Shotgun metagenomic sequencing improves cross‐kingdom diagnosis of mycetoma

Mycetoma is a chronic cutaneous and subcutaneous infection caused by bacteria (actinomycetoma) or fungi (eumycetoma).1 Delayed or inaccurate microbiological identification may lead to prolonged empirical therapy, unnecessary surgery or inappropriate treatment. Conventional diagnosis based on culture and targeted molecular assays is limited by slow growth, commensal overgrowth and restricted taxonomic coverage, given the 70 implicated species.2, 3 Following the recognition of mycetoma as a neglected tropical disease1 and the inclusion of mycetoma-associated fungi in the WHO priority pathogens list,4 improved cross-kingdom diagnostic strategies are needed. We evaluated untargeted shotgun metagenomic sequencing (SMg)5, 6 against standard-of-care (SoC) methods as part of a combined diagnostic strategy for suspected mycetoma, aiming to improve microbiological confirmation and guide treatment decisions. Nineteen samples (6 grains, 8 fresh tissues, 5 formalin-fixed paraffin-embedded tissues FFPE) from 10 patients with suspected mycetoma were retrospectively analysed (Figure 1a). SoC included grain colour, direct microscopy, histopathology, culture and Sanger DNA barcoding. Overall, SoC identified a causative pathogen in 6/19 samples (31.6%), with a significantly higher diagnostic yield in grains than in tissues (83.3% vs. 7.7%, p = 0.003) (Table 1). Culture contributed to only 2/14 fresh samples, both eumycetomas. ITS2 sequencing confirmed these two but frequently detected non-causative fungi, while 16S rRNA sequencing confirmed Actinomadura madurae in 4/9 actinomycetoma-classified samples. SMg identified a mycetoma-causative taxon in 9/19 samples (47.4%), with an overall concordance of 84.2% (16/19) (Figure 1b). All SoC-positive cases were also positive by SMg for the same species (A. madurae, Madurella mycetomatis, Fusarium solani species complex). Among the 13 SoC-negative samples, SMg identified additional putative causative fungi in 3 cases (23.1%): M. mycetomatis (grain), M. mycetomatis with Aspergillus fumigatus (FFPE) and Fusarium sp. (fresh tissue). Commensal fungi were detected in 13/19 samples (68.4%) and non-mycetoma bacteria in 7/19 (36.8%). In three cases with suppurative inflammation, SMg identified additional putative pathogens in two (Figure 1c). These findings highlight the ability of SMg to detect pathogens missed by conventional methods, particularly in samples with low microbial burden, and to reveal possible co-infections, as illustrated by A. fumigatus and Fusarium sp., the latter identified in a patient with confirmed actinomycetoma. Diagnostic performance was strongly influenced by sample type, with grains providing the highest yield. SMg detected a causative pathogen in all grains (6/6) but in only 3/13 tissues (p = 0.003). Reads per million (RPM) values reflected the microbial burden, with a median RPM of 16.3 in grains versus 1.4 in tissues. FFPE samples showed the lowest yield, consistent with DNA fragmentation and paraffin embedding increasing background contamination.7 Among the 10 samples negative by both SoC and SMg, most were tissue biopsies with histopathological features of inactive disease or possible sampling limitations. Overall, SMg showed strong agreement with SoC and increased diagnostic yield. Although NGS has previously been applied to mycetoma diagnosis using targeted approaches that improved fungal detection in black grains,8 these methods remain restricted to a single kingdom. In contrast, SMg enables simultaneous detection of bacterial and fungal mycetoma agents from a single assay, providing comprehensive cross-kingdom identification. Beyond causative pathogens, SMg also identified skin-associated microbiota in most samples and additional opportunistic bacteria in suppurative lesions, consistent with microbial communities previously reported in eumycetoma grains9 and other chronic cutaneous infections.10 Although limited by a small sample size and retrospective design, our findings support a pragmatic diagnostic strategy: grains should be prioritized whenever possible, and SMg should be considered when SoC is inconclusive, particularly in tissue or FFPE samples. Prospective studies are needed to define sensitivity, specificity and cost-effectiveness, especially in endemic settings where access to advanced diagnostics remains limited. Declaration of Generative AI and AI-assisted technologies in the writing process: All content was written by the authors. Spelling corrections, English grammar and turns of phrase were revised with the help of ChatGPT-5 (OpenAI) and DeepL translator. The author(s) take full responsibility for the final content of the publication. The authors have nothing to report. The authors declare no conflict of interest. This retrospective non-interventional study used anonymized residual samples collected as part of routine diagnostic care and did not require formal ethics committee approval according to national regulations. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki. All samples were collected as leftovers from routine clinical testing in the Parasitology–Mycology department, and no additional procedures or interventions were performed for research purposes. The raw sequencing data underlying this article cannot be shared publicly for the privacy of individuals who participated in the study. The microbial sequencing data will be shared on a reasonable request to the corresponding author.