The Human Microbiome in Cancer

A mini-review of microbiome optimization in integrative oncology

By Tina Kaczor, ND, FABNO

Printer Friendly PagePrinter Friendly Page

 

 

This paper is part of NMJ's 2017 Oncology Special Issue. Read the paper or download the full issue here.

Abstract

The gut microbiome is fast becoming a central focus in understanding the effectiveness of chemotherapies and immunotherapies in cancer care. In addition, various other organs have their own distinct microbiomes. These microbiomes—and imbalances in them—are associated with cancers of the head and neck, larynx, lung, breast, pancreas, esophagus, gallbladder, and colon/rectum. It is incumbent on integrative practitioners to maintain an evidence-based approach to optimizing health through microbiome optimization. Human microbiomes can be considered micro-ecosystems, and their benefit to the host may be best reflected in their complexity.

We now understand the human microbiome to be intimately linked to both cancer causation and outcomes of cancer treatment. Recent advances in technology and bioinformatics are allowing for elucidation of the microbiome established in various tissues. These microbiomes are being better appreciated as micro-ecosystems whose relationship with the human host can dictate health or disease.

Introduction

According to the World Health Organization, singular organisms are directly responsible for causing up to 15% of cancers worldwide.1 Now microbiota imbalances—from species-specific to phylum-wide—are also implicated in cancer etiology. Cancers of the colon, breast, larynx, esophagus, pancreas, head and neck, lung, and gallbladder have been associated with imbalances of their resident microbiota.2-7 These relationships are slowly being unraveled as the complex, entwined biology of the host-biota symbiosis is better understood as a micro-ecosystem in which the health or disease state of its host is wholly dependent.8

There is also a growing appreciation for the role of the microbiota in general, and the gut flora specifically, in cancer treatment outcome.9 The idea that chemotherapy works by directly assaulting cancer cells has been an oversimplification. We now know that chemotherapies depend on the immune system to home in on tumors, and immune stimulation may account for much of chemotherapy’s “tumor kill.”10 This immune response is intimately dependent on the gut microbiome.11

While we cannot recommend specific probiotics or prebiotics, we are in a position to optimize microbiota diversity in an effort to create the healthiest micro-ecosystem possible.

The microbiota of the body comprise bacteria, fungi, archaea, and viruses, with bacterial organisms dominating the populations in healthy people. The taxonomy (identification of organisms by phyla, family, species, etc.) of organisms between organs is called spatial diversity or beta variation. For example, the colon is generally dominated by phyla: Firmicutes and Bacteroidetes. From the mouth to the stomach, however, Firmicutes and Actinobacteria are the dominant phyla. Thus, each organ can be considered as having its own unique micro-ecology. In addition to spatial diversity, these micro-ecologies also undergo temporal changes, varying over the lifespan and shifting according to lifestyle habits and seasons of the year.12-14

Taxonomy can help us understand human microbiota. However, we cannot fully understand the complexity of interactions between the microbiota and its human host by merely identifying the organisms. It is the net functions of the microbiota that the human body relies on. For example, in one study the taxonomy of the mouth and gut varied dramatically by phylum, but the functional aspects, such as carbohydrate metabolism, purine metabolism, and cofactor/vitamin biosynthesis (measured through metagenomics) were remarkably similar.15

Again, the idea of the microbiome as a micro-ecosystem is apt. Comparing function over phylum is analogous to comparing the complex ecology of a healthy rainforest to a healthy deciduous forest. They may have little in common taxonomically, but the net effect of their functions, such as oxygen production, water filtration, and organic matter decay, is very similar. Much like the human microbiome, forest ecosystems rely on high levels of diversity for their health, and it is only after perturbations in this diversity (eg, invasive species, clear cutting) that we begin to discover the consequences.

A systems biology approach, together with the ability to discern form and function of the microbiome with technology only recently available, has led to a renaissance of sorts in medicine, and oncology is no exception. Practitioners of integrative oncology should be ready to incorporate the support with the most evidence to maximize outcomes of conventional treatments for our patients. For example, proper sleep, hygiene, plant-based diets, and exercise all optimize the health of the human microbiota.16-18 This review describes some of the most recent data furthering our understanding of the microbiota in cancer care.

Human Microbiota and Cancer Therapies

In 2013 the journal Science published 2 seminal studies on the gut microbiota and conventional cancer therapies. In the first experiment, by Viaud and colleagues, cyclophosphamide (a chemotherapy drug) was given to tumor-bearing mice.19,20 Cyclophosphamide is a known cytotoxic agent and an immune-modulating drug that can affect T lymphocytes (T cells).21 The administration of cyclophosphamide caused a translocation of gram-positive commensal bacteria from the gut to secondary lymph tissues (mesenteric nodes and spleen). This stimulated a measurable anticancer immune response, specifically through engaging T helper 17 (Th17) cells and memory T cells. When mice were treated with antibiotics, the tumor response to cyclophosphamide was greatly reduced.

To be clear, chemotherapy’s ability to affect immune function through antigen stimulation and T cell activation has been known for some time.22 That commensal organisms translocated intact to stimulate this response is paradigm-shifting. This concept should give us pause when we contemplate the long-held assumptions of separation between our body’s internal organs and our microbiota and make us reconsider the indiscriminant use of antibiotics during chemotherapy or radiation treatment.22

In 2016, Pflug et al conducted a retrospective analysis of 800 cyclophosphamide-treated chronic lymphocytic leukemia (CLL) patients and 122 cisplatin-treated lymphoma patients to see if outcomes were affected by antibiotics specific for gram-positive bacteria.23 Consistent with Viaud’s observations in mice, the patients with CLL who received antibiotics (n=45) had significantly worse outcomes. Progression-free survival (PFS) was 14.1 months, and median overall survival (OS) was 56.1 months for patients who received antibiotics, compared to 44.1 months (P<0.001) and 91.7 months (P<0.001) for those who did not. Outcomes were similar for lymphoma patients treated with cisplatin. Patients who received antibiotics (n=21) had a PFS of 2.3 months, compared to 11.5 months for those who did not (P=0.001). The authors concluded, “Our data supports a potential negative impact of anti-Gram-positive antibiotics on the anticancer activity of cyclophosphamide and cisplatin in a clinical setting.”23

This clinical outcome data certainly informs us on the use of antibiotics with cyclophosphamide and platinum drugs. However, we are still left wondering if a damaged mucosal lining may augment immune stimulation, or even facilitate translocation of gram-positive bacteria. Jacob Schor, ND, FABNO, posed this conundrum in this journal shortly after Viaud’s research in 2013: Should a patient’s increased gut permeability during chemotherapy be treated or not?24 There is no doubt that chemotherapeutics damage the epithelial layer of the gut, with those inducing mucositis the worst culprits.25,26 Does this damage somehow optimize the interface of the microbiota with the underlying immune reaction at the epithelial interface? Do those with “leaky gut” have better response to therapies? Or, is there an amount of intestinal permeability that is optimal, and a point at which it is so severe it poses more threat than possible benefit?

Another 2013 study published in Science was aptly titled “Commensal Bacteria Control Cancer Response to Therapy by Modulating the Tumor Microenvironment.”27 This study, by Iida and colleagues, was actually 2 experiments. Mice with transplanted tumors were treated either with chemotherapy (oxaliplatin) or with an immune treatment (CpG-oligonucleotide immunotherapy). In each case tumor response to the anticancer treatment required the presence of commensal bacteria in the gut. Mechanistically, Iida and colleagues demonstrated that commensal flora is requisite to obtain the oxidative burst or the inflammatory necrosis induced by oxaliplatin or the immune therapy, respectively.

How does the gut microflora manage to affect distant tumors? Just below the intestinal epithelial layer dendritic cells sample antigens or whole organisms and present these to secondary lymph tissue.28 At the tumor itself, tumor-associated myeloid (TAM) cells communicate with nearby T cells, largely through Toll-like receptors.29 These myeloid cells are highly plastic cells that can differentiate into various phenotypes involved in tumor progression or regression, depending on the molecular signals received.30,31 Experiments have shown that many of the signals that myeloid cells use to elicit immune response/recognition of the tumor require the presence of intact gut microbiota.10,32 In the experiments already mentioned, Iida and colleagues showed TAMs were not capable of producing the requisite oxidation for oxaliplatin cytotoxicity when the gut was sterilized. In Viaud’s cyclophosphamide experiment, myeloid cells gave rise to increased effector T cells and memory cells, and this reaction did not take place when the gut was sterile. In addition, experiments with whole body irradiation plus adoptive T cell transfer demonstrate that TAMs lead to cytotoxic lymphocytes expansion and ensuing cytotoxicity, but only in the presence of normal gut microbiota.33

Checkpoint Blockade Agents and Microbiota

Immunotherapies have been in the news due to the emergence of checkpoint inhibitors approved for several tough-to-treat cancers. Programmed death-1 (PD-1) receptor, programmed death ligand-1 (PD-L1), and cytotoxic T lymphocyte–associated protein-4 (CTLA-4) are checkpoint blockade proteins expressed by cancer cells and cells in tumor stroma. Monoclonal antibodies that target these checkpoint blockade proteins include pembrolizumab (Keytruda), nivolumab (Opdivo), atezolizumab (Tecentriq), avelumab (Bavencio), durvalumab (Imfinzi), and ipilimumab (Yervoy).

Normally, once engaged, PD-1, PDL-1, or CTLA-4 act as brakes to halt the immune attack of cancer cells. When one of the monoclonal antibodies binds to the checkpoint protein, the brake is rendered ineffective, which allows the immune response to proceed.

Several studies in mice have assessed the effect of various gut flora on checkpoint inhibitor antibody efficacy. In one study, Bacteroides fragilis reduced CTLA-4 blockade–induced colitis and promoted the maturation of intratumoral dendritic cells.34 These effects would seem to be beneficial to a patient, making the treatment more tolerable and perhaps more effective.

However, this mouse study should be interpreted with caution. Colitis severity may be associated with treatment efficacy. In a small study of metastatic melanoma patients (n=26), fecal microbiota composition was assessed at baseline and before each infusion of ipilumimab.35 Patients whose baseline microbiota was enriched with Faecalibacterium genus and other Firmicutes (cluster A; n=12) had longer PFS (P=0.0039) and OS (P=0.051) than patients whose baseline microbiota was driven by Bacteroides (cluster B; n=10). Those with Firmicutes-dominant flora also had fewer immunosuppressive regulatory T cells (Tregs), less colitis, and overall better benefit of treatment.35

In another study of tumors in mice, Bifidobacterium spp (breve, longum, and adolescentis) enhanced dendritic cell activation and improved response to PD-1 inhibitor treatment.36 In Iida’s experiment mentioned above, commensals that augmented immunotherapy included Ruminococcus and Alistipes shahii, while Lactobacillus fermentum dampened immunotherapy.

While we continue to sort out what bacterial composition is optimal with checkpoint inhibition, recent human data suggests that a diversity of flora may be of particular benefit. In a large cohort (n=221) of metastatic melanoma patients, most of whom had received PD-1 inhibitors (n=105), there was higher microbiota diversity in treatment responders vs nonresponders (P=0.03).37 There was also a higher abundance of the class Clostridiales in responders but higher Bacteroidales in nonresponders. The authors concluded that diversity (P=0.009; hazard ratio [HR]=7.67) and abundance of specific bacteria (P=0.007; HR=3.88) was associated with improved PFS to anti-PD-1 therapy.

The first clinical study to specifically look at the gut flora and checkpoint inhibitor outcomes was published in 2015. Dubin and colleagues found that metastatic melanoma patients with Bacteroidetes-dominant gut flora undergoing treatment with a CTLA-4 inhibitor had less colitis than those without Bacteroidetes dominance.38 Colitis was also more prevalent in those whose biome had fewer genes involved in polyamine transport and B-vitamin synthesis. Bacteroidetes are bile acid–resistant bacteria that are found more often in those consuming diets higher in animal proteins and fats. Theoretically, suggesting patients consume more animal-based products may lead to less colitis with treatment. But at what cost?

Clinical Considerations

For now, precise recommendations regarding what species of probiotics, which prebiotics, and what type of diet optimize gut flora during checkpoint blockade treatment is not clear. While side effect reduction—such as lessening colitis by enhancing Bacteroidetes—is often the goal, data suggests this could undermine the treatment response itself.

One consistent finding in both rodents and humans is that antibiotics blunt the immune response induced by the checkpoint blockade agents.19,27,39 A recent study by Derosa and colleagues found that antibiotic use during treatment with PD-1 inhibitors can affect outcomes adversely. In a retrospective analysis of 175 patients receiving PD-1 inhibitor therapy for a variety of cancers, almost a third received an antibiotic (mostly beta lactamases or fluoroquinolones) within 1 month before or 2 months after starting treatment with a PD-1 inhibitor. Overall, those who took an antibiotic had lower PFS and lower OS (3.4 months vs 5.2 months, P<0.013, and 12.2 months vs 20.8 months, P<0.001). This held true when patients were stratified by cancer type as well.

Antibiotics are sometimes a necessary component of care, but they should be reserved as a last resort in those undergoing cancer treatment.40 As integrative practitioners we must help our patients maintain a strong defensive immune system, through reminders about hygiene, sleep, exercise, laughter, nutrient repletion, and a healthy diet, as the foundation of care to avoid any need for antibiotics.

While we cannot recommend specific probiotics or prebiotics, we are in a position to optimize microbiota diversity in an effort to create the healthiest micro-ecosystem possible. As reviewed above, evidence implies that diversity may be one of the keys to immune recognition of cancer cells and efficacy of anticancer treatments.41

Perhaps one of the most profound studies suggesting that gut flora diversity affects cancer outcomes was published in 2014 by Taur and colleagues.42 Stool samples were obtained from 80 stem cell transplant recipients at the time of stem cell engraftment, and patients were followed over the course of the next few years. At 3 years post-transplant, survival was 36%, 60%, and 67% for the low, intermediate, and high diversity groups, respectively (P=0.019). In this study bacterial diversity was an independent risk factor for mortality.

In the end, microbiota diversity, like ecological diversity, may be the ultimate measure of a healthy micro-ecosystem. If this is true, then a highly varied diet may be the dietary recommendation.

Conclusion

Currently, dysbiosis of the human microbiome in many organs is associated with variety of cancers. It is only a matter of time before microbiome profiles are used as clinical tools that will stratify cancer risk or assist early detection.43 Of course, association of organ dysbiosis with cancer is not synonymous with causation. However, corroborative evidence, such as the increased incidence of breast cancer in women with high antibiotic use, certainly strengthens the case.44-47

We are just beginning to define an optimal microbiome that will positively affect outcomes of chemotherapy or immunotherapies. Given the early data, it appears likely that commensal organisms are integrally involved in a tumor’s response to many cancer treatments. In addition, bacterial diversity appears to be associated with better treatment response. One piece of evidence-based medicine that may affect outcomes is the use of antibiotics as a last resort during treatment. Evidence to date does not support use of high-dose, specific probiotic strains to improve cancer treatment outcomes. A healthy omnivorous diet, such as the Mediterranean diet, along with exercise and proper sleep, may be the best, though less-than-precise, prescription to complement conventional therapies.

About the Author

Tina Kaczor, ND, FABNO, is editor-in-chief of Natural Medicine Journal and a naturopathic physician, board certified in naturopathic oncology. She received her naturopathic doctorate from National University of Natural Medicine, and completed her residency in naturopathic oncology at Cancer Treatment Centers of America, Tulsa, Oklahoma. Kaczor received undergraduate degrees from the State University of New York at Buffalo. She is the past president and treasurer of the Oncology Association of Naturopathic Physicians and secretary of the American Board of Naturopathic Oncology. She has been published in several peer-reviewed journals. Kaczor is based in Portland, Oregon.

References

  1. World Health Organization. Cancer prevention. http://www.who.int/cancer/prevention/en/. Accessed August 31, 2017.
  2. Gao Z, Guo B, Gao R, Zhu Q, Qin H. Microbiota disbiosis is associated with colorectal cancer. Front Microbiol. 2015;6:20.
  3. Sinha R, Ahn J, Sampson JN, et al. Fecal microbiota, fecal metabolome, and colorectal cancer interrelations. PLoS One. 2016;11(3):e0152126.
  4. Xuan C, Shamonki JM, Chung A, et al. Microbial dysbiosis is associated with human breast cancer. PLoS One. 2014;9(1):e837.
  5. Sheflin AM, Whitney AK, Weir TL. Cancer-promoting effects of microbial dysbiosis. Curr Oncol Rep. 2014;16(10):406.
  6. Kantono M, Guo B. Inflammasomes and cancer: the dynamic role of the inflammasome in tumor development. Front Immunol. 2017;8:1132.
  7. Wang H, Altemus J, Niazi F, et al. Breast tissue, oral and urinary microbiomes in breast cancer. Oncotarget. 2017;8:88122-88138.
  8. Raes J, Bork P. Molecular eco-systems biology: towards an understanding of community function. Nat Rev Microbiol. 2008;6(9):693-699.
  9. Viaud S, Daillère R, Boneca IG, et al. Harnessing the intestinal microbiome for optimal therapeutic immunomodulation. Cancer Res. 2014;74(16):4217-4221.
  10. Karin M, Jobin C, Balkwill F. Chemotherapy, immunity and microbiota--a new triumvirate? Nat Med. 2014;20(2):126-127.
  11. Bordon Y. Tumour immunology: anticancer drugs need bugs. Nat Rev Immunol. 2014;14(1):1.
  12. Jandhyala SM, Talukdar R, Subramanyam C, Vuyyuru H, Sasikala M, Nageshwar Reddy D. Role of the normal gut microbiota. World J Gastroenterol. 2015;21(29):8787-8803.
  13. Hisada T, Endoh K, Kuriki K. Inter- and intra-individual variations in seasonal and daily stabilities of the human gut microbiota in Japanese. Arch Microbiol. 2015;197(7):919-934.
  14. Davenport ER, Mizrahi-Man O, Michelini K, Barreiro LB, Ober C, Gilad Y. Seasonal variation in human gut microbiome composition. PLoS One. 2014;9(3):e90731.
  15. Lozupone CA, Stombaugh JI, Gordon JI, Jansson JK, Knight R. Diversity, stability and resilience of the human gut microbiota. Nature. 2012;489(7415):220-230.
  16. Sheflin AM, Melby CL, Carbonero F, Weir TL. Linking dietary patterns with gut microbial composition and function. Gut Microbes. 2017;8(2):113-129.
  17. Chandrakumaran H, Safdar A, Sager M, Nazli A, Akhtar M. Regular exercise shapes healthy gut microbiome. J Bacteriol Mycol Open Access. 2016;3(3):63.
  18. Zhang SL, Bai L, Goel N, et al. Human and rat gut microbiome composition is maintained following sleep restriction. Proc Natl Acad Sci. 2017;114(8):E1564-E1571.
  19. Viaud S, Saccheri F, Mignot G, et al. The intestinal microbiota modulates the anticancer immune effects of cyclophosphamide. Science. 2013;342(6161):971-976.
  20. Madondo MT, Quinn M, Plebanski M. Low dose cyclophosphamide: mechanisms of T cell modulation. Cancer Treat Rev. 2016;42:3-9.
  21. Scurr MJ, Pembroke T, Bloom A, et al. Low-dose cyclophosphamide induces anti-tumor T-cell responses which associate with survival in metastatic colorectal cancer. Clin Cancer Res. 2017:clincanres.0895.
  22. Apetoh L, Ghiringhelli F, Tesniere A, et al. Toll-like receptor 4-dependent contribution of the immune system to anticancer chemotherapy and radiotherapy. Nat Med. 2007;13(9):1050-1059.
  23. Pflug N, Kluth S, Vehreschild JJ, et al. Efficacy of antineoplastic treatment is associated with the use of antibiotics that modulate intestinal microbiota. Oncoimmunology. 2016;5(6):e1150399.
  24. Schor J. Leaky gut and chemo: to treat or not to treat? Nat Med J. 2014;6(21).
  25. Melichar B, Hyšpler R, Dragounová E, Dvořák J, Kalábová H, Tichá A. Gastrointestinal permeability in ovarian cancer and breast cancer patients treated with paclitaxel and platinum. 2007;7:155.
  26. van Vliet MJ, Harmsen HJM, de Bont ESJM, Tissing WJE. The role of intestinal microbiota in the development and severity of chemotherapy-induced mucositis. PLoS Pathog. 2010;6(5):e1000879.
  27. Iida N, Dzutsev A, Stewart CA, et al. Commensal bacteria control cancer response to therapy by modulating the tumor microenvironment. Science. 2013;342(6161):967-970.
  28. Hooper LV, Littman DR, Macpherson AJ. Interactions between the microbiota and the immune system. Science. 2012;336(6086):1268-1273.
  29. Goldszmid RS, Dzutsev A, Viaud S, Zitvogel L, Restifo NP, Trinchieri G. Microbiota modulation of myeloid cells in cancer therapy. Cancer Immunol Res. 2015;3(2):103-109.
  30. Schouppe E, De Baetselier P, Van Ginderachter JA, Sarukhan A. Instruction of myeloid cells by the tumor microenvironment: open questions on the dynamics and plasticity of different tumor-associated myeloid cell populations. Oncoimmunology. 2012;1(7):1135-1145.
  31. Elliott LA, Doherty GA, Sheahan K, Ryan EJ. Human tumor-infiltrating myeloid cells: phenotypic and functional diversity. Front Immunol. 2017;8:86.
  32. Gorjifard S, Goldszmid RS. Beating cancer with a gut feeling. Cell Host Microbe. 2015;18(6):646-648.
  33. Paulos CM, Wrzesinski C, Kaiser A, et al. Microbial translocation augments the function of adoptively transferred self/tumor-specific CD8+ T cells via TLR4 signaling. J Clin Invest. 2007;117(8):2197-2204.
  34. Vétizou M, Pitt JM, Daillère R, et al. Anticancer immunotherapy by CTLA-4 blockade relies on the gut microbiota. 2015;350(6264):1079-1084.
  35. Chaput N, Lepage P, Coutzac C, et al. Baseline gut microbiota predicts clinical response and colitis in metastatic melanoma patients treated with ipilimumab. Ann Oncol. 2017;28(6):1368-1379.
  36. Sivan A, Corrales L, Hubert N, et al. Commensal Bifidobacterium promotes antitumor immunity and facilitates anti–PD-L1 efficacy. Science. 2015;350(6264):1084-1089.
  37. Wargo JA, Gopalakrishnan V, Spencer C, et al. Association of the diversity and composition of the gut microbiome with responses and survival (PFS) in metastatic melanoma (MM) patients (pts) on anti-PD-1 therapy. J Clin Oncol. 2017;35(15):3008
  38. Dubin K, Callahan MK, Ren B, et al. Intestinal microbiome analyses identify melanoma patients at risk for checkpoint-blockade-induced colitis. Nat Commun. 2016;7:10391.
  39. Saleh K, Khalife-saleh N, Kourie HR. Is gut microbiome a predictive marker to response to immune checkpoint inhibitors? Immunotherapy. 2017;9(11):865-866.
  40. Derosa L, Routy B, Mezquita L, et al. Antibiotics prescription to decrease progression-free survival (PFS) and overall survival (OS) in patients with advanced cancers treated with PD1/PDL1 immune checkpoint inhibitors. J Clin Oncol. 2017;35(15_suppl):3015.
  41. Helwick C. Gut bacteria may enhance, or hamper, response to anti–PD-1 agents. The ASCO Post. http://www.ascopost.com/issues/april-10-2017/gut-bacteria-may-enhance-or-hamper-response-to-anti-pd-1-agents/. Published April 10, 2017. Accessed September 4, 2017.
  42. Taur Y, Jenq RR, Perales M-A, et al. The effects of intestinal tract bacterial diversity on mortality following allogeneic hematopoietic stem cell transplantation. Blood. 2014;124(7):1174-1182.
  43. Zeller G, Tap J, Voigt AY, et al. Potential of fecal microbiota for early-stage detection of colorectal cancer. Mol Syst Biol. 2014;10(11):766.
  44. Knekt P, Adlercreutz H, Rissanen H, Aromaa A, Teppo L, Heliövaara M. Does antibacterial treatment for urinary tract infection contribute to the risk of breast cancer? Br J Cancer. 2000;82(5):1107.
  45. Garcia Rodriguez LA, Gonzalez-Perez A. Use of antibiotics and risk of breast cancer. Am J Epidemiol. 2005;161(7):616-619.
  46. Deming-Halverson SL, Hodgson ME, D’Aloisio A, Shore D, Sandler D. Antibiotic use and breast cancer risk: results from the Sister Study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res. 2017;77(13).
  47. Sørensen HT, Skriver MV, Friis S, McLaughlin JK, Blot WJ, Baron JA. Use of antibiotics and risk of breast cancer: a population-based case–control study. Br J Cancer. 2005;92(3):594.