Infectious Agents and Cancer

“[H]uman papilloma virus, hepatitis B virus, hepatitis C virus, Epstein-Barr virus, human herpes virus 8, human T-cell lymphotropic virus 1, human immunodeficiency virus, Merkel cell polyomavirus, Helicobacter pylori, Opisthorchis viverrini, Clonorchis sinensis, Schistosoma haematobium […] are recognized as carcinogens and probable carcinogens by [the] International Agency for Research on Cancer (IARC). They are not considered in this book […] The aim of this monograph is to analyze associations of other infectious agents with cancer risk […] virology is not considered in our monograph: although there are some viruses that can be connected with cancer but are not included into the IARC list (John Cunningham virus, herpes simplex virus-1 and -2, human cytomegalovirus, simian virus 40, xenotropic murine leukemia virus-related virus), we decided to leave them for the virologists and to concentrate our efforts on other infectious agents (bacteria, protozoa, helminths and fungi) […] To the best of our knowledge, this is the first book devoted to this problem”

…

Here’s what I wrote on goodreads:

“This book is written by three Russian researchers, and you can tell; the language is occasionally hilariously bad, but it’s not too difficult to figure out what they’re trying to say. The content partially made up for the poor language, as the book covers quite a bit of ground considering the low page count.”

I gave the book two stars. I’m glad they wrote the book, because it covered some stuff I didn’t know much about. I think I’m closer to one star than three, but it’s mostly because it’s terribly written, not because I have major objections to the coverage as such. What I mean by this is that they talk about a lot of studies and they include a lot of data – they’re scientists who write about scientific research, they just happen to be Russian scientists who are not very good at English. It’s terribly written, but the stuff is interesting.

…

As mentioned above there are quite a few viruses which we know may lead to cancer in humans. I’ve recently read a lot of stuff about this topic as it was covered in both Boffetta at el. and also rather extensively in part 5 of the Sexually Transmitted Diseases text, which covered sexually transmitted viral pathogens (that section of the book was with its 230 pages actually a ‘book-length section’; it was significantly longer than this book is..). I’ve even covered some of that stuff here on the blog, e.g. here. I may incidentally write more about these things and related stuff later, as I’m quite far behind in terms of my intended coverage of the STD book at the moment.

Anyway, viruses aren’t the only bad guys around. So these guys decided to write a book about some other infectious diseases affecting humans, and how these infectious diseases may relate to cancer risk. As they point out in the book, “there is only one bacterium, Helicobacter pylori, which is recognized by IARC as an established human carcinogen.” After reading this book you’ll realize that there are some others which perhaps look a bit suspicious as well. In some cases a lot of studies have been done and you have both animal-models, lab-analyses, case-control studies, cohort studies, … In other cases you have just a few small studies to judge from. As is always the case when people have a close look at epidemiological research, this stuff is messy. Sometimes studies that looked really convincing turn out to not replicate in larger samples, sometimes dramatically different effect sizes are found in different areas of the world (which may of course both be interpreted as an indicator that the ‘true’ effect sizes are different in the different subpopulations, or it may be interpreted as a result e.g. of faulty study design which makes those Swedish data look really fishy..), sometimes different results can be explained by differences in data quality/type of data applied/etc. (classic cases are different effects based on whether you rely on self reports or biological disease markers, and different results from analyses of bacterial cultures vs PCR analyses), and so on and so forth. There are a lot of details, and they cover them in the book. I occasionally see people criticize epidemiological research online on the grounds that many (‘all?’) results published in this area are just random correlations without any deeper meaning. Sometimes this criticism may well be warranted, and the authors of this book certainly in some cases seem to go quite a bit further than I would do based on the same data. But there’s another part of the story here. When you start out with a couple of case-control studies indicating that guys with cancer type X are more likely to have positive lab cultures for this specific micro-organism, that may not be a big deal. But perhaps then a few microbiologists show up and tell you that it would actually make a lot of sense if there was a connection here (and they might start talking about fancy stuff like various ‘modulations of host immune responses’, ‘inflammatory markers’, ‘the role of nitric oxides’, …). They conduct some studies as well and perhaps one of the things they find is that the observed cancer grades in the patients seem to depend quite a lot upon which of the pathogen subtypes the individual happen to be infected with (perhaps suddenly also providing an explanation for some previously surprising negative results in specific cases). And then perhaps you get a couple of animal studies that show that these animals get cancer when you infect them with these bugs and don’t treat the infection. Perhaps you have a few more studies as well in different populations, because Chinese people get cancer too, and you start seeing that people around the world who happen to be infected with these bugs are all more likely to get cancer, compared to the locals who are not infected (…or perhaps not, and then it just gets more fun…). This process goes on for a while, until at some point it starts getting really hard to think these positive correlations are all just the result of random p-value hunting done by bored researchers who don’t know what else to do with their time, and you start asking yourself if perhaps this idea is not as stupid as it was when you first encountered it. Most of the time the process stops before then because the proposed link isn’t there, but modern epidemiology is not just random collections of correlations.

In the context of the specific infectious diseases covered in the book the people who have in some sense the final say in these things (the IARC) think we’re not quite there yet, but you have some cases where some different lines of evidence all seem to indicate that a link may be present and relevant. It would be highly surprising to me if in 20 years time we’d have realized that none of the infectious diseases they talk about in this book are at all involved in cancer pathogenesis. A related point is that most likely we’ve missed some ‘true connections’ along the way, and will continue to do so in the future, because even if a link is there, it’s sometimes really hard to find it and easy to overlook it, for many different reasons.

I have quoted a bit from the book below and added some comments here and there. I have corrected some of the spelling/language errors the authors made to ease reading; if a word is placed in brackets, it’s an indicator that I’ve replaced a misspelled word by the correct one (‘they meant to use’). The authors do not even have any clue how and when to use the word ‘the’, often using it when it’s not needed and forgetting to use it in cases where it is needed, which made quoting from the book painful. Read it for the content.

…

“Chronic inflammation substantially increases the probability of neoplastic transformation of the surrounding cells, inducing mutations and epigenetic alterations by the activity of inflammatory molecules […] through the formation of free radicals and DNA damage […] Since infectious agents persisting in the organism may cause chronic inflammation, they can also promote local carcinogenesis. […] Chronic inflammation can also specifically affect the functioning of [an] organ, for instance, promoting cholelithiasis and urolithiasis that increase the time of exposure of the gallbladder, bile ducts, urinary bladder and ureters to chemical carcinogens and carcinogenic bacteria. […] In addition to […] metabolic and immune mechanisms, a number of bacteria […] and protozoa […] [produce] or [contain] in their cell wall their own toxins […] possessing [carcinogenic] activity, affecting cell-cell interactions, intracellular signal transduction or induction of mutations and epigenetic alterations that can influence vital cell processes (apoptosis, proliferation, survival, growth, differentiation, invasion). Intracellular protozoan (Toxoplasma gondii) may induce resistance to multiple mechanisms of apoptosis […]. So, bacterial and [protozoan toxins] may function like initiating or like promoting agents.”

“Typhoid fever, which is a systemic infection caused by Salmonella enterica serovar Typhi (S. typhi), is a major health problem in developing countries. There are approximately 21.6 million cases of typhoid fever worldwide and an estimated 200,000 deaths every year. It is known that S. typhi may colonize the gallbladder, causing […] chronic inflammation. Welton et al. (1979) were the very first to [establish] an association between the typhoid-carrier state and death due to malignancies of the hepatobiliary tract. They recruited 471 U.S. carriers of [S. typhi] , matched them with 942 controls and demonstrated that chronic typhoid carriers died of hepatobiliary cancer six times more often than the controls. […] The absence of basic research analyzing the carcinogenic properties of S. typhi does not allow placing it in the short list of the infectious agents that may be a cause of cancer development but are not included in the IARC roster, but this bacterium undoubtedly should be [on] the extended list. If [basic] studies on cell lines and animal models [support] the results of [the] epidemiological investigations, S. typhi can be placed [on] the short list.” [I included this in part because it is one of several examples in the book of how even strong correlations and high relative risks are not considered sufficient on their own by epidemiologists to settle matters. Some relative risks in other studies have been even higher – a study on gall-bladder cancer found an RR of 12.7].

“Tuberculosis (TB), a destructive disease [affecting] the lungs […] is a major global health burden, with about nine million of new cases and 1.1 million deaths annually. When the host protective immunity fails to control M. tuberculosis growth, progression to active disease occurs. […] According to the data of the last comprehensive systematic review and [meta-analysis] published by Brenner et al. (2011), there were 30 studies […] conducted in North America, Europe and Asia, which investigated the association of tuberculosis on lung cancer risk with adjustment for smoking. The relative risk (RR) of lung cancer development among patients with TB history was 1.76 (95% CI = 1.49–2.08).”

“22 studies from North America, Europe and East Asia [have] investigated the association between pneumonia and lung cancer risk while adjusting for smoking […] A significant increase in lung cancer risk was observed among all studies (RR = 1.43, 95% CI = 1.22–1.68). […] To sum up, there are basic as well as extensive epidemiological evidence that С. pneumoniae may cause lung cancer” [However effect sizes seem to be different in different countries. I was skeptical about this one in part because a non-smoker’s absolute risk of getting lung cancer is very low, meaning that relative risks in the neighbourhood reported above although statistically significant probably are clinically insignificant. How pneumonia and smoking interact seems to me a much more important question. Then again we haven’t got an explanation for all of the non-smoking-related lung cancers yet, and they are caused by something, so it’s also not like researching this is a complete waste of time.]

“Primary infection with C. trachomatis [Chlamydia], the most prevalent sexually transmitted bacterium worldwide with an estimated 90 million new cases occurring each year, is often asymptomatic and may persist for several months or years. The first study analyzing possible association of C. trachomatis with cervical cancer was carried out by Schachter et al. (1975) who assessed the prevalence of antibodies to TRIC (trachoma-inclusion conjunctivitis) agents in women with cervical dysplasia and in women attending selected clinics […]. According to this investigation, antibodies to chlamydiae were identified in 77.6% of the women with dysplasia or cervical cancer whereas antichlamydial antibodies were less prevalent in the other clinic populations. Four years later, Paavonen et al. (1979) obtained [similar] results in 93 of patients with cervical dysplasia comparing them to the controls. […] Smith et al. (2001, 2002) examined 499 women with incident invasive cervical cancer cases and 539 control patients from Brazil and the Philippines, detecting that C. trachomatis increased risk of squamous cervical cancer among HPV-positive women (OR=2.1; 95% CI=1.1–4.0). The results were similar in both countries.” [As I recently pointed out elsewhere, “Chronic infection with HPV is a necessary cause of cervical cancer. Using sensitive molecular techniques, virtually all tumours are positive for the virus.” But as this finding (and other related findings) indicate, other infectious processes may play a role as well in HPV-related cancers. Synergistic effects are common in this area (recall for example also the herpes simplex virus-HIV link).]

“Trichomonas vaginalis (T. vaginalis), a protozoan parasite, is the causative agent of trichomoniasis, the most common nonviral sexually transmitted disease in humans. This parasite has a worldwide distribution and it infects 250–350 million people worldwide. [wiki says ~150 mil, but these guesstimates should always be taken with a grain of salt. Either way it affects a lot of people] […] Zhang et al. (1995) observed a relationship between T. vaginalis infection and cervical cancer in [their] prospective study in a cohort of 16,797 Chinese women. T. vaginalis-infection correlated with higher cervical cancer risk (RR=3.3, 95% CI=1.5–7.4). In a large cohort study conducted in Finland by Viikki et al. (2000) T. vaginalis was associated with a high RR of cervical cancer, 6.4 (95% CI = 3.7–10) and SIR [standardized incidence ratio]=5.5 (95% CI=4.2–7.2s), respectively. […] Mekki and Ivić (1979), detected that T. vaginalis were of a significantly smaller diameter in invasive carcinoma and carcinoma in situ in comparison with dysplasia. In the control group with trichomoniasis alone, the diameter of T. vaginalis was twice as large as that in carcinoma and larger compared to dysplasia, indicating that small forms of T. vaginalis are more carcinogenic than large ones. […] To sum up, there are basic as well as epidemiological evidence that T. vaginalis may be a cause of cervical and prostate cancer […] For cervical cancer it is evident, for prostate cancer it is arguable. According to our criteria, it is possible to include it in the short list of the infectious agents that may be a cause of cancer development but are not placed in the IARC roster.”

“At the moment of publication, IARC [recognizes] Schistosoma haematobium [and] [S. mansoni], Opisthorchis viverrini, and Clonorchis sinensis as causative agents of cancer, leaving a possibility to enlarge this list by [Schistosoma japonicum] [and] Opisthorchis felineus.” [The authors think the list should be enlarged even more, but I did not find their helmith data/coverage very convincing (not much research has been done in this area), so I decided not to cover these things here].