Correction, 2026-10-10. This piece analyzes the same source as the analysis dated 2026-10-07 (regenerative-endodontics-apexification-148-teeth-six-month): Jiang, Bai and Chen (2026), Frontiers in Medicine, DOI 10.3389/fmed.2026.1945140. Publishing it as a separate dated piece contravened this site’s dedup contract, one source, one analysis. The 2026-10-07 piece, including its own dated correction, stands as the record for this source; this page is retained so nothing already linked disappears, but its content should be read as a second account of the same study, not new evidence.
What was tested
Huan Jiang and colleagues report a retrospective single-center cohort of 148 immature permanent teeth in 148 children treated for periapical periodontitis between January 2022 and December 2025, published in Frontiers in Medicine in October 2026 (open access). The study was approved by the ethics committee of the Sichuan Hospital of Stomatology (No. 2026010008). Each child contributed one tooth. Teeth were allocated non-randomly by the attending pediatric endodontists: 74 to regenerative endodontic procedures (REP) and 74 to apexification. The apexification group received calcium hydroxide paste, replaced at three-month visits until a radiographic apical barrier formed, then conventional obturation. The REP group had minimal instrumentation with 10/15 K files, 3% hydrogen peroxide and saline irrigation, a one-to-two-week calcium hydroxide dressing, then a second visit at which periapical bleeding was induced 2 to 3 mm beyond the apex, a blood clot was allowed to form for 10 to 15 minutes, and the canal orifice was sealed with 2 to 3 mm of ProRoot MTA. Outcomes at six months used a composite clinical-plus-radiographic grade (markedly effective, effective, or invalid), plus measured root canal wall thickness and root length, sulcus bleeding index, plaque index, probing depth, gingival crevicular fluid bFGF and VEGF by ELISA, and complications. Prognostic analysis split all 148 teeth into good (n = 122) and poor (n = 26) prognosis groups and ran multivariate logistic regression.
What they found
The primary efficacy result favored REP. The total effective rate (markedly effective plus effective) was 90.54% (67 of 74) in the REP group versus 74.32% (55 of 74) after apexification (chi-squared 6.719, P = 0.009), with invalid outcomes in 9.46% versus 25.68% of teeth.
Radiographic measurements moved in the same direction. Root canal wall thickness went from 2.03 plus or minus 0.25 mm to 3.00 plus or minus 0.46 mm in the REP group and from 2.04 plus or minus 0.18 mm to 2.78 plus or minus 0.35 mm after apexification (between-group P = 0.001). Root length increased from 8.08 plus or minus 0.22 mm to 8.76 plus or minus 0.42 mm with REP versus 8.14 plus or minus 0.27 mm to 8.58 plus or minus 0.34 mm with apexification (P = 0.004). Periodontal indices at six months were lower with REP: sulcus bleeding index 1.47 plus or minus 0.30 versus 1.78 plus or minus 0.33, plaque index 1.37 plus or minus 0.27 versus 1.76 plus or minus 0.29, and probing depth 3.23 plus or minus 0.63 mm versus 3.99 plus or minus 0.71 mm (all P < 0.001). Gingival crevicular fluid bFGF was 40.66 plus or minus 4.43 pg per microliter with REP versus 36.09 plus or minus 4.71 after apexification, and VEGF was 41.93 plus or minus 6.44 versus 38.18 plus or minus 4.57 pg per microliter (both P < 0.001). Complications (alveolar bone inflammatory absorption, root fracture, or gingival abscess) occurred in 3 of 74 REP teeth (4.05%) versus 10 of 74 (13.50%) (chi-squared 4.132, P = 0.042).
The prognosis analysis identified the two groups’ baseline difference. Teeth with a poor prognosis had less REP (26.92% versus 54.92%, P = 0.009) and larger apical lesions (3.11 plus or minus 0.43 mm versus 2.90 plus or minus 0.39 mm, P = 0.016). In the regression, apexification carried 4.081 times the odds of a poor prognosis versus REP (95% CI 1.532 to 10.873, P = 0.005), and each 1 mm increase in apical lesion range multiplied the odds by 4.980 (95% CI 1.531 to 16.197, P = 0.008). Age and Nolla developmental stage, forced into the model as confounders, were not significant; model fit was modest (Nagelkerke R-squared 0.247).
What it does not show
This is a retrospective, non-randomized, single-center comparison, and the authors say so directly: allocation was chosen by clinicians, so the 4.081 odds ratio for apexification may partly reflect baseline differences between teeth assigned to each treatment rather than a pure treatment effect. The “total effective rate” is an author-defined composite of symptom resolution and radiographic appearance, not a validated patient-reported or histologic outcome. Six months is short: nothing here addresses the durability, fracture resistance, or long-term survival of these teeth, and the root-length gain (0.68 mm over 6 months in the REP group versus 0.44 mm with apexification, a 0.24 mm difference in gain) is a statistical difference, not proof of restored physiology. No vitality, sensibility, or innervation testing was reported, so the work does not establish that regenerated vital pulp formed; REP outcomes in general can include repair tissue rather than true pulp. The bFGF and VEGF elevations are associative biomarker shifts in crevicular fluid, not evidence of a mechanism. And the poor-prognosis subgroup has only 26 teeth, so the regression’s wide confidence intervals reflect limited power.
Where it sits in the field
For the pulp and dentin repair program, this is routine-of-care evidence rather than a research advance: it compares two established treatments in the exact indication both are used for, necrotic immature permanent teeth in children, and finds the biologic approach superior on short-term composite outcomes. Its incremental value over earlier retrospective series is the explicit prognostic framing, and the lesion-size finding is the practically useful part: preoperative periapical lesion extent, measurable on routine imaging, was the dominant modifiable-looking predictor of a poor six-month outcome. The result is consistent with the direction of the existing clinical literature on regenerative endodontics, so it shifts no program tier; the current field assessment stands at /field/.
Where we differ from the coverage
We found no press or popular coverage of this paper to differ from. Against the paper itself, we bound two framings. First, the abstract’s phrase “favorable short-term outcomes” is right to stress short-term; the six-month window and composite grade support a clinical-efficacy claim, not a regeneration claim, and we have avoided the paper’s occasional “tooth regeneration” phrasing, which this data does not demonstrate. Second, the discussion’s statement that “the risk of poor prognosis in children receiving apexification was 4.081 times higher” is the authors’ own number, but it comes from an observational model with R-squared of 0.247, and the authors themselves attach the confounding caveat; we quote it with that caveat attached rather than as a treatment effect.
Provenance: grounded in the full open-access text of Jiang, Bai and Chen, Frontiers in Medicine, October 2026, DOI 10.3389/fmed.2026.1945140, read in full from the publisher site; every number above was checked against the article text and tables. Method and sourcing standard at /method/.