What was measured, and in what

Mokhtari and colleagues, in “Laser-induced modulation of root dentin surface for growth factor release and enhanced stem cell response” (Lasers in Medical Science, published 2026-08-04, DOI 10.1007/s10103-026-04971-9), asked whether laser irradiation can condition root dentin to release its matrix-bound TGF-beta1 as well as, or better than, the standard chemical chelator EDTA. Regenerative endodontics relies on that release: TGF-beta1 sequestered in dentin is one of the signals thought to help recruit and differentiate the stem cells a treated canal depends on, and it is only one factor among several (BMPs, IGF, and others), the only one this paper measured.

This is a purely in vitro tooth-block study. There is no animal arm and no pulp-regeneration endpoint. The substrate was mid-root dentin blocks (5 by 5 by 2 mm) cut from freshly extracted single-rooted human teeth. Cells were SHED (stem cells from human exfoliated deciduous teeth) from a single institutional stock, seeded directly onto the treated blocks. Following the American Association of Endodontists guideline, every group except the untreated control was first irrigated with 1.5% NaOCl for 5 minutes; lasers were applied after that.

The design crossed three lasers at three energy densities: Nd:YAG (1064 nm, pulsed) and diode (980 nm and 635 nm), each at 24, 60, and 120 J/cm2, giving nine laser groups plus three controls (untreated, NaOCl only, and the NaOCl then 17% EDTA positive control). TGF-beta1 release was measured by ELISA at n=5 per group; viability (MTT) and day-7 ALP activity at n=3 per group; surface changes by SEM.

The core result: one setting beat EDTA

TGF-beta1 released into buffer differed strongly across the twelve groups (one-way ANOVA F(11,48) = 54.467, p < 0.0001). The means, in pg/mL:

  • Nd:YAG 60 J/cm2: 566.14 (SD 55.65). The single highest group.
  • Nd:YAG 24 J/cm2: 385.28 (SD 65.55).
  • Nd:YAG 120 J/cm2: 284.06 (SD 25.11).
  • Diode 980 at 60 J/cm2: 410.29 (SD 74.47).
  • NaOCl then EDTA (positive control): 378.10 (SD 60.91).
  • Untreated (negative control): 20.93 (SD 8.55).
  • NaOCl only: 11.36 (SD 5.59).

The headline comparison is Nd:YAG at 60 J/cm2 versus the EDTA positive control: a mean difference of 188.0 pg/mL in the laser’s favor, p = 0.0217 (Games-Howell). That is the one setting in the study that significantly outperformed EDTA. NaOCl alone did not raise TGF-beta1 above untreated dentin at all (p = 0.641), which is why the chemical chelation step matters.

The cell numbers: viable, and higher early ALP

No group was cytotoxic. At 72 hours, SHED viability on dentin conditioned with Nd:YAG at 24 and 60 J/cm2, diode 980 at 24 and 120 J/cm2, and diode 635 at 24 J/cm2 was significantly higher than on both the NaOCl-only and the EDTA surfaces (p < 0.05). The EDTA surface was statistically indistinguishable from untreated dentin for viability at every time point (p > 0.05), a hint that strong demineralization does not by itself make a friendlier surface for cells.

Early differentiation, read as day-7 ALP activity, tracked the TGF-beta1 result: Nd:YAG at 60 J/cm2 gave the highest ALP (1.746 U/mL), significantly above the EDTA control (1.230 U/mL, p < 0.05), while all laser groups exceeded the untreated control (0.371 U/mL, p < 0.001). The authors are explicit that ALP is an indirect marker and, in their words, “should be interpreted as an early indicator of differentiation potential rather than definitive evidence of odontogenesis.”

Where the dose-response breaks

The non-obvious finding is that more energy was not better and wavelength mattered. For Nd:YAG, the moderate 60 J/cm2 setting beat its own 120 J/cm2 setting by 282.1 pg/mL (p = 0.0014), and 120 J/cm2 fell back to statistical parity with EDTA (p = 0.2556). The 635 nm diode went the wrong way as energy rose, from 363.75 pg/mL at 24 J/cm2 down to 148.52 at 60 and 100.25 at 120. SEM explains part of this: at 120 J/cm2 both Nd:YAG and the 980 nm diode caused “surface melting, resolidification, and occasional microcracks,” and the authors attribute the loss at high fluence to thermal denaturation of the protein and melted peritubular dentin occluding the tubules. There is also a split in what the best release setting can do. EDTA does two jobs in the chemical step: it releases growth factor and it removes the smear layer. In this study only the two 120 J/cm2 laser settings (Nd:YAG and 980 nm) matched EDTA on smear-layer removal, and those are the same damaged, low-release settings. So Nd:YAG at 60 J/cm2 beat EDTA on TGF-beta1 release but did not match EDTA on smear clearance. The effective window is narrow, and the outcome is a combined effect of wavelength, power, and delivery, not a single tunable dial: the authors caution that their design “does not allow attribution of the results to a single variable.”

Honest boundary

Read plainly, this is a bench result about how much TGF-beta1 diffuses out of a dentin block into buffer over 24 hours, plus a favorable but early cell-surface response from one SHED source. It is not pulp regeneration, not odontoblast differentiation confirmed by mineralization or DSPP and DMP-1 expression, and not an in vivo result. Releasing more growth factor into buffer is a proxy for the biology that matters, not the biology itself.

Several limits bound the claim, some the authors flag themselves: a single growth factor was measured; the sample sizes are small (n=5 for release, n=3 for viability and ALP); surface analysis was qualitative, with no quantitative roughness measurement; tooth-to-tooth heterogeneity could shift dentin thickness, tubule density, and laser absorption; and, notably for a paper whose entire argument turns on fluence, laser output power was not verified with a calibrated power meter, relying instead on the manufacturer’s stated values. The 635 nm arm also raised energy by lengthening exposure rather than power, adding a thermal-time confound the authors acknowledge. Generalizability past one SHED stock is untested, and translation to in vivo models, in the authors’ words, “is essential to confirm the regenerative potential.”

No lay, university, or trade-press coverage of this specific paper exists as of publication, so there is no divergent framing to correct here. The framing to watch for if coverage does appear is the familiar “lasers release healing growth factors to regrow teeth”: this study does not support that at all. It measured growth factor release from dentin, not tooth regrowth, and even the release claim holds only at one narrow setting.

Where it sits in the field

Within thirdteeth this maps to the pulp-dentin-repair program (see /programs/pulp-dentin-repair/), route repair. It is a dentin-conditioning method study: a candidate for the growth-factor-release part of the chemical step of a regenerative endodontic procedure, not a full replacement for EDTA (which still cleared the smear layer better here) and not a new route to a tooth. On the evidence ladder it sits well behind the human pulp-repair data in that program and behind the many EDTA and irrigant comparisons that precede it; its contribution is a specific, testable parameter claim (Nd:YAG, 60 J/cm2) that another lab can now try to reproduce, ideally against a calibrated power meter and a broader donor set. For how we rank and bound this kind of work, see /field/.

Provenance: numbers verified against the primary source, read in full from the PMC full text and the eight supplementary statistical tables; scoring and boundary rules per /method/.