What was registered
A three-arm randomized trial of the self-assembling peptide P11-4 (Curodont Repair, vVARDIS, Switzerland; oligopeptide-104) for enamel carious white spot lesions was first posted on ClinicalTrials.gov on 23 June 2026 as NCT07664046, lead sponsor the Medical University of Warsaw. The registry lists the study as recruiting, with actual start 2 February 2026, estimated primary completion 31 October 2027, and estimated study completion 14 December 2027. The full protocol was published on 7 October 2026 in Frontiers in Oral Health (volume 7, open access, CC BY 4.0) by Turska-Szybka and colleagues in the Department of Paediatric Dentistry. The abstract reports ethics approval from the Bioethics Committee of the Medical University of Warsaw (KB/1/2026, 12 January 2026) and a planned recruitment window of 1 February to 31 July 2026.
The design
Seventy-five participants aged over 10 and under 18, each with at least two active enamel white spot lesions on smooth surfaces of permanent teeth (ICDAS II code 1 or 2, noncavitated, vital pulp), will be randomized 1:1:1. Group A receives a single Curodont Repair application plus fluoride varnish (Duraphat, 22,600 ppm F) every three months; Group B receives Curodont Repair at baseline and again at the three-month visit plus the same varnish schedule; Group C receives varnish alone. Randomization uses block size 3 with sequentially numbered opaque sealed envelopes, and allocation is drawn in a second step so the enrolling investigator cannot know it in advance. The trial is single-masked at the level of the outcomes assessor: participants are told their assigned group after randomization, two investigators deliver the intervention, do the clinical examination, and record adverse events unblinded, and a third investigator, blinded throughout, performs all quantitative light-induced fluorescence (QLF) imaging. Sample size rests on a two-sided alpha of 0.05 and 80% power: about 21 evaluable participants per group, inflated to 25 per group for an anticipated 15% attrition. The effect-size assumption, roughly a 20-unit between-group difference in lesion volume at six months (minus 21.92 plus or minus 29.28 versus minus 1.90 plus or minus 16.27), is taken from Güven et al. and compares P11-4 against fluoride, not single against repeated application; the authors state plainly that no prior evidence exists for the trial’s actual primary contrast and the number is a pragmatic planning assumption. Analysis uses linear mixed-effects models with random intercepts for patient and for lesion nested within patient.
The paper and the registry disagree on the primary endpoint
This is the part worth reading carefully. The protocol paper’s stated primary endpoint is the change in the QLF-derived lesion volume parameter, ΔQ, from baseline to the six-month time point, comparing repeated versus single P11-4 application; fluorescence loss, ΔF, is a secondary endpoint there. The registry record instead lists two primary outcomes, both at twelve months: change in ΔF and change in ΔQ from baseline to twelve months, with the six-month ΔQ change appearing only among secondaries. So the prespecified primary comparison, the time point, and the number of primary endpoints differ between the two public records of the same trial. The dates also diverge: the abstract says study completion is expected in July 2027, while the registry estimates primary completion at the end of October 2027 and final completion in mid-December 2027. The within-paper site name is inconsistent as well (University Dental Center in one section, University Clinical Center in two others).
What it does not show
A protocol is a plan, not a result: no participant data exist yet, and this piece reports nothing about efficacy. The question the trial can answer is also narrower than the field’s headline interest. P11-4 is a remineralization adjunct for early, noncavitated enamel lesions; it works on existing teeth with existing white spots, not on regrowing tooth structure, so it sits outside the five program routes this site tracks and is recorded here as a dated trial event only. Within that lane, the trial compares application schedules of an already-marketed agent against varnish, in one center, in adolescents with smooth-surface lesions; children under 10, cavitated lesions, and patients in fixed appliances are all excluded, so the result will not generalize to the orthodontic white spot population that drives much of the clinical demand. The modest sample, single site, and envelope-based concealment are serviceable but not strong.
Where we differ from the coverage
We found no press or popular coverage of this registration to differ from. Against the records themselves, we note the endpoint and date divergences above, which neither the abstract nor the registry entry flags. We also decline to treat the registry enrollment of 75 as fixed: the registry marks it as estimated. And where the abstract frames the trial as answering the “optimal application protocol,” the honest boundary is narrower, a dosing-schedule comparison for one product in one age band and lesion type, with a primary endpoint definition that remains unsettled between the two public versions of the protocol.
Provenance: grounded in the ClinicalTrials.gov record NCT07664046 (API record read in full on 9 October 2026), the publisher’s abstract and full protocol text of Turska-Szybka et al., Frontiers in Oral Health 7, 2026, DOI 10.3389/froh.2026.1932140, and the Crossref metadata; every number above was checked against at least two of those three sources. Method and sourcing standard at /method/.