A lateral flow-based portable platform for determination of reproductive status of cattle

Our objective was to develop and validate a tool integrating a disposable fluorescence-based lateral flow immunoassay (LFIA) coupled with a portable imaging device for estimating circulating plasma concentrations of progesterone (P4). First, we developed and optimized a competitive LFIA test strip to measure P4 in bovine plasma. The LFIA design included a sample pad, a conjugate pad that stores R-phycoerythrin-antiP4 conjugates, a glass-fiber spacer pad, a nitrocellulose membrane with printed test and control lines, and a cellulose-fiber absorbent pad. To perform a test, 20 mu L of plasma and 50 mu L of running buffer were added on the sample pad. After 3 min, 45 mu L of running buffer was added to initiate sample flow. After allowing 15 min to stabilize the colorimetric signal, strips were introduced in an LFIA portable reader wirelessly linked to a laptop to determine P4 concentration based on test-to-control-line signal (T/C ratio). In a series of experiments (n = 6), the ability of the LFIA to differentiate plasma samples with >= 1 or <1 ng/mL of P4 was evaluated. For each experiment, a calibration curve was constructed using plasma with known concentrations of P4 (0.1 to 3.7 ng/mL; n = 5). The resulting linear equation was then used to determine a T/C ratio cutoff to differentiate samples with >= 1 or <1 ng/mL of P4. In addition, to evaluate the ability of the platform to assign samples to P4 concentration groups without a calibration curve for individual batches, we performed a receiver operating characteristic analysis to identify a single cutoff value for T/C ratio that could potentially be used for all batches. Overall, calibration curves showed a linear relationship between T/C ratio and P4 levels (mean coefficient of determination = 0.74; range 0.42 to 0.99). Next, plasma samples from lactating dairy cows (n = 58) were tested in triplicate to determine the ability of the LFIA system to differentiate plasma samples with >= 1 or <1 ng/mL of P4 using a RIA for P4 as reference test. Overall, the LFIA assay correctly classified 90% of the samples, with 97% sensitivity, 83% specificity, 85% positive predictive value, and 96% negative predictive value. Agreement between the tests was substantial (kappa = 0.79; 95% confidence interval 0.64 to 0.95). When using a single cutoff value for T/C ratio selected by receiver operating characteristic analysis, sensitivity and specificity to determine CL presence were 97 (95% confidence interval 82 to 99) and 79% (95% confidence interval 60 to 92), respectively. These data suggest that the developed portable LFIA system can accurately differentiate plasma samples with >= 1 or <1 ng/mL of P4.