Impact of etonogestrel implant use on T-cell and cytokine profiles in the female genital tract and blood.

BACKGROUND: While prior epidemiologic studies have suggested that injectable progestin-based contraceptive depot medroxyprogesterone acetate (DMPA) use may increase a woman's risk of acquiring HIV, recent data have suggested that DMPA users may be at a similar risk for HIV acquisition as users of the copper intrauterine device and levonorgestrel implant. Use of the etonogestrel Implant (Eng-Implant) is increasing but there are currently no studies evaluating its effect on HIV acquisition risk.
OBJECTIVE: Evaluate the potential effect of the Eng-Implant use on HIV acquisition risk by analyzing HIV target cells and cytokine profiles in the lower genital tract and blood of adult premenopausal HIV-negative women using the Eng-Implant.
METHODS: We prospectively obtained paired cervicovaginal lavage (CVL) and blood samples at 4 study visits over 16 weeks from women between ages 18-45, with normal menses (22-35 day intervals), HIV uninfected with no recent hormonal contraceptive or copper intrauterine device (IUD) use, no clinical signs of a sexually transmitted infection at enrollment and who were medically eligible to initiate Eng-Implant. Participants attended pre-Eng-Implant study visits (week -2, week 0) with the Eng-Implant inserted at the end of the week 0 study visit and returned for study visits at weeks 12 and 14. Genital tract leukocytes (enriched from CVL) and peripheral blood mononuclear cells (PBMC) from the study visits were evaluated for markers of activation (CD38, HLA-DR), retention (CD103) and trafficking (CCR7) on HIV target cells (CCR5+CD4+ T cells) using multicolor flow cytometry. Cytokines and chemokines in the CVL supernatant and blood plasma were measured in a Luminex assay. We estimated and compared study endpoints among the samples collected before and after contraception initiation with repeated-measures analyses using linear mixed models.
RESULTS: Fifteen of 18 women who received an Eng-Implant completed all 4 study visits. The percentage of CD4+ T cells in CVL was not increased after implant placement but the percentage of CD4+ T cells expressing the HIV co-receptor CCR5 did increase after implant placement (p = 0.02). In addition, the percentage of central memory CD4+ T-cells (CCR7+) in CVL increased after implant placement (p = 0.004). The percentage of CVL CD4+, CCR5+ HIV target cells expressing activation markers after implant placement was either reduced (HLA-DR+, p = 0.01) or unchanged (CD38+, p = 0.45). Most CVL cytokine and chemokine concentrations were not significantly different after implant placement except for a higher level of the soluble lymphocyte activation marker (sCD40L; p = 0.04) and lower levels of IL12p70 (p = 0.02) and G-CSF (p<0.001). In systemic blood, none of the changes noted in CVL after implant placement occurred except for decreases in the percentage CD4 T-cells expressing HLA-DR+ T cells (p = 0.006) and G-CSF (p = 0.02).
CONCLUSIONS: Eng-Implant use was associated with a moderate increase in the availability of HIV target cells in the genital tract, however the percentage of these cells that were activated did not increase and there were minimal shifts in the overall immune environment. Given the mixed nature of these findings, it is unclear if these implant-induced changes alter HIV risk.

Introduction

As approximately 40% of all pregnancies worldwide are unintended [1], prevention of unintended pregnancy is a public health priority. Over 200 million women worldwide use progestin-containing contraception (HC) either alone or in combination with estrogen to achieve their family planning goals [2]. Some research suggests that HC may contribute to the spread of HIV by increasing susceptibility to infection [3]. The greatest concern had been with Depot Medroxyprogesterone acetate (DMPA) where a meta-analysis of nine studies estimated a significant increase in HIV risk of 40% with DMPA compared to non-HC use [3]. Notably, these findings have not been consistently demonstrated by all studies, and all the studies were observational thus prone to potential confounding. Further, the findings have been challenged by the results from the Evidence for Contraceptive Options and HIV outcomes (ECHO) study [4], a large randomized trial that found no significant increase in HIV risk among users of DMPA compared to copper intrauterine device and levonorgestrel implant users. While these data are reassuring, gaps remain in our understanding of the relative HIV risk with contraceptive use compared to non-use and the impact of other progestin-containing contraceptive methods. Notably, data on the etonogestrel implant (Eng-Implant), are limited despite increasing rates of global use of these contraceptive implants.

There are several potential non-contraceptive effects from the use of potent steroid hormones. High levels of estrogen and progesterone during pregnancy are associated with a shift from a TH1 to a TH2-dominant immune profile, dampening the pro-inflammatory pathways, and increasing susceptibility to certain disease conditions (e.g., influenza, malaria, listeria) while reducing the severity of others (e.g., multiple sclerosis, rheumatoid arthritis) [5-9]. This natural phenomenon lends credence to the scientific premise of immune changes with hormone concentrations. HIV risk could thus be amplified by an increased representation of cells expressing HIV co-receptors within the female genital tract or trafficking of HIV target cells to the genital mucosa. CD4+ T-cells expressing the cell surface receptor C-C chemokine receptor type 5 (CCR5, the primary HIV co-receptor) are among the first cells to be infected and subsequently the virus can spread to regional lymph nodes [10]. Within the lower genital tract mucosa, the number and type of cellular targets, primarily CD4+ T-cells expressing CCR5, predict susceptibility to HIV infection [11, 12]. The functional properties exhibited by CD4+ T-cells influence susceptibility to HIV infection, specifically the expression of activation-associated molecules (markers such as HLADR, CD38) have been associated with increased risk of HIV acquisition [13-16].

Prior studies have evaluated the immune effects of DMPA on these cellular markers, with some, but not all, noting increases in key HIV target cell markers [17-23]. No studies have explored the effect of the Eng-Implant on these key cellular markers of HIV risk nor evaluated larger epidemiologic data to explore any cohorts to evaluate for an association between Eng-Implant use and HIV acquisition. Given this gap, we aimed to prospectively examine the effect of Eng-Implant initiation on the systemic and lower genital tract mucosal immune environment, with a focus on HIV target cells. The Eng-Implant is a long-acting highly effective progestin-only contraceptive method containing a 3rd generation progestin. Although Eng-Implant has less glucocorticoid activity compared to medroxyprogesterone in DMPA and less endogenous estrogen inhibition, we hypothesize that given the sustained progestin exposure over time, there will still be some immunologic changes within the genital tract to suggest increased susceptibility with use.

Materials and methods

Study population and recruitment

This was a prospective study to evaluate the effect of three months of Eng-Implant use on HIV target cells and inflammatory markers in the lower genital tract and systemic circulation. This manuscript is the first to evaluate one of the primary study objectives of a larger cohort study of three contraceptive methods registered at clinicaltrials.gov, NCT02357368. Women recruited into the larger cohort could initiate the Eng-Implant, the Levonorgestrel Intrauterine device or DMPA based on their preference. For this analysis, we will focus on the results from all individuals selecting the Eng-Implant, as the differences in baseline characteristics and relatively small sample size limit our power to make comparisons between methods. Women interested in initiating a new contraceptive method were recruited from the metro-Atlanta area via community-based postings or local referral from clinics. We enrolled eligible women between ages 18–45, who experienced normal menses (22–35 day intervals) for at least three cycles, had an intact uterus and cervix, and were HIV uninfected (determined by point-of-care rapid test using OraQuick®). Participants could not have used HC or copper intrauterine device (IUD) in the previous 6 months, had any signs of an STI on clinical examination at time of enrollment and needed to be medically eligible to initiate their selected contraceptive method (for this analysis Eng-Implant) based on CDC medical eligibility criteria for contraceptive use and clinical judgment. Approval for this study was obtained from the Emory IRB and Grady Research Oversight Committee prior to study initiation. Written informed consent was obtained from all participants and all laboratory researchers and technicians were blinded to contraceptive exposure.

Study procedures/clinical visits

The primary exposure of interest was the Eng-Implant (Nexplanon, Merck & Co, Inc) [24]. We scheduled four study visits for each participant, two visits prior to contraceptive initiation and two visits approximately three months after Eng-Implant administration. Study visits were scheduled with the goal of pre-contraceptive sample collection at both the luteal (visit 1- target of 21 days after last menstrual period with window of 17 days to onset of next menses) and the follicular (visit 2-target three days after completion of menses, with window up to 14 days after onset of menses) phases of the menstrual cycle, based on a self-reporting of date of the last menstrual period. The Eng-Implant was placed at completion of visit 2. Post-contraceptive sampling collection occurred two weeks apart approximately three months after contraceptive initiation: visit 3 with target 12 weeks after (window 11–14 weeks after contraceptive initiation) and visit 4 with target 14 days after visit 3 (window of 12 to 16 days). The a priori goal was to compare the baseline (with follicular and luteal variation accounted for) with the post contraception (with 2 visits to account for variations over a 2 week period where endogenous hormonal changes may occur) results. We requested participants to abstain from vaginal intercourse for 24 hours prior to each visit to minimize the risk of contamination of genital tract samples by semen.

Specimen collection

During a speculum examination, we collected a cervicovaginal swab for sexually transmitted infections (DrySwab™, Lakewood Biochemical Company). This was followed by a cervicovaginal lavage (CVL) collection with a lavage from the cervix, vaginal walls and posterior fornix with 10 ml of phosphate-buffered saline (PBS) for approximately 60 seconds as per the protocol described by the Microbicide Trials Network (https://vimeo.com/224957115/00cb72fed6) with details previously described [25]. To enhance cellular yield, CVL was performed twice. We collected blood in 8 mL sodium citrate-containing CPT tubes (BD Biosciences). CVL allows enrichment of target cells positioned at the apical lumen in proximity to exposure with a lower risk of tissue trauma from sampling that would cause bleeding and contamination of phenotyping. CVL or luminal cells are not imbedded within the tissue, persist within a harsh environment, and have a reduced cell yield compared with other sampling approaches, but CVL provides an accurate means of tissue resident phenotyping at the site of sexually transmitted exposure. In experiments where luminal T cells are analyzed separately from T cells embedded in the tissue, these two populations have been shown to be very similar phenotypically and functionally. [26-31] Microscopically, it has been shown that luminal T cells remain closely associated with the apical face of the epithelium. [32-36]Several studies have shown these luminal T cells are viable, capable of recognizing and responding to antigen, and play a critical role in immunity at mucosal sites[28, 35, 37–41]. Luminal T cells are sufficient to provide significant protection even when T cells located in the underlying tissues are not present[41], thus although you may not find a large number of T cells in CVL, these cells can be critical for barrier protection. With our methodology we have high viability of the cells (70–90%) and although cell count numbers for leukocytes are low, these counts are within the range of other sampling methodologies [25, 42].

Covariate assessment

The following covariates were measured at each visit. 1) Semen presence in CVL: we detected semen presence using the Abacus ABAcard p30 test to detect prostate specific antigen (PSA). 2) Presence of sexually transmitted infections (STIs): we collected a cervicovaginal swab for STIs (DrySwab™, Lakewood Biochemical Company) at each visit prior to CVL from cervical os. DNA was extracted using the Qiagen DNA Mini Kit and used to amplify targets from Neisseria gonorrhoeae, Chlamydia trachomatis, Mycoplasma genitalium, Trichomoniasis vaginalis, and Herpes simplex virus types 1 and 2, using two real-time duplex PCR assays and Qiagen Rotor-Gene Q real-time PCR instrument. Qiagen Rotor-Gene Q Series software was used to analyze data. These multiplex PCR assays were performed in the Division of Sexually Transmitted Diseases Laboratory Reference and Research Branch at the US Centers for Disease Control and Prevention. 3) Bacterial Vaginosis (BV): we determined the presence of BV by Nugent score criteria [43] from gram stains prepared from CVL. Prior comparison data from our lab between CVL smears and swab among 37 sample pairs were highly correlated (r>0.88, p<0.0001) with categorical interpretations in agreement for all slides. Scores above six are considered consistent with BV. 4) Blood presence in the CVL: we defined blood presence qualitatively with a urine dipstick test detecting > = 8000 RBC/μl. This cut off was selected for inclusion of potential systemic blood contamination, however exploration and use of other cut-off values did not meaningfully alter our study findings.

Immune marker assessment

Specimens were placed in a cooler with ice immediately after collection and transported to the Division of HIV/AIDS Prevention Laboratory Branch at Centers for Disease Control and Prevention (CDC) within four hours of collection for processing, cellular isolation and characterization. Blood was separated into plasma and peripheral blood mononuclear cells (PBMCs) by centrifugation in CPT tubes as instructed by manufacturer. After collecting blood plasma, PBMCs were collected from the CPT tube and washed with PBS prior to staining. CVL specimens were enriched for leukocytes using Percoll gradient centrifugation as previously described [25]. Plasma and CVL supernatant aliquots were stored at -80°C until analysis. Cellular characterization was performed at that time on CVL leukocytes and PBMCs via flow cytometry. Viable leukocytes were distinguished using Zombie Fixable Viability Kits (Biolegend) then blocked for non-specific staining with anti-CD16/32 Fc (BioXcell).

The primary outcome of interest was the proportion of CD4 cells with CCR5 expression. Secondary outcomes evaluated were: 1) CD4/CD8 T-cell ratio to measure the changes in T-cell homeostasis, 2) the expression of activation markers CD38 and HLA-DR, peripheral tissue retention marker CD103[44] and trafficking marker CCR7 on CD4 T-cells or CD4 CCR5+ T-cells, 3) the differentiation of lymphocyte memory CD4 and CD8 T-cell phenotypes (Naïve T-cell (TNA): CD45RAhi and CCR7hi; central memory T-cells (TCM): CD45RAlo and CCR7hi; effector memory T-cells (TEM): CD45RAlo and CCR7lo; and effector memory RA expressing T-cells (TEMRA): CD45RAhi and CCR7lo). These outcomes are quantified for each individual at every time point.

Cells were stained with the following fluorochrome conjugated antibodies: CD3 (V450, UCHT1), CD4 (Alexa Fluor 700®, RPA-T4), CD8 (BV510, RPA-T8), CCR7 (PE-CF594, 150503), CCR5 (PE, 3A9), CD103 (FITC, Ber-ACT8) (BD Biosciences), CD38 (PE/Cy7, HIT2), CD45 (BV650™, H130), CD45RA (BV605™, HI100), HLA-DR (BV785™, L243) (Biolegend). Stained samples were run on an LSRII flow cytometer and acquired using FACS DIVA software (BD Immunocytochemistry Systems, San Jose, CA, USA) and analyzed using FlowJo software (TreeStar, Inc.). Cellular measurements were analyzed as a percentage of CD4+ T cells, or CD4+ CCR5+ T cells expressing a given marker or combination of markers. For accurate measurement of CCR5 expression frequency on CD4 T cells, CCR5 gating was set against matched-naïve CD4 T cells from PBMCs as previously described [45].

Soluble immune mediators from the CVL supernatant and plasma were evaluated using Luminex technology with xPONENT software (Luminex Corporation) with all samples tested in duplicate on a 96 well plate containing seven standards, two quality controls and 39 samples using a customized multi-analyte panel (HCYTOMAG-60K-18 MILLIPLEX Human Cytokine panel, Millipore). The panel contained selected proinflammatory, inhibitory and chemotactic soluble cytokine and chemokines [IL-1b, IL-6, IL-12 (p70), IFN-a2, IFN-g, IL-1a, IL-17, IL-2, TNF-a, IL-4, GM-CSF, G-CSF sCD40, MIP-1a, MIP-1b, IP-10, IL-8, Fractalkine (CX3CL1)]. Using the sigmoid standard curve from the Millipore Analyst 5.1, a regression curve was extrapolated from the raw data individually for each cytokine. For samples below the level for quantification, we used half the lower limit of detection.

Statistical methods

Visits were dichotomized into pre-implant use (visits 1 and 2) and post-implant use (visits 3 and 4). Any outcome (cytokine, cellular marker) value below the limit of detection was assigned a value of half of the lowest measured value for that outcome. Only samples with greater than 100 viableCD3+ cells extracted were included in the analyses, a similar approach to Lajoie et al [46]. To evaluate for potential associations that may be confounding our interpretation of study findings, we conducted separate logistic mixed models to assess the association of implant status (pre, post) with CVL visit characteristics (semen, STI, blood, BV). Models contain covariate of interest, a random intercept for subject, and variance components variance structure. This statistical approach was selected for evaluation of longitudinal data with repeated measurements of the same patient over time [47]. Separate generalized linear mixed models with a gamma distribution and log link were used to assess the association of each cytokine (IL-1b, IL-6, IL-12 (p70), IFN-a2, IFN-g, IL-1a, IL-17, IL-2, TNF-a, IL-4, GM-CSF, G-CSF sCD40), chemokine (MIP-1a, MIP-1b), chemotactic cytokine (IP-10, IL-8, Fractalkine) and cellular marker (CD4 CCR5, CD4 CD38, CD4 HLA-DR 2, CD4 CD103, CD4 CCR7) outcome with implant use (pre, post). Models included implant use, a random intercept for subject and variance components covariance structure. All models were stratified by tissue type (CVL, blood). CVL models additionally included presence of semen, presence of blood, STI, and BV status as covariates. Model-based estimates and 95% confidence intervals of estimated mean outcome level by implant use were back-transformed (exponentiated) to produce estimated arithmetic means on the original scale. Similarly, the estimated arithmetic mean ratio (AMR) and 95% confidence interval of post-implant use to pre-implant use was produced by exponentiating the coefficient for implant use. Linear mixed regression models were used to assess whether the distribution of lymphocyte memory cells into four mutually exclusive groups (TNA, TCM, TEM, TEMRA) varied by implant use. Models contained memory cell type, implant use, and memory cell type * implant use interaction term. The Type 3 F test of the interaction term is reported as well as model-based estimates and 95% confidence intervals for memory cell type and implant use. The models included robust variance estimates and compound symmetry covariance structure by subject grouped by memory cell type nested within implant use and were stratified by CD4 and CD8 T-cell phenotypes and tissue type. Model fit was assessed through residual plots. To reduce the potential impact from multiple comparisons on false discovery, we interpreted our results with an adjusted p-value using a Benjamini and Hochberg false discovery rate of 0.1 for each set of analyses within specimen type and cytokine/cellular marker sets. Memory cell type analyses set α = 0.05. Analyses were conducted in SAS v9.4.

Results

Eighteen women enrolled in the study and completed both pre-contraceptive visits, 16 women completed visit 3 (88.9%) and 15 completed all 4 visits (83.3%). All pre-contraceptive luteal and follicular samples were collected during the appropriate windows described in the study methods, with Visit 3 and Visit 4 conducted at a median of 84 days (Q1: 83, Q3: 86.5 days) and 105 days (Q1: 98, Q3: 111) post-contraceptive initiation. Women were predominately African-American (83%), unmarried (83%), and young (median age 24 years) (Table 1). CVLs collected from 53 (79% of all visits) visits contained greater than 100 viable CD3+ T-cells and were subsequently included in this analysis. There were no associations between having fewer than 100 CD3+ T-cells on the analysis and visit number (data not shown). STIs were diagnosed by PCR at 28 (42%) visits, and BV diagnosed by Nugent score at 27 (42%) visits (Table 2). There were no significant differences in any of the visit level covariates between before and after the implant placement.

Table 1

Cohort characteristics of n = 18 women enrolled in study.

	n	%
Age, years (median (Q1,Q3))	23.7	(23.2, 30.2)
Race
African-American	15	(83.33)
Other	3	(16.67)
Ethnicity: Hispanic	1	(5.56)
Marital Status
Married/cohabitating	3	(16.67)
Single/divorced/widowed	15	(83.33)
Education
<High school diploma	4	(22.22)
High school diploma/GED	5	(27.78)
Some college	6	(33.33)
Associate’s degree/Technical certification	1	(5.56)
Bachelor’s degree	2	(11.11)
Annual Income
<$10,000	7	(38.89)
$10,000-$24,999	5	(27.78)
$25,000-$50,000	4	(22.22)
Don’t know/refuse	2	(11.11)

Table 2

Time-varying characteristics of CVL specimens by study visit.

Characteristic	Visit 1 (n = 18)		Visit 2 (n = 18)		Visit 3 (n = 16)		Visit 4 (n = 15)
	n	%	n	%	n	%	n	%
Semen	4	(22.2)	5	(27.8)	2	(12.5)	3	(20.0)
STI	5	(27.8)	7	(38.9)	9	(56.3)	7	(46.7)
BV	6	(33.3)	8	(47.1)	7	(50.0)	6	(40.0)
Blood	1	(5.6)	4	(22.2)	2	(12.5)	1	(6.7)
Viable CD3 lymphocyte count <100	6	(33.3)	3	(16.7)	2	(12.5)	3	(20.0)
	Median	IQR	Median	IQR	Median	IQR	Median	IQR
viable CD3 lymphocyte number	548.5	245, 2769	1118	185, 2330	1629	844, 4121	1036.5	423, 3288.5
CD4 lymphocyte number*	328	140, 1340	494	128, 998	769	346, 1990	574	76, 1418

IQR = Interquartile range,

* among samples with viable CD3 lymphocyte count ≥ 100.

In the lower genital tract, we noted a significant increase in the proportion of CD4 cells expressing CCR5 after implant placement compared to measures taken prior to placement [AMR 1.56, 95% CI 1.09–2.24] (Table 3). Furthermore, there was a decrease in the CD4/CD8 ratio [AMR 0.70, 95% CI 0.53, 0.94], consistent with a significant increase in the proportion of CD8+ T-cells following Eng-Implant use [AMR 1.27, 95% CI: 1.06–1.53]. Eng-Implant use resulted in a decreased proportion of genital tract CD4+ T-cells expressing HLA-DR [AMR 0.58, 95% CI 0.38, 0.90] and CD4 CCR5+ T-cells expressing HLA-DR [AMR 0.54, 95% CI 0.34, 0.85]; however there were no significant changes in the expression of the activation marker CD38. Notably, all of these findings except for the decreased CD4 CCR5+ T-cells expressing HLA-DR remained significant after adjusting the alpha for multiple comparisons. Additionally, Eng-Implant use significantly changed the distribution of T-cell subtypes among both the CD4 (p = 0.004, Fig 1A) and CD8 T-cells (p = 0.023, Fig 1B), with an observed shift away from effector memory subtype.

Table 3

Estimated cellular marker levels in the CVL for implant users, adjusting for repeated measures and covariates.

	Pre-Implant	Post-Implant		Arithmetic Mean Ratio Post-Implant/ Pre-Implant
Cellular Marker	Estimate* (95%CI)	Estimate* (95% CI)	p-value	(95% CI)
% of CD3 + T-cells expressing:
CD4+	51.07 (41.45, 62.93)	46.92 (38.41, 57.32)	0.373	0.92 (0.76, 1.11)
CD8+	21.60 (16.64, 28.03)	27.48 (21.28, 35.49)	0.013	1.27 (1.06, 1.53)
CD4/CD8 ratio	2.60 (1.81, 3.74)	1.83 (1.29, 2.61)	0.019	0.70 (0.53, 0.94)
% of CD4+ T-cells expressing:
CCR5+	17.76 (10.64, 29.65)	27.75 (16.69, 46.14)	0.017	1.56 (1.09,2.24)
CD38+	32.37 (24.56, 42.66)	40.26 (31.06, 52.19)	0.111	1.24 (0.95, 1.63)
HLA-DR+	25.69 (15.21, 43.38)	15.01 (8.96, 25.13)	0.018	0.58 (0.38, 0.90)
CD103+	6.58 (3.09, 14.02)	11.32 (5.81, 22.08)	0.084	1.72 (0.92, 3.21)
% of CD4+ CCR5+ T-cells expressing:
CCR7+	32.37 (22.02, 47.57)	38.46 (27.11, 54.56)	0.290	1.19 (0.85, 1.65)
CD38+	65.01 (51.84, 81.52)	60.12 (49.54, 72.95)	0.453	0.92 (0.75, 1.14)
HLA-DR+	50.98 (31.32, 82.98)	27.45 (17.98, 41.89)	0.009	0.54 (0.34, 0.85)

Generalized linear mixed model with a random intercept for participant, variance components covariance structure, gamma distribution, log link Restricted to CD3 count>100

* Back-transformed estimate (arithmetic mean)

** P-value is for adjusted model. Bold indicates significant after Benjamini–Hochberg correction

Fig 1

a. CVL CD4 memory cell distribution before and after implant use, adjusting for time-varying semen presence by PSA, sexually transmitted infections, bacterial vaginosis, presence of blood in the sample and repeated measures, p = 0.004. b. CVL CD8 memory cell distribution before and after implant use, adjusting for time-varying semen presence by PSA, sexually transmitted infections, bacterial vaginosis, presence of blood in the sample and repeated measures p = 0.023. Memory cell phenotypes included Naïve T-cell (CD45RAhi and CCR7hi); Tcm = central memory T-cells (CD45RAlo and CCR7hi; Tem = effector memory T-cells (CD45RAlo and CCR7lo) and TEMRA = effector memory RA expressing T-cells (CD45RAhi and CCR7lo).

Among the PBMCs, there were minimal changes in the distribution of T-cell phenotypes, with no noted changes in T-cell ratios or CCR5 co-receptor expression (Table 4). There was a significant increase in CCR7 expression on CCR5+ CD4 T-cells [AMR 1.30, 95% CI 1.08, 1.57] and decreased HLA-DR on CD4 T-cells [AMR 0.81, 95% CI 0.70, 0.90]. These findings remained significant with the adjusted alpha value. There was a significant difference in distribution of memory cell phenotype among the CD4 T-cells (p = 0.014) with an observed shift towards more naïve cells and reduced effector memory cells (Fig 2A and 2B).

Table 4

Estimated cellular marker levels in PBMC for implant users, adjusting for repeated measures.

	Pre-Implant	Post-Implant		Arithmetic Mean Ratio Post-Implant/ Pre-Implant
Cellular Marker	Estimate* (95% CI)	Estimate* (95% CI)	p-value**	(95% CI)
% of CD3 + T-cells expressing:
CD4+	66.48 (62.58, 70.61)	66.24 (62.25, 70.49)	0.857	1.00 (0.96, 1.04)
CD8+	25.10 (21.75, 28.97)	25.69 (22.14, 29.80)	0.662	1.02 (0.92, 1.14)
CD4/CD8 ratio	2.80 (2.22, 3.54)	2.77 (2.17, 3.55)	0.929	0.99 (0.81, 1.21)
% of CD4+ T-cells expressing:
CCR5+	3.46 (2.64, 4.53)	3.97 (2.99, 5.27)	0.276	1.15 (0.89, 1.48)
CD38+	15.76 (12.23, 20.32)	15.41 (11.85, 20.04)	0.800	0.98 (0.82, 1.17)
HLA-DR+	2.78 (2.32, 3.33)	2.26 (1.87, 2.73)	0.006	0.81 (0.70, 0.94)
CD103+	0.14 (0.07, 0.28)	0.12 (0.05, 0.24)	0.351	0.81 (0.52, 1.28)
% of CD4+ CCR5+ T-cells expressing:
CCR7+	25.81 (20.96, 31.77)	33.65 (27.15, 41.70)	0.007	1.30 (1.08, 1.57)
CD38+	24.58 (20.21, 29.88)	27.02 (22.08, 33.05)	0.290	1.10 (0.92, 1.31)
HLA-DR+	22.22 (19.52, 25.29)	21.08 (18.41, 24.14)	0.496	0.95 (0.81, 1.11)

Generalized linear mixed model controlling for time-varying semen presence by PSA, sexually transmitted infections, bacterial vaginosis, presence of blood in the sample and repeated measures with a random intercept for participant, variance components covariance structure, gamma distribution, log link. Restricted to CD3 count>100

* Back-transformed estimate (arithmetic mean)

** P-value is for adjusted model. Bold indicates significant after Benjamini–Hochberg correction

Fig 2

a. PBMC CD4 memory cell distribution before and after implant use, adjusting for repeated measures, p = 0.014. b. PBMC CD8 memory cell distribution before and after implant use, adjusting for repeated measures, p = 0.536. Memory cell phenotypes included Naïve T-cell (CD45RAhi and CCR7hi); Tcm = central memory T-cells (CD45RAlo and CCR7hi; Tem = effector memory T-cells (CD45RAlo and CCR7lo) and TEMRA = effector memory RA expressing T-cells (CD45RAhi and CCR7lo).

Overall, lower genital tract cytokine expression was similar before and after implant initiation (Fig 3, S1 Appendix), with significant reductions only noted for GSCF [AMR 0.54, 95% CI 0.39–0.74, p = 0.0004] and IL12p70 [AMR 0.69, 95% CI 0.51–0.94, p = 0.0212] and a significant increase in sCD40L [AMR 1.46 95% CI 1.02, 2.08, p = 0.0380]. Among these, only GCSF remained significant after adjusting the alpha for the multiple comparisons. Similarly minimal changes in plasma cytokine concentrations were observed following Eng-Implant use, with only a significant reduction in GCSF [AMR 0.84, 95% CI 0.72, 0.97, p = 0.0207] (Fig 4, S2 Appendix), however this was no longer significant after adjusting alpha for multiple comparisons.

Fig 3

Forest plot of arithmetic mean ratio of cytokine levels post-implant compared to pre-implant use with 95% confidence intervals, adjusting for time-varying semen presence by PSA, sexually transmitted infections, bacterial vaginosis, presence of blood in the sample and repeated measures.

Fig 4

Forest plot of arithmetic mean ratio of cytokine levels post-implant compared to pre-implant use with 95% confidence intervals, adjusting for repeated measures.

Discussion

The results of our study highlight few changes in the lower genital tract inflammatory environment following Eng-Implant initiation. While several studies have evaluated genital tract immune changes after use of other hormonal contraceptive methods [17–23, 48–61], to our knowledge, no published study has evaluated Eng-Implant. We report an increase in the proportion of CD4 T-cells expressing the co-receptor CCR5 at the genital mucosa with Eng-Implant use that could be associated with increased risk of HIV infection; however, not all our findings relate a clear picture of increased susceptibility. For example, implant placement reduced the frequency of the activation marker HLA-DR, but not CD38, among CD4 T-cells. The clinical significance of these findings on HIV acquisition is unclear. Further, while there were minimal changes in soluble immune markers in the lower genital tract, these changes suggest a slight increase in local immune suppression with reduced concentrations of proinflammatory cytokines, a finding similarly noted with DMPA [54]. As we observed some shifts in T-cell populations in the genital tract associated with Eng-Implant use, we cannot eliminate the potential that Eng-Implant has an effect on HIV acquisition. This finding is important when interpreting the results of the recently conducted ECHO study [4], as they did not find a significant difference in HIV acquisition between DMPA users and users of the copper intrauterine device and the levonorgestrel implant. The ECHO results are encouraging that DMPA did not differ from these other methods in relation to HIV risk. The ECHO study was powered to detect a clinically significant increased risk of 50% and conclusions regarding other methods cannot be made. Furthermore, while we evaluated the etonogestrel and not levonorgestrel implant, we find some changes in immunologic markers with unclear impact on susceptibility. While small changes in individual risk with a contraceptive method use should not alter eligibility for use [62] of a contraceptive method, with increasing global utilization of many of the longer-acting contraceptive methods, it is important for research to identify even subtle differences that may influence counselling for high-risk individuals and have public health importance.

The increased expression of CCR5 at the genital mucosa may reflect infiltration of T-cells or a direct effect of the implant on the expression of CCR5 [63]. The increased expression of CD103, coupled with the shift from the canonical effector memory phenotype towards a central or migratory memory subtype, suggests that infiltration and retention are drivers of this shift [25]. The CD4/CD8 T-cell ratio further supports that the Eng-Implant use is influencing the trafficking patterns of immune cells. The increased frequency of CD8 T-cells at the genital mucosa is provocative and clue potential alterations in local inflammation. Prior research suggests that effector CD8 T-cells cannot enter into the vagina without CD4-T cell permission in the form of activation-associated cytokines [33]. While our cytokine findings do not fully support this finding, it is possible that soluble cytokine measurements may not detect this mechanism.

A decrease in GCSF was noted in the genital tract after implant initiation. A similar reduction in GCSF was also observed in plasma (although not significant after adjusting for multiple comparisons). This finding of a small yet significant reduction in GCSF may signify an alternative pathway associated with altered HIV susceptibility through damaged mucosa. Granulocyte colony-stimulating factor (GCSF) may induce an inflammatory reaction enhancing neutrophil function. With receptors on granulosa cells, GCSF has been implicated in ovulation and thus could be downregulated in the setting of ovulation inhibition associated with implant use [64]. GCSF is also associated with wound healing and has been associated with faster healing from genital ulcerations [65], GCSF stimulates the proliferation and differentiation of cells that participate in acute and chronic inflammation and immune responses including mature leukocytes, macrophages, and dendritic cells [66]. This potential mechanism of altered immune response should be further explored to determine if clinically significant.

The mechanism by which progestin contraceptives may be influencing immune expression in the genital tract is not fully elucidated. Progestins may act via alteration of gene expression after binding to and activating intracellular steroid receptors [63], which vary based on different tissue cell types. Gonadal hormones can regulate the expression of numerous genes involved in multiple cellular functions [67, 68] with the effects modified by cell type, presence of other hormones and transcription factors, and their binding potential for progestins by other steroid receptors besides the progesterone receptor can result in agonist or antagonist activity. Further, biological effect can vary based upon the dose of progestin. High progestin levels can cause thickening of cervical mucus that creates a barrier to sperm assent, suppress ovulation and alter the endometrial lining. Progestin-containing contraceptive methods can differ by their mode of delivery, length of effectiveness, global availability, degree of endogenous hormone and ovulation inhibition and type of progestin they contain with varying degrees of estrogenic, androgenic, anti-androgenic, glucocorticoid and anti-mineralocorticoid activity [69, 70]. For example, medroxyprogesterone (MPA), a synthetic progestin in DMPA, has potent glucocorticoid (GC) activity compared with weaker GC activity for ENG, where levoneorgestrel (LNG) has no GC activity. Given the differences among different contraceptive methods, it is important to understand the immunologic effect of varying types of contraception to understand their potential and relative impact on reproductive health and immunity. Notably, even among individuals using the same contraceptive, the serum progestin concentrations can vary widely and these differences may influence the effect of the particular contraceptive [63]. The Eng-Implant users may have variability in serum concentrations with implant use as well as tissue level exposure and tissue responsiveness via steroid receptors. Understanding the individual level factors that influence both systemic hormonal concentrations and mucosal level response to the hormone is critical to provide guidance for individual level counseling.

While other studies have not evaluated the Eng-Implant, prior studies evaluating the effects of DMPA have conflicting results [17-22]. For example, while two studies did not see changes in the vaginal HIV target cells with DMPA use [18, 23] one other found significantly higher frequencies of CCR5+ CD4+ T-cells (relative risk: 3.92) compared to non-users [22]. A recent cross sectional analysis comparing 15 DMPA users to 20 non-hormonal contraceptive users found higher levels of activated T-cells and a higher proportion of CD4+CCR5+ T-cells among DMPA users on tissue biopsy samples, however this increase was not noted among the cervical mononuclear cells obtained via cytobrush and cervical spatula [46]. Some of the inconsistencies may be related to the differences in study population, study methodology (cross sectional versus longitudinal), timing of sample collection in relation to luteal or follicular phases or timing in relation to hormonal contraception, sample collection approach or laboratory methodology. Importantly, vaginal immune parameters are influenced by many factors and quite variable within and between individuals. This variability may account for some of the discrepancies in DMPA studies and highlight the need to interpret the results of our study in the context of future research among different study populations and exploring individual level factors that may account for variable responses.

A strength of our study design is that we captured two time points over the course of four weeks both before and after implant placement to capture the overall environment given changes over a cycle with endogenous hormonal exposure. Given our small sample size, these results should be interpreted cautiously. While our sample size limits our power to evaluate subtle immunologic changes, the changes that we do identify highlight the need for larger, more robust studies to determine if these changes influence a woman’s susceptibility to HIV infection. The longitudinal nature of this study allows us to control for measured and unmeasured biases that occur in cross sectional studies that are the predominant study type in the field. Although we excluded women from participation with clinical evidence of any infection at baseline, several women had asymptomatic infections diagnosed or acquired infections over the course of the study. As individuals with sexually transmitted infections, bacterial vaginosis, and recent semen exposure, factors independently known to alter HIV susceptibility, were not excluded from this analysis, but rather the time-varying presence of these exposures were controlled for in our final models, we feel these findings are likely more representative of real-world findings. Although BV, STIs and semen may modify HIV susceptibility, larger studies are needed for adequate power to analyze the potential effect modification of these risk factors. Although heterogeneity in the endogenous hormonal response to the contraceptive is possible and we did not measure and control for endogenous hormonal levels, we selected to include 2 time points post initiation to help control for some of that variability. As there are known variations in local immune factors with these infections, the inclusion of these women may have contributed to reduced power for detecting a difference in some study outcomes. As women are self-selected, individual differences that could underlie differential responses to contraceptive exposure may limit the generalizability of our results. Importantly, given the wide range of variability in the number of cells from the CVL that are collected, we are evaluating the proportion of cells expressing different cellular markers and not the number of total cells present that express these markers. Lastly, as we are also reporting on markers of HIV susceptibility, any extrapolation to qualify the degree that these factors may alter true susceptibility is limited.

There are multiple benefits of contraceptives beyond fertility-control including reduced abortion, maternal and neonatal morbidity and HIV perinatal transmission. Our findings relating Eng-Implant with HIV susceptibility markers are subtle with unclear clinical impact, and consistent with the results ECHO trial findings. Informed decision-making must include information about the superior typical-use effectiveness of long-acting reversible contraceptive methods, such as the Eng-Implant. Additionally, informed consent requires that we share information on the lack of clear evidence on increased risk of HIV susceptibility with all hormonal contraceptive methods with the promotion of dual method use with condoms.

(DOCX)

Click here for additional data file.

(DOCX)

Click here for additional data file.

29 Oct 2019

PONE-D-19-23634

Impact of etonogestrel implant use on T-cell and cytokine profiles in the female genital tract and blood

PLOS ONE

Dear Dr Haddad,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

ACADEMIC EDITOR:

We would appreciate receiving your revised manuscript by Dec 13 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Manish Sagar, MD

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

Additional Editor Comments:

Please see comments below from the reviewers.

As an editor, I am a bit worried about using cells from vaginal lavage. Often cell numbers are low especially live cells. This can make it difficult to ascertain phenotypes with certainty

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

Reviewer #3: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This manuscript reports the results from a prospective study investing the potential risk of etonogestrel implant use on change of T-cell and cytokine profiles in the female genital tract and blood. I have below comments or questions.

In Statistical Methods, for analysis of lymphocyte memory cells with four group types, please make it clear if each subject would have certain % of each of the 4 types of memory cells at each time point or each subject would only have one type of the 4 types of memory cells at each time point.

Per the recent publication (Nature Communications 10, Article number: 3753 (2019)), semen may have effect to alter the risk of HIV acquisition. It would be informative in this study to add some analysis to investigating the potential interaction effect between semen presence/absence and etonogestrel implant use.

In Tables 2 and 3, are the p-values the FDR adjusted p-values or the original p-values?

Please provide more information in a table to show the % and N at each time points for semen presence, STI, blood and BV.

Reviewer #2: Please justify the use of vaginal lavage samples for cell phenotyping, and provide cell numbers and cytokine concentrations in addition to percent change.

Explain how your data relates to data from DMPA studies; vaginal immune parameters are highly variable and could account for differences between studies.

Reword conclusions to remove the implication that data from your study suggests that HIV susceptibility is enhanced by Eng treatment.

Reviewer #3: This article describes the results of a prospective, observational study of changes in HIV target cells and cytokine concentrations associated with use of the ENG contraceptive implant. Few data have been published on the association of ENG implant use and HIV risk to date. With implant use increasing rapidly in areas of high HIV prevalence (most notably in sub-Saharan Africa), the study addresses an important gap in the scientific literature on ENG implant use and HIV susceptibility. In their analysis, the authors evaluated changes in immunological markers in cervicovaginal fluid and blood PBMCs and identified associations between ENG implant use and increases in the proportion of CD4+ cells expressing CCR5 but decreases in the CD4:CD8 T cell ratio in the genital tract only. Results of analyses assessing markers of T cell activation were inconsistent in both tissue types and few changes in cytokine concentrations were identified after ENG implant initiation. This study’s strengths lie in its prospective design which accounts for potential variation in immunological markers at different stages of the menstrual cycle and adequate ascertainment and control of important covariates (STIs, BV). The study is limited by its small sample size as only 15 participants had data from all four pre/post implant study visits. Given this limitation and the fact that these data are drawn from an observational design, which could further be confounded by unmeasured factors, results should be interpreted carefully.

Major comments

• Given the observational nature of this study, the authors should avoid causal language in the interpretation of their results, such as how implant use had an “effect on”, “results in” or “led to” changes in immunological markers. It’s highly recommended that this language be amended throughout the paper.

• Additionally, caution should be taken when interpreting changes in immunological markers as changes in HIV risk. Based on the final two sentences in the first paragraph of the discussion, it sounds as if the authors are suggesting that counseling around implant use should be altered based on the results of this study. This may be a step too far without evidence linking implant use with HIV acquisition. If anything, these results suggest that implant use may alter immunologic factors that have been shown to be associated with HIV risk, and highlight the need for further evidence. Please consider revising.

• The introduction could benefit from a clearer statement of the authors’ hypotheses and the biological or epidemiological evidence to support these hypotheses. Is there reason to believe from these studies that ENG exposure would increase HIV risk or alter immunity in the genital tract? Why is this research question worth pursuing aside from the current lack of evidence?

• Clarification of the sampling is needed in the methods. Were all implant users from the parent study included? If not, how were ENG implant users sampled and is there reason to believe there is selection bias from the sampling?

• Please provide rationale for the decision to impute the lowest measured value for BLQ results rather than using an approach that is independent of measured values (e.g half of the LLOQ). Would this not bias your results?

Minor comments

• In the introduction, it is worth noting that all of the studies that were conducted prior to ECHO were observational

• In the Results, consider adding the median/range of number of days from LMP that samples were collected for each pre-implant time point, and median/range number of days from implant insertion for the two post-implant time points. It is unclear whether you were able to collect the samples within the defined target luteal/follicular periods.

• What was the purpose of the models assessing the association between implant status with the CVL visit characteristics? Please clarify.

• In the models assessing the distribution of lymphocyte memory cells, a linear model seems inappropriate with a categorical outcome. Can you clarify what your dependent variable was in these models or further explain your rationale for the linear model?

• Could you please elaborate on the types of STIs diagnosed in the first paragraph of the Results?

• Did occurrence of detecting <100 CD3+ T cells vary significantly by the four visit types? If so, explain how this might influence your results.

• It is interesting that only a decrease in GCSF was noted in the genital tract after implant initiation. A similar reduction in GCSF was also observed in plasma (although not significant after adjusting for multiple comparisons). Consider discussing the role of this cytokine in the discussion and how it may relate to ENG exposure from a biological perspective.

• In the third paragraph of the discussion, the sentence beginning “Progestins can regulate the expression of numerous genes involved in multiple cellular functions…” should have a citation.

• Author group for citation 4 (ECHO paper) needs to be corrected

• It would be helpful to spell out the memory cell classifications in Figures 1 and 2 either in the legend or the description

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: Lisa Haddad review for PLOSOne.docx

Click here for additional data file.

31 Jan 2020

REVIEWER COMMENTS:

Additional Editor Comments:

Please see comments below from the reviewers.

As an editor, I am a bit worried about using cells from vaginal lavage. Often cell numbers are low especially live cells. This can make it difficult to ascertain phenotypes with certainty

While we recognize your concern regarding CVL, we and several others have successfully characterized the FGT from CVL. CVL allows enrichment of target cells positioned at the apical lumen in proximity to exposure with a lower risk of tissue trauma from sampling that would cause bleeding and contamination of phenotyping. These CVL enriched lymphocytes are phenotypically and functionally shown to be comparable to those resident at the underlying tissue. With our methodology we have high viability of the cells (70-90%) and although cell count numbers for leukocytes are low, these counts are within the range of other sampling methodologies. We have added clarification to highlight this in the methods section. (Page 7, 3rd paragraph)

Reviewer #1: This manuscript reports the results from a prospective study investing the potential risk of etonogestrel implant use on change of T-cell and cytokine profiles in the female genital tract and blood. I have below comments or questions.

In Statistical Methods, for analysis of lymphocyte memory cells with four group types, please make it clear if each subject would have certain % of each of the 4 types of memory cells at each time point or each subject would only have one type of the 4 types of memory cells at each time point.

At each time point, for each individual, the cells are classified into 4 sub-types based on their marker expression. This clarification has been added to the methods section. (Page 9, 5th paragraph)

Per the recent publication (Nature Communications 10, Article number: 3753 (2019)), semen may have effect to alter the risk of HIV acquisition. It would be informative in this study to add some analysis to investigating the potential interaction effect between semen presence/absence and etonogestrel implant use.

We recognize that semen, similar to STIs and BV, may modify the impact of the implant on HIV susceptibility, however our study was not powered to be able to evaluate effect modification. We altered our discussion as follows and added this citation to the paper:

“As individuals with sexually transmitted infections, bacterial vaginosis, and recent semen exposure, factors independently known to alter HIV susceptibility, were not excluded from this analysis, but rather the time-varying presence of these exposures were controlled for in our final models, we feel these findings are likely more representative of real-world findings. Although BV, STIs and semen may modify HIV susceptibility, larger studies are needed for adequate power to analyze the potential effect modification of these risk factors.” )(Page 16, 6th paragraph)

In Tables 2 and 3, are the p-values the FDR adjusted p-values or the original p-values?

These are the adjusted p-values. This detail has been highlighted. (page 23 and 25, table 2 and 3)

Please provide more information in a table to show the % and N at each time points for semen presence, STI, blood and BV.

We have added Table 2 with the time varying characteristics at each visit. (Page 22, Table 2)

Reviewer #2: Please justify the use of vaginal lavage samples for cell phenotyping, and provide cell numbers and cytokine concentrations in addition to percent change.

As mentioned above, CVL allows enrichment of lymphocytes positioned at the apical lumen in proximity to exposure with a lower risk of tissue trauma from sampling that would cause bleeding and contamination of phenotyping. These CVL enriched lymphocytes are phenotypically and functionally shown to be comparable to those resident at the underlying tissue. The cellular yield for the CVL procedure can be variable and thus we rely on the percentages to aid in quantifying the comparable phenotype of the cellular populations. We have added detail in our methods section to highlight why we selected CVL for our analyses. 9table 7, 3rd paragraph)

Further we have added in median and IQR cell counts to table 2 and added supplementary tables that include the cytokine concentrations. We believe however that the proportion expressing these markers are more appropriate method of evaluation and have maintained our analyses as such. This approach is consistent with other studies utilizing CVL given variability in the number of cells retrieved with this sampling approach. In our discussion, we comment on this limitation. We do however believe that proportion of cells expressing different phenotypic markers highlight the characteristic quality of the immune response.

Explain how your data relates to data from DMPA studies; vaginal immune parameters are highly variable and could account for differences between studies.

Thank you for this comment. We have added the following to our discussion to address this point:

“Importantly, vaginal immune parameters are influences by many factors and quite variable within and between individuals. This variability may account for some of the discrepancies in DMPA studies and highlight the need to interpret the results of our study in the context of future research among different study populations and exploring individual level factors that may account for variable responses.” (Page 16, 5th paragraph)

Reword conclusions to remove the implication that data from your study suggests that HIV susceptibility is enhanced by Eng treatment.

We agree that the conclusion of enhanced susceptibility with Eng treatment should be restated. Our conclusions were written to highlight that we cannot make this statement. We have rewritten our conclusion to further reduce the potential for overinterpreting our findings: “Our findings relating Eng-implant with HIV susceptibility markers are subtle with unclear clinical impact, and consistent with the results ECHO trial findings.” (Page 17)

Reviewer #3: This article describes the results of a prospective, observational study of changes in HIV target cells and cytokine concentrations associated with use of the ENG contraceptive implant. Few data have been published on the association of ENG implant use and HIV risk to date. With implant use increasing rapidly in areas of high HIV prevalence (most notably in sub-Saharan Africa), the study addresses an important gap in the scientific literature on ENG implant use and HIV susceptibility. In their analysis, the authors evaluated changes in immunological markers in cervicovaginal fluid and blood PBMCs and identified associations between ENG implant use and increases in the proportion of CD4+ cells expressing CCR5 but decreases in the CD4:CD8 T cell ratio in the genital tract only. Results of analyses assessing markers of T cell activation were inconsistent in both tissue types and few changes in cytokine concentrations were identified after ENG implant initiation. This study’s strengths lie in its prospective design which accounts for potential variation in immunological markers at different stages of the menstrual cycle and adequate ascertainment and control of important covariates (STIs, BV). The study is limited by its small sample size as only 15 participants had data from all four pre/post implant study visits. Given this limitation and the fact that these data are drawn from an observational design, which could further be confounded by unmeasured factors, results should be interpreted carefully.

We agree with your last comment. To highlight this point we have specifically added the following sentence to our discussion. “Given our small sample size, these results should be interpreted cautiously” (Page 16, 6th paragraph)

Major comments

• Given the observational nature of this study, the authors should avoid causal language in the interpretation of their results, such as how implant use had an “effect on”, “results in” or “led to” changes in immunological markers. It’s highly recommended that this language be amended throughout the paper.

We have altered the language to reduce causal language

• Additionally, caution should be taken when interpreting changes in immunological markers as changes in HIV risk. Based on the final two sentences in the first paragraph of the discussion, it sounds as if the authors are suggesting that counseling around implant use should be altered based on the results of this study. This may be a step too far without evidence linking implant use with HIV acquisition. If anything, these results suggest that implant use may alter immunologic factors that have been shown to be associated with HIV risk, and highlight the need for further evidence. Please consider revising.

We have revised the text to reduce overinterpretation of the findings while highlighting the need for research in this area.

• The introduction could benefit from a clearer statement of the authors’ hypotheses and the biological or epidemiological evidence to support these hypotheses. Is there reason to believe from these studies that ENG exposure would increase HIV risk or alter immunity in the genital tract? Why is this research question worth pursuing aside from the current lack of evidence?

We recognize that our hypothesis was not clearly stated. We have added to our introduction “The Eng-Implant is a long-acting highly effective progestin-only contraceptive method containing a 3rd generation progestin. Although Eng-Implant has less glucocorticoid activity compared to medroxyprogesterone in DMPA and less endogenous estrogen inhibition, we hypothesize that given the sustained progestin exposure over time, there will still be some immunologic changes within the genital tract to suggest increased susceptibility with use.” (Page 4, 3rd paragraph)

• Clarification of the sampling is needed in the methods. Were all implant users from the parent study included? If not, how were ENG implant users sampled and is there reason to believe there is selection bias from the sampling?

All implant users from the parent study were included. Clarification to methods has been added. (Page 6)

• Please provide rationale for the decision to impute the lowest measured value for BLQ results rather than using an approach that is independent of measured values (e.g half of the LLOQ). Would this not bias your results?

We apologize for not including this detail. For samples below the level for quantification, we used half the lower limit of detection. (Page 10)

Minor comments

• In the introduction, it is worth noting that all of the studies that were conducted prior to ECHO were observational

This has been added (Page 3, 1st paragraph introduction)

• In the Results, consider adding the median/range of number of days from LMP that samples were collected for each pre-implant time point, and median/range number of days from implant insertion for the two post-implant time points. It is unclear whether you were able to collect the samples within the defined target luteal/follicular periods.

All pre-contraceptive samples were collected during the appropriate window described in the study methods, specifically 17 days to onset of next menses and less than 14 days after first day of last menstrual period for luteal and follicular phases, respectively. Visit 3 was conducted at a median of 84 days (Q1: 83. Q3: 86.5 days) post-contraceptive initiation and Visit 4 was conducted at a median of 105 days (Q1:98, Q3: 111) post-contraceptive initiation. This detail has been added to the text. (Page 11, 1st paragraph of results)

What was the purpose of the models assessing the association between implant status with the CVL visit characteristics? Please clarify.

We compared characteristics of the visits before and after implant use to ensure there were no other factors we needed to consider in our analysis that may be confounding our study findings. This has been added to methods (Page 10, 8th paragraph)

• In the models assessing the distribution of lymphocyte memory cells, a linear model seems inappropriate with a categorical outcome. Can you clarify what your dependent variable was in these models or further explain your rationale for the linear model?

The outcomes within each memory cell subtype were proportion of cells with that characteristic profile. This proportion was continuous. (Page 11)

• Could you please elaborate on the types of STIs diagnosed in the first paragraph of the Results?

We have added detail to describe the specific STIs diagnosed. (Page 8)

• Did occurrence of detecting <100 CD3+ T cells vary significantly by the four visit types? If so, explain how this might influence your results.

There was no difference in the occurrence of low cell counts by visit type. This has been clarified in the results (Page 11, 1st paragraph)

• It is interesting that only a decrease in GCSF was noted in the genital tract after implant initiation. A similar reduction in GCSF was also observed in plasma (although not significant after adjusting for multiple comparisons). Consider discussing the role of this cytokine in the discussion and how it may relate to ENG exposure from a biological perspective.

We have added the following paragraph to the discussion:

“ A decrease in GCSF was noted in the genital tract after implant initiation. A similar reduction in GCSF was also observed in plasma (although not significant after adjusting for multiple comparisons). This finding of a small yet significant reduction in G-CSF may signify an alternative pathway associated with altered HIV susceptibility through damaged mucosa. Granulocyte colony-stimulating factor (G-CSF) may induce an inflammatory reaction enhancing neutrophil function. With receptors on granulosa cells, G-CSF has been implicated in ovulation and thus could be downregulated in the setting of ovulation inhibition associated with implant use. GCSF is also associated with wound healing and has been associated with faster healing from genital ulcerations, G-CSF stimulates the proliferation and differentiation of cells that participate in acute and chronic inflammation and immune responses including mature leukocytes, macrophages, and dendritic cells. This potential mechanism of altered immune response should be further explored to determine if clinically significant.” (Page 14)

• In the third paragraph of the discussion, the sentence beginning “Progestins can regulate the expression of numerous genes involved in multiple cellular functions…” should have a citation.

Citations have been added with sentence altered to reflect these relate to gonadal hormones.

• Author group for citation 4 (ECHO paper) needs to be corrected

Citation corrected

• It would be helpful to spell out the memory cell classifications in Figures 1 and 2 either in the legend or the description

This has been added (Page 24 and 26)

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see ourPrivacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool,https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

Lisa Haddad et al. studied the effects of Nexplanon contraceptive implant use on vaginal and systemic immune parameters associated with HIV acquisition. This report is timely due to the current controversy about whether depoprovera (DMPA) and other types of hormonal contraception affect susceptibility to HIV-1 infection. Like DMPA, the active component in Nexplanon, etonogestrel (Eng), is a synthetic progestin that targets the glucocorticoid receptor in addition to the progesterone receptor. This is the first report on the potential effects of Eng on biological factors underlying HIV risk. The investigators used a similar approach to earlier investigations done by others on the effects of DMPA on vaginal and systemic immunity. It should be noted that these studies to date have not reached a consensus on the effects of DMPA on vaginal immunity. [ Byrne et al (2016) detected increased numbers and frequency of CCR5+ CD4+ T cells in endocervical cytobrush samples from women on DMPA, but detected a similar increase during the luteal phase of the menstrual cycle in women that did not use DMPA. They also reported no significant effect on 14 vaginal cytokine concentrations. Smith-McCune et al (2017) detected no increase in the number or frequency of CCR5+ CD4+ T cells in endocervical cytobrush samples, but found increased concentrations of MCP1 and IFN alpha 2 in endocervical mucus. Mitchell et al (2014) found no effect of DMPA on numbers or frequency of CCR5+ CD4+ cells in ectocervical biopsies. Other large cohort studies have detected higher concentrations of RANTES and other chemokines in vaginal fluid from women on DMPA,]

This study reports a significant increase in the percent of T cells expressing CCR5 and the central memory phenotype (CCR7) in cells recovered from cervicovaginal lavage (CVL) fluid of women using Eng implants.

Unlike previous studies, Haddad et al. used lymphocytes recovered from cervicovaginal lavage (CVLs) samples for phenotyping. This is a major limitation because lymphocytes in CVL samples are usually present in low in numbers and have poor viability. In this study, samples with fewer than 100 CD3+ T cells (20%) were excluded from analysis, but actual cell counts were not provided. Flow analysis is not accurate for samples with low cell counts (<20,000 cells). CVL mononuclear cell counts should be provided as the number of HIV target cells is a more meaningful endpoint than % HIV target cells.

As noted above, we have added detail in our methods section to highlight why we selected CVL for our analyses. Further we have added in cell counts and cytokine concentrations in the tables. We believe however that the proportion expressing these markers are more appropriate method of evaluation and have maintained our analyses as such. In our discussion, we comment on this limitation. We do however believe that proportion of cells expressing different phenotypic markers highlight the characteristic quality of the immune response. (Pages 22, 23 and 29)

The investigators reported collecting vaginal samples during the luteal and follicular phase of the menstrual cycle which is important because Byrne et al. reported a significant increase in CCR5+ CD4+ cells during the luteal phase. Did the investigators determine whether this was the case in their study before they combined these samples for the analysis?

We did compare the Luteal and follicular findings and found no significant differences at these time points. Importantly our sample may not have been large enough to find such a difference, as possibly smaller than the effect of the implant. Notably, if there were a difference and we combined the groups, this would have biased our findings towards the null hypothesis. We have added this to our results and discussion sections (Page 11, 1dt paragraph, page 16, 5th paragraph)

Do women using Eng implants have cycling levels of endogenous progesterone and estrogen, and were these included as covariates in the analysis?

In this study we did not have endogenous hormonal concentrations available for comparison. We have added the following to the discussion: Although heterogeneity in the endogenous hormonal response to the contraceptive is possible and we did not measure and control for endogenous hormonal levels, we selected to include 2 time points post initiation to help control for some of that variability (Page 17, 6th paragraph)

Concentrations of cytokines in CVL fluid were also studied, and a modest but statistically significant increase in sCD40L, and decreases in Il-12 and G-CSF were noted. These data were presented as relative percentages (before and after ENG use). It would be helpful to also have the cytokine concentrations to provide a context for the potential physiological significance of these findings.

The cytokine values have been added as appendix tables. (See appendix Tables)

Since vaginal immune parameters are highly variable and there is no clear pattern of differences in vaginal parameters among women using progestin injections or implants for contraception, the differences in vaginal environment variables described in this report that are attributed to Eng use should be interpreted with caution. Conclusions such as: “Eng implant use led to a moderate increase in the availability of HIV target cells in the genital tract” and “ It is unclear if these implant induced changes would be any less safe than other contraceptives with regard to HIV risk” should be reworded.

We have changed the wording throughout the manuscript to minimize the perception of any conclusion regarding increased risk.

21 Feb 2020

PONE-D-19-23634R1

Impact of etonogestrel implant use on T-cell and cytokine profiles in the female genital tract and blood

PLOS ONE

Dear Dr Haddad,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

We would appreciate receiving your revised manuscript by Apr 06 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Manish Sagar, MD

Academic Editor

PLOS ONE

Additional Editor Comments (if provided):

Editor’s comments

1) Please update this link: www.mtnstopshiv.org/node/773.

2) Please provide a citation for the following statement: “These CVL enriched lymphocytes are phenotypically and functionally shown to be comparable to cells resident at the underlying tissue.”

3) Please clarify are these only live cells, as assessed by Zombie staining: “CVLs collected from 53 (79% of all visits) visits contained greater than 100 CD3+ T-cells and were subsequently included in this analysis.”

4) In Table 3 please correct: % of CD4+ CCR5% T-cells expressing: to % of CD4+ CCR5+ T-cells expressing

5) Given the reviewer’s comments, please include the total number of viable CD4+ T cells at the different visits in Table 2.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: (No Response)

Reviewer #2: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: For the Figures 1 and 2, what’s the dependent variable in the linear mixed model? Please provide the model structure of the equation.

Reviewer #2: I appreciate that the authors, in response to reviewers’ comments, provided more details about the limitations of their study, provided more rigorous statistical analyses, and downplayed the potential implications of their results for HIV transmission.

However, I don’t think they adequately addressed the reviewers concerns about the use of CVL samples as a source of genital tract lymphocytes. The two references that were added to bolster the claim that CVLs provide adequate samples (Iyer SS 2017, Swaims-Kolmeier 2016), and others (McKinnon LR 2014, Archary D 2015) clearly indicate that the cell yield in CVLs is very low compared with other sampling techniques (cytobrush, biopsy). In McKinnon, the median CD4 cell count in CVLs was 89 vs. 1,170 for cytobrush samples. Furthermore, I did not find references for the claims (page 17) that cells in CVLs are phenotypically and functionally comparable to resident cells in underlying tissue; there are few CD4 T cells in the normal vaginal epithelium (Pudney J 2005), and cell viability is often an issue with CVL samples because the acidic vaginal pH can kill lymphocytes in minutes (Olmsted SS 2005). These issues remains a major weakness of this paper.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

27 Feb 2020

Editor’s comments

1) Please update this link: www.mtnstopshiv.org/node/773.

We have changed this link to reference a video of the CVL procedure used: https://vimeo.com/224957115/00cb72fed6

2) Please provide a citation for the following statement: “These CVL enriched lymphocytes are phenotypically and functionally shown to be comparable to cells resident at the underlying tissue.”

We have changed the sentence slightly, expanded the text to further clarify the role of luminal cells and added several citations to support these statements. “CVL or luminal cells are not imbedded within the tissue, persist within a harsh environment, and have a reduced cell yield compared with other sampling approaches, but CVL provides an accurate means of tissue resident phenotyping at the site of sexually transmitted infection exposure. In experiments where luminal T cells are analyzed separately from T cells embedded in the tissue, these two populations have been shown to be very similar phenotypically and functionally. 1-6 Microscopically, it has been shown that luminal T cells remain closely associated with the apical face of the epithelium. 7-11Several studies have shown these luminal T cells are viable, capable of recognizing and responding to antigen, and play a critical role in immunity at mucosal sites3,10,12-16. Luminal T cells are sufficient to provide significant protection even when T cells located in the underlying tissues are not present16, thus although you may not find a large number of T cells in CVL, these cells can be critical for barrier protection”.

3) Please clarify are these only live cells, as assessed by Zombie staining: “CVLs collected from 53 (79% of all visits) visits contained greater than 100 CD3+ T-cells and were subsequently included in this analysis.”

Yes, we performed and gated for cell viability following discrimination of single cell lymphocytes using the Biolegend Zombie Yellow™ Fixable Viability Kit as mentioned in the methods thus only considered viable leukocytes for the analyses. We adjusted the above to add that these contained greater than 100 viable CD3+ T-cells.

4) In Table 3 please correct: % of CD4+ CCR5% T-cells expressing: to % of CD4+ CCR5+ T-cells expressing

Thank you for noting this. Correction has been made in both Table 3 and Table 4

5) Given the reviewer’s comments, please include the total number of viable CD4+ T cells at the different visits in Table 2.

We have added this to table 2.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #2: (No Response)

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: (No Response)

Reviewer #2: Partly

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: (No Response)

Reviewer #2: Yes

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: (No Response)

Reviewer #2: Yes

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: (No Response)

Reviewer #2: Yes

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: For the Figures 1 and 2, what’s the dependent variable in the linear mixed model? Please provide the model structure of the equation.

The dependent variable is (continuous) percentage of cells in each mutually exclusive memory cell class.

Reviewer #2: I appreciate that the authors, in response to reviewers’ comments, provided more details about the limitations of their study, provided more rigorous statistical analyses, and downplayed the potential implications of their results for HIV transmission.

However, I don’t think they adequately addressed the reviewers concerns about the use of CVL samples as a source of genital tract lymphocytes. The two references that were added to bolster the claim that CVLs provide adequate samples (Iyer SS 2017, Swaims-Kolmeier 2016), and others (McKinnon LR 2014, Archary D 2015) clearly indicate that the cell yield in CVLs is very low compared with other sampling techniques (cytobrush, biopsy). In McKinnon, the median CD4 cell count in CVLs was 89 vs. 1,170 for cytobrush samples. Furthermore, I did not find references for the claims (page 17) that cells in CVLs are phenotypically and functionally comparable to resident cells in underlying tissue; there are few CD4 T cells in the normal vaginal epithelium (Pudney J 2005), and cell viability is often an issue with CVL samples because the acidic vaginal pH can kill lymphocytes in minutes (Olmsted SS 2005). These issues remains a major weakness of this paper.

Thank you for this comment. We have expanded our response as noted above and hope this is now adequate. “CVL or luminal cells are not imbedded within the tissue, persist within a harsh environment, and have a reduced cell yield compared with other sampling approaches, but CVL provides an accurate means of tissue resident phenotyping at the site of sexually transmitted infection exposure. In experiments where luminal T cells are analyzed separately from T cells embedded in the tissue, these two populations have been shown to be very similar phenotypically and functionally. 1-6 Microscopically, it has been shown that luminal T cells remain closely associated with the apical face of the epithelium. 7-11Several studies have shown these luminal T cells are viable, capable of recognizing and responding to antigen, and play a critical role in immunity at mucosal sites3,10,12-16. Luminal T cells are sufficient to provide significant protection even when T cells located in the underlying tissues are not present16, thus although you may not find a large number of T cells in CVL, these cells can be critical for barrier protection”.

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us atfigures@plos.org. Please note that Supporting Information files do not need this step.

1. Kohlmeier JE, Miller SC, Woodland DL. Cutting edge: Antigen is not required for the activation and maintenance of virus-specific memory CD8+ T cells in the lung airways. J Immunol 2007;178:4721-5.

2. Richter MV, Topham DJ. The alpha1beta1 integrin and TNF receptor II protect airway CD8+ effector T cells from apoptosis during influenza infection. J Immunol 2007;179:5054-63.

3. Bivas-Benita M, Gillard GO, Bar L, et al. Airway CD8(+) T cells induced by pulmonary DNA immunization mediate protective anti-viral immunity. Mucosal Immunol 2013;6:156-66.

4. Ye F, Turner J, Flano E. Contribution of pulmonary KLRG1(high) and KLRG1(low) CD8 T cells to effector and memory responses during influenza virus infection. J Immunol 2012;189:5206-11.

5. Macdonald DC, Singh H, Whelan MA, et al. Harnessing alveolar macrophages for sustained mucosal T-cell recall confers long-term protection to mice against lethal influenza challenge without clinical disease. Mucosal Immunol 2014;7:89-100.

6. Slutter B, Pewe LL, Kaech SM, Harty JT. Lung airway-surveilling CXCR3(hi) memory CD8(+) T cells are critical for protection against influenza A virus. Immunity 2013;39:939-48.

7. Wands JM, Roark CL, Aydintug MK, et al. Distribution and leukocyte contacts of gammadelta T cells in the lung. J Leukoc Biol 2005;78:1086-96.

8. Nakanishi Y, Lu B, Gerard C, Iwasaki A. CD8(+) T lymphocyte mobilization to virus-infected tissue requires CD4(+) T-cell help. Nature 2009;462:510-3.

9. Wu T, Hu Y, Lee YT, et al. Lung-resident memory CD8 T cells (TRM) are indispensable for optimal cross-protection against pulmonary virus infection. J Leukoc Biol 2014;95:215-24.

10. Laidlaw BJ, Zhang N, Marshall HD, et al. CD4+ T cell help guides formation of CD103+ lung-resident memory CD8+ T cells during influenza viral infection. Immunity 2014;41:633-45.

11. Hu Y, Lee YT, Kaech SM, Garvy B, Cauley LS. Smad4 promotes differentiation of effector and circulating memory CD8 T cells but is dispensable for tissue-resident memory CD8 T cells. J Immunol 2015;194:2407-14.

12. Hogan RJ, Zhong W, Usherwood EJ, Cookenham T, Roberts AD, Woodland DL. Protection from respiratory virus infections can be mediated by antigen-specific CD4(+) T cells that persist in the lungs. J Exp Med 2001;193:981-6.

13. Ely KH, Roberts AD, Woodland DL. Cutting edge: effector memory CD8+ T cells in the lung airways retain the potential to mediate recall responses. J Immunol 2003;171:3338-42.

14. Kohlmeier JE, Cookenham T, Roberts AD, Miller SC, Woodland DL. Type I interferons regulate cytolytic activity of memory CD8(+) T cells in the lung airways during respiratory virus challenge. Immunity 2010;33:96-105.

15. Horvath CN, Shaler CR, Jeyanathan M, Zganiacz A, Xing Z. Mechanisms of delayed anti-tuberculosis protection in the lung of parenteral BCG-vaccinated hosts: a critical role of airway luminal T cells. Mucosal Immunol 2012;5:420-31.

16. McMaster SR, Wilson JJ, Wang H, Kohlmeier JE. Airway-Resident Memory CD8 T Cells Provide Antigen-Specific Protection against Respiratory Virus Challenge through Rapid IFN-gamma Production. J Immunol 2015;195:203-9.

3 Mar 2020

Impact of etonogestrel implant use on T-cell and cytokine profiles in the female genital tract and blood

PONE-D-19-23634R2

Dear Dr. Haddad,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Manish Sagar, MD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

12 Mar 2020

PONE-D-19-23634R2

Impact of etonogestrel implant use on T-cell and cytokine profiles in the female genital tract and blood

Dear Dr. Haddad:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Manish Sagar

Academic Editor

PLOS ONE