Updated: 16 Sept 2011
by: Charlotte Haug. Editor-in-chief of The Journal of the Norwegian Medical Association
A shorter version of this article was published in New Scientist magazine on 1 October 2011
We need to talk about HPV vaccination – seriously
Should we vaccinate girls against HPV? Michele Bachmann's blundering aside, there is no clear answer
Whenever a new medical intervention comes onto the scene, doctors have hard choices to make. Do we really know enough about its potential benefits and risks to recommend using it? If in doubt, should we err on the side of caution or on the side of hope?
These are difficult questions because medical knowledge is usually incomplete and ambiguous. They are especially difficult to answer for drugs that may prevent disease in the future, particularly when those drugs are given to otherwise healthy people. Vaccines, particularly the human papillomavirus (HPV) vaccine, are a case in point.
The HPV vaccine is designed to prevent cervical cancer, which currently kills around 250,000 women a year worldwide. It works by inducing immunity to HPV, the leading cause of the disease and the world’s most common sexually transmitted infection: around 79 per cent of people catch it at some point in their lives. Almost all infections are quickly cleared by the immune system, but in a few women infection persists and can progress to precancerous lesions and eventually cervical cancer.
The first HPV vaccine came on the market in 2006 and public health officials in the US, Canada, Australia and several European countries now recommend vaccination of preadolescent girls.
But despite claims that the vaccine will cut cancer deaths by two-thirds or more, its overall effectiveness in the prevention of cervical cancer remains unknown and will not be known for decades.
The theory behind the vaccine is sound: if HPV infection can be prevented, cervical cancer will not occur. But in practice the issue is more complex.
Why? First, there are more than 100 different types of HPV, at least 15 of which cause cancer. But the vaccine protects only against the two most important cancer-causing strains, HPV-16 and HPV-18, though it may also offer partial protection against some closely related strains (The Lancet, vol 374, p 301). But the remaining strains still cause cancer.
Second, though clinical trials have shown that the vaccine reduces the incidence of precancerous lesions, we cannot say for sure that this will translate into cutting cervical cancer and deaths 20 to 40 years in the future.
There are many other gaps in our knowledge. How long does the vaccine provide protection without a booster? Does it affect natural immunity against HPV, and with what consequences? Can we really be sure that the vaccine protects preadolescent girls when proper clinical trials have been carried out only in women aged 16 to 24?
Another critical question is what vaccination does to the uptake of cervical screening. As the vaccines protect against only some of the cancer-causing strains, women must continue cervical screening. But vaccinated women may feel protected and therefore be less likely to go for screening.
Resolving these essential questions will require decades of observation of large numbers of women.
There is another serious question that may be answered sooner: what effect will the vaccine have on the other cancer-causing strains of HPV? Nature never leaves a void, so if HPV-16 and HPV-18 are suppressed by an effective vaccine, other strains of the virus will take their place. The question is, will these strains cause cervical cancer?
Results from clinical trials are not encouraging. Vaccinated women show an increased number of precancerous lesions caused by strains of HPV other than HPV-16 and HPV-18. The results are not statistically significant, but if the trend is real – and further clinical trials should tell us in a few years – there is reason for serious concern.
Yet another worry is side effects. In the US, the Vaccine Adverse Events Reporting System (VAERS) has received many reports of adverse effects following HPV vaccination. The vast majority are mild, such as headaches and nausea, but there are some reports of severe reactions including anaphylactic shock, Guillain-Barré syndrome, pancreatitis and thrombosis. Of 12,424 adverse reactions reported between June 2006 and December 2008, 32 were fatal. About 23 million girls received a vaccination in that period. Overall, the rate of adverse effects is 54 per 100,000 doses, about the same as for other vaccines (The Journal of the American Medical Association, vol 302, p 750).
VAERS is a voluntary reporting system with serious limitations – it is impossible, for example, to be sure that the reported effects were actually caused by the vaccine. Only systematic studies will be able to distinguish the true extent of side effects of the vaccine.
There has been pressure on policy-makers worldwide to introduce the HPV vaccine. But how can they make rational choices about a medical intervention that might do good in the distant future, but might also do harm, with no prospect of an answer for decades? How should parents, doctors or anyone else decide whether it is a good thing to give a young girl the vaccine?
One way forward is to build a mathematical model of the disease and use it to test the benefits of vaccination. Doing so is extremely complex, however. The model has to factor in the natural history of HPV infection, the effect of the vaccine over a lifetime, the effect on other HPV strains, the effect of the vaccine on natural immunity against HPV, the sexual behaviour of the girls and women and their partners, and finally, cervical-cancer screening practices.
The modelling has been done (The New England Journal of Medicine, vol 359, p 821) and suggests that vaccinating 12-year-old girls is worth it. However, the model includes a number of highly optimistic assumptions. Among these are that the vaccine offers lifelong protection with no need for a booster, that it has the same effect on young girls as older women, that HPV-16 and HPV-18 are not replaced by other cancer-causing strains, that vaccinated women continue getting screened for cervical cancer, and that natural immunity against HPV is unaffected.
Are these assumptions reasonable? We do not know. What we do know is that if they are not, vaccination is less effective and may even be less effective than screening alone. And what if the vaccine has long-term side effects?
With so many essential questions unanswered, there is good reason to be cautious about introducing large-scale HPV vaccination. Instead, we should concentrate on getting answers.
One thing’s for sure though. Those answers are unlikely to emerge from the likes of Michele Bachmann.
Source: New Scientist