Comment on SCHIM VAN DER LOEFF et al. 2019 – Should female sex workers be offered HPV vaccinationffam

68 Aufrufe

PDF an Freunde weiterleiten:

Ihre IP-Adresse wird aus Sicherheitsgründen gespeichert um kriminelle Aktivitäten und unerlaubten Spam zu unterbinden. Leiten Sie nur E-Mails weiter, wo der Empfänger mit dem Versand auch einverstanden ist.

Comment on: 
 
SCHIM VAN DER LOEFF et al. (2019): 
 
Ä6KRXOGIHPDOHVH[ZRUNHUVEHRIIHUHG+39YDFFLQDWLRQ"³ 
 
 
 
 
In May 2019, the first ever article about the pros and cons of HPV vaccination of FSWs in 
a  peer-reviewed  scientific  journal was  published LQ Ä+XPDQ 9DFFLQHV 	
�,�P�P�X�Q�R�W�K�H�U�D�S�H�X�W�L�F�V����������-  8: 1544 ± 1548), entitled: 
 
Ä6KRXOGIHPDOHVH[ZRUNHUVEHRIIHUHG+39YDFFLQDWLRQ?³ 
 
The authors, M.F. SCHIM VAN DER LOEFF, A. VORSTERS, E. MARRA, P. VAN 
DAMME and A. HOGEWONING, belong to the Public Health Service of Amsterdam, the 
University  of  Amsterdam  (Public  Health  Research  Institute,  Amsterdam  Infection  & 
Immunity Institute, Department of Dermatology) and the ÄCentre for the Evaluation of 
Vaccination³ at the  University of Antwerpen.   
 
The article of SCHIM VAN DER LOEFF et al. is the first paper about the question of HPV 
vaccination of FSWs in a peer-reviewed journal. There were some former papers with 
respect to the implementation of HPV vaccination for FSWs, e.g. about a trial in Lima/Peru 
%52:1HWDOEXWWKH\GLGQ¶WFRQVLGHULQGHWDLOWKHSURVDQGFRQVRI
such vaccination and the question whether vaccination is still effective at all when it is 
administered to a highly pre-exposed adult female population like FSWs. The article of 
SCHIM VAN DER LOEFF is restricted to female sex workers; however, many aspects are 
applicable to MSWs and transgender SWs as well. 
 
SCHIM VAN DER LOEFF et al. note the high HPV risks of FSWs, the limited protection 
of condoms, the limited use of condoms (in part depending on the sex work setting) and the 
young age (~ 50 % < 26 years) of many FSWs. 
 
 
The authors mention the following arguments in favor of vaccination: 
 
ƔKLJK risks of HPV infection, the precursors of cervical cancer, and thus cervical cancer in 
FSWs; higher risk for anogenital warts and ± possibly/probably ± anogenital cancers   
ƔXQOLNHOLQHVVWKDW)6:VDUHDOUHDG\YDFFLQDWHGDJDLQVW+39 
ƔWKHH[FHOOHQWLPPXQRJHQLFLW\RIWKHYDFFLQHHYHQLQZRPHQZKRZHUHSUHYLRXVO\H[SRVHG
to HPV types from the vaccine  
ƔUHFXUUHQFHUDWHVDIWHUVXUJLFDOWUHDWPHQWRI&,1DUHlower in vaccinated than in non-
vaccinated women, so even infected FSW may profit from vaccination 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-2- 
 
 
ƔWKHHIILFDF\RIWKHYDFFLQHLQPLG-adult women (26 ± 45 years) against persistent infections 
and CIN2+, as demonstrated by several RCTs (randomized control trials), at least if the 
women are not infected with the HPV types from the vaccine at the time of vaccination. 
Moreover, seropositivity against HPV 16/18 as a marker for a previous infection by these 
+39W\SHVGRHVQ¶WVHHPWRLPSDLUWKHHIILFDF\RIYDFFLQDWLRQDORW 
ƔFSWs who may be currently infected with one or two HPV types from the vaccine may 
still profit from protection against the other HPV types of the vaccine. For example, 64 % 
of  FSWs  in  Amsterdam are ³QDwYH´ ZLWK UHJDUG WR +39 (PCR-negative and  sero-
negative), and these percentages are even higher for other hrHPV types from the vaccine. 
ƔSURWHFWLRQRIFOLHQWVWKHLURWKHUVH[XDOSDUWQHUVDQGKHUGLPPXQLW\ZLWKHIIHFWVRQWKH
general population. Vaccine-induced neutralizing antibodies on the surface of the female  
genital tract prevent further transmission to clients (even in FSWs who are infected). The 
authors calculated that about 500.000 HPV transmissions can be prevented each year, if all 
~ 5000 FSWs in Amsterdam are vaccinated.  
 
 
The authors note the following arguments against vaccination (discussed  in detail 
below): 
 
ƔWKH$WULDOZKLFKVKRZHGQRstatistically significant efficacy of Gardasil vaccination 
against persistent anal infections in a heavily pre-exposed adult cohort, dominated by HIV-
positive MSM   
ƔZRPHQZKRDOUHDG\FOHDUHGD+39LQIHFWLRQPD\EHSURWHFWHGE\QDWXUDOLPPXQLW\DQG
may thus be able to clear a reinfection even without being vaccinated in the meantime 
ƔQRWDOO+39LQIHFWLRQVLQZRPHQUHVXOWLQVHUXPDQWLERGLHVVRWKHUHPD\EHwomen/FSWs 
presumed to be HPV-naïve (PCR-negative and seronegative) who were in fact infected in 
the past and cleared this infection without leaving detectable titres of antibodies, so the 
SHUFHQWDJHRI³WUXO\´QDwYH)6:IRU+39LQ$PVWHUGDPPD\be overestimated 
ƔWKHXQUHVROYHGUROHRIODWHQF\,WLVQRWH[SHFWHGWKDWYDFFLQH-induced antibodies are able 
to clear latent infections (the virus is intracellular and antibodies cannot bind to it). It is 
unknown whether vaccination can suppress recrudescence of latent infections. 
ƔDFDVH-control study of CIN 2+ prevalence in vaccinated vs. non-vaccinated women which 
investigated the role of different doses of Gardasil (³1 or 2´ versus ³3 and more´) and ages 
at  the  time  of  the  first  vaccine  dose  showed no (strictly  speaking:  only  a  small and 
insignificant)  effect  in  those  who  got  the  first  dose  at  an  age  of  21  years  or  beyond 
(SILVERBERG et at 2018).  
 
In their discussion, SCHIM VAN DER LOEFF et al. pointed out that three different aspects 
of HPV vaccination of FSWs have to be considered: the impact on an individual FSW, on 
the population of FSWs, and on the general population. ³7KHUHDUHQRFOHDU-cut data about 
the effectiveness of HPV vaccination of highly, pre-exposed women, while indirect evidence 
VXJJHVWVVRPHGHJUHHRIHIIHFWLYHQHVV´(SCHIM VAN DER LOEFF, p. 1546).  Furthermore, 
the authors state that HPV vaccination of FSWs ³FRXOGSURYLGHKHDOWKEHQHILWVIRUWKHPDOH
and femDOHSRSXODWLRQDWODUJH´ 
 
On an individual base, they pointed out ³QRWDOO6:DUHWKHVDPH´and in individual cases 
the balance between pros and cons may favor vaccination, for example in women who 
 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-3- 
 
 
 recently started sex work or plan to do so in the future and who had little vaginal/anal sexual 
experience in the past ³WKHQ+39YDFFLQDWLRQVKRXOGFHUWDLQO\EHUHFRPPHQGHG´   
 
The authors see the need for more data and more research, including studies on the long-
term effect of vaccination in highly pre-exposed women, but acknowledge that a RCT in a 
SW population to study the effectiveness of the vaccine in this group would be difficult. The 
article concludes with the sentence: ³,QLQGLYLGXDOFDVHV+39YDFFLQation of SW may be 
UHFRPPHQGHG´ 
 
 
 
Comment:  
  
The article is very interesting and important, because it is the first article which discusses 
the pros and cons of HPV vaccination of FSWs in a peer-reviewed scientific journal. It took 
13 years after the implementation of the HPV vaccine until the first article on this subject 
was published. This means a big progress in this field and a base for further discussion and 
research.   
 
I agree with all of the content of the paper. This is not a big surprise, because we used nearly 
completely the same literature. I especially agree with the three different perspectives on the 
subject in the discussion: (i) that HPV vaccination of as many FSWs as possible has a 
favorable effect on the population at large, the clients of FSWs and the private partners of 
the clients and FSWs (in the sense of herd immunity), (ii) that not all FSWs are the same and 
that there are FSWs to whom FSW vaccination should be (strongly) recommended (need for 
balance of pros and cons for each individual FSW), and (iii) that there is currently no 
evidence to recommend HPV vaccination to all FSWs in general, furthermore (iv) that there 
is a lack of direct evidence for or against HPV vaccination of FSWs because there are no 
such studies in the FSW population, and (v) that it is very difficult and (depending on the 
type of trial, e.g. RCT) even impossible to carry out HPV vaccination trials within the FSW 
population, at least as far as the effectiveness of the vaccine for FSWs is concerned (of course 
one can carry out trials with regard to the acceptance of the vaccine by FSWs and similar 
subjects, see BROWN et al.).  
 
Therefore it makes no sense to wait until such studies are published (because this will 
probably never  happen),  and  one  has  to  weigh  the available  evidence pro  or  contra 
vaccination of an individual FSW from other studies or study populations. So the evidence 
on which one has to rely is only the best available evidence (this means: indirect evidence), 
not  the  best  (theoretically)  possible  evidence.  It  makes  no  sense  to  wait  until  the 
(theoretically) possible best (direct) evidence is available, while in the meantime many 
FSWs are exposed to HPV risks and disease burden which can be reduced (at least in part) 
by vaccination. 
 
 
 
 
 
 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-4- 
 
 
Oral/oropharyngeal risks 
 
However, there is an additional important aspect which is not mentioned in the SCHIM VAN  
DER LOEFF et al. article and which may influence the balancing of pros and cons in a lot 
of individual cases.  
 
It is well established meanwhile that unprotected oral sex may result in oral/oropharyngeal 
HPV contamination and infection. If this infection becomes persistent, it may result in 
oral/oropharyngeal cancer (abbr.: OOPC; mostly tonsils, base of the tongue) after many 
years or decades, and there is no secondary prevention/early detection method available for 
these cancers in the moment (contrary to cervical cancer). It is well established now (with 
several dozens of original and review papers), (i) that oral sex (fellatio and cunnilingus; the 
role of tongue kissing is still unclear and at most very small) is the dominant source of 
HRHPV/HPV16  infection  in  the  mouth  and  oropharynx,  (ii)  that  the  prevalence  of 
oral/oropharyngeal HRHPV is strongly correlated with the number of oral sex partners (more 
strongly with this number in the last 1 ± 2 years than the lifetime number due to the clearance 
of most of these infections), and (iii) that the risk of HPV-associated oral/oropharyngeal 
cancer (OOPC) correlates with the lifetime number of oral sex partners and the amount of 
lifetime oral sex experience. (Oral LRHPV correlates less strongly with oral sex practices; 
studies which look IRU³DOORUDO+39W\SHVWRJHWKHU´DQGZKLFKDUHQRWUHVWULFWHGWR+5+39
or HPV16 may thus underestimate the role of oral sex with regard to HPV infections of 
clinical significance, or may even fail to find results of statistical significance). For example, 
in unvaccinated men and women from the NHANES population (US) who had oral sex 
within the last 12 months, the oral prevalence of HPV16 and/or 18 was found to be 0.4 % in 
persons with 1 lifetime oral sex partner, 1.2 % in persons with 2 ± 24 lifetime oral sex 
partners, but 6.5 % in persons with more than 24 lifetime oral sex partners (GUPTA et al. 
2019; combined for men and women; ~ 57 % women). 
 
There is a strong and continuing rise in the incidence of HPV-associated OOPC since a few 
decades in North America and Europe. The most impressive data come from Scandinavia 
(especially Sweden) and US. 
 
However, because Cancer Registry GDWDGRQ¶WGLVWLQJXLVKEHWZHHQ+5+39-positive und 
HRHPV-negative OOPC at the same anatomical site, the rise in HPV-associated OOPC can 
be easily overlooked when working with Cancer Registry data, because at the same time, the 
incidence of HRHPV-negative OOPC cases, mainly (nearly all) caused by smoking, is 
decreasing, since smoking habits decrease since many years in the general population in 
developed countries. The combined effect (rise of HRHPV-associated OOPC, but decrease 
of smoking-induced cancers) may result in a stagnation of OOPC incidence, and so one can 
easily miss that there is a problem with HPV-associated OOPC. As Aimee Kreimer from the 
National Cancer Institute (NCI) (US) pointed out in some of her papers, the incidence of 
HPV-associated OOPCs (combined for men and women) will exceed the number of incident 
cervical cancers in the in the US very soon. (For review of the most recent knowledge in the 
field of epidemiology see SHEWALE and GILLISON 2019).  
 
HPV-associated OOPCs mostly affect the oropharynx and especially tonsils and the base of 
the tongue (i.e. the epithelium in context with lymphatic tissue); however, they may also 
occur at many different sites in the oral cavity, though the relative portion of HPV-associated  
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-5- 
 
 
cancers among all cancers of the oral cavity is smaller than it is in the oropharynx (where 
HPV is the dominant cause meanwhile), and the combination of smoking and heavy alcohol 
drinking is still the most important cause for cancers of the oral cavity.  
 
HRHPV and smoking(+ alcohol) utilize different molecularbiological pathways to induce 
cancer in the oral cavity and oropharynx. This means that persistent HRHPV infection in the 
oral  cavity/oropharynx is not  solely a carcinogenic cofactor (like alcohol) to  smoking; 
HRHPV, especially HPV 16, itself is a primary carcinogen in the mouth and oropharynx, 
and so even nonsmokers/nondrinkers may get this cancer type though it is very unlikely for 
them to get HRHPV-free cancers at these sites, especially in the oropharynx.   
 
In US it is an often cited observation, cited in many papers about HRHPV-associated oral 
infections and cancers in mouth and oropharynx, that these HPV-associated OOPCs are 
found not so rarely in younger, non-smoking, socially/financially better situated white men 
compared to smoking/alcohol-induced cancers which are more often found in older, socially 
more deprived, heavily smoking and drinking men. Women are affected too, but less often 
than men. 
 
Nevertheless, smoking is a cocarcinogen on the HRHPV-associated pathway because it 
reduces  the  natural  immunity  against  HPV  and  thus  delays  or  impedes  clearance  and 
promotes the progression towards cancer. Smoking also enhances viral shedding on the 
infected mucosal site and thus enhances infectiousness and the risk for sexual partners. So 
smoking is in fact an additional risk factor also for HPV-associated OOPCs, and smokers 
are found to have significantly higher oral HPV 16/18 prevalences (e.g., GUPTA et al. 2019), 
but not on its own (smoking-specific) pathway, but smoking strengthens the HPV-associated 
pathway to cancer. 
 
Both fellatio and cunnilingus may result in oral/oropharyngeal HPV contamination and 
infection. However, the pro-act risk of cunnilingus seems to be higher than that of fellatio 
because  of  a  higher  viral  load  on  the  female  genital  mucosa  and  in  genital secretions 
compared to the surface of the skin and mucosa of the penis. This may explain the higher 
prevalence of oral/oropharyngeal HRHPV infections in healthy men as well as the higher 
percentage of men among all patients with HRHPV-associated OOPCs compared to women. 
Furthermore, due to  genital  HPV contamination  or infection earlier in life, the natural 
immunity against oral/oropharyngeal infections seems to work better in women than in men, 
thus explaining in part the differences in the prevalence of oral/oropharyngeal HRHPV 
infections and OOPCs between men and women (contrary to genital contaminations or 
LQIHFWLRQVLQZRPHQSHQLOHLQIHFWLRQVGRQ¶WVHHPWRLQGXFHDFOLQLFDOO\UHOHYDQWGHJUHHRI
humoral or antibody-mediated natural immunity, and they rarely induce the production of 
antibodies. This may also explain why penile HRHPV prevalence shows a tendency to rise 
with increasing age, whereas cervical HRHPV prevalence is decreasing; e.g. BROUWER et 
al. 2019).  
 
However, even women are at risk of HPV-associated OOPCs, and smoking seems to impair 
the effectiveness of naturally acquired immunity from former genital contaminations or 
infections.  Thus  oral/oropharyngeal  HRHPV  and  OOPC  is  not  restricted  to  men  who 
practiced unprotected oral sex; it also affects women with a risky sexual lifestyle, but with  
 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-6- 
 
 
higher risks for smoking women due to impaired natural immunity (which nonsmoking 
women may gather as a consequence of their sexual lifestyle). 
 
Since unprotected oral sex is an important service in sex work, maybe the most important in 
some countries or sex work settings or for some (or many?) individual FSWs (though it is 
prohibited by law now in Germany as far as fellatio on men is concerned, but unprotected 
cunnilingus is not forbidden), the oral risks have also to be considered in the individual 
balance for the recommendation for (or against) HPV vaccination for an individual FSW (or 
MSW).  Contrary  to  cervical  cancer,  there  is  currently  no  secondary  prevention/early 
detection  program  available  for  HPV-associated  OOPC  or  precancer,  what  leaves 
vaccination as the only way of prevention (apart from consistently protected oral sex; see 
GUPTA et al. 2019).  
 
Though it is forbidden now in Germany, and the fine for the client may be (theoretically) as 
much as 50.000 Euro, unprotected fellatio is still practiced frequently. Some FSWs welcome 
this aspect of the law (Prostituiertenschutzgesetz) because it offers to them a good reason to 
decline  unprotected  fellatio.  However,  many  FSWs  have  plausible  reasons  why  they 
continue to practice unprotected fellatio, and why some of them really like to do so. These 
reasons are discussed in a German forum-based paper DQGFDOOHG³WKH RUDOVH[GLOHPPD´ 
 
)XUWKHUPRUH LQ VHWWLQJV ZKHUH )6:V GRQ¶W ZRUN DORQH PDQ\ )6:V OLNH WR SHUIRUP
threesomes and/or lesbian games/actions with colleagues. Due to the high prevalence of 
HRHPV on the genital mucosa of unvaccinated FSWs, and the higher viral shedding on 
female genital mucosa compared to the penis of infected men (and even more in smokers), 
this  practice  (cunnilingus  on  colleagues)  results  in  a  higher per-act  risk  of  acquiring 
oral/oropharyngeal HRHPV compared to fellatio on male clients. 
 
As a consequence, FSWs who practice unprotected oral sex with clients (and even more, 
when  they  practice  cunnilingus  on  colleagues)  are  without  doubt  highly  exposed to 
oral/oropharyngeal HRHPV contaminations and they are one of the subpopulations with the 
highest risk for incident and persistent oral/oropharyngeal HRHPV infections, and, as a 
consequence, HPV-associated OOPC. Among all women, these FSWs are probably the 
group with the highest risk at all, especially if they are smokers.   
  
Fortunately nearly all HRHPV-associated OOPCs are caused by HPV 16, with a small 
contribution of HPV 18; the effect of all other HPV types together on OOPC is negligible, 
at  least  in Europe  and  North  America.  Thus  even  the  bivalent  vaccine  covers  nearly 
completely the spectrum of HRHPV types which are relevant in the mouth and oropharynx 
(much more specifically than at the cervix), though the nonavalent vaccine offers a small 
advantage since warts of the anogenital type may also occur in the mouth (though rarely), 
and a few cases of OOPC seem to be associated with HPV 6.  
 
There are no studies which prove that (former) FSWs have a higher risk of HPV-associated 
OOPC. Such studies would be unpractical due to the long time (1 ± >3 decades) between 
HPV  infection  and  symptomatic  OOPC,  so  most  affected  women  would  have  stopped 
working in the sex industry a long time before cancer diagnosis. And many former FSWs 
repress their former involvement in sex work psychologically after they stopped it, and 
therefore ZRQ¶WGLVFORVHLWPDQ\\HDUVRUGHFDGHVODWHUZKHQFDQFHULVGLDJQRVHG7KXVZH 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-7- 
 
 
have only indirect evidence that FSWs who practice unprotected oral sex will have a higher 
risk of HPV-associated OOPC many years later (though HPV-16-E6-antibodies might be a 
proxy for that risk, because they occur many years before cancer diagnosis, but probably 
more reliably in non-smokers than in smokers), but this indirect evidence is strong due to 
the high number of studies and papers which demonstrated the association between oral sex 
experience (e.g., number of oral sex partners), oral HPV 16 infection, and HPV-associated 
OOPC.     
 
Though FSWs who practice unprotected oral sex are heavily exposed to risks of oral HPV 
contamination, their oral/oropharyngeal prevalence of HPV 16 is much smaller than at the 
cervix or anus (as far as one knows in the moment). There are only a few studies with respect 
to oral HPV 16 prevalence in FSWs (and no studies concerning persistent oral infections in 
FSWs), and it was found to be about 2 % with the maximum in a small study from Hungary 
with 5.9 % HPV 16.  
 
 
 
Oral HPV 16 prevalence in FSWs: 
 
Europe: 
CANADAS et al. 2004 (Spain): 2.1 % (N = 188) 
MAREK 2014 (Hungary): 5.9 % (small study; N = 34; HPV 16 prevalence data  
    not published/personal communication; control group: 0 % HPV 16) 
 
Outside Europe: 
MATSUSHITA et al. 2011 (Japan): 0 % HPV 16, but 1.0 % HPV 18 and 5.1 % 
    HRHPV altogether (note: the relative portion of carcinogenic HPV types differs in 
    East Asia from Europe and North America) (N = 196) 
BROWN et al. 2011 (Peru): 0.54 % (N = 184) (note: oral sex 95 % protected by  
    condoms) 
CHATTERJEE et al. 2001 (Calcutta, India): 29.0 % HPV 16 and/or 18 (N = 69);  
     however, there are some doubts about the adequacy of the methods of this study, so   
     this study should not be considered as very relevant (too many false-positives or simple  
     contaminations?) 
 
  
Nevertheless, estimating 2 ± 6 % as the relevant oral HPV 16 prevalence in FSWs in Europe,  
this prevalence is up to seven- to twentyfold compared to women of the general population, 
where it was found to be 0.3 % in the very large NHANES population IURP86'¶628=$
et al. 2014; GILLISON et al. 2012;  SONAWANE et al. 2017).  
 
So,  as  expected,  FSWs have  indeed  a  much  higher  risk  of  oral/oropharyngeal  HPV16 
infection compared to other women. In most studies on FSWs it is also not clear how much 
unprotected oral sex was involved at all. In Peru, where oral sex was performed protected 
almost every time, oral HPV 16 prevalence was not much higher than in NHANES (0.54 vs. 
0.3 %). Even in the Hungarian Study, many FSWs said they use condoms for oral sex. So 
there is at least some possibility that oral HPV 16 prevalence might be even higher in FSW 
populations who are used to practice oral sex without protection regularly. However, from 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-8- 
 
 
 WKHIHZDYDLODEOHGDWDH[FHSWWKHYHU\SUREOHPDWLF³ROG´VWXG\E\&HATTERJEE et al.) it 
is evident that oral HPV 16 prevalence is (much?) smaller in FSWs than their cervical or 
anal prevalence, and this is also in accordance with oral vs. cervical prevalences in the 
general population. The same applies to oral HPV 16 prevalence in MSM compared to anal 
HPV 16. So there are no reasons to assume that oral HPV 16 prevalence in FSWs might be 
as high or even higher than their cervical HPV 16 prevalence. As indicated by data from 
people from the NHANES population with 25 and more lifetime oral sex partners (see 
BROUWER et al. 2019), the oral HPV 16/18 prevalence seems to level out in cases of very 
high  risk  exposition; i.e.  in  highly  exposed groups,  oral  HPV  16/18  prevalence  rises 
disproportionately if the (already high) risk increases further. There is probably some kind 
RI ³VDWXUDWLRQ´ DQG D PD[LPDO SUHYDOHQFH OHYHO +RZHYHU GXH WR WKH ODFN RI HQRXJK
epidemiological data for high risk (= highly exposed) persons, and also some heterogeneity 
in the methods of oral/oropharyngeal sampling and HPV detection, it is so far impossible to 
quantify this maximal risk level, and it may be different depending on sex, age, smoking or 
HIV status, since all of them influence natural immunity against oral/oropharyngeal HPV.  
 
Anyway, with regard to vaccination it is important to note that oral prevalence is (much?) 
lower than the genital prevalence in women of the general population and also in FSWs, and 
this may explain why the effectiveness of HPV vaccination (mostly Gardasil) in NHANES 
was shown to be stronger with regard to oral infections (HPV 16/18) (risk reduction: 90 ± 
100 %) than for cervicovaginal infections (risk reduction: 80 %) (age group: 18 ± 24 years; 
NHANES; BROUWER et al. 2019).      
 
These differences between cervical and oral/oropharyngeal prevalence are also expected for 
KLVWRORJLFDOUHDVRQV+39FRQWDPLQDWLRQGRHVQ¶WUHVXOWLQHYLWDEO\LQWUXH+39LQIHFWLRQ
HPV needs micro-cracks or microabrasions of the mucosal epithelium to get access to the 
basal cell layer of the epithelium. Whereas the cervical epithelium is (in part) only one-
layered,  and  the  anal  epithelium,  though  thicker,  often  hurt  by  micro-cracks  or 
microabrasions (if not from anal sex, then from defecation or cleaning procedures, or some 
grade of inflammation as a consequence of these micro-traumata or eczemas), the epithelium 
in the mouth and pharynx is thicker and much more resistant, except for some problematic 
areas in the crypts of the tonsils (which are therefore much more frequently affected by HPV 
16 and HPV-associated cancer than other sites in the mouth and pharynx). In summary, it 
has  to  be  expected  that  the  oral/oropharyngeal  HPV  16  prevalence  in  the  mouth  and 
oropharynx  is  smaller  compared  to  the  cervical  or  anal  site  even  in  heavily  exposed 
individuals  like  FSWs. Oral  contamination  with  HPV  16  results  more rarely  in  a  true 
infection compared to cervical or anal contamination.   
 
This is an important aspect with regard to vaccination: all available vaccines cover the most 
relevant HPV types for OOPC in Europe and North America (to at least 95 %), and since the 
oral/oropharyngeal HPV 16 prevalence in FSWs seems to be (much?) smaller than 10 %, 
WKH³WRR-late-DUJXPHQW´ZKLFKPD\DSSO\WRDVXEVWDQWLDOSRUWLRQRI)6:VZLWKUHJDUGWR
cervical and/or anal infection, does therefore not apply so strongly to the prevention of oral 
infections by the vaccine. And, most important in this context, a detailed analysis of the 
Costa Rica Vaccine Trial showed that a prevalent infection at one mucosal site (e.g., cervical 
RUDQDODWWKHWLPHRIYDFFLQDWLRQGRHVQ¶WLPSDLUWKHHIIHFWLYHQHVVRIWKHYDFFLQDWLRQDJDLQVW
new infections at  other  mucosal  sites (e.g. oral) which were uninfected at  the time of 
vaccination (BEACHLER et al. 2016).  
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-9- 
 
 
The Costa Rica Vaccine Trial (CVT) (HERRERO et al. 2013) proved that the bivalent 
vaccine protected young women (18 ± 26 years at the time of the first dose of the vaccine) 
against prevalent oral HPV16/18 infection four years later (risk reduction 93.3 % for HPV 
16+18), and the ACTG A5298 trial cited in SCHIM VAN DER LOEFF et al. demonstrated 
the effectiveness of Gardasil 4 against persistent (!) oral infections of the four vaccine types 
even  in  older  and  immunocompromised  (HIV-infected)  individuals,  mostly  men  (point 
estimate: 88 % risk reduction of persistent oral infections). Due to small numbers of oral 
infections, the confidence interval around this point estimate was large; however, in spite of 
the small numbers of persistent oral infections statistical significance was reached. It is also 
LPSRUWDQWWKDWWKLV³RUDOVXFFHVV´ZDVGLVFRYHUHGLQWKHVDPHVWXG\ZKLFKGHPRQVWUDWHGWKH
inefficacy (or low efficacy) of the vaccine against persistent anal infections in this heavily 
pre-exposed population. Thus the study design was able to demonstrate this fundamental 
difference between the effects of vaccination with regard to the anal versus oral site in this 
very special population. This is a second proof for the result from the CVT (BEACHLER et 
DOWKDWDQLQIHFWLRQDWRQHPXFRVDOVLWHGRHVQ¶WLPSDLUWKHHIIHFWLYHQHVVRIWKHYDFFLQH
at other currently uninfected mucosal sites.  
 
Also several papers based on the NHANES population in US demonstrated a high efficacy 
(about 80 ± 100 %) of HPV vaccination (usually Gardasil 4) against prevalent oral infections 
in men and women, but the age at the time of vaccination is not so clear in this population 
(HIRTH et al. 2017; SONAWANE et al. 2017; CHATURVEDI et al. 2018; BROUWER et 
DO+RZHYHUDYDLODEOHGDWDVXJJHVWWKDWPRVWSDUWLFLSDQWVJRWWKHYDFFLQH³WRRODWH´
i.e.  18  years  or  older  (mean  age  for  vaccination  was  18.5  years  in  the  data  from 
CHATURVEDI et al.), especially if one respects that sexual life in US often starts with oral 
sex (see GUPTA et al. 2019 for discussion).  
 
And a cross-sectional study from UK (MEHANNA et al. 2018) showed a 91 % risk reduction 
of oral/oropharyngeal HPV 16 infections in vaccinated girls and young women (N = 243; 12 
± 24 years, median: 18.6 years) compared to unvaccinated (0.5 % vs. 5.6 %; p = 0.04). Most 
or all of the vaccinated girls had received Gardasil (the vaccine applied by the UK NHS for 
the vaccine program of school girls). All participants of this study got tonsillectomy for non-
malignant reasons. Oral samples for HPV testing were collected by oral rinse, tongue base 
and  pharyngeal  wall  brushes,  and  by  examination  of  the  tonsil  tissue  following 
tonsillectomy.  This  extensive  sampling,  much  more  extensive  than  in  any  other  study 
concerning oral HPV prevalence, may contribute to the unusually high HPV 16 prevalence 
of 5.6 % in the unvaccinated girls and young women among the tonsillectomy patients (see 
below).  
 
This  study  not  only  demonstrates  the  high  efficacy  of  the  vaccine  to  reduce 
oral/oropharyngeal HPV 16 prevalence, but it also showed signs of herd immunity. All 69 
boys in this study were free of oral/oropharyngeal HPV 16, though no boy had gotten the 
vaccine (HPV 16: unvaccinated boys: 0.0 %; unvaccinated girls: 5,6 %). 78 % of the girls 
were vaccinated, which may explain the protection of the boys and young men.  
 
The high oral prevalence of HPV16 in unvaccinated girls and young women from the general 
population  who  got  tonsillectomiced  in  the  MEHANNA  study  contrasts  to  all  known 
epidemiological data for women from the general population (e.g., the very large NHANES 
data set). However, besides the most extensive oral/oropharyngeal sampling ever done,  
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-10- 
 
 
young people with hypertrophic or chronically infected tonsils may be much more prone to 
oropharyngeal  HPV  infections  (and  their  persistence)  than  those  with  healthy  tonsils, 
because in these cases HPV particles have easy access to the basal layers of the epithelium 
(³entry port´). The high prevalence may thus be a consequence of longer lasting tonsil 
pathology which may have made the tonsils more susceptible to HPV infection in the case 
of HPV contamination. In fact, it was already shown that people who were tonsillectomiced 
in former times have a reduced risk of oral/oropharyngeal HPV infection (BEACHLER et 
al. 2015). So it is not unexpected that people with chronically infected and/or hypertrophic 
tonsils may have a higher risk for HPV infection in the oropharyngeal area, or persistence 
of HPV at that site. Chronic inflammation may impair clearance and promote persistence, 
what may also explain the high prevalence in tonsillectomy patients.  
 
Nevertheless, since all participants of that study suffered from tonsillar disease which gave 
UHDVRQIRUWRQVLOOHFWRP\WKHVHFRQVLGHUDWLRQVGRQ¶WDIIHFWWKHTXDOLW\RIHYLGHQFHDVIDUDV
the effectiveness of the vaccine is concerned; they only explain while people with chronic 
tonsillitis or tonsillar hypertrophy may be another high risk group for oropharyngeal HPV 
(16). If one acknowledges this aspect, one can consider the MEHANNA study as another 
³ULVNJURXSVWXG\´RQRUDOYDFFLQHHIILFDF\RI*DUGDVLOOLNHWKH$WULDOZKLFKZDV
also a risk group study, but for very different reasons).   
 
Nevertheless,  the ACTG A5298 trial  is  the most  important  trial in  the context  of oral 
infections in risk groups so far because it is the only trial which looked after persistent oral 
infections (only they matter with regard to cancer risk) and since it was carried out in a 
heavily exposed and pre-exposed population. Because HIV infection increases the risks of 
incident and prevalent oral HPV infections, the A5298 study population may tentatively 
reflect the oral risk level of FSWs. In fact, some studies on HIV infected populations (mostly 
MSM) demonstrated oral HPV 16 prevalences close to the 5.9 % in the small Hungarian 
FSW study or the 6.5 % HPV16/18 in NHANES men and women who had 25 or more oral 
sex partners in their lifetime (GUPTA et al. 2019). So there are good reasons to assume a 
close proximity of the A5298 trial to the oral HPV risk situation of FSWs who practice 
unprotected oral sex with their clients (and possibly colleagues), and these favorable A5298 
data (Gardasil) are supplemented by the 93.3 % efficacy of Cervarix against prevalent oral 
HPV 16/18 infections four years after the first vaccine dose in women aged 18-26, i.e. in an 
age group when many FSWs start their sex work (in accordance with the own experiences 
of the authors in the first chapter of their article).   
 
So there is already some evidence in favor of HPV vaccination for FSWs who practice 
unprotected oral sex with their clients (and colleagues). The efficacy of the vaccine against 
HPV-associated OOPC in (former) FSWs will never be demonstrated in trials due to the 
methodological limitations mentioned above. In the absence of cytological precursors like 
CIN or AIN in the oropharynx and due to the long latency until OOPC develops, there is a 
general consensus that the prevention of persistent infections is an acceptable endpoint in 
HPV vaccination studies ZLWKUHJDUGWRWKHRUDOHIIHFWLYHQHVVRIWKHYDFFLQH³7KH3ULPDU\
�(�Q�G�S�R�L�Q�W�V��I�R�U��3�U�R�S�K�\�O�D�F�W�L�F��+�3�9��9�D�F�F�L�Q�H��7�U�L�D�O��&�R�P�P�L�W�W�H�H�����,�$�5�&��8��6���1�D�W�L�R�Q�D�O��&�D�Q�F�H�U�
Institute) (LOWY et al. 2015). The A 5298 (WILKIN et al. 2018) trial is so far the only trial 
in accordance with this recommendation. 
 
 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-11- 
 
 
Taken this into account, we already have some good evidence to offer HPV vaccination to 
FSWs who like to practice unprotected oral sex, and it is not probable that we will get better 
evidence  in  the  near  future.  Of  course  more  studies  concerning  the  oral  HPV  16 
epidemiology in FSWs are urgently needed. For example, it would be interesting to see the 
correlation between oral HPV 16 prevalence and the frequency of unprotected oral sex in 
the last 1 or 2 years, at best distinguishing between fellatio and cunnilingus (on colleagues). 
Such a study would be feasible with one-time-examination (mouth wash/gurgle sample and 
questionnaire)  and  thus  not  methodologically  impossible.  However,  since  unprotected 
fellatio is forbidden in Germany, such studies cannot be performed in Germany since those 
FSWs who continue to offer unprotected oral sex to their clients in spite of the law would 
not admit this in the questionnaire due to the legal restrictions, making any study of this kind 
useless or at least prone to extreme bias. However, in spite of the small number of available 
studies, we already have better indirect evidence for the protective effect of HPV vaccination 
of  FSWs  against  (persistent)  oral  infections  than  against  cervical  or  anal  infections  in 
potentially/probably heavily pre-exposed FSWs.         
 
 
Oral HPV risks for clients 
 
There is a second important aspect in the context of oral sex. Many clients like to practice 
unprotected cunnilingus on FSWs (dental dams are used extremely rare, about 1 %). Some 
)6:VGRQ¶WDOORZFXQQLOLQJXVWKH\PD\IHDUPLFUR-traumata on the genital mucosa which 
may facilitate STI acquisition, or they tend to suffer from candidiasis after unprotected 
cunnilingus. However, many FSWs like cunnilingus, either for their own pleasure or to relax 
DQGWREHDEOHWRVWD\DZKLOH³SDVVLYH´GXULQJ³SDLGWLPH´,HVWLPDWHWKDWDERXW�� 50 % 
of all paid contacts with FSWs in Germany involve some cunnilingus. These estimates are 
based on two independent methods in a clients¶ forum in Germany: (i) a questionnaire (35 ± 
45 % cunnilingus per contact; n = 578 participants) and (ii) the detailed analysis of forum 
reports from FOLHQWVDERXWVLQJOHFRQWDFWVZLWKD)6:VRFDOOHG³IXFN-UHSRUWV´!
cunnilingus per contact; n = 184 reports). The latter may be biased towards longer or more 
complex meetings with FSWs ZKLFKDUHPRUH³ZRUWK´WREHUHported in the forum) what 
may explain the difference between the results of the questionnaire and the analysis of the 
reports. However, only 15 % of the participants of the questionnaire said that they never lick 
on FSWs. Dental dams were used by about 1 % of the clients (n = 578). 
 
Due to the high genital prevalence of HRHPV in unvaccinated FSWs (and the usually high 
concordance between cervical und vulvar infections), these clients have a high risk of oral 
HRHPV contamination and infection. As mentioned above, the viral load on female genital 
mucosa is assumed to be higher than on the surface of the penis; moreover, many FSWs are 
heavy smokers and smoking is known to increase viral shedding at any infected mucosal 
site.  Men  also  miss  a  possible  naturally  generated  immunity  as  a  result  of  genital 
contamination or infection which may generate some humoral or antibody-based protection 
also at the oral site in women (see '¶628=$HWDO$PRQJDOO+39-associated 
cancers  in  men,  oropharyngeal  cancer  is  the most  prevalent  (RKI  2018a),  much  more 
prevalent than anal or penile cancer. 
 
 
 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-12- 
 
 
And it is also well established that husbands of women who had higher-grade CIN or cervical 
cancer in their life history, have a higher risk of OOPC (e.g. CHANCELLOR et al 2017; 
HEMMINKI K et al. 2000).  
 
Thus, men who like to practice unprotected cunnilingus on unvaccinated FSWs belong to 
the top risk groups for prevalent and persistent oral/oropharyngeal HPV 16/18 infections, 
and, as a consequence, HPV-associated OOPC. Depending on the frequency of this practice, 
they may even surpass HIV-infected MSM with regard to their risk of HPV-associated 
OOPC. 
 
From a public health view it might be desirable to recommend HPV vaccination to FSWs 
who like to be licked by their clients, as well as for their own protection (clients with HPV 
in saliva might infect the genital mucosa of FSWs, though this way seems to be less effective 
than the opposite direction from vulvar mucosa to mouth and pharynx, due to higher viral 
load there compared to saliva) as for the protection of their clients. 
 
Since (i) recommendation of HPV vaccination to FSWs (or at least subgroups of FSWs) is 
not established so far, and since (ii) the probability, that any given FSW may have received 
the vaccine in the past is very small (I agree fully to what SCHIM VAN DER LOEFF wrote 
in their paper about that subject, and this is also corroborated by own experiences), I go so 
far  to  recommend  HPV  vaccination  to  those  clients  who  like  to  practice  unprotected 
cunnilingus on FSWs. The immunogenicity of the HPV vaccine is already proved for adult 
men, even for Gardasil which results in lower antibody titres than Cervarix (for Gardasil 4: 
PINTO et al. 2016); the same applies to the A5298 trial in an age group that was on average 
older (median: 47 years) than that of the PINTO study. However, in the male PINTO cohort 
(Gardasil 4), by months 18 and 30, oral antibodies were only detectable in some participants 
(HPV-16, 39.8% and 29.6%; HPV-18, 10.7% and 4.6% of individuals) (PARKER et al. 
7KLVGRHVQ¶WPHDQWKDWRUDOSURWHFWLRQYDQLVKHGGXULQJWKLVWLPH,Iprotection against 
prevalent oral infection vanishes so quickly, the epidemiological data which support a high 
efficacy of Gardasil alone (NHANES men) or mixed use of Gardasil and Cervarix (but 
mostly dominated by Gardasil) like NHANES (women) or the MEHANNA tonsillectomy 
study (Gardasil 4) would be impossible (the UK National Health Service program for HPV 
vaccination  of  girls  used  Gardasil).  In  the  absence  of  detectable  oral  antibodies,  oral 
contamination might booster antibody production before a productive, stable infection is 
established in vaccinated individuals.     
 
Before the results of the A5298 trial were available, it seemed obvious to recommend 
Cervarix to clients of FSWs if the primary purpose of vaccination was oral sex. It is well-
known that Cervarix generates higher antibody titres (measured as geometric median) than 
Gardasil, and it was also shown that the antibody titre in the oral fluid is 2 ± 3 orders of 
magnitude smaller than in blood serum (PINTO et al. 2016). Though antibodies against HPV 
16 were present in the oral fluid of 96 % of adult men after Gardasil(4) vaccination (PINTO 
et al. 2016), the titres were so low that one might question whether they are fully protective 
WKH3,172VWXG\GLGQ¶WORRNIRURUDOLQIHFWLRQVLWZDVDERXWDQtibody titres in serum and 
oral fluid).  
 
With regard to Cervarix, we already know about its protective effect against prevalent oral 
infections in young adults from the CVT since the end of 2012 (HERRERO et al. 2013).  
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-13- 
 
 
Before A5298, NHANES-based studies and MEHANNA et al., evidence for efficacy against 
oral  infections  was  better  for  Cervarix  than  for  Gardasil,  and  the  low  antibody  titres 
following Gardasil vaccination caused some doubts about its oral/oropharyngeal efficacy.   
 
This gave reason to recommend Cervarix (and safe some money) to men who like to lick on 
FSWs (as main reason for vaccination), because higher antibody titres in the oral fluid can 
be expected after Cervarix vaccination. It was found in the Gardasil trial with adult men 
(PINTO et al. 2016, PARKER et al. 2019), that, on an individual base, the titres of antibodies 
in oral fluid and blood serum are closely correlated to one another (as expected).  
 
However, with the results of the A5298 trial and also some NHANES data (in US, nearly all 
vaccinated people got Gardasil, and as long as Cervarix was available there, it was not 
approved for men, so 100 % of vaccinated men should have gotten Gardasil), Gardasil may 
also be recommended to men who like to practice cunnilingus on FSWs. Because of the 
limited protection of condoms against HPV, the protective effect of Gardasil against genital 
warts (if not currently infected) may also be interesting for clients. There was a report from 
a FSW in a forum who was interested in anogenital warts and counted the clients with such 
lesions. She reported anogenital warts on every seventh client; however, no one knows 
whether  her  diagnosis  was  correct  or  whether  she  misclassified  some  other  lesions  as 
anogenital warts. Nevertheless, this report shows that anogenital warts are in fact a problem 
for clients of FSWs.  
 
Meanwhile, the role of the HPV vaccine for the prevention of HPV-associated oral and 
oropharyngeal  cancer  is  regarded  as  so  important  and  conclusive  that  there  are 
recommendations  in  the  US  that  also  dentists  should  promote  and  offer  vaccination 
(American Dental Association  2012 and 2018; National  HPV Vaccination Roundtable) 
(KEPKA D et al. 2019).  
 
 
 
Herd immunity 
 
As mentioned in the article from SCHIM VAN DER LOEFF et al, there are also aspects of 
herd immunity. But not only for clients and other women (partners of the clients) and the 
general population, but also for FSWs themselves. The more FSWs are vaccinated, the 
smaller the risk for the remaining unvaccinated FSWs or \RXQJEHJLQQHUVZKRGLGQ¶WKDG
already the chance for vaccination, due to lack of consultation/recommendation or money.  
 
It is well established that herd immunity works well with regard to HPV vaccination. Though 
the best example for that is the nearly complete eradication of genital warts in the young 
generation in Australia (though only girls had been vaccinated, boys/young men had an 
enormous profit from that), there are also examples from other countries with comparatively 
high vaccination rates in schoolgirls (e.g. Sweden, UK; for systematic review and meta-
analysis see DROLET et al. 2019).  
 
,QDVH[XDOQHWZRUNWKRVHZKRGLGQ¶WJHWWKHYDFFLQHSURILWIURPYDFFLQDWLRQRIRWKHUV7KLV
�G�R�H�V�Q���W��Z�R�U�N��R�Q�O�\��L�Q��Y�H�U�\��W�L�J�K�W��V�H�[�X�D�O��Q�H�W�Z�R�U�N�V���E�X�W��W�R��D��O�R�Z�H�U��H�[�Went and with some 
temporal delay even in very loose and weak networks. So, for example, a reduction in the  
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-14- 
 
 
prevalence  of  genital  warts  in  young  homosexual  men  in  Australia  was  found  as  a 
consequence of the vaccination of most of the schoolgirls (FAIRLEY et al. 2012).  
 
FSWs from the same sex work setting (e.g. brothel, club), but even in the same local 
(geographical) area are interconnected with  one another in  a more or less  tight  sexual 
network not only via direct interaction (lesbian games, threesomes) inside their sex work 
setting, but also (and to a larger extent) via their clients, who often contact several FSWs in 
the same setting or local area, if not at the same day, then sometime in the past or future. 
This results in a tight sexual network within the same setting and a less tight network in the 
local area between FSWs and clients, with private partners of the clients and of the FSWs in 
the periphery of that network. Infected private male partners of FSWs may serve as a 
reservoir for HPV reinfection or HPV reloading of FSWs (ping-pong effect), especially if 
WKH\GRQ¶WXVHFRQGRPVDQGWKXVLPSDLUFOHDUDQFHRISUHYDOHQW+39LQIHFWLRQVZKLFKWKHQ
may become persistent) or regression of CIN lesions. The lower the HPV burden in this local 
(setting-specific or regional) sexual network, the lower the future HPV risks for those FSWs 
ZKRGLGQ¶WJHWYDFFLQDWLRQ 
 
 
 
Conclusions 
 
In summary, I agree with 6&+,09$1'(5/2())¶V discussion that the recommendation 
for (or against) HPV vaccination must be balanced for each individual FSW, and vaccination 
should only be recommended when this balance favors vaccination. Unprotected oral sex, 
active  (on  clients  and  colleagues)  and/or  passive  (cunnilingus  by  clients)  should  be 
considered in favor of vaccination. In contrast to cervical cancer, there is currently no 
secondary prevention/early detection program  available against HPV-associated OOPC, 
OHDYLQJYDFFLQDWLRQDVWKHRQO\ZD\RISUHYHQWLRQLIRQHGRHVQ¶WDFFHSWWRSUDFWLFHDQ\IRUP
of oral sex consequently with barrier protection (see GUPTA et al. 2019) (though even in 
the case of consequently protected oral sex some residual risks remain, for example (i) 
licking at the scrotum which may also be infected by HPV ± as proven by genital warts on 
the scrotum; (ii) self-inoculation from the genital/vulvar to the oral site in case that the 
condom is not changed when fellatio is practiced following vaginal intercourse, e.g. to 
improve  or  regain  erection; (iii) erroneously  turning  around  the  dental  dam  or  other 
removable barriers during cunnilingus so that the contaminated side of the barrier gets in 
close contact to the tongue).  
 
The bandwidth how much an individual FSW may profit from vaccination, is large: On one 
side there are very young, maybe 18 yrs. old beginners in sex work with no or little sexual 
experience before, as SCHIM VAN DER LOEFF et al. already discussed in their paper (they 
really exist, e.g. from rural areas and traditional families in Romania or Bulgaria). Because 
of the high HPV risks of sex work (and the limited protective effect of condoms, estimated  
to be only 50 ± 70 %), those FSWs will probably profit much more from HPV vaccination 
WKDQDQHDUO\\UVYDFFLQDWHGJLUOIURPWKHJHQHUDOSRSXODWLRQZKRZRQ¶WHQJDJHLQ
sex work later in life. Some studies found up to 10-fold risks of cervical high-risk-HPV 
prevalence in FSWs compared to age-matched controls (e.g. TOUZE et al. 2001), and a (not 
peer-reviewed) review about the relative risks of cervical pathologies in FSWs found a four- 
to fivefold risk of higher-grade dysplasia like CIN2+ or HSIL in FSWs compared to women  
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-15- 
 
 
of the general population (own results in a forum-based paper; includes 8 studies with 6965 
FSWs and even more controls).   
 
Thus, a young beginner in sex work with no or little sexual experience in the past is expected 
to profit much more from vaccination than girls from the general population, though such 
cases may be rare. In these cases, vaccination of the beginner may prevent much more HPV-
DVVRFLDWHGGLVHDVHEXUGHQWKDQWKH³HDUO\´\UVYDFFLQDWLRQRID\RXQJJLUORIWKH
general population with average sexual behavior/risks in later life can do.    
 
However, in countries where SW have to be counselled before starting sex work (like 
Germany), it is (at least theoretically) feasible to recommend HPV vaccination (and to give 
the first dose of the vaccine) before sex work is started. Since antibody titres rise quickly 
after that dose, protection against HPV will start during the second week after injection. 
 
On the other extreme boundary of that bandwidth are experienced FSWs in their late 30ths, 
40ths or 50ths with a long history of sex work. If they participated in cervical/genital cancer 
prevention/early detection programs on a regular base and never had any HPV-associated 
pathology,  they  have  obviously  no  problems  to  avoid  or  clear  incident  genital  HPV 
infections and have an effective natural immunity against HPV. And maybe they offer only 
protected oral sex and/or want to stop sex work not so long time in the future. Their profit 
IURPYDFFLQDWLRQLVSUREDEO\YHU\VPDOORUQHDUO\]HURDQGGRHVQ¶WRXWZHLJKWKHFRVWVDQG
inconveniences and possible (local/systemic) side effects of vaccination, whoever pays for 
it, even if is for free. Also the HPV risks for their clients are expected to be much smaller 
(compared to young FSWs), though they still exist. Thus vaccination of such FSWs would 
have only smaller effects on client protection and herd immunity.  
        
These two examples show the extreme boundaries of the large bandwidth. Most FSWs will 
have to be categorized somewhere between these extremes; as SCHIM VAN DER LOEFF 
et al. wrote: ³QRWDOO6:DUHWKHVDPH´  
 
There are many criteria which may have to be taken into account to estimate whether an 
individual FSW will profit more or less from vaccination, and whether vaccination should 
be recommended to this individual FSW, or whether the expected profit is probably too small 
if one compares it with costs, inconveniences, side effects and risks? Such criteria may be: 
 
Ɣ�F�D�O�H�Q�G�D�U��D�J�H�����Y�H�U�\��\�R�X�Q�J��� 
Ɣsexual experience in the past (lifetime and especially during the last 1 ± 2 years, the usual  
      clearing time of HPV infections) 
Ɣ�W�L�P�H��L�Q��V�H�[��Z�R�U�N��K�L�V�W�R�U�\��R�I��V�H�[��Z�R�U�N��Q�X�P�E�H�U��R�I����Y�D�J�L�Q�D�O����F�O�L�H�Q�W�V��L�Q��W�K�H��S�D�V�W� 
Ɣplans for the future: how long does she want to continue sex work? 
Ɣresults of former cervical cancer screenings, history of HPV infections or HPV-associated  
      pathologies (also as a proxy for natural immunocompetence against HPV)   
Ɣoffered sexual services and non-genital (for example oral) HPV risks?  
      (e.g., unprotected oral sex;  
       unprotected fellatio on clients,  
       unprotected lesbian action with other FSWs;  
       unprotected cunnilingus from clients) 
 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-16- 
 
 
Ɣsmoking (reduces the chance of spontaneous clearance of HPV infections at any mucosal  
       site, enhances viral shedding and infectiousness) 
Ɣcombination of smoking and unprotected oral sex 
 
if available: 
Ɣcurrent cervical HPV status (if HPV 16 is present without pathology, it is an 
      open question whether it is recommendable to delay vaccination until this infection 
      cleared. This rises also the question whether cervical HPV testing before vaccination 
      should be recommended to active/heavily pre-exposed FSWs?)   (see below) 
Ɣcurrent HPV-associated pathology (according to most trials, HPV vaccination reduces the  
      risk of recurrence after treatment/removal of CIN2+, especially if administered as  
      adjuvant to surgical treatment) 
 
 
Based on these (and maybe some additional) criteria, it might be theoretically possible to 
develop a score to grade the individual strength of recommendation for or against HPV 
vaccination for an individual FSW. However, there are not enough epidemiological data to 
weigh out each criterion against the other ones, and to validate and calibrate the score. 
Furthermore, there are secondary criteria like access or participation in cervical cancer 
VFUHHQLQJWKRVHZKRGRQ¶WSDUWL�F�L�S�D�W�H��P�D�\��S�U�R�I�L�W��P�R�U�H��I�U�R�P��Y�D�F�F�L�Q�D�W�L�R�Q��E�H�F�D�X�V�H��W�K�H�\��G�R�Q���W�
have the chance of early detection) or psychological aspects. A FSW may feel better during 
work when she knows that she got the vaccine and is protected against new infections, 
including genital warts which are not life-threatening, but very unpleasant for FSWs. She 
did the best  she  could  do  against  HPV-associated  diseases  (besides  cervical  cancer 
screening), and this may relax psychologically, independent of how much she can profit 
from vaccination according to objective calculations. Thus vaccination may contribute to 
psychological health of FSWs who worry about infection risks, and psychological health is 
an important and often debated matter for FSWs.  
 
So, in summary, I agree with SCHIM VAN DER /2())¶V paper that the primary question 
is not whether to recommend HPV vaccination either to all FSWs or to none of them, but to 
recommend HPV vaccination to those FSWs (and beginners of SW) who will profit most 
from it, or who want it on their own (maybe for psychological relief because they want to 
do the best against HPV what is possible and to feel protected, as a contribution to their 
psychic health). 
 
And of course the same applies to MSW or transgender SW (though some criteria must be 
adapted), but also to some male clients of FSWs as already described above. 
 
 
Costs of vaccination 
 
A main obstacle are the high costs of vaccination; those FSWs who might profit the most 
(like the young beginners) might have the strongest (financial) limitations to get the vaccine 
because they have no money ± RWKHUZLVHPRVWRIWKHPZRXOGQ¶WVWDUWVH[ZRUNDWDOO,WLV
well established meanwhile that most FSWs would accept the vaccine and wish to get it, if 
it is for free, are at least much cheaper than it actually is (e.g. Peru: BROWN et al. 2012; 
BROWN et al. 2013; Netherlands: MARRA et al. 2017; China: HONG et al. 2013;  
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-17- 
 
 
Cambodia:  WADHERA  et  al.  2015;  Hungary:  MAREK  et  al.  2014;  more  indirectly: 
Antwerp: PELEMAN 2015). 
 
So, in real life, beyond science and the different grades of scientific evidence, the balance is 
QRWVRPXFKEHWZHHQWKH³SURVDQGFRQV´EDVHGRQVFLHQWLILFHYLGHQFHLQWKHZRUVWFDVHDQ
individual FSW may profit only very little, but her clients and their private partners as well 
as her own prLYDWHSDUWQHUVFDQSURILWEXWPXFKPRUHEHWZHHQWKH³SURV´DQGWKHKLJKFRVWV
�7�K�H����S�U�R�V����P�X�V�W��R�X�W�Z�H�L�J�K��W�K�H��K�L�J�K��F�R�V�W�V���D�Q�G��W�K�L�V��L�V��G�L�I�I�L�F�X�O�W��W�R��H�V�W�L�P�D�W�H��R�U��F�D�O�F�X�O�D�W�H��R�Q��D�Q�
individual base, especially in a population which is prone to financial problems and different 
degrees of pre-exposition to HPV.  
 
But as already mentioned above, also secondary and subjective criteria have to be taken into 
account. From the public health view, offering vaccination for FSWs for free would be 
optimal in order to reach many FSWs (including those who would profit at most like the 
young beginners) and to generate some impact on herd immunity in their sexual network 
and the general population.  
 
One idea based on some data from the ARBYN Cochrane review (ARBYN et al. 2018) is 
that one might reduce the number of vaccine doses in financially deprived young FSWs (< 
25 years) to one or two doses, but the evidence for this reduction is much better for Cervarix 
than for Gardasil (in general, if all three doses are expected to be taken, one should favor 
Gardasil 9 for FSWs because of several advantages). But it is not so clear from these 
prospective trials whether the protective effect of 1 or 2 doses Cervarix lasts for as many 
years as that of three doses (a matter which is still open to debate in general), but many 
financially deprived young women practice sex work only for a few years until their financial 
situation improved, so even a limited protective effect (as far as the duration of the protection 
is concerned) of a reduced, suboptimal vaccination schedule might be (nearly) as protective 
for them as three doses.  
 
However, since median antibody titres are higher following Cervarix vaccination, favorable 
results from Cervarix trials (like the efficacy of 1 or 2 doses in young women < 25 years) 
cannot be applied directly to Gardasil vaccination. In fact, the case-control study from 
SILVERBERG et al. warns that 1 or 2 doses of Gardasil offer much less protection than 3 
doses even in teenagers (aged 14 ± 17 years) if one looks at the prevalence/incidence of 
CIN2+ or 3+ about 10 years after the first dose (the mean age of the women was 26.3 years 
at the time of study examination). In the Healthplan Setting in the San Francisco Bay area 
of the SILVERBERG study, only Gardasil (4) was applied; the study stopped too early to 
include participants who got Gardasil 9. Thus it is very evident now that a dose reduction (1 
or 2 doses) of Gardasil cannot be recommended even in teenagers (beyond 14 years) and 
very  young  adults.  This  conclusion  is  also  corroborated  by  several  Gardasil  studies 
mentioned in the discussion part of the SILVERBERG paper.      
 
Nevertheless, because of many advantages Gardasil 9 should be preferred for FSWs, but 
then there is no discussion about the number of vaccine doses needed, independent of age. 
Thus the high costs are the most important hindrance for vaccination.  
 
 
 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-18- 
 
 
But there are several possibilities to reduce the costs without dose reduction. Foreign FSWs 
might get the vaccine for lower costs in their home country; however this would delay 
vaccination until the next visit at home and may thus reduce its effectiveness due to infection 
risks in the meantime. It would be optimal for them to get the vaccine (at least the first dose) 
EHIRUHWKH\WUDYHOWRZHVWHUQFRXQWULHVIRUVH[ZRUNEXWWKH\SUREDEO\GRQ¶WNQRZDERXWWKH
�D�G�Y�D�Q�W�D�J�H�V��R�I��W�K�H��Y�D�F�F�L�Q�D�W�L�R�Q��D�Q�G��D�O�V�R��Z�R�Q���W��Kave the money for that in their home country, 
even if is cheaper there than in western countries.        
 
Some Health Insurances in Germany pay for HPV vaccination on a voluntary base until the 
age of 26 (so-FDOOHG³6DW]XQJVOHLVWXQJ´+RZHYHUDQRIILFLDO recommendation for FSWs 
and other adult persons with high HPV risks from Public Health Authorities (in Germany: 
Robert-Koch-Institut;  STIKO)  would  considerably  increase  the  chance  that  Health 
Insurances pay for vaccination. Such a recommendation must not contrast to our views 
(6&+,09$1'(5/2())¶VSDSHU and this comment), that the decision pro or contra 
vaccination should be made on an individual balance and that there is too little evidence in 
IDYRU RI D ³JHQHUDO UHFRPPHQGDWLRQ´ ZKLFK ILWV WR DOO LQGLYLGual  FSWs.  The  official 
recommendation can point to the need for such an individual balance.  
 
Furthermore, cost effectiveness is an important aspect for Public Health authorities. The 
calculations for an age-mixed FSW cohort, based on the 7 years end-result of VIVANE and 
adapted to the cervical HPV16/18 prevalence in FSWs, hints in favor of cost effectiveness 
(see  below).  However,  if  the  authorities  still  have  doubts  or  want  to  improve  cost 
effectiveness, they might reduce positive recommendations (based on individual balance) to 
those FSWs (or adult women at high risks in general) with a negative HPV16-DNA result at 
the cervix before the first vaccine dose, and/or certain (younger) age groups of FSWs and 
other highly exposed risk groups. Though this is suboptimal since such recommendations 
would exclude some FSWs/persons who might profit from vaccination too (e.g. FSWs who 
are HPV-16-DNA positive at the cervix, but practice a lot of unprotected oral sex), but any 
recommendation in favor of HPV vaccination of FSWs or subgroups of FSWs is better than 
no recommendation at all with regard to the question whether health insurances may pay the 
vaccination. With an official recommendation in the background, it would be much easier 
for FSWs to get the vaccination from their health insurance, provided they are member of a 
health iQVXUDQFHDWDOOPDQ\IRUHLJQ)6:VGRQ¶WKDYHDQ\KHDOWKLQVXUDQFH or maybe only 
a cheap travel health insurance which pays in case of acute illness, but not for vaccination).  
 
In case the FSWs have to pay vaccination by their own, they should keep the bills and fiscal 
officers  should  accept  these  costs  as  professional  outlay  if  FSWs  make  a  tax  return 
declaration. The professional reason for that vaccination can be compared (from a fiscal 
office point of view) to Hepatitis B vaccination for medical professions (which applies to 
FSWs too).  
     
Another important open question is whether it is wise to recommend a cervical HPV test (at 
least for HPV 16) in FSWs or sexually experienced/pre-exposed women who want to start 
VH[ZRUNLQRUGHUWRUHVSHFWWKHUHVXOWVRIWKLVWHVWLQWKHEDODQFHRIWKH³SURVDQGFRQV´Of 
FRXUVHWKHWHVWPXVWEHDEOHWRGHWHFW+39VSHFLILFDOO\DQGQRWRQO\³+39´RU³+5+39´
in general). Meanwhile, there are also self-tests on the market, and vaginal smears seem to 
be  at  least  as  effective  (and  probably  more  effective)  than  cervical  smears  to  detect 
infections.  
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-19- 
 
 
It is not generally recommended to do such a test before vaccination or to make vaccination 
dependent on it, but the situation may be very different in a (heavily) pre-exposed population 
like active FSWs, compared to the vaccination routine for children and adolescents of the 
general population. The combined data from several studies mentioned in the ARBYN 
review clearly show that a current infection with HPV 16/18 at the time of vaccination 
reduces the protective effect of the vaccine against HPV16/18-associated CIN2+ within the 
follow-up period of (in most studies) 3 to < 4 years to nearly zero, whereas seropositivity 
alone (i.e. in the absence of a current infection) results only in a small reduction of the 
vaccine efficacy with regard to the same endpoint. In other words: clearance of an infection 
which  was  present  before  vaccination  (but  no  longer  during vaccination)  results  in  an 
HIIHFWLYHQHVVRIWKHYDFFLQHFORVHWRWKDWLQQDwYHRUSUHVXPHGWREHQDwYH³SVHXGR-QDwYH´
women. This rises the question whether it is better to delay vaccination until the current HPV 
16 infection is cleared in women who are HPV 16 ±positive (cervical) at the time when 
vaccination is intended. If one follows this concept, a cervical HPV(16) test prior to the first 
dose should be recommended to heavily pre-exposed women. A negative test would then 
PHDQDELJ³SUR´LQIDYRURI vaccination. A positive test would result in the need to rebalance 
the decision on vaccination, especially the time of vaccination. There still may be reasons 
not to delay the vaccination until cervical HPV 16 cleared, e.g. protection with regard to 
unprotected oral sex as mentioned above, or protection of clients who perform unprotected 
cunnilingus (since they have a high risk when it is already proven that the genital area of the 
FSW is infected by HPV 16). It would be wise to recommend to these FSWs thaWWKH\GRQ¶W
allow unprotected cunnilingus to their clients until the LQIHFWLRQFOHDUHGLIWKH\GRQ¶WWDNH
the vaccine: 
 
Firstly for the protection of their clients, but also for the currently infected FSW herself: 
frequent  cunnilingus  might  delay  HPV  clearance  due  to  micro-traumata  or  low-grade 
inflammation (caused by some components/antigens from the saliva or secondary to micro-
traumata).  
 
,GRQ¶WWKLQNWKDWFHUYLFDO+39WHVWLQJEHIRUHYDFFLQDWLRQVKRXOGEHUHFRPPHQGHGLQ
every case when HPV vaccination of active FSWs or beginners is considered. But it may 
help to improve the accuracy of the balance in individual cases when the balance is less clear. 
But there is one important exception: if a HPV-type from the vaccine is present at the cervix 
and associated with a pathology which has to be treated locally/surgically, vaccination is 
recommended in order to reduce the risk of recurrence (as mentioned also in SCHIM VAN 
DER LOEFF et al.).     
 
 
Missing studies from the FSW population  
 
,GRQ¶WH[SHFWthat we get better evidence (pro or contra) in the next years. The highly mobile, 
internationally working and often anonymous FSW population is not available for long 
lasting RCTs. Such RCTs would have to last at least 7 years or (better) longer. The problem 
with most of the published prospective trials with young or middle-aged adult women was 
the  short  duration  of  3 ± 4  years (see  ARBYN  et  al.,  where  all  of  these  studies  are 
summarized in the several tables). It is known that it takes about 4 years (or more) from a 
new (incident) infection (which later became persistent) to develop into CIN2+ (e.g. VINK 
et al. 2011: mean 3.75 years) or CIN3 (e.g. RKI 2008b: 3 ± 6 years, i.e. mean ~ 4.5 years). 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-20- 
 
 
 It is well-known WKDWYDFFLQDWLRQGRHVQ¶WDIIHFWWKHQDtural history and fate of an infection 
which is already present at the time of vaccination. So what should one expect 3 or 4 years 
after the first vaccine dose in potentially/probably/heavily pre-exposed women, if CIN2+ or 
CIN3 is the endpoint? In fact: nothing or something very little and insignificant.  
 
When a trial is cut after 3 or 4 years, one may be able to see a reduction of incident, prevalent 
or persistent infections of HPV types from the vaccine (and some cross protection) in the 
vaccine group, maybe also reductions of CIN1 (however, one should respect that CIN1 is 
not so tightly correlated with HRV 16/18 or HRHPV in general as CIN2+), but one cannot 
expect a significant reduction in CIN2+ or CIN3 because all or most CIN2+ which occur in 
this time window of the first 3 ± 4 years after the first vaccine dose develop due to infections 
which were already present at the time of vaccination. Even if the trial involved a cervical 
HPV test at start, it could have been false-negative in those cases who develop CIN2+ or 
CIN3 during the first ~ 3 years after the first dose. The accuracy of these tests is not 100 %, 
or the prevalent infection was in a latent stage at the time of testing.   
 
Thus  a  reduction  of  CIN2+/CIN3  cannot  be  expected  before  year  4  or  5  in 
potentially/probably (heavily) pre-exposed women. In fact, the final results of the VIVIANE 
trial (WHEELER et al. 2016) (with an average age of 37 at the time of vaccination and a 
pre-exposition with regard to HPV16/18 which was heavier than what is expected in the 
general population) found two cases of CIN3+ in the vaccine group and six cases in the 
control group in the years 5 ± 7, while there was no profit from the vaccine with regard to 
CIN2+ and CIN3+ in the years 1 ± 4. This means a risk reduction of 67 % for CIN 3+ 
(irrespective of  HPV type) in the vaccine group in the years 5 ± 7, close to that what can be 
expected from a vaccine which is only bivalent (70 % or 70 ± 80 %, but not more for the 
distant endpoints).   
 
But the need for local treatment at the cervix was reduced by 62 % in the years 5 ± 7, 
compared to the control group (10 versus 26 cases), and this difference was significant. Both 
endpoints (CIN 3, need for local treatment) are of high practical relevance for FSWs. Maybe 
DSUHYDOHQWRUSHUVLVWHQWLQIHFWLRQRU&,1GRHVQ¶WRUVKRXOGQ¶WZRUU\WKHPDORWDWOHDVW
�L�Q����\�R�X�Q�J�H�U����\�H�D�U�V����E�X�W��D��&�,�1����R�U��W�K�H��Q�H�H�G��I�R�U��W�U�H�D�W�P�H�Q�W��D�W��W�Ke cervix is an important 
undesirable event and means some or a lot of inconveniences, costs and also risks.   
 
Contrary to local treatment needs (-  62 %; sign.), due to the small numbers of CIN 3+ cases 
in this age group, including the control group (the VIVIANE women had reached a mean 
age of 44 years when the trial stopped), the difference in CIN3 (2 : 6) is not significant. The 
VWXG\ZDVXQGHUSRZHUHGWRILQGDVLJQLILFDQWGLIIHUHQFHIRU&,1LQWKLV³ROG´DJHJURXS
(where the incidence of CIN2+ or 3+ is lower than in younger age groups even in the absence 
of vaccination), or the trial stopped too early to make the difference significant. However, 
there were significant reductions of CIN1+ (irrespective of HPV type) in months 48 ± 84 (- 
48.1 %) in the total vaccinated cohort (i.e. irrespective of HPV pre-exposition), of CIN 1+ 
(HPV16/18-associated) at month 84 in the total vaccinated cohort for efficacy (i.e. only pre-
exposition to either HPV 16 or 18 was allowed) (-75.5 %) and non-significant reductions of 
CIN 1+(16/18) (-75.5 %) and CIN 2+(16/18) (-80.4 %) in the same cohort at month 48 and 
a nearly significant reduction (missing significance just a little) of 78.2 % for CIN2+(16/18). 
Taken all of these results into account, it is highly improbable that the reduction of CIN 3 (-
67 %) in years 5 ± 7, though formally insignificant, is the result of a random. 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-21- 
 
 
Thus the VIVIANE trial made it clear that vaccine trials in adult, (possibly/probably/heavily) 
pre-exposed women should last much longer than four years if one wants to look at endpoints 
of clinical relevance like CIN2+/CIN3+ or the need for treatment at the cervix. But those are 
the endpoints which are really interesting and relevant for the FSWs themselves. It can be 
modelled from these final results of the VIVIANE trial, that less than 25 FSWs of an age-
mixed cohort of FSWs (but younger than VIVIANE on average) would have to be vaccinated 
to prevent one case of need for treatment at the cervix in the years 1 ± 7 after vaccination, 
and  thus  much less  than  25  to  prevent  one  case  of  need  for  treatment  lifelong  after 
vaccination  (own  results  in  a  forum-based  paper).  The  model  is  based  on table  4  in 
WHEELER et al; it is calculated in table 4 that 143 local cervical treatments can be prevented 
by Cervarix per 100.000 women-years of the study population (mean age at first vaccine 
dose: 37 years) in the first seven years after the first vaccine dose, i.e. 143 per (100.000 : 7) 
= 14.283 women within seven years, i.e. 1 per 100 women of the study population within 
the first seven years after the first dose.   
 
The model calculation for an age-mixed cohort of FSWs is based (i) on 15 % cervical HPV16 
or 18 prevalence at the time of vaccination (according to SOOHOO et al. for FSWs from 
Europe, but also in full agreement with own data from Amsterdam from SCHIM VAN DER 
LOEFF¶VJURXS [MARRA et al. 2018]), contrary to ³only´ 4  % in the total vaccinated cohort 
from VIVIANE, and (ii) a 4.5-fold risk for treatment needs at the cervix in unvaccinated 
FSWs compared to unvaccinated control women (based on 4- to 5-fold risk for CIN2+ or 
HSIL  in  the  meta-analysis  from  8  studies  with  6965  FSWs  and  even  more  controls 
mentioned above). 
 
In an analogous way one may calculate that (much?) less than 35 FSWs of an age-mixed 
cohort (but on average significantly younger than VIVIANE) have to be vaccinated to 
SUHYHQWRQHFDVHRI&,1&,1LQ\HDUDQGEH\RQGOLIHORQJ+RZPXFK³OHVVWKDQ´
will depend on the age structure of the cohort (the younger, the less). But this model is less 
robust because it is based on small numbers (2 : 6 cases of CIN3 in years 5 ± 7 in VIVIANE), 
whereas the database for the calculation of the prevented treatment needs is much more 
robust and includes statistical significant components (the reduction of treatment needs in 
years 5 ± 7 was found to be significant).   
 
Because RCTs (especially of the recommended duration > 7 years) are not possible in the 
FSW population, one might consider case-control studies with one-time examination (maybe 
even  anonymous/pseudonymous)  to  compare  the  HPV  infection  or  PAP/CIN  status  of 
vaccinated FSWs to those who are unvaccinated. This might be interesting, but has also a 
ORWRIOLPLWDWLRQVDQGPD\ UHVXOWLQ ³ZURQJ´ FRQFOXVLRQVZLWK UHJDUGWRWKHORQJ-term 
effectiveness and advantages of the vaccination. First, HPV-DNA (e.g., 16/18) prevalence 
is not a good indicator since it may even represent a fresh contamination from a client which 
may not result in a true (stable) infection. Persistent infection (6 or 12 months) is a much 
better endpoint with higher clinical relevance and avoiding the contamination bias; however 
it is not available following one-time examination. The next endpoint (after prevalent HPV 
infection) which is available in one-time examination is PAP grading/CIN.  
 
Second, as the SCHIM VAN DER LOEFF group found in a former paper (MARRA et al. 
2017), one cannot rely on the anamnestic history of the FSWs; most of who pretended to 
have gotten HPV vaccination were unvaccinated and probably confounded HBV and HPV  
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-22- 
 
 
vaccination. Therefore the HPV vaccination status would have to be verified by serological 
testing if one needs accurate information for a case-control trial.  
 
Third, if the FSWs have problems to recall correctly whether they got the HPV vaccine, it 
would be even more difficult for them to remember the exact timing of vaccination. But this 
is very crucial: As mentioned above, a reduction of CIN 2+ in probably/heavily pre-exposed 
adult women can only be expected after about 4 years after the first vaccine dose. CIN 2+ 
which are discovered within the first ~ 3 ± 4 years after the first dose are probably due to 
earlier infections which were already present at the time of the first dose. (In a case-control 
study, one has no information about HPV status at the time of vaccination). To analyze the 
long-term efficacy, one should exclude all FSWs who got the vaccine less than 4 years ago, 
or analyze and present the results (e.g. < 3 years, 3 ± 4 years; 4 years and beyond; time since 
first dose) separately. 
 
Since the answers of the FSWs about the time of their vaccination cannot be verified in most 
cases (except there are bills or so), this is a fundamental obstacle for such a study. If one 
GRHVQ¶WUHVSHFWWKHa\HDUVFXW-off in respect to the time since vaccination, such a study 
may result in the conclusion that HPV vaccination of FSWs is ineffective (as far as cervical 
CIN2+ status or need for treatment are concerned). And with regard to CIN 2+, one should 
look for the underlying HPV type and present the results for CIN2+ associated with HPV 
types from the vaccine separately. Otherwise effects of the vaccine may be overlooked or 
WRRPXFKGLOXWHGLIRQO\³DOO&,1´LUUHVSHFWLYHRI+39W\SHDUHSUHVHQWHG&,1ZLWK
HPV from the vaccine, especially HPV 16/18, have a higher risk of progression than any 
CIN2+, so the distinction between CIN2+ (16/18) and CIN2+ (irrespective of HPV type) is 
not only of academic interest or a trick to improve the statistics of vaccine efficacy in trials 
funded by the vaccine industry, but it is an important matter of clinical and prognostic 
UHOHYDQFHDQGWKXVLWLVMXVWLILHGWRGRVR$WEHVWUHVXOWVIRU³DOO&,1´LUUHVSHFWLYHRI
HPV type) and CIN2+ (16/18) should be presented both.  
 
Even if the vaccine is highly effective in pre-exposed women, no significant reduction in 
CIN2+ could be expected during the first 3 ± 4 years, so one has to look what happens beyond 
4 years, and most RCTs with post-adolescent women failed to do so. Anyway, though the 
study  concept  described  above  (one-time  examination,  PAP  pathology  or  CIN status, 
verification of HPV vaccination status by serological tests, documentation of the exact time 
of vaccination) could work in theory even with FSWs, with CIN2+ or (better) CIN3+ as 
endpoints and the need WRGLIIHUHQWLDWHEHWZHHQ³IUHVK´- \HDUVDQG³ROG´\HDUVDQG
more)  vaccination,  one  would  need  very  high  numbers  of  study  participants  to  avoid 
underpowering, thus even this kind of one-time examination study is practically impossible.  
 
In summar\,GRQ¶WH[SHFWEHWWHUHYLGHQFHZLWKUHJDUGWRWKHVXEMHFWRI+39YDFFLQDWLRQRI
FSWs in the next years, at least as far as trials within the FSW population are concerned. So 
one will have to keep on relying on indirect evidence in the next years and probably in the 
next decades. 
 
 
 
 
 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-23- 
 
 
Contra arguments from SCHIM VAN DER LOEFF et al. 
 
There are some aspects one may  consider in  the context  of the five contra arguments 
mentioned in the paper. 
 
First, the A5298 trial (WILKIN et al. 2018) is indeed disappointing with regard to persistent 
anal  infections.  There  was  only  a  non-significant  risk  reduction  of  about  one  quarter. 
However, this was a small study and the small effect (one quarter) might have been become 
significant in a larger study population. So it is still open whether the effect of the vaccine 
with  regard  to  persistent  anal  infections  in  a  heavily  pre-exposed  population  is  truly 
zero/completely negligible or whether it is small, but still of interest for populations with 
high anal risks (~ 25 % risk reduction may be better than nothing for people at high risk).  
 
0RUHRYHULWLVXQFOHDUZKHWKHUWKHVHGLVDSSRLQWLQJ³DQDO´UHVXOWVZRXOGDSSO\WRFHUYLFDO
infections in FSWs too. There is of course a strong epidemiological correlation between 
cervical and anal infections in women in general and especially in FSWs (e.g., MARRA et 
al. 2018). 
 
Because only about 16 % of FSWs practice anal sex with clients (due to a recent worldwide 
meta-analysis; OWEN et al. 2019), which also applies to our area (own data), most anal 
infections in FSWs may be due to smear infections from the genital area, i.e. autoinoculation. 
The ACTG A5298 population was dominated by HIV-infected MSM. The natural history of 
anal HPV infections in MSM may differ from the natural history of cervical infections in 
women or FSWs (without HIV), respectively. (Micro-)traumata and inflammations delay or 
impair clearance, and they may also stimulate reactivation of latent infections or infections 
which already entered the process of clearance, but had not yet completed it, thus stopping 
or reversing this process. The anal area is prone to micro-traumata and (maybe low-grade) 
inflammation or even some grade of eczemas, even in the absence of anal sex, and even 
more in the presence of anal sex. Since the A5298 trial involved mostly HIV-infected MSM, 
LWLVSUREDEOHWKDWPDQ\RIWKHPSUDFWLFHGDQDOVH[ZKHWKHUSURWHFWHGRUXQSURWHFWHGGRHVQ¶W
matter in the context of mechanical micro-traumata and trauma-induced inflammation).  
 
The cervical aUHDLVDPXFKPRUH³SURWHFWHG´DUHDHYHQLQ)6:VLIWKH\ZRUNFRQVLVWHQWO\
with condoms) where HPV infections may have the chance to clear with less disturbance 
and interruptions than in the anal area of MSM. However, this question is still open, and 
these are only theoretical considerations.  
 
But there is another aspect, and this one is based on studies. SCHIM VAN DER LOEFF et 
al.  consider the A5298 trial as the only trial so far in a heavily pre-exposed population. 
However, there are vaccination RCTs in young adult women where a substantial portion of 
the participants were HPV16/18-negative, but HPV16/18-seropositive at the time of the first 
dose. So this subpopulation among the participants was pre-exposed to an extent of 100 %. 
According to the results of these prospective studies, summarized and combined in the 
ARBYN Cochrane review (table 11), the effectiveness of the vaccine in these pre-exposed 
women was quite good and only a little less than in naïve (or pseudo-naïve) women from the 
same RCTs. (I FDOOWKHVHZRPHQ³SVHXGR-QDwYH´women who were infected in the past, but 
GLGQ¶WJHQHUDWHDQWLERGLHVRUWKHDQWLERGLHVWLWUHVDUHORVWRUWRRORZPHDQZKLOHEHORZWKH
�O�L�P�L�W�V��R�I��G�H�W�H�F�W�L�R�Q���V�R��W�K�D�W��W�K�H�V�H��Z�R�P�H�Q��D�U�H��P�L�V�F�O�D�V�V�L�I�L�H�G��D�V����Q�D�w�Y�H�����E�X�W��L�Q��I�D�F�W��W�K�H�\��D�U�H�Q���W� 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-24- 
 
 
truly naïve). The results from the RCTs mentioned in the ARBYN review are confined to 
WKHFHUYLFDOPXFRVDOVLWH7KHVH³FHUYLFDO´UHVXOWVDUHWKXVLQFRQWUDVWWRWKHUHVXOWVIURPWKH
anal site of the A5298 trial. And the endpoint used in these studies (CIN2+) is of higher 
clinical  relevance  than  persistent  infection.  These  differences  may  be  explained  by 
differences in the natural history of cervical infections versus anal infections in MSM as 
discussed above.    
 
Nevertheless, the study design of the A5298 trial, which produced disappointing results 
concerning persistent anal infections, showed favorable effects with regard to oral infections. 
7KLV XQGHUOLQHV WKH JUDGH RI HYLGHQFH IRU WKHVH ³RUDO UHVXOWV´ LQ VSLWH RI WKHLU ODUJH
confidence intervals. 
 
The second contra argument in the SCHIM VAN DER LOEFF paper is the suggestion that 
women who once cleared an earlier HPV infection in the past, will also do so in the future. 
As the authors point out in their paper, vaccination of these women will still have a protective 
effect for their clients,  which attenuates this contra argument a little.  I agree that the 
probability that these women can clear a reinfection of the same HPV type is higher than in 
women who were not pre-exposed to that type, but there is no guarantee that they will do so. 
Otherwise, there would be no cases of new infections, new persistent infections or new 
CIN2+ in the RCTs among those women who were seropositive but HPV-DNA-negative at 
the time of vaccination, even in the unvaccinated or placebo-vaccinated control group. So, 
in fact, reinfection (or reactivation) may happen also in women who seemed to have cleared 
the former infection, and who were able to generate natural antibody titres which reduce the 
risk of reinfection to some extent, depending on the titre (BEACHLER et al. 2016, see 
below).    
 
Furthermore,  the age-dependent incidence  of  cervical  cancer  shows  two  or  even  three 
maxima in different age groups, which may point to a reduction of natural immunity against 
HPV with increasing age. Since cancer is diagnosed at least 7 ± 8 years after infection, and 
in most cases about 20 years later, the decrease in natural immunity must have started many 
years before the second and possibly third maximum of cancer incidence (time of diagnosis).   
 
Second, many FSWs are smokers, or intensify pre-existing smoking during sex work. Other 
FSWs are at first nonsmokers, but started smoking sometime after starting sex work. So even 
if they were able to clear infections in the past when they were nonsmokers, it is unsure 
whether they well be able to do so while heavily smoking? Smoking impairs clearance and 
promotes persistence and progression. 
 
BEACHLER et al. (2016) showed the effects of natural antibody titres and the risk of 
reinfection  (which  may,  of  course,  also  mean  reactivation  of  a  non-diagnosed  latent 
infection). It is clear from these results that women who were able to clear the infection (or 
to reduce the infection to a latent stage) may suffer from reinfection (or reactivation), and 
the probability is dependent on the natural antibody titre (as expected). As a consequence, 
women with cleared infection and low natural antibody titre have a higher risk of reinfection 
(and may profit more from vaccination) than women with high (> median) natural titres. 
These  effects  were  also  confirmed  in  an  analysis  of  the  Costa  Rica  Vaccine  Trial 
(BEACHLER et al. 2015). In seropositive (but HPV-DNA-negative) women (18-26 years) 
at the time of vaccination, the effectiveness of the vaccine against prevalent infections was  
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-25- 
 
 
high when antibody titres were low; in contrast, those with high natural antibody titres did 
profit only a little from vaccination. Not all details were discussed in this paper, maybe due 
to limitations of the length of the article by the journal. If one analyses the data from the 
tables, it becomes evident that the weak effect of the vaccination in those with high natural 
titres before vaccination is due to a much smaller basic risk of reinfection (or reactivation) 
in this group. As expected, high natural antibody titres generate a high-grade (though not 
complete) natural protection against reinfection (or reactivation?), and in this subpopulation 
of seropositive women with high natural titres, vaccination offers only little additional effect. 
In women with lower titres, the risk oIUHLQIHFWLRQRUUHDFWLYDWLRQGLGQ¶WGLIIHUDORWIURP
seronegative women (cf. BEACHLER et al. 2016), and thus the efficacy of the vaccine was 
close to that in seronegative women.  
 
Theoretically, in HPV-DNA-negative (potentially/probably/heavily) pre-exposed women, a 
test of seropositivity (as least for HPV 16) and, in case of a positive result, the quantification 
of the antibody titre, might have some predictive value with regard to the effectiveness of   
vaccination. However, due to the rarity of high natural antibody titres in a total cohort of 
ZRPHQLQWHQGHGWRJHWYDFFLQDWHGTXDQWLWDWLYHDQWLERG\WHVWLQJEHIRUHYDFFLQDWLRQGRHVQ¶W
seem  to  be  cost-effective.  And  if  one  considers  any  pre-testing  in 
(potentially/probably/heavily) pre-exposed women before vaccination in order to improve 
WKHTXDOLW\RIWKHEDODQFHRI³SURVDQGFRQV´ or the cost effectiveness of vaccination, then 
cervical  HPV(16)-testing  is  of  much  greater  relevance  than  qualitative  or  quantitative 
antibody testing (however, antibody testing may be recommended to FSWs who want to get 
the vaccine or are recommended to do so but are unsure whether they got it already in the 
past.  In  case  of  a  positive  result,  i.e.  an  antibody  titre  which  is  indicative  of  former 
vaccination,  they  save  the  money  for  vaccination.  In  such  cases,  quantitative  or  semi-
quantitative antibody testing is certainly cost effective).  
 
The third contra argument concerns those pre-H[SRVHGZRPHQZKRDUH³SVHXGR-QDwYH´
according to the definition mentioned above. I think this is not a matter of great concern. As 
shown in the ARBYN review and the CVT, in young HPV-DNA-negative women with low 
DQWLERG\WLWUHVWKHYDFFLQHLVQHDUO\DVHIIHFWLYHDVLQ³QDwYHSVHXGR-QDwYH´ZRPHQ6R
why should the vaccine work less effective in pre-exposed women without natural antibodies 
(or below the limits of detection) compared to those with detectable low antibody titres?   
 
The fourth  contra  argument  is  an  important  one,  because  the  relative  importance  of 
latency/reactivation versus true reinfection is understood badly and the prevalence of latency 
may be actually underrated. But this problem applies to all RCTs with potentially/probably 
pre-exposed  adult  women.  As  far  as  a  high  effectiveness  of  the  vaccine  (e.g.  against 
prevalent or persistent infections or CIN2+ by HPV16/18) in adult women who were HPV-
DNA-negative at the time of vaccination is shown in WKHVH5&7VLWGRHVQ¶WPDWWHUDORW
whether  the  vaccine  prevented  reinfection  or  reactivation  of  latent  infections.  A  latent 
infection whLFKQHYHUUHDFWLYDWHVSUREDEO\GRHVQ¶WSURJUHVVLQWKHGLUHFWLRQRIFDQFHUDQG
since there is no viral shedding during latency, there is also no risk for clients. Some other 
vaccines which are in development (e.g. against Herpes simplex 2, another infection with 
some relevance for FSWs) are aimed, beside prevention of primary infections, to suppress 
UHDFWLYDWLRQLQLQIHFWHGSHRSOH+69GRHVQ¶WFOHDUDQGFDQQRWEHHUDGLFDWHGIURPDQ
infected person. Once a person is infected, one can only try to suppress reactivation by 
antiviral agents or, maybe in the future, vaccination). The same applies to Varicella 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-26- 
 
 
vaccination in older people to prevent shingles. Since the HPV vaccine is unable to clear a 
prevalent infection, it is highly plausible that it is also unable to clear a latent infection. 
However, if it can prevent reactivation (what is still a matter of debate), it LVVWLOO³HIIHFWLYH´
and the question of latency is of minor relevance.  
 
Fifth,  the  SILVERBERG  case-control  study  is  an  important  one,  but  there  are 
methodological  constraints.  As  pointed  out  above,  in  (possibly/probably)  pre-exposed 
women, an effect of vaccination against endpoints like CIN2+ or CIN 3 cannot be expected 
prior to 3 or 4 years after the first dose of the vaccine, and the available evidence suggests a 
cut-off at 4 years after that dose. As has already been done with the final results of the 
VIVIANE trial (where the participants were comparatively more pre-exposed due to the 
inclusion criteria and their mean age of 37 years at the time of the first dose than women 
from the general population, though much less pre-exposed than active FSWs) (WHEELER 
et al. 2016), one should distinguish between events like CIN2+ or CIN3 which occur in the 
first ~ 4 years after the first dose (< 4 years) and those which occur later (4 years and beyond). 
For example, the VIVIANE trial found no reduction of CIN 2+ (irrespective of HPV type) 
at all in the totally vaccinated cohort (i.e. the cohort including the pre-exposed women) 
compared to controls within months 0 ± 48 (time interval), but 33.8 % in month 48 (point of 
time). Confined to CIN2+ associated with HPV 16 or 18, the risk reduction was about 80 % 
at month 48 and also at month 84, so the cut-off is expected to be somewhere between 3 and 
4 years (see table below).  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-27- 
 
 
_______________________________________________________________ 
 
Results from WHEELER et al. (VIVIANE trial; eff.: efficacy of Cervarix):  
 
 
CIN 1 +  
0 ± 48 months (irresp.*; TVC)**:         eff.: 14.8 %  (n.s.)  
48 months (16/18***; TVCE****):     eff.: 75.5 % (n.s.) 
48 ± 84 months (irresp.; TVC):             eff.: 48.1 % (sign.)  
84 months (16/18; TVCE):                   eff.: 75.5 % (sign.) 
 
CIN 2+ 
0 ± 48 months (irresp*.; TVC)**:         eff.: -1.2 %  (n.s.) 
48 months (16/18***; TVCE****):     eff.: 80.4 % (n.s.) 
48 ± 84 months (irresp.; TVC):             eff.: 33.8 % (n.s.). 
84 months (16/18; TVCE):                   eff.: 78.2 % (n.s., but nearly sign., CI: -13.1 to 98.0) 
 
CIN3+ 
There are too few cases (0 or 1) for calculations at the time points 48 and 84 months.  
0 ± 48 months (irresp*.; TVC)**:         eff.:   none (48 vs. 39 cases among controls) 
48 ± 84 months (irresp.; TVC):             eff.:   67 % (2 vs. 6 cases; n. s.). 
 
Local cervical treatment: 
0 ± 48 months (irresp*.; TVC)**:         eff.:  7.2 % (n.s.)  (78 vs. 84 cases) 
48 - 84 months (irresp.; TVC):              eff.:  61.9 % (sign.; 10 vs. 26 cases). 
__________________________________________________________________ 
 
*irrespective of HPV type 
** total vaccinated cohort, i.e. irrespective of HPV pre-exposition at the time of vaccination 
*** only CIN which are associated with HPV 16 and/or 18 
**** total vaccinated cohort for efficacy (excludes women who were pre-exposed to both HPV 16 
and 18 at the time of vaccination; pre-exposition like HPV-DNA or seropositivity to one of the two 
types  was  allowed;  thus,  TVC  is  more  inclusive  than  TVCE  and  TVC  resembles  a  cohort  of 
potentially pre-exposed women without any testing before vaccination) 
 
 
 
The results show obvious differences with regard to vaccine efficacy between 0 ± 48 months 
on one side and 48 months, 48 ± 84 months and 84 months on the other side. The study was 
probably underpowered to find significant differences in all of the categories from 48 months 
RQZKLFKPD\LQSDUWDOVRUHVXOWIURPWKH³ROG´DJHVWUXFWXUHRIWKHFohort and thus 
comparatively low risks for events like CIN2+ or treatment needs even in the control group. 
As shown my McCLUNG et al (2019), in the United States the number of diagnosed CIN2+ 
(HPV16/18) in women aged 40-44 years (which corresponds to the mean age of the women 
in the last years of VIVIANE) was about 75 % less than in women between 20 and 29 years 
at  the  time  when  HPV  vaccine  was  introduced  and  before  the  first  effects  on  CIN2+ 
prevalence could be expected in those age groups (year 2008).    
 
 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-28- 
 
 
The VIVIANE data suggest that in such a (potentially/probably pre-exposed) population a 
cut-off close to 48 months should be chosen if one wants to identify protective effects of the 
vaccine for endpoints like CIN2+ or, better, CIN2+ (16/18). Before that point of time, no or 
only negligible effects can be expected with regard to endpoints of cytological/histological 
relevance like CIN 2+ because all or nearly all of the CIN2+ which occur during the first 3  
years after the first vaccine dose, and many of the CIN 2+ which occur in year 4, will result 
from HPV infections which were already present at the time of vaccination, especially if 
HPV status at the time of vaccination is not examined or considered.   
 
The longer the time since vaccination, the higher the relative (protective) effect of the 
vaccine compared to unvaccinated controls, and the higher the cumulative incidence of 
events which could be prevented by the vaccine. This is demonstrated very well in figure 2 
LQ:+((/(5HWDO³FX�P�X�O�D�W�L�Y�H��L�Q�F�L�G�H�Q�F�H��R�I��&�,�1�����L�U�U�H�V�S�H�F�W�L�Y�H��R�I��+�3�9��W�\�S�H�����
where the number of prevented events increased on a constant rate, starting after about 2 ± 
2.5 years, without any signs of weakening or saturation towards the end of the study (7 
years), so that it is very plausible that the cumulative incidence of prevented CIN 1+ would 
have continued to increase by the same rate if the trial had lasted longer. It is also plausible 
that the same applies to CIN 2+ or 3+, though this is not shown in that graph probably due 
to the smaller number of events and some temporal delay compared to CIN 1+. It is logical 
that vaccine-induced reduction of CIN 1+ starts earlier than for CIN2+ or CIN3+. 
 
At the time of study examination in SILVERBERG et al., cases (CIN2+ or CIN3+, resp.) 
and controls had a mean age of 26.3 or 26.4 years, resp. After adjustment, SILVERBERG et 
al. found a non-significant 15 % reduction of CIN 3+ in those who got the first dose at an 
age of 21 years and beyond, a significant 41 % reduction of CIN3+ in those who got the first 
dose at 18 ± 20 years, and of 73 % in those who got the first dose at 14 ± 17 years, provided 
that they got three (or more) doses. With one or two doses, the reductions were only 10 % 
\HDUV�� 20 years) and 21 % (14 ± 17 years). 
 
Another important limitation of the SILVERBERG study is that it refers to all CIN2+/CIN3+ 
(irrespective of HPV type), and the efficacy of the vaccine might have been much more 
pronounced if there had been made calculations based only on CIN2+/CIN3+ associated 
with HPV 16/18 VHHIRUH[DPSOHWKHGLIIHUHQFHVEHWZHHQWKH³&,1LUUHVS´DQG³&,1
��������categories in the table above with data from the VIVIANE trial).  
 
Taking into account the time span between a new infection (which could be prevented by 
the vaccine) and the occurrence of CIN2+ or CIN3+ which may result from that infection, 
what should one expect for those participants who had a mean age of 26.3 or 26.4 years at 
the time of study examination and who got the first dose at an age of 21 years and beyond 
(only those who got the first dose during the last six months before study examination were 
excluded from the study)? Only few of these women will have spent significant longer times 
than 4 years after vaccination, and those few women may contribute to the 15 % risk 
reduction of CIN3+ and 8 % risk reduction of CIN2 + in the 3-dose group. It would be very 
interesting to see results separately for those (among those vaccinated at 21 years and 
beyond) who got the vaccine less than 3 years ago compared to those who got it 4 years or 
more ago (assuming a cut-off somewhere between 3 and 4 years), but these calculations 
where not carried out. Additionally, as already mentioned above, it would be also very 
interesting to see these calculations (and all results) restricted to CIN 2+ (16/18) and CIN 3+  
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-29- 
 
 
(16/18). This matters a lot, since CIN2+ (16/18) and CIN3+ (16/18) have a worse prognosis 
than CIN2+ (non-16/18) or CIN3+ (non-16/18). 
 
Without these missing information and calculations, the results of SILVERBERG et al. are 
completely in line with what has to be expected from such a study design. However, this 
study is very valuable because it finally demonstrates that one or two doses of Gardasil are 
not enough. This was also suggested by former studies, but now it is definitely clear, even 
with respect to younger age groups beyond 14 years.     
 
Thus, in a retrospective study like that of SILVERBERG et al., it is important to respect 
thoroughly the time of the first vaccine dose and the time interval between first dose and 
study examination. If this is not done, this will attenuate the effectiveness of the vaccine in 
possibly/probably  pre-exposed  women because  of CIN2+/CIN  3  which  result  from 
infections which were already present at the time of vaccination.  
 
Moreover, since the cervical HPV status was not established at the time of vaccination in the 
SILVERBERG study (as a consequence of the retrospective design), it is not possible in 
such  a  study  to  exclude  those  who  were  infected  at  that  time, as  could  be  done  by 
CASTELLSAGUE et al. 2011.  
 
The SILVERBERG study, when considered in combination with the results from the RCTs 
with young adult women as summarized in the ARBYN paper, is thus another argument in 
favor  of  HPV(16-)testing  of  (probably/heavily)  pre-exposed  women/FSWs  prior  to 
vaccination in those cases when there are no extra-genital risks which favor HPV vaccination 
on their own (e.g., unprotected oral sex risks). For example, there was a 88.7 % reduction in 
the combined incidence of persistent infections, CIN or extragenital lesions like genital warts 
associated with HPV 6/11/16/18 after a median time of 4.0 years after Gardasil vaccination 
in 24 ± 45 years old women (age at the time of vaccination) in a RCT for the per-protocol 
cohort  (seronegative  and  PCR-negative  at  the  time  of  vaccination;  3  doses) 
(CASTELLSAGUE et al. 2011). So there is little doubt that those FSWs would profit from 
vaccination who fulfill the criteria of such a cohort (PCR-negative and seronegative), but, as 
discussed above, being PCR-negative is much more important than being seronegative since 
seropositivity in the absence of PCR-positivity reduces the efficacy of the vaccine only to a 
minor extent (compared to PCR-positivity).         
 
In spite of these limitations QRGLIIHUHQWLDWLRQEHWZHHQ³\RXQJ´DQG³ROG´YDFFLQDWLRQLH
�����R�U����\�H�D�U�V��Y�H�U�V�X�V�������\�H�D�U�V���Q�R��V�H�S�D�U�D�W�H��U�H�V�X�O�W�V��I�R�U��&�,�1����D�Q�G��&�,�1����D�V�V�R�F�L�D�W�H�G��Z�L�W�K�
HPV 16/18), which all attenuate the efficacy of the vaccine, the SILVERBERG paper 
recommends catch-up vaccination for 18 to 20 year old women with 3 doses, so in any way 
it is compatible with recommendations in favor of vaccination for very young FSWs or 
young beginners in sex work. Looking at the somewhat disappointing results for this age 
group (-32 % CIN 2+ and ±41 % CIN 3+ following three or more doses), one must recall 
that these reductions apply to all CIN2+/3+ irrespective of HPV type and that (in accordance 
with other trials) the reduction quote can be expected to be (much?) higher if the calculation 
was restricted to CIN2+/CIN3+ associated with HPV 16/18, but the association with special 
HPV types was not part of the study protocol.     
 
 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-30- 
 
 
In summary, there is substantial evidence that there is a not very small (but difficult to be 
calculated) portion of FSWs who may profit only (very) little from vaccination as far as their 
genital/cervical site (and, per smear infection, their anal site) is concerned. However, since 
it is extremely improbable that a FSW had been pre-exposed to all HPV types of Gardasil 9, 
there should be at least a small profit in any case, if Gardasil 9 is chosen for vaccination.  
 
However, the least genital profit can be expected for those who are actually infected by HPV 
16. Thus cervical HPV(16)-testing in heavily pre-exposed women like FSWs may have an 
important predictive value with regard to the effectiveness of the vaccine in that individual 
case and may improve the SUHFLVLRQRIWKHEDODQFHRI³SURVDQGFRQV´DQGDOVRWKHFRVW-
effectiveness of vaccination. 
 
So SCHIM VA1'(5/2())¶V DUJXPHQWV³FRQWUD´+39YDFFLQDWLRQDUHPRVWO\DUJXPHQWV
in favor of HPV(16)-pretesting of probably/heavily pre-exposed women (FSWs), and the 
result of this test should then become an important aspect of the balance suggested in the 
SDSHU%XWDVPHQWLRQHGDERYH,GRQ¶WWKLQNWKDW+39-pretesting is recommended in all 
FDVHVRI³DGXOW´RU)6:YDFFLQDWLRQ,ZRXOGrecommend it only in those cases where it 
would have a decisive and crucial effect on the balance. As pointed out above, if a FSW likes 
to perform (or get performed) unprotected oral sex, the balance is in favor of vaccination, 
independent of genital HPV status. In any case, cervical cancer screening is important for all 
FSWs (and former FSWs), whether vaccinated or not, and, as pointed out above, the risk of 
CIN2+ in possibly/probably pre-exposed women stays nearly the same in the first four years 
after vaccination compared to unvaccinated women. HPV vaccination, for whatever reason 
it  is  recommended  (e.g.  unprotected  oral  sex),  should  always  be  supplemented  by 
information concerning the importance of cervical screening. 
 
In summary, I agree with SCHIM VA1'(5/2())¶V paper that the recommendation of 
HPV vaccination to FSWs should be based on a balanced individual decision and not on a 
general pro- or contra-statement which applies to all FSWs. However, since I also consider 
the risks of unprotected oral sex for FSWs themselves as well as for their clients, I see the 
balance more often in  favor of vaccination.  But there is  also no doubt that the contra 
arguments the authors present in their paper show important limitations concerning the 
anogenital effectiveness of the vaccine in (possibly/probably/heavily) pre-exposed women 
like active FSWs; however, some of these limitations can be overcome if one considers 
cervical  HPV(16)-testing  before  vaccination  in  selected  individual  cases  as  part  of  the 
balanced decision. 
 
 
 
References 
 
 
ARBYN M et al., Cochrane Database Syst Rev. 2018; 5:CD009069.  
           doi: 10.1002/14651858.CD009069.pub3.  
BEACHLER DC et al. (2015), J Natl Cancer Inst. 2015; 108 (1). pii: djv302. doi:  
           10.1093/jnci/djv302 
BEACHLER DC et al. (2016), J Infect Dis. 2016; 213 (9): 1444 ± 1454 
BROWN B et al. (2011), Int J STD AIDS 2011; 22 (11) : 655 ± 658 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.
-31- 
 
 
BROWN B et al. (2012), Vaccine 2012; 30: 2309 ± 2314 
BROWN B et al. (2013), Int J STD AIDS 2013; 25 (7): 531 - 535 
BROUWER AF et al., Prev Med. 2019; 123: 288 ± 298 
CANADAS MP et al., J Clin Microbiol 2004; 42 (3): 1330 - 1332 
CASTELLSAGUE X et al., Br J Cancer 2011; 105: 28 ± 37 
CHANCELLOR JA et al., Community Dent Oral Epidemiol. 2016; 45 (1): 20 ± 34 
CHATTERJEE R et al., J Environ Pathol Toxicol Oncol 2001; 20 (2) : 113 ± 117 
CHATURVEDI AK et al., J Clin Oncol. 2018; 36 (3): 262 ± 267 
DROLET M et al., Lancet 2019; 394 (10197): 497 ± 509 
'¶628=$G. et al. (2014), PLOS ONE 2014; 9 (1): e86023 
'¶628=$*HWDO-,QIHFW'LV 1893 ± 1896 
FAIRLEY C et al., Abstract, 28th International Papillomavirus Conference 2012, San  
          Juan; Abstract Book Epidemiology/Public Health p. 249 
GILLISON ML et al., JAMA 2012; 307: 693 - 703 
GUPTA A et al., Oral Dis. 2019; 25 (2): 609 ± 616 
HEMMINKI K et al., Eur J Cancer Prev. 2000; 9: 433 ± 437 
HERRERO R et al., PloS ONE 2013: 8 (7): e68329. 
HIRTH JM et al., Vaccine 2017; 35: 3446 ± 3451 
HONG Y et al., BMC Public Health 2013; 13: 696 
KEPKA D et al., Prev Med Rep. 2019; 15:100957; doi: 10.1016/j.pmedr.2019.100957 
LOWY DR et al., Lancet Oncol 2015; 16: e226-e233 
MAREK E et al., Europ J Cancer Care 2014; 23: 65 ± 75 
MARRA E et al. (2017), Sex Transm Dis. 2017; 44 (12): 756 ± 762 
MARRA E et al. (2018), J Infect. 2018; 76 (4): 393 - 405 
MATSUSHITA K et al, Jpn J Infect Dis. 2011; 64 (1): 34 - 39 
McCLUNG NM et al., MMWR Morb Mortal Wkly Rep. 2019; 68(15): 337±343 
MEHANNA H et al.; Clin Infect Dis. 2018; doi: 10.1093/cid/ciy1081. 
OWEN BN e al.; AIDS Behav. 2019; doi: 10.1007/s10461-019-02477-w 
PARKER KH et al., Vaccine 2019; 37(21): 2864 ± 2869 
PELEMAN K, http://www.ghapro.be/nl/documents/GHAPRO_2015_jaarverslag.pdf, p. 21  
PINTO LA et al., J Infect Dis. 2016; 214 (8): 1276 ± 1283 
RKI (Robert Koch Institute) 2018a; Epid. Bulletin 26/2018: 233 ± 251 
RKI (Robert Koch Institute) 2018a; Epid. Bulletin 27/2018: 255 ± 259 
SHEWALE JB, GILLISON ML, Curr Opin Virol. 2019; 39: 33-40 
SILVERBERG MJ et al., Lancet Child Adolesc Health 2018; 2 (10): 707 ± 714 
SOOHOO M et al., Open AIDS J 2013; 7: 58 - 66 
SONAWANE K et al., Ann Intern Med. 2017; 167(10):714-724 
TOUZE A et al., J. Clin Microbiol. 2001; 39 (12): 4344-4348 
VINK A et al., Abstract (P-04.17), 27th International Papilloma Virus Conference 2011, 
Berlin   
WADHERA P., Int J STD AIDS 2015; 26: 893 ± 902 
WHEELER CM et al., Lancer Infect Dis. 2016; 16 (10): 1154 ± 1168 
WILKIN TJ et al., Clin Infect Dis. 2018; 67 (9): 1339 - 1346  
 
 
 
 
Eingestellt über www.PDF-ins-Internet.de - Haftung für Inhalt und Inhaber aller Rechte ist der Puplisher
Kontaktdaten und Anbieterkennung des Puplishers/Autors entnehmen Sie bitte dem PDF-Archives auf www.PDF-ins-Internet.de.

Comment on SCHIM VAN DER LOEFF et al. 2019 – Should female sex workers be offered HPV vaccinationffam

PDF & Publisher Info