Accuracy of rapid point-of-care serological tests for leprosy diagnosis: a systematic review and meta-analysis

Romero, Carmen Phang; Castro, Rodolfo; do Brasil, Pedro Emmanuel A; Pereira, Daniella R; Pinheiro, Roberta Olmo; Toscano, Cristiana M; de Oliveira, Maria Regina Fernandes

doi:10.1590/0074-02760220317

BACKGROUND

Leprosy is a chronic infectious disease, still endemic in many countries that may lead to neurological, ophthalmic, and motor sequelae if not treated early. Access to timely diagnosis and multidrug therapy (MDT) remains a crucial element in the World Health Organization’s strategy to eliminate the disease as a public health problem.

OBJECTIVES

This systematic review aims to evaluate the accuracy of rapid point-of-care (POC) tests for diagnosis of leprosy.

METHODS

Searches were carried out in electronic databases (PubMed, EMBASE, CRD, Cochrane Library and LILACS) in April 2021 for patients with suspicion or confirmatory diagnostic of leprosy, classified in multibacillary (MB) or paucibacillary (PB) cases, performing rapid POC serological tests compared to clinical evaluation, smear microscopy and immunohistochemistry analysis. Methodological quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies tool (QUADAS-2). A meta-analysis was undertaken to generate pooled estimates of diagnostic parameters, presenting sensitivity, specificity and diagnostic odds ratio (DOR) values. The review protocol was registered at PROSPERO, CRD # 42014009658.

FINDINGS

From 893 potentially relevant references, 12 articles were included reporting 16 diagnostic tests accuracy studies with 5395 individuals enrolled. Meta-analysis of NDO-LID and PGL-I tests data in MB patients showed sensitivity and specificity [95% confidence interval (CI)] of 0.83 (0.71-0.91), 0.91 (0.72-0.97); and 0.92 (0.86-0.96), 0.93 (0.78-0.98); respectively, with high heterogeneity among the studies.

MAIN CONCLUSIONS

Our results can inform policymakers regarding the possibility of implementing accurate, rapid POC tests for leprosy in public health services, especially within primary health care.

Key words:
accuracy; diagnosis; leprosy; Mycobacterium infections; point-of-care; systematic review

[1] + Corresponding author: carmenprc@gmail.com / carmen.romero@cdts.fiocruz.br

	Inclusion	Exclusion
Population	Patients with clinical diagnosis of leprosy Only endemic controls were considered.	Animal model testing. Patients with comorbidities: human immunodeficiency virus (HIV), tuberculosis.
Intervention	Point-of-care (POC) serological tests (defined as the index test).	Laboratory serological tests for research purposes enzyme-linked immunosorbent assay (ELISA), polymerase chain reaction (PCR). Cytological, histological or immuno-histopathological diagnostic studies.
Comparator	Conventional diagnosis: based on clinical signs and symptoms; skin lesion compatible with leprosy and with permanent loss of thermal, painful and/or tactile sensitivity, with or without thickened nerves. A positive smear of an intradermal smear is considered a confirmatory test; however, a negative result does not exclude the diagnosis of leprosy. Ridley and Jopling (R&J) classification for diagnosis and classification	Only classification of cases by operational classification (CO) based on the count of skin lesions, typifying patients with less than five lesions as paucibacillary (PB) and with five or more lesions as multibacillary (MB).
Outcomes	Primary: sensitivity (Se), specificity (Sp), positive likelihood ratio (LR+) and negative likelihood ratio (LR-). Clinical: early detection of leprosy.	No information on the number of patients, positive and negative test results, primary outcomes: sensitivity, specificity; neither information that could allow their calculation.
Study design	Primary studies of diagnostic accuracy, including those with reliable abstracts, tables and additional information that allow extracting data from the study (2x2t able).	Technology development studies. Case report.

Characteristics	Number of studies n (%)
Manufacturer
Orange Life (Rio de Janeiro) Brazil	5 (31.25)
CTK Biotech (San Diego CA)	2 (12.5)
KIT (Biomedical Research) The Netherlands	2 (12.5)
Standard Diagnostics (Yongin South Korea)	1 (6.25)
InBios International Inc. Seattle WA USA	1 (6.25)
Yonsei University and Standard Diagnostics Inc (Suweon Republic of Korea)	1 (6.25)
OmegaTeknika Limited (Dublin Ireland) and Royal Tropical Institute (KIT Amsterdam Netherlands)	1 (6.25)
Amrad ICT Diagnostics (Brookvale) NSW	1 (6.25)
UPCON; Labrox, Finland	1 (6.25)
Instituto de Patologia Tropical e Saúde Pública (IPTSP), Universidade Federal de Goiás	1 (6.25)
Antigen - Index test
Phenolic glycolipid-I (PGL-I)	5 (31.25)
NDO-LID conjugate	9 (56.25)
PGL-I + (IP-10, CCL4 and CRP)	1 (6.25)
35-kD test	1 (6.25)
Publication year
2019	2 (12.5)
2018	2 (12.5)
2017	2 (12.5)
2016	2 (12.5)
2014	3 (18.75)
2012	2 (12.5)
2008	1 (6.25)
2003	1 (6.25)
1999	1 (6.25)
Multicentric investigation
Yes	6 (37.5)
No	10 (62.5)
Study setting
Primary care	2 (12.5)
Secondary care	7 (43.75)
Tertiary care	7 (43.75)
Country were study conducted^*
Brazil^*	8 (50.0)
Bangladesh^*	1 (6.25)
Colômbia	1 (6.25)
China^*	1 (6.25)
Ethiopia^*	1 (6.25)
Ghana^*	1 (6.25)
India	1 (6.25)
Indonesia^*	1 (6.25)
Nepal^*	3 (18.75)
Philippines^*	5 (31.25)

Study, year	Cohort	Index test	Biological sample	Standard reference	N° participants enrolled	N° participants included in analysis^*	% MB
Leturiondo AL, 2019	Brazil	NDO-LID conjugate	Serum	CO + Bacilloscopy + R&J	780	701	70.76%
Leturiondo AL, 2019	Brazil	Phenolic glycolipid-I (PGL-I)	Serum	CO + Bacilloscopy + R&J	701	70.76%
Góis RS, 2018	Brazil	NDO-LID conjugate	Whole blood	CO + Bacilloscopy	140	125	31.43%
Hooij AV, 2018	Brazil, China, Ethiopia	PGL-I + (IP-10, CCL4, CRP)	Whole blood	CO + R&J	715	467	63.00%
Frade MAC, 2017	Brazil	NDO-LID + Smart Reader	Whole blood	CO + R&J	434	288	88.37%
Hooij AV, 2017	Philippines, Bangladesh	NDO-LID conjugate	Serum	CO + R&J + BI	434	434	74.82%
Duthie MS, 2016¹	Philippines	NDO-LID conjugate¹	Whole blood	CO + Bacilloscopy + R&J	635	102	NA
Duthie MS, 2016²	Philippines	NDO-LID conjugate²	Whole blood	CO + Bacilloscopy + R&J	102	NA
Duthie MS, 2014¹	Philippines	NDO-LID conjugate¹	Serum	CO + Bacilloscopy + R&J	384	333	77.04%
Duthie MS, 2014³	Philippines	NDO-LID conjugate³	Serum	CO + Bacilloscopy + R&J	333	77.04%
Duthie MS, 2014	Colombia, Philippines	NDO-LID conjugate	Whole blood	CO + Bacilloscopy + R&J	388	299	74.49%
Stefani M, 2012⁴	Brazil, Nepal	Phenolic glycolipid-I (PGL-I)⁴	Both (Serum+WB)	CO + Bacilloscopy + R&J	363	239	50.00%
Stefani M, 2012⁵	Brazil, Nepal	Phenolic glycolipid-I (PGL-I)⁵	Both (Serum+WB)	CO + Bacilloscopy + R&J	363	139	49.64%
Parkash O, 2008	India	Phenolic glycolipid-I (PGL-I)	Whole blood	CO + Bacilloscopy	209	172	17.01%
Bührer-Sékula S, 2003	Brazil, Indonesia, Philippines, Ghana	Phenolic glycolipid-I (PGL-I)	Both (Serum+WB)	CO	739	433	57.29%
Roche P, 1999	Nepal	35-kD test card^*	Serum	R&J	174	97	55.56%