are invalid.[1]
Both current registries are Completed and have results: U.S. enrollment 93, last update September 19, 2019; European enrollment 74, last update October 12, 2021. Neither has attached protocol/SAP documents. This review read the complete official PMC author manuscript and pertinent FDA methods, data-quality, efficacy and safety assessments. The separate journal appendix/protocol/SAP and participant-level data were not obtained or reanalyzed; FDA's descriptions of those plans are not a claim that this review inspected their complete originals.[1][3][4][6]
FDA approval and material identity
FDA approved the finished Scenesse implant on October 8, 2019 under NDA 210797 to increase pain-free light exposure in adults with a history of phototoxic reactions from EPP. Drugs@FDA lists the August 12, 2024 labeling supplement. The current DailyMed page is updated May 11, 2026, while the clinical prescribing information is revised August 2024; those are different dates, not two approvals.[5][7][8]
The label specifies a controlled-release implant containing afamelanotide acetate, with the strength expressed as afamelanotide, and a defined polymer matrix. Chemical identity, salt/amount basis, finished form and release behavior are separate fields. This approval does not transfer to a powder, vial, research-use label or “Melanotan I” listing.[5] The approved-drug versus research-material guide and peptide salts and amount basis guide explain why.
What these studies do not prove
These studies do not establish prevention of every phototoxic reaction, treatment of every porphyria, improvement in liver complications, unrestricted sunlight safety, or an effect in excluded populations. Small rare-disease samples, behavior-dependent diaries, partial unmasking, missing observations, nominal secondary testing and source discrepancies limit interpretation.
They also do not establish a complete long-term safety profile or authenticate another material's identity, purity, stability, manufacture or exposure. The favorable product-specific regulatory decision and the methodological limitations can both be true. Use the trial and FDA status guide to distinguish a published study from a current regulatory record.
Frequently asked questions
What did the paper's U.S. primary analysis report?
Median pain-free direct-sunlight exposure was 69.4 versus 40.8 hours in 89 people with diary data; the paper reports P=0.04.
Why does the current label give different U.S. medians?
It gives 64.1 versus 40.5 hours, matching FDA's broader 93-person analysis rather than the paper's 89-person diary analysis.
Why are FDA's P values 0.055 and 0.091?
They come from two different missing-diary assumptions in the 93-person randomized-and-dispensed set. Neither analysis includes the placebo participant excluded before dispensing.
Was the European primary endpoint unchanged from the original protocol?
No. FDA documents a change to the analysis plan before that study was unblinded, informed by findings from another completed trial.
Did both trials show fewer phototoxic reactions?
The European paper comparison favored active treatment at the nominal level. The U.S. per-person comparison was not statistically significant.
Is afamelanotide FDA approved?
The finished Scenesse implant is approved for the specific adult EPP indication. Approval does not eliminate the label's contraindications, warnings or evidence limitations.
Does approval validate “Melanotan I” powder?
No. A shared name does not establish an equivalent finished drug, formulation, release profile, quality or regulatory status.
Primary and official sources
- Primary paper: DOI 10.1056/NEJMoa1411481; official PMC author manuscript
- PubMed: PMID 26132941
- ClinicalTrials.gov: U.S. CUV039, NCT01605136
- ClinicalTrials.gov: European CUV029, NCT00979745
- FDA Scenesse label, revised August 2024; current DailyMed Scenesse label
- FDA: NDA 210797 multidisciplinary review
- FDA: NDA 210797 approval letter
- Drugs@FDA: Scenesse application and labeling history
Review and corrections
Sources checked September 2, 2026 (America/Edmonton) / September 3, 2026 (UTC). Serum Research is responsible for this editorial summary. No independent clinical, statistical, scientific, analytical, or regulatory review is claimed. This is not medical advice, treatment guidance, or a material-quality or equivalence conclusion.
Read how we review evidence, browse the study records or research library, and contact us with a correction and its source.