Metagenomic next-generation sequencing in patients with fever of unknown origin: A comprehensive systematic literature review and meta-analysis

Metagenomic Next-generation Sequencing in Patients With Infectious Meningoencephalitis: A Comprehensive Systematic Literature Review and Meta-analysis

Background

We aimed to assess the accuracy, clinical efficacy, and limitations of metagenomic next-generation sequencing (mNGS) for diagnosing infectious meningoencephalitis.

Methods

We performed a systematic literature review and meta-analysis of studies that evaluated the performance of mNGS to determine the cause of infectious meningoencephalitis. We explored PubMed, Cumulative Index to Nursing and Allied Health, Embase, Cochrane Central Register of Controlled Trials, ClinicalTrials.gov, and Web of Science up to 12 November 2024. To perform a meta-analysis, we calculated the pooled diagnostic odds ratio (DOR) for mNGS and for conventional microbiological tests (CMTs) compared to the clinical diagnosis.

Results

Thirty-four studies met the inclusion criteria, with mNGS-positive rates ranging from 43.5% to 93.5% for infectious meningoencephalitis. The meta-analysis included 23 studies with 1660 patients. The pooled sensitivity was 0.70 (95% confidence interval [CI], .67-.72), and its specificity was 0.93 (95% CI, .92-.94). The DOR for mNGS was 26.7 (95% CI, 10.4-68.8), compared to 12.2 (95% CI, 3.2-47.0) for CMTs. For tuberculosis meningoencephalitis, mNGS demonstrated a pooled sensitivity of 0.67 (95% CI, .61-.72) and specificity of 0.97 (95% CI, .95-.99), with a DOR of 43.5 (95% CI, 7.4-256.6).

Conclusions

Our review indicates that mNGS can be a valuable diagnostic tool for infectious meningoencephalitis, offering high sensitivity and specificity. mNGS's superior DOR compared to that of CMTs highlights its potential for more accurate diagnoses and targeted interventions. Further research is needed to optimize which patients and at what point in the diagnostic process mNGS should be used.

Enhancing fever of unknown origin diagnosis: machine learning approaches to predict metagenomic next-generation sequencing positivity

Objective

Metagenomic next-generation sequencing (mNGS) can potentially detect various pathogenic microorganisms without bias to improve the diagnostic rate of fever of unknown origin (FUO), but there are no effective methods to predict mNGS-positive results. This study aimed to develop an interpretable machine learning algorithm for the effective prediction of mNGS results in patients with FUO.

Methods

A clinical dataset from a large medical institution was used to develop and compare the performance of several predictive models, namely eXtreme Gradient Boosting (XGBoost), Light Gradient-Boosting Machine (LightGBM), and Random Forest, and the Shapley additive explanation (SHAP) method was employed to interpret and analyze the results.

Results

The mNGS-positive rate among 284 patients with FUO reached 64.1%. Overall, the LightGBM-based model exhibited the best comprehensive predictive performance, with areas under the curve of 0.84 and 0.93 for the training and validation sets, respectively. Using the SHAP method, the five most important factors for predicting mNGS-positive results were albumin, procalcitonin, blood culture, disease type, and sample type.

Conclusion

The validated LightGBM-based predictive model could have practical clinical value in enhancing the application of mNGS in the etiological diagnosis of FUO, representing a powerful tool to optimize the timing of mNGS.

Application of metagenomic next-generation sequencing in the diagnosis of pathogens in patients with diabetes complicated by community-acquired pneumonia

To explore the clinical utility and optimal timing of metagenomic next-generation sequencing (mNGS) in diagnosing pathogens in patients with diabetes complicated by community-acquired pneumonia (CAP). The study included 50 hospitalized patients diagnosed with diabetes complicated by CAP who underwent conventional microbiological testing (CMT) and mNGS using bronchoalveolar lavage fluid. Among the 50 cases, 16% presented no respiratory symptoms. There were significant increases in inflammatory markers such as C-reactive protein, erythrocyte sedimentation rate, and interleukin-6, with patchy imaging changes being the most prevalent. The positive rates for pathogen detection by mNGS and CMTs were 78 and 21% (P < 0.05). The mNGS was significantly better than the CMTs in the detection of rare pathogens such as Anaerobes, Chlamydia psittaci, Legionella pneumophila, Mycobacterium bovis, Aspergillus fumigatus, and Pneumocystis japonicus (P < 0.05). After clinical interpretation, 85% (22/26) of viruses, 24% (9/37) of bacteria, and 25% (2/8) of fungi were non-pathogen organisms by mNGS. There was a significant difference in the rates of adjustment in anti-infection treatment strategies based on the pathogen detection results from CMTs and mNGS, which were 2 and 46%, respectively (P < 0.05). We found that mNGS was superior to CMTs in terms of the positive rate of pathogen detection, detecting mixed infection incidence, rare pathogen detection rates, and the adjustment of treatment strategies. However, mNGS results need to be interpreted in the context of the clinic.