Oral microbiome and lesion-associated microbiological findings in medication-related osteonecrosis of the jaw: a systematic review and evidence map.

BACKGROUND: Medication-related osteonecrosis of the jaw (MRONJ) may involve microbial colonization, anaerobic organisms, and biofilm formation, but whether it has a reproducible microbial signature remains unclear. We mapped the human microbiological evidence in MRONJ. METHODS: The protocol was registered in PROSPERO (CRD420261432571). Reporting followed PRISMA 2020, PRISMA-S, and Synthesis Without Meta-analysis guidance. Six sources were searched from inception in two waves through 25 July 2026, supplemented by citation searching. Original human microbiological studies of MRONJ were eligible. Community-level evidence was classified as core, other MRONJ microbiological evidence as supportive, and evidence from related jawbone conditions as contextual. Five prespecified domains included explicit MRONJ-specific Actinomyces events and denominators, strict-anaerobe evidence, and direct biofilm evidence. Design-specific JBI checklists and STORMS-informed reporting items were applied. Meta-analysis was conditional on prespecified compatibility criteria. RESULTS: Of 4,044 records, 71 reports representing 67 independent studies were included: 51 MRONJ studies, comprising 13 core and 38 supportive studies, and 16 contextual studies. Community-level findings were not sufficiently comparable for pooling. Among the 51 MRONJ studies, criteria were met by 19 for strict-anaerobe evidence, seven for viral evidence, six for explicit Actinomyces events and denominators, three for direct biofilm evidence, and two for fungal evidence; 32 contributed to at least one domain. Nine method-specific Actinomyces observations from six studies reported raw detection proportions of 5.1%-65.7% by culture, 82.9%-96.4% by PCR or quantitative PCR, and 36.4%-100.0% by histology, using patient- or specimen-level denominators. Diagnostic criteria, antibiotic timing, negative controls, and data availability were inconsistently reported. No domain met the compatibility criteria for meta-analysis. CONCLUSIONS: The MRONJ microbiological evidence is methodologically heterogeneous and dominated by observational, lesion-derived studies. Detection proportions varied substantially with method and denominator unit, limiting their interpretation as prevalence estimates. Repeated reports of anaerobic organisms and Actinomyces, together with sparse direct biofilm evidence, describe lesion-associated findings of undetermined specificity rather than a reproducible MRONJ-specific signature. Standardized comparators and sampling, explicit denominators, contamination controls, extractable numerical results, and accessible sequence data are needed for future synthesis and clinical evaluation. SYSTEMATIC REVIEW REGISTRATION: https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD420261432571, identifier CRD420261432571.