Abstract

We sought to determine whether oral fluid can be used to assess serum human papillomavirus (HPV) antibody status by enrolling women who had received a prophylactic HPV-16 vaccine in a new follow-up study. After the prophylactic HPV-6/11/16/18 vaccine was licensed in the United States, we administered it to consenting participants. With serologic findings used as the reference standard, The sensitivity of oral fluid was 49.6% (95% confidence interval [CI], 42.0%–57.3%) before and 100% (95% CI, 92.0%–100%) after administration of the quadrivalent vaccine. Oral fluid may have the potential to be used for monitoring of prophylactic HPV vaccines in the future

Oral fluid is viewed as an attractive alternative to serum for antibody detection in epidemiologic studies. In contrast to venipuncture, oral fluid sampling is simple, noninvasive, and painless for the participants. It also does not require medically trained personnel, and participants can collect the specimen on their own [1]. The translocation of immunoglobulin G (IgG) from blood to extracellular fluid occurs most notably in the dental capillary bed, and the transudate can be obtained from fluid lying in the dental-gingival crevice [2]. This serous fluid is called oral mucosal transudate (OMT). OMT is substantially richer in IgG than saliva and constitutes a potentially valuable specimen to reflect the status of serum IgG [2]

Previous studies have shown that OMT human papillomavirus (HPV)–specific IgG levels in natural infection are low and only modestly correlate with serum HPV-specific IgG levels [3–7]. These findings are thought to be due to the dilution of the transudated IgG from serum into the oral fluid. Serum HPV-16 IgG levels induced by prophylactic HPV vaccines are several-fold higher than those induced by natural infection with HPV-16 [8–10]. Therefore, we hypothesized that HPV-16 IgG levels in OMT may strongly correlate with those in serum among vaccinated women. To test this hypothesis we conducted a study among women who had received prophylactic vaccines consisting of HPV-16 L1 viruslike particles

MethodsBetween October 1998 and November 1999, 2391 women were enrolled in a multicenter, double-blind, phase IIb, randomized, controlled trial of a prophylactic HPV-16 L1 viruslike particle vaccine in the United States. Details of that study can be found elsewhere [9]. Of 2391 participants in the trial, 500 women were enrolled in Seattle, Washington. Beginning in February 2006, all 500 women were offered participation in a new extended follow-up study with up to 3 visits occurring every 6 months to assess the long-term efficacy of the monovalent vaccine. The institutional review board of the University of Washington approved the study. One aim of this study, the focus of this report, was to assess the use of OMT in lieu of serum for assessment of HPV-16 IgG among women who had previously received the monovalent vaccine

After the quadrivalent vaccine was licensed in the United States in 2006, we offered it to all study participants. Blood specimen collection began in March 2006, with 10 mL of blood drawn for assessment of HPV-16 IgG in serum. OMT collection began in June 2006. Approximately 0.5–0.8 mL of OMT was obtained for assessment of HPV-16 IgG in oral fluid. An OraSure device (OraSure Technologies) was used to collect OMT specimens. The collection pad from the kit was handed to the participant. The participant was instructed to place the pad between the gum and cheek and rub it back and forth along the gum line until the pad was moist. The pad was left stationary against the gum for 2–5 min and then placed into the liquid in the specimen collection vial for shipment to the study-designated laboratory

Serum and OMT specimens were defrosted, and the liquid was collected by centrifugation (4000 rpm for 5 min at 4°C in an Eppendorf 5810R centrifuge) into 2-mL freezer vials for storage at −70°C until testing. HPV-16 L1 was synthesized by Blue Heron Biotechnology to maximize expression in Escherichia coli. This sequence was subsequently cloned into a modified pGex-4T vector to express L1 proteins with glutathione S-transferase (GST) fused at the N-terminus and an 11-aa epitope fused to the C-terminus. Use of the optimized sequence increased L1 protein expression detected by Western blot analysis; however, the level of L1 protein expression, measured by antibodies that recognized conformation-dependent epitopes, did not increase (data not shown, sequence available on request)

The detection of antibodies to HPV-16 L1 was performed according to the methods of Waterboer et al [11, 12], with modified incubation conditions. Compared with conventional serologic assays, this method requires less time and lower sample volumes without loss of sensitivity [11]. As such, this method is suited for large seroepidemiologic studies in which testing can be conducted on several samples under almost identical conditions. In brief, HPV-16 L1 and BKV VP1 proteins were expressed as GST fusion proteins in Rosetta cells (EMD Biosciences). An epitope-tagged version of GST was also expressed. Cells were lysed by 2 passes through a Microfluidizer (model M-110S; Mirofluidics). Polystyrene microspheres (beads) containing a unique combination of fluorescent dyes (MiraiBio) were covalently coupled with glutathione-linked casein (Sigma). Each protein preparation was bound to a different bead set and incubated for 1 h at room temperature with shaking. The beads were washed 3 times with phosphate-buffered saline containing 1% casein, combined, and diluted in the same buffer (final concentration, 55,000 beads/mL of each type)

Human serum samples were diluted 1:50 in the blocking buffer in polypropylene plates. Oral fluid specimens were diluted 1:2 in the same buffer. After blocking for 1 h at room temperature with shaking, 50 mL of diluted serum or diluted oral fluid was mixed with an equal volume of bead mixture in a 96-well filter plate (MultiScreenHTS; Millipore). Plates were incubated overnight at 4°C. The next day the plates were removed from the refrigerator and incubated with shaking for 1 h before washing. Incubations with biotin–anti-human IgG and streptavidin-phycoerythrin were performed as described elsewhere [11], and plates were read on a Bio-Plex 200 Instrument (Bio-Rad Laboratories) after calibration for the more sensitive readings. Controls included serum and oral fluid specimens [3] from women with no previous sex partners. Antibody reactivity is reported as median fluorescence intensity (MFI). Anti–HPV-16 positivity was defined as an MFI value ⩾2 standard deviations above that of women with no previous sex partners; the cutoff points for serum and OMT anti–HPV-16 positivity were determined to be 2318 and 1262 MFI, respectively

The sensitivity and specificity of OMT testing, with serologic results used as the reference standard, were calculated before and after administration of the quadrivalent vaccine. Because specimens collected before administration of the quadrivalent vaccine were not independent (ie, a single woman may have contributed >1 pair of specimens), confidence intervals (CIs) were calculated using the clustered sandwich estimator of variance. We also sought to determine whether OMT and serum can be used to classify the participants in the monovalent vaccine trial according to their vaccination status. We developed receiver operating characteristic (ROC) curves to determine the sensitivity and specificity of OMT and serum for several cutoff points, treating vaccination status as the reference variable. The ROC analysis was performed using the information obtained at enrollment. All analyses were conducted using Stata software (version 10; Stata)

ResultsA total of 291 women were enrolled between March 2006 and May 2008. Fifteen women came only for 1 visit and before we had started collecting OMT specimens; therefore, their data were excluded from the analyses. Of the remaining 276 participants, 139 women received monovalent vaccine, and 137 received placebo. The mean age of participants in both groups was 29 years (range, 25–33 years). More than 70% of participants in both groups were white

A total of 520 pairs of specimens were collected from 276 participants during visits before administration of the quadrivalent HPV vaccine. A total of 279 pairs of specimens were collected from 139 monovalent vaccine recipients. Values for serum and OMT anti–HPV-16 reactivity in monovalent vaccine recipients were 10.6 and 5.7 times higher, respectively, than those in placebo recipients. Among the monovalent vaccine recipients, the anti–HPV-16 reactivity of serum was approximately 7.7 times higher than that of OMT (Figure 1). In this group, 268 serum specimens (96.1%) and 133 OMT specimens (47.7%) were anti–HPV-16 positive; corresponding serum specimens of all positive OMT specimens (n=133) were also positive, and corresponding OMT specimens of all negative serum specimens (n=11) were also negative. The sensitivity and specificity of OMT testing, with serologic results considered the reference standard, were 49.6% (95% CI, 42.0%–57.3%) and 100% (95% CI, 76.1%–100%), respectively

Figure 1

Reactivity of antibody to human papillomavirus 16 in serum and oral mucosal transudate (OMT) specimens before and after administration of the quadrivalent vaccine

Figure 1

Reactivity of antibody to human papillomavirus 16 in serum and oral mucosal transudate (OMT) specimens before and after administration of the quadrivalent vaccine

ROC analysis indicated that the OMT and serum anti–HPV-16 reactivity cutoff points of 763 and 2045 MFI accurately classified the vaccination status of the highest proportion (79.8% and 89.9%, respectively) of participants in the monovalent vaccine trial. The sensitivity and specificity of OMT for determining the vaccination status corresponding to the cutoff point of 763 MFI were 74.1% (95% CI, 66.7%–81.5%) and 85.5% (95% CI, 80.3%–92.0%), respectively. The sensitivity and specificity of serum for determining the vaccination status corresponding to the cutoff point of 2045 MFI were 98.6% (95% CI, 96.6%–100%) and 81.0% (95% CI, 74.4%–87.7%), respectively. The areas under the curve for OMT and serum were 0.84 (95% CI, 0.79–0.89) and 0.95 (95% CI, 0.92–0.97), respectively (Figure 2)

Figure 2

Receiver operating characteristic (ROC) curves for the performance of oral mucosal transudate (A) and serum (B) in classifying participants in the monovalent vaccine trial by their vaccination status

Figure 2

Receiver operating characteristic (ROC) curves for the performance of oral mucosal transudate (A) and serum (B) in classifying participants in the monovalent vaccine trial by their vaccination status

A total of 145 participants chose to receive the quadrivalent vaccine, of whom 81 (36 monovalent vaccine recipients and 45 placebo recipients) came for 1 visit after vaccination with the quadrivalent vaccine. The mean follow-up time after administration of the first dose of the quadrivalent vaccine was 6.0 months (range, 2.5–7.8 months). After administration of this vaccine, anti–HPV-16 reactivity in serum and OMT specimens rose substantially (Figure 1). All serum and OMT specimens of monovalent vaccine recipients were anti–HPV-16 positive. In this group, the sensitivity of OMT testing for detection of HPV-16 IgG was 100% (95% CI, 92.0%–100%); specificity could not be determined owing to lack of anti–HPV-16–negative serum specimens

DiscussionTo our knowledge, this is the first study to investigate the utility of oral fluid sampling in lieu of venipuncture for the determination of serum antibody status among women vaccinated with prophylactic HPV vaccines. Eight and a half years after vaccination with the monovalent HPV-16 vaccine, OMT was 100% specific, but not highly sensitive, for the determination of serum HPV-16 IgG status. It is conceivable that the levels of transudated IgG into OMT were still not high enough to be detected by the methods used, resulting in a low sensitivity of OMT testing

After administration of the quadrivalent vaccine, there was a several-fold increase in anti–HPV-16 reactivity in serum and OMT. The antibody reactivity was measured within a short period of time (ie, 6 months) after vaccination. The utility of OMT testing for the determination of serum HPV-16 IgG status will likely depend on how long the levels of serum HPV-16 IgG are sustained after vaccination. It has been shown that, a few years after administration of the quadrivalent vaccine, serum HPV-16 IgG titers remain substantially higher than those in natural infection [10, 13]. Therefore, it is conceivable that the OMT HPV-16 IgG status may accurately reflect the serum HPV-16 IgG status for at least a few years after vaccination

This study is subject to some limitations. Antibody responses after administration of the quadrivalent vaccine in some participants may well have been beyond the maximum values that could be reported by the assay. Therefore, it was not possible to calculate accurately the difference or correlation between serum and OMT antibody levels after administration of the quadrivalent vaccine. In addition, our study was limited to a relatively homogenous population of women in Seattle; as the prophylactic HPV vaccines are introduced to different populations, it may be important to evaluate the performance of OMT testing in those groups as well

During the past decade, oral fluid sampling has increasingly gained attention as an alternative to venipuncture in assessing population immunity as well as in designing and evaluating vaccination programs. Widespread application of this method to epidemiologic studies of HPV in vaccinated populations requires a number of developments: the sensitivity of assays needs to be improved, validated positive and negative controls need to be developed and provided, standardized cutoff points for positivity need to be defined, and the adequacy of collected specimens needs to be verified [4, 14, 15]. With such developments, oral fluid collection may have the potential to be used in the monitoring of prophylactic HPV vaccines

References

1
Fitchen
JH
Future applications of oral fluid specimen technology
Am J Med
 , vol. 
102
 (pg. 
21
-
5
)
2
Cordeiro
ML
Turpin
CS
McAdams
SA
A comparative study of saliva and OraSure oral fluid
Ann NY Acad Sci
 , vol. 
694
 (pg. 
330
-
1
)
3
Buchinsky
FJ
Carter
JJ
Wipf
GC
Hughes
JP
Koutsky
LA
Comparison of oral fluid and serum ELISAs in the determination of IgG response to natural human papillomavirus infection in university women
J Clin Virol
 , vol. 
35
 (pg. 
450
-
3
)
4
Cameron
JE
Snowhite
IV
Chaturvedi
AK
Hagensee
ME
Human papillomavirus–specific antibody status in oral fluids modestly reflects serum status in human immunodeficiency virus–positive individuals
Clin Diagn Lab Immunol
 , vol. 
10
 (pg. 
431
-
8
)
5
Marais
DJ
Best
JM
Rose
RC
, et al.  . 
Oral antibodies to human papillomavirus type 16 in women with cervical neoplasia
J Med Virol
 , vol. 
65
 (pg. 
149
-
54
)
6
Passmore
JA
Marais
DJ
Sampson
C
, et al.  . 
Cervicovaginal, oral, and serum IgG and IgA responses to human papillomavirus type 16 in women with cervical intraepithelial neoplasia
J Med Virol
 , vol. 
79
 (pg. 
1375
-
80
)
7
Marais
DJ
Sampson
C
Jeftha
A
, et al.  . 
More men than women make mucosal IgA antibodies to Human papillomavirus type 16 (HPV-16) and HPV-18: a study of oral HPV and oral HPV antibodies in a normal healthy population
BMC Infect Dis
 , vol. 
6
 pg. 
95
 
8
Harper
DM
Franco
EL
Wheeler
CM
, et al.  . 
Sustained efficacy up to 4.5 years of a bivalent L1 virus-like particle vaccine against human papillomavirus types 16 and 18: follow-up from a randomised control trial
Lancet
 , vol. 
367
 (pg. 
1247
-
55
)
9
Mao
C
Koutsky
LA
Ault
KA
, et al.  . 
Efficacy of human papillomavirus-16 vaccine to prevent cervical intraepithelial neoplasia: a randomized controlled trial
Obstet Gynecol
 , vol. 
107
 (pg. 
18
-
27
)
10
Villa
LL
Ault
KA
Giuliano
AR
, et al.  . 
Immunologic responses following administration of a vaccine targeting human papillomavirus types 6, 11, 16, and 18
Vaccine
 , vol. 
24
 (pg. 
5571
-
83
)
11
Waterboer
T
Sehr
P
Michael
KM
, et al.  . 
Multiplex human papillomavirus serology based on in situ–purified glutathione S-transferase fusion proteins
Clin Chem
 , vol. 
51
 (pg. 
1845
-
53
)
12
Waterboer
T
Sehr
P
Pawlita
M
Suppression of non-specific binding in serological Luminex assays
J Immunol Methods
 , vol. 
309
 (pg. 
200
-
4
)
13
Olsson
SE
Villa
LL
Costa
RL
, et al.  . 
Induction of immune memory following administration of a prophylactic quadrivalent human papillomavirus (HPV) types 6/11/16/18 L1 virus-like particle (VLP) vaccine
Vaccine
 , vol. 
25
 (pg. 
4931
-
9
)
14
Cameron
SO
Carman
WF
The use of the OraSure collection device for hepatitis virus testing in health care settings
J Clin Virol
 , vol. 
34(Suppl
  
1):S22–8
15
Nokes
DJ
Enquselassie
F
Nigatu
W
, et al.  . 
Has oral fluid the potential to replace serum for the evaluation of population immunity levels?. a study of measles, rubella and hepatitis B in rural Ethiopia
Bull World Health Organ
 , vol. 
79
 (pg. 
588
-
95
)
Potential conflicts of interest: The University of Washington has received funding from Merck through the University of Washington to support human papillomavirus vaccine research conducted by L.A.K. D.A.G. has served as a consultant and member of the speakers bureau for Merck
Presented in part: Meeting of the American Association for Cancer Research, Denver, Colorado, April 2009 (abstract 2152)
Financial support: National Institute of Allergy and Infectious Diseases (grant R01-AI38383)