Prevalence of human papillomavirus infection by number of vaccine doses among US women

INTRODUCTION

Over a decade after the introduction of the human papillomavirus (HPV) vaccine in the United States (US), only 51.1% of adolescents have completed the vaccine series while a greater number (68.1%) received at least one dose.¹ The suboptimal series completion rate in the US is partly attributable to the barriers, including unawareness of or forgetting the need for additional doses, lack of insurance coverage or health care professional recommendations, and less frequent contact with the medical system.^2,3 To simplify the recommendations, trials are evaluating the efficacy of a single-dose regimen.⁴ Herein, we investigated HPV infection prevalence among women by number of vaccine doses received.

METHODS

We analyzed the National Health and Nutritional Examination Survey (NHANES) (2009–2016) data, which is a cross-sectional, stratified, multistage probability sample of the US population. Demographics and immunization history were self-reported and collected by trained interviewers during a home interview. Sexual behavior data were self-reported by participants in the Mobile Examination Center (MEC). Participants provided self-collected cervicovaginal swab specimens. The specimens were evaluated by polymerase-chain-reaction followed by type-specific hybridization. Details of the survey questionnaire, sample collection, and laboratory methods are available elsewhere.⁵

We identified women aged 18–26 years at the time of survey participation with non-missing HPV vaccination and HPV test data. Nationally representative estimates for prevalence and the representative population counts were computed using NHANES sampling weights. The survey weight-adjusted Wald F-test was used to examine the difference in the prevalence of HPV infection [4-valent vaccine types (HPV-6/11/16/18); cross-protection types (HPV-31/33/45); and other high-risk types (HPV-35/39/51/52/56/58/59/68)] by the number of doses received. The differences in predicted probability for HPV-6/11/16/18 by vaccine dose(s) and by the levels of risk factors were estimated using a multivariable logistic regression model. The model was adjusted simultaneously for age as a linear term, race/ethnicity, age at sexual debut, and lifetime number of male sex partners. Statistical significance was tested at P<0.05. All analyses were performed with SAS 9.4 software (SAS Institute) using SAS PROC SURVEY procedures, which included weight, cluster, and strata statements, to incorporate sampling weights and to adjust for the complex survey design. This study was deemed exempt from review by the institutional review board of the University of Texas Health Science Center.

RESULTS

The study sample included a total of 1620 women of which 1004 were unvaccinated and 616 received at least one dose of HPV vaccine: 106 received 1 dose, 126 received 2 doses, and 384 received 3 doses. The majority (56.5%) of the women were White and the mean age was 22.2 years.

Compared to unvaccinated women (prevalence of 12.5% [95% CI, 9.7% to 15.3%]), HPV-6/11/16/18 infection was significantly lower among women who received one dose (2.4% [CI, 0.0% to 4.9%]), two doses (5.1% [CI, 0.8% to 9.5%]), or three doses (3.1% [CI, 0.9% to 5.3%]) of HPV vaccine (Table 1). There was no significant difference in prevalence for one vs two dose(s) (P=0.12) or one vs three doses (P=0.70). Differences were not statistically significant for cross protection (except for 2 doses vs unvaccinated and 1 dose vs 2 doses) and other high-risk HPV types.

Table 1:

Prevalence of genital HPV infection among unvaccinated and HPV-vaccinated (by number of doses) women aged 18–26 years, NHANES 2009–2016

	Vaccinateda (1 dose)	Vaccinateda (2 doses)	Vaccinateda (3 doses)	Unvaccinated	P Valueb
HPV-6/11/16/18 (4-valent vaccine-type infection)
n/N	4/106	7/126	14/384	111/1004	<0.001C/<0.01d/<0.001e/0.12f/0.40g/0.70h
Weighted Prevalence (%)	2.4 (0.0–4.9)	5.1 (0.8–9.5)	3.1 (0.9–5.3)	12.5 (9.7–15.3)
Weighted n/N	22 459/924 276	61 684/1 200 402	121 940/3 919 600	1 179 961/9 440 387
HPV 31/33/45 (cross-protection types)
n/N	11/106	3/126	22/384	57/1004	0.15c/<0.05d/0.61e/<0.05f/0.11g/0.26h
Weighted Prevalence (%)	10.7 (3.5–18.0)	2.8 (0.0–6.0)	6.3 (3.2–9.4)	5.4 (3.7–7.1)
Weighted n/N	99 328/924 276	33 843/1 200 402	248 469/3 919 600	511 238/9 440 387
Other HR HPV i
n/N	22/106	34/126	109/384	254/1004	0.61c/0.52d/0.61e/0.39f/0.89g/0.84h
Weighted Prevalence (%)	22.7 (12.4–32.9)	28.1 (17.8–38.4)	27.2 (20.1–33.5)	25.2 (21.6–28.8)
Weighted n/N	209 527/924 276	337 309/1 200 402	1 064 296/3 919 600	2 378 757/9 440 387

HPV = human papillomavirus; HR = high-risk; n = participants with infection; N = total survey participants; NHANES = National Health and Nutrition Examination Survey

^aDose information was available for 616 women: 106 received one dose, 126 received two doses, and 384 received three doses.

^bP values for survey weight adjusted Wald F test

^c1 dose vs unvaccinated

^d2 doses vs unvaccinated

^e3 doses vs unvaccinated

^f1 vs 2 dose(s)

^g2 vs 3 doses

^h1 vs 3 doses

ⁱOther HR HPV include type-35/39/51/52/56/58/59/68

In adjusted analysis, the predicted probability (PP) of HPV-6/11/16/18 infection was higher in unvaccinated (7.4% [CI, 7.1% to 7.7%]) women compared with women who received one (2.3% [CI, 1.9% to 2.8%), two (5.7 [CI, 5.1% to 6.2%]), or three (3.1% [CI, 2.9% to 3.4%]) dose(s) (Table 2). Black women had a greater PP (10.8%) of HPV-6/11/16/18 compared to white women (6.6%). The PP was also higher for women with >5-lifetime number of male sex partners (11.6%) than women with 0–5 lifetime number of male partners (3.3%).

Table 2:

Difference in predicted probabilities of HPV-6/11/16/18 (4-valent vaccine-type) infection by risk factor among women, NHANES 2009–2016

Risk Factors	Genital HPV-6/11/16/18 infectiona
	Predicted probability (95% CI), %	Difference in Predicted Probabilityb (95% CI), %
HPV vaccine dose(s)
0 (unvaccinated)	7.4 (7.1 to 7.7)	Reference
1	2.3 (1.9 to 2.8)	−5.0 (−5.6 to −4.5)
2	5.7 (5.1 to 6.2)	−1.7 (−2.4 to −0.1)
3	3.1 (2.9 to 3.4)	−4.3 (−4.6 to −4.0)
Race/ethnicity
White	6.6 (6.3 to 6.8)	Reference
Black	10.8 (10.3 to 11.3)	4.2 (3.7 to 4.8)
Others	5.9 (5.6 to 6.3)	−0.6 (−1.1 to −0.2)
Age at sexual debut, years
<15	7.2 (6.9 to 7.5)	Reference
≥15	6.1 (5.8 to 6.4)	−1.1 (−1.5 to −0.2)
Lifetime number of male sex partners
0–5	3.3 (3.1 to 3.4)	Reference
>5	11.6 (11.3 to 12.0)	8.3 (8.0 to 8.7)

HPV = human papillomavirus; NHANES = National Health and Nutrition Examination Survey; Others = Mexican American, Other Hispanic, or other races, including multiracial

^aModel was simultaneously adjusted for variables in the table and age as a linear term. The model included all women with non-missing data on all listed variables (n=1315)

^bDifferences in predicted probability reflect the risk relative to the reference group adjusted for variables in the model and age as a linear term

DISCUSSION

Our study suggests that US women who received one dose of the HPV vaccine may have gained similar protection against vaccine-type infections compared to those who received additional doses. These findings support previous observational studies and posthoc analyses of vaccine trials that demonstrated comparable effectiveness of one dose to two or three doses.⁴ The limitations of our study include a cross-sectional design that precluded us from evaluating the timing of HPV vaccination compared to potential exposure. Furthermore, any conclusion regarding efficacy by individual number of doses cannot be drawn given the self-reported immunization history that may be prone to recall bias.⁶ If ongoing trials confirm sufficient efficacy and sustained duration of protection from a single-dose regimen, vaccine initiation (as opposed to the series completion) will become a more achievable metric of population coverage.