The present cross-sectional study was conducted on Mosqueiro Island, a district belonging to Belém, the capital of Pará state, Brazilian Amazon, which has a total population of approximately 30,000 inhabitants. The field studies were performed during a 1-week period from May 12 to 16, 2014. SC was randomly selected from four public schools (three elementary and one middle school). Newly diagnosed cases in children were followed up by visits to the households to examine HC. Blood was obtained from all SC and HC to assess antibody titers to the three different antigens. Earlobe SSSs were obtained from 79/105 new cases from the field to perform RLEP qPCR. As a control group, leprosy patients newly diagnosed at Dr. Marcello Candia Reference Unit in Sanitary Dermatology of the State of Pará (URE) located in Marituba, Pará, and their corresponding HC were examined, and blood samples were obtained.

This study conforms to the Declaration of Helsinki and was approved by the Institute of Health Sciences Research Ethics Committee from Pará Federal University (CAAE 26765414.0.0000.0018 CEP-ICS/UFPA). All individuals involved agreed voluntarily to participate and signed an informed consent form after receiving information about the study. Parents of minors or responsible adults signed their consent, allowing them to participate. All data analyzed were anonymized to protect the privacy of participants.

All subjects were examined and evaluated by experienced leprologists. The diagnosis of leprosy was based solely on well-accepted clinical signs and symptoms, including the detection of hypopigmented or infiltrated skin lesions with loss of sensation assessed with standard graded Semmes–Weinstein monofilaments (20) and the palpation of peripheral nerves to identify characteristic pain associated with inflammation or swelling. Each newly diagnosed patient was classified into one of the five clinical forms defined by Ridley–Jopling, except in cases where there was only a single hypopigmented macule without nerve involvement, indicating an early indeterminate form of leprosy. Indeterminate (I) and tuberculoid-tuberculoid (TT) forms were operationally categorized as PB, while all other forms (borderline-tuberculoid [BT], borderline-borderline [BB], BL, and LL) were grouped together as MB. Primary neural leprosy (PNL) was diagnosed if there was nerve enlargement associated with functional or sensory loss without any skin lesions. If only one nerve was affected, the case was classified as PB, whereas two or more enlarged nerves indicated MB. The DG was ranked from 0 to 2 (DG0, no disability; DG1, loss of sensation in the hand or foot; DG2, visible damage, deformity or disability) as determined during a clinical examination of sensory-motor functions based on a WHO standardized neurological evaluation (21).

All participants had blood samples collected in EDTA tubes after clinical evaluation. For antibodies titration, high-affinity polystyrene Immulon IV 96-well ELISA plates (Dynex Technologies, Chantilly, VA, USA) were coated overnight with 50 ng/50 µl of carbonate/bicarbonate sodium buffer with NDO-BSA, LID-1, or NDO-LID. The plates were blocked with blocking solution (phosphate-buffered saline [PBS], pH 7.2 with 1% bovine serum albumin [BSA], and 0.05% Tween-20 [T]) for 1 h. Serum samples were diluted 1:300 with a blocking solution and tested in duplicate for 2 h of incubation. The secondary anti-human IgM (Sigma-Aldrich, St. Louis, MO, USA, A0420) antibodies were used for NDO-BSA (dilution 1:10,000), while secondary anti-human IgG (Rockland, Gilbertsville, PA, USA, #609-1316 2 mg) were used for LID-1 (dilution 1:20,000), and both antibodies were used for NDO-LID. The reaction was developed by the insertion of an antibody conjugated with peroxidase in PBS/T for 1.5 h, and then the peroxidase color substrate o-phenylenediamine dihydrochloride (SIGMAFAST™ OPD, Sigma-Aldrich P9187, USA) was added to each well for 15 min. To stop the reaction, 50 µl/well of H₂SO₄ at 4 N was added and immediately read at 490 nm using an MRX Revelation 4.25 microplate reader (Dynex Technologies, Chantilly, VA, USA).

To control intra- and inter-test variations, positive (LL patient) and negative (non-endemic control [NEC]) controls were added to each plate. All samples were tested in duplicate plus a blank control well coated only with BSA.

Cutoff values were established using the mean of results plus 3 times SD of optical density (OD) of health endemic and non-endemic subjects. Results were considered from the mean of the duplicates less the blank value. Cutoff for anti-NDO-BSA and anti-LID-1 was 0.295 and for anti-NDO-LID was 0.475.

A wide range of molecular targets were evaluated for the detection of M. leprae (7); one of them is an RLEP family, which, despite a variable structure and unknown function, has 29 repetitions scattered on the M. leprae chromosome; for this reason, the RLEP region is easier to amplify, expanding the possibility of detection even with small amounts of sample (22). SSS samples of both earlobes were collected, immediately fixed in 1 ml of 70% ethanol, and stored for up to 30 days at room temperature before extraction.

DNA extraction was performed following the manufacturer’s recommendations (Qiagen DNeasy Blood and Tissue Kit, Germantown, MD, USA). RLEP detection was established with 10 ng of total DNA using 5′-GTGAGGGTAGTTGTT-3′ (LP1) and 5′-GGTGCGAATAGTT-3′ (LP2) primers (IDT, Integrated DNA Technologies, IA, USA) amplifying a 129-bp fragment detected by SYBR Green fluorescent DNA-binding dye using Applied Biosystems 7500 Real-Time PCR Systems. Each reaction was conducted in a total of 20 µl (4 μl of DNA template, 5 μl of water, 10 μl of SYBR, and 1 μl of primers). The reaction occurred with the following specifications: Uracil-DNA glycosylase (UDG) at 50°C for 2 min, prior 95°C for 2 min for initial denaturation followed by 45 cycles, each cycle consisting of denaturation at 95°C for 15 s, annealing at 58°C for 15 s and extension at 72°C for 1 min, and melting curve running in each experiment. To determine qPCR positivity, a standard curve was prepared from purified M. leprae DNA extracted from nude mouse footpads, and then five points of standard dilution (1:50, 1:100, 1:1,000, 1:2,000, and 1:5,000) of this DNA were included in each plate, plus three negative points as a template control (see Supplementary Figure 1 and Supplementary Table 1 ).

A melt curve was used to analyze the specificity of the amplified products. Samples were considered positive when the fluorescent signal intersected the automatically calculated threshold line. Amplifications with a cycle threshold (Ct) ≤ 45 were considered positive for RLEP. Pearson’s coefficient (r²) is calculated for each plate, and the plate results only were accepted if r² ≥ 0.95. We perform qPCR experiments according to the MIQE checklist (see Supplementary Table 2 ).

A two-tailed Mann–Whitney U test was used for the comparison of antibody titers between two groups. All results were considered statistically significant when p < 0.05. To verify the antigen diagnostic performances, we calculated the area under the curve (AUC) by receiver operating characteristic (ROC) curve. The results above 0.7 were considered able to discriminate HC from patients.

This study was performed in accordance with the recommendations of the Brazilian National Ethics Committee (CONEP) guidelines approved by Pará Federal University Ethics Committee number CAAE 26765414.0.0000.0018 with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki. The protocol was approved by the Pará Federal University Ethics Committee.

A total of 894 individuals from Mosqueiro Island were surveyed, 706 SC were examined for clinical signs/symptoms of leprosy ( Figure 1 ), and 65 (9.2%) were SC diagnosed as a new case of leprosy (SC_NC), presenting mainly hypochromic skin lesions with loss of sensation, with the number of lesions ranging from 0 to 9 (median 2). Subsequently, 188 HC of SC_NC were evaluated at their homes during follow-up visits (range 1–9 and median of 2 HC per household), and 40 (21.2%) HCs were diagnosed as a new case of leprosy (HC_NC). HC_NC showed a median of three (range 0–21) skin lesions, and nine (22.5%) presented with disability grade 1 (DG1) ( Table 1 ). In summary, active case finding diagnosed leprosy in 105 subjects (105/894, 11.7%), with 65 SC and 40 HC. Although 19 (18.1%) presented with DG1, there was no case with disability grade 2 (DG2).

NDO-BSA, LID-1, and NDO-LID were quantified by ELISA for each group (SC, SC_NC, HC, and HC_NC), and seropositivity was characterized based on the cutoff defined for each antigen. There was no significant difference between the SC median versus the SC_NC median for any of the antigens (p-value: 0.254, 0.955, and 0.282 for NDO-BSA, LID-1, and NDO-LID, respectively), with an equivalent OD range within each group. HC and HC_NC also did not show differences for the three antigens tested (p-value: 0.968, 0.480, and 0.262 for NDO-BSA, LID-1, and NDO-LID, respectively). The seropositive percentages were slightly higher for HC than SC for NDO-BSA (70.9% and 68.8%, respectively) and NDO-LID (29.5% and 20.3%, respectively). Similarly, the seropositivity profile of newly diagnosed cases was the highest for NDO-BSA (73.3%), followed by NDO-LID (33.3%) and LID-1 (5.7%).

To ratify the clinical diagnosis, amplification of the RLEP region using qPCR was performed on 79/105 new cases diagnosed in the field. Out of the 49 SC_NC, 43 (87.7%) were found to be positive, while of the 30 HC_NC, 25 (83.3%) were positive. In total, 68/79 (86.07%) new cases were found to be positive, with a Ct median of 41.1 cycles ( Figure 2A ).

In comparison to the field diagnosis from Mosqueiro, we randomly selected 25 patients (24 MB and 1 PB) who were diagnosed at Dr. Marcello Candia Reference Unit in Sanitary Dermatology of the State of Pará (URE). These patients, new cases from URE (URE_NC), come from the surrounding region, referred by other basic health units or physicians for confirmation of suspected leprosy. Often, these individuals exhibit more advanced disease due to delays in diagnosis or misdiagnosis (23), which can manifest itself with an increased bacillary index, additional nerve damage, and increased levels of disability. Along with these 25 patients from URE, 54 of their HCs (HC_URE) were also enrolled. Upon examination, five of these (9.2%) were found to have clinical symptoms of leprosy, and these HCs of cases from URE were diagnosed as a new case (HC_URE-NC).

The antibody titers against the three antigens were determined for all PB (n = 17) and MB (n = 88) cases from Mosqueiro (SC_NC and HC_NC) versus PB (n = 1) and MB (n = 24) cases from URE. All cases from Mosqueiro demonstrated antibody titers significantly lower than those from URE cases, especially for anti-LID-1 and anti-NDO-LID antibodies ( Figure 2B ). When the titers were analyzed based on clinical form, there was a clear increase in the mean OD as subjects progressed from TT to LL, with 100% of BL (n = 5) and LL (n = 5) patients being positive against all three antigens ( Figure 3 ).

Comparing exclusively subjects without leprosy from Mosqueiro and URE (SC, HC, and HC_URE), we observed that only anti-NDO-BSA demonstrated different levels of antibody titers between groups, with a high median on HC from Mosqueiro, followed by SC and HC_URE from URE ( Figure 4 ).

To evaluate the validity of NDO-BSA, LID-1, and NDO-LID as diagnostic tools, parameters of sensitivity, specificity, and accuracy were calculated for both classes of leprosy patients. No biomarker detected all cases of leprosy, independent of the disease progression at the time of diagnosis. The ability to identify leprosy cases through serology (sensitivity) was higher for NDO-BSA, followed by NDO-LID and LID-1. Anti-NDO-BSA antibodies presented the highest sensitivity either in the field or at URE; however, specificity and accuracy significantly decreased for early leprosy cases diagnosed in the field, limiting anti-NDO-BSA applicability as a diagnostic biomarker for cases in the initial stages. Anti-NDO-LID displayed equivalence to anti-NDO-BSA in all percentages for well-defined patients, while for field subjects, the sensitivity was notably lower. On the other hand, anti-LID-1 was the most effective for differentiating subjects with or without leprosy (specificity), even for cases diagnosed with initial clinical manifestations, showing 100% specificity for URE cases and 95% specificity for Mosqueiro cases. Unfortunately, anti-LID-1 antibodies showed very limited sensitivity, particularly for early diagnosis ( Table 2 ).

Analyzing the correlation of laboratory test parameters by ROC curve, we verified a large difference in ROC curve power comparing well-defined URE cases or oligosymptomatic field cases. URE subjects presented a high AUC for all biomarkers, indicating the applicability of the use of these biomarkers for well-defined leprosy patients. In contrast, anti-NDO-BSA, anti-LID-1, or anti-NDO-LID antibodies did not discriminate serological cases and non-cases based on the AUC test ( Figure 5 ).