Diagnostic accuracy of the lactate stress test for detecting mitochondrial disorders: Systematic review and meta-analysis

Abstract

Due to their variable phenotypes, mitochondrial disorders (MDs) can be difficult to diagnose. The absolute load lactate stress test (LSTA) and the relative load lactate stress test (LSTR) have been shown to be useful screening tools for the detection of MDs. In this study, we aimed to perform a meta-analysis to evaluate the diagnostic accuracy of these tests in detecting MDs. The study protocol was registered with PROSPERO (no. CRD42022331710). We performed a comprehensive search of PubMed, Web of Science and Scopus from January 10th, 2022 to July 27th, 2022 and included case-control and cohort diagnostic studies that targeted participants with MDs and used LSTA and/or LSTR as index tests. Two reviewers worked separately to compile information from selected articles. Risk of bias and applicability were assessed using the QUADAS-2 tool. Sensitivity and specificity, as well as diagnostic odds ratios (DORs) and area under the curve (AUC) were calculated using Meta-DiSc 2.0 and Stata software. The analysis included 14 studies with a total of 1064 participants, divided into six studies with 793 participants for LSTA and eight studies with 271 participants for LSTR. For LSTA the meta-analysis gave a pooled sensitivity of 0.67 (95 % CI 0.62, 0.72), a specificity of 0.93 (95 % CI 0.85, 0.97), DOR of 26.63 (95 % CI 10.99, 64.52), and AUC of 0.70 (95 % CI 0.66, 0.74). For LSTR, the pooled sensitivity was 0.52 (95 % CI 0.33, 0.70), specificity 0.94 (95 % CI 0.79, 0.99), DOR 18.14 (95 % CI 2.99, 109.85), and the AUC 0.80 (95 % CI 0.76, 0.83). LSTA and LSTR showed as screening tests moderate sensitivity and high specificity for MD diagnosis, particularly for LSTR. The choice of test may depend on the patient's individual aerobic capacity and motor skills and the availability of equipment.

Keywords: Absolute load lactate stress test; Diagnostic accuracy; Meta-analysis; Mitochondrial disorders; Relative load lactate stress test.

Fig. 1

Flow diagram of study selection process. n, number of studies.

Fig. 2

Risk of bias and applicability concerns summary of included studies using the QUADAS-2 tool. QUADAS-2, Quality Assessment of Diagnostic Accuracy Studies.

Fig. 3

Risk of bias and applicability concerns graph of included studies using the QUADAS-2 tool.

Fig. 4

Forest plot of sensitivity of individual studies for LSTA (A) and LSTR (B). TP, true positive; FP, false positive; FN, false negative; TN, true negative; 95 % CI, 95 % confidence interval.

Fig. 5

Forest plot of specificity of individual studies for LSTA (A) and LSTR (B). TP, true positive; FP, false positive; FN, false negative; TN, true negative; 95 % CI, 95 % confidence interval.

Fig. 6

Sensitivity, specificity, and summary receiver operating characteristic (SROC) curves LSTA (A) and LSTR (B) in the diagnosis of MDs. A, LSTA. 1: study by Petty RKH et al.; 2: study by Dengler R et al.; 3: study by Finsterer J and Milvay E. (2002); 4: study by Finsterer J and Milvay E. (2004); 5: study by Hanisch F et al.; 6: study by Kurihara M et al. B, LSTR. 1: study by Nashef L and Lane RJM; 2: study by Dandurand et al.; 3: study by Taivassalo T et al. (2001); 4: study by Jensen T et al.; 5: study by Jeppesen TD et al.; 6: study by Taivassalo T et al. (2003); 7: study by Hammarén E et al.; 8: study by Løkken N et al. AUC, area under the curve.

Fig. 7

Sensitivity, specificity and summary receiver operating characteristic (SROC) curves of overall LST for MDs 1: study by Petty RKH et al.; 2: study by Nashef L and Lane RJM; 3: study by Dandurand et al.; 4: study by Dengler R et al.; 5: study by Taivassalo T et al. (2001); 6: study by Finsterer J and Milvay E. (2002); 7: study by Jensen T et al.; 8: study by Jeppesen TD et al.; 9: study by Taivassalo T et al. (2003); 10: study by Finsterer J and Milvay E. (2004); 11: study by Hammarén E et al.; 12: study by Hanisch F et al.; 13: study by Løkken N et al.; 14: study by Kurihara M et al.

Fig. 8

Overall LST Deeks' funnel plot for detecting publication bias. 1: study by Petty RKH et al.; 2: study by Nashef L and Lane RJM; 3: study by Dandurand et al.; 4: study by Dengler R et al.; 5: study by Taivassalo T et al. (2001); 6: study by Finsterer J and Milvay E. (2002); 7: study by Jensen T et al.; 8: study by Jeppesen TD et al.; 9: study by Taivassalo T et al. (2003); 10: study by Finsterer J and Milvay E. (2004); 11: study by Hammarén E et al.; 12: study by Hanisch F et al.; 13: study by Løkken N et al.; 14: study by Kurihara M et al.

Fig. 9

LSTA Deeks' funnel plot for detecting publication bias. 1: study by Petty RKH et al.; 2: study by Dengler R et al.; 3: study by Finsterer J and Milvay E. (2002); 4: study by Finsterer J and Milvay E. (2004); 5: study by Hanisch F et al.; 6: study by Kurihara M et al.

Fig. 10

LSTR Deeks' funnel plot for detecting publication bias. 1: study by Nashef L and Lane RJM; 2: study by Dandurand et al.; 3: study by Taivassalo T et al. (2001); 4: study by Jensen T et al.; 5: study by Jeppesen TD et al.; 6: study by Taivassalo T et al. (2003); 7: study by Hammarén E et al.; 8: study by Løkken N et al. AUC, area under the curve.