From the Departments of aEpidemiology and bOncology, McGill University, Montreal, QB; cDepartment of Epidemiology, University of California, Los Angeles, CA; and dIFC-National Research Council, Pisa, Italy.
Editors’ note: Related articles appear on pages 635 and 638.
Correspondence: Eduardo L. Franco, Division of Cancer Epidemiology, McGill University, 546 Pine Ave West, Montreal, QC, Canada H2W 1S6. E-mail: email@example.com.
The Nobel Prize Committee awarded last year's Physiology and Medicine prize to 3 scientists who identified the causes of 2 diseases that kill millions of people worldwide: HIV/AIDS and cancer of the cervix. Harald zur Hausen received half of the prize “for his discovery of human papilloma viruses causing cervical cancer” and Françoise Barré-Sinoussi and Luc Montagnier shared the other half of the award “for their discovery of human immunodeficiency virus.”1
The knowledge that stemmed directly or indirectly from the vigorous research by these 3 outstanding scientists will prevent the countless premature deaths caused by these diseases. As epidemiologists working in public health on the global scene, we applaud the Nobel Committee's decision. We hasten to add, however, that in this era in which advances in medical knowledge can come only from strong multidisciplinary approaches, the laurels of recognition to those who initiate or sustain the basic science discoveries are grounded on a foundation of public health science. Our arguments, summarized later, are self-serving. The 4 of us are card-carrying epidemiologists who have witnessed how the monumental work of the 3 Nobel laureates was validated and brought to the attention of the public and the pharmaceutical and biotechnology sector by a massive public health research undertaking championed by epidemiologists.
That the Nobel Committee has chosen not to add the names of such champions is more or less understandable. The Committee has a track record of rewarding primarily the forebears of the scientific domains that led to the medical advances. There are many candidates competing for science's highest accolade, and there is only one Nobel Prize given each year in Medicine. Packaging 2 diseases into the 2008 award may have helped the Nobel Committee deal with the backlog of eminent scientists who must be recognized with as little delay as possible (the rules of the award state that the prize cannot be given posthumously). Without questioning the Nobel's modus operandi (the very glamour of the prize rests in the organization's aura of tradition and independence in reaching its decisions), we feel that the readers of Epidemiology will find comfort in being reminded of the contribution of their discipline in the accomplishments that led to a Nobel recognition in 2008, and that ultimately enabled the Nobel-Prize-winners’ work to have impact.
Over history, epidemiologic studies have made many of the observations that were necessary for getting the right hypotheses to bear fruit in terms of health interventions. Robert Koch was given the Nobel Prize in 1905 for his work in tuberculosis, but he could also have received the prize for his isolation of the Vibrio cholerae. Had the Nobel Committee awarded the prize for cholera, according to their current practice they would have left out John Snow who elegantly demonstrated where the infectious agent came from, how it spread and how it could be prevented. One of the lines of inquiry of the last century with most influence on public health intervention was the demonstration of the carcinogenic role of tobacco smoking. The list of pioneer epidemiologists who championed this research includes such well-known names as Doll, Hill, Peto, Wynder, and Graham, as well as others no less illustrious. Each of these epidemiologists received substantial recognition from the research community, but not a Nobel Prize.
Does the Nobel Committee focus on the elucidation of mechanisms and on downstream, proximal causes rather than upstream determinants? This is the perception that comes across to the epidemiology community. From a broader professional viewpoint, the discoveries of the agents that cause AIDS and cervical cancer set in motion a cascade of other basic science discoveries in molecular biology, pathology, and immunology. Such advances augmented the body of evidence contributed by epidemiologists to form the knowledge that was eventually translated into successful health interventions, and to useful insights that now propel other scientific lines of inquiry.
The role of epidemiology in helping to establish HPV infection as the cause of cervical cancer is a virtual anthology of hard-learned lessons for our discipline. Every year, cervical cancer affects half a million women worldwide, about 80% of whom live in extreme poverty in sub-Saharan Africa, Latin America, and Southeast Asia. Not uncommonly, women with cervical cancer die before the age of 45 years, and they have a proportionally higher number of children than other women of the same age. A large body of epidemiologic work that began in the 1950s2 clearly demonstrated that the disease was related to sexual practices, which pointed the way to the search for a sexually transmitted agent. On the basis of these epidemiologic observations, zur Hausen postulated in the 1970s3 a role for HPV in cervical cancer, primarily on the basis that HPV DNA could be found in most of cervical cancer biopsies.
Already in the early 1970s epidemiologists had suggested infections as the possible cause of cervical cancer,4,5 but 10 years later there was much skepticism concerning the role of HPV infection in cervical cancer. Part of the problem came from the staying power of the Herpes hypothesis, which seemed to satisfy the quest for a sexually transmitted agent.6 Also, early molecular epidemiologic studies that attempted to test the causal role of HPV failed because they used imperfect tools to detect HPV DNA. In the mid to late 1980s, much of what could then be called molecular epidemiologic evidence was largely incoherent and cast doubt on the notion that HPV was the likely culprit. Because HPV infection status was often misclassified in the early studies (due to inaccurate hybridization techniques), (1) the magnitude of the association between HPV and cervical cancer was weak, (2) HPV was only weakly associated with sexual practices, and (3) when the association between sexual activity and disease risk was controlled for the presence of HPV, the relation did not disappear.7,8 Adding to the confusion and fuelling skepticism was the finding that HPV infection seemed to affect most adults.9 This finding was later retracted.10 Embarrassingly enough, these problematic findings came to light at the same time that molecular biologists obtained strong proof of concept for a carcinogenic role for HPV, showing that some viral oncogenes interfered directly with the genes that control cellular proliferation.11,12
Luckily, epidemiologists quickly regrouped following these initial observations and gained experience with the novel molecular genetic tools that were emerging in the early 1990s, such as the polymerase-chain-reaction and signal-amplification techniques. This allowed the epidemiologic community to turn the page on the earlier incoherent findings due to measurement error, and to recognize the full extent of the causal relations at play.13–15 From that point on, progress came quickly; case-control and cohort studies using modern laboratory techniques demonstrated that infection by certain genotypes of HPV is one the strongest cancer risk factors ever found.16–19 Subsequent work also produced precise estimates of relative risks that defined the HPV genotypes that had to be targeted for prevention.20 In 1995, HPVs 16 and 18 were classified as group 1 carcinogens by an expert panel convened by the International Agency for Research on Cancer,21 on which several epidemiologists served.
This classification had an enormous impact in subsequent policy. The biotechnology industry began to develop HPV tests to improve the accuracy of cervical cancer screening. Also emboldened by the emerging consensus, 2 pharmaceutical companies, Merck and GlaxoSmithKline (known then as SmithKline Biologicals) began the massive research and development that, hundreds of millions of dollars and 10 years later, led to the licensing of the first prophylactic vaccines against HPV and the diseases that it causes.
Did epidemiologists help or did they delay the scientific process that led to today's novel HPV-based interventions to prevent cervical cancer? This debate took partisan overtones in 2001 in an exchange between zur Hausen and some of us. In his opinion (presumably shared by a large number of basic scientists), the knowledge that could have propelled successful biomedical interventions was already available in the late 1980s.22 In our opinion, “in the late 1980s it seemed prudent … to wait until consistent epidemiologic data supported the inferences from molecular biologic studies.”23 Clearly, the standard for proving a causal role for HPV in cervical cancer became one of the most stringent in all of cancer prevention.
In hindsight, the early missteps of epidemiology could not have been avoided. What is remarkable, however, was how fast the lessons were learned. This led to a vigorous multidisciplinary collaboration to solve the puzzle that brought us today's cervical cancer control interventions and strengthened the case for the meritorious Nobel Prize to Professor zur Hausen. He, more than most basic scientists, now appreciates the value of epidemiology in inspiring discoveries and in providing the translational roadmap for successful health interventions. In fact the HPV-cervical cancer story underscores the importance of multiple disciplines in recognition of disease pathogenesis. In this process epidemiologic observations often “start the ball rolling!”
Epidemiologists can take pride in the fact that they identified the main upstream causes of cervical cancer at an early stage. A causal conclusion concerning the role of HPV could not have come at an earlier stage without violating the discipline's most basic tenets in establishing etiologic relations. Kirkegard said that life can only be understood in retrospect but you have to live life forward in time. In hindsight, HPV could have been classified as a carcinogen at an earlier stage, but such an early decision would have come at the expense of suppressing the emergence of a body of epidemiologic evidence that now serves as the foundation for promising cervical cancer prevention technologies.
2. Wynder EL, Cornfield J, Schroff PD, Doraiswami KR. A study of environmental factors in carcinoma of the cervix. Am J Obstet Gynecol
3. zur Hausen H. Human papilloma viruses and their possible role in squamous cell carcinomas. Curr Top Microbiol Immunol.
4. Rotkin ID. A comparison review of key epidemiological studies in cervical cancer related to current searches for transmissible agents. Cancer Res
5. Beral V. Cancer of the cervix: a sexually transmitted infection? Lancet
6. Franco EL. Viral etiology of cervical cancer: a critique of the evidence. Rev Infect Dis
7. Franco EL. The sexually transmitted disease model for cervical cancer: incoherent epidemiologic findings and the role of misclassification of human papillomavirus infection. Epidemiology
8. Schiffman MH, Schatzkin A. Test reliability is critically important to molecular epidemiology: an example from studies of human papillomavirus infection and cervical neoplasia. Cancer Res
9. Tidy JA, Parry GC, Ward P, et al. High rate of human papillomavirus type 16 infection in cytologically normal cervices. Lancet.
10. Tidy J, Farrell PJ. Retraction: human papillomavirus subtype 16b. Lancet.
11. Dyson N, Howley PM, Münger K, Harlow E. The human papilloma virus-16 E7 oncoprotein is able to bind to the retinoblastoma gene product. Science
12. Scheffner M, Werness BA, Huibregtse JM, Levine AJ, Howley PM. The E6 oncoprotein encoded by human papillomavirus types 16 and 18 promotes the degradation of p53. Cell
13. Muñoz N, Bosch FX, de Sanjosé S, et al. The causal link between human papillomavirus and invasive cervical cancer: a population-based case-control study in Colombia and Spain. Int J Cancer
14. Koutsky LA, Holmes KK, Critchlow CW, et al. A cohort study of the risk of cervical intraepithelial neoplasia grade 2 or 3 in relation to papillomavirus infection. N Engl J Med
15. Schiffman MH, Bauer HM, Hoover RN, et al. Epidemiologic evidence showing that human papillomavirus infection causes most cervical intraepithelial neoplasia. J Natl Cancer Inst
16. Muñoz N, Bosch FX, de Sanjosé S, Shah KV. The role of HPV in the etiology of cervical cancer. Mutat Res
17. Kjaer SK, van den Brule AJ, Bock JE, et al. Human papillomavirus–the most significant risk determinant of cervical intraepithelial neoplasia. Int J Cancer
18. Liaw KL, Glass AG, Manos MM, et al. Detection of human papillomavirus DNA in cytologically normal women and subsequent cervical squamous intraepithelial lesions. J Natl Cancer Inst
19. Herrero R, Hildesheim A, Bratti C, et al. Population-based study of human papillomavirus infection and cervical neoplasia in rural Costa Rica. J Natl Cancer Inst
20. Muñoz N, Bosch FX, de Sanjosé S, et al. International Agency for Research on Cancer Multicenter Cervical Cancer Study Group. Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med
21. IARC Working Group. Human Papillomaviruses. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans
. Vol. 64. Lyon: International Agency for Research on Cancer; 1995.
22. zur Hausen H. Cervical carcinoma and human papillomavirus: on the road to preventing a major human cancer. J Natl Cancer Inst.
23. Bosch FX, Muñoz N, de Sanjosé S, et al. Re: Cervical carcinoma and human papillomavirus: on the road to preventing a major human cancer. J Natl Cancer Inst