Impact of a carrageenan gel on viral load of genital human papillomavirus infections in sexually active women: Findings from the Carrageenan‐gel Against Transmission of Cervical Human papillomavirus (CATCH) trial

Previous research has shown that women's use of a carrageenan gel reduces the risk of acquiring genital human papillomavirus (HPV) infections but does not help to clear existing ones. Although gel use may not result in complete clearance, it may decrease the viral load of HPV infections. We tested this hypothesis in the Carrageenan‐gel Against Transmission of Cervical Human papillomavirus (CATCH) randomized controlled trial. Participants of the CATCH study were selected for viral load testing if they had completed the first four study visits and tested positive for HPV42 or HPV51 in at least one of these visits. HPV42 and HPV51 were chosen as they were among the most abundant low‐ and high‐risk types, respectively, in the study sample. We measured viral load with a type‐specific real‐time polymerase chain reaction. Results were displayed using summary statistics. Of 461 enrolled participants, 39 were included in the HPV42 analysis set and 56 in the HPV51 analysis set. The median time between visits 1 and 4 was 3.7 months. The viral load (copies/cell) of HPV42 ranged from <0.001 to 13 434.1, and that of HPV51 from <0.001 to 967.1. The net median change in HPV42 viral load over all four visits was −1.04 copies/cell in the carrageenan and −147 copies/cell in the placebo arm (Wilcoxon rank sum test, p = 0.26). There was no net median change in HPV51 viral load over all four visits in either arm (p = 0.45). The use of a carrageenan‐based gel is unlikely to reduce the viral load of HPVs 42 or 51.


| INTRODUCTION
Carrageenan, a sulfated polysaccharide, has been studied for its anti-HPV (human papillomavirus) activity in vitro, in vivo, and in clinical studies. 1 Results from preclinical studies were consistent with a protective effect of carrageenan against HPV.Carrageenan's potent anti-HPV activity was first demonstrated in cells 2 and later confirmed in a mouse model. 35][6][7][8] In a post hoc substudy of a trial originally designed to test carrageenan's efficacy against human immunodeficiency virus, a lower cervical HPV prevalence in the carrageenan compared to the placebo arm was found at the trial end, but only among the most compliant users (adjusted odds ratio, aOR: 0.62 [95% confidence interval, CI: 0.41−0.94],n = 348). 4ere were, however, no baseline or intermediate measurements to assess HPV status.An observational study reported that a carrageenan-based gel may accelerate clearance of existing HPV infection (aOR: 4.9 [95% CI: 1.60−15.1],n = 75), 5 which may occur due to carrageenan interfering with the interaction between HPV and the basement membrane, 2 thereby reducing the local spread of infectious foci and leading to a decrease in HPV viral load.

Results of the Carrageenan gel Against Transmission of Cervical
Human papillomavirus (CATCH) trial were consistent with a protective effect of a carrageenan-based gel compared to a placebo against the acquisition of HPV in sexually active women (HR: 0.63 [95% CI: 0.49−0.81],n = 429). 8However, among HPV-positive participants at baseline, no HPV clearance effect was observed (HR: 1.03 [95% CI: 0.77−1.38],n = 240).To investigate why carrageenan influenced incidence but not clearance, we hypothesized that carrageenan may lower the viral load (i.e., the local proliferation) of HPV infection(s) but not lead to complete clearance.In this context, the viral loads of HPVs 42 and 51 were quantified and compared between study visits and by arm.

| Analytic sample
In the interest of study efficiency, participants were included if they had attended all of the first four study visits and at least one of their cervicovaginal specimens was positive for HPV42 or HPV51.Samples were restricted to those obtained at the first four visits during which the gel was used most frequently as participants were instructed to use around 5−10 mL of gel every other day for the first month, independent of intercourse, and before/after sexual intercourse for the duration of their participation.
Viral load testing was based on type-specific real-time polymerase chain reactions (PCR) as described previously. 14We tested for viral load of all four samples from each subject with at least one documented (via the linear array assay) HPV42 or HPV51 infection across the four visits.We tested negative samples for these types because type-specific PCR tends to have much lower thresholds of molecular sensitivity than amplification assays based on consensus or degenerate primer designs. 15Therefore, they may detect HPV DNA even in samples that are negative for HPV based on amplification assays.Samples were first tested for the presence of HPV inhibitors using an internal control. 16Then, using a Light Cycler PCR and detection system (Roche Molecular Systems), samples were tested for viral load of the HPV type they were positive for in at least one sample.For each sample, cycle thresholds were compared to those of a titration curve by serial dilution of HPV42 or HPV51 plasmids in 75 ng of human genomic DNA (Roche Diagnostics) in 10 mM Tris-HCl.The total number of cells was estimated from the quantification of β-globin, where one cell was equivalent to two copies of β-globin. 17Viral load was expressed as the number of HPV copies divided by the total number of cells.

| RESULTS
Of the 461 participants randomized in the CATCH trial, 99 had fewer than four visits and were excluded (Figure 1).Among these, 323 and 306 participants were HPV42-negative and HPV51negative (based on the linear array) across visits 1−4, respectively.
Overall, 80 participants were included: 39 in the HPV42 analysis set and 56 in the HPV51 analysis set.Fifteen participants were positive for both HPV types.There were non-zero viral loads in 55.6% (20/36) of samples that were HPV42-negative and 12.3% (9/73) of samples that were HPV51-negative.All samples that were HPV42-or HPV51-positive had a non-zero viral load for the corresponding HPV type.Of note, the prevalence of other HPV types across the four visits is shown in Table S1.
The baseline characteristics of participants are shown in Table 1.The median age was 20.8 years, with median age at first intercourse being 17 years.Participants were mostly Canadian (45.1%), single (82.5%), never smokers (61.3%), had 12−20 lifetime sex partners (30%), had one sexual partner in the past month (51.3%), and had received the HPV vaccine (62.5%).The majority (83.5%) of participants tested positive for any HPV type at baseline.

Table 2 displays the positivity and viral load results of HPVs 42
and 51 across visits.HPV42 positivity ranged from 60.0% to 91.7% in both arms.At baseline, HPV42 positivity was similar between arms (78.6% in the carrageenan vs. 79.2% in the placebo arm), while that of HPV51 was slightly different (83.3% in the carrageenan vs. 71.1% in the placebo arm).Overall and across both arms, HPV42 viral load (copies/cell) ranged from <0.001 to 13 434.1, and HPV51 viral load ranged from <0.001 to 967.1.At baseline, the median HPV42 viral load was higher in the placebo than the carrageenan arm (73.9 vs. 1.5), whereas for the median HPV51 viral, it was higher in the carrageenan than the placebo arm (0.0960 vs. 0.0265).In the carrageenan arm, the prevalence and viral load of both HPV types were lower at visit two than at visit one (HPV42: 60.0% vs. 78.6%,median viral load 0.5 vs. 1.5;HPV51: 66.7% vs. 83.3%,median viral load 0.0141 vs. 0.0960).In the placebo arm, prevalence and median HPV51 viral load was also lower at visit two compared to visit one (63.2% vs. 71.1%,0.0114 vs. 0.0264).F I G U R E 1 Selection of participants for viral load testing in the CATCH trial.Of the 461 participants randomized in the CATCH trial, 362 had completed the first four study visits.Among these, 39 participants were positive for HPV42, and 56 participants were positive for HPV51 at at least one of the four visits based on the linear array.Fifteen participants tested positive for both HPVs 42 and 51 and were included in the two analysis sets.A total of 318 samples were tested for viral load (with 60 samples being tested for both HPV42 and 51): 39 participants in the HPV42 analysis set x 4 samples/ participant plus 56 participants in the HPV51 analysis set x 4 samples/participant, 15 participants x 4 samples/participants, minus 2 samples that were mishandled/invalid.CATCH, carrageenan-gel against transmission of cervical human papillomavirus; HPV, human papillomavirus.

Table 3 displays the median difference in
To the best of our knowledge, this is the first analysis to assess the effect of a carrageenan-based gel on the viral load of HPV infections in sexually active women.Assessing viral load of HPVs 42 and 51 across four study visits (baseline to approximately 3 months postbaseline), a slight decrease in HPV42 viral load was found between visits 1 (baseline) and 2 (1 month later) in the carrageenan arm (when we expect compliance to be highest and gel use most frequent), but this was not maintained over all study visits, and there were no major changes in viral loads of HPV51.Taken together, these results suggest it is unlikely that the use of a carrageenan-based gel reduces the viral load of either HPV type.
The major strength of the current analysis is the study design and the selection criteria of the analytical sample, which allowed us to efficiently compare HPV viral load between participants using a carrageenan gel and those using a placebo gel longitudinally.The analytical sample was based on the first four visits, which would typically occur within 3 months of each other (actual median time between visits 1 and 2, 2 and 3, and 3 and 4 was 0.5, 1.2, and 3.7 months, respectively).We considered this time frame to be optimal to capture an effect of carrageenan if there was one.We chose two of the most prevalent HPV types to assess viral load for study efficiency since the assumed mechanism of protection conferred by carrageenan is not type-specific.HPV42 and HPV51 were also highly prevalent representative types of low-risk and high-risk HPV types, respectively.Carrageenan was previously shown to inhibit HPV42 and HPV51 in vitro. 18In addition, participants' samples were tested for viral load regardless of HPV-positivity, showing that some samples negative for HPV by the linear array had non-zero values for viral load.The conclusions of the study were nonetheless limited by the small sample size; in total, 80 participants contributed 318 samples.
Although no previous studies have assessed viral load in specimens from participants using a carrageenan or placebo gel, there has been research on HPV positivity and related viral load in heterosexual couples 14 and in women with other vaginal infections. 19 heterosexual couples, the median HPV42 viral load was higher than the median HPV51 viral load in both men (HPV42: 31.9 at visit 1 and 24.8 at visit 2, HPV51: 7.6 at visit 1 and 10.3 at visit 2) and women (HPV42: 5.2 at visit 1 and 11.5 at visit 2, HPV51: 3.0 at visit 1 and 0.153 at visit 2). 14Another study that assessed the viral load of HPV16 and HPV52 reported no differences in HPV viral load when stratifying by participants' positivity for different vaginal infections. 19 appears that neither one of these studies 14,19 assessed the viral load of samples that tested negative for HPV based on the linear array assay and a real-time PCR assay, respectively.
Currently, an ongoing feasibility trial to assess the effect of a carrageenan-based mouthwash on oral HPV infection, as opposed to genital HPV infection, is in the recruitment stage. 20This trial will T A B L E 1 Baseline characteristics of participants.T A B L E 2 HPV42 and HPV51 positivity and summary statistics of related viral load (copies/cell) by study arm and visit number.provide more evidence (for or against) carrageenan's anti-HPV effect.
Given that viral load is thought to be related to HPV infectiousness, 21 the findings from the current analysis may encourage additional research in this area, in particular, large-scale studies to assess carrageenan's effect on viral load that include more frequent use of carrageenan.
T A B L E 3 Median difference and net change in HPV42 and HPV51 viral loads (copies/cell) between/over visits.
Summary statistics of viral load were calculated, including the median, interquartile range, range, and geometric mean due to the viral load distributions being right-skewed.Viral loads below the threshold of assay detection were imputed using the lowest value of viral load for that HPV type divided by 2 (16 and 64 samples in the HPV42 and HPV51 analysis sets, respectively).For one participant with missing viral load data at visit 3, the average viral load between neighboring visits was used.The difference in viral load was calculated for each participant by subtracting the viral load at visit 2 from visit 1, where a negative median viral load would indicate a decrease in viral load at that visit compared to the baseline visit.A net change in viral load was also calculated for each participant by summing the difference in viral load between visits 1 and 2, visits 1 and 3, and between visits 1 and 4. The median net change in viral load was then calculated for the carrageenan and placebo arms.The median difference and net change in viral load between/over visits were compared by the study arm using the Wilcoxon rank sum test; p-values < 0.05 were considered significant.Analyses were performed in Stata version 17.
HPV42 and HPV51 viral loads between/over visits.The median difference in HPV42 viral loads between visits 1 and 2 in the carrageenan arm compared to that in the placebo arm was significantly different (−0.04 vs. 2.01 copies/cell, p = 0.03).Between visits 1 and 2, the median difference in HPV51 viral load was similar in the carrageenan compared to the placebo arm (−0.004 vs. 0.00 copies/cell, p = 0.21).The net median change in viral load over all four visits corresponded to a net decrease in HPV42 viral load (carrageenan: −1.04, placebo: −147 copies/cell).This change was not significant between groups (p = 0.26).No net decrease in HPV51 viral load was observed.