From the Dipartimento di Epidemiologia, Istituto Nazionale per le Malattie Infettive ‘Lazzaro Spallanzani', Rome, aServizio di Epidemiologia, Centro di Riferimento Oncologico, Aviano, bIstituto di Ricerche Farmacologiche ‘Mario Negri’ and Istituto di Statistica e Biometria, Università degli Studi, Milan, Italy, and cField and Intervention Study Unit, International Agency for Research on Cancer, Lyon, France.
Requests for reprints to: D. Serraino, Istituto Nazionale per le Malattie Infettive ‘Lazzaro Spallanzani', I.R.C.C.S, via Portuense, 292, 00142 Roma, Italy.
Received: 27 July 2001;
revised: 3 December 2001; accepted: 13 December 2001.
Sponsorship: Supported by the Ministero della Sanità, Istituto Superiore di Sanità, Roma, through the III National AIDS Research Project grant number 20C.15 and grant number 20C/1.1
Invasive cervical cancer (ICC) was the third and last malignancy, after Kaposi's sarcoma (KS) and non-Hodgkin's lymphomas (NHL), included in the list of AIDS-defining illnesses (ADI) in 1993 . At that time, an association between cervical neoplasia and HIV infection had only started to emerge and it seemed to be restricted to non-invasive cervical tumours [2–5]. An excess of invasive and pre-invasive cervical neoplasias among HIV-infected women was later confirmed in Europe and the USA [6–9]. Such excess was, however, much weaker than that found for KS and NHL (i.e., an increase of 5- to-10-fold, compared with approximately 1000-fold and 100-fold for KS and NHL respectively) [5,10].
During the last decade, an increasing proportion of AIDS cases has been registered among women, as compared with men [11,12]. In an analysis of AIDS surveillance data on women with AIDS reported to the Italian AIDS Registry between 1993 and 1995, the frequency of ICC was nearly threefold higher among injecting drug users (IDU) than among those infected through heterosexual contacts , suggesting that sexual behaviour (e.g., prostitution) and/or lack of appropriate cervical cancer screening may favour the occurrence of ICC in this risk category.
In this paper, we took advantage of AIDS surveillance data in Europe between 1993 and 1999 in order to elucidate the prevalence and major correlates of ICC as an ADI.
The 30 June 2000 update of the European Non Aggregate AIDS Data Set (ENAADS) was analysed . Information regarding ENAADS, which contains anonymous information for all individuals whose AIDS diagnosis fitted the 1993 European AIDS-Surveillance case definition , and its limitations has been already published [10,11]. Briefly, 38 countries of the World Health Organization European Region provided standard individual information on AIDS cases to the European Centre for the Epidemiological Monitoring of AIDS, Paris, France. For the sake of statistical analysis, i.e., to avoid extreme random effects, only countries with at least 50 cases of AIDS reported in women 20–49 years of age were considered. Thus, the present investigation was based on data from 15 countries (i.e., Denmark, Ireland, Norway, Sweden and the UK in northern Europe; Austria, Belgium, France, Germany, the Netherlands and Switzerland in central Europe; Greece, Italy, Portugal and Spain in southern Europe). The analysis focused on women diagnosed with AIDS between January 1993 and December 1999. To reduce the effect of incomplete notification and reporting delay AIDS cases diagnosed between January and June 2000 were not considered.
A woman was considered to be affected by ICC if such neoplasm was listed as one of the four ADI that can be reported. HIV transmission categories were hierarchically classified as: (i) IDU, when the use of injecting drugs was mentioned; (ii) heterosexual with an IDU partner, when HIV was reported to have been acquired through heterosexual intercourse with an IDU; (iii) heterosexual originating from a HIV-epidemic country; (iv) heterosexual with a non-IDU partner; (v) other, when the woman acquired HIV infection through exposure to contaminated blood or blood products or when the mode of HIV infection acquisition was not ascertained.
Country-specific ICC incidence rates for 1995 in women aged 20–49 years, standardized on the world standard population, were derived from Ferlay et al.  for the same 15 European countries. In Austria, France, Germany, Italy, Spain and Switzerland the cancer registration system did not cover the whole population. In each of these countries, a national estimate of ICC incidence rates was obtained by pooling together information from the various regional cancer registries.
The association between selected characteristics and the presence of ICC as an ADI was assessed by means of odds ratios (OR) and 95% confidence intervals (CI), calculated by means of unconditional logistic regression equations . OR were adjusted for age group, geographic area, year of AIDS diagnosis, and HIV-transmission category.
Correlations between ICC country-specific incidence rates and selected AIDS-related variables (i.e., the proportion of ICC as an AIDS-defining illness; the proportion of IDU women among AIDS cases; and the male : female ratio in IDU as a proxy of the potential number of sexual partners of IDU women within the IDU community) were estimated using the Pearson correlation coefficient (r) weighted for the number of AIDS cases in each country.
Overall, ICC was present among the ADI of 576 (2.5%) of the 23 054 women aged 20–49 years diagnosed with AIDS between 1993 and 1999 in the 15 European countries (Table 1). ICC was thus more common than KS (455 cases, 2.0%), although less so than NHL (676 cases, 2.9%; data not shown). ICC as an ADI was found in only 20 women aged 50 years or more (1.1% of AIDS cases; data not shown).
Women aged 30 years or more showed a significantly increased ICC risk (crude OR, 1.4) than younger women (Table 1). Compared to women from northern Europe (where 0.8% had ICC as an ADI), women from central (2.2%) and southern Europe (3.0%) had significantly elevated risks of ICC (crude OR, 2.7 and 3.8, respectively;Table 1).
There was a modest increase in the proportion of ICC after 1995 (OR, 1.2;Table 1). Compared to heterosexual women with a non-IDU partner, IDU women had an OR of ICC as an ADI of 1.6, whereas women who acquired HIV infection through blood or through other unspecified routes showed an OR of 0.5. No increased risk of ICC was found among women who reported an IDU partner and those who originated from HIV-endemic countries. The inclusion of all variables in Table 1 in a multivariate analysis did not modify substantially the strength of the associations considered, except for some reduction in the OR for southern Europe (from 3.8 to 3.1;Table 1).
The apparent role of IDU was consistent in the three geographical areas: IDU women showed a more elevated prevalence of ICC than heterosexuals in northern (OR, 2.4; 95% CI, 0.6–10.3), central (OR, 1.9; 95% CI, 1.2–3.0), and southern Europe (OR, 1.3; 95% CI, 1.0–1.8). Similarly, the reduced risk associated with acquisition of HIV infection through blood or other unspecified routes was geographically consistent (data not shown).
Fig. 1 shows the correlation between the prevalence of ICC and incidence rates of ICC in the female population age 20–49 years of the 15 European countries considered. Several countries with low (e.g., Spain or Belgium) or intermediate (e.g., France or Italy) incidence rates of ICC in the general female population showed higher percentages of AIDS-associated ICC than countries with elevated rates (e.g., Denmark and Germany). Overall, ICC incidence rates showed a negative correlation (r, −0.59;P = 0.02) with the prevalence of ICC as an ADI (Fig. 1).
To evaluate further the relationship between ICC and injecting drug use, the proportion of ICC in each country was correlated with the percentage of IDU among female AIDS cases (Fig. 2a), or with the male : female ratio among IDU with AIDS, as a proxy of sexual promiscuity of IDU women (Fig. 2b). Elevated correlation coefficients emerged for both variables (r, 0.70 and r, 0.74, respectively) .
Since the inclusion of ICC among AIDS-defining illnesses in 1993, 2.5% of HIV-infected women age 20–49 years in 15 major European countries presented with this neoplasm as an ADI. The prevalence of ICC as an ADI was inversely correlated with the incidence of ICC in the general female population of the major European countries examined. These findings suggest that HIV-infected women represent a separate group whose risk level does not reflect the country-specific background risk of ICC, but, rather, the specific characteristics of the AIDS epidemic (i.e., the predominance of IDU among AIDS patients) and the efficacy of cervical cancer screening programs. The increased prevalence of ICC among HIV-infected women is, in fact, chiefly determined by two factors, i.e., increased risk of acquisition and persistence of human papillomavirus infection (i.e., the presumed cause of cervical tumours) , which has been shown to be more common among IDU , and the inadequacy of cytological screening programs to interrupt the progression from pre-invasive to invasive cervical neoplasias . As far we know, the pattern of utilization of Papanicolaou smear has not been expressly studied among IDU women, although evidence exists that when cervical cancer is detected in IDU women who were not known to be HIV-positive, it is generally an advanced disease .
A two- to-sevenfold higher risk of ICC in IDU women than in other HIV transmission categories has already been shown in Italy and the USA [6–8]. Although specific information on sexual habits is not available in AIDS surveillance data, it is worth noting that an elevated male : female ratio among IDU with AIDS (as a proxy of the potential number of sexual partners of IDU women within the IDU community ) was correlated with an increased prevalence of ICC at AIDS diagnosis. The association with the area of living in Europe was, however, stronger than the one with HIV transmission group, pointing to the importance of the ‘screening’ factor. The first organized nationwide cytological screening programs in Europe were established in the 1970s in Nordic countries, followed by England and Wales in the 1980s [20,21]. It was only in the mid 1990s that population-based screening programmes gained grounds in Southern and Central European countries, too .
AIDS surveillance data have important limitations [10,11], including different degrees of completeness of AIDS surveillance in different European countries  and, specific to ICC, variations in the frequency of gynaecological and cytological examinations of HIV-infected women. Furthermore, AIDS surveillance data do not include information of pre-invasive cervical neoplasias nor on ICC that may have occurred after AIDS diagnosis. However, after the adoption in 1993 of an AIDS definition in the USA based on the CD4 T cell count, European countries offer the only opportunity to monitor AIDS presentation patterns, including AIDS-associated cancers, at population level.
In conclusion, the prevalence of ICC as AIDS-defining illness seems to be, in Europe, a marker of two major features: the predominance of IDU among AIDS patients and the failure of cervical screening programs in southern and central Europe to reach under-privileged groups of women. In the light of the increase in life expectancy among HIV-infected women and the good curability of early cervical tumours, specific efforts should be made to offer routinely cervical screening tests to women, especially IDU, at first detection of HIV infection and at follow-up examinations.
The authors thank The European Centre for the Epidemiological Monitoring of AIDS, Paris, France and I. Calderan and L. Mei for editorial assistance.
1. Ancelle-Park RA. Expanded European AIDS case definition [letter]. Lancet 1993, 341: 441.441.
2. Maiman M, Fructher RG, Guy L, Cuthill S, Levine P, Serur E. Human immunodeficiency virus infection and invasive cervical carcinoma. Cancer 1993, 71: 402–406.
3. Maiman M, Fructher RG, Serur E, Remy JC, Feuer G, Boyce J. Human immunodeficiency virus infection and cervical neoplasia. Gynecol Oncol 1990, 38: 377–382.
4. International Agency for Research on Cancer. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans
. Vol. 64, Human Papillomaviruses
. Lyon: IARC; 1995.
5. International Agency for Research on Cancer. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Vol. 67, Human Immunodeficiency Viruses and Human T-Cell Lymph tropic Viruses.
Lyon: IARC; 1996.
6. Franceschi S, Dal Maso L, Arniani S. et al
. Risk of cancer other than Kaposi's sarcoma and non-Hodgkin's lymphoma in persons with AIDS in Italy. Br J Cancer 1998, 78: 966–970.
7. Serraino D, Carrieri MP, Pradier C. et al
. Incidence of invasive cervical cancer among women with, or at risk for, HIV infection. Int J Cancer 1999, 82: 334–337.
8. Frisch M, Biggar RJ, Goedert JJ. Human papillomavirus-associated cancers in patients with human immunodeficiency virus infection and acquired immunodeficiency syndrome. J Natl Cancer Inst 2000, 92: 1500–1510.
9. Ahdieh L, Muñoz A, Vlahov D, Trimble CL, Timpson LA, Shah K. Cervical neoplasia and repeated positivity of human papillomavirus infection in human immunodeficiency virus-seropositive and -seronegative women. Am J Epidemiol 2000, 151: 1148–1157.
10. Dal Maso L, Serraino D, Franceschi S. Epidemiology of HIV-associated malignancies in developed and developing countries. Eur J Cancer 2001, 37: 1188–1201.
11. Dal Maso L, Lo Re A, Franceschi S, La Vecchia C. AIDS incidence rates in Europe and in the USA, 1990–94. J Epidemiol Biostat 1997, 2: 81–86.
12. European Centre for the Epidemiological Monitoring of AIDS. HIV/AIDS Surveillance in Europe. Mid-year Report, number 63.
Paris: European Centre for the Epidemiological Monitoring of AIDS; 2000.
13. Serraino D, Napoli PA, Zaccarelli M, Alliegro MB, Pezzotti P, Rezza G. High frequency of invasive cervical cancer among female injecting drug users with AIDS in Italy [letter]. AIDS, 1996, 10: 1041–1042.
14. Ferlay J, Bray F, Sankila R, Parkin MD. EUCAN: Cancer Incidence, Mortality and Prevalence in the European Union. 0, version 0.0. IARC CancerBase No. 4
. Lyon: IARC Press; 1999.
15. Breslow NE, Day NE. Statistical methods in cancer research. Vol. I. The analysis of case-control studies. IARC Scientific Publication number 32.
Lyon: IARC; 1980.
16. Walboomers JMM, Jacobs MV, Manos MM. et al
. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999, 189: 12–19.
17. Rezza G, Giuliani M, Serraino D. et al
. Risk factors for cervical presence of human papillomavirus DNA among women at risk for HIV infection. Epidemiol Infect 1998, 121: 173–177.
18. Maiman M, Fruchter RG, Clark M, Arrastia CD, Matthews R, Gates EJ. Cervical cancer as an AIDS-defining illness. Obstet Gynecol 1997, 89: 76–80.
19. Booth RE, Kwiatkowski CF, Chitwood DD. Sex related HIV risk behaviors: differential risks among injection drug users, crack smokers, and injection drug users who smoke crack. Drug Alcohol Depend 2000, 58: 219–226.
20. Van Ballegooijen A, van den Akker-van Marle E, Patnick J. et al
. Overview of important cervical cancer screening process values in European Union (EU) countries, and tentative predictions of the corresponding effectiveness and cost-effectiveness. Eur J Cancer 2000, 36: 2177–2188.
21. Franceschi S, Herrero R, La Vecchia C. Cervical cancer screening in Europe: what next? Eur J Cancer 2000, 36: 2272–2275.
22. Ajdacic-Gross V, Zellweger U, Wang J, Fleerackers Y, Somaini B. How complete is AIDS surveillance in Europe? An eagle eye comparison with mortality data. J Epidemiol Comm Health 2001, 55: 52–56.
© 2002 Lippincott Williams & Wilkins, Inc.