Evaluation of the HPV 18 antibody response in Gardasil^® vaccinees after 48 mo using a pseudovirion neutralization assay

Abstract

The pseudo-neutralization assay (PsV) described in the current report allows for the creation of HPV18 pseudovirions in order to evaluate whether the antibody responses elicited following vaccination with Gardasil^® are sufficient to neutralize the activity of these pseudovirions in vitro. The PsV assay evaluates a broader antibody response than the HPV competitive Lumniex Immunoassay (cLIA), which monitors the presence of a single neutralizing epitope. We employed two different approaches to the HPV18-PsV assay: one using standard dilutions of heat inactivated serum from vaccinated subjects, as is typically reported in the literature, and the other using heat inactivated serum from which IgG antibodies were purified and quantitated as an attempt to reduce assay background and achieve a level of quantitation greater than that afforded through simple dilution. Here, we show that after 48 mo, Gardasil^® vaccinated subjects from three groups defined by HPV 18 cLIA titer have detectable HPV18 neutralizing antibodies as measured by either approach in the HPV18-PsV assay. These data support the observed sustained HPV18 protection against persistent infection and disease with the absence of breakthrough cases in Gardasil^® vaccinees and suggests that neutralizing antibodies are present although they may no longer be detectable by cLIA.

Keywords: :

• Gardasil®
• HPV
• pseudovirion
• vaccine

Prophylactic vaccination with GARDASIL^® [quadrivalent HPV type 6/11/16/18 VLP vaccine (qHPV)] has demonstrated significant clinical efficacy for the prevention of persistent infection and disease attributable to vaccine HPV types.1-4 The vaccine's efficacy has been based on detection of HPV DNA in infected tissue specimens and, to date, no serological correlate of protection has been established. Vaccination with qHPV elicits a humoral immune response measured in clinical trials by multiplexed competitive luminex-based immunoassays (cLIA) with HPV-type specific neutralizing monoclonal antibodies. In clinical trials with 4 y follow up, HPV 18 cLIA antibody titers have been found to decline over time,5 in some cases, to levels near or below cLIA assay serostatus set cutoffs. No breakthrough cases, however, of HPV 18 related disease have occurred to date. Here, we attempt to explore this apparent inconsistency by utilizing an HPV 18 pseudovirion neutralization assay (18-PsV) to determine whether subjects who had cLIA detectable HPV 18 antibodies after vaccination maintain non cLIA-detectable antibody levels that are capable of neutralizing HPV 18 over time.

Sera from four groups of subjects vaccinated with the qHPV vaccine were evaluated for the presence of HPV 18 neutralizing antibodies in the current report. Ten subjects per group were randomly selected from the qHPV vaccinated arm of Phase 3 clinical trials in which at least 48 mo of follow up had been completed. Groups were as follows: Sera from subjects whose HPV 18 cLIA titers were: (1) below the lower limit of quantitation (LLOQ) at Day 0, (2) below the LLOQ at month 48 (M48), (3) between LLOQ and the assay serostatus cutoff at M48, or (4) between the serostatus cutoff and the HPV 18 geometric mean titer (GMT) for M48. All sera were heat inactivated and were then evaluated in the HPV 18-PsV assay. The same sera were tested both directly with standard dilutions as well as after IgG purification and subsequent quantitative addition of IgG to each 18-PsV assay.

In brief, 293TT cells were pre-plated in a 96-well cell culture plate with neutralization assay media at a concentration of 30,000 cells/well and incubated for 2–4 h at 37°C. A preparation of HPV 18 pseudovirions was titrated to determine the appropriate innoculum for use in the 18-PsV assay. The HPV18 pseudovirion stock used throughout this study was diluted 1:1000 in neutralization assay media. Sera were prepared at dilutions or concentrations 5× the desired final and were added to diluted HPV 18 pseudovirions to obtain the final assay concentrations in a constant 100ul volume. For assays using dilutions of heat-inactivated sera, neutralization was measured at final dilutions of 40, 200, 400, 2000, 4000 to reflect the dilutions used in the cLIA assays. For assays using sera purified IgG, all IgG specimens were diluted to 2 mg/ml to standardize. Two-fold serial dilutions resulting in purified IgG antibody final assay concentrations of 400 µg/ml through 0.195 µg/ml were used. Serum or purified IgG antibodies and HPV18-PsV were incubated together on ice for 1 h before applying to pre-plated 293TT cells. The cells were incubated at 37°C for approximately 72 h. Cell culture supernatants were then collected and the presence of SEAP was measured via chemiluminescence using Clonetech Great EscAPe Alkaline Phosphatase Detection 2.0 kit and a Dynex Luminometer with glow endpoint Relative Light Unit (RLU) readings at 0.2 s per well. Minimum and Maximum SEAP values were calculated as the mean from two no-sera/no-PsV wells and four PsV-only wells per plate, respectively. The 50% SEAP level was calculated by subtracting half of the difference of the Max SEAP and Min SEAP from the Max SEAP. For heat inactivated serum dilution evaluations, the 50% neutralizing titer is defined as the reciprocal of the highest dilution at which the RLUs are lower than the 50% SEAP level. For purified IgG antibody evaluations, the 50% neutralizing concentration is defined as the lowest concentration at which the RLUs are lower than the 50% SEAP level. All specimens were evaluated at least twice by each method for confirmation and reproducibility of results. If the overall qualitative positive/negative neutralization result was not confirmed in the second test, a third pseudoneutralization assay was performed to confirm the qualitative result for that specimen.

The results from the pseudovirion neutralizing assays performed were that HPV 18 neutralizing antibodies were detected in all four groups of qHPV vaccinees using diluted whole sera. Subjects numbered 1–40 were randomly assigned across the groups. Table 1 contains the 50% neutralization titers of the two replicates of each subject categorized by group. A consensus qualitative positive or negative 18-PsV neutralizing result for each group is summarized at the bottom of the table. Those that require a dilution titer of < 40 to achieve 50% neutralization were determined to be not neutralizing. All of the M48 qHPV vaccinees regardless of titers determined by the cLIA were shown to neutralize HPV 18 PsVs. Three out of 10 samples from day 0 vaccinees with cLIA titers below LLOQ neutralized. In 25 of 40 specimens (62.5%), discordant titer results were observed as highlighted in bold. Additionally, 12 specimens required a third test to achieve the concordant qualitative results recorded. Thus although the HPV 18-PsV appears to have greater sensitivity than the cLIA assay, there are challenges to produce consistent, reproducible results.

Table 1. HPV18-PsV 50% neutralization* titers by group in two replicates per specimen using dilutions of heat inactivated sera

Day 0 < LLOQ		M48 < LLOQ		LLOQ < M48 < Cutoff		Cutoff < M48 < GMT
Pat. #	50% neutralization titer	Pat. #	50% neutralization titer	Pat. #	50% neutralization titer	Pat. #	50% neutralization titer
6	40/40	2	400/400	5	20/200	1	4000/2000
7	< 40/ < 40**	4	40/400	10	4000/200	3	200/400
12	< 40/ < 40	8	200/400	11	4000/4000	18	2000/4000
15	< 40/ < 40	9	4000/400	13	400/400	19	200/4000
20	40/4000	17	4000/4000	14	200/400	21	200/4000
34	< 40/ < 40	22	40/4000	16	200/40	23	400/4000
35	< 40/ < 40	24	200/400	26	4000/400	28	4000/2000
36	< 40/ < 40	25	4000/400	27	4000/2000	31	200/200
38	< 40/ < 40	30	400/2000	29	4000/2000	33	40/40
39	40/40	32	40/2000	37	40/400	40	40/200
	3 of 10 neutralize		10 of 10 neutralize		10 of 10 neutralize		10 of 10 neutralize

* Defined as reciprocal of the highest dilution at which the RLUs are lower than the 50% SEAP level; **A titer with a dilution of < 40 is considered not neutralizing.

HPV 18 neutralizing antibodies were also detected in the IgG antibody purified samples. Concentrations of purified IgG antibodies recovered from the 40 serum specimens ranged from 30.16 to 7.75 mg/ml with a median concentration of 15.75 mg/ml. Table 2 contains the 50% neutralization titers of the two replicates of each subject categorized by group. A consensus qualitative positive or negative 18-PsV neutralizing result for each group is summarized at the bottom of the table. Those that require a concentration of IgG > 400 µg/mL to achieve 50% neutralization were determined to be not neutralizing. Seven of ten qHPV vaccinees whose cLIA titers were below the cLIA LLOQ at M48 scored positive for neutralization. In both groups 3 and 4, M48 qHPV vaccinees with cLIA titers either between the LLOQ and serostatus cutoff or between the serostatus cutoff and study HPV 18 GMT, 9 of 10 scored positive for neutralization. No neutralization was observed in the group 1 vaccinees with cLIA titers below the LLOQ at day 0. In 14 of 40 specimens (35%), discordant neutralizing concentrations were observed as highlighted in bold. Additionally, 15 specimens required a third test to achieve the concordant qualitative results recorded. Thus purifying the IgG antibodies from the serum presumably results in a loss of a subset of neutralizing antibodies as the sensitivity of detecting neutralizing potential via the HPV 18 PsV is decreased compared with the HPV 18-PsV assay run with serum.

Table 2. HPV18-PsV 50% neutralization concentrations (µg/ml)* by group in two replicates per specimen using purified IgG

Day 0 < LLOQ		M48 < LLOQ		LLOQ < M48 < Cutoff		Cutoff < M48 < GMT
Pat. #	50% neutralization concentration	Pat. #	50% neutralization concentration	Pat. #	50% neutralization concentration	Pat. #	50% neutralization concentration
6	> 400/ > 400**	2	400/400	5	100/100	1	100/200
7	> 400/ > 400	4	200/400	10	400/100	3	400/200
12	> 400/ > 400	8	200/200	11	50/12.5	18	400/200
15	> 400/ > 400	9	400/400	13	400/200	19	200/200
20	> 400/ > 400	17	200/400	14	400/100	21	200/200
34	> 400/ > 400	22	200/12.5	16	400/400	23	> 400/ > 400
35	> 400/ > 400	24	> 400/ > 400	26	100/400	28	400/400
36	> 400/ > 400	25	> 400/ > 400	27	> 400/ > 400	31	400/200
38	> 400/ > 400	30	400/400	29	400/400	33	200/200
39	> 400/ > 400	32	> 400/ > 400	37	40/400	40	40/200
	0 of 10 neutralize		7 of 10 neutralize		9 of 10 neutralize		9 of 10 neutralize

* Defined as the lowest concentration tested at which the RLUs are lower than the 50% SEAP level; **A concentration > 400 µg/ml is considered not neutralizing.

Results of testing dilutions of heat-inactivated sera demonstrated that neutralizing antibodies to HPV 18 are present in all qHPV vaccinated subjects 48 mo after vaccination regardless of cLIA titer (Table 1). Even those patients whose cLIA titer was lower than both the cLIA serostatus cutoff and the lower limit of quantitation had detectable neutralizing titers in the 18-PsV assay. However, in the HPV18-PsV assay, at low dilutions of sera, three of 10 patients in the Day 0 group (not yet vaccinated) that were determined to be HPV 18 negative by cLIA (< LLOQ) also scored positive in the HPV 18 PsV.

Using IgG purified from the same sera set in the 18-PsV assay showed that the majority of vaccinated individuals (25 of 30, 83.3%) had detectable neutralizing antibodies regardless of cLIA titer. As was expected, more patients had detectable neutralizing antibody responses via 18-PsV assay if they had a cLIA titer above the LLOQ of the cLIA assay (18 of 20) than those whose cLIA titers were below the LLOQ (7 of 10). In utilizing purified IgG, however, none of 10 patients in the Day 0 group showed neutralizing response correlating well with the recorded negative status determined by the cLIA assay.

When comparing the use of standard heat-inactivated sera vs. heat-inactivated sera purified IgG specimens in the 18-PsV assay, we saw that IgG purification did improve the consistency of the titer results. Of the 40 specimens evaluated, fewer of those which were IgG purified produced discordant titer results between replicates tested, 35% in purified vs 62.5% in heat-inactivated sera. Also, fewer IgG purified sera needed to be repeated in a third test to achieve a consensus qualitative result than those tested with whole serum dilutions. At low dilutions of sera vs. high concentrations of sera purified IgG there were apparent differences in neutralization detection exemplified in the Day 0 specimens. All Day 0 specimens tested were cLIA negative. None of the ten Day 0 specimens showed neutralization with sera purified IgG, however, three of the ten specimens did appear to neutralize with low dilutions of sera. The additional step of IgG purification, removal of sera background activity, may allow for more specificity in the assay. This increase in specificity may, however, be coupled with a decrease in sensitivity. The PsV assay run with serum dilutions allows for any antibody isotype (IgG, IgA, IgM, etc.) to neutralize any neutralizing epitope. The serum purified IgG PsV assay assesses only the IgG antibody isotype neutralization. All of the vaccinated individuals at M48 had a neutralizing response using sera only; while only 25 of 30 had a neutralizing response using sera purified IgG. The sera purified IgG neutralization responses corresponded well with the cLIA titer groups.

Neutralization assays are often referred to as a “gold standard,” as a demonstration of functional antibody response. Because HPV is not easily grown in culture, pseudoneutralization assays have been developed as a surrogate. However, the 18-PsV assay does suffer from some serious limitations. One such limitation is variation and difficulty in standardizing. In 2008, Dessy et al.6 reported 46.22% CV in their attempts to validate the HPV18-PsV assay. We evaluated the %CV over 20 separate assays in wells containing controls for the maximum SEAP (PsV only, no serum/Ab) and minimum SEAP (no PsV, no serum/Ab) secretion in 293TT cells. Over the 20 separate assay plates evaluated, %CV for the maximum SEAP wells was 46.83%. For minimum SEAP wells, the variation was much higher with a CV = 89.95%. Assay variability is also demonstrated in the amount of discordant titer results or concentrations observed within replicates of the same specimens by using either sera or IgG purified 18-PsV methods. The average %CV within a plate for the maximum and minimum SEAP wells across these same 20 assays was less than the variance between assays. Other limitations to the use of the 18-PsV assay include the time it takes to run the assays (over 3 d), difficulty of PsV preparation and standardization and the cost of SEAP detection reagents and equipment. Assays for less-expensive colorimetric detection in this same platform have been developed, but were not evaluated here.

In conclusion, our results show that the majority of vaccinated subjects tested at M48 from all three groups defined by HPV 18 cLIA titer [(1) below end-of-study mean titer, (2) below the cLIA serostatus cutoff, and (3) below the cLIA LLOQ) have detectable neutralizing antibodies to HPV 18]. The question of antibody titers required to ensure neutralizing protection against HPV 18 remains unanswered, mainly due to continued high clinical efficacy of the qHPV vaccine. However, this study draws out the point that the assays used to measure neutralizing antibodies have different properties, measure slightly different subsets of the immune response and have unique sensitivities and specificities and these characteristics must be considered upon evaluating immunoassay data. This data does support the observed sustained HPV 18 protection against persistent infection and disease with the absence of breakthrough cases in qHPV vaccinees and suggests that neutralizing antibodies as measured by the 18-PsV are present although they may no longer be detectable by cLIA.

Acknowledgments

We thank Scott Vuocolo for assistance with the preparation and writing of this manuscript.

Disclosure of Potential Conflicts of Interest

C.C.R. and R.W. are employees of Merck, Sharp and Dohme and may own stock and/or stock options.