Infection, immunoregulation, and cancer
Transcript of Infection, immunoregulation, and cancer
� 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011 141
Graham A. W. Rook
Angus DalgleishInfection, immunoregulation, andcancer
Authors’ addresses
Graham A. W. Rook1, Angus Dalgleish2
1Department of Infection, University College London (UCL),
London, UK.2St George’s Healthcare NHS Trust, London, UK.
Correspondence to:
Graham A. W. Rook
Department of Infection
University College London (UCL)
46 Cleveland Street
London W1T 4JF, UK
Tel.: +44 20 7679 9489
Fax: +44 20 7607 0016
e-mail: [email protected]
Acknowledgements
The authors declare no conflicts of interest.
Immunological Reviews 2011
Vol. 240: 141–159
Printed in Singapore. All rights reserved
� 2011 John Wiley & Sons A/S
Immunological Reviews
0105-2896
Summary: As man has moved rapidly from the hunter–gatherer envi-ronment to the living conditions of the industrialized countries, the inci-dences of some cancers have increased alarmingly. Recent increases areusually attributed to dietary changes or to altered exposures to putativecarcinogens associated with the modern lifestyle. However, the changesin cancer incidence parallel similar increases in non-neoplastic chronicinflammatory disorders (inflammatory bowel disease, allergies, andautoimmunity), and the epidemiology is often strikingly similar. Thisparallel is worth exploring, because the increases in chronic inflamma-tory disorders are at least partly explained by immunoregulatory defectsresulting from diminished exposure to microorganisms that co-evolvedwith mammals and developed a role in driving immunoregulatory cir-cuits (the hygiene hypothesis). Dysregulated chronic inflammation candrive oncogenesis and also provides growth and angiogenic factors thatenhance the subsequent proliferation and spread of tumor cells. Thus, amodern failure to downregulate inappropriate inflammation couldunderlie increases in some cancers in parallel with the increases inchronic inflammatory disorders. This possibility is supported by recentwork showing that in some circumstances regulatory T cells protectagainst cancer, rather than aggravating it, as previously assumed. Agreater understanding of these interactions might pave the way toimproved microbe-based immunotherapies.
Keywords: cancer, immunoregulation, Treg, hygiene hypothesis, infection, IL-10,TGF-b
Introduction
As man has moved rapidly from the hunter–gatherer environ-
ment to the living conditions of the rich industrialized coun-
tries, the incidences of some cancers have increased
alarmingly. Some cancers (< 20%) are attributed to onco-
genic changes or chronic inflammation caused by infections,
but most of the recent increases have been attributed by
oncologists to dietary changes or to altered exposures to puta-
tive carcinogens associated with the modern lifestyle. Interest-
ingly, the changes in cancer incidence have been paralleled by
even larger increases in a number of non-neoplastic chronic
inflammatory disorders (inflammatory bowel disease, aller-
gies, and autoimmunity), and the details of the epidemiology
are often strikingly similar. This parallel is worth exploring,
because chronic inflammation can contribute to oncogenesis
and often provides growth and angiogenic factors which
enhance the proliferation and spread of tumor cells. A failure
to regulate inflammation could underlie not only the chronic
inflammatory disorders but also some cancers.
If this hypothetical parallel were found to be real, it would
provide important pathogenetic clues, because a significant
part of the increase in chronic inflammatory disorders is
now thought to be attributable to diminished exposure to a
range of microorganisms and parasites that either co-evolved
with humans or accompanied humans from at least as far
back as the Paleolithic era. These organisms have been pres-
ent throughout human evolution as commensals (notably in
the intestinal microbiota), environmental ‘pseudocommen-
sals’, causative agents of subclinical infections or asymptom-
atic carrier states, and helminth infections. Modern revisions
of the ‘Hygiene’ or ‘Old Friends’ hypothesis suggest that
such organisms have been tasked by co-evolutionary pro-
cesses with establishing the ‘normal’ levels of immunoregu-
lation within the immune system. Therefore, when exposure
to the Old Friends is reduced, there is a relative failure to
terminate inappropriate inflammatory responses, leading to
increases in the chronic inflammatory disorders listed above
(1). Here, in view of the crucial roles of inflammation in
cancer, we explore the potential relevance of this concept to
the changing incidences of several cancers in developed
countries.
Infection as a cause of cancer
There have been numerous excellent reviews of the role of
infections as causes of cancer (2–6). In contrast, there has
been much less discussion of the possible roles of infections
in the prevention of cancer, and in this review we seek to
explore both aspects in parallel.
Table 1 illustrates most of the organisms associated with can-
cer development, classified as far as possible according to how
cancer is promoted (though there is much overlap). One
purpose of the table is to demonstrate that inflammation is a
crucial aspect of oncogenesis attributable to infection, while
very few organisms cause cancer by direct infection and trans-
formation of target cells. The lymphotropic oncogenic viruses
Epstein–Barr virus (EBV), human herpesvirus 8 (HHV8), and
human T-lymphotropic virus 1 (HTLV-1), are examples of the
latter, and these viruses directly infect a subset of lymphoid
cells, which they can then transform either directly or
indirectly by transactivation of oncogenes, as in the case of
HTLV-1.
Little is known about the role in cancer causation of the
truly intracellular infections that form the major focus of this
Table 1. The major infectious agents that trigger cancer
Mechanism Organism Cancer References
Infect and transform lymphoid cells Epstein–Barr virus (EBV) Burkitt’s lymphoma* (4)Human Herpesvirus 8 (HHV8) Kaposi sarcomaHuman T-lymphotropic virus 1 (HTLV-1) T-cell leukemia
Transformation HPV Cervical cancer (2)Inflammation and partial integration HBV Hepatocellular carcinoma (2)Chronic inflammation and oncogenic proteins Hepatitis C virus (HCV) Hepatocellular carcinoma (4)Chronic stimulation of lymphocytes bypathogen antigens and ⁄ or autoantigens
Helicobacter pylori Gastric lymphoma (3)Campylobacter jejuni � Immunoproliferative small intestinal dis. (29)Borrelia burgdorferi Cutaneous lymphoma (3)Chlamydia psittaci Ocular lymphoma (30)Hepatitis C virus (HCV)� Spleen lymphoma (3)
Chronic inflammation Bacteroides fragilis (enterotoxigenic) Colorectal cancer (146, 147)Malaria (cofactor) Burkitt’s lymphoma (7)Tuberculosis Lung cancer (9, 10)Schistosomiasis Bladder cancer (2, 5)Liver flukes Cholangiocarcinoma (5)Salmonella spp. Gallbladder cancer (8)Helicobacter pylori Gastric and esophageal adenocarcinoma (113, 115)
Chronic inflammation; possible associations Propionibacterium acnes? Prostate (148)Mycoplasma spp. ? Prostate (149)?? unknown prostatitis ?? Prostate
Immunosuppression HIV EBV+ CNS lymphomas (4)HHV8+ sarcoma (Kaposi)HPV+ ano-genital cancers
*EBV also associated with nasopharyngeal carcinoma, Hodgkin lymphoma and non-Hodgkin lymphoma.�C. jejuni produce CdtB (cytolethal distending toxin B) that causes direct DNA damage.�HCV also encodes oncogenic proteins.
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
142 � 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011
volume, though there are exceptions. Malaria can act as a co-
factor for EBV-induced oncogenesis, because Plasmodium falcipa-
rum antigens such as PfEMP1 can induce reactivation of EBV
(7). The increased viral load and the concomitant polyclonal
B-cell activation with enhanced B-cell survival may augment
the risk of endemic Burkitt’s lymphoma in children living in
malaria-endemic areas (7). Similarly, although gallbladder
cancer is relatively rare, the highest incidences are seen in
populations where chronic infections like Salmonella typhi and S.
paratyphi are prevalent, though obesity and multiparity are also
consistently associated (8). More controversially, it has been
reported that autopsies of patients with lung cancer frequently
reveal the simultaneous presence of tuberculous lesions in the
lungs (9). This observation might be meaningful, because it
has been reported that there is an increased risk of lung cancer
following a diagnosis of TB among farmers in Xuanwei,
China. The risk was greatest within 5 years of diagnosis but
remained elevated (approximately threefold) for more than
10 years. This association was seen in men and women and
was not due to confounding effects of tobacco use or indoor
environmental smoke (10). Nevertheless, the situation is
likely to be more complex, because M. tuberculosis has immuno-
regulatory effects that differ according to whether latent or
progressive infection has developed, and this is discussed
later.
In view of the paucity of direct evidence for involvement of
intracellular infections in either preventing or causing onco-
genesis (excluding viruses which are outside the remit of this
article), certain other organisms will be used to illustrate their
potential role. In particular, the involvement of Helicobacter pylori
in gastric lymphoma and gastric and esophageal adenocarci-
noma is of striking interest. Most H. pylori localize within the
gastric mucus layer and do not directly interact with host cells.
However, some organisms adhere to gastric epithelial cells
and occasionally are internalized by these cells. Using fluores-
cent in situ hybridization (FISH) with probes for H. pylori 16S
rRNA and cagA, H. pylori was detected inside metaplastic, dys-
plastic, and neoplastic epithelial cells (11). These findings are
compatible with the view that an intracellular location is
involved in the carcinogenic effect and fits with the observa-
tion that therapy aimed at eliminating H. pylori can induce
regression of precancerous lesions (11).
Chronic inflammation
Inflammation is present in tumors and is fundamental to their
development and spread. Inflammation can enhance mutation
(12), growth, vascularization (13), and metastasis (Fig. 1).
Tumor necrosis factor-a (TNF-a) may be particularly impor-
tant (14), and it has been found to be secreted by a wide vari-
ety of tumor cells, including B and T-cell lymphoma, myeloid
and lymphoid leukemias, carcinomas of the breast, colon,
lung, ovary, pancreas, prostate, cervix, and glioblastoma, and
neuroblastoma (reviewed in 15). There have been excellent
recent reviews of these issues, which will therefore only be
outlined briefly here (14–18).
Background: inflammation in cancer
First, tumors contain ‘smoldering’ inflammation (including
release of TNF-a, as discussed above) that is not obviously
related to any external inflammatory stimulus (17). This is
partly because several important oncogenes such as RAS and
MYC activate signaling pathways involved in inflammation
(17). In addition, commonly seen mutations in tumor-sup-
pressor genes such as Von Hippel-Lindau tumor suppressor
(VHL), transforming growth factor-b (TGF-b), and phospha-
tase and tensin homologue (PTEN), can all lead to increased
activation of transcription factors involved in inflammation
and vascularization, particularly nuclear factor-jB (NF-jB),
hypoxia inducible factor 1a (HIF-1a), and signal transducer
and activator of transcription 3 (STAT3) (19). Moreover, many
carcinomas secrete factors that upregulate fibronectin and
recruit vascular endothelial growth factor receptor 1 (VEG-
FR1)-positive hematopoietic progenitors and inflammatory
Fig. 1. Some potential roles of infection in preventing and causingcancer (excluding direct transformation, which is discussed briefly inthe main text). Epidemiological studies indicate that transient infectionmay provide some protection against cancer. However, chronic infectionleading to chronic inflammation can initiate oncogenesis, and also, bysubsequently providing growth and angiogenic factors, potentiate exist-ing cancers.
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
� 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011 143
cells to sites of tumor growth. These cells may also secrete fac-
tors that directly activate recruited myeloid cells to produce
tumor-promoting cytokines such as TNF-a and IL-6 (14, 20).
Increased rates of DNA damage and defective repair
The mutation rate in inflamed tissues is higher than in normal
tissues, though lower than in tumors (21). A number of
mechanisms and pathways have been implicated in the
increased mutation rate. Growth factors, cytokines, and
chemokines produced by inflammatory cells and by the tumor
itself induce increased expression of transcription factors such
as c-Myc and also increase the activity of NF-jB, HIF-1a, and
STAT3. These transcription factors alter the expression of hun-
dreds of target genes related to cell growth, apoptosis, inva-
sion, and inflammation. These processes accelerate the
intrinsic mutation rate in cancer cells, at least partly by
increasing levels of reactive oxygen species (ROS) and nitro-
gen species.
ROS released from inflammatory cells can cause double
strand breaks in the DNA of cancer cells. Moreover nitric
oxide (NO) and the inflammatory cytokine interleukin-6 (IL-
6) increase the activity of DNA methyltransferase, resulting in
extensive methylation of cytosine bases including those in
CpG islands that can be located in promoters. This can lead to
the silencing of promoters and loss of expression of the genes
encoding essential mismatch-repair genes, such as hMLH1.
This epigenetic change leads in turn to faulty DNA repair,
resulting in microsatellite insertions and deletions within the
coding regions of tumor suppressor genes. (microsatellites are
normal, common, repeated sequences of DNA from 1to 6 bp,
often located in promoter regions). The consequent microsat-
ellite instability in tumor suppressor genes can lead to uncon-
trolled cell division and tumor growth (22). Among the genes
that contain microsatellites in their coding regions are those
encoding the TGF-b receptor 2, insulin-like growth factor 2
receptor, and BCL2-associated X protein (BAX).
Another intriguing mechanism of DNA damage is ectopic
expression of activation-induced cytidine deaminase (AID).
This DNA ⁄RNA editing enzyme normally induces hypermuta-
tion of the immunoglobulin loci in B cells. However, it can be
ectopically induced in tumors by a variety of inflammatory
cytokines. When ectopically expressed, rather than targeting
immunoglobulin loci in B cells, AID produces mutations and
translocations through induction of double strand breaks in
the DNA of other cell types. Ultimately chromosomal instabil-
ity results in abnormal segregation of chromosomes and aneu-
ploidy (12).
Angiogenesis and metastasis
In healthy adults, the growth of new vessels is largely confined
to the female genital tract and to sites of wound repair. It is
regulated by a balance between angiogenic factors such as vas-
cular endothelial growth factor A (VEGFA), fibroblast growth
factors (FGFs), platelet-derived growth factor (PDGF), and
epidermal growth factor (EGF), and inhibitors of angiogenesis
such as thrombospondin 1, angiostatin, endostatin, and
tumstatin. The inflammatory cascades in tumors (Fig. 1),
whether the intrinsic effects of the oncogenic mutations in the
tumor cells themselves (17), secondary release of mediators
from recruited inflammatory cells (20), or inflammation due
to infection (23), can all lead to an imbalance, where inappro-
priate angiogenesis is encouraged and metastasis is facilitated
(24–26). In a classic experiment, fibrosarcoma cells showed
enhanced metastasis to the lungs in the presence of exogenous
TNF, while neutralization of endogenous TNF led to a signifi-
cant decrease in the number of pulmonary metastases (26,
27). Angiogenesis is further promoted by infiltration of
monocytes expressing tyrosine kinase with immunoglobulin-
like and EGF-like domains 2 (TIE-2). These TIE-2-expressing
monocytes (TEMs) constitute a small pro-angiogenic mono-
cyte subset that circulates in the mouse and human peripheral
blood. TIE-2 is a cell-surface receptor for angiopoietins, and
these cells are preferentially recruited to tumors and other sites
of angiogenesis (18).
Additional oncogenic effects of infection
All the mechanisms described in the previous section can
occur in tumors in the absence of infection. In the current
context, the important point is that infections induce similar
changes and so can initiate and promote tumors. Moreover
infections can encourage oncogenesis by additional mecha-
nisms described in this section.
Infection-induced DNA abnormalities
The immune response to infection can lead to local release of
ROS and nitrogen species, thereby increasing the likelihood of
oncogenic mutations. Interestingly, methylation changes
affecting CpG islands in promoters of tumor suppressor genes
like those discussed above are also found in a model of Helico-
bacter infection (28).
Chronic stimulation of lymphoid cells
Chronic stimulation by microbial antigens or autoantigens is
associated with a distinct subset of extranodal lymphomas
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
144 � 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011
arising from B cells of the marginal zone of mucosa-associ-
ated lymphoid tissue (MALT) or of the spleen (3, 29, 30).
The implicated organisms include H. pylori, Campylobacter jejuni,
Borrelia burgdorferi, Chlamydia psittaci, and hepatitis C virus
(HCV). For example, H. pylori-infected gastric mucosal cells
produce proinflammatory cytokines (such as lymphotoxin
beta) and B-cell homing factors (such as BCA-1), leading to
the emergence of MALT in the gastric mucosa. While some
of the chronic stimulation is likely to be attributable to ongo-
ing responses to microbial antigens, in several cases it seems
that the major stimulus comes from an autoimmune response
triggered by the infection. For example, neoplastic B cells
from gastric MALT lymphoma are not specific for H. pylori
antigens but rather for autoantigens found in the gastric
mucosa (3). These autoreactive B cells are thought to receive
cognate help from H. pylori-specific T cells displaying cross-
reactivity with gastric autoantigens (31, 32). Indeed autoim-
munity, even when not directly attributable to a known
infection (e.g. Sjogren’s syndrome) is associated with an
increased risk of lymphomas. Thus, several mechanisms may
account for the finding that in at least 60% of patients, eradi-
cation of H. pylori can cause regression of low-grade MALT
lymphoma, and of early stage high-grade lymphomas of the
stomach (33). These mechanisms include reduction in bacte-
rial antigen load, removal of an ‘adjuvant’ stimulus for
autoimmunity, and elimination of abnormally functioning H.
pylori-infected cells.
Similar arguments apply to several of the other organisms
implicated in the induction of extranodal lymphomas. C. jejuni
is associated with Guillain–Barre syndrome, which is caused
by cross-reactive antibodies that bind both to C. jejuni lipo-
oligosaccharides and to gangliosides in the nervous system
(34). Similarly, the OspA protein of B. burgdorferi is structurally
homologous to human lymphocyte function antigen-1
(LFA1) and may play a role in the autoimmune manifestations
and chronic activation of the immune system that leads occa-
sionally to lymphoma (35).
Damage, atrophy, metaplasia, dysplasia, and cancer
A further mechanism is involved when chronic infections lead
to chronic tissue damage. Detailed histopathological analysis
reveals the sequence of events that occurs in H. pylori-induced
adenocarcinoma of the stomach. First there is chronic active
non-atrophic gastritis, which is followed by focal loss of
glands (atrophy) particularly at the antrum-corpus junction,
with fibrous stromal tissue replacing the lost glands. Next
there is metaplasia, with the appearance of cells initially
resembling various types of differentiated small intestinal epi-
thelial cells and later resembling colonic mucosa. These events
might be an early phase of dysplasia. Dysplasia is characterized
by atypical changes in nuclear morphology and tissue archi-
tecture. At some point, adenocarcinoma develops in these
lesions. Some investigators believe that the initial stages of
inflammation and atrophy create an abnormal microenviron-
ment favoring engraftment of bone marrow-derived stem cells
(BMDCs) that replace the depleted local tissue stem cells.
These BMDCs are in a distinct epigenetic state and seem sus-
ceptible to mutation and deviation away from the pathway of
complete differentiation, resulting in progressive loss of dif-
ferentiation, uncontrolled replication, and eventual neoplastic
invasive behavior (36). If this is correct, then infection-
induced chronic inflammation can cause cancer by attrition of
local stem cells, and their replacement with more dysplasia-
susceptible BMDCs.
Infections may also reduce cancer
In addition to contributing oncogenic effects in the ways
described above, infections may also reduce the risk of cancer
(Fig. 1). The mechanisms fall into three broad categories. First,
infection might boost the host’s own immune response to
cancers because of adjuvant and immunostimulatory effects.
Second, cross-reactivity between microorganisms and tumor
antigens may lead to the priming of anti-tumor effector cells.
Third, infections can modulate immunoregulatory circuits, as
recognized in relation to the Hygiene or Old Friends hypothe-
sis. These three mechanisms are discussed below. It is not
always possible to know which particular one is involved, and
it is likely that many infections induce more than one of
them.
Adjuvant effects
Although chronic infection, by causing chronic inflammation,
is likely to drive oncogenesis, transient exposure to microbial
components such as endotoxin (LPS) or transient infections
would be expected to enhance immune responses (Fig. 1). In
the mid-19th century, Coley (37) had noted remission of can-
cer in patients developing severe infections, and he attempted
to use bacterial extracts as a therapy, with some evidence of
success. This work inspired an epidemiological study of peo-
ple developing melanomas. The study concluded that the risk
of melanoma was inversely related to the numbers of their
recorded infections, particularly when accompanied by high
fevers (38). Patients with histories of one, two to three and
four or more infections had odds ratios of 0.66, 0.63, and
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
� 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011 145
0.32 respectively, and the trend was statistically significant
(P = 0.004) (38).
LPS, like many microbial components, is a powerful adju-
vant. This quality has been thought to explain why workers in
the cotton industry in Shanghai have reduced incidences of
several cancers. Their heavy exposure to LPS might be protec-
tive against cancers of the esophagus, stomach (39), breast
(40), and lung (41). Similarly, dairy workers, who are
exposed to a multitude of microbial components including
LPS, have been shown to have lower incidence of lung cancer
(42). In the 1970s, a number of authors claimed that BCG
vaccination protected against leukemia and other hematologi-
cal tumors, though the validity of the findings remains uncer-
tain (43). To date, one can only speculate as to what effect
latent tuberculosis, affecting one third of the world’s popula-
tion, has on the incidence of cancer.
Cross-reactivity
It appears that glycosylation changes are early consequences of
oncogenesis, and glycoproteins on tumor cells bear oligosac-
charides that are markedly different from those found on pro-
teins produced by normal cells (44). It is suggested that these
may sometimes resemble the oligosaccharide structures of
bacteria (45), with more exposed galactose, N-acetylgalactos-
amine, and mannose (Fig. 1). This may go some way to
explain why bacterial infections may be protective in cancer,
as they might prime immune responses to cross-reactive
epitopes on tumor cells.
Another documented cross-reactive antigen is the HERV-
K-MEL epitope expressed in many melanomas. It is encoded
by a human endogenous retrovirus of the HERV-K family.
Sequences predicted to cross-react with this epitope were
identified in BCG, vaccinia virus used for the smallpox vac-
cine, and the 17D yellow fever vaccine, all of which appear to
afford some protection against development of melanoma
(45–47).
Modulation of immunoregulatory circuits
It was pointed out above that dairy workers might be pro-
tected from ⁄have lower risk of developing lung cancer (42).
An adjuvant effect of LPS is a possible explanation, but others
may be more relevant. Exposure to cowsheds also protects
from allergic disorders (48), and this observation forms one
of the strongest epidemiological findings underpinning the
Hygiene hypothesis, outlined in the Introduction. One might
speculate that the ‘cowshed’ effect downregulates Th2
responses, terminates inappropriate chronic (potentially
oncogenic) inflammation, and so simultaneously reduces
oncogenesis, and improves the potential for Th1-dominated
anti-cancer responses. The implications of the Hygiene
hypothesis for cancer are expanded in the next section.
Hygiene hypothesis
The Hygiene hypothesis seeks to explain the rapid simulta-
neous increases in the incidences of several types of chronic
inflammatory disease in the rich developed countries. These
include autoimmunity [e.g. multiple sclerosis (MS) and Type
1 diabetes] and inflammatory bowel diseases (IBDs) (e.g. Cro-
hn’s disease and ulcerative colitis) (49). All of these chronic
inflammatory disorders show evidence of defective immuno-
regulation. Regulatory defects have been found in individuals
suffering from allergic disorders (50), some autoimmune dis-
eases (51, 52), and probably in IBD too (53, 54). This sug-
gests that in developed countries there is a generalized defect
in regulation of the immune system that is manifested in dif-
ferent individuals as a failure to terminate inappropriate
responses to allergens (allergy), self (autoimmunity such as
MS or Type 1 diabetes), or gut contents (IBD). This can be
explained by an environmental effect operating at the popula-
tion level, with superimposed genetic effects operating at the
level of the individual (Fig. 2). We know that a generalized
dysfunction of immunoregulatory mechanisms can lead to
simultaneous increases in these diverse types of pathology,
because genetic defects of forkhead box protein 3 (Foxp3), a
transcription factor that plays a crucial role in the development
and function of regulatory T cells (Tregs), leads to a syndrome
known as X-linked autoimmunity–allergic dysregulation syn-
drome that includes aspects of allergy, autoimmunity, and
enteropathy (55). A recent reformulation of the Hygiene
hypothesis (Old Friends Hypothesis) suggests that certain
organisms that were constantly present throughout mamma-
lian evolutionary history had to be tolerated and so co-evolved
a role in driving regulatory pathways within the immune sys-
tem. The depletion of these organisms from the modern envi-
ronment is thought to be contributing to the increase in
immunoregulatory disorders (1).
The organisms involved in the Hygiene hypothesis
Common infections of childhood do not protect against aller-
gies (56–58), Type 1 diabetes (59), or IBD (60–62). Rather
epidemiological, experimental, and clinical studies have indi-
cated that the important organisms in this context are rela-
tively benign organisms such as lactobacilli, helminths,
orofecally transmitted infections inducing carrier states, and
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
146 � 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011
saprophytes from mud and untreated water (1). All have
long-standing associations with humans but have become
depleted by modern lifestyles. These organisms are listed and
referenced in Table 2. They were all present from very early in
human evolution and became depleted during the second epi-
demiological transition: during industrialization, the con-
struction of modern cities, concrete, tarmac roads, and
separation from mud animals and feces that had always previ-
ously provided contact with them (Fig. 3).
Pro-inflammatory effects of previously beneficial genetic
factors
The depletion of regulation-inducing organisms leads to a fur-
ther complication. In the developed countries, genetic factors
that were not previously causing disease can assume negative
roles when the Old Friends are removed. In areas with the
highest burdens of immunoregulation-inducing organisms,
accumulation of compensatory single nucleotide polymor-
phisms (SNPs) or other genetic variants tended to increase
release of pro-inflammatory cytokines (63–65) or enabled
enhanced Th2 responses to help control parasite burdens
(66). However, as soon as the immunoregulatory organisms
(Figs 2 and 3) are removed, these SNPs lead to exaggerated
responses, and become risk factors for chronic inflammatory
disorders (Fig. 4).
Hygiene hypothesis and cancer
As summarized above, the Hygiene hypothesis suggests that
the modern lifestyle is leaving the immune system with inade-
quate immunoregulatory mechanisms, causing, in some indi-
viduals, increased susceptibility to a variety of chronic
inflammatory disorders. As many cancers are increasing in
incidence too, it is interesting to speculate about a possible
connection. It is usually assumed that cancer is associated with
excessive immunoregulation rather than too little, because
many experimental systems have shown detrimental effects of
Tregs in established tumors. However in view of the clear role
of inflammation in oncogenesis, there is the obvious possibil-
ity that mechanisms described by the Old Friends Hypothesis
may be protective by terminating inappropriate inflammation,
Fig. 2. The Hygiene or Old Friends hypothesis. The organisms with which mammals co-evolved (see Fig. 3) had to be tolerated, and mammals arenow in state of evolved dependence on their ability to drive immunoregulatory pathways. In developed countries, immunoregulation is thereforedecreased, and gene–environment interactions cause a subset of people living in developed countries to develop chronic inflammatory disorders. Thenature of the disorder (if any) depends on the genetic background of the individual.
Table 2. Some organisms implicated in the ‘Old Friends’ hypothesis
Organism References
Hepatitis A virus (150, 151)Helminths (152, 153)Helicobacter pylori (117)Salmonella (154)Gut microbiota (123, 155)Mycobacterium tuberculosis (156, 157)Toxoplasma gondii (57, 151, 158)(unknown ? microbial)* (48, 159)
*Protection from cowsheds and homes contaminated by animal feces.
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
� 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011 147
which is a contributing factor to the initiation of some cancers
and to the accelerated growth of others (Fig. 1). It might be
very significant that non-steroidal anti-inflammatory agents,
such as cycloxygenase-2 (COX-2) inhibitors, reduce both the
risk of developing certain cancers (such as colon and breast
cancer) and the mortality caused by these same cancers (67).
Fig. 3. The Darwinian background to the Hygiene or Old Friends hypothesis. The organisms implicated by epidemiology, clinical studies, in vitroexperiments, and animal models are shown highlighted in bold font on the right. These organisms accompanied man at least as far back as the Paleo-lithic era and were part of mammalian evolutionary history. They were still present, often in larger quantities, after the first epidemiological transition(Neolithic). Epidemiological studies indicate that the new sporadic infections picked up from domesticated animals following animal husbandry inthe Neolithic are not so relevant. The important organisms were progressively lost from the start of the second epidemiological transition in the early19th century.
present from the paleolithic Genetics
Potential susceptibility to Th1/Th17 or Th2 disorders in the
Compensatory SNP or alleles 1) pro-inflammatory cytokines
2) IgE
Environmental triggers
on their role in induction of immunoregulation
progressively from mid 18th C as contact withsoil, animals and faeces diminished
Delayed exposure to viruses ?
Allergies, MS, IBD, type 1 diabetes etc and some cancers?
Diet, obesity
Factors not usually sufficient to cause disease
Vit D deficiency,
Pollutants (dioxins) etc
Fig. 4. The additional effects of compensatory single nucleotide polymorphisms (SNPs) and of aggravating environmental changes. In parts ofthe world with high levels of immunoregulatory Old Friends, there has been selection for compensatory SNPs that restores the inflammatory response.This has established an equilibrium that is disturbed when the Old Friends are lost during the second epidemiological transition. The same SNPs thenbecome risk factors for chronic inflammatory disorders. This situation is likely to be aggravated by other modern environmental changes that also tendto increase inflammatory responses, such as obesity, vitamin D deficiency, and pollutants such as dioxins that drive Th17 cells.
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
148 � 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011
This suggestion would be strengthened if there were evidence
that anti-inflammatory cytokines of the types associated with
immunoregulatory pathways were found to be protective.
Effects of IL-10 on cancer
In IL-10) ⁄ ) mice, severe inflammation occurs in the gut as
soon as colonization with normal microbiota occurs. This
inflammation is mediated by CD4+ Th1-like cells and acti-
vated macrophages and is attenuated by administering anti-
IFN-c antibody. Eventually, the inflammatory lesions resemble
early adenocarcinomas. Restoration of IL-10 is protective:
IL-10) ⁄ ) mice treated with exogenous IL-10 daily from 3 weeks
of age remained free of disease. Similarly, administration of
IL-10 prevented the colitis that occurs in 100% of SCID mice
transplanted with CD45RBhi CD4+ effector T cells. Interest-
ingly, administration of IL-10 to adult IL-10) ⁄ ) mice after the
colitis was established did not cure the colitis but it still very
significantly reduced the incidence of adenocarcinomas (68).
These results imply that in addition to the anti-inflamma-
tory effect, IL-10 is also involved in other protective anti-
tumor mechanisms, and further studies do indeed confirm
this finding. First, there is strong evidence that IL-10 encour-
ages NK cell-mediated tumor cell killing by downregulating
MHC class I, modulating NK cells, and inhibiting angiogenesis
(69–72). It has also been reported that systemic administra-
tion of IL-10 induced an effective specific immune response
against established tumors in mice. IL-10-treated mice were
immune to subsequent rechallenge with 10-fold larger num-
bers of the same but not a different tumor (73). This work in
mice with laboratory tumor cell lines might be too artificial
for extrapolation to humans, but it does imply that IL-10
encouraged an adaptive immune response. In conclusion, it is
not surprising that the presence of high levels of IL-10 in the
serum or within the tumor can be a favorable prognostic indi-
cator for some tumors but an unfavorable one for others (72).
Effects of TGF-b on cancer
The effects of TGF-b on cancer remain controversial. Many
tumors secrete TGF-b, a potent suppressor of the cytotoxic
T-cell response, a crucial anti-tumor effector mechanism,
thereby promoting tumor growth (74, 75). In addition, there
is evidence to suggest that Tregs inhibit the anti-tumor func-
tions of NK cells by expressing membrane-bound TGF-b
(76). Moreover, TGF-bs are involved in epithelial to mesen-
chymal transdifferentiation. The switch of carcinoma cells to
an invasive phenotype has some similarities to this normal
process, and TGF-b can induce epithelial to mesenchymal
transition in vitro in numerous types of cancer cells (77). Simi-
larly, control of angiogenesis is a normal function of TGF-b,
partly via induction of VEGF, which explains why knockout of
the genes encoding TGF-b1 or TGFbRII is lethal to embryos.
It is therefore not surprising that neutralizing antibodies to
TGF-b can reduce angiogenesis in tumors (77). Hence, there
are several ways in which TGF-b might promote the growth
of established tumors.
TGF-b may, however, have a different role in tumor initia-
tion. The broad anti-inflammatory effects of TGF-b might be
expected to counter inflammation-induced oncogenesis
described earlier. Similarly, TGF-b is a potent inhibitor of
growth of epithelial cells and plays a role in the homoeostasis
of healthy epithelial surfaces (77). Indeed mutations in
human TGFBR2 and in SMAD2 and SMAD4, which are
involved in transduction of signals from the activated TGF-b
receptors, have been found in both sporadic and inherited
colon cancer. Experiments in TGF-b) ⁄ ) mice cast some light
on the mechanism. Germ-free TGF-b) ⁄ )RAG2) ⁄ ) mice do
not develop colitis or cancer. However, when exposed to H.
hepaticus, both colitis and cancer were observed (78, 79). Inter-
estingly, RAG2) ⁄ ) mice, which retain the ability to secrete
TGF-b, remain cancer-free even in the presence of H. hepaticus,
indicating that TGF-b can block oncogenic effects of the
H. hepaticus-induced inflammation.
In conclusion, the available data are compatible with the
view that TGF-b may protect against the initiation of cancer
via its effects on cellular homeostasis and its anti-inflamma-
tory roles. In contrast, TGF-b might promote growth of
tumors once these are established.
Effects of Tregs on cancer
There have been numerous reviews of the detrimental role of
Tregs in progressive established cancer (75), and other
reviews of the ways in which cancer therapy might be
improved by drugs that oppose the functions or development
of Tregs (80). However, since regulatory cytokines can be
shown to have protective effects against the induction of can-
cers and, in the case of IL-10, some protective effects even
after the tumor has formed, we should expect to find evidence
that Tregs can protect against oncogenesis, perhaps particu-
larly in relation to cancers that are known to be associated
with uncontrolled chronic inflammation (Fig. 5).
Beneficial role for Tregs in animal models of cancer
An example of the protection from oncogenesis by Tregs was
seen in C57BL ⁄6 mice infected with Helicobacter felis. This
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
� 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011 149
organism induces Th1- and IFN-c-dependent gastric atrophy,
a premalignant lesion reminiscent of that which can occur in
humans infected with H. pylori. However, when H. felis-
infected mice were co-infected with Heligmosomoides polygyrus, a
natural murine nematode parasite with a strictly enteric life
cycle, there was minimal gastritis, and the premalignant atro-
phic lesions were not seen (81). The co-infection did not
eliminate inflammation. Indeed, 16 weeks after infection gas-
tric colonization with H. felis was greater in the group co-
infected with H. polygyrus, particularly in the antrum, with
extension into the corpus mucosa. However, the inflamma-
tion in the co-infected mice was quite different, with
strikingly reduced IFN-c and increased expression of IL-10,
TGF-b, and some Th2 cytokines (81). A similar effect was
revealed using Helicobacter hepaticus, another natural enteric bac-
terial pathogen of mice. 129 ⁄ SvEv RAG-2-deficient mice were
infected with H. hepaticus by gastric gavage. Many of these ani-
mals develop IBD and colon cancer. However, when the
H. hepaticus-infected mice were treated at 1, 3, or 12 months
after infection with adoptive transfer of CD4+CD45RBloCD25+
regulatory cells, they had reduced severity of IBD and
significantly lower risk of colon cancer (82), but only if the
Tregs came from IL-10-competent donors. When Tregs from
IL-10) ⁄ ) mice were used, there was an increased incidence
of mucinous tumors (82). Thus, these models show that
induction of Tregs by co-infection with a helminth (a mecha-
nism discussed above in relation to the Hygiene hypothesis)
or passive transfer of Tregs can convert an inflammatory
response from a pattern that causes cancer to one that does
not.
The same group also investigated the APCMin ⁄ + mouse. APC
(adenomatosis polyposis coli) is a human gene that causes
colorectal cancer when it is mutated. APCMin mice have a
point mutation in the murine homologue, and they develop
multiple intestinal adenomas and mammary tumors. How-
ever, the number of epithelial adenomas is reduced when het-
erozygote APCMin ⁄ + mice are treated with CD4+CD25+ Tregs,
but as in the H. felis model discussed above, only if the Tregs
come from IL-10-competent donors (83).
Further work has confirmed the requirement for IL-10
but also revealed that not all Tregs have the same ability to
inhibit oncogenesis (Fig. 6). Tregs transferred from donors
that had been preinfected with H. hepaticus protected against
development of cancer in several models (23). The models
were: (i) colorectal cancer (CRC) developing from inflam-
matory bowel disease in H. hepaticus-infected RAG2) ⁄ )
129 ⁄ SvEv mice; (ii) CRC and mammary adenocarcinoma in
H. hepaticus-free APCMin ⁄ + mice given pro-inflammatory
effector T cells; and (iii) spontaneous mammary adenocar-
cinoma in FBV Her2 ⁄ Neu mice. In each case, even when
no Helicobacter was present in the recipient (i.e. the cancer
was spontaneous, or triggered by transfer of pro-inflamma-
tory T cells), only Tregs from Helicobacter-experienced mice
were protective (Fig. 6). When derived from H. hepaticus-
naive ‘hygienic’ animals, they exacerbated the cancer in the
first two models and had no effect in the third one. The
Treg from animals that have experienced infection
Beneficial roles of Treg Detrimental roles of Treg
Treg in tumour stroma or local lymphoid aggregates block development of protective cytotoxic lymphocyte (CTL) response.
Treg block oncogenesis in site of chronic inflammation.
Treg in tumour bed reduce inflammation, and so block release of growth and angiogenic factors.
IL-10: Anti-inflammatory Increase NK activity Inhibit angiogenesis TGF-ββ: Anti-inflammatory Inhibit epithelial growth
Treg from ‘hygienic’ animals
IL-10: Inhibit protective anti-tumour response
TGF-ββ: Inhibit CTL
Fig. 5. The heterogeneous roles of Tregs and anti-inflammatory cytokines in cancer. It is well documented that Tregs can oppose protectiveresponses to cancer. It is now emerging that Tregs of certain poorly defined phenotypes, in certain locations, can be protective against both initiationand progression of cancer. Tregs can terminate chronic inflammation and so stop oncogenesis. In existing tumors, Tregs can block release of growthand angiogenic factors, but to achieve this, they need to be in the tumor bed itself. Tregs located in surrounding lymphoid aggregates are more likelyto inhibit generation of the protective T-cell response. We still know little about the control of the localization or phenotype of these Tregs, but in ani-mal models (see Fig. 6), Tregs from infected animals can have the relevant properties, whereas Treg from very clean specific-pathogen free (SPF) ani-mals do not.
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
150 � 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011
exacerbation might have been due to inadequate release of
IL-10, to conversion to Th17, or to a change in homing.
This implies a functional difference in the Tregs, indepen-
dent of their specificity.
In conclusion, this series of experiments has shown that a
gut-associated organism (H. hepaticus) can induce and modulate
Tregs in ways that endow them with the ability to control can-
cer-promoting inflammation throughout the body (i.e. in the
breast as well as the gut) via an IL-10-dependent mechanism.
This constitutes experimental evidence for the relevance of the
Hygiene hypothesis to cancer.
Location and function of Tregs in cancer
The strategies that are effectively anti-tumor in the models
discussed in the previous section also caused a change in the
location of Foxp3+ Treg cells. Tumor progression was associ-
ated with Foxp3+ cells in lymph nodes and in tissue surround-
ing the tumor, whereas remission was associated with Foxp3+
cells in the tumor itself (23). Female FVB strain MMTV-HER-
2 ⁄neu (ErbB2) transgenic mice have a genetically determined
tendency to develop mammary tumors resembling those seen
in 30% of humans with breast cancer. Growth of the tumors
requires ongoing inflammation. Working with mice already
bearing tumors, it was found that administration of anti-TNF
significantly decreased subsequent tumor growth. Interest-
ingly, the anti-TNF also caused accumulation of Foxp3+ cells
in the tumor itself, whereas in untreated animals with more
rapidly progressing tumors, there were no Foxp3+ cells in the
tumor but instead an accumulation in the lymph nodes and
tissue surrounding the tumor. These authors observed a simi-
lar phenomenon in the APCMin ⁄ + mouse model (23).
This observation might be relevant to human disease. The
assessment of numbers of circulating Tregs or of Tregs infil-
trating the tumor itself has been used as a prognostic indicator
in human cancers, though the correlations found depend on
the tumor studied. Their presence has been claimed to be a
negative prognostic indicator in carcinomas of the lung, pan-
creas, stomach, liver, and ovary, whereas Tregs appeared ben-
eficial in B-cell lymphoma, head and neck, and colon
carcinoma, though in other studies numbers of Tregs have not
correlated with changes in survival (reviewed in 84). This het-
erogeneity of results suggests that something is being missed,
and the animal models discussed above suggest that the locali-
zation of the Tregs might be crucial. A study that counted
infiltrating Tregs in breast cancer concluded that high num-
bers of these cells were detrimental to survival (85), but a fur-
ther study that recorded the precise location of the Tregs
concluded that the presence of Foxp3+ Tregs within the tumor
bed did not affect the patients’ clinical evolution, whereas
their presence in lymphoid-enriched areas correlated with a
higher risk of relapse, a shorter relapse-free survival, and
reduced overall survival (84). This finding agrees with the
animal models described above (23) and might suggest that
Tregs in the lymphoid aggregates can inhibit the development
of effector cells such as CTLs, whereas Tregs in the tumor bed
can usefully inhibit the tumor-promoting effects of inflamma-
tion discussed earlier.
The microanatomical location is also crucial when consider-
ing the role of tumor-infiltrating macrophages. These corre-
late with improved survival if present in the tumor ‘islets’,
among the tumor cells, but are linked to a poor prognosis if
present in the surrounding stromal tissue (86). It is possible
that stromal macrophages support tumor growth by enhanc-
ing formation of tumor stroma and angiogenesis, whereas
macrophages in contact with tumor cells are cytotoxic, and so
limit tumor growth.
As a further example, in a murine model of colon adenoma-
tous polyposis, there is focal mastocytosis that promotes
tumor initiation and progression (87, 88). Adoptively trans-
ferred Tregs suppress this mastocytosis. In contrast, the
endogenous Tregs that are present in the polyps in large num-
bers show proinflammatory activity and promote differentia-
tion and expansion of mast cells (87, 88). It is known that
Tregs can be incapacitated by exposure to TNF (89) or IL-6
(90), so it may be that when Tregs fail to suppress inflamma-
tion, perhaps because TNF and IL-6 are being released sponta-
neously from the tumor cells (and in this model, by the mast
cells) their function can be perverted. Indeed, Treg cells that
expand in adenomatous polyps no longer produce IL-10 and
Infect with Helicobacter hepaticus
Treg-deficient effector T cells (CD4+CD45RBhi)
C57Bl APCmin+/–
129 Rag2–/–
Colorectal cancer
Mammary adenocarcinoma
Transfer of Treg from infected mice is protective
Fig. 6. Examples of beneficial effects of Tregs in animal models ofcancer. Transfer of Tregs from animals infected with Helicobacter hepaticusinto animals developing the cancers was protective, even when this organ-ism was not involved in the cancer concerned and was not present in therecipient. By contrast, Tregs from specific pathogen-free animals were notprotective, indicating that the precise phenotype of the Treg is crucial.
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
� 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011 151
instead switch to production of the pro-inflammatory
cytokine IL-17 (87).
Treg modulation of inappropriate inflammatory
mechanisms in cancer
Tregs not only inhibit inflammatory responses but also influ-
ence the balance between different effector cell types and
selectively alter cytokine secretion profiles. For instance,
Foxp3+ Tregs can functionally differentiate to acquire an abil-
ity to specifically control Th1 or Th2 type immune responses
by modifying their expression of Th1-related or Th2-related
transcription factors (91). Thus, some natural Foxp3+ Tregs
express T-bet, which increases when they are stimulated in a
Th1 cytokine environment. This causes upregulation of C–X–
C chemokine receptor 3 and promotes recruitment to sites of
type 1 inflammation. Foxp3 also controls expression of the
transcription factor IFN regulatory factor 4 (IRF4), which is
required for Th2 differentiation. Deletion of IRF4 from Tregs
impairs their ability to control Th2- but not Th1-mediated
inflammation (92). Therefore not only is the location of the
Treg crucial but also their precise function. They need to be
able to inhibit the types of T cell that are detrimental to tumor
immunity such as Th2 cells. The inflammatory cells that infil-
trate tumors are often not well adapted to the role of driving
Th1- and CTL-mediated anti-tumor responses and show a Th2
bias. The dendritic cells tend to have an immature phenotype,
and the myelomonocytic cells express the alternative M2 phe-
notype (19). M2 macrophages are driven by signals such as
IL-4, IL-13, glucocorticoids, and IL-10. They express low
IL-12 but high IL-10, IL-1decoyR, IL-1ra, and scavenger
receptors (SR-A and mannose receptors). These cells promote
Th2 responses, scavenge debris, and drive angiogenesis and
tissue remodeling as discussed earlier (18).
Thus although there is strong evidence that Th1-mediated
inflammation has a particular tendency to initiate oncogenesis
(68), once a tumor is established, a switch to a Th2 pattern
may reduce anti-tumor mechanisms, while the Th2-associated
repair and angiogenic functions facilitate growth (18, 19). In
this context, it is interesting that several of the organisms
involved in the Hygiene hypothesis have the ability to drive
Tregs that can suppress Th2 responses. This includes M. vaccae
(93), various helminths (94–96), and H. pylori (discussed in
detail below).
Epidemiology of cancer in relation to infection
An examination of the epidemiology of some of the cancers
that increase markedly in incidence as developing countries
adopt the lifestyle of rich industrialized countries reveals
trends that mimic those that led to the invocation of the
Hygiene hypothesis to explain the rising incidence of chronic
inflammatory disorders. For instance, age-specific incidence
rates for specific cancers in Asians correlate with the state of
economic development of their country of residence. Inci-
dences are much lower in Asians living in India compared to
those living in the UK or USA (97).
Hodgkin’s lymphoma
The risk of classical Hodgkin’s lymphoma (HL) is inversely
associated with socioeconomic status (98). In the West, HL is
most prevalent among whites, followed by blacks and Hispan-
ics, with the lowest incidence in Asians. Nevertheless,
although incidence is low in Asian subgroups (presumably for
genetic reasons), rates in Asians residing in the US are almost
double the rates in Asians living in Asia, so there are clearly
environmental factors that partly compensate for this resis-
tance when Asians move to the USA (99). Early exposure to
other children at nursery school and day care seems to
decrease the risk of HL in young adults, and the risk is lowest
among those with multiple older but not younger siblings.
Having older siblings is associated with earlier exposure to
common childhood pathogens and Old Friends (100), and
this is precisely the observation that, in relation to allergies,
led to the Hygiene hypothesis (101). We know that delayed
exposure to some infections that are usually harmless in neo-
nates can result in serious disease (e.g. polio or hepatitis A
virus) and delayed exposure to viruses has also been consid-
ered by some authors in relation to the rise in Type 1 diabetes
(102). There is evidence that HL is associated with EBV,
which appears to increase the risk fourfold. Clearly the vast
majority of individuals with EBV do not develop HL (100,
103), and this association does not explain the socioeconomic
or family size data.
Acute lymphatic leukemia of childhood
The epidemiology of acute lymphatic leukemia (ALL) of
childhood also shows striking parallels with the epidemiolog-
ical findings that gave rise to the original version of the
Hygiene hypothesis (101, 104). A study in Northern Califor-
nia provided preliminary evidence that protection from ALL
is proportional to the number and frequency of social
contacts (105). A large population-based case–control study
[The UK Childhood Cancer Study (UKCCS)] revealed further
evidence that social contacts in infancy can reduce risk of
childhood ALL (106). Interestingly, early playgroup
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
152 � 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011
attendance appeared to give some protection against not only
ALL but also to a limited extent against other pediatric can-
cers such as central nervous system tumors (106, UKCCS
investigators, unpublished data quoted in 104). Equally strik-
ing is the fact that an analysis of data from 40 countries
revealed a very significant positive correlation between inci-
dences of ALL and Type 1 diabetes. Both diseases also corre-
lated positively with GDP (107). A similar analysis had
previously revealed a correlation between Type 1 diabetes
and allergic disorders (108).
Carcinoma of the stomach or esophagus
The story of H. pylori is particularly instructive, because we
know which organism is involved and we have some insight
into the mechanisms. The majority of epidemiologic studies
have reported a decreased incidence of gastric cancer in Africa
compared to other countries where H. pylori infection is ende-
mic. This is often described as ‘The African Enigma’. Looking
at IgG subclasses as an approximate measure of the relative
balance of Th1 and Th2-mediated response it was noted that
the citizens of Soweto (an economically impoverished town-
ship near Cape Town, South Africa) had a more Th2-biased
response to H. pylori than did people from Europe or Australia
(discussed in 109). An appealing hypothesis is that the
inflammatory effect of H. pylori in man is modified by simulta-
neous helminth infections, as demonstrated in experimental
models outlined earlier (23, 81). The milder chronic inflam-
mation or increased immunoregulatory networks seen in the
presence of helminths might mitigate the risk of stomach
cancer (Fig. 7).
Gastric ulcer disease due to H. pylori appeared in the 19th
century in the USA and Europe, and its prevalence peaked in
the early 20th century. Duodenal ulcer disease appeared and
peaked 10–30 years later (110, 111). This again could be due
to changes in immunoregulation; peptic ulcer disease is asso-
ciated with reduced expression of Foxp3, IL-10, and TGF-b in
gastric biopsies from H. pylori-infected donors (112). The
modern lifestyle may be associated with poorly controlled
Th1-orientated inflammation, causing more acute local dam-
age and also causing the infection to be more frequently con-
fined to the antrum of the stomach, whereas the typical
situation observed in patients in developing countries is a
milder pan-gastritis (113).
Modern hygiene and antibiotics are leading to populations
where many individuals do not carry H. pylori at all. Some
authors believe that this total absence of colonization by
H. pylori may be responsible for recent increases in the inci-
dence of short- and long-segment Barrett’s esophagus and its
malignant complications (114, 115). The reasoning here is
that in the absence of the complex biological effects of H.
pylori, which include mucosal atrophy, acid production by the
stomach is excessive, leading to chronic inflammatory esopha-
gitis and carcinogenesis (Fig. 7).
Loss of helminths ------------> --- Loss of H. pylori (antibiotics) --->
Normal (ancient) Late 20th C
Gastric atrophy - B12 deficiency - Iron deficiency - Some adeno-
carcinoma in old age
Peptic ulcer disease shows sharp peak
Sustained high acid, oesophageal adenocarcinoma
Progressive loss of Treg, more allergic disorders
Mid 19th–early 20th C
Fig. 7. The changing role of Helicobacter pylori. Human studies and animal models suggest that in the presence of co-infection with Treg-inducinghelminths, the inflammatory response induced by H. pylori is attenuated but involves the whole stomach. Adenocarcinoma is a late consequence insome individuals. As helminths are progressively eliminated, a more vigorous response occurs, and peptic ulceration is seen in individuals whose Tregresponse to H. pylori itself is suboptimal, while H. pylori becomes confined to the pylorus. More recently, the situation has been complicated by the factthat long-term infection with H. pylori results in decreased acid production. Thus, total loss of H. pylori due to antibiotics results in sustained high levelsof acid production and consequent esophageal adenocarcinoma (modified and redrawn from reference 113).
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
� 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011 153
It is important to note that H. pylori is itself a potent inducer
of Tregs, both locally in the stomach and duodenum (112,
116), and systemically (116), though Tregs are reduced in
patients who develop ulcers (112). The changes in malignant
and non-malignant manifestations of H. pylori described above
have coincided with the striking increases in allergic disorders
that prompted the Hygiene hypothesis. Interestingly, epide-
miological studies indicate that infection with H. pylori protects
against allergic disorders (57, 117), and similar effects are
seen in laboratory models (118, 119). Thus H. pylori may pro-
vide an interesting example of the parallel between the
hygiene of Old Friends hypothesis as applied to chronic
inflammatory disorders and its application to cancer.
Colon cancer
Worldwide, colorectal cancer incidence rates increase as coun-
tries undergo the second epidemiological transition to eco-
nomic development and remain high in longstanding,
economically developed countries such as Japan, Australia,
Western Europe, and North America (120). Part of this
increase is associated with the increase in inflammatory bowel
disease (IBD), in which the Hygiene hypothesis is already
implicated. As outlined above, helminth infection can reduce
colorectal cancer by inducing anti-inflammatory immunoreg-
ulation in a mouse model (81), and helminths have a thera-
peutic effect in human IBD (121, 122). However, other
components of the microbiota also have important immuno-
regulatory functions (123). The microbiota can be altered by
inappropriate diet, and obesity is associated with a distinct
pattern of changes in the gut microbiota (124). Similarly,
antibiotics have profound effects on the nature and diversity
of the intestinal microbiota (125). After a severe transient gut
infection, a different stable ecosystem can take over (dysbio-
sis), which can keep the host alive, but leads to illnesses such
as irritable bowel syndrome and IBD (126), with consequent
susceptibility to inflammation-related cancer.
Hygiene hypothesis, depression, and cancer
Depression provides another important potential link between
the Hygiene hypothesis and cancer. Depression predicts
increased mortality in cancer patients. This is true whether
depression is diagnosed before or after cancer diagnosis and
remains after controlling for confounding medical variables
(127). An analysis of 25 independent studies reported that
mortality rates were up to 25% higher in patients experienc-
ing depressive symptoms (P < 0.001) and up to 39% higher
in patients diagnosed with major or minor depression
(P = 0.03) (128). Chronically raised levels of circulating pro-
inflammatory cytokines can lead to depressive symptoms, and
there is a subset of individuals with depression in whom the
mood disorder may be caused by the same failure to terminate
inappropriate inflammatory responses that lies behind the
increases in somatic inflammatory disorders (allergies, IBD,
autoimmunity) frequently discussed in relation to the
Hygiene hypothesis (129, 130). Several anti-inflammatory
agents such as anti-TNF and NSAIDs have now been found to
exert anti-depressant effects (131). Interestingly, in an animal
model heat-killed Mycobacterium vaccae, a potent inducer of Tregs
(93), had anti-depressant effects similar to those of Prozac,
and a subset of serotonergic neurons involved in this effect
was identified (132). We can therefore postulate that by
inducing regulatory mechanisms and reducing circulating
pro-inflammatory cytokines, the microbial exposures involved
in the Hygiene hypothesis may exert a supplementary
anti-cancer effect, secondary to improvements in mood and
coping.
Microorganisms and immunotherapy of cancer
The arguments and data presented above suggest that
microbes or their components could have a role to play in
cancer prevention by exerting immunoregulatory effects that
suppress chronic oncogenic inflammation. Similarly, these
preparations might aid treatment of established cancer by
altering the balance of effector mechanisms, suppressing those
that are unhelpful or detrimental (Th2), and driving effective
anti-tumor effector pathways. In principle it might be possible
to drive subsets of Tregs that block inflammation in the tumor
bed (and so limit vascularization, and the provision of growth
factors), without blocking the generation of effector cells in
the lymphoid tissue (84). This aspect has not yet been investi-
gated.
Many research groups are trying to use microorganisms or
their components as adjuvants and as carriers of tumor target
antigens. The organisms selected are often intracellular infec-
tions. Thus, Listeria monocytogenes is being developed as a carrier
of tumor antigens that enters APCs (133, 134), and compo-
nents of Toxoplasma gondii are being used in a dendritic
cell-based therapeutic cancer vaccine (135). However, most
studies on the use of microorganisms for immunotherapy in
human have used mycobacteria.
Cancer immunotherapy with mycobacteria
The best-established use of immunotherapy is intravesical
BCG for bladder cancer. The mode of action is unclear, but a
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
154 � 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011
recent review of 80 randomized trials and 11 meta-analyses
concluded that intravesical BCG, although more toxic, is more
effective than mitomycin-C for the prophylaxis of tumor
recurrence in high-risk patients, although there is no signifi-
cant difference in overall survival. Chemotherapy and BCG
seem to be equally effective in preventing recurrence in inter-
mediate risk groups (136). BCG has been tested in many other
cancer types and is still sporadically used in melanoma. An
examination of all reported trials available found a trend to
enhanced survival in patients with stage III malignant mela-
noma and cutaneous non-visceral metastases given intrale-
sional BCG (137).
More recently, cancer studies have been concentrating on a
different mycobacterium (M. vaccae) that has unusual immu-
nological properties. Used as a heat-killed preparation, it
drives Tregs and inhibits Th2 (93) but encourages cytotoxic
T cells (138) and NK activity (139). Encouraged by several
studies of heat-killed M. vaccae in animal models of various
cancers, by Phase II studies on intradermal M. vaccae in pros-
tate cancer (140), and in inoperable non-small cell lung can-
cer and mesothelioma (141), an open label Phase III trial
was performed in a total of 419 patients with several types
of lung cancer, mostly squamous cell carcinoma and adeno-
carcinoma (142). Deficiencies in the study design led to poor
compliance, with only 53% receiving more than two injec-
tions of M. vaccae, and the analysis plan did not allow for
independent appraisal of the different cancer types. Overall
there was no survival advantage in the group receiving
M. vaccae, but patients showed significantly less deterioration
in their Global Health Status score (P = 0.02), particularly
those parameters related to mood, coping, and depression
(see previous section). Meanwhile, a separate analysis of
compliant patients confirmed that while treatment with M.
vaccae gave no survival benefit in squamous cell carcinoma, in
sharp contrast, survival of compliant adenocarcinoma patients
receiving M. vaccae was increased by a mean of 135 days
(P = 0.0009, Kaplan–Meier log rank test) (143). This analy-
sis fell outside the original analysis plan of the Phase III trial
and must therefore be treated with caution, but it has been
sufficient to motivate further trials that are now being
planned.
Heat-killed M. vaccae was also tested as an immunotherapeu-
tic agent for patients with metastatic, postnephrectomy, renal
cell carcinoma. Sixty patients received injections of M. vaccae
intradermally on entry and after 2, 4, 8, 12, 16, 20, and
24 weeks. Subsequently, they received injections every
8 weeks. Survival was compared with that of historical con-
trols who had been treated either with biological response
modifiers (IL-2, IFN-a) or with chemotherapy. Those treated
with M. vaccae alone survived as long as those receiving IL-2 or
IFN-a, and both treatment groups survived longer than those
on chemotherapy (P < 0.001) (144). In a second small study,
36 patients were randomized to receive treatment with IL-2
alone or IL-2 plus M. vaccae. Those receiving IL-2 alone had sig-
nificantly more adverse events than those receiving M. vaccae
plus IL-2 (P < 0.001) (144). This combination therapy was
prompted by a previous pilot study suggesting that the combi-
nation of M. vaccae and IL-2 might be more beneficial than
M. vaccae alone in melanoma (145). A small biotech startup
(Immodulon; http://www.immodulon.com/index.php) is
currently setting up trials of cancer immunotherapy using a
very similar organism.
Conclusions
In this review, we have brought together information from
several different disciplines. Until now, the thinking that has
been applied to the increases in chronic inflammatory disor-
ders in rich developed countries has been entirely different
from the thinking applied to similar increases in many can-
cers, despite rather precise epidemiological parallels. Here we
have discussed the ability of infections to cause cancer but
also pointed out that the immunodysregulation that accom-
panies the Western lifestyle might also contribute both to
oncogenesis and to subsequent inability to control or elimi-
nate incipient cancers. A greater understanding of these inter-
actions might pave the way to improved microbe-based
immunotherapies. Interestingly, the intracellular infections
are central to this analysis and particularly to therapeutic
studies currently in progress, and these results may provide
valuable new tools.
An obvious limitation of this discussion is the fact that most
of the evidence that Treg and anti-inflammatory cytokines can
sometimes protect against development or progression of
cancer comes from animal models. We cannot yet state with
certainty that the same is true for any human cancers. Never-
theless, the situation is changing, and as attention is focused
more precisely on the exact location and phenotype of the
Treg in human disease, the heterogeneity is beginning to
emerge, and evidence of protective roles in some cancers is
becoming suggestive. We hope that this review will encour-
age further explorations of this area which has the potential to
influence both the prevention and treatment of those cancers
for which it proves relevant.
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
� 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011 155
References
1. Rook GAW. 99th Dahlem Conference on
Infection, Inflammation and ChronicInflammatory Disorders: Darwinian medi-
cine and the ‘hygiene’ or ‘old friends’hypothesis. Clin Exp Immunol
2010;160:70–79.2. Parkin DM. The global health burden of
infection-associated cancers in the year2002. Int J Cancer 2006;118:3030–3044.
3. Suarez F, Lortholary O, Hermine O, Lecuit
M. Infection-associated lymphomas derivedfrom marginal zone B cells: a model of anti-
gen-driven lymphoproliferation. Blood2006;107:3034–3044.
4. de Martel C, Franceschi S. Infections andcancer: established associations and new
hypotheses. Crit Rev Oncol Hematol2009;70:183–194.
5. Herrera LA, Benitez-Bribiesca L, Mohar A,Ostrosky-Wegman P. Role of infectious dis-
eases in human carcinogenesis. Environ MolMutagen 2005;45:284–303.
6. van Lier EA, van Kranen HJ, van Vliet JA,Rahamat-Langendoen JC. Estimated number
of new cancer cases attributable to infectionin the Netherlands in 2003. Cancer Lett
2008;272:226–231.7. Chene A, et al. Endemic Burkitt’s lymphoma
as a polymicrobial disease: new insights onthe interaction between Plasmodium falciparum
and Epstein–Barr virus. Semin Cancer Biol2009;19:411–420.
8. Randi G, Franceschi S, La Vecchia C. Gall-bladder cancer worldwide: geographical
distribution and risk factors. Int J Cancer2006;118:1591–1602.
9. Dacosta NA, Kinare SG. Association of lungcarcinoma and tuberculosis. J Postgrad Med
1991;37:185–189.10. Engels EA, et al. Tuberculosis and subse-
quent risk of lung cancer in Xuanwei,China. Int J Cancer 2009;124:1183–1187.
11. Semino-Mora C, Doi SQ, Marty A, Simko V,
Carlstedt I, Dubois A. Intracellular and inter-stitial expression of Helicobacter pylori viru-
lence genes in gastric precancerousintestinal metaplasia and adenocarcinoma. J
Infect Dis 2003;187:1165–1177.12. Colotta F, Allavena P, Sica A, Garlanda C,
Mantovani A. Cancer-related inflammation,the seventh hallmark of cancer: links to
genetic instability. Carcinogenesis2009;30:1073–1081.
13. O’Byrne KJ, Dalgleish AG, Browning MJ,Steward WP, Harris AL. The relationship
between angiogenesis and the immuneresponse in carcinogenesis and the progres-
sion of malignant disease. Eur J Cancer2000;36:151–169.
14. Balkwill F. Tumour necrosis factor and can-cer. Nat Rev Cancer 2009;9:361–371.
15. Aggarwal BB, Shishodia S, Sandur SK,
Pandey MK, Sethi G. Inflammation andcancer: how hot is the link? Biochem
Pharmacol 2006;72:1605–1621.16. O’Byrne KJ, Dalgleish AG. Chronic immune
activation and inflammation as the cause ofmalignancy. Br J Cancer 2001;85:473–483.
17. Porta C, et al. Cellular and molecular path-ways linking inflammation and cancer.
Immunobiology 2009;214:761–777.
18. De Palma M, Naldini L. Tie2-expressingmonocytes (TEMs): novel targets and vehi-
cles of anticancer therapy? Biochim BiophysActa 2009;1796:5–10.
19. Mantovani A, Allavena P, Sica A, Balkwill F.Cancer-related inflammation. Nature
2008;454:436–444.20. Kim S, et al. Carcinoma-produced factors
activate myeloid cells through TLR2 to stim-ulate metastasis. Nature 2009;457:102–
106.21. Bielas JH, Loeb KR, Rubin BP, True LD,
Loeb LA. Human cancers express a mutatorphenotype. Proc Natl Acad Sci USA
2006;103:18238–18242.22. Jones PA, Baylin SB. The epigenomics of
cancer. Cell 2007;128:683–692.23. Erdman SE, et al. Unifying roles for regula-
tory T cells and inflammation in cancer. IntJ Cancer 2010;126:1651–1665.
24. Geiger TR, Peeper DS. Metastasis mecha-nisms. Biochim Biophys Acta 2009;
1796:293–308.25. Abdelrahim M, Konduri S, Basha R, Philip
PA, Baker CH. Angiogenesis: an update andpotential drug approaches (review). Int J
Oncol 2010;36:5–18.26. Wu Y, Zhou BP. TNF-alpha ⁄ NF-kap-
paB ⁄ Snail pathway in cancer cell migrationand invasion. Br J Cancer 2010;102:639–
644.27. Orosz P, Echtenacher B, Falk W, Ruschoff J,
Weber D, Mannel DN. Enhancement of
experimental metastasis by tumor necrosisfactor. J Exp Med 1993;177:1391–1398.
28. Niwa T, et al. Inflammatory processes trig-gered by Helicobacter pylori infection cause
aberrant DNA methylation in gastric epithe-lial cells. Cancer Res 2010;70:1430–1440.
29. Lecuit M, et al. Immunoproliferative smallintestinal disease associated with Campylobac-
ter jejuni. N Engl J Med 2004;350:239–248.30. Ferreri AJ, et al. Evidence for an association
between Chlamydia psittaci and ocular adnexallymphomas. J Natl Cancer Inst 2004;
96:586–594.31. Hussell T, Isaacson PG, Crabtree JE, Spencer
J. The response of cells from low-gradeB-cell gastric lymphomas of mucosa-associ-
ated lymphoid tissue to Helicobacter pylori.Lancet 1993;342:571–574.
32. D’Elios MM, Bergman MP, Amedei A,
Appelmelk BJ, Del Prete G. Helicobacter pyloriand gastric autoimmunity. Microbes Infect
2004;6:1395–1401.33. Chen LT, et al. Long-term results of anti-
Helicobacter pylori therapy in early-stage gastrichigh-grade transformed MALT lymphoma. J
Natl Cancer Inst 2005;97:1345–1353.34. Yuki N. Human gangliosides and bacterial
lipo-oligosaccharides in the development of
autoimmune neuropathies. Methods MolBiol 2010;600:51–65.
35. Singh SK, Girschick HJ. Lyme borreliosis:from infection to autoimmunity. Clin
Microbiol Infect 2004;10:598–614.36. Correa P, Houghton J. Carcinogenesis of
Helicobacter pylori. Gastroenterology2007;133:659–672.
37. Nauts HC, McLaren JR. Coley toxins – thefirst century. Adv Exp Med Biol 1990;
267:483–500.38. Kolmel KF, et al. Infections and melanoma
risk: results of a multicentre EORTC case-control study. European Organization for
Research and Treatment of Cancer. Mela-noma Res 1999;9:511–519.
39. Wernli KJ, et al. Occupational risk factorsfor esophageal and stomach cancers among
female textile workers in Shanghai, China.Am J Epidemiol 2006;163:717–725.
40. Ray RM, et al. Occupational exposures andbreast cancer among women textile workers
in Shanghai. Epidemiology 2007;18:383–392.
41. Astrakianakis G, et al. Lung cancer riskamong female textile workers exposed to
endotoxin. J Natl Cancer Inst 2007;99:357–364.
42. Mastrangelo G, Grange JM, Fadda E, FedeliU, Buja A, Lange JH. Lung cancer risk: effect
of dairy farming and the consequence ofremoving that occupational exposure. Am J
Epidemiol 2005;161:1037–1046.
43. Hoover RN. Bacillus Calmette-Guerin vacci-nation and cancer prevention: a critical
review of the human experience. Cancer Res1976;36:652–654.
44. Drake PM, et al. Sweetening the pot: addingglycosylation to the biomarker discovery
equation. Clin Chem 2010;56:223–236.45. Grange JM, Krone B, Stanford JL. Immuno-
therapy for malignant melanoma – tracingAriadne’s thread through the labyrinth. Eur
J Cancer 2009;45:2266–2273.46. Krone B, Kolmel KF, Henz BM, Grange JM.
Protection against melanoma by vaccinationwith Bacille Calmette-Guerin (BCG) and ⁄ or
vaccinia: an epidemiology-based hypothesison the nature of a melanoma risk factor and
its immunological control. Eur J Cancer2005;41:104–117.
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
156 � 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011
47. Mastrangelo G, et al. Does yellow fever 17Dvaccine protect against melanoma? Vaccine
2009;27:588–591.48. Riedler J, et al. Exposure to farming in early
life and development of asthma and allergy:a cross-sectional survey. Lancet
2001;358:1129–1133.49. Bach JF. The effect of infections on suscepti-
bility to autoimmune and allergic diseases.N Engl J Med 2002;347:911–920.
50. Akdis M, et al. Immune responses in healthyand allergic individuals are characterized by
a fine balance between allergen-specific T-regulatory 1 and T-helper 2 cells. J Exp Med
2004;199:1567–1575.
51. Viglietta V, Baecher-Allan C, Weiner HL,Hafler DA. Loss of functional suppression by
CD4+CD25+ regulatory T cells in patientswith multiple sclerosis. J Exp Med
2004;199:971–979.52. Kriegel MA, Lohmann T, Gabler C, Blank N,
Kalden JR, Lorenz HM. Defective suppressorfunction of human CD4+ CD25+ regulatory
T cells in autoimmune polyglandular syn-drome type II. J Exp Med 2004;199:1285–
1291.53. Kraus TA, Toy L, Chan L, Childs J, Mayer L.
Failure to induce oral tolerance to a solubleprotein in patients with inflammatory
bowel disease. Gastroenterology2004;126:1771–1778.
54. Powrie F, Read S, Mottet C, Uhlig H, MaloyK. Control of immune pathology by regula-
tory T cells. Novartis Found Symp2003;252:92–98.
55. Wildin RS, Smyk-Pearson S, Filipovich AH.Clinical and molecular features of the
immunodysregulation, polyendocrinopa-thy, enteropathy, X linked (IPEX) syn-
drome. J Med Genet 2002;39:537–545.56. Benn CS, Melbye M, Wohlfahrt J, Bjorksten
B, Aaby P. Cohort study of sibling effect,infectious diseases, and risk of atopic der-
matitis during first 18 months of life. BMJ2004;328:1223.
57. Matricardi PM, et al. Exposure to foodborneand orofecal microbes versus airborne
viruses in relation to atopy and allergic
asthma: epidemiological study. Br Med J2000;320:412–417.
58. Dunder T, Tapiainen T, Pokka T, Uhari M.Infections in child day care centers and later
development of asthma, allergic rhinitis,and atopic dermatitis: prospective follow-
up survey 12 years after controlled random-ized hygiene intervention. Arch Pediatr
Adolesc Med 2007;161:972–977.59. Cardwell CR, Carson DJ, Yarnell J, Shields
MD, Patterson CC. Atopy, home environ-ment and the risk of childhood-onset type
1 diabetes: a population-based case-con-trol study. Pediatr Diabetes 2008;9:191–
196.
60. Amre DK, et al. Investigating the hygienehypothesis as a risk factor in pediatric onset
Crohn’s disease: a case-control study. Am JGastroenterol 2006;101:1005–1011.
61. Bernstein CN, Rawsthorne P, Cheang M,Blanchard JF. A population-based case
control study of potential risk factors forIBD. Am J Gastroenterol 2006;101:993–
1002.62. Koloski NA, Bret L, Radford-Smith G.
Hygiene hypothesis in inflammatory boweldisease: a critical review of the literature.
World J Gastroenterol 2008;14:165–173.63. Fumagalli M, et al. Parasites represent a
major selective force for interleukin genes
and shape the genetic predisposition toautoimmune conditions. J Exp Med
2009;206:1395–1408.64. Fredericks CA, et al. Healthy young women
with serotonin transporter SS polymorphismshow a pro-inflammatory bias under resting
and stress conditions. Brain Behav Immun2010;24:350–357.
65. Barreiro LB, Quintana-Murci L. From evolu-tionary genetics to human immunology:
how selection shapes host defence genes.Nat Rev Genet 2010;11:17–30.
66. Barnes KC, Grant AV, Gao P. A review of thegenetic epidemiology of resistance to para-
sitic disease and atopic asthma: commonvariants for common phenotypes? Curr
Opin Allergy Clin Immunol 2005;5:379–385.
67. Cuzick J, et al. Aspirin and non-steroidalanti-inflammatory drugs for cancer preven-
tion: an international consensus statement.Lancet Oncol 2009;10:501–507.
68. Berg DJ, et al. Enterocolitis and colon cancerin interleukin-10-deficient mice are associ-
ated with aberrant cytokine production andCD4(+) TH1-like responses. J Clin Invest
1996;98:1010–1020.69. Kundu N, Beaty TL, Jackson MJ, Fulton AM.
Antimetastatic and antitumor activities ofinterleukin 10 in a murine model of breast
cancer. J Natl Cancer Inst 1996;88:536–541.
70. Zheng LM, et al. Interleukin-10 inhibits
tumor metastasis through an NK cell-depen-dent mechanism. J Exp Med 1996;
184:579–584.71. Kundu N, Fulton AM. Interleukin-10 inhib-
its tumor metastasis, downregulates MHCclass I, and enhances NK lysis. Cell Immunol
1997;180:55–61.72. Moore KW, de Waal Malefyt R, Coffman
RL, O’Garra A. Interleukin-10 and the inter-leukin-10 receptor. Annu Rev Immunol
2001;19:683–765.73. Berman RM, Suzuki T, Tahara H, Robbins
PD, Narula SK, Lotze MT. Systemic adminis-tration of cellular IL-10 induces an effective,
specific, and long-lived immune response
against established tumors in mice. J Immu-nol 1996;157:231–238.
74. Inge TH, Hoover SK, Susskind BM, BarrettSK, Bear HD. Inhibition of tumor-specific
cytotoxic T-lymphocyte responses by trans-forming growth factor beta 1. Cancer Res
1992;52:1386–1392.75. Khazaie K, von Boehmer H. The impact of
CD4+CD25+ Treg on tumor specific CD8+ Tcell cytotoxicity and cancer. Semin Cancer
Biol 2006;16:124–136.76. Ghiringhelli F, Menard C, Martin F, Zitvogel
L. The role of regulatory T cells in the con-trol of natural killer cells: relevance during
tumor progression. Immunol Rev
2006;214:229–238.77. Derynck R, Akhurst RJ, Balmain A. TGF-beta
signaling in tumor suppression and cancerprogression. Nat Genet 2001;29:117–129.
78. Engle SJ, Hoying JB, Boivin GP, Ormsby I,Gartside PS, Doetschman T. Transforming
growth factor beta1 suppresses nonmeta-static colon cancer at an early stage of
tumorigenesis. Cancer Res 1999;59:3379–3386.
79. Engle SJ, et al. Elimination of colon cancerin germ-free transforming growth factor
beta 1-deficient mice. Cancer Res2002;62:6362–6366.
80. Nizar S, et al. T-regulatory cell modulation:the future of cancer immunotherapy? Br J
Cancer 2009;100:1697–1703.81. Fox JG, et al. Concurrent enteric helminth
infection modulates inflammation and gas-tric immune responses and reduces helicob-
acter-induced gastric atrophy. Nat Med2000;6:536–542.
82. Erdman SE, et al. CD4(+)CD25(+) regula-tory lymphocytes require interleukin 10 to
interrupt colon carcinogenesis in mice. Can-cer Res 2003;63:6042–6050.
83. Erdman SE, et al. CD4+CD25+ regulatorylymphocytes induce regression of intestinal
tumors in ApcMin ⁄ + mice. Cancer Res2005;65:3998–4004.
84. Menetrier-Caux C, Gobert M, Caux C. Dif-ferences in tumor regulatory T-cell localiza-
tion and activation status impact patient
outcome. Cancer Res 2009;69:7895–7898.85. Bates GJ, et al. Quantification of regulatory
T cells enables the identification of high-riskbreast cancer patients and those at risk of
late relapse. J Clin Oncol 2006;24:5373–5380.
86. Welsh TJ, Green RH, Richardson D, WallerDA, O’Byrne KJ, Bradding P. Macrophage
and mast-cell invasion of tumor cell isletsconfers a marked survival advantage in non-
small-cell lung cancer. J Clin Oncol2005;23:8959–8967.
87. Gounaris E, et al. T-regulatory cells shiftfrom a protective anti-inflammatory to a
cancer-promoting proinflammatory
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
� 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011 157
phenotype in polyposis. Cancer Res 2009;69:5490–5497.
88. Colombo MP, Piconese S. Polyps wrap mastcells and Treg within tumorigenic tentacles.
Cancer Res 2009;69:5619–5622.89. Wu AJ, Hua H, Munson SH, McDevitt HO.
Tumor necrosis factor-alpha regulation ofCD4+CD25+ T cell levels in NOD mice.
Proc Natl Acad Sci USA 2002;99:12287–12292.
90. Goodman WA, Levine AD, Massari JV, Sug-iyama H, McCormick TS, Cooper KD. IL-6
signaling in psoriasis prevents immune sup-pression by regulatory T cells. J Immunol
2009;183:3170–3176.
91. Sakaguchi S, Wing K, Yamaguchi T. Dynam-ics of peripheral tolerance and immune reg-
ulation mediated by Treg. Eur J Immunol2009;39:2331–2336.
92. Sakaguchi S, Wing K, Onishi Y, Prieto-Mar-tin P, Yamaguchi T. Regulatory T cells: how
do they suppress immune responses? IntImmunol 2009;21:1105–1111.
93. Zuany-Amorim C, et al. Suppression of air-way eosinophilia by killed Mycobacterium
vaccae-induced allergen-specific regulatoryT-cells. Nat Med 2002;8:625–629.
94. Wilson MS, Taylor MD, Balic A, Finney CA,Lamb JR, Maizels RM. Suppression of aller-
gic airway inflammation by helminth-induced regulatory T cells. J Exp Med
2005;202:1199–1212.95. Maizels RM, Yazdanbakhsh M. Immune reg-
ulation by helminth parasites: cellular andmolecular mechanisms. Nat Rev Immunol
2003;3:733–744.96. van den Biggelaar AH, et al. Long-term
treatment of intestinal helminths increasesmite skin-test reactivity in Gabonese school-
children. J Infect Dis 2004;189:892–900.97. Rastogi T, et al. Cancer incidence rates
among South Asians in four geographicregions: India, Singapore, UK and US. Int J
Epidemiol 2008;37:147–160.98. Gutensohn N, Cole P. Childhood social
environment and Hodgkin’s disease. N EnglJ Med 1981;304:135–140.
99. Glaser SL, Hsu JL. Hodgkin’s disease in
Asians: incidence patterns and risk factors inpopulation-based data. Leuk Res 2002;
26:261–269.100. Mani H, Jaffe ES. Hodgkin lymphoma: an
update on its biology with new insights intoclassification. Clin Lymphoma Myeloma
2009;9:206–216.101. Strachan DP. Hay fever, hygiene, and
household size. Br Med J 1989;299:1259–1260.
102. Filippi CM, Estes EA, Oldham JE, von Her-rath MG. Immunoregulatory mechanisms
triggered by viral infections protect fromtype 1 diabetes in mice. J Clin Invest
2009;119:1515–1523.
103. Cartwright RA, Watkins G. Epidemiology ofHodgkin’s disease: a review. Hematol Oncol
2004;22:11–26.104. Greaves M. Infection, immune responses
and the aetiology of childhood leukaemia.Nat Rev Cancer 2006;6:193–203.
105. Ma X, et al. Daycare attendance and risk ofchildhood acute lymphoblastic leukaemia.
Br J Cancer 2002;86:1419–1424.106. Gilham C, et al. Day care in infancy and risk
of childhood acute lymphoblastic leukae-mia: findings from UK case-control study.
BMJ 2005;330:1294.107. Feltbower RG, McKinney PA, Greaves MF,
Parslow RC, Bodansky HJ. International par-
allels in leukaemia and diabetes epidemiol-ogy. Arch Dis Child 2004;89:54–56.
108. Stene LC, Nafstad P. Relation betweenoccurrence of type 1 diabetes and asthma.
Lancet 2001;357:607.109. Fox JG, Wang TC. The ‘‘African Enigma’’–
another explanation (letter). Nat Med2000;6:1297–1298.
110. Baron JH, Sonnenberg A. Period- andcohort-age contours of deaths from gastric
and duodenal ulcer in New York 1804-1998. Am J Gastroenterol 2001;96:2887–
2891.111. Sonnenberg A. Causes underlying the birth-
cohort phenomenon of peptic ulcer: analy-sis of mortality data 1911-2000, England
and Wales. Int J Epidemiol 2006;35:1090–1097.
112. Robinson K, et al. Helicobacter pylori-inducedpeptic ulcer disease is associated with inade-
quate regulatory T cell responses. Gut2008;57:1375–1385.
113. Atherton JC, Blaser MJ. Coadaptation of Heli-cobacter pylori and humans: ancient history,
modern implications. J Clin Invest2009;119:2475–2487.
114. Vaezi MF, et al. CagA-positive strains ofHelicobacter pylori may protect against Barrett’s
esophagus. Am J Gastroenterol 2000;95:2206–2211.
115. Cover TL, Blaser MJ. Helicobacter pylori inhealth and disease. Gastroenterology
2009;136:1863–1873.
116. Lundgren A, et al. Mucosal FOXP3-express-ing CD4+ CD25high regulatory T cells in
Helicobacter pylori-infected patients. InfectImmun 2005;73:523–531.
117. Reibman J, et al. Asthma is inversely associ-ated with Helicobacter pylori status in an urban
population. PLoS ONE 2008;3:e4060.118. Codolo G, et al. The neutrophil-activating
protein of Helicobacter pylori down-modulatesTh2 inflammation in ovalbumin-induced
allergic asthma. Cell Microbiol 2008;10:2355–2363.
119. D’Elios MM, et al. Helicobacter pylori, asthmaand allergy. FEMS Immunol Med Microbiol
2009;56:1–8.
120. Center MM, Jemal A, Smith RA, Ward E.Worldwide variations in colorectal cancer.
CA Cancer J Clin 2009;59:366–378.121. Summers RW, Elliott DE, Urban JF Jr,
Thompson R, Weinstock JV. Trichuris suistherapy in Crohn’s disease. Gut
2005;54:87–90.122. Summers RW, Elliott DE, Urban JF Jr,
Thompson RA, Weinstock JV. Trichuris suistherapy for active ulcerative colitis: a ran-
domized controlled trial. Gastroenterology2005;128:825–832.
123. Mazmanian SK, Round JL, Kasper DL. Amicrobial symbiosis factor prevents intesti-
nal inflammatory disease. Nature
2008;453:620–625.124. Turnbaugh PJ, et al. A core gut microbiome
in obese and lean twins. Nature2009;457:480–484.
125. Dethlefsen L, Huse S, Sogin ML, RelmanDA. The pervasive effects of an antibiotic on
the human gut microbiota, as revealed bydeep 16s rRNA sequencing. PLoS Biol
2008;6:e280.126. Marshall JK. Post-infectious irritable bowel
syndrome following water contamination.Kidney Int Suppl 2009;112:S42–S43.
127. Pinquart M, Duberstein PR. Depression andcancer mortality: a meta-analysis. Psychol
Med 2010;40:1797–1810.128. Satin JR, Linden W, Phillips MJ. Depression
as a predictor of disease progression andmortality in cancer patients: a meta-analysis.
Cancer 2009;115:5349–5361.129. Rook GAW, Lowry CA. The hygiene
hypothesis and psychiatric disorders. TrendsImmunol 2008;29:150–158.
130. Rook GAW, Lowry CA. The hygienehypothesis and affective and anxiety disor-
ders. In: Rook G ed. The Hygiene Hypothe-sis and Darwinian Medicine. Basel:
Birkhauser, 2009:189–220.131. Muller N. COX-2 inhibitors as antidepres-
sants and antipsychotics: clinical evidence.Curr Opin Investig Drugs 2010;11:31–42.
132. Lowry CA, et al. Identification of animmune-responsive mesolimbocortical
serotonergic system: potential role in regu-
lation of emotional behavior. Neuroscience2007;146:756–772.
133. Gravekamp C, Paterson Y. Harnessing Listeriamonocytogenes to target tumors. Cancer Biol
Ther 2010;9:257–265.134. Kim SH, Castro F, Paterson Y, Gravekamp C.
High efficacy of a Listeria-based vaccineagainst metastatic breast cancer reveals a
dual mode of action. Cancer Res2009;69:5860–5866.
135. Motamedi M, Arab S, Moazzeni SM, KhamisAbadi M, Hadjati J. Improvement of a den-
dritic cell-based therapeutic cancer vaccinewith components of Toxoplasma gondii. Clin
Vaccine Immunol 2009;16:1393–1398.
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
158 � 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011
136. Shelley MD, Mason MD, Kynaston H. Intra-vesical therapy for superficial bladder can-
cer: a systematic review of randomised trialsand meta-analyses. Canc Treat Rev
2010;36:195–205.137. Tan JK, Ho VC. Pooled analysis of the
efficacy of Bacille Calmette-Guerin (BCG)immunotherapy in malignant melanoma.
J Dermatol Surg Oncol 1993;19:985–990.
138. Skinner MA, Yuan S, Prestidge R, Chuk D,Watson JD, Tan PLJ. Immunization with
heat-killed Mycobacterium vaccae stimulatesCD8+ cytotoxic T cells specific for macro-
phages infected with Mycobacterium tuberculosis.
Infect Immun 1997;65:4525–4530.139. Mendes R, et al. Clinical and immunological
assessment of Mycobacterium vaccae (SRL172)with chemotherapy in patients with malig-
nant mesothelioma. Br J Cancer 2002;86:336–341.
140. Hrouda D, Baban B, Dunsmuir WD, KirbyRS, Dalgleish AG. Immunotherapy of
advanced prostate cancer: a phase I ⁄ II trialusing Mycobacterium vaccae (SRL172). Br J Urol
1998;82:568–573.141. O’Brien ME, et al. A randomized phase II
study of SRL172 (Mycobacterium vaccae) com-bined with chemotherapy in patients with
advanced inoperable non-small-cell lungcancer and mesothelioma. Br J Cancer
2000;83:853–857.142. O’Brien ME, et al. SRL172 (killed Mycobacte-
rium vaccae) in addition to standard chemo-therapy improves quality of life without
affecting survival, in patients with advancednon-small-cell lung cancer: phase III results.
Ann Oncol 2004;15:906–914.
143. Stanford JL, Stanford CA, O’Brien ME,Grange JM. Successful immunotherapy with
Mycobacterium vaccae in the treatment of ade-nocarcinoma of the lung. Eur J Cancer
2008;44:224–227.144. Patel PM, et al. An evaluation of a prepara-
tion of Mycobacterium vaccae (SRL172) as animmunotherapeutic agent in renal cancer.
Eur J Cancer 2008;44:216–223.145. Nicholson S, et al. A randomized phase II
trial of SRL172 (Mycobacterium vaccae) + ⁄ -low-dose interleukin-2 in the treatment of
metastatic malignant melanoma. MelanomaRes 2003;13:389–393.
146. Wu S, et al. A human colonic commensal
promotes colon tumorigenesis via activationof T helper type 17 T cell responses. Nat
Med 2009;15:1016–1022.147. Toprak NU, et al. A possible role of Bactero-
ides fragilis enterotoxin in the aetiology ofcolorectal cancer. Clin Microbiol Infect
2006;12:782–786.148. Cohen RJ, Shannon BA, McNeal JE, Shannon
T, Garrett KL. Propionibacterium acnes asso-ciated with inflammation in radical prosta-
tectomy specimens: a possible link to cancerevolution? J Urol 2005;173:1969–1974.
149. Namiki K, et al. Persistent exposure to Myco-plasma induces malignant transformation of
human prostate cells. PLoS ONE 2009;4:e6872.
150. Umetsu DT, McIntire JJ, DeKruyff RH. TIM-1, hepatitis A virus and the hygiene theory
of atopy: association of TIM-1 with atopy. JPediatr Gastroenterol Nutr 2005;40(Sup-
pl.):S43.151. Matricardi PM, Rosmini F, Panetta V, Ferrig-
no L, Bonini S. Hay fever and asthma in
relation to markers of infection in the Uni-ted States. J Allergy Clin Immunol
2002;110:381–387.152. Yazdanbakhsh M, Kremsner PG, van Ree R.
Allergy, parasites, and the hygiene hypothe-sis. Science 2002;296:490–494.
153. Correale J, Farez M. Association betweenparasite infection and immune responses in
multiple sclerosis. Ann Neurol 2007;61:97–108.
154. Pelosi U, et al. The inverse association ofsalmonellosis in infancy with allergic rhino-
conjunctivitis and asthma at school-age: alongitudinal study. Allergy 2005;60:626–
630.
155. Bjorksten B, Naaber P, Sepp E, Mikelsaar M.The intestinal microflora in allergic Estonian
and Swedish 2-year-old children. Clin ExpAllergy 1999;29:342–346.
156. Shirakawa T, Enomoto T, Shimazu S,Hopkin JM. The inverse association between
tuberculin responses and atopic disorder.Science 1996;275:77–79.
157. Obihara CC, et al. Mycobacterium tuberculosisinfection may protect against allergy in a
tuberculosis endemic area. Clin Exp Allergy2006;36:70–76.
158. Wagner A, et al. Immunoregulation byToxoplasma gondii infection prevents allergic
immune responses in mice. Int J Parasitol2009;39:465–472.
159. McDade TW, Rutherford J, Adair L, KuzawaCW. Early origins of inflammation: micro-
bial exposures in infancy predict lower lev-els of C-reactive protein in adulthood. Proc
Biol Sci 2010;277:1129–1137.
Rook & Dalgleish Æ Infection, immunoregulation, and cancer
� 2011 John Wiley & Sons A/S • Immunological Reviews 240/2011 159