Different Challenges in Eliminating HPV16 Compared to Other Types: A Modeling Study

The Jou nal o In ec ious Diseases 336 • JID 2017:216 (1 Augus ) • Baussano e al The Jou nal o In ec ious Diseases® 2017;216:336–44 Di e en Challenges in Elimina ing HPV16 Compa ed o O he Types: A Modeling S udy IacopoBaussano,1 Ful ioLazza a o,1,2 GuglielmoRonco,3 Ma iLeh inen,4,5 JoakimDillne ,5 and Sil iaF anceschi1 1In e na ional Agency o Resea ch on Cance , Lyon, F ance; 2Uni o Cance Epidemiology, AOU Ci à della Salu e e della Scienza, Hospi al o Tu in, and 3Depa men o Cance Sc eening, Cen e o Epidemiology and P e en ion in Oncology, Tu in, I aly; 4Uni e si y o Tampe e, Finland; and 5Depa men o Labo a o y Medicine, Ka olinska Ins i u e, Huddinge, Sweden Backg ound. Human papilloma i us (HPV) accina ion is s ill no eaching many high- isk popula ions. HPV16/18 accines o e c oss-p o ec ion agains o he ypes, o example, HPV45. Bo h di ec accine e icacy and indi ec he d p o ec ion con ibu e o accina ion e ec i eness. Me hods. We used a dynamic ansmission model, calib a ed o ce ical sc eening da a om I aly, o es ima e accina ion e ec- i eness agains HPV16 and HPV45 in ec ion, assuming o HPV45 ei he 95% o lowe c oss-p o ec ion. Resul s. Basic ep oduc i e numbe was smalle (2.1 s 4.0) and hence accine e ec i eness and he d p o ec ion s onge o HPV45 han o HPV16. The la ges di e ence in he educ ion o in ec ion p e alence in women <35 yea s old was ound a 70% co e age in gi ls-only accina ion p og ams (99% s 83% o o al p o ec ion o HPV45 and HPV16, espec i ely, mainly owing o s onge he d p o ec ion, ie, 37% s 16%). In gende -neu al accina ion, he la ges di e ence was a 40% co e age (he d p o ec ion, 54% s 28% o HPV16 and HPV45, espec i ely). Wi h ≥80% co e age, e en 50% c oss-p o ec ion would educe HPV45 by ≥94%. Conclusions. The cha ac e is ics o indi idual high- isk HPV ypes s ongly in luence he d p o ec ion and de e mine he le el o co e age and c oss-p o ec ion equi ed o educe o elimina e he in ec ion h ough HPV accina ion. HPV16 in ec ion and ela ed cance s a e he mos di icul o elimina e. Keywo ds. he d e ec ; HPV accina ion; c oss-p o ec ion; ce ical cance con ol. Cu en ly licensed i uslike pa icle accines agains human pap- illoma i us (HPV) a e nea ly 100% e icacious in he p e en ion o in ec ion om accine a ge ed HPV ypes [1]. Bo h he bi a- len (2V) and quad i alen (4V) accines a ge he high- isk (HR) ypes HPV16 and HPV18, which accoun o app oxima ely 70% o all ce ical cance s wo ldwide [2], whe eas he newe 9- alen accine also a ge s HR HPV31/33/45/52/58 [3], aising he p o- po ion o p e en able ce ical cance s o app oxima ely 90% [2]. Howe e , indings om andomized con olled ials [4–6] and popula ion-based su eys conduc ed a e he implemen a ion o HPV accina ion p og ams [7–10] showed ha he 2V accine and, o a lesse ex en , he 4V accine also o e some c oss-p o- ec ion agains o he HR ypes ha a e phylogene ically ela ed o HPV16 o HPV18, such as HPV31, HPV33, and HPV45. Adequa e accina ion co e age in adolescen gi ls (o in gi ls and boys, ie, gende -neu al accina ion) is a key condi ion o he a o able impac o accina ion. Minimal co e age h esh- olds o HPV con ol o elimina ion depend on he e icacy o he accine in accina ed indi iduals (di ec p o ec ion) and he s eng h o he d p o ec ion, ha is, he indi ec p o ec ion agains he in ec ion among un accina ed indi iduals [11]. In a sexually ansmi ed in ec ion, such as HPV in ec ion, he d p o ec ion is go e ned by he p obabili y o in ec ion ansmis- sion, he du a ion o he in ec ion, and sexual ac i i y pa e n, which a ies in di e en popula ions [11]. As a esul , he o e - all e ec i eness o HPV accina ion a a popula ion le el, ha is, he sum o accine e icacy and he d p o ec ion, is popula ion speci ic and, wi hin he same popula ion, ype speci ic. Empi ical da a ha e p o ided ea ly e idence o subs an ial he d p o ec ion in he ew yea s a e he s a o HPV acci- na ion p og ams [1, 7–10]. In he p esen epo , we used he In e na ional Agency o Resea ch on Cance ’s de e minis ic ansmission dynamic model [12, 13] o es ima e long- e m e ec i eness and he d p o ec ion by co e age, sepa a ely o gi ls-only and gende -neu al accina ion p og ams. The ocus is on HPV16, he mos p e alen [14] and mos ca cinogenic [15] ype, which is he mos ap o pe sis [5, 16], and HPV45, as an example o ano he ela i ely common ca cinogenic ype a ge ed by one bu no all a ailable HPV accines. METHODS S udy Popula ion and Assump ions Abou Sexual Beha io and Vaccina ion We simula ed a popula ion wi h a sex-equal and asso a i e sexual beha io cha ac e is ic o many high-income coun ies, MAJOR ARTICLE © The Au ho 2017. Published by Ox o d Uni e si y P ess o he In ec ious Diseases Socie y o Ame ica. This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion-NonComme cial-NoDe i s licence (h p://c ea i ecommons.o g/licenses/ by-nc-nd/4.0/), which pe mi s non-comme cial ep oduc ion and dis ibu ion o he wo k, in any medium, p o ided he o iginal wo k is no al e ed o ans o med in any way, and ha he wo k is p ope ly ci ed. Fo comme cial e-use, please con ac [email p o ec ed]. DOI: 10.1093/in dis/jix299 Recei ed 11 Ap il 2017; edi o ial decision 15 June 2017; accep ed 30 June 2017; published online July 5, 2017. Co espondence: I. Baussano, MD, MSc, PhD, In e na ional Agency o Resea ch on Cance , 150 Cou s Albe Thomas, 69372 Lyon Cedex 08, F ance ([email p o ec ed]). XX XXXX He d E ec o HPV Vaccina ion • JID 2017:216 (1 Augus ) • 337 ha is, wi h ela i ely simila sexual beha io in women and men, including simila age be ween pa ne s [13]. In pa icu- la , we simula ed he sexual beha io o he I alian popula ion as epo ed in a na ionwide popula ion-based su ey [17]. The obse ed and modeled age-speci ic p e alence o HPV16 and HPV45 in he p e accina ion e a is shown in Figu e 1A. The sim- ula ed popula ion was open s able and s a i ied by age ( ange, 10–70 yea s) and 3 classes o sexual ac i i y (high, in e media e, and low), cha ac e ized by an age-speci ic numbe o new sex- ual pa ne s pe yea (Supplemen a y Table S1). Sexual mixing is measu ed on a scale om 0 ( ully asso a i e, ie, like wi h like) o 1 ( andom mixing). In he p esen s udy, we se he asso a- i e mixing o sexually ac i i y classes o age g oups o ei he 0.30 (high) o 0.75 (low) as shown in Table 1. We also assumed ha sexual ac i i y did no s a be o e age 14 yea s and ha all women and men we e ini ially suscep ible o HPV16 and HPV45. Vaccine e icacy agains HPV16 and HPV45, when a ge ed by he accine, was se o be 95% [5]. The s eng h o c oss-p o- ec ion agains HPV45 in andomized con olled ials [4–6] and popula ion-based su eys [7, 9, 10, 19] is lowe and less consis en ly epo ed, and we he e o e in es iga ed 3 possible le els: 70%, 50%, and 30%. Finally, we assumed ha immuni y o HPV16 and HPV45 ( a ge ed and c oss-p o ec ion) om 3 o 2 doses (a 6-mon h in e als) was he same and li elong in bo h sexes [6, 20], and ha accine co e age in gende -neu al accina ion p og ams was he same in boys and gi ls. Model Pa ame e iza ion and Calib a ion In a p e ious a icle [12] we epo ed he calib a ion and alida- ion p ocess o ou model o 13 HR HPV ypes. B ie ly, 100 000 se s o pa ame e alues we e gene a ed by independen ly sampling, o each pa ame e , a uni o m dis ibu ion wi hin a p especi ied ange o alues, using a La in hype cube algo- i hm. Each se o alues was used o gene a e a model-based age-speci ic cu e o p e alence o each HPV ype. Finally, each model’s ou pu was compa ed wi h he obse ed age-spe- ci ic p e alence o each HR HPV ype, by calcula ing binomial log likelihood. We i ed ou model o HPV16 and HPV45 age-speci ic p e alence cu es in I alian women [12, 18, 21] by calib a ing he a e age pe sis ence (de e mined by he a e o clea ance) o HPV16 and HPV45 in ec ions [5]. Figu e 1B shows he pe capi a annual clea ance a es o inciden HPV16 and HPV45 in ec ion, as es ima ed by calib a ing he base-case model o ype and age-speci ic p e alence cu es om I aly. Ea ly HPV45 clea ance a e was highe (2.7 pe pe son-yea on a e age, o 22% pe mon h) han HPV16 clea ance a e (1.5 pe 7 6 5 4 P e alence, % 3 2 1 0 Ra e o Clea ance pe Pe son-Yea 0 .5 1 1.5 2 2.5 3 25–29 30–34 35–39 40–44 Age, y 0.511.5 2 2.5 3 3.544.555.5 6 6.5 Time Since In ec ion, y Obse ed da a (95% CIs) HPV-16 HPV-45 HPV-16 HPV-45 HPV-16 HPV-45 Model 45–49 50–54 55–59 Figu e 1. Age-speci ic p e alence o human papilloma i us (HPV)16 and HPV45 in he p e accina ion e a, including obse ed da a [18] and model ou pu s [12] (A) and a es o clea ance acco ding o model ou pu s (B). CIs, con idence in e als. 338 • JID 2017:216 (1 Augus ) • Baussano e al pe son-yea o 12% pe mon h). Clea ance a es o bo h ypes dec ease o e ime and con e ge a app oxima ely 0.14 pe pe - son-yea a e 6 yea s since in ec ion (Supplemen a y Table S2), in acco dance wi h empi ical e idence [22, 23]. Di e en pa ame e alues we e ei he assumed o calib a ed acco ding o a base-case scena io and a ious sensi i i y anal- yses (Table 1). In he base-case scena io, we se (1) he a e age du a ion o in ec ion o be 11 and 5 mon hs o HPV16 and HPV45, espec i ely; (2) he p obabili y o ansmission pe sexual pa ne ship o bo h HPV16 and HPV45 o be 70% in bo h sexes [12, 24, 25]; and (3) he p obabili y o de eloping ype-speci ic immuni y a e in ec ion clea ance, o bo h ypes, o be 30% in women and 0% in men [26]. Fo bo h HPV16 and HPV45, we also assessed he sensi i i y o model ou pu s o al e na i e sexual mixing pa e ns o cing asso a i e mixing by sexually ac i i y class and by age o be al e na i ely high (ie, 0.3) o low (ie, 0.75). A la ge numbe o di e en scena ios we e conside ed o HPV45 o assess he impac o di e en le els o c oss-p o ec ion om accines ha do no a ge HPV45 and he obus ness o model ou pu s o he unce ain y o ype-spe- ci ic na u al his o y, ha is, du a ion o in ec ion and immuni y a e clea ance [26] (Table 1). Fo each sensi i i y analysis, we used he single bes - i ing se o pa ame e s co esponding o he p especi ied assump ions. De ails abou he model s uc u e and calib a ing p ocess a e epo ed in he Supplemen a y Da a. Model-Based Analyses Fo each scena io, we calcula ed he basic ep oduc i e numbe (R0, ie, he a e age numbe o seconda y in ec ions esul ing om 1 case o ei he HPV16 o HPV45 in ec ion in a o ally Table 1. Model Pa ame e s Rela ed o Sexual Beha io , Vaccine Pe o mance, and In ec ion o HPV16 and HPV45 Pa ame e HPV16 HPV45 Base Case Sensi i i y Analysis by Sexual Asso a i e Mixing Base Casea Sensi i i y Analysis by In ec ion Du a ionb Immuni y A e Clea ancec Sexual Asso a i e Mixing High Low High Low New sexual pa ne s pe yea , mean, No. 1. 1 1. 1 1. 1 1. 1 1. 1 1. 1 1. 1 1. 1 Mixing be ween sexual ac i i y classesd 0.75 0.3 0.75 0.75 0.75 0.75 0.3 0.75 Mixing be ween age g oupsd 0.3 0.3 0.75 0.3 0.3 0.3 0.3 0.75 Vaccine e icacy, % 95 95 95 95 o less 95 o less 95 o less 95 o less 95 o less Du a ion o accine p o ec ion Li elong Li elong Li elong Li elong Li elong Li elong Li elong Li elong Assumed o calib a ed In ec ion du a ion, mean, mo 11e11 11 5e11 3e5 5 T ansmission p ob- abili y pe sexual pa ne ship, % 70 70 70 70 25e70 70 70 Immuni y a e in ec ion clea ance in women, % 30 30 30 30 30 0 30 30 Calcula ed Basic ep oduc i e numbe (R0) 4.1 5.6 2.9 2 1. 5 1. 3 2.7 1. 4 Gende -neu al accina ion co e - age su icien o elimina ion, % 75 81 67 49 38 21 62 30 Abb e ia ion: HPV, human papilloma i us. aSame ype-speci ic ansmission p obabili y as HPV16. bSame ype-speci ic in ec ion du a ion as HPV16. cAbsence o immuni y a e HPV45 clea ance. dThis is a measu e o he endency o indi iduals wi h simila sexual ac i i y o o m sexual pa ne ships. I is measu ed on a scale whe e ully and andomly asso a i e (ie, like-wi h-like) mixing co esponds o alues o 0 and 1, espec i ely. ePa ame e s ha ha e been calib a ed. Pa ame e s ha ha e been assumed. He d E ec o HPV Vaccina ion • JID 2017:216 (1 Augus ) • 339 suscep ible popula ion), using he nex -gene a ion ma ix me hod [27]. Fo model p edic ions, ou comes we e measu ed as he pe cen age p e alence educ ion (%PR) in HPV16 and HPV45 compa ed wi h no accina ion in women aged 15–34 yea s a pos accina ion equilib ium (ie, app oxima ely 50 yea s a e he in oduc ion o accina ion). The p ima y ou come was o e all accine e ec i eness. He d p o ec ion was he di e ence be ween o e all e ec i eness, es i- ma ed by he model, and di ec accine e icacy, es ima ed by mul iplying accina ion co e age by 95% e icacy o by he le el o c oss-p o ec ion agains HPV45, as epo ed. Gi ls-only and gende -neu al accina ion p og ams we e sepa a ely assessed. RESULTS In he base-case scena io, R0 es ima e is subs an ially la ge o HPV16 han o HPV45 (4.1 and 2.0, espec i ely) (Table 1), and his di e ence has a la ge in luence on he e ec i eness o ac- cina ion. Figu e 2A and 2B show he ype-speci ic %PR by co - e age om a accine a ge ing bo h ypes in ei he a gi ls-only o a gende -neu al accina ion p og am, espec i ely. The a ea be ween he cu e o o e all e ec i eness agains HPV45 and HPV16 and he s aigh cu e o 95% accine e icacy (com- mon o he 2 ypes) ep esen s he d p o ec ion. In a gi ls-only accina ion p og am, he la ges di e ence in o e all e ec- i eness be ween HPV16 and HPV45 s eadily inc eases up o app oxima ely 70% co e age (%PR, 99% o HPV45 s 83% o HPV16). The di e ence be ween ypes is en i ely accoun ed o by he la ge con ibu ion o he d p o ec ion o HPV45 han HPV16 (33% s 16% o he %PR, espec i ely) (Figu e 2A). In a gende -neu al accina ion p og am, he la ges di e ence in %PR is al eady eached a app oxima ely 40% co e age (92% o HPV45 s 66% o HPV16), wi h he d p o ec ion accoun - ing o mo e han hal o he o al %PR o HPV45 (Figu e 2B). Co e age o 75% and 49%, espec i ely, he e o e seems su i- cien in gende -neu al accina ion o he elimina ion HPV16 and HPV45 among women aged 15–34 yea s. Figu e 3 shows he e ec i eness o a accine ha o e ed only 50% c oss-p o ec ion agains HPV45 by co e age le el. A 70% co e age in gi ls-only accina ion, he %PR is 59% and he d p o ec ion accoun s o 24% o his e ec . I boys we e also ac- cina ed, he %PR om c oss-p o ec ion would inc ease o 86%, o which 51% is om he d p o ec ion. Sensi i i y analyses, o example, assuming no di e ence in du a ion o in ec ion be ween HPV16 and HPV45 o no immuni y a e clea ance o HPV45, had li le impac on he g ea e di icul y in elimina ing HPV16 compa ed wi h HPV45 shown by he base-case scena io (Table 1). Likewise, an inc ease and a dec ease in asso a i e mixing be ween sexual classes o age g oups inc eased and dec eased, espec- i ely, he co e age le el necessa y o elimina e bo h HPV16 and HPV45 (Table 1). Table 2 shows he o e all e ec i eness and he d p o ec- ion agains HPV16 and HPV45 acco ding o co e age le el, 100 90 80 Rela i e Reduc ion in P e alence, % 70 60 50 40 30 20 10 0 100 9080706050 Co e age, % Vaccine e icacy = 95% A B Vaccine e icacy = 95% Gi ls-only accina ion O e all p o ec ion s Di ec p o ec ion s HPV-45 HPV-16 HPV-16 and HPV-4 5 Gende -neu al accina ion O e all p o ec ion s Di ec p o ec ion s HPV-45 HPV-16 HPV-16 and HPV-45 403020100 100 90 80 Rela i e Reduc ion in P e alence, % 70 60 50 40 30 20 10 0 100 9080706050 Co e age, % 403020100 Figu e 2. Rela i e educ ion in p e alence o human papilloma i us (HPV)16 and HPV45, among all women aged 15–34 yea s a e accina ion o 11-yea -old gi ls (A) o gi ls and boys (B), by co e age and ype o p o ec ion. O e all p o ec ion s HPV-45 C oss-p o ec ion = 50% Gende -neu al accina ion Gi ls-only accina ion Di ec p o ec ion s HPV-45 100 90 80 Rela i e Reduc ion in P e alence, % 70 60 50 40 30 20 10 0 100 9080706050 Co e age, % 403020100 Figu e 3. Rela i e educ ion o p e alence o human papilloma i us (HPV)45 om 50% c oss-p o ec ion among all women aged 15–34 yea s a e accina ion o 11-yea -old gi ls only and gende -neu al accina ion, by co e age and ype o p o ec ion. 340 • JID 2017:216 (1 Augus ) • Baussano e al inclusion o boys, and, o HPV45, di e en assump ions abou accine e icacy and ype-speci ic na u al his o y (see Table 1). Compa ed wi h he base-case model, sensi i i y analyses p oduced highe %PR o nea ly all di e en combina ions o ac o s, mainly owing o s onge he d p o ec ion. The da a in Table 2 he e o e con i m he gene al inding ha HPV45 is eas- ie o elimina e han HPV16 and ha he highe he co e age, he lowe he con ibu ion o accina ing boys in addi ion o gi ls. Figu e 4A and 4B show he %PR o HPV45 om gende -neu- al accina ion by co e age in 2 hypo he ical popula ions wi h sexual mixing pa e ns di e en om ha in he base-case scena io. In a popula ion wi h high asso a i e mixing (Figu e 4A), he cu es o HPV45 and HPV16 become close , and he di e ence in he d p o ec ion diminishes. Con e sely, he di - e ence in %PR be ween he 2 HPV ypes in a popula ion wi h low asso a i e mixing (Figu e 4B) is e en la ge han in he base-case scena io (Figu e 2B), and he d immuni y inc eases. This di e ence eaches a maximum a 30% co e age (al eady su icien o elimina e HPV45). Figu e 5 p esen s a compa ison o he %PR o HPV45 ha can be achie ed by a accine’s 50% c oss-p o ec ion by co e age and sexual mixing. The impac o such a accine is much s on- ge in a popula ion wi h low asso a i e mixing; in such a pop- ula ion, HPV45 elimina ion is seen a 60% co e age whe eas he same co e age app oxima ely hal es HPV45 p e alence in a popula ion wi h high asso a i e mixing. Supplemen a y Table S3 shows de ailed es ima es o %PR and he d p o ec ion o gi ls-only accina ion and se e al le els o c oss-p o ec ion. Changes in R0 es ima e unde lie he indings o sensi i - i y analyses. The assump ion o he same in ec ion du a ion o HPV16 and HPV45, and o no acqui ed immuni y a e HPV45 clea ance, sligh ly dec eases he HPV45 R0 ( om 1.5 o 1.3) (Table 1). Changes in he asso a i eness o sexual mixing p oduce opposi e e ec s (Table 1). In a highly asso a i e pop- ula ion, he R0 inc eases o HPV16 ( om 4.1 in he base-case model o 5.6) and HPV45 ( om 2.0 o 2.7). Con e sely, in a popula ion wi h low asso a i e mixing, he R0 dec eases o 2.9 o HPV16 and 1.4 o HPV45, hus explaining he as p e- dominance o he he d e ec in Figu e 4B. DISCUSSION Ou p esen epo highligh s he impo an ole o he d p o- ec ion in he elimina ion o HPV in ec ion, especially o HPV ypes less p one o long- e m pe sis ence han HPV16. I also shows ha o any le el o popula ion co e age, and also in he p esence o less han 95% accine e icacy, he success o a accina ion p og am is la ge he lowe he p e accina ion p e alence o an indi idual HPV ype. A ≥60% co e age in gende -neu al p og ams, o ins ance, e en 50% c oss-p o- ec ion could elimina e HPV45. This phenomenon is mainly due o la ge he d p o ec ion o HR ypes such as HPV45 ha a e less able han HPV16 o induce long- e m pe sis ence [5, 16]. In ac , o any in ec ion, he magni ude o he d p o ec ion Table 2. Pe cen age Reduc ion in HPV16 and HPV45 P e alence Among Women Aged 15–34 Yea s A e Gi ls-Only o Gende -Neu al Vaccina ion Compa ed Wi h No Vaccina ion by Vaccine E icacy, Co e age, and Biological Scena ios Vaccine E icacy, % Co e age, % Reduc ion in P e alence, % (He d P o ec ion, %) Gi ls-Only Vaccina ion Gende -Neu al Vaccina ion Base Case Sensi i i y Analysis (HPV45) by Base Case Sensi i i y Analysis (HPV45) by HPV16 HPV45aIn ec ion Du a ionb Immuni y A e Clea ancecHPV16 HPV45aIn ec ion Du a ionb Immuni y A e Clea ancec 95 40 51 (13) 63 (25) 93 (55) 100 (62) 67 (29) 90 (52) 99 (61) 100 (62) 60 74 (17) 89 (32) 99 (42) 100 (43) 90 (33) 100 (43) 100 (43) 100 (43) 80 92 (16) 100 (24) 100 (24) 100 (24) 100 (24) 100 (24) 100 (24) 100 (24) 70d40 … 47 (19) 80 (52) 84 (56) … 69 (41) 97 (69) 100 (72) 60 … 68 (26) 95 (53) 100 (58) … 96 (54) 100 (58) 100 (58) 80 … 88 (32) 99 (43) 100 (44) … 100 (44) 100 (44) 100 (44) 50d40 … 34 (14) 63 (43) 63 (43) … 51 (31) 51 (31) 99 (79) 60 … 50 (20) 84 (54) 89 (59) … 73 (43) 73 (43) 100 (70) 80 … 65 (25) 94 (54) 100 (60) … 94 (54) 94 (54) 100 (60) 30d40 … 21 (9) 40 (28) 40 (28) … 31 (19) 66 (54) 68 (56) 60 … 31 (13) 58 (40) 58 (40) … 46 (28) 85 (67) 94 (76) 80 … 41 (17) 72 (48) 74 (50) … 60 (36) 94 (70) 100 (76) Abb e ia ion: HPV, human papilloma i us. aSame ype-speci ic ansmission p obabili y as HPV16. bSame ype-speci ic in ec ion du a ion as HPV16. cAbsence o immuni y a e HPV45 clea ance. dLe el o c oss-p o ec ion. He d E ec o HPV Vaccina ion • JID 2017:216 (1 Augus ) • 341 agains an indi idual HPV ype is go e ned by i s ansmission po en ial (R0), which depends on he du a ion o he in ec- ious pe iod [11, 28]. Ou es ima e o R0 o HPV16 (4.1) was la ge han ha o HPV45 (2.0) in ou base-case scena io ha assumes he same ansmission p obabili y and immuni y a e clea ance and ound an app oxima ely 2- old longe du a ion o HPV16 han o HPV45 in ec ion. Al e na i e hypo heses ha assumed no di e ence in in ec ion du a ion be ween he 2 ypes and no na u al immuni y o HPV45 showed, an e en g ea e di e ence be ween he 2 ypes. Ano he enhance o he d p o ec ion is accina ing boys in addi ion o gi ls. The bene i o gende -neu al accina- ion is he e o e mos impo an i ei he co e age o accine e icacy a e subop imal, o example, wi h co e age <60% o he elimina ion o HPV16 using an HPV16 a ge ing accine o wi h pa ial c oss-p o ec ion only agains HPV45. We also p ojec ed he e ec s o di e en accina ion scena ios on he e en ual decline o ce ical cance due o HPV16 and HPV45 in a coun y such as I aly (Supplemen a y Table S3). As o he co esponding in ec ions, he elimina ion o HPV16- ela ed ce ical cance equi es he highes le el o accine e icacy and co e age, whe eas pa ial c oss-p o ec ion may be su icien o elimina e ce ical cance associa ed wi h HPV45 and possibly o he HR HPV ypes ha may sha e wi h HPV45 a sho e in ec ion du a ion and less abili y o p oduce cance han HPV16 and HPV18. We hink ha he assump ions used in he base-case scena io bes e lec ed he indings o s udies on HPV na u al his o y, o example, lack o subs an ial he e ogenei y be ween ca cino- genic HPV ypes in ansmission a es [12, 24, 25, 29] and in he p obabili y o de eloping ype-speci ic immuni y a e in ec ion clea ance [26]. In con as , he du a ion o ca cinogenic HPV in ec ion has been shown o be he e ogeneous by ype [5, 16], wi h HPV16 he ype he mos ap o pe sis , hus i s g ea e abili y o induce long- e m pe sis ence and malignan ans o - ma ion han o he HR ypes [14, 30]. Indeed, he es ima e o he di e ence in he a e age du a ion o HPV16 and HPV45 (11 and 5 mon hs, espec i ely) calib a ed by he base-case scena io was e en la ge han in p e ious wo k, o example, 12 and 8 mon hs, acco ding o he la ges published me a-analysis [16]. In ac , he a e age du a ion o HPV16 in ec ion was gene ally e alua ed o e ela i ely sho ollow-up pe iods and p edom- inan ly among cy ologically no mal women [16], o was un- ca ed by de ec ion and ea men o HPV16-associa ed lesions. Ne e heless, we pe o med se e al sensi i i y analyses based on assump ions abou he na u al his o y o HPV45 o sexual mixing ha di e ed om hose used in he base-case scena io. They con i med he g ea e di icul y o elimina ing HPV16 compa ed wi h HPV45. Only high asso a i eness in sexual mixing be ween sexual classes and age g oups aised he HPV45 R0 su icien ly o a enua e he di e ence in ype-speci ic he d immuni y. Vaccine e icacy = 95% Vaccine e icacy = 95% O e all p o ec ion s Asso a i e sexual mixing: high Di ec p o ec ion s HPV-45 HPV-16 HPV-16 and HPV-45 Asso a i e sexual mixing: low O e all p o ec ion s Di ec p o ec ion s HPV-45 HPV-16 HPV-16 and HPV-45 100 A B 90 80 Rela i e Reduc ion in P e alence, % 70 60 50 40 30 20 10 0 100 9080706050 Co e age, % 403020100 100 90 80 Rela i e Reduc ion in P e alence, % 70 60 50 40 30 20 10 0 100 9080706050 Co e age, % 403020100 Figu e 4. Rela i e educ ion o p e alence o human papilloma i us (HPV)16 and HPV45, among all women aged 15–34 yea s a e accina ion o 11-yea -old gi ls and boys in a popula ion wi h high (A) o low (B) asso a i e sexual mixing, by co - e age and ype o p o ec ion. O e all p o ec ion s HPV-45 Di ec p o ec ion s HPV-45 C oss-p o ec ion = 50% Asso a i e sexual mixing: low Asso a i e sexual mixing: high 100 90 80 Rela i e Reduc ion in P e alence, % 70 60 50 40 30 20 10 0 100 9080706050 Co e age, % 403020100 Figu e 5. Rela i e educ ion o p e alence o human papilloma i us (HPV) 45 om 50% c oss-p o ec ion among all women aged 15–34yea s a e accina ion o 11-yea -old gi ls and boys in popula ions wi h high o low asso a i e sexual mixing, by co e age and ype o p o ec ion. 342 • JID 2017:216 (1 Augus ) • Baussano e al The di e ences be ween HPV16 and HPV45 he ein epo ed ag ee wi h he indings o B isson e al [31, 32], who showed s onge accine e ec i eness agains HPV11/6 ypes han agains HPV16/18 owing o di e ences in du a ion o in ec- iousness and, hence, in R0. Indeed, hese ela ionships be ween HPV ypes may be gene alizable, because he R0 ange calcu- la ed by ou model (2.1–4.0 in he baseline scena io) plausibly e lec s hose o he mos impo an ca cinogenic HPV ypes. Ou es ima es all wi hin he ange o o he sexually ansmi ed in ec ions, such as human immunode iciency i us [33], syph- ilis [34], gono hea, and chlamydia [35]. Con e sely, R0 alues o sexually ansmi ed in ec ions a e subs an ially lowe han hose o childhood epidemic in ec ions (eg, measles, whooping cough, ubella, and poliomyeli is), which equi e >90% co e - age o be con olled [36]. The s eng hs o he p esen s udy include he use o a ali- da ed ansmission model o ep esen changes in HPV p e - alence. T ansmission models can cap u e he dynamics o in ec ion ci cula ion [11] in a popula ion and ha e he dis inc ad an age o including he e ec o he d immuni y a ibu - able o accina ion [37]. We could also de i e es ima es o he pa ame e s go e ning he na u al his o y o HPV16 and HPV45 in ec ions om he calib a ion o a la ge clinical ial conduc ed in I aly [18], whe eby we we e able o p edic accu- a ely he incidence o HPV in ec ion in HPV nega i e women. This allowed us o p o ide a ange o unce ain y o each pa ame e es ima e [12]. The limi a ions o he p esen s udy mainly de i e om he unce ain ies ha emain in some o he model assump ions. C oss-p o ec ion agains HPV45 in clinical ials, o ins ance, anged be ween 8% and 79% and was consis en ly highe o 2V han o 4V accine [5, 6]. Popula ion-based s udies om pos accina ion su eys also showed a pa ial e icacy agains HPV31/33/45 as a combined end poin o app oxima ely 40% o 4V [9] and 50% o 2V [7, 10] accine. P elimina y da a also sugges ha c oss-p o ec ion agains HPV31/33/45 is compa- able o 2 o 3 doses [6, 7, 9, 10] and may be o long du a ion [6, 20], and ha p ecance ous lesions de i ing om non accine HPV ypes a e also subs an ially educed by 2V accina ion [38]. Ob iously, he sexual beha io o women and men is pop- ula ion-speci ic, and i s desc ip ion is always an o e simpli i- ca ion because accu a e in o ma ion on sexual ne wo ks (eg, sequen ial o concu en sexual pa ne ships) a e e y di icul o ob ain. In conclusion, he cha ac e is ics o indi idual HR HPV ypes s ongly in luence he d immuni y and de e mine he le el o co e age and ype-speci ic accine e icacy (including c oss-p o ec ion) ha a e equi ed o educe o elimina e he in ec ion h ough HPV accina ion. HPV16 is ha de o elim- ina e han HPV45 and, p obably, any o he ype. Ou indings a e pa icula ly ele an o low- and middle-income coun ies ha a e especially challenged by p og amma ic di icul ies [39] and inc eases in he cos o accines acco ding o he numbe o a ge ed ypes [40]. Supplemen a yDa a Supplemen a y ma e ials a e a ailable a The Jou nal o In ec ious Diseases online. Consis ing o da a p o ided by he au ho s o bene i he eade , he pos ed ma e ials a e no copyedi ed and a e he sole esponsibili y o he au ho s, so ques ions o com- men s should be add essed o he co esponding au ho . No es Acknowledgmen s. We hank Blandine de Fleu ieu and Simopekka Vänskä o hei commen s and help ul discussion ega ding his s udy. Financial suppo . This wo k was suppo ed by he Eu opean Commission’s Se en h F amewo k P og amme, FP7- HEALTH-2013 (g an 603019), by he Canadian Ins i u es o Heal h Resea ch Founda ion (g an 334069 o I.B., F.L., and S. F.), and by he Bill & Melinda Ga es Founda ion (g an OPP1053353 o I.B., F.L., and S.F.). Po en ial con lic s o in e es s. J. D. has ecei ed p e ious g an s o his ins i u ion om Me ck, a manu ac u e o HPV ac- cines. All o he au ho s epo no po en ial con lic s o in e es . All au ho s ha e submi ed he ICMJE Fo m o Disclosu e o Po en ial Con lic s o In e es . Con lic s ha he edi o s conside ele an o he con en o he manusc ip ha e been disclosed. Re e ences 1. Chow EP, Danielewski JA, Fehle G, e al. Human papil- loma i us in young women wi h Chlamydia achoma is in ec ion 7yea s a e he Aus alian human papilloma i- us accina ion p og amme: a c oss-sec ional s udy. Lance In ec Dis 2015; 15:1314–23. 2. Se ano B, de Sanjosé S, Tous S, e al. Human papilloma i- us geno ype a ibu ion o HPVs 6, 11, 16, 18, 31, 33, 45, 52 and 58 in emale anogeni al lesions. Eu J Cance 2015; 51:1732–41. 3. Jou a EA, Giuliano AR, I e sen OE, e al; B oad Spec um HPV Vaccine S udy. A 9- alen HPV accine agains in ec- ion and in aepi helial neoplasia in women. N Engl J Med 2015; 372:711–23. 4. Malagón T, D ole M, Boily MC, e al. C oss-p o ec i e e i- cacy o wo human papilloma i us accines: a sys ema ic e iew and me a-analysis. Lance In ec Dis 2012; 12:781–9. 5. Leh inen M, Dillne J. Clinical ials o human papillo- ma i us accines and beyond. Na Re Clin Oncol 2013; 10:400–10. 6. K eime AR, S uy F, Del Rosa io-Raymundo MR, e al; Cos a Rica Vaccine T ial S udy G oup Au ho s; PATRICIA S udy G oup Au ho s; HPV PATRICIA P incipal In es iga o s/Co-P incipal In es iga o Collabo a o s; GSK Vaccines Clinical S udy Suppo G oup. E icacy o ewe He d E ec o HPV Vaccina ion • JID 2017:216 (1 Augus ) • 343 han h ee doses o an HPV-16/18 AS04-adju an ed ac- cine: combined analysis o da a om he Cos a Rica Vaccine and PATRICIA T ials. Lance Oncol 2015; 16:775–86. 7. Came on RL, Ka anagh K, Pan J, e al. Human papilloma- i us p e alence and he d immuni y a e in oduc ion o accina ion p og am, Sco land, 2009–2013. Eme g In ec Dis 2016; 22:56–64. 8. D ole M, Béna d É, Boily MC, e al. Popula ion-le el impac and he d e ec s ollowing human papilloma i us accina- ion p og ammes: a sys ema ic e iew and me a-analysis. Lance In ec Dis 2015; 15:565–80. 9. Tab izi SN, B o he on JM, Kaldo JM, e al. Assessmen o he d immuni y and c oss-p o ec ion a e a human pap- illoma i us accina ion p og amme in Aus alia: a epea c oss-sec ional s udy. Lance In ec Dis 2014; 14:958–66. 10. Cuschie i K, Ka anagh K, Moo e C, Bha ia R, Lo e J, Pollock KG. Impac o pa ial bi alen HPV accina ion on accine- ype in ec ion: a popula ion-based analysis. B J Cance 2016; 114:1261–4. 11. Ga ne GP. Role o he d immuni y in de e mining he e ec o accines agains sexually ansmi ed disease. J In ec Dis 2005; 191(suppl 1):S97–106. 12. Baussano I, El s öm KM, Lazza a o F, e al. Type-speci ic human papilloma i us biological ea u es: alida ed mod- el-based es ima es. PLoS One 2013; 8:e81171. 13. Baussano I, Lazza a o F, B isson M, F anceschi S. Human papilloma i us accina ion a a ime o changing sexual beha io . Eme g In ec Dis 2016; 22:18–23. 14. Schi man M, Doo ba J, Wen zensen N, e al. Ca cinogenic human papilloma i us in ec ion. Na Re Dis P ime s 2016; 2:16086. 15. Guan P, Howell-Jones R, Li N, e al. Human papilloma i- us ypes in 115,789 HPV-posi i e women: a me a-anal- ysis om ce ical in ec ion o cance . In J Cance 2012; 131:2349–59. 16. Rosi ch AF, Koshiol J, Hudgens MG, e al. Pa e ns o pe - sis en geni al human papilloma i us in ec ion among women wo ldwide: a li e a u e e iew and me a-analysis. In J Cance 2013; 133:1271–85. 17. Signo elli C, Pasqua ella C, Limina RM, e  al. Thi d I alian na ional su ey on knowledge, a i udes, and sex- ual beha iou in ela ion o HIV/AIDS isk and he ole o heal h educa ion campaigns. Eu J Public Heal h 2006; 16:498–504. 18. Ronco G, Gio gi-Rossi P, Ca ozzi F, e al; New Technologies o Ce ical Cance sc eening (NTCC) Wo king G oup. E icacy o human papilloma i us es ing o he de ec ion o in asi e ce ical cance s and ce ical in aepi helial neo- plasia: a andomised con olled ial. Lance Oncol 2010; 11:249–57. 19. Meshe D, Soldan K, Leh inen M, e al. Popula ion-le el e ec s o human papilloma i us accina ion p og ams on in ec ions wi h non accine geno ypes. Eme g In ec Dis 2016; 22:1732–40. 20. Sa aeian M, Po as C, Pan Y, e al; CVT G oup. Du able an i- body esponses ollowing one dose o he bi alen human papilloma i us L1 i us-like pa icle accine in he Cos a Rica Vaccine T ial. Cance P e Res (Phila) 2013; 6:1242–50. 21. Ca ozzi F, De Ma co L, Gillio-Tos A, e al; NTCC Wo king G oup. Age and geog aphic a iabili y o human papillo- ma i us high- isk geno ype dis ibu ion in a la ge un ac- cina ed popula ion and o accina ion impac on HPV p e alence. J Clin Vi ol 2014; 60:257–63. 22. Plumme M, Schi man M, Cas le PE, Mauco -Boulch D, Wheele CM; ALTS G oup. A 2-yea p ospec i e s udy o human papilloma i us pe sis ence among women wi h a cy ological diagnosis o a ypical squamous cells o unde e - mined signi icance o low-g ade squamous in aepi helial lesion. J In ec Dis 2007; 195:1582–9. 23. Rod íguez AC, Schi man M, He e o R, e  al; P oyec o Epidemiológico Guanacas e G oup. Rapid clea ance o human papilloma i us and implica ions o clinical ocus on pe sis en in ec ions. J Na l Cance Ins 2008; 100:513–7. 24. Bogaa ds JA, Xi idou M, Coupé VM, Meije CJ, Wallinga J, Be kho J. Model-based es ima ion o i al ansmissibili y and in ec ion-induced esis ance om he age-dependen p e alence o in ec ion o 14 high- isk ypes o human papilloma i us. Am J Epidemiol 2010; 171:817–25. 25. Vänskä S, Au anen K, Leino T, e al. Impac o accina- ion on 14 high- isk HPV ype in ec ions: a ma hema i- cal modelling app oach. PLoS One 2013; 8:e72088. 26. Beachle DC, Jenkins G, Sa aeian M, K eime AR, Wen zensen N. Na u al acqui ed immuni y agains subse- quen geni al human papilloma i us in ec ion: a sys ema ic e iew and me a-analysis. J In ec Dis 2016; 213:1444–54. 27. Vynnycky E, Whi e RG. Sexually ansmi ed in ec ions: an in oduc ion o in ec ious disease modelling. Ox o d, UK: Ox o d Uni e si y P ess, 2010:223–81. 28. Baussano I, F anceschi S, Plumme M. In ec ion ans- mission and ch onic disease models in he s udy o in ec- ion-associa ed cance s. B J Cance 2014; 110:7–11. 29. Bu chell AN, Cou lée F, Tellie PP, Hanley J, F anco EL. Geni al ansmission o human papilloma i us in ecen ly o med he e osexual couples. J In ec Dis 2011; 204:1723–9. 30. Schi man M, He e o R, Desalle R, e al. The ca cinogenic- i y o human papilloma i us ypes e lec s i al e olu ion. Vi ology 2005; 337:76–84. 31. B isson M, Van de Velde N, Boily MC. Di e en popula- ion-le el accina ion e ec i eness o HPV ypes 16, 18, 6 and 11. Sex T ansm In ec 2011; 87:41–3. 32. B isson M, an de Velde N, F anco EL, D ole M, Boily MC. Inc emen al impac o adding boys o cu en human pap- illoma i us accina ion p og ams: ole o he d immuni y. J In ec Dis 2011; 204:372–6. 344 • JID 2017:216 (1 Augus ) • Baussano e al 33. Nsubuga RN, Whi e RG, Mayanja BN, Sha e LA. Es ima ion o he HIV basic ep oduc ion numbe in u al Sou h Wes Uganda: 1991–2008. PLoS One 2014; 9:e83778. 34. G assly NC, F ase C, Ga ne GP. Hos immuni y and syn- ch onized epidemics o syphilis ac oss he Uni ed S a es. Na u e 2005; 433:417–21. 35. B unham RC, Nagelke ke NJ, Plumme FA, Moses S. Es ima ing he basic ep oduc i e a es o Neisse ia gon- o hoeae and Chlamydia achoma is: he implica ions o acqui ed immuni y. Sex T ansm Dis 1994; 21:353–6. 36. Ande son RM, May RM. Di ec ly ansmi ed in ec ions diseases: con ol by accina ion. Science 1982; 215:1053–60. 37. B isson M, Béna d É, D ole M, e  al. Popula ion-le el impac , he d immuni y, and elimina ion a e human papilloma i us accina ion: a sys ema ic e iew and me a-analysis o p edic ions om ansmission-dynamic models. Lance Public Heal h 2016; 1:e8–17. 38. Leh inen M, Paa onen J, Wheele CM, e al; HPV PATRICIA S udy G oup. O e all e icacy o HPV-16/18 AS04-adju an ed accine agains g ade 3 o g ea e ce ical in aepi helial neo- plasia: 4-yea end-o -s udy analysis o he andomised, dou- ble-blind PATRICIA ial. Lance Oncol 2012; 13:89–99. 39. B uni L, Diaz M, Ba ionue o-Rosas L, e al. Global es i- ma es o human papilloma i us accina ion co e age by egion and income le el: a pooled analysis. Lance Glob Heal h 2016; 4:e453–63. 40. Wu T, Hu YM, Li J, e al. Immunogenici y and sa e y o an E. coli-p oduced bi alen human papilloma i us ( ype 16 and 18) accine: a andomized con olled phase 2 clinical ial. Vaccine 2015; 33:3940–6.