Cost-effectiveness of quadrivalent human papillomavirus (HPV) vaccination in Mexico: A transmission dynamic model-based evaluation

Vaccine (2007) 26, 128—139 available at www.sciencedirect.com journal homepage: www.elsevier.com/locate/vaccine Cost-effectiveness of quadrivalent ...
2 downloads 2 Views 706KB Size
Vaccine (2007) 26, 128—139

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/vaccine

Cost-effectiveness of quadrivalent human papillomavirus (HPV) vaccination in Mexico: A transmission dynamic model-based evaluation夽 Ralph P. Insinga a,∗, Erik J. Dasbach a, Elamin H. Elbasha a, Andrea Puig b, Luz Myriam Reynales-Shigematsu c a

Department of Health Economic Statistics UG1C-60, Merck Research Laboratories, P.O. Box 1000, North Wales, PA 19454-1099, USA b Health Care Systems Department, The Wharton School, Colonial Penn Center, 3641 Locust Walk, University of Pennsylvania, Philadelphia, PA 19104, USA c Departamento de Investigaci´ on sobre Tabaco, Instituto Nacional de Salud P´ ublica, Av. Universidad 655, Col. Santa Mar´ıa Ahuacatitl´ an, Cuernavaca, Morelos, Mexico Received 20 February 2007; received in revised form 17 October 2007; accepted 18 October 2007 Available online 20 November 2007

KEYWORDS Human papillomavirus; Vaccine; Cost-effectiveness analysis

Summary We examined the potential health outcomes and cost-effectiveness of quadrivalent human papillomavirus (HPV) 6/11/16/18 vaccination strategies in the Mexican population using a multi-HPV type dynamic transmission model. Assuming similar cervical screening practices, with or without vaccination, we examined the incremental cost-effectiveness of vaccination strategies for 12 year-old females, with or without male vaccination, and temporary age 12—24 catch-up vaccination for females or both sexes. The most effective strategy therein was vaccination of 12-year-olds, plus a temporary 12—24-year-old catch-up program covering both sexes; whereby HPV 6/11/16/18-related cervical cancer, high-grade cervical precancer, and genital wart incidence was reduced by 84—98% during year 50 following vaccine introduction. Incremental cost-effectiveness ratios in the primary analyses ranged from ∼$3000 (U.S.) per quality-adjusted life year (QALY) gained for female vaccination strategies to ∼$16000/QALY for adding male vaccination with catch-up. © 2007 Elsevier Ltd. All rights reserved.

Background 夽 This study was supported by Merck & Co., Inc., where authors Insinga, Dasbach and Elbasha are currently employed. ∗ Corresponding author at: Merck & Co., Inc., UG1C-60, P.O. Box 1000, North Wales, PA 19454-1099, USA. Tel.: +1 267 305 7992; fax: +1 267 305 6455. E-mail address: ralph [email protected] (R.P. Insinga).

Cervical cancer is the second most common cancer among women worldwide [1]. In Mexico, a national cervical cancer screening program was initiated in 1974 [2]. However, the disease remains the country’s leading cause of female cancer-related mortality [3] and resulted in 4241 deaths in

0264-410X/$ — see front matter © 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.vaccine.2007.10.056

Cost-effectiveness of HPV vaccination in Mexico 2004 [4]. Human papillomavirus (HPV) infections are the primary cause of cervical, [5] anal, penile, vaginal and vulvar cancers, [6] anogenital warts [7] and recurrent respiratory papillomatoses [8,9] and also have been observed among a fraction of head and neck cancers [10]. HPV types 16 and 18 play a major role in the development of anogenital cancers, while types 6 and 11 are observed in the vast majority of anogenital warts and recurrent respiratory papillomatoses [11]. A quadrivalent vaccine (Gardasil® , Merck & Co., Inc.) targeting HPV types 6, 11, 16 and 18 was recently approved for use in Mexico. In implementing the vaccine, policymakers in Mexico will seek to understand the long-term health and economic impact of vaccination within the population and the cost-effectiveness of alternate vaccination strategies. We recently developed a multi-HPV type dynamic transmission model and demonstrated that a quadrivalent (HPV 6, 11, 16, 18) vaccine can be cost-effective when administered among U.S. females and males aged 12—24 [12]. The cost-effectiveness of HPV vaccination in Mexico has not been previously evaluated. In this paper, we have adapted our model to examine the health outcomes and cost-effectiveness of quadrivalent HPV vaccination in the Mexican population.

Methods A full description of the development, structure and selection of inputs for the dynamic transmission model is reported elsewhere [12,13]. Here, we briefly describe the model structure and selection of inputs relevant to the population in Mexico (Table 1).

Demographic and sexual mixing model structure and inputs The model simulates aging and all-cause mortality over time within the Mexican population ages 12 and over. We set the population size in the model to 100,000 individuals. The heterosexually mixing population is divided into 17 age groups (i.e., 12—14, 15—17, 18—19, 20—24, 25—29, 30—34, 35—39, 40—44, 45—49, 50—54, 55—59, 60—64, 65—69, 70—74, 75—79, 80—84, and 85+) and individuals within each age group die at a Mexican age and sex-specific all-cause mortality rate (net of cervical cancer mortality for females) [4,14]. The model simulates the transmission of HPV infection within the population as determined by the course of sexual mixing; a feature which allows for estimating both the direct and indirect (i.e., herd immunity) benefits of vaccination. Hence, we stratified each age group into three sexual activity groups, defined according to the rates of sexual partner change per unit time: low (0—1 per year), medium (2—4 per year), and high (5+ per year). In the absence of a representative Mexican data source, U.S. data were used to estimate the relative partner acquisition rate within each sexual activity group [15]. However, we utilized data from a national survey in Mexico to estimate the mean number of sexual partners per year by age group [16]. Assortativeness of sexual mixing between age and sexual activity groups was modeled with parameters of 0.6 (for age) and 0.7 (for sexual

129 activity group) [15]. In the general, the more heterogeneous the mixing (i.e., there is greater mixing between those of different ages and sexual activity classes) the greater the spread of infection and resultant disease in the population. The probability of HPV transmission given sexual contact with an infected partner was modeled as a rate per partnership (0.8 for male to female; 0.7 for female to male); [12] consistent with rates assumed in earlier HPV modeling work [17].

Epidemiologic and clinical model structure and inputs The natural course of HPV 6, 11, 16 and 18 infection among women in Mexico was modeled based on data derived from a comprehensive international literature review of the natural history of HPV infection and disease [18]. These international epidemiologic data were also used in the prior U.S. cost-effectiveness analysis [12], as there were found to be a small number of studies on type-specific HPV infection and disease natural history in general that were suited to our modeling purposes, much less for any particular country, and the natural course of incident HPV infections was hypothesized to be similar across populations. The pooling of data on the natural history of incident HPV infection and disease from multiple international settings is consistent with previous literature reviews and models of HPV natural history [19—22]. The clinical management of HPV disease is known to vary by country, however, and we used Mexican data sources to estimate the age-specific frequency of cervical cytologic screening [23] and hysterectomy [24], the proportion of the female population never receiving cervical cancer screening [25], and age and stage-specific cervical cancer mortality [26]. Consistent with data reported from a recent retrospective costing study conducted in Mexico, women diagnosed with cervical intraepithelial neoplasia (CIN) grade 3 were assumed to undergo treatment, while women diagnosed with CIN 1 and 2 were passively managed (Luz Myriam Reynales-Shigematsu, unpublished data). We derived data for other clinical parameters such as the sensitivity and specificity of cervical cytology [27—31] and colposcopy [32], treatment cure rates for CIN [33] and the persistence of HPV infection following treatment [34] from the international literature. These clinical parameters were generally anticipated to be less variable by country, and Mexico-specific studies were either absent from the literature or less methodologically suited to the data format required for model inputs [18] than sources obtained from the international literature. For instance, although we did not encounter studies describing population-based efficacy of treatment for cervical disease for Mexico, it is interesting to note that 5-year absolute survival rates among cervical cancer patients are quite similar between Mexico (66.6%), [26] and the U.S. (67.8%), as seen with the Surveillance Epidemiology and End Results (SEER) program from 1998—2004 [35]. Over the past several years, there has been a substantial increase in the rate of cervical cancer screening in Mexico [36]. For instance, an estimated 100,000 Pap smears were performed in Mexico in 1974, compared to 2.5 million in

130 Table 1

R.P. Insinga et al. Model parameters

Parameter

Estimate

Mean number of sexual partners per year, by age [16]a 12—14 years 15—17 years 18—29 years 30—49 years ≥50 years

0.1 0.3 1.9 1.1 1.0

Hysterectomy for non-HPV-related conditions by age, % per year [24] 15—24 years 25—29 years 30—34 years 35—39 years 40—44 years 45—54 years ≥55 years

0.01 0.03 0.12 0.36 0.67 0.58 0.20

Cervical cytology screening by age, % per year (excluding women with prior hysterectomy) [23,24] 18—24 years 11.9 (11.9) 25—34 years 22.6 (22.6) 35—44 years 25.2 (25.9) 45—64 years 21.4 (24.3) 65+ years 8.9 (10.4) Women never screened, % [25] 12.0 Conventional cytology sensitivity, % [28] For CIN 1 For ≥CIN 2/3 Conventional cytology specificity, % [27,29—31] % of women diagnosed with CIN undergoing treatmentb CIN 1/2 CIN 3

28 59 97 0% 100%

Cervical cancer mortality by age and stage, % by year [26]c for localized/regional/distant cervical cancers 15—39 years 13.5/20.2/80.8 40—49 years 6.6/10.0/39.9 50—59 years 9.0/13.6/54.3 60—69 years 10.0/15.0/60.2 ≥70 years 13.0/19.5/78.5 Costs of diagnosing and treating HPV disease (2005 Mexican Pesos)b Genital wart treatmentd Conventional cytology screening exam Colposcopy Cervical biopsy CIN treatment Localized cervical cancer treatment Regional cervical cancer treatment Distant cervical cancer treatment

2,000 183 490 204 15,849 76,842 86,846 85,022

Other model parameters [12]e



HPV: human papillomavirus; CIN: cervical intraepithelial neoplasia. a The mean number of sexual partners per year for ages 12—17 was based on assumptions made for the U.S. population [12] as Mexican data were unavailable. b Luz Myriam Reynales-Shigematsu,unpublished data. c Cervical cancer mortality rates from Flores-Luna et al. [26] were reported by age group and by stage individually, but not by both age and stage together. Based on the reported data, mean 12-month stage-specific mortality rates were estimated for each age group by multiplying each 12-month stage-specific rate by the ratio of the age-specific to the overall (across all ages) 5-year mortality rate reported in the paper. d Investigator estimate. e Model parameters specific to Mexico are summarized in the table. Other model parameters for the natural history of HPV infection, CIN and cervical cancer, cervical cancer symptom development, test characteristics of colposcopy, care seeking behavior for genital warts, efficacy of disease treatment, persistence of HPV following treatment and quality of life weights are identical to those reported for the U.S. model as described elsewhere [12].

Cost-effectiveness of HPV vaccination in Mexico 1994 and 6.5 million in 2004. As a result of this increase in screening, particularly among younger and middle-aged women, [25] it is expected that cervical cancer incidence rates in the future will be falling, regardless of the introduction of HPV vaccination. To account for changing disease incidence resulting from this rise in screening coverage, we ran the model simulating a large increase in cervical cancer screening, to current levels (using data for 2004), in the decade prior to the introduction of HPV vaccination, based on data reported by Mexico’s Instituto Nacional de Salud P´ ublica [23]. We also assumed that 12.0% of Mexican women will remain unscreened throughout their lifetimes, based on data from the 2003 Encuesta Nacional de Salud Reproductiva [25]. Thus, cervical screening was not modeled for 12.0% of the female population in our model (with cervical disease detected only based on symptoms at the cancerous stage), with Mexican cervical cancer screening rates applied over the remaining denominator of women at each age with an intact cervix. The resultant output was then compared to current cervical cancer data for Mexico during model calibration [4].

HPV vaccination strategies and characteristics Assuming similar cervical cancer screening utilization, with or without HPV vaccination, we examined model epidemiologic and economic output for six strategies in the reference case analysis: (1) no vaccination; (2) vaccination of females at age 12; (3) vaccination of males and females at age 12; (4) vaccination of females at age 12, with a temporary 5-year catch-up program for females ages 12—24; (5) vaccination of males and females at age 12, with a temporary 5-year catch-up program for females ages 12—24; (6) vaccination of males and females at age 12, with a temporary 5-year catch-up program for males and females ages 12—24. In the reference case, we assumed that three-dose vaccine coverage for females and males at age 12 would linearly increase from 0 to 70% during the first 5 years following vaccine introduction and remain constant thereafter. It was assumed that only previously unvaccinated individuals would participate in the temporary catch-up program, and that three-dose coverage among this group would linearly increase from 0 to 50% during the first 5 years following vaccine introduction and then fall to 0%. In practice this meant that the model predicted that approximately 70% of all individuals aged 12—24-years during this 5-year window would be vaccinated during the first 5 years following vaccine introduction with both 12-year-old and catch-up vaccination. As actual vaccine coverage rates will not be known until implementation, other coverage levels were examined in sensitivity analyses. Based on clinical trial data we assumed a vaccine efficacy against incident HPV 6, 11, 16 and 18 infections of 90%, [37] with an efficacy of 95.2% against incident cervical disease and 98.9% against genital warts due to these HPV types [38]. We did not model other conditions (e.g., vaginal or vulvar cancers) caused by these four HPV types, or infection and disease caused by HPV types not targeted by the vaccine. Because the vaccine is prophylactic, we assumed no therapeutic benefits against HPV infections prevalent at the time of vaccination or for women with prior infection

131 and clearance of a given HPV type. The duration of vaccinederived protection is currently unknown, however, to date, no waning in efficacy has been observed through 5 years of follow-up [39]. Consistent with the range of values examined in prior cost-effectiveness studies of HPV vaccines, [21,40—42] in our reference case analysis, we assume a lifetime duration of vaccine protection against HPV 6/11/16/18 and examine results assuming a 10-year duration of protection in sensitivity analyses.

Economic data Costs were estimated from the perspective of the Mexican healthcare system. The direct medical costs associated with the diagnosis and treatment of cervical disease were based on a recent micro-costing study conducted within Mexico’s National Public Health Institute (Luz Myriam ReynalesShigematsu, unpublished data). A Mexican population-based study for the costs associated with the management of genital warts was not found. As an approximation, we assumed that the cost per case of genital warts in Mexico, relative to that for CIN 3 (15,849 pesos), was proportionately similar to what has been observed based on costing studies in the U.S. population data (about 1:8) [43,44] and assigned a cost of 2000 pesos per case of genital warts. We assumed a cost for the three-dose vaccine series of 2640 pesos, which is about $240 U.S., based on an exchange rate of 11 Mexican pesos per $1 U.S. [45]. Non-medical costs such as a patient’s lost wages and transportation costs were not included in the analysis. Quality-adjusted life years (QALYs) were estimated based on health utilities as reported for the U.S. model as Mexico-specific data were not available [12]. In addition to utilities for HPV disease states, age and sex-specific utility weights were also incorporated for individuals without HPV disease to account for the impact of co-morbid conditions, as described in our previous work [12]. All costs were adjusted to 2005 Mexican pesos using inflation rates reported by the Mexican Central Bank [46] and both costs and effects were discounted at a 3% annual rate.

Model simulations and validation All model equations and inputs were programmed in Mathematica (Wolfram Research, Champaign, IL, v5.2.0.0) [47]. The predictive validity of the model was assessed by comparing model output to available data on the health burden of HPV disease in Mexico. For instance, assuming that approximately 70% of cervical cancer deaths in Mexico are caused by HPV 16 and 18-related disease, [48] the estimated female HPV 16/18-related cervical cancer mortality rate for Mexico across females of all ages in 2004 was 5.6 per 100,000; [4,14] with a projected rate of 5.5 per 100,000 in year 1 of the model in the absence of vaccination. Population-based data on the incidence of CIN and genital warts in Mexico have not been reported and constitute an area for future research. Thus, we were unable to calibrate the model to population data for these events. The model examined the impact of introducing HPV vaccination over a time horizon of 100 years. Epidemiologic outputs included HPV prevalence, the incidence of CIN, gen-

132

R.P. Insinga et al.

Figure 1 Incidence of clinically diagnosed CIN 2/3 due to HPV 6/11/16/18 infection among females ages 12 and over by vaccination strategy. Incidence rates for HPV 6/11/16/18-related CIN 2/3 per 100,000 female population are depicted for each vaccination strategy during the 100 years following the introduction of vaccination.

ital warts and cervical cancer, and cervical cancer mortality. Economic outputs included costs, QALYs and incremental costs/QALY for each strategy. Model inputs were varied in sensitivity analyses. Inputs varied in one-way sensitivity analyses included the duration of vaccine protection and level of vaccine coverage; model parameters which were found to be influential in our prior cost-effectiveness analysis work [12]. We also varied the proportion of clinically diagnosed genital warts and their costs, given the lack of population-based data for this disease endpoint for Mexico. Finally, we examined a ‘‘pessimistic scenario’’ in a multi-way sensitivity analysis in which the duration of vaccine protection, incidence and costs of genital warts, and health utility decrements for cervical disease and genital warts were simultaneously set to values less favorable to the cost-effectiveness of quadrivalent HPV vaccination than those assumed in the reference case.

Results Health outcomes For each of the six vaccination strategies, Figs. 1—4 depict the model projected HPV 6, 11, 16 and 18-related incidence of clinically diagnosed CIN 2/3 and cervical cancer for females, and genital warts for males and females (among individuals ages 12 and over), during the 100 years following the introduction of HPV vaccination. For all disease endpoints examined, we found the most clinically effective strategy to be vaccination of 12-yearold females and males combined with a temporary female and male 12—24-year-old catch-up program. The model projected long-term reductions in HPV 6/11/16/18-related disease incidence during year 50 following the introduction of HPV vaccination were 90, 86, 98 and 98% for CIN 2/3, cervical cancer and male and female genital warts,

Figure 2 Incidence of clinically diagnosed cervical cancer due to HPV 16/18 infection among females ages 12 and over by screening and vaccination strategy. Incidence rates for HPV 16/18-related cervical cancers per 100,000 female population are depicted for each vaccination strategy during the 100 years following the introduction of vaccination.

Cost-effectiveness of HPV vaccination in Mexico

133

Figure 3 Incidence of clinically diagnosed genital warts due to HPV 6/11 infection among males ages 12 and over by vaccination strategy. Incidence rates for HPV 6/11-related genital warts per 100,000 male population are depicted for each vaccination strategy during the 100 years following the introduction of vaccination.

respectively. We found that reductions in genital wart incidence following vaccine introduction occurred sooner across all vaccine strategies than reductions in CIN 2/3 and cervical cancer. For instance, among each sex-specific population ages 12 and over, under the most comprehensive vaccination strategy examined, the incidence of HPV 6/11/16/18-related CIN 2/3 was estimated to be reduced by 28 per 100,000 (39%), cervical cancer by 0.4 per 100,000 (4%), male genital warts by 152 per 100,000 (93%) and female genital warts by 150 per 100,000 (93%), during year 10 following vaccine introduction. The addition of a temporary 5-year catch-up program was found to yield substantially greater health benefits over the short and medium terms than strategies involving vaccination of 12-year olds alone. For example, over a 25 year time frame, estimated reductions in the cumulative incidence of HPV 6/11/16/18-related CIN 2/3, cervical cancer and male and female genital warts were 134, 254, 67 and

70% larger respectively, when administering both 12-yearold female vaccination and a temporary age 12—24 female catch-up vaccination program, than when vaccinating 12year-old females alone. Over the longer term, the annual incidence of disease was comparable between otherwise similar programs with and without temporary catch-up vaccination, as members of the initial catch-up cohort exited the model. However, relatively lower disease incidence with catch-up was observed to persist over the longer term when catch-up vaccination was modeled as a permanent rather than a temporary strategy (results not shown).

Cost-effectiveness analysis Incremental cost-effectiveness ratios for the reference case analysis are presented in Table 2, where strategies are listed in order of increasing effectiveness (total discounted

Figure 4 Incidence of clinically diagnosed genital warts due to HPV 6/11 infection among females ages 12 and over by vaccination strategy. Incidence rates for HPV 6/11-related genital warts per 100,000 female population are depicted for each vaccination strategy during the 100 years following the introduction of vaccination.

134 Table 2

R.P. Insinga et al. Cost-effectiveness analysis of HPV 6/11/16/18 vaccination

Strategiesb

No vaccination 12-year-old female vaccination 12-year-old female & male vaccination 12-year-old female + 12—24-year-old temporary female catch-up vaccination 12-year-old female & male + 12—24-year-old temporary female catch-up vaccination 12-year-old female & male + 12—24-year-old temporary female & male catch-up vaccination

Incrementala

Discounted total Costs (pesos)

QALYs

Costs (pesos)

QALYs

Pesos/QALY

$U.S./QALY

81,927,144 110,704,625 150,342,112

2,694,316 2,695,278 2,695,560

— 28,777,482 39,637,487

— 962 282

— 29,905 Dominated

— 2,719 Dominated

123,143,801

2,695,649

12,439,175

371

33,530

3,048

163,562,812

2,695,870

40,419,011

221

183,297

16,663

180,016,177

2,695,959

16,453,365

90

183,717

16,702

HPV: human papillomavirus; QALY: quality-adjusted life year. a Incremental costs, QALYs and cost/QALY ratios are compared to the preceding non-dominated strategy. Strategies are listed in order of increasing effectiveness. A strategy will be dominated if a more effective strategy is relatively less costly or has a lower cost-effectiveness ratio. b All strategies are in addition to cervical cancer screening according to current rates observed within Mexico.

QALYs). Strategies are dominated if there is a more effective strategy that is relatively less costly or has a lower incremental cost-effectiveness ratio. Incremental costs, QALYs and cost/QALY ratios for a given strategy are compared to the preceding non-dominated strategy. Compared to no vaccination, a 12-year-old female vaccination program was found to have an incremental cost-effectiveness ratio of 29,905 pesos/QALY (U.S. $2719/QALY). The addition of a temporary female 12—24year-old catch-up program to this strategy was of similar cost-effectiveness. The most effective strategy, involving vaccination of both males and females at age 12 and a temporary 12—24-year-old catch-up program for both sexes yielded an incremental cost-effectiveness ratio of 183,717 pesos/QALY (U.S. $16,702/QALY) when compared to 12-yearold vaccination of both sexes with a 12—24-year-old female catch-up program.

(∼75% total reduction in the economic burden of genital warts) resulted in cost-effectiveness ratios that were 20—30% higher than those observed for the reference case analysis. In the ‘‘pessimistic scenario’’ parameter values for the duration of vaccine protection (10 years), genital wart burden (50% reduction in genital wart diagnoses and costs per case) and health utility values for cervical disease and genital warts (0.97 for all states) were simultaneously set to values less favorable to the cost-effectiveness of HPV vaccination than those assumed in the reference case. This resulted in incremental cost-effectiveness ratios for each vaccination strategy that were approximately 2.5—5.5 times higher than those observed in the reference case analysis (range: 139,158—456,350 pesos/QALY; U.S. $12,651—41,486/QALY).

Cumulative population benefits Sensitivity analyses Cost-effectiveness ratios from one-way sensitivity analyses are presented in Table 3. Parameters varied reflected those of particular policy relevance and/or for which considerable uncertainty exists. Reducing the assumed duration of vaccine protection from lifelong to 10 years resulted in the vaccination of 12-year-old females alone becoming a dominated strategy, and higher cost-effectiveness ratios for the remaining vaccination strategies (range: 89,000—245,000 pesos/QALY; U.S. $8000—22,000/QALY). Assuming a lower vaccination coverage rate of 20% for all individuals included within each strategy resulted in lower cost-effectiveness ratios, particularly for strategies involving male vaccination. Conversely, the reverse was observed with higher vaccine coverage rates of 85%. Reducing both the proportion of clinically diagnosed genital warts and their costs by 50%

Table 4 provides an estimation of the number of cases of HPV disease that could be prevented in Mexico over the next 25 and 100 years with quadrivalent HPV vaccination according to the four non-dominated vaccination strategies in the reference case analysis (12-year-old females only, 12-year-old females + 12—24-year-old female temporary catch-up, 12-year-old females and males + 12—24-yearold female temporary catch-up, 12-year-old females and males + 12—24-year-old female and male temporary catchup). We assumed a current population size for Mexico among individuals ages 12+ of 40,067,453 for males and 41,193,835 for females, and an annual population growth rate of 1% for this projection [14]. Over a 100 year time horizon following the introduction of quadrivalent HPV vaccination, for the most effective strategy of 12-year-old female and male vaccination with a temporary 12—24-year-old catch-

Strategies

Sensitivity analyses for cost-effectiveness of HPV 6/11/16/18 vaccination b

No vaccination 12-year-old female vaccination 12-year-old female & male vaccination 12-year-old female + 12—24-year-old temporary female catch-up vaccination 12-year-old female & male + 12—24-year-old temporary female catch-up vaccination 12-year-old female & male + 12—24-year-old temporary female & male catch-up vaccination

Mexican peso cost/QALY ratio ($U.S.)a Reference-case analysis

10-year duration of vaccine protection

20% vaccine coveragec

85% vaccine coveragec

50% reduction in genital wart diagnoses and costs per case

Pessimistic scenariod

— 29,905 (2,719) Dominated

— Dominated Dominated

— 27,478 (2,498) Dominated

— 31,153 (2,832) Dominated

— 37,991 (3,454) Dominated

— Dominated Dominated

33,530 (3,048)

89,682 (8,153)

29,156 (2,651)

36,257 (3,296)

40,589 (3,690)

139,158 (12,651)

183,297 (16,663)

Dominated

69,506 (6,319)

Dominated

Dominated

449,182 (40,835)

183,717 (16,702)

245,745 (22,340)

89,541 (8,140)

365,515 (33,229)

238,622 (21,693)

456,350 (41,486)

Cost-effectiveness of HPV vaccination in Mexico

Table 3

HPV: human papillomavirus; QALY: quality-adjusted life year. a Incremental cost/QALY ratios are compared to the next most effective non-dominated strategy. A strategy will be dominated if a more effective strategy is relatively less costly or has a lower cost-effectiveness ratio. b All strategies are in addition to cervical cancer screening according to current rates observed within Mexico. c Coverage rate applies to all ages, with a linear ramp-up to coverage level over 5 years as described for reference case analyses. d The pessimistic scenario simultaneously assumes values for duration of vaccine protection (10 years), genital wart burden (50% reduction in both the proportion of clinically diagnosed genital warts and the costs per case) and health utilities (0.97 for all genital wart and cervical disease health states) less favorable to the cost-effectiveness of quadrivalent HPV vaccination than those assumed in the reference case.

135

21,760,238 (100%)

up program, it was estimated that more than 17 million cases of genital warts (among both males and females), 3 million cases of CIN 2/3, 400,000 cases of cervical cancer and 150,000 cervical cancer deaths could be prevented in Mexico. Across the strategies examined, genital warts accounted for 80—82% of total HPV disease cases prevented over 100 years; a figure which rose to 87—91% over a shorter 25 year time frame. Within 25 years following the vaccine’s introduction, it was estimated that the addition of a temporary 5-year catch-up program for females ages 12—24 to a 12-year old female vaccination program could increase the number of HPV disease cases prevented by 73% (1.1 million additional cases) and cervical cancer deaths avoided by 292% (3788 additional deaths prevented). Adding vaccination of males at age 12, along with age 12—24 temporary male catch-up vaccination, would further expand the number of HPV disease cases prevented by over 30% (800,000 additional cases) and cervical cancer deaths avoided by 23% (1165 additional deaths prevented). The disproportionately larger reductions in HPV disease cases over 100 years as compared to 25 years were in part related to the increasing population impact of the vaccine in reducing HPV disease over time, as well as the progressively larger size of the population over time as the result of projected growth.

Discussion HPV: human papillomavirus; CIN: cervical intra-epithelial neoplasia. a Percentages refer to the proportion of all cases prevented for a particular strategy and time horizon.

3,463,056 (15.9%) 17,895,606 (82.2%) 399,012 (11.5%) 3,037,691 (87.9%)

18,008 (0.5%)

3,454,711 (100%)

401,576 (1.8%)

21,280,665 (100%) 3,395,984 (16.0%) 17,530,083 (82.4%) 350,623 (11.3%) 2,723,682 (88.2%)

15,423 (0.5%)

3,089,729 (100%)

354,598 (1.7%)

15,261,474 (100%) 287,740 (1.9%) 2,738,618 (17.9%) 12,235,117 (80.2%) 317,347 (12.0%) 2,313,491 (87.5%)

14,570 (0.6%)

2,645,409 (100%)

248,806 (1.8%) 2,439,820 (17.5%) 11,234,402 (80.7%) 135,678 (8.9%)

12-year-old female vaccination 12-year-old female + 12—24-year-old temporary female catch-up vaccination 12-year-old female & male + 12—24-year-old temporary female catch-up vaccination 12-year-old female & male + 12−24-year-old temporary female & male catch-up vaccination

1,391,155 (90.9%)

4,036 (0.3%)

1,530,869 (100%)

Cervical cancer n (%) CIN 2/3 n (%) Genital warts n (%) Total n (%) Genital warts n (%)a

CIN 2/3 n (%)

Cervical cancer n (%)

100 year time horizon

Reference case projection of cumulative HPV 6/11/16/18-related disease cases potentially prevented in Mexico over the next 25 and 100 years

25 year time horizon Strategies

Table 4

13,923,028 (100%)

R.P. Insinga et al.

Total n (%)

136

Consistent with our prior results reported for the U.S. population, [12] data from this analysis suggest that a quadrivalent HPV 6/11/16/18 vaccine can yield substantial benefits in reducing the incidence of cervical cancer, CIN and genital warts in Mexico. In fact, the per capita incremental projected health benefits associated with each vaccination strategy, in terms of quality-adjusted life years gained, were estimated to be approximately twice as large for Mexico as for the United States [12]. This differential is in large part driven by the greater opportunity for the prevention of cervical cancer in Mexico, due to lower cervical cancer screening rates, with resultant rates of cervical cancer estimated to be approximately 3 times higher in Mexico than the U.S. Because cervical cancers can take many years to develop, in the first 25 years following the introduction of vaccination, the largest expected reductions in disease incidence were estimated to derive from the prevention of HPV 6/11related genital warts. Although non-malignant, genital warts can have a negative psychosocial impact [49] and, in some instances, may be characterized by itching, burning and tenderness at the wart site and anal, urethral or vaginal bleeding or discharge [50]. Their treatment often requires multiple patient visits [43] with therapies involving genital burning, freezing or surgical excision, which can be painful and are frequently unsuccessful in preventing wart recurrence [51]. The addition of a temporary 5-year catch-up vaccination program for 12—24-year-old males and females was found to provide substantially greater reductions in HPV disease over the short and medium-term than the vaccination of 12-year-olds alone. This was especially evident for the prevention of cervical cancer where, for instance, adding a temporary female catch-up program more than tripled the

Cost-effectiveness of HPV vaccination in Mexico number of cervical cancer cases prevented over 25 years, compared to solely vaccinating 12-year-old girls. This was seen in part because it can take many years for cervical cancers to develop and vaccination of the age 12—24 catch-up cohort occurred closer in time to preventing cancer development for those females relative to for routinely vaccinated 12-year-olds. In the reference case cost-effectiveness analysis, U.S. dollar equivalent cost per QALY ratios for female vaccination strategies were relatively similar to our previous estimates for quadrivalent HPV vaccination in the U.S. ($2500—$4500/QALY) [12]. However, the incremental costeffectiveness of adding vaccination for males to a female vaccination program had a lower ratio for Mexico (U.S. ∼$16,000/QALY versus $40,000—45,000/QALY). To our knowledge, there have not been prior cost/QALY studies published for Mexican health interventions in the area of HPV disease and only a small number of evaluations in other clinical areas [52,53]. Some analysts have considered a cost-effectiveness threshold of less than three times the gross domestic product (GDP) per capita per disabilityor quality-adjusted life year gained for developing world countries [54—56]. The estimated GDP per capita for Mexico in 2005 was U.S. $10,000 [57]. The reference case and sensitivity analysis cost-effectiveness ratios for female only vaccination strategies fall well below three times this value ($30,000/QALY). The reference case incremental cost-effectiveness ratios for adding male vaccination also fall below this threshold, while sensitivity analysis results assuming an 85% coverage rate for 12-year-old males are just above ($33,229/QALY). However, there are reasons to regard all of the cost-effectiveness results presented in this analysis as conservative, as will be subsequently discussed. Our analysis has several limitations. First, we have not modeled the health benefits and cost offsets associated with other HPV diseases which may result from HPV 6, 11, 16 and 18 infections such as anal, penile, vaginal, vulvar and head and neck cancers and recurrent respiratory papillomatoses. Inclusion of these diseases would further improve the costeffectiveness profile for quadrivalent HPV vaccination and represents an area for future work. Second, we have not characterized the productivity losses associated with HPV disease resulting from lost labor earnings due to morbidity or premature mortality. A recent study reported annual productivity losses associated with cervical cancer mortality in the U.S. in excess of $1 billion [58]. Although the U.S. Panel on cost-effectiveness in Health and Medicine has recommended against the inclusion of productivity losses directly within cost/QALY estimations, [59] their characterization can assist in more completely describing the economic impact of health interventions. For instance, the Canadian Agency for Drugs and Technology in Health accepts the estimation of productivity losses as a separate analysis from cost/QALY data in reimbursement submissions [60]. Third, population data are not currently available on the annual age-specific incidence of cervical cancer, CIN or genital warts in Mexico. This limited our ability to calibrate the model for these endpoints. As noted previously, we were able to reasonably calibrate to current cervical cancer mortality rates observed in Mexico [4,14]. The natural history parameters for HPV infection and disease utilized in the

137 model were based on international literature and identical to those we previously used in our U.S. cost-effectiveness model, which was calibrated to population data for genital warts, CIN and cervical cancer [12]. Fourth, population data on the costs associated with all follow-up care for an incident episode of genital warts were not available. The cost figure (2000 pesos) assumed in the analysis is consistent with the relative magnitude of genital wart costs observed elsewhere, [43,44] however, more accurate estimation of this cost represents an area for future research. Sensitivity analyses examining the impact of reducing both the proportion of clinically diagnosed genital warts and the costs per case found that model assumptions concerning these parameters exerted a moderate influence on overall cost-effectiveness results. Fifth, we have not considered potential HPV type crossprotection or competitive release, however these issues can be addressed in the future as empirical data warrant. Preliminary data have suggested that HPV vaccines might confer some degree of cross-type protection against certain HPV types not specifically targeted by vaccines [61]. Finally, both the incidence of HPV disease and costs of care may vary by healthcare provider and geographic area within Mexico [62,63]. In conclusion, we have demonstrated that quadrivalent HPV 6/11/16/18 vaccination of 12-year-old females and males with a temporary age 12—24 catch-up vaccination program can be a highly effective strategy for reducing the incidence of cervical cancer, CIN and genital warts in Mexico, while retaining current cervical cancer screening practices. Vaccination strategies for both females and males appeared cost-effective based on cited per capita GDPbased thresholds in the reference case analysis [54—56]. In sensitivity analyses, the cost-effectiveness of femaleonly vaccination strategies was relatively robust, however, the incremental cost-effectiveness of male vaccination appeared less cost-effective at very high coverage levels for female vaccination. These data support the implementation of quadrivalent HPV vaccination as a major advance in the prevention of cervical disease and genital warts in Mexico.

Acknowledgments The authors thank Dr. Eduardo Lazcano-Ponce of the Instituto Nacional de Salud P´ ublica in Mexico for helpful comments on a draft version of this manuscript.

References [1] Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005;55(2):74—108. [2] Lazcano-Ponce EC, Moss S, Alonso dR, Salmeron CJ, Hernandez AM. Cervical cancer screening in developing countries: why is it ineffective? The case of Mexico. Arch Med Res 1999 May;30(3):240—50. [3] Ferlay J, Bray F, Pisani P, Parkin DM. GLOBOCAN 2002: Cancer incidence, mortality and prevalence worldwide. IARC Cancerbase No. 5 version 2.0, IARC Press, Lyon, 2004. [4] Distribuci´ on de defunciones por c´ ancer c´ ervico uterino seg´ un entidad federativa de residencia habitual de la mujer y grupos de edad, 2004. Direcci´ on General Adjunta de Salud Reproductiva, INEGI/Ssa, Aguascalientes, Mexico, 2004.

138 [5] Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, et al. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999;189(1):12—9. [6] Carter JJ, Madeleine MM, Shera K, Schwartz SM, CushingHaugen KL, Wipf GC, et al. Human papillomavirus 16 and 18 L1 serology compared across anogenital cancer sites. Cancer Res 2001;61(5):1934—40. [7] Brown DR, Schroeder JM, Bryan JT, Stoler MH, Fife KH. Detection of multiple human papillomavirus types in Condylomata acuminata lesions from otherwise healthy and immunosuppressed patients. J Clin Microbiol 1999;37(10):3316—22. [8] Pou AM, Rimell FL, Jordan JA, Shoemaker DL, Johnson JT, Barua P, et al. Adult respiratory papillomatosis: human papillomavirus type and viral coinfections as predictors of prognosis. Ann Otol Rhinol Laryngol 1995;104(10 Pt 1):758—62. [9] Rimell FL, Shoemaker DL, Pou AM, Jordan JA, Post JC, Ehrlich GD. Pediatric respiratory papillomatosis: prognostic role of viral typing and cofactors. Laryngoscope 1997;107(7):915—8. [10] Kreimer AR, Clifford GM, Boyle P, Franceschi S. Human papillomavirus types in head and neck squamous cell carcinomas worldwide: a systematic review. Cancer Epidemiol Biomarkers Prev 2005;14(2):467—75. [11] Dunne EF, Markowitz LE. Genital human papillomavirus infection. Clin Infect Dis 2006;43(5):624—9. [12] Elbasha EH, Dasbach EJ, Insinga RP. Model for assessing human papillomavirus vaccination strategies. Emerg Infect Dis 2007;13(1):28—41. [13] Dasbach EJ, Elbasha EH, Insinga RP. Mathematical models for predicting the epidemiologic and economic impact of vaccination against human papillomavirus infection and disease. Epidemiol Rev 2006;28:88—100. [14] Proyecciones de la poblaci´ on de M´ exico, 2000—2050. Consejo Nacional de Poblaci´ on. M´ exico, DF: CONAPO, 2002. [15] Laumann EO, Gagnon JH, Michael RT, Michaels S. The social organization of sexuality. Chicago, IL: University of Chicago Press; 1994. [16] Primera encuesta nacional sobre sexo: estudio de opinion en viviendas. 2004. Consulta Mitofsky, Aguascalientes, Mexico. http://www.consulta.com.mx; Accessed: August 2006. [17] Hughes JP, Garnett GP, Koutsky L. The theoretical populationlevel impact of a prophylactic human papilloma virus vaccine. Epidemiology 2002;13(6):631—9. [18] Insinga RP, Dasbach EJ, Elbasha EH. Epidemiologic natural history and clinical management of human papillomavirus (HPV) disease: A critical review of the literature in the development of an HPV dynamic transmission model; 2006. Under review. [19] Cantor SB, Atkinson EN, Cardenas-Turanzas M, Benedet JL, Follen M, MacAulay C. Natural history of cervical intraepithelial neoplasia: a meta-analysis. Acta Cytol 2005;49(4): 405—15. [20] Goldie SJ, Grima D, Kohli M, Wright TC, Weinstein M, Franco E. A comprehensive natural history model of HPV infection and cervical cancer to estimate the clinical impact of a prophylactic HPV-16/18 vaccine. Int J Cancer 2003;106(6):896—904. [21] Myers ER, McCrory DC, Nanda K, Bastian L, Matchar DB. Mathematical model for the natural history of human papillomavirus infection and cervical carcinogenesis. Am J Epidemiol 2000;151(12):1158—71. [22] Ostor AG. Natural history of cervical intraepithelial neoplasia: a critical review. Int J Gynecol Pathol 1993;12(2):186—92. [23] Louftalot M. Avances del programa de detecci´ on oportuna del c´ ancer cervical en M´ exico. 3-3-2005. Presented at the XI Congreso de Investigaci´ on en Salud P´ ublica, Cuernavaca, Mexico: also available at: http://www.insp.mx/2005/xicongreso/ ponencias/as/j/1.ppt. [24] Unpublished data. Instituto Mexicano de Seguridad Social, D.F., Mexico, 2004.

R.P. Insinga et al. [25] Encuesta Nacional de Salud Reproductiva 2003: Principales Resultados. Secretaria de Salud, D.F., Mexico, 2003, www.letraese.org.mx/sxencuestasaludreproductiva.ppt. [26] Flores-Luna L, Salazar-Martinez E, Escudero-De los RP, Gonzalez-Lira G, Zamora-Munoz S, Lazcano-Ponce E. Prognostic factors related to cervical cancer survival in Mexican women. Int J Gynaecol Obstet 2001;75(1):33—42. [27] Baldauf JJ, Dreyfus M, Lehmann M, Ritter J, Philippe E. Cervical cancer screening with cervicography and cytology. Eur J Obstet Gynecol Reprod Biol 1995;58(1):33—9. [28] Bigras G, de Marval F. The probability for a Pap test to be abnormal is directly proportional to HPV viral load: results from a Swiss study comparing HPV testing and liquid-based cytology to detect cervical cancer precursors in 13,842 women. Br J Cancer 2005;93(5):575—81. [29] Coste J, Cochand-Priollet B, De Cremoux P, Le Gales C, Cartier I, Molinie V, et al. Cross sectional study of conventional cervical smear, monolayer cytology, and human papillomavirus DNA testing for cervical cancer screening. BMJ 2003;326(7392):733. [30] Hockstad RL. A comparison of simultaneous cervical cytology, HPV testing, and colposcopy. Fam Pract Res J 1992;12(1): 53—60. [31] Mann W, Lonky N, Massad S, Scotti R, Blanco J, Vasilev S. Papanicolaou smear screening augmented by a magnified chemiluminescent exam. Int J Gynaecol Obstet 1993;43(3): 289—96. [32] Mitchell MF, Schottenfeld D, Tortolero-Luna G, Cantor SB, Richards-Kortum R. Colposcopy for the diagnosis of squamous intraepithelial lesions: a meta-analysis. Obstet Gynecol 1998;91(4):626—31. [33] Flannelly G, Langhan H, Jandial L, Mana E, Campbell M, Kitchener H. A study of treatment failures following large loop excision of the transformation zone for the treatment of cervical intraepithelial neoplasia. Br J Obstet Gynaecol 1997;104(6):718—22. [34] Cruickshank ME, Sharp L, Chambers G, Smart L, Murray G. Persistent infection with human papillomavirus following the successful treatment of high grade cervical intraepithelial neoplasia. BJOG 2002;109(5):579—81. [35] SEER*Stat 6.3.5 — SEER Cancer Incidence Limited-Use Database, 1973—2004 and Associated Populations in Binary Format for Use by SEER*Stat. 4-16-2007. Bethesda, MD, U.S.A., Cancer Statistics Branch, National Cancer Institute, National Insititutes of Health. [36] Avances del programa de deteccion oportuna del cancer cervical en Mexico. Direccion de Prevencion y Control del Cancer en la Mujer, Centro de Equidad de Genero y Salud Reproductiva, Secretaria de Salud, D.F., Mexico, March 2005. [37] Koutsky LA, Ault KA, Wheeler CM, Brown DR, Barr E, Alvarez FB, et al. A controlled trial of a human papillomavirus type 16 vaccine. N Engl J Med 2002;347(21):1645—51. [38] Gardasil Quadrivalent Human Papillomavirus (Types 6, 11, 16, 18) Recombinant Vaccine. 4-4-2007. Product Approval Information, U.S. Food and Drug Administration, Center for Biologics Evaluation and Research, Rockville, MD. [39] Villa LL, Costa RL, Petta CA, Andrade RP, Paavonen J, Iversen OE, et al. High sustained efficacy of a prophylactic quadrivalent human papillomavirus types 6/11/16/18 L1 virus-like particle vaccine through 5 years of follow-up. Br J Cancer 2006;95(11):1459—66. [40] Goldie SJ, Kohli M, Grima D, Weinstein MC, Wright TC, Bosch FX, et al. Projected clinical benefits and cost-effectiveness of a human papillomavirus 16/18 vaccine. J Natl Cancer Inst 2004;96(8):604—15. [41] Sanders GD, Taira AV. Cost-effectiveness of a potential vaccine for human papillomavirus. Emerg Infect Dis 2003;9(1): 37—48.

Cost-effectiveness of HPV vaccination in Mexico [42] Stratton KR, Durch JS, Lawrence RS, editors. Vaccines for the 21st century: a tool for decision making. Washington, D.C.: Institute of Medicine, National Academy Press; 1999. [43] Insinga RP, Dasbach EJ, Myers ER. The health and economic burden of genital warts in a set of private health plans in the United States. Clin Infect Dis 2003;36(11):1397—403. [44] Insinga RP, Glass AG, Rush BB. The health care costs of cervical human papillomavirus-related disease. Am J Obstet Gynecol 2004;191(1):114—20. [45] Treasury Reporting Rates of Exchange. Financial Management Service: A bureau of the United States Treasury http://www.fms.treas.gov/intn.html#rates. Accessed: January 3, 2007. [46] Inflacion. Banco de Mexico http://www.banxico.org.mx/ PortalesEspecializados/inflacion/inflacion.html. Accessed: August 2, 2006. [47] Elbasha EH, Dasbach EJ, Insinga RP. A multi-type HPV transmission model assessing the economic impact of a quadrivalent HPV vaccine. 2006. Under review. [48] Munoz N, Bosch FX, Castellsague X, Diaz M, de Sanjose S, Hammouda D, et al. Against which human papillomavirus types shall we vaccinate and screen? The international perspective. Int J Cancer 2004;111(2):278—85. [49] Maw RD, Reitano M, Roy M. An international survey of patients with genital warts: perceptions regarding treatment and impact on lifestyle. Int J STD AIDS 1998;9(10):571—8. [50] Chuang TY, Perry HO, Kurland LT, Ilstrup DM. Condyloma acuminatum in Rochester, Minn., 1950—1978. I. Epidemiology and clinical features. Arch Dermatol 1984;120(4):469—75. [51] Wiley DJ, Douglas J, Beutner K, Cox T, Fife K, Moscicki AB, et al. External genital warts: diagnosis, treatment, and prevention. Clin Infect Dis 2002;35(Suppl 2):S210—24. [52] Arredondo A, Rangel R, de Icaza E. Cost-effectiveness of interventions for end-stage renal disease. Rev Saude Publica 1998;32(6):556—65. [53] Garcia-Contreras F, Nevarez-Sida A, Constantino-Casas P, AbudBastida F, Garduno-Espinosa J. Cost-effectiveness of chronic

139

[54]

[55]

[56]

[57]

[58]

[59]

[60]

[61]

[62]

[63]

hepatitis C treatment with thymosin alpha-1. Arch Med Res 2006;37(5):663—73. Gonzalez-Pier E, Gutierrez-Delgado C, Stevens G, BarrazaLlorens M, Porras-Condey R, Carvalho N, et al. Priority setting for health interventions in Mexico’s System of Social Protection in Health. Lancet 2006;368(9547):1608—18. Lieu TA, McGuire TG, Hinman AR. Overcoming economic barriers to the optimal use of vaccines. Health Aff (Millwood) 2005;24(3):666—79. Resch SC, Salomon JA, Murray M, Weinstein MC. Costeffectiveness of treating multidrug-resistant tuberculosis. Plos Med 2006;3(7):1048—57. The world factbook: Mexico. United States Central Intelligence Agency: https://cia.gov/cia//publications/factbook/ geos/mx.html#Econ. Accessed: December 6, 2006. Insinga RP. Annual productivity costs due to cervical cancer mortality in the United States. Womens Health Issues 2006 Sep;16(5):236—42. Gold MR, Siegel JE, Russell LB, Weinstein MC, editors. Costeffectiveness in health and medicine. New York, NY: Oxford University Press; 1996. Guidelines for the economic evaluation of health technologies: Canada. 3rd ed. Ottawa: Canadian Agency for Drugs and Technologies in Health; 2006. Harper DM, Franco EL, Wheeler CM, Moscicki AB, Romanowski B, Roteli-Martins M, 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 2006;367(9518): 1247—55. Gomez-Jauregui J. Costs and quality of the timely detection of cervix uteri cancer at a public clinica and at a non-government organization. Salud Publica Mex 2001;43(4):279—88. Palacio-Mejia LS, Rangel-Gomez G, Hernandez-Avila M, Lazcano-Ponce E. Cervical cancer, a disease of poverty: mortality differences between urban and rural areas in Mexico. Salud Publica Mex 2003;45(Suppl 3):S315—25.

Suggest Documents