2017 data published in peer-reviewed journal

Click below to be redirected to our open-access article published in June 2017 in PLoS NTD. 

Loa Antibody RApid test

The Loa Antibody Rapid Test detects antibodies directed against Loa loa-SXP-1, a specific and validated marker of L. loa infection. The test was evaluated on serum samples from patients infected with L. loa (n=109) and other helminths (n=187), as well as on uninfected controls (n=78). When read with the naked eye, the test was 94% sensitive for L. loa infection and was 100% specific when sera from healthy endemic and non-endemic controls or from those with S. stercoralis infections were used as the comparators. When sera of patients with O. volvulus or W. bancrofti were used as the comparators, the specificity of the LFA was 82% and 87% respectively. A companion smartphone reader allowed measurement of the test line intensities and establishment of cutoff values. With a cutoff of 800 reader units, the assay sensitivity decreased to 56%, but the specificity increased to 97% for O. volvulus and 100% for W. bancrofti-infected individuals.  

Analytical sensitivity

The analytical sensitivity of the Loa Antibody Rapid Test was assessed with pooled L. loa sera, either undiluted or serially diluted with negative serum from uninfected North American individuals. The signal was measured after 20 minutes with the Cellmic smartphone reader [Cat# 63-0001]. When tested with undiluted pooled Loa sera, the assay gave a sharp test line (1113 RUs), comparable in visual intensity to the control line (Figure 1). The signal remained sharp after a 5-fold and 10-fold dilution (914 and 890 RUs, respectively). Upon further dilutions of the test sample, the test line intensity gradually decreased. At a 100-fold dilution, the signal was clearly visible but weaker (447 RUs). The limit of eye detection was reached at ~ an 800-fold dilution (100 RUs).

Figure 1. Analytical sensitivity of the Loa Antibody Rapid Test. The test was run as per the General Procedure, with 5 µL of undiluted pooled L. loa sera, (left), 5 µL pooled L. loa sera serially diluted in negative delipidized human serum with dilution factors of up to 1:1600 (center), or 5 µL pure negative delipidized serum (right). The test lines were quantified with the smartphone reader. The data is reported in reader units (RUs).

Figure 1. Analytical sensitivity of the Loa Antibody Rapid Test. The test was run as per the General Procedure, with 5 µL of undiluted pooled L. loa sera, (left), 5 µL pooled L. loa sera serially diluted in negative delipidized human serum with dilution factors of up to 1:1600 (center), or 5 µL pure negative delipidized serum (right). The test lines were quantified with the smartphone reader. The data is reported in reader units (RUs).

 

Quantification of specific anti-L. loa antibody levels in clinical samples

The Loa Antibody Rapid Test was evaluated at the National Institutes of Health in the Laboratory of Parasitic Diseases using a set of cryopreserved human sera. For logistic reasons, the tests were let dry at the end of the run and were rewetted before scoring. The drying/rewetting process decreases the test line intensity by up to 20% but still gives reproducible and interpretable results.

Figure 2 summarizes the clinical data. Most samples of sera from those with L. loa infection gave sharp signals, with a median value of 845 RUs. Onchocerciasis samples cross-reacted to a small degree with Ll-SXP-1 but there was a clear differentiation in the RUs between the two groups (P value < 0.0001 in a Mann-Whitney U Test). The onchocerciasis group gave substantially weaker signals and median value of 21 RUs – either not detectable or barely detectable with the naked eye. Similarly, the bancroftian filariasis samples partially cross-reacted with the L. loa samples, but were even weaker with a median value of 10 RUs (undetectable by the naked eye, P < 0.0001). Strongyloides samples as well as endemic and non-endemic normal controls were all negative by eye (P < 0.0001). The control lines were tightly grouped with a median value of 1321 RUs (See Supplementary Figure 1).

 
Figure 2. Control and test line intensities by disease. The scatter plot shows individual data points, with median values as blue horizontal bars. The median value for the L. loa samples was 845. Two possible cutoff values are shown at 100 and 800 RUs as dotted lines. Ov = O. volvulus, Wb = W. bancrofti, Ss = S. stercoralis, EN = Endemic Normals, NEN = Non-Endemic Normals.

Figure 2. Control and test line intensities by disease. The scatter plot shows individual data points, with median values as blue horizontal bars. The median value for the L. loa samples was 845. Two possible cutoff values are shown at 100 and 800 RUs as dotted lines. Ov = O. volvulus, Wb = W. bancrofti, Ss = S. stercoralis, EN = Endemic Normals, NEN = Non-Endemic Normals.

 

 

Sensitivity and Specificity of the TEST using clinical serum samples

When read with the naked eye, the Loa Antibody Rapid Test was 94% sensitive for L. loa. The specificity was 82% vs. O. volvulus, 84% vs. W. bancrofti, and 100% vs. S. stercoralis, endemic normal, and non-endemic normal samples (Table 1). The LFA reader offers the remarkable advantage of establishing an objective cutoff value. By adjusting the cutoff value, the assay results can be tuned to modulate the sensitivity and specificity to the needs of the end-users (Table 1). A cutoff of 100 RUs is just above what can be detected by the unaided eye and provides similar results as to the visual scoring: 88% sensitivity, 100% specificity when compared to normal, uninfected sera, and 83–84% specificity towards onchocerciasis and lymphatic filariasis. With a cutoff of 200 RUs, the sensitivity is 84% and the specificity towards O. volvulus and W. bancrofti is 88% and 90%, respectively. By increasing the cutoff to 800 RUs, the sensitivity decreases to 56%, but the specificity towards other filarial species increases to 97­–100%. Receiver Operating Characteristic (ROC) curves (Figure 3, and Supporting Information) allow to examine in more detail the effect of the cutoff on the sensitivity and specificity.

Table 1. Sensitivity and Specificity at Different Cutoff Values. The cutoff values are expressed in Reader Units (RUs). Sensitivity and specificity values are reported in % with 95% confidence intervals in brackets. Ov = O. volvulus, Wb = W. bancrofti, Ss = S. stercoralis, EN = Endemic Normals, NEN = Non-Endemic Normals.

Table 1. Sensitivity and Specificity at Different Cutoff Values. The cutoff values are expressed in Reader Units (RUs). Sensitivity and specificity values are reported in % with 95% confidence intervals in brackets. Ov = O. volvulus, Wb = W. bancrofti, Ss = S. stercoralis, EN = Endemic Normals, NEN = Non-Endemic Normals.

 
Figure 3. Sensitivity and specificity versus O. volvulus as a function of the cutoff value.

Figure 3. Sensitivity and specificity versus O. volvulus as a function of the cutoff value.