Comparison of sampling methods to measure HIV RNA viral load in female genital tract secretions

Abstract

Problem: How does menstrual cup (MC) compare to other genital sampling methods for HIV RNA recovery?

Method of study: We compared HIV RNA levels between MC, endocervical swab (ECS), and ECS-enriched cervicovaginal lavage (eCVL) specimens in 51 HIV-positive, antiretroviral therapy-naive women at enrollment, 3 and 6 months, with order rotated by visit. Paired comparisons were analyzed with McNemar's exact tests, signed-rank tests, and an extension of Somer's D for pooled analyses across visits.

Results: MC specimens had the highest proportion of quantifiable HIV VL at enrollment and month 3, but more MC specimens (n=12.8%) were insufficient for testing, compared with ECS (2%, P=0.006) and eCVL (0%, P<0.001). Among sufficient specimens, median VL was significantly higher for MC (2.62 log₁₀ copies/mL) compared to ECS (1.30 log₁₀ copies/mL, P<0.001) and eCVL (1.60 log₁₀ copies/mL, P<0.001) across visits.

Conclusion: MC may be more sensitive than eCVL and CVS, provided insufficient specimens are reduced.

Keywords: cervicovaginal lavage; endocervical swab; female genital tract sampling; genital HIV viral load; menstrual cup.

FIGURE 1

Comparison in sensitivity to detect HIV RNA concentrations in genital tract specimens collected using different sampling methods. MC, menstrual cup (red); ECS, endocervical swab (blue); eCVL, swab-enriched cervicovaginal lavage (green) with A=enrollment, B=month 3, and C=month 6 visits. Each circle’s area is proportional to the number of samples with quantifiable HIV RNA and their subset, with solid lines indicating the sample that was taken first. Values within circles indicate percentage of total (all women) with quantifiable HIV RNA; the values in brackets indicate the number of total. Values outside of circles indicate the percentage (and number in brackets) of samples, where HIV RNA was not detectable/not quantifiable

FIGURE 2

Comparison of HIV RNA concentrations in menstrual cup (MC), endocervical swab (ECS), and enriched cervicovaginal lavage (eCVL) samples, by sampling order. At enrollment, the specimen collection order was as follows: MC/ECS/eCVL (left panel). At month 3, the order was ECS/eCVL/MC (middle panel). At month 6, the order was eCVL/MC/ECS (right panel). Each dot represents the HIV RNA concentration in each woman’s sample; gray lines link matched sample sets for each woman. Wilcoxon rank-signed test was performed to compare viral loads between sample methods for matched sets, and P-values <0.017 were considered significant

FIGURE 3

Correlation between HIV RNA concentrations detected in matching MC, ECS, and eCVL genital samples, according to sampling order. Each dot represents the HIV RNA concentration in each woman’s matching sample pairs. Spearman correlation was performed, and P-values <0.017 were considered significant. Dotted line represents the line of equivalence (x=y)