New approaches for developing biomarkers of hormonal contraceptive use

Abstract

To identify biomarkers of hormonal contraceptive (HC) use in urine and saliva, we conducted a pilot study with 30 women initiating levonorgestrel (LNG) containing combined oral contraceptives (COCs) or depot medroxyprogesterone acetate (DMPA) (15/group). Based on established COC pharmacokinetics, we collected serum and urine samples before COC ingestion and during Days one and three of use, or before DMPA injection and on Days 21 and 60 post-injection. We used liquid chromatography-tandem mass spectrometry (LC-MS/MS) to measure serum/urine LNG and MPA. LNG was undetectable at baseline (specificity 100%); post ingestion, most urine samples had detectable LNG levels (sensitivity: 80% 6 h post Dose one, 93% 6 h post Dose three). We used a DetectX LNG immunoassay kit and showed 100% sensitivity measuring urine LNG. Urine MPA levels were undetectable in 14/15 women at baseline (specificity 91%); post-injection all urine samples had detectable MPA levels (sensitivity: 100% days 21 and 60). Results suggest urine sampling can be used to identify a biomarker of LNG and MPA use. Based on evidence from other steroidal hormonal studies showing changes affecting the transcriptome profile of saliva at 24 h, we used the same (COC, DMPA) timepoints to collect saliva. We performed transcriptome analysis and detected several differentially expressed genes in DMPA users' saliva on Days 21 and 60 compared to baseline; none among COC users. We plan further research of differential gene expression in saliva as a HC biomarker of DMPA use, and will explore longer periods of COC use and saliva collection times, and application of microRNA sequencing to support using saliva as a COC biomarker.

Figure 1

Study Flow Diagram. Consort Diagram showing that 41 potential participants were screened and 30 were enrolled, out of which 15 per group (COC or Depo).

Figure 2

(A) Levonorgestrel (LNG) concentrations measured by LCMS/MS method in serum and urine samples were collected on days-1 and 3 of the COC treatment. No individuals showed detectable LNG in serum or urine at baseline (Pre) but all showed an increase at 6 h post ingestion on both days (day-1 and 3). Only eight of 15 participants had detectable levels at all three time points post baseline. Six participants had LNG levels below the level of quantification at the trough timepoint (24 h post dose 2), three of whom also had undetectable levels 6 h after their first dose. (B) Levonorgestrel (LNG) concentrations measured by Enzyme Immuno assay (EIA) method in urine samples collected on day-1 and 3 of the COC treatment. No individuals showed detectable LNG in serum or urine at baseline (Pre) and showed detectable levels at all time points post ingestion on both days (day-1 and 3). (C) Medroxyprogesterone (MPA) concentrations measured by LC–MS/MS method in serum and urine samples collected on day 0 (Baseline) and 21 and 60 days after depo injection. Four individuals had detectable MPA in serum prior to receiving their injection, two of whom also had detectable MPA in urine. The two women with undetectable urine MPA but detectable serum MPA had levels below that required for contraceptive efficacy (< 0.2 ng/mL). One individual with undetectable serum MPA, had detectable urine MPA, with the level in the urine slightly above the serum threshold level (< 0.2 ng/mL). We detected MPA in serum and urine of all participants at 21- and 60-days post injection.

Figure 3

(A–C) A bi-clustering heatmap showing the visual expression profile of the top differentially expressed genes sorted by their adjusted p-value by plotting their log2 transformed expression values in samples (generated by using the Wald test). (A) shows the pre- versus D21 comparison where we observed 10 differentially expressed genes, out of which nine were upregulated and one was downregulated. (B) shows pre-versus D60 comparison, we saw five differentially expressed genes; one was downregulated and four were upregulated. (C) shows the last comparison between D21 versus D60, there were a total of 50 differentially expressed genes out of which 32 were downregulated and 18 were upregulated.

Figure 4

(A–C) Volcano plots to visualize the expression profile of the top differentially expressed genes sorted by their adjusted p-value by plotting their log2 transformed expression values. Downregulated DEGs are noted in red and upregulated DEGs are noted in blue.