Health impact and cost-effectiveness of implementing gender-neutral vaccination with the 9-valent HPV vaccine in Hong Kong

ABSTRACT

Human papillomavirus (HPV) can cause several diseases, including cancers, in both sexes. In January 2020, the Hong Kong government launched a school-based vaccination program for girls 10–12 years of age with the 9-valent HPV (9vHPV) vaccine for the prevention of HPV-related diseases; however, boys were not included. The current study estimated the potential health and economic impact of a routine gender-neutral vaccination (GNV) approach compared with the current female-only vaccination (FOV) strategy. We used a dynamic transmission model, adapted to Hong Kong. The model estimates changes in HPV-related disease incidence and mortality, treatment costs (in 2019 Hong Kong dollars), quality-adjusted life years (QALY), and incremental cost-effectiveness ratios (ICERs) over a 100-year time horizon. The base case analysis compared FOV with the 9vHPV vaccine with routine GNV (coverage rate 70%) for the prevention of HPV-related diseases. Compared with a FOV approach, routine GNV with the 9vHPV vaccine is predicted to provide greater reductions in cumulative HPV-related disease incidence and mortality, as well as lower HPV-related treatment costs. In the base case analysis, the ICER was $248,354 per QALY for routine GNV. As compared with FOV, routine GNV fell below the cost-effectiveness ceiling of $382,046/year for Hong Kong. These results highlight the potential value of a routine GNV program with the 9vHPV vaccine among 12-year-olds in Hong Kong to reduce the public health and economic burden of HPV-related diseases.

Introduction

Human papillomavirus (HPV) can lead to the development of HPV-related diseases, including cervical, vulvar, and vaginal cancers and precancers in women; penile cancer and precancers in men; and anogenital cancers, head and neck cancer, genital warts, and recurrent respiratory papillomatosis (RRP) in both sexes.^1,2 While a declining trend in the incidence and prevalence of HPV infection and HPV-related diseases has been observed since the introduction of HPV vaccinations for girls in many countries,^3,4 the burden of HPV-related diseases among men remains high.⁵ Excluding cervical cancer, HPV-attributable cancers were more frequent in men than in women (approximately 69,000 vs. 55,000 new cases); this was primarily driven by a substantially higher number of oropharyngeal cancer cases (34,000 vs. 8,100).⁶ In addition, the incidence of HPV-related diseases, such as oropharyngeal cancer and anal cancer, have been increasing rapidly among men, further highlighting the unmet need for preventing HPV-related diseases.^7,8 For example, the age-standardized incidence of oropharyngeal cancer in males in the United States (approximately 8.5 cases per 100,000) now exceeds the incidence of cervical cancer in females (approximately 7 cases per 100,000).⁹

In Hong Kong, the incidence of anal cancer (0.4 per 100,000) and penile cancer (0.4 per 100,000) is comparable to other Asian regions (i.e., Japan, South Korea, Malaysia, Singapore, and certain cities/regions of China), but oropharyngeal cancer incidence is higher (2.0 vs. 0.4–1.5 per 100,000) and continues to increase.^10–12 Additionally, the incidence of newly diagnosed genital warts among adult men in Hong Kong (292.2 per 100,000 person-years)¹³ is higher than incidence rates reported in other countries, including Australia, Canada, Germany, Italy, the Netherlands, Spain, the United Kingdom, and the United States.¹⁴ These data highlight the burden of HPV-related disease, which is high among men in Hong Kong.

In addition to cervical cancer screening programs, many countries have implemented HPV vaccination programs to reduce the incidence of cervical cancer and other HPV-related diseases.^15–17 However, screening programs for non-cervical cancers (i.e., oropharyngeal cancers) and HPV vaccination programs for males are limited.¹⁸ Given that males have a high incidence of HPV infection and HPV-related disease and continue to be at risk of subsequent infections,^5,19–21 there is a strong rationale to include men in HPV immunization programs. HPV infection and disease incidence are high in men and remain high through the mid-adult years.²² The burden of oral HPV infection is also higher in men than in women, particularly in older men (i.e., age >50 years).²³ Furthermore, HPV-related oropharyngeal cancer incidence continues to increase among men despite decreasing trends observed for cancers related to smoking and alcohol use (i.e., lung cancer and oral cavity cancers).²⁴ Finally, as antibodies acquired from natural HPV infection do not appear to protect men against subsequent infection or disease, men rarely develop immunity following natural HPV infection.²⁵ Taken together, a gender-neutral vaccination (GNV) approach can directly protect males against HPV infection and HPV-related diseases, and may further reduce HPV-related diseases in females. As such, a number of countries have recently transitioned from female-only vaccination (FOV) programs to a GNV approach in an effort to reduce the incidence and burden of HPV-related diseases and cancers.^17,26,27 Therefore, it is important to determine the public health and economic impact of a GNV approach with the 9vHPV vaccine.

The four most prevalent HPV types associated with cervical cancer in Hong Kong are HPV 16, 18, 52, and 58.^28,29 These HPV types are targeted by the 9-valent HPV vaccine (9vHPV; Gardasil® 9, Merck & Co., Inc., Rahway, NJ, USA), which targets HPV types 6, 11, 16, 18, 31, 33, 45, 52, and 58 to provide protection against approximately 90% of all HPV-related cancers and >90% of genital warts cases.^1,30,31 In January 2020, the Hong Kong government launched a school-based vaccination program for girls in primary school grades 5 and 6 (ages 10–12 years) with the 9vHPV vaccine for the prevention of HPV-related diseases.³² A recent analysis predicted that routine 9vHPV FOV administered to 12-year-old girls along with catch-up vaccination among 13- to 18-year-old females in Hong Kong is expected to reduce the incidence of cervical cancer, anal cancer, cervical intraepithelial neoplasia (CIN) 1, CIN 2/3, vaginal intraepithelial neoplasia 2/3 (VaIN 2/3), and genital warts (both females and males) and can be considered cost-effective.³³ The objective of this modeling analysis was to assess the public health impact and cost-effectiveness of routine GNV compared with routine FOV with the 9vHPV vaccine in Hong Kong.

Materials and methods

Model design

A previously validated and published HPV-type dynamic transmission mathematical model simulating the natural history of HPV infections and estimating the cost associated with HPV-related diseases in the US population^34,35 was adapted to the Hong Kong setting. The model estimates the health and economic impact of 9vHPV routine GNV, with or without catch-up vaccination, compared with 9vHPV FOV.

The model structure and assumptions have been previously described in detail.^34,35 Briefly, individuals enter the model when they are born at a gender-specific and sexual activity-specific rate and move between successive age groups at an age- and gender-specific rate per year. These individuals then exit the model when they die (also at an age- and gender-specific per capita death rate per year). An additional age- and stage-dependent death rate was applied to patients with cervical cancer, but additional risk of death was assumed for those with other HPV-related diseases (i.e., CIN 1/2/3, vaginal/vulvar cancers, VaIN 2/3, vulvar intraepithelial neoplasia, anal cancer, head and neck cancer, penile cancer, or genital warts). HPV transmission and the occurrence of CIN, cervical cancer, external genital warts, and other HPV-related disease was also simulated in the model. A similar natural history structure was followed for progression of individuals from acquisition of an infection to disease, as assumed in previous models for HPV 16/18.³⁶ The patient population within the model was divided into distinct epidemiological categories based on each patient’s status with respect to infection, disease, screening, and treatment.^34,35

Target population

The target population for 9vHPV routine GNV and 9vHPV FOV is adolescents aged 12 years in primary school. The target population for 9vHPV routine GNV plus catch-up vaccination is girls and boys aged 13–18 years.

Scenarios

In the base case, scenarios that were assessed in this analysis are: (i) 9vHPV FOV (12-year-old girls only), (ii) 9vHPV routine GNV (12-year-old girls and boys) versus 9vHPV FOV, and (iii) 9vHPV routine GNV (12-year-old girls and boys) plus catch-up vaccination (13- to 18-year-old females and males) versus 9vHPV FOV.

Model compartments

The age-structured mathematical model³⁴ comprised demographic variables describing the age structure of the population, behavioral variables describing sexual activity, and epidemiologic variables describing transmission of HPV infection and any HPV-related diseases.

Model inputs

Input variables of the epidemiological model are divided into demographics, sexual behavior, disease and treatment patterns, and screening. A detailed summary of these parameter groups is provided in Table 1 and S1 Table along with the data sources.

Table 1.

Model input parameters for disease, treatment, and screening variables.

Input parameter		Data source
Disease variables
Population level hysterectomy rates by age group, % per year
15–24	0.0005%	Hong Kong Special Administrative Region: Census and Statistics Deparatment37
25–29	0.0037%
30–34	0.0128%
35–39	0.0335%
40–44	0.1081%
45–54	0.1828%
≥55	0.0981%
Treatment variables, % of femalesa
CIN/CIS cases treated
CIN 1	3.8%	Hong Kong Special Administrative Region: Census and Statistics Deparatment37
CIN 2	95.5%
CIN 3	95.5%
CIS	95.5%
VaIN/CIS cases treated
VaIN 1	30%	Model calibration
VaIN 2/3	100%
CIS	100%
Genital warts treated
Male	75%	Model calibration
Female	75%
Screening variables
Cervical cancer screening
Percent of females receiving a follow-up screening test after abnormal PAP smear diagnosis	82.5	Hong Kong Center for Health Protection38
Percent of females receiving gynecological cancer screening tests at least once every 3 years	13.04	Hong Kong Department of Health39
Percent of females screened for cervical cancer in the past year by age group, year
0–17	0	Hong Kong Center for Health Protection;38 Hong Kong Department of Health39
18	0
19	0
20–24	0
25–26	2.34
27–29	2.34
30–34	4.17
35–39	4.59
40–44	4.87
45–49	4.81
50–54	4.86
55–59	4.75
60–64	3.83
65–69	0.45
70–74	0.45
75–79	0.45
80–84	0.45
>85	0.45
Performance of diagnostic PAP screening and colposcopy
PAP specificity	0.94	Elbasha et al. 201035
Colposcopy sensitivity	0.96
Colposcopy specificity	0.48
CIN Screening
Performance of diagnostic PAP screening by CIN stage
CIN 1	0.28	Elbasha et al. 201035
CIN 2	0.59
CIN 3	0.59

CIN, cervical intraepithelial neoplasia; CIS, carcinoma in situ; VaIN, vaginal intraepithelial neoplasia.

^aTreatment variables presented as the percentage of the female population with cervical cancer recognizing their symptoms and seeking treatment.

Demographic and sexual behavior variables

Population figures and annual all-cause mortality for the general population by sex and age were retrieved from Hong Kong-specific data and are summarized in detail in S1 Table. An overall Hong Kong population of 7,336,600 individuals was included in the model, which comprised 3,961,200 females and 3,375,400 males.⁴⁰ Sexual behavior variables were also retrieved from Hong Kong-specific data when available;^41,42 these included the annual mean number of sexual partners, the proportions of the population at low (mean number of sexual partners per year of 0–1), medium (2–4 sexual partners per year), and high (5 or more sexual partners per year) sexual activity risk, and the annual mean number of sexual partners by sexual activity risk (S1 Table).

Disease and treatment patterns

Model input variables to simulate the natural history of the disease as well as disease and treatment patterns utilized Hong Kong-specific data when available or were estimated through model calibration (Table 1).

The crude incidence rate of cervical cancer in Hong Kong was 12.7 per 100,000 person-years in 2015.⁴³ Incidence rates of vaginal/vulvar cancer, head and neck cancer (both sexes), anal cancer (both sexes), and penile cancer were obtained from published sources.^37,44–46

The crude annual mortality rate of cervical cancer in Hong Kong was 4.3 per 100,000 person-years in 2015.⁴³ Annual mortality rates for each HPV-related cancer, stratified by age and stage (local, regional, and distant disease), were obtained from published sources.^44–46

The proportion of females receiving a follow-up screening test after abnormal PAP smear diagnosis, age-specific cervical cancer screening tests in the past year, and the proportion of females receiving gynecological cancer screening tests at least once every 3 years were estimated based on Hong Kong-specific data (Table 1).^38,39 The performance of diagnostic cervical screening and of PAP screening for CIN are also summarized in Table 1.

The proportions of the populations with HPV-related diseases (including cervical cancer and precancerous lesions [CIN 1/2/3 and carcinoma in situ (CIS)], vaginal cancer and precancerous lesions [VaIN 1/2/3], vulvar cancer, anal cancer [both sexes], head and neck cancer [both sexes], penile cancer, and genital warts [both sexes]) recognizing their symptoms and seeking treatment, and the proportions who were treated, were estimated through model calibration (Table 1). Annual Hong Kong hysterectomy rates were obtained (Table 1).³⁷

The prophylactic efficacy or the degree of protection offered by the vaccine was based on clinical trial data.^47–52 The duration of protection against HPV genotypes 6, 11, 16, 18, 31, 33, 45, 52, and 58 was assumed to be lifelong (except for the 20-year duration of protection in the sensitivity analysis), and herd immunity was assumed. The use of a dynamic model structure allows herd immunity to be taken into consideration (i.e., an indirect protection of non-vaccinated people by the vaccinated people due to a reduction of the circulating virus).

Vaccine strategies and coverage rates

Assumed vaccine coverage rates (VCRs) for the scenarios assessed in this analysis were 1) 70% VCR for 9vHPV FOV (12-year-old girls only), 2) 70% VCR for 9vHPV routine GNV (12-year-old girls and boys), and 3) 70% VCR for 9vHPV routine GNV (12-year-old girls and boys) and 30% VCR for catch-up vaccination (13- to 18-year-old females and males). The 70% VCR was chosen for 9vHPV FOV and 9vHPV routine GNV vaccination strategies because these routine strategies were assumed to be carried out in a school-based manner. Furthermore, an interim target VCR of 70% was set by the Hong Kong government after the introduction of the female school vaccination program.⁵³ The 30% VCR was chosen for the catch-up vaccination strategy because the HPV catch-up program is more likely to be a public-private partnership with private clinics rather than school-based, and as a result, a lower vaccine uptake rate is expected.

Cost inputs

Cost parameters used in the model included cost per episode of care, cost of vaccination, and cost of screening and diagnostic tests; costs were reported in 2019 Hong Kong dollars (HKD). A detailed summary of all costs used in the model is provided in Table 2.

Table 2.

Cost parameters for HPV-related diseases from the healthcare payer perspective.

Parameter	Cost (2019 HKD)	Data source
Costs for diagnosis and treatment per episode of care
Female-only conditions
Cervical cancer, local disease	225,790	Private hospital costs
Cervical cancer, regional disease	267,790
Cervical cancer, distant disease	80,000
CIN 1	5,790
CIN 2	18,790
CIN 3	18,790
Vaginal cancer, local disease	162,945
Vaginal cancer, regional disease	169,870
Vaginal cancer, distant disease	50,000
VaIN 2	18,790
VaIN 3 and CIS	18,790
Vulvar cancer, local disease	131,340
Vulvar cancer, regional disease	193,030
Vulvar cancer, distant disease	50,000
Male-only conditions
Penile cancer, local disease	157,325
Penile cancer, regional disease	196,000
Penile cancer, distant disease	130,000
Gender-neutral conditions
Anal cancer, local disease	245,260
Anal cancer, regional disease	258,235
Anal cancer, distant disease	12,000
Head and neck cancer, local disease	190,220
Head and neck cancer, regional disease	310,220
Head and neck cancer, distant disease	60,000
Genital warts	1,110	Social Hygiene Clinic – Department of Health HKSAR
RRP	214,952	Private hospital costs
Cost of 9vHPV vaccination
Per dosea	1,390	Private hospital costs
Screening and diagnostic tests (female-only)
Screening (test and visit)	1,210	The Family Planning Association of Hong Kong41
Colposcopy	4,080	Private hospital costs Hong Kong Special Administrative Region: Census and Statistics Deparatment37
Biopsy	4,580

CIN, cervical intraepithelial neoplasia; CIS, carcinoma in situ; HKD, Hong Kong Dollar; HPV, human papillomavirus; RRP, recurrent respiratory papillomatosis; VaIN, vaginal intraepithelial neoplasia.

^aThe cost per dose includes the cost of the vaccine and the administrative fee.

The base model adopted the public price of the 9vHPV vaccine in Hong Kong ($1,390 HKD). The cost per dose includes the cost of the vaccine and the administrative fee.

The estimated costs per episode of care of each HPV-related disease, defined as the cost of management from diagnosis to resolution of the case, are summarized in Table 2. Input costs were derived from prices charged to a private case in a public hospital in Hong Kong (data on file). It is assumed that the private charges are equivalent to the cost incurred to the public healthcare provider of the government of Hong Kong. Costs for screening by PAP smear and cytology with specialist consultation were set at $1,210 HKD, colposcopy with histopathological exam at $4,080 HKD, and colposcopy-guided biopsy with histopathological examination at $4,580 HKD.

All costs are from a healthcare payer perspective. A discount rate of 3% for both costs and quality-adjusted life years (QALYs) was applied to the model.

Health-related quality of life

The health utility values for the Hong Kong general population and for those with HPV-related diseases were derived from Elbasha et al. (2010) (S2 Table).³⁵

Health and economic model outputs

The base case analysis compared routine FOV with the 9vHPV vaccine with two primary scenarios: (i) routine GNV with the 9vHPV vaccine and (ii) routine GNV with the 9vHPV vaccine plus catch-up vaccination. Health and economic model calculations considered the prevention of all HPV-related diseases, including head and neck cancers and penile cancer.

The mathematical model simulated and tracked various outcomes, including the cumulative incidence and number of prevented cases of, and number of averted deaths attributable to, HPV-related diseases, including precancers (i.e., CIN 1, CIN 2, CIN 3, VaIN 2/3), cervical cancer, vaginal/vulvar cancer, anal cancer, head and neck cancers, penile cancer, and genital warts.

The model also estimated the cumulative reductions in HPV-related treatment costs and QALYs. The incremental cost-effectiveness ratio (ICER) was estimated as the ratio of incremental vaccination costs to incremental QALYs. Cost-effectiveness was determined at a ceiling threshold of $382,046 HKD or 1.0 times the gross domestic product (GDP) per capita of Hong Kong in 2019; this threshold reflected the maximum willingness of decision-makers to pay for an additional QALY.⁵⁴

Model calibration and validation

Calibration was performed to ensure consistency between epidemiologic model inputs and real-world observations. The targets of the calibration were data on incidence and mortality rates of HPV-related diseases. The predictive validity of the model was assessed by calibrating model predictions to the observed data on the incidence and mortality of cervical cancer in Hong Kong (see Table 1) that was attributable to HPV types 16 and 18 as well as 31, 33, 45, 52, and 58. The model was also calibrated to the incidence rate of genital warts estimated from available data.

The model was calibrated using cervical cancer recognition and transmission probability parameters as well as screening rates. At the beginning of the calibration process, all the parameters were set to the values from the US model.³⁴ Then, cervical cancer recognition parameters and screening rates were adjusted to converge on the observed ratio between the cervical cancer incidence rate and the cervical cancer mortality rate. Subsequently, the per sexual partnership transmission probability of HPV types 16 and 18 was adjusted. Finally, the per sexual partnership transmission probability of HPV types 31, 33, 45, 52, and 58 was adjusted to achieve the correct incidence and mortality attributed to these five types.

Sensitivity analysis

Deterministic one-way sensitivity analyses were conducted to assess the sensitivity of ICER values to the inclusion of HPV-related head and neck cancers but exclusion of HPV-related penile cancer. Additional sensitivity analyses were conducted to assess other variables that have been shown to be impactful to cost-effectiveness, including VCR of 30%, 60%, and 80%, treatment costs±10%, vaccine price±10%, QALY±5%, and the percentage of all cervical cancers attributed to HPV types 16, 18, 31, 33, 45, 52, and 58.

Horizon

Outcome measures were reported over a time horizon of 100 years because this was an appropriate time frame from which the system approached steady state and most benefits and costs of vaccination could be realized. The vaccination programs as well as current cancer screening programs were assumed to be in effect for the entire 100 year time horizon.

Results

HPV-related disease incidence

Results for the base case analysis of HPV-related disease incidence including head and neck cancer and penile cancer are reported in Table 3. Compared to routine FOV with the 9vHPV vaccine, the model predicted that routine GNV with the 9vHPV vaccine would avert an additional 1,013 cervical cancer (4.5% reduction), 182 CIN 1 (7.4% reduction), 643 CIN 2/3 (6.9% reduction), 106 female anal cancer (4.6% reduction), 334 male anal cancer (14.6% reduction), 129 female head and neck cancer (4.8% reduction), 1,193 male head and neck cancer (16.2% reduction), 569 penile cancer (24.9% reduction), 38,982 female genital warts (33.2% reduction), and 196,515 male genital warts (51.1% reduction) cases over 100 years. The model further predicted that adding catch-up vaccination to routine GNV with the 9vHPV vaccine would lead to even greater disease reductions.

Table 3.

Cumulative reduction in HPV-related disease incidence and cost for routine 9vHPV GNV with or without catch-up vaccination versus 9vHPV FOV in Hong Kong over 100 years (base case analysis).

HPV-related disease	Number of cases averted (% reduction) versus 9vHPV FOV
	Routine 9vHPV GNVa	Routine 9vHPV GNVa plus catch-up vaccinationb
Anal cancer (females)	106 (4.6)	167 (7.3)
Anal cancer (males)	334 (14.6)	390 (17)
Cervical cancer	1,013 (4.5)	1,818 (8.1)
CIN 1	182 (7.4)	308 (12.5)
CIN 2/3	643 (6.9)	1,090 (11.7)
Genital warts (females)	38,982 (33.2)	45,921 (39.1)
Genital warts (males)	196,515 (51.1)	210,691 (54.8)
Head and neck cancers (females)	129 (4.8)	197 (7.4)
Head and neck cancers (males)	1,193 (16.2)	1,370 (18.6)
HPV 6/11-related CIN 1	147 (30.9)	168 (35.4)
Penile cancer	569 (24.9)	606 (26.5)
Vaginal cancer	2 (3.2)	5 (6.4)
VaIN 2/3	3 (3.9)	5 (7.5)
Vulvar cancer	5 (3.1)	10 (6.3)
Total cost savings from reductions in cumulative incidence, 2019 HKD	$161,931,828 (4.3)	$242,362,835 (6.5)

CIN, cervical intraepithelial neoplasia; FOV, female-only vaccination; GNV, gender-neutral vaccination; HKD, Hong Kong Dollar; HPV, human papillomavirus; VaIN, vaginal intraepithelial neoplasia.

^a12-year-old girls and boys with 70% coverage.

^b13- to 18-year-old females and males with 30% coverage.

For anal cancer, genital warts, and head and neck cancers, routine GNV generally produced greater decreases in prevalence in males than in females (Table 3). Figure 1 depicts the reduction over time in HPV 16/18-related head and neck cancer (Figure 1a), HPV 16/18-related penile cancer (Figure 1b), and HPV 16/18/31/33/45/52/58-related anal cancer (Figure 1c) cases among males for routine GNV with the 9vHPV vaccine with or without catch-up vaccination, compared to routine FOV with the 9vHPV vaccine over 100 years. Greater reductions in the incidence of these cancers were observed for routine GNV with the 9vHPV vaccine with or without catch-up vaccination versus routine FOV with the 9vHPV vaccine.

Figure 1.

Estimated incidence of selected HPV-Related cancers among males under three scenarios: 9vHPV Routine GNV, 9vHPV routine GNV plus catch-up vaccination, and 9vHPV FOV over 100 years. (a) HPV 16/18-related head and neck cancer among males; (b) HPV 16/18-related penile cancer among males; (c) HPV 16/18/31/33/45/52/58-related anal cancer among males.

HPV-related disease mortality

As a result of reductions in the cumulative incidence of HPV-related diseases, the model predicted that routine GNV with the 9vHPV vaccine would result in cumulative reduction of HPV-related deaths over 100 years (Table 4), with the greatest reduction in deaths from head and neck cancers (412 for males and 37 for females), cervical cancer (335), anal cancer (67 for males and 33 for females), and penile cancer (58). The model further predicted that adding catch-up vaccination to routine GNV with the 9vHPV vaccine would lead to even greater reductions in HPV-related deaths.

Table 4.

Cumulative reduction in HPV-related deaths for routine 9vHPV GNV with or without catch-up vaccination compared with 9vHPV FOV in Hong Kong over 100 years (base case analysis).

HPV-related disease	Number of deaths averted (% reduction) versus 9vHPV FOV
	Routine 9vHPV GNVa	Routine 9vHPV GNVa plus catch-up vaccinationb
Anal cancer (females)	33 (4.4)	52 (7.0)
Anal cancer (males)	67 (13.9)	79 (16.3)
Cervical cancer	335 (4.2)	608 (7.5)
Head and neck cancers (females)	37 (4.5)	57 (7.0)
Head and neck cancers (males)	412 (15.2)	475 (17.6)
Penile cancer	58 (23.2)	62 (24.8)
Vaginal cancer	1 (3.0)	2 (6.1)
Vulvar cancer	2 (3.0)	3 (6.0)

FOV, female-only vaccination; GNV, gender-neutral vaccination; HPV, human papillomavirus.

^a12-year-old girls and boys with 70% coverage.

^b13- to 18-year-old females and males with 30% coverage.

For anal cancer and head and neck cancers, the number of deaths averted was greater in males than in females (Table 4). Figure 2 shows the reduction over time in HPV 16/18-related head and neck cancer (Figure 2a), HPV 16/18-related penile cancer (Figure 2b), and HPV 16/18/31/33/45/52/58-related anal cancer (Figure 2c) deaths among males for routine GNV with the 9vHPV vaccine with or without catch-up vaccination versus routine FOV with the 9vHPV vaccine over 100 years. Larger reductions in mortality from these cancers were evident for the routine GNV strategy, compared with a FOV approach.

Figure 2.

Estimated HPV-Related cancer deaths among males under three scenarios: 9vHPV routine GNV, 9vHPV routine GNV plus catch-up vaccination, and 9vHPV FOV over 100 years. (a) HPV 16/18-related head and neck cancer deaths among males; (b) HPV 16/18-related penile cancer deaths among males; (c) HPV 16/18/31/33/45/52/58-related anal cancer deaths among males.

HPV-related treatment costs

In the base case analysis including head and neck cancer and penile cancer, the model predicted a reduction in HPV-related treatment cost of $161,931,828 HKD when routine GNV with the 9vHPV vaccine was implemented instead of routine FOV with the 9vHPV vaccine (4.3% reduction) and $242,362,835 HKD for routine GNV with the 9vVPV vaccine plus catch-up vaccination (6.5% reduction versus routine FOV with the 9vHPV vaccine) (Table 3).

Figure 3 shows the reduction in estimated HPV treatment costs by HPV genotype for routine GNV with the 9vHPV vaccine versus routine FOV with the HPV vaccine (Figure 3a) or routine GNV with the 9vHPV vaccine plus catch-up vaccination versus routine FOV with the 9vHPV vaccine over 100 years (Figure 3b). These figures show that greater cost savings were predicted for routine GNV with the 9vHPV vaccine with or without catch-up vaccination; treatment-related cost savings were primarily associated with HPV types 6, 11, 16, and 18.

Figure 3.

Estimated treatment costs avoided over 100 years by HPV genotype for (a) routine 9vHPV GNV versus 9vHPV FOV and (b) routine 9vHPV GNV plus catch-up vaccination versus 9vHPV FOV.

HPV vaccination strategy cost-effectiveness

Results of the base case analysis for cost-effectiveness including head and neck cancer and penile cancer are reported in Table 5. Compared with routine FOV with the 9vHPV vaccine, routine GNV with the 9vHPV vaccine without and with catch-up vaccination were associated with incremental costs ($378.82 and $434.82 HKD, respectively) and QALYs (0.00153 and 0.00252 years, respectively). ICERs were highest for cervical cancer only and decreased upon serial addition of each type of cancer (vaginal, vulvar, anal, head and neck, and penile) and genital warts to the model, showing that cost-effectiveness increases when all HPV-related diseases are considered (data not shown). The ICERs for routine GNV with the 9vHPV vaccine without and with catch-up vaccination compared to routine FOV with the 9vHPV vaccine were $248,354 and $172,424 HKD/QALY, respectively.

Table 5.

Cost-effectiveness analysis of routine 9vHPV GNV strategies compared with 9vHPV FOV^a.

Scenario	Costs/person (HKD)	QALYs/person (year)	Incremental costs (HKD)	Incremental QALYs (year)	ICER (HKD/QALY)
Base case analysis (including head and neck and penile cancers)
9vHPV FOVb	$1,968.60	27.25764	–	–	–
Routine 9vHPV GNVc,d	$2,347.42	27.25917	$378.82	0.00153	$248,354
Routine 9vHPV GNV plus catch-up vaccinationd,e	$2,403.42	27.26016	$434.82	0.00252	$172,424
Sensitivity analysis (including head and neck cancer, excluding penile cancer)
9vHPV FOVb	$1,951.85	27.25805	–	–	–
Routine 9vHPV GNVc,d	$2,332.19	27.25953	$380.34	0.00149	$255,987
Routine 9vHPV GNV plus catch-up vaccinationd,e	$2,388.36	27.26053	$436.51	0.00248	$176,154

FOV, female-only vaccination; GNV, gender-neutral vaccination; HKD, Hong Kong Dollar; HPV, human papillomavirus; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life-year.

^aCosts per person include screening, vaccination costs and treatment costs. These costs are per person, not just per person vaccinated.

^b12-year-old girls at 70% coverage.

^c12-year-old girls and boys at 70% coverage.

^dIncremental costs, QALYs, and ICERs with respect to 9vHPV FOV.

^e13- to 18-year-old females and males at 30% coverage.

Costs are presented in 2019 Hong Kong Dollars.

Sensitivity analysis

When a sensitivity analysis was conducted where HPV-related penile cancer was excluded, total reductions in HPV-related healthcare costs were $150,772,649 HKD for routine GNV with the 9vHPV vaccine (4.2% reduction versus routine FOV with the 9vHPV vaccine) and $229,909,438 HKD for routine GNV routine 9vHPV plus catch-up vaccination (6.3% reduction versus 9vHPV FOV; data not shown). This resulted in ICERs of $255,987/QALY for routine GNV with the 9vHPV vaccine versus 9vHPV FOV and $176,154/QALY for routine GNV routine 9vHPV plus catch-up vaccination versus 9vHPV FOV (Table 5). Additional sensitivity analyses are shown in S3 Table and S4 Table.

Discussion

Among men, the burden of HPV-related diseases is high both globally and in Hong Kong, highlighting the need for men to be included in HPV immunization programs. The efficacy, immunogenicity, and safety of the 9vHPV vaccine has been studied previously in Asian subjects.⁵⁵ Here we report the potential public health and economic impact of routine GNV with the 9vHPV vaccine compared with routine FOV with the 9vHPV vaccine in Hong Kong. Using a validated dynamic transmission model, a routine GNV approach with or without catch-up vaccination is projected to reduce HPV-related disease incidence and costs, compared with a FOV approach, and would be cost-effective in Hong Kong. Routine GNV with the 9vHPV vaccine is projected to reduce HPV-related diseases beyond cervical cancer, including vaginal, vulvar, and anal cancers, as well as genital warts; impacting not just the female population but also males who are at an increased risk of burdensome HPV-related diseases, such as head and neck and penile cancers. Additional disease reduction in both males and females would also be achieved with a routine GNV approach plus catch-up vaccination.

The current results indicating greater and earlier protection of both males and females with routine GNV versus FOV in Hong Kong are consistent with previous modeling studies,^56,57 including those that assumed female coverage rates that were higher (i.e., up to 85%) than the current analysis (70%).^3,58–61 A recent study adapted to Spain reported that compared to FOV of 11–12-year-olds at a coverage of 77.8%, routine GNV with the 9vHPV vaccine with the same coverage in girls and 55% in boys of the same age was predicted to reduce the number and percentage of male genital warts (607,659 [44.3%]) and anal cancer (822 [10.8%]) cases, as well as anal cancer deaths (247 [9.9%]) over 100 years.⁵⁸ In females, routine 9vHPV GNV was also predicted to reduce the number and percentage of genital warts (197,194 [29.2%]), CIN 1 (7,088 [6.4%]), CIN 2/3 (21,406 [6.1%]), cervical cancer (3,369 [3.6%]), vaginal cancer (71 [2.5%]), and anal cancer (151 [3.5%]) cases, as well as cervical cancer (720 [3.0%]), vaginal cancer (8 [2.3%]), and anal cancer deaths (45 [3.1%]) compared to FOV over 100 years. The benefits of routine 9vHPV GNV on male anal cancer cases and both male and female genital warts cases were also predicted to occur more rapidly than with FOV. Similar findings have been reported in studies conducted in the Italian and French settings at ~70% and 60% predicted coverage rates, respectively.^57,62 Taken together, a predicted benefit in both genders with routine GNV compared to FOV at high assumed female coverage rates is noteworthy because demonstrating added benefit through male vaccination becomes more difficult as female coverage rates increase.^60,62

To our knowledge, this is the first study in the Asia-Pacific region that evaluated the impact of 9vHPV routine GNV on additional HPV-related diseases including head and neck cancers and penile cancer caused by HPV. Although these diseases are not currently included in the 9vHPV indication,⁶³ there is evidence demonstrating the efficacy of HPV vaccines in reducing persistent HPV infections at specific anatomic sites, such as persistent external genital infection⁴⁸ and persistent oral infection.^64,65 Furthermore, reports from many countries in the international literature have indicated that HPV vaccines may protect against these diseases.^22,66–69 Indeed, 90% of head and neck cancers and 85% of penile cancers are attributable to the 9vHPV vaccine types,¹ and thus would appear to be preventable with widespread 9vHPV vaccination. Therefore, there is a rationale for including head and neck and penile cancers in modeling studies to assess the true health and economic impact of HPV vaccination programs.

Reductions in the incidence of HPV-related diseases for males were projected, implying that males were afforded direct benefits as a result of a GNV approach. Indeed, school-based GNV with the 4-valent HPV vaccine in Singapore was found to provide direct benefit to males with regard to the prevention of genital warts.⁷⁰ Previous modeling studies evaluating the potential impact of GNV in European countries have also demonstrated direct benefits to males with regard to reductions in genital warts and HPV-related cancers, including anal cancer, head and neck cancers, and penile cancer.^61,71 An HPV-type dynamic transmission model of HPV vaccination in Europe predicted that 4vHPV GNV would reduce the number of male HPV 16/18-related carcinoma cases (anal, head and neck, and penile) by 65%, compared with 4vHPV FOV (at 70% coverage).⁷¹ In another study, using a dynamic compartmental model of HPV vaccination in Sweden, implementation of a GNV or FOV program (at 80% coverage) was predicted to reduce the number of male HPV-related cancer cases (anal, oropharyngeal, and penile cancers) by 84% and 69%, respectively, compared with no vaccination.⁶¹

Similar to other modeling studies that assessed the impact of a GNV approach, the current model accounted for herd immunity with GNV, which leads to reduction in HPV-related diseases in both sexes.^56,59,61,71 However, the distribution of cancer cases prevented and life years gained among women, heterosexual men, and men who have sex with men depends on VCRs, with most gains attributed to women at low to medium coverage and increasing gains attributed to men who have sex with men at high coverage.⁵⁹

From a cost-effectiveness perspective, routine GNV with the 9vHPV vaccine was predicted to be cost-effective in the Hong Kong setting compared to routine FOV; greater cost-effectiveness was observed when catch-up was included. In the base case analysis that included head and neck and penile cancers, the ICER was $248,354 HKD/QALY for routine GNV with the 9vHPV vaccine compared with routine FOV – a value well within the cost-effectiveness threshold of $382,046 HKD/QALY (1.0 times the GDP). When penile cancer was excluded, the model still predicted that routine GNV would be cost-effective versus a routine FOV approach, with an ICER of $255,987 HKD/QALY. The current results predicting that routine GNV with the 9vHPV vaccine would be cost-effective compared to a FOV approach in Hong Kong are consistent with recent evaluations that assessed the cost-effectiveness of implementing 9vHPV GNV programs in Spain and France.^57,58 When interpreting the results presented here, it is important to note that while GDP-based cost-effectiveness ratios are informative, they should not be the only measure used for decision-making processes. In 2001, the World Health Organization’s Commission on Macroeconomics in Health suggested cost-effectiveness thresholds based on multiples of a country’s GDP. However, more recent publications have noted that GDP-based thresholds lack country specificity and that cost-effectiveness data should be used alongside other considerations, including budget impact and feasibility considerations, rather than in isolation based on a single threshold value.^72,73

Our analysis has some limitations. Our model did not include all of the potential health benefits (e.g., prevention of RRP, neonatal morbidity, and mortality related to cervical lesions) or economic benefits (e.g., indirect disease-related healthcare costs, such as work productivity) of HPV vaccination, thus, likely underestimating the value of a routine GNV approach. Input costs in the present manuscript were derived from prices charged to a private case in a public hospital in Hong Kong. Prices charged to public cases in Hong Kong are heavily subsidized by the government, therefore, using these prices in public hospitals does not accurately reflect the actual costs. The prices charged to private cases are closer to the actual costs, although these are likely to be slightly overestimated. Our model assumed a female vaccination coverage rate of 70%, which is higher than many countries with national HPV vaccination programs are able to achieve. Further, the value of adding males to routine HPV vaccination programs becomes harder to demonstrate; as female coverage increases, the high female coverage rate assumed in the current analysis may underestimate the true impact observed in a real-world setting with lower actual female coverage rates. In our sensitivity analyses, we assess the impact of vaccination rate of 30%, 60%, and 80%. Similarly, our model did not include disease transmission among males who have sex with males and did not assess possible changes to cervical cancer screening over the next 100 years. The calibrated Hong Kong model may overestimate the number of deaths due to undiagnosed cervical cancer, thus overestimating the QALYs gained due to deaths avoided. This may result in underestimating the ICER by no more than 15% which results in ICERS that are still well below the 1.0 times the GDP threshold utilized here.

Our model-based analysis suggests that routine GNV with the 9vHPV vaccine is projected to have a substantial public health impact for both men and women and is also likely to be cost-effective in reducing HPV-related diseases including head and neck cancer and penile cancer, compared with the current routine FOV approach with the 9vHPV vaccine in Hong Kong. The model results highlight the value of a routine GNV program among 12-year-olds to reduce the burden of HPV-related diseases in Hong Kong.

Funding Statement

This study was sponsored by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA.

Disclosure statement

Sally Shuk Yee Cheng, Danny Hsu, Queenie Wing-Lei Wong, Andrew Pavelyev, Isaya Sukarom, and Kunal Saxena are employees of Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA and shareholders in Merck & Co., Inc., Rahway, NJ, USA. Tak Hong Cheung has no conflict of interest to report.

Supplementary data

Supplemental data for this article can be accessed on the publisher’s website at https://doi.org/10.1080/21645515.2023.2184605.