Adjunctive antimicrobial photodynamic therapy in the nonsurgical treatment of periodontitis: an umbrella review

Abstract

Aim

Antimicrobial photodynamic therapy (aPDT) has been proposed as a locally acting adjunct to subgingival instrumentation for periodontitis. This umbrella review aimed to synthesize and critically appraise systematic reviews of adjunctive aPDT during active nonsurgical periodontal therapy, with emphasis on methodological confidence, clinical importance, and overlap of the evidence base.

Method

PubMed/MEDLINE, the Cochrane Library, Epistemonikos and Google Scholar were searched from inception through 24 July 2026, supplemented by backward and forward citation searching. For Google Scholar, the first 200 relevance-ranked results were screened, and exact-title and author searches were used to identify additional review reports. Systematic reviews with or without meta-analysis were eligible when they evaluated aPDT added to subgingival instrumentation and reported separable clinical evidence for adults receiving active nonsurgical treatment. Reporting followed PRISMA 2020 and PRIOR. Methodological confidence was appraised with AMSTAR 2; ROBIS domains were used as an interpretive framework rather than as a formal domain-by-domain rating tool. One reviewer completed selection, extraction and appraisal using a prespecified form and a structured second source-check pass for eligibility and numerical data.

Results

Fifteen systematic reviews published from 2010 to 2025 were included. After recalculation using the AMSTAR 2 decision rules, four reviews had high methodological confidence, three had moderate confidence, three had low confidence, and five had critically low confidence. The 2024 Cochrane review included 50 randomized trials and 1407 adults. During active treatment at 6 months, adjunctive aPDT produced an additional PPD reduction of 0.52 mm (95% CI 0.31–0.74; 15 studies, 452 participants) and CAL gain of 0.44 mm (95% CI 0.24–0.64; 13 studies, 414 participants); both were very low-certainty and judged unlikely to be clinically important. BOP improved by 5.72 percentage points (95% CI 1.62–9.81; 5 studies, 171 participants), also with very low certainty. No aPDT-related adverse effects were reported in 24 active-treatment studies involving 639 participants, while pocket closure, satisfaction, quality of life and economic outcomes were sparsely reported.

Conclusion

Adjunctive aPDT may produce small short-term improvements in PPD and CAL after nonsurgical periodontal therapy, but the most reliable evidence remains very low certainty, and the observed mean differences have not been shown to translate into patient-important or treatment-changing benefit. Current evidence does not support routine adjunctive use; standardized trials should prioritize pocket closure, patient-reported outcomes, long-term effectiveness and economic outcomes.

PROSPERO registration

CRD420261468530

Introduction

Periodontitis is a chronic, dysbiotic and inflammation-mediated disease that destroys the periodontal ligament and alveolar bone and remains a major cause of tooth loss and oral disability [1–3]. Global estimates indicate that more than one billion people are affected by severe periodontitis, and the absolute burden is expected to rise with population growth and ageing [4, 5]. Contemporary classification integrates stage, grade, extent and complexity [2, 3]. Treatment is organized as a stepwise process encompassing behavioural change, supragingival biofilm control, risk-factor management, subgingival instrumentation, reassessment, periodontal surgery when indicated and long-term supportive care [6, 7].

Periodontitis develops through a reciprocal interaction between a dysbiotic subgingival microbial community and a susceptible host inflammatory response [8]. Mechanical disruption of the subgingival biofilm and removal of calculus remain central to active therapy; however, root anatomy, furcation involvement, deep pockets, tissue penetration of microorganisms, operator access and microbial recolonization may restrict the response achieved by instrumentation alone [6, 7]. These limitations have encouraged investigation of locally delivered adjuncts that might enhance microbial reduction without routine systemic antibiotic exposure.

Antimicrobial photodynamic therapy is based on the interaction of a photosensitizer, light of an appropriate wavelength and molecular oxygen. Photo-excitation generates reactive oxygen species, including singlet oxygen and free radicals, which oxidize microbial membranes, proteins and nucleic acids [9, 10]. The concept was applied to oral bacteria in early laboratory work and subsequently expanded across dental infections [11, 12]. In periodontal pockets, commonly studied photosensitizers include methylene blue, toluidine blue O, phenothiazine chloride and indocyanine green. Light wavelength, power, fluence, exposure time, pre-irradiation interval, fibre geometry, pocket irrigation, number of sessions and timing relative to instrumentation vary substantially. These parameters influence photosensitizer uptake, oxygen availability, optical penetration and the balance between photodynamic and photothermal effects [9, 10, 12].

Systematic reviews of adjunctive aPDT have been published since 2010, initially synthesizing small trial sets [13, 14]. Subsequent reviews considered expanding or overlapping randomized evidence bases [15–17]. Later meta-analyses continued to examine clinical outcomes and methodological differences [18–20]. Additional syntheses evaluated clinical outcomes across heterogeneous interventions and trial designs [21–23]. More recent reviews continued this evaluation and included the comprehensive 2024 Cochrane assessment [24–26]. Evidence synthesis continued the following year [27]. Their conclusions are discordant. Several meta-analyses reported statistically significant short-term improvements in PPD and CAL, while other reviews found no additional benefit or judged the evidence too uncertain for routine practice. Repeated inclusion of the same primary trials can create an appearance of replication even when the independent evidence base has expanded only modestly [22, 26].

An umbrella review is therefore appropriate because it can answer questions that a single primary-study meta-analysis cannot: how review methods differ, how often the same trials are reused, why pooled conclusions conflict, and whether statistically detectable effects correspond to outcomes that alter clinical management. The present umbrella review therefore evaluated systematic reviews of aPDT as an adjunct to active nonsurgical periodontal therapy. Its distinctive contribution beyond the 2024 Cochrane review is to compare review-level methodological confidence, map recurrent use of the evidence base, trace the evolution of conclusions over time, and separate statistical significance from clinically meaningful benefit. The objective was to synthesize clinical, microbiological, inflammatory, safety and patient-centred outcomes and to identify implications for practice and future research.

Materials and methods

Review design and reporting

This study was conducted as an umbrella review, with the systematic review as the principal unit of analysis. The design was informed by methodological guidance for overviews of reviews and was reported in accordance with PRISMA 2020 and the PRIOR statement [28–30]. The review is registered in PROSPERO (CRD420261468530), and the PRIOR checklist is provided as an additional file. AMSTAR 2 was used to appraise methodological confidence [31]. ROBIS domains informed interpretation of limitations in eligibility, identification and selection, data collection and appraisal, and synthesis; a formal domain-by-domain ROBIS rating was not generated [32]. GRADE certainty terminology was reproduced only when it had been explicitly reported by an included review; no de novo umbrella-level GRADE rating was assigned [33, 34].

Review question

Among adults receiving active nonsurgical treatment for periodontitis, does aPDT used as an adjunct to subgingival instrumentation, compared with instrumentation alone or sham/placebo aPDT, improve clinical periodontal outcomes, microbiological or inflammatory outcomes, patient-reported outcomes and safety? The full eligibility criteria are given in Table 1.

Table 1.

Eligibility criteria for the umbrella review.

Domain	Eligibility criterion
Population	Adults with clinically diagnosed periodontitis undergoing active nonsurgical periodontal treatment. Reviews using former chronic-periodontitis terminology or the current stage/grade framework were eligible.
Intervention	aPDT delivered during or after subgingival instrumentation, irrespective of photosensitizer, wavelength, light source or number of sessions.
Comparator	Subgingival instrumentation alone or with sham/placebo aPDT.
Outcomes	PPD, CAL, bleeding on probing (BOP), gingival recession, plaque or gingival indices, pocket closure, microbiological outcomes, inflammatory biomarkers, adverse events, discomfort, satisfaction, quality of life or cost.
Review design	Systematic review with or without meta-analysis that predominantly synthesized randomized or controlled clinical trials. Mixed-scope reviews were eligible when active nonsurgical periodontitis findings were separable.
Core exclusions	Narrative reviews; primary studies; peri-implant-only reviews; surgical-only reviews; supportive-periodontal-therapy-only reviews; reviews restricted to diabetes, HIV, smokers, aggressive/molar-incisor periodontitis or a single current stage/grade; reviews comparing aPDT only with systemic antibiotics; and reviews restricted to one photosensitizer when their purpose was protocol-specific rather than general clinical effectiveness.

Information sources and search strategy

PubMed/MEDLINE, the Cochrane Library, Epistemonikos and Google Scholar were searched from inception to 24 July 2026. Backward reference searching of eligible reviews and forward citation searching were used to identify additional review reports and updated versions. Search terms combined controlled vocabulary and free-text terms for periodontitis, periodontal disease, photodynamic therapy, photochemotherapy, photodisinfection, antimicrobial photodynamic therapy, systematic review and meta-analysis. No language restriction was applied at the search stage. In Google Scholar, the first 200 results ranked by relevance were screened, followed by exact-title and author searches for potentially eligible reviews. Source-specific syntax, field restrictions, the search date and citation-search procedures are provided in Supplementary Table 1. The review-selection diagram begins with the deduplicated candidate set used for full-text eligibility assessment, providing a stable denominator for the documented selection process.

Selection process

Titles, abstracts and bibliographic records were screened against the predefined scope, followed by full-text assessment of potentially eligible systematic reviews. Mixed-scope reviews were retained only when findings for active nonsurgical treatment of periodontitis around natural teeth were separable. Updated reviews were retained alongside earlier reviews because the objective was to evaluate the evolution and repetition of the evidence; interpretation was weighted toward the most recent, most comprehensive and methodologically strongest reviews. One reviewer completed study selection, extraction and initial appraisal using a prespecified form. A second structured pass was then used to re-apply eligibility criteria to every included and excluded full-text report and to recheck every numerical estimate used in the text and tables against the source report.

Data extraction

A structured extraction form captured authorship, publication year, search period, review scope, number and design of primary studies, participant characteristics, intervention protocols, comparators, follow-up, outcomes, effect measures, 95% confidence intervals, heterogeneity statistics, risk-of-bias methods, certainty assessment, adverse events, conflicts of interest and review conclusions. Where multiple time points were reported, 3-, 6- and 12-month findings were extracted separately. Mean differences expressed in millimetres were prioritized over standardized mean differences because they are more directly interpretable clinically. Numerical estimates were rechecked against the abstract, full text or publisher record. A detailed review-level evidence table is provided in Supplementary Table 2.

Clinical relevance framework

Clinical relevance was interpreted prospectively at the synthesis stage using treatment-relevant endpoints rather than a newly imposed universal millimetre threshold. Priority was given to pocket closure, the proportion of sites reaching PPD ≤ 4 mm without BOP, need for periodontal surgery, tooth retention, adverse effects, discomfort, satisfaction, oral-health-related quality of life and cost. When reviews reported only continuous PPD or CAL effects, absolute mean differences and 95% confidence intervals were presented and interpreted alongside risk of bias, heterogeneity, duration of follow-up and any explicit clinical-importance judgement made by the highest-confidence review. In particular, the 2024 Cochrane review explicitly judged its 6-month PPD and CAL effects as not clinically important [26].

Methodological confidence and risk of bias

AMSTAR 2 was applied using its 16 domains and seven critical domains: protocol registration, adequacy of the literature search, justification for excluded studies, risk-of-bias assessment, appropriateness of meta-analytic methods, consideration of risk of bias in interpretation and assessment of publication bias [31]. Overall confidence was recalculated directly from the original AMSTAR 2 decision rules: no more than one non-critical weakness = high; more than one non-critical weakness with no critical flaw = moderate; one critical flaw = low; and more than one critical flaw = critically low. Item-level judgements are reported in Supplementary Table 3, with decisive critical-domain explanations in Supplementary Table 4. ROBIS domains were used only to structure interpretation of potential bias in review conduct, not to generate a separate formal ROBIS rating.

Primary-study overlap

Primary-study overlap was evaluated by comparing review search dates, study counts, trial-design descriptions and identifiable recurrent primary-study citations reported in included-study tables or reference lists. Corrected covered area (CCA) was prespecified as the preferred quantitative measure when a consistent review-by-primary-study matrix could be constructed [35]. Three reviews did not provide consistently separable active-treatment trial lists, and the reporting unit varied across reviews between trials, publications and treatment comparisons. A single global CCA was therefore not calculated because combining non-equivalent units would create a misleadingly precise statistic. Instead, overlap was characterized as recurrent or substantial only when repeated primary trials and nested historical evidence bases were directly identifiable; the review-level basis for this interpretation is summarized in Supplementary Table 5 [35–37].

Data synthesis

A second-order meta-analysis was not performed. Pooling review-level estimates would have double-counted participants, mixed split-mouth and parallel-group designs, combined different effect measures and follow-up periods, and obscured differences in aPDT protocols. Split-mouth trials require paired analysis because outcomes within the same participant are correlated, and inappropriate treatment of such data can distort precision [38]. Discordant reviews were prioritized in a fixed sequence: (1) AMSTAR 2 confidence; (2) later search date when confidence was equal; (3) broader directly relevant randomized-trial coverage; (4) design-appropriate handling of split-mouth data; (5) formal outcome-level certainty assessment; and (6) direct relevance to active nonsurgical treatment. A lower-confidence review was not allowed to override a higher-confidence review solely because it reported a larger effect. Earlier reviews were retained to show how estimates and conclusions changed over time.

Results

Selection of systematic reviews

After deduplication and bibliographic screening, 28 potentially eligible systematic-review reports underwent full-text assessment. Thirteen were excluded. Among the five restricted-population or disease-phenotype reviews, three addressed aggressive periodontitis, HIV or diabetes-specific populations [39–41]. The remaining two addressed diabetes-specific or grade-specific disease [42, 43]. Two reviews were limited to supportive periodontal therapy or residual pockets [44, 45]. Two reviews compared aPDT principally with systemic antibiotics [46, 47]. Among the four single-photosensitizer reviews, one focused on methylene blue and two on indocyanine green [48–50]. The fourth focused on riboflavin [51]. Fifteen reviews met the criteria for the core synthesis [13–15]. The reconciled review-selection process is shown in Fig. 1, and the full-text exclusion log is provided in Supplementary Table 6.

Fig. 1. Review-level selection flow from the deduplicated candidate set.

Exclusion categories are reconciled with Supplementary Table 2.

Characteristics of included reviews

The characteristics of the included systematic reviews are given in Table 2. The 15 included reviews were published between 2010 and 2025 [13–27]. Twelve contained a meta-analysis and three were narrative systematic reviews. The evidence base expanded from four to five trials in the earliest reviews to 50 randomized trials and 1407 adults in the 2024 Cochrane review. Follow-up was usually 3 or 6 months, with relatively little evidence beyond 12 months. Split-mouth designs and small parallel-group trials were common. Intervention protocols differed in photosensitizer, wavelength, power, fluence, pre-irradiation contact time, fibre position and number of applications.

Table 2.

Characteristics and principal findings of included systematic reviews.

Review	Search/methods	Primary studies	Scope	Principal finding	AMSTAR 2
Azarpazhooh et al. 2010 [13]	Search to May 2009	5 studies	PDT alone or adjunct to SRP	PPD MD 0.25 mm (95% CI 0.04–0.45); CAL MD 0.34 mm (95% CI 0.05–0.63); early evidence base remained small.	Critically low
Atieh, 2010 [14]	Search to Oct 2008	4 RCTs; 101 participants	Adjunctive aPDT + SRP	Possible short-term benefit; substantial imprecision	Critically low
Sgolastra et al. 2013 [15]	Multiple databases	Small early RCT base	PDT alone or adjunctive	3-month PPD/CAL benefit; no clear 6-month benefit	Critically low
Sgolastra et al. 2013 [16]	PRISMA/Cochrane methods	14 RCTs	Adjunctive aPDT + SRP	PPD MD 0.19 mm (95% CI 0.07–0.31); CAL MD 0.37 mm (95% CI 0.26–0.47).	Low
Joseph et al. 2017 [17]	6 databases	20 RCTs/quasi-RCTs	PDT alone or adjunctive	Potential benefit, but high protocol heterogeneity	Critically low
Azaripour et al. 2018 [18]	Systematic review/meta-analysis	26 RCTs	Adjunctive aPDT + SRP	PPD MD 0.37 mm; CAL MD 0.33 mm	Low
Chambrone et al. 2018 [19]	AAP best-evidence review	Mixed periodontal indications; separable nonsurgical data	Adjunctive aPDT	Benefits inconsistent; patient-centred evidence limited	Moderate
Salvi et al. 2020 [20]	PROSPERO registered	2 comparable aPDT studies for 6-month PPD	Untreated periodontitis	6-month PPD WMD 0.35 mm (95% CI –0.04 to 0.73); two comparable studies; no statistically significant benefit.	High
Moro et al. 2021 [21]	Systematic review/meta-analysis	RCT evidence	Chronic periodontitis; adjunctive aPDT	Small short-term clinical benefit; heterogeneity remained	Moderate
Dalvi et al. 2021 [22]	PROSPERO registered	31 eligible; 18 meta-analysed	Chronic/aggressive terminology	31 eligible studies; 18 meta-analysed; design-stratified analyses found no consistent PPD/CAL benefit.	High
Pardo et al. 2023 [23]	Systematic review/meta-analysis	14 controlled trial reports	PPD ≥ 5 mm	3-month PPD SMD –0.76 with very high heterogeneity (I² = 88%).	Low
Souza et al. 2024 [24]	5 databases; 2 reviewers	4 RCTs	External pocket irradiation protocol	No clear clinical benefit over SRP alone	High
Mahdizade Ari et al. 2024 [25]	PROSPERO CRD42021267008; broad search	53 articles	Broad chronic-periodontitis PDT literature	Narrative benefit suggested; no pooled certainty estimate	Critically low
Jervøe-Storm et al. 2024 [26]	Cochrane review; search to 14 February 2024; duplicate processes; RoB 2/GRADE	50 RCTs; 1,407 adults	Active/supportive periodontal and peri-implant treatment	6-month PPD MD 0.52 mm (95% CI 0.31–0.74) and CAL MD 0.44 mm (95% CI 0.24–0.64); both very low certainty.	High
Wang et al. 2025 [27]	Systematic review/meta-analysis	18 RCTs	Adjunctive PDT + SRP	18 RCTs; PPD SMD –0.55 (95% CI –0.97 to –0.13; I² = 76.2%); statistical benefit with substantial heterogeneity.	Moderate

Methodological confidence

After recalculation from the item-level AMSTAR 2 judgements using the original decision rules, four reviews were rated high confidence, three moderate, three low and five critically low (Table 3). Salvi et al. and Dalvi et al. were rated high confidence [20, 22]. Souza et al. and the Cochrane review were also rated high confidence [24, 26]. Moderate-confidence reviews were Chambrone et al. Moro et al. and Wang et al. [19, 21, 27]. Low-confidence reviews were Sgolastra et al. (adjunctive meta-analysis), Azaripour et al. and Pardo et al [16, 18, 23]. Critically low reviews included Azarpazhooh et al. Atieh and Sgolastra et al. (broad review) [13–15]. Joseph et al. and Mahdizade Ari et al. were also critically low [17, 25]. The most consequential downgrades reflected multiple failures in critical domains, particularly protocol registration, excluded-study justification and publication-bias assessment. Item-level judgements and decisive critical-domain explanations are provided in Supplementary Tables 3 and 4.

Table 3.

Summary of methodological confidence.

AMSTAR 2 confidence	Reviews	Main basis for judgement
High	Salvi et al. [20]; Dalvi et al. [22]; Souza et al. [24]; Jervøe-Storm et al. [26]	No critical flaw and no more than one non-critical weakness in the item-level AMSTAR 2 assessment.
Moderate	Chambrone et al. [19]; Moro et al. [21]; Wang et al. [27]	No critical flaw, with more than one non-critical weakness.
Low	Sgolastra et al. (adjunctive meta-analysis) [16].; Azaripour et al. [18].; Pardo et al. [23]	One critical flaw, principally absence of a justified excluded-study list.
Critically low	Azarpazhooh et al. [13]; Atieh [14].; Sgolastra et al. (broad review) [15].; Joseph et al. [17]; Mahdizade Ari et al. [25]	More than one critical flaw, most often involving protocol registration, excluded-study justification and/or publication-bias assessment.

Primary-study overlap

Recurrent primary-study overlap was evident across the review series. The small trial corpus synthesized in 2010 continued to contribute to later reviews, while subsequent reviews added newer, predominantly small trials to the historical evidence base [13, 22, 26]. Because three reviews did not provide consistently separable active-treatment trial lists and because study units were reported inconsistently as trials, publications or treatment comparisons, a single global CCA was not computed. The overlap conclusion is therefore based on directly identifiable repeated trials, chronological nesting of search windows and the marked contrast between the growth in review count and the more modest growth in independent adequately powered parallel-group trials. Review count should not be interpreted as independent replication.

Probing pocket depth

PPD was the most frequently pooled outcome. Azarpazhooh et al. reported an additional PPD reduction of 0.25 mm (95% CI 0.04–0.45) [13]. Sgolastra et al. reported a mean difference of 0.19 mm (95% CI 0.07–0.31) [16]. Azaripour et al. reported approximately 0.37 mm additional reduction [18]. Salvi et al. found a nonsignificant 6-month weighted mean difference of 0.35 mm (95% CI –0.04 to 0.73) from two sufficiently comparable studies [20]. Dalvi et al. found no consistent additional PPD benefit when split-mouth and parallel-group trials were analysed separately [22]. Pardo et al. reported a larger standardized effect at 3 months (SMD –0.76), but heterogeneity was very high (I² = 88%), limiting interpretability [23].

The 2024 Cochrane review estimated an additional 0.52 mm reduction in PPD at 6 months during active periodontal treatment (95% CI 0.31–0.74; 15 studies, 452 participants; I² = 97%) [26]. The certainty was very low because contributing trials had high or unclear risk of bias and inconsistency was substantial; the authors judged the difference not clinically important. Wang et al. subsequently reported a statistically significant PPD effect (SMD -0.55, 95% CI –0.97 to –0.13; I² = 76.2%) [27]. Taken together, the review-level evidence supports a possible small short-term effect, but not a stable or clinically established incremental benefit.

Clinical attachment level

CAL findings broadly paralleled PPD. Azarpazhooh et al. reported an additional CAL gain of 0.34 mm (95% CI 0.05–0.63) [13]. Sgolastra et al. reported 0.37 mm (95% CI 0.26–0.47) [16]. Azaripour et al. reported approximately 0.33 mm additional gain [18]. Dalvi et al. found no consistent benefit after separating trial designs [22]. The Cochrane estimate was 0.44 mm at 6 months (95% CI 0.24–0.64; 13 studies, 414 participants), again with very low certainty and an explicit judgement that the difference was not clinically important [26]. Wang et al. reported a favourable standardized CAL effect with substantial heterogeneity [27]. The consistency of direction across several reviews must therefore be balanced against small absolute magnitude, risk of bias, short follow-up and recurrent study overlap.

Bleeding, plaque and gingival inflammation

Bleeding and plaque outcomes were less consistently reported and were measured using different definitions and indices. The Cochrane review estimated an additional 5.72-percentage-point reduction in BOP at 6 months during active treatment (95% CI 1.62–9.81; 5 studies, 171 participants; I² = 98%), with very low certainty and no clinically important difference [26]. Several earlier reviews reported short-term reductions in bleeding or gingival inflammation, whereas others found no additional effect after instrumentation [15, 17, 21]. Plaque findings are especially difficult to attribute to aPDT because they are strongly influenced by oral-hygiene instruction, behavioural adherence and co-interventions.

Pocket closure and clinically meaningful endpoints

Few reviews reported pocket closure, the proportion of sites with PPD ≤ 4 mm without bleeding, need for periodontal surgery or tooth-level treatment escalation. These endpoints are more closely aligned with contemporary periodontal treatment goals than small mean changes across all sites [6]. The absence of consistent pocket-closure data prevents a direct assessment of whether adjunctive aPDT changes subsequent clinical management. Mean differences below 0.5 mm may be statistically detectable in pooled analyses but may not translate into fewer residual pathological pockets or reduced need for surgery.

Microbiological and inflammatory outcomes

Microbiological outcomes were heterogeneous in sampling site, timing, target organisms and reporting scale. Some trials reported transient reductions in Porphyromonas gingivalis, Aggregatibacter actinomycetemcomitans or total bacterial load, while others found no additional reduction beyond instrumentation [15, 21, 25]. Methylene-blue- and indocyanine-green-specific reviews reported protocol-dependent signals but did not establish a generalizable antimicrobial regimen [48–50]. A riboflavin-specific review likewise did not establish a generalizable regimen [51]. Inflammatory biomarkers, including interleukins, tumour necrosis factor and matrix metalloproteinases, were reported by relatively few small aPDT studies, and no reproducible biomarker pattern established a durable aPDT-specific host-modulating effect.

Safety, patient-reported outcomes and cost

Adverse-event reporting did not identify a clear short-term safety signal during active treatment. In the 2024 Cochrane review, 24 active-treatment studies involving 639 participants reported no adverse effects attributed to aPDT, and the certainty for the reported adverse-effects outcome was moderate [26]. Adverse-event ascertainment, however, was often passive or briefly described. No active-treatment studies in the Cochrane review reported pocket closure at 6 months, participant satisfaction or quality of life. Participant discomfort and economic outcomes were also rarely incorporated across the review literature, limiting assessment of treatment value when the expected incremental clinical effect is small.

The full outcome-level synthesis and clinical interpretation are presented in Table 4, and the principal explanations for discordant review conclusions are summarized in Table 5. Detailed review-level estimates, confidence intervals, trial counts, participant counts, follow-up and heterogeneity statistics are provided in Supplementary Table 5.

Table 4.

Outcome-level synthesis and clinical interpretation.

Outcome	Direction and magnitude	Certainty reported by included reviews / evidence strength	Clinical interpretation
PPD	Usually favoured aPDT. High-confidence Cochrane estimate at 6 months: MD 0.52 mm (95% CI 0.31–0.74; 15 studies, 452 participants; I² = 97%) [26]	Cochrane: very low certainty; no de novo umbrella-level GRADE rating assigned.	Small short-term mean effect; the Cochrane review judged the difference not clinically important.
CAL	High-confidence Cochrane estimate at 6 months: MD 0.44 mm (95% CI 0.24–0.64; 13 studies, 414 participants) [26]	Cochrane: very low certainty; no de novo umbrella-level GRADE rating assigned.	Modest mean gain; clinical impact on pocket closure or subsequent treatment remains unresolved.
BOP/gingival inflammation	Cochrane 6-month BOP: MD 5.72 percentage points (95% CI 1.62–9.81; 5 studies, 171 participants; I² = 98%) [26]	Cochrane: very low certainty.	Difference was not judged clinically important; definitions varied across reviews.
Plaque indices	Occasional favourable short-term effects [18, 25, 27]	No uniform outcome-level certainty rating across reviews.	Strongly influenced by oral-hygiene instruction, adherence and co-interventions.
Microbiological outcomes	Mixed and usually transient [15, 17, 21, 22, 25, 48–51]	No de novo umbrella-level GRADE rating assigned.	No standardized microbial endpoint or durable suppression established.
Adverse events	Twenty-four active-treatment studies (639 participants) in the Cochrane review reported no aPDT-related adverse effects [26]	Cochrane: moderate certainty for the reported adverse-effects outcome.	No clear short-term safety signal in reported studies; ascertainment was often limited.
Pocket closure/QoL/satisfaction/cost	Cochrane reported no pocket-closure data at 6 months and no satisfaction or quality-of-life data; economic outcomes were rarely evaluated [19, 26]	No umbrella-level certainty label assigned.	Major patient-centred and treatment-decision evidence gap.

Table 5.

Main explanations for discordant conclusions.

Source of discordance	Effect on interpretation
Evolution of the evidence base	Early estimates were based on four or five trials and were unstable; later reviews added trials but many remained small and short-term.
Primary-study overlap	Successive reviews repeatedly synthesized the same early RCTs, so review count overstated independent replication.
Trial design	Split-mouth and parallel-group designs were sometimes pooled without transparent handling of within-person correlation [23]
Protocol heterogeneity	Photosensitizer, wavelength, fluence, pre-irradiation contact time, fibre position and number of sessions differed markedly.
Population heterogeneity	Baseline pocket depth, smoking, diabetes, disease severity and maintenance status varied.
Outcome/time-point selection	Reviews pooled different indices, change scores and follow-up periods, and used both mean and standardized mean differences.
Risk-of-bias interpretation	Some positive conclusions did not adequately reflect trial limitations, imprecision or absence of certainty grading.

Discussion

This umbrella review identified 15 systematic reviews spanning 15 years of evidence synthesis. The direction of effect often favoured adjunctive aPDT for short-term PPD reduction and CAL gain, but absolute improvements expressed in millimetres were generally small. The highest-confidence and most comprehensive review estimated 6-month effects of 0.52 mm for PPD and 0.44 mm for CAL, both with very low certainty and both judged not clinically important [26]. The central inference is therefore not that aPDT lacks biological activity, but that current randomized evidence has not established a reproducible patient-important incremental benefit when aPDT is added to adequately performed nonsurgical periodontal therapy.

The synthesis also clarifies why the literature can appear more favourable than the underlying independent evidence. Multiple reviews repeatedly pooled overlapping small trials, analytical choices differed, and substantial protocol heterogeneity prevents aPDT from being treated as a single standardized intervention. Statistical significance in an average PPD or CAL difference does not establish improved pocket closure, reduced need for surgery, tooth retention, better quality of life or favourable economic value.

The trajectory of review conclusions is instructive. The earliest reviews appropriately described the evidence as insufficient or imprecise [13, 14]. The 2013 and 2018 meta-analyses reported small favourable PPD and CAL estimates [16, 18]. Later reviews that separated trial designs or imposed stricter comparability criteria found smaller or nonsignificant effects [20, 22, 24]. The Cochrane review integrated the largest randomized evidence base and showed that a favourable point estimate can coexist with very low certainty and limited clinical importance [26]. The 2025 meta-analysis again found statistical significance but did not resolve heterogeneity [27]. This pattern supports prioritizing certainty, clinical relevance and trial independence rather than counting positive review conclusions.

Possible effect modification by baseline pocket depth and treatment protocol deserves attention, but it is not yet sufficiently consistent to define a preferred indication. In the AAP best-evidence review, some subgroup analyses suggested larger additional PPD reductions in deeper sites than in moderate pockets, while Pardo et al. found marked protocol-dependent differences across photosensitizer/wavelength subgroups [19, 23]. These signals arise from small, heterogeneous trial subsets and were not converted into a reproducible protocol recommendation by the highest-confidence evidence. Accordingly, the present synthesis does not identify a specific photosensitizer, wavelength, irradiation time or number of sessions as clinically superior.

Behavioural change, effective oral-hygiene instruction, smoking cessation, glycaemic control when relevant, subgingival instrumentation and supportive periodontal care remain the foundation of treatment [6, 7, 52]. The EFP S3 guideline did not support routine adjunctive use of lasers or aPDT because the evidence was heterogeneous and of limited certainty [6, 20]. The present umbrella review supports that position. Current evidence does not justify routine addition of aPDT for all patients with periodontitis, and aPDT should not replace meticulous instrumentation or established risk-factor control.

Clinical use should be framed as an optional adjunct rather than a substitute for established periodontal therapy. When discussing aPDT, clinicians should explain that the average incremental benefit is small, the optimal protocol is unresolved, long-term and patient-centred outcomes are limited, and formal cost-effectiveness evidence is sparse. Contextual reviews excluded from the core dataset because they compared aPDT primarily with systemic antibiotics reported small heterogeneous datasets and do not establish therapeutic equivalence [46, 47]. These comparisons should not position aPDT as a validated substitute when systemic antimicrobials are specifically indicated.

Contextual protocol-specific reviews excluded from the core synthesis further illustrate the difficulty of generalization. Methylene blue, indocyanine green and riboflavin differ in absorption spectra, tissue penetration and potential photothermal contribution [48–51]. A positive finding with one device or photosensitizer cannot be transferred automatically to another. The economic evidence is insufficient to determine cost-effectiveness; future evaluations should measure equipment and consumable costs, chair time, staff training, patient burden and downstream treatment requirements alongside clinical outcomes.

Recent studies of conventional nonsurgical periodontal therapy provide useful biological context for inflammatory outcomes. Standardized NSPT has been associated with reductions in circulating CRP, IL-1β, IL-6 and TNF-α across periodontal stages and grades, and a 2026 randomized secondary analysis in people with type 2 diabetes reported lower hsCRP at 6 months after NSPT than after oral-hygiene instruction alone [53, 54]. These findings support the systemic inflammatory relevance of periodontal treatment but do not establish an aPDT-specific systemic effect.

Future trials should use prospectively registered protocols, concealed parallel-group randomization, adequate sample-size calculations and blinded outcome assessment where feasible. Split-mouth designs should be reserved for questions in which carry-across effects are unlikely and should be analysed using paired methods [38]. Trials should report photosensitizer concentration, pre-irradiation contact time, wavelength, power, fluence, fibre geometry, irradiation duration, number and timing of sessions, operator training and all co-interventions. Current stage and grade terminology should replace isolated use of chronic/aggressive labels [2, 3].

Follow-up should extend to at least 12 months and include the proportion of sites achieving PPD ≤ 4 mm without bleeding, pocket closure, need for surgery, tooth retention, discomfort, satisfaction, oral-health-related quality of life, adverse events and cost-effectiveness [6, 52]. Microbiological studies should prespecify clinically relevant targets, standardize sampling and account for clustering of sites within teeth and patients. Complete numerical change-score data should be reported. Future systematic reviews should quantify primary-study overlap whenever study units can be harmonized and avoid redundant second-order pooling [29, 35, 37]. They should explicitly link conclusions to risk of bias, outcome-level certainty and treatment-relevant endpoints (Table 6) [31, 33, 34].

Table 6.

Minimum design requirements for future trials.

Domain	Recommendation
Design	Adequately powered, multicentre parallel-group RCTs with allocation concealment and assessor blinding.
Population	Stage- and grade-defined periodontitis; prespecified stratification by baseline pocket depth, smoking and diabetes.
Intervention reporting	Photosensitizer, concentration, wavelength, power, fluence, fibre geometry, pre-irradiation contact time, irradiation time and number of sessions.
Comparator	Standardized subgingival instrumentation with sham light/photosensitizer where feasible.
Primary outcomes	Pocket closure and PPD ≤ 4 mm without BOP at patient and site levels.
Secondary outcomes	CAL, need for surgery, tooth retention, adverse events, discomfort, satisfaction, quality of life and cost.
Follow-up	Minimum 12 months, with longer supportive-care observation.
Analysis	Account for clustering and paired data; publish complete change scores and prespecified subgroup analyses.

Strengths and limitations

This umbrella review incorporated the previously omitted Chambrone and Souza reviews [19, 24]. It applied PRIOR alongside PRISMA 2020 and used an explicit fixed hierarchy for interpreting discordant reviews [28, 29]. AMSTAR 2 confidence was recalculated directly from item-level judgements and distinguished from outcome-level GRADE certainty, without inappropriate second-order pooling [31, 33, 34]. The review also distinguished broad effectiveness reviews from reviews restricted to special populations, supportive therapy, alternative comparators or single photosensitizers. The synthesis foregrounded clinical importance, patient-centred outcomes and recurrent use of the same primary evidence.

The review-level search combined PubMed/MEDLINE, the Cochrane Library, Epistemonikos, Google Scholar and bidirectional citation searching rather than stand-alone subscription-platform searching in Embase, Scopus and Web of Science. Screening, extraction and appraisal were completed by one reviewer using prespecified criteria and structured repeat source checks. These features increase susceptibility to retrieval, selection and extraction error compared with fully independent duplicate processes. A global CCA was not computed because the source reviews inconsistently reported trials, publications and treatment comparisons as the unit of evidence; overlap was therefore interpreted from directly identifiable repeated trials, study counts and search chronology. Some reviews also used older chronic/aggressive terminology or mixed active and supportive therapy, requiring extraction of separable subgroups. The clinical conclusions consequently give greatest weight to the highest-confidence review rather than to the number of concordant reviews.

Conclusion

Across 15 systematic reviews, adjunctive aPDT was associated with small short-term improvements in PPD and CAL after nonsurgical periodontal therapy. The estimates were heterogeneous, underlying trials were commonly small or at risk of bias, and recurrent primary-study overlap reduced the independence of review-level findings. The 2024 Cochrane review rated the principal 6-month clinical evidence as very low certainty and judged the apparent PPD and CAL benefits not clinically important. No clear short-term safety signal emerged from reported active-treatment studies, but pocket closure, patient-reported outcomes, long-term effectiveness and economic outcomes remain insufficiently characterized. Current evidence does not support routine addition of aPDT to adequately performed subgingival instrumentation. Standardized and adequately powered trials are required to determine whether a clinically meaningful benefit exists for clearly defined patient groups and protocols.

Ethics approval

Ethical approval was not required for this study because it is an umbrella review based exclusively on publicly available published literature and did not involve direct participation of human participants or animals.

Author contributions

Conceptualization, VM, KSD.; Methodology, VM, SP; Software, KSD; Validation, AM; Formal Analysis, SP; Investigation, VM; Resources, AM; Data Curation, KSD; Writing Original Draft Preparation, VM, SP; Writing—Review And Editing, KSD, AM; Visualization, SP; Supervision, VM; Project Administration, AM; Funding Acquisition: NA. All Authors Have Read And Agreed To The Published Version Of The Manuscript.

Funding

The authors declare that no funds, grants, or other support were received during the preparation or conduct of this study. Open access funding provided by Dr. DY Patil Vidyapeeth, Pune (Deemed to be University).

Data availability

All data generated or analysed during this study are included in this published article and its supplementary materials.

Competing interests

The authors declare no competing interests.

Consent for publication

Not applicable. This manuscript does not contain any identifiable individual participant data or images.

Informed consent to participate

Not applicable, as this study did not involve human participants.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41405-026-00503-3.