Keywords
Carbohydrate-insulin model - gut microbiome - hyperinsulinemia - obesity - obesity-related
cancers - sarcopenic obesity
Introduction
The world of obesity is viewed through the lens of cardiometabolic disorders. It is
less well appreciated that obesity confers significant risk of specific type of cancers
and that obesity will be the leading cause of cancer in the coming years.
The Obesity Pandemic
The world is facing an obesity pandemic. The World Health Organization estimates that
obesity rates across the globe have tripled since 1975, and in 2016, more than 1.9
billion adults were overweight and of these, over 650 million were obese. The Gulf
region has been particularly affected by this epidemic, with an estimated 30%–40%
of the population being overweight or obese. Based on 2016 data, the CIA World Factbook
identifies that Kuwait is the “fattest” country in the Gulf with almost 40% of population
being obese; Oman at 27% has the lowest percentage of obese adults in the GCC (ranked
29th globally).
The cardiometabolic risks associated with obesity, such as Type 2 diabetes mellitus
(T2DM), hypertension, fatty liver (nonalcoholic fatty liver disease [NAFLD]), hypertension,
and coronary artery disease, are well known. It is less well appreciated that obesity
increases the risk of several types of cancer.
Obesity-Related Cancers
It is estimated that 9% of the cancer burden in North America, Europe, and Middle
East in 2013 was obesity related.[1] Mendelian randomization studies have placed the risk of obesity-related cancers
(JRCs) even higher.[2] This prevalence is likely to have grown since, especially after control of competing
causes of cancer such as smoking and infection. In 2015, tobacco smoking contributed
to the largest proportion of cancer cases in the UK, closely followed by overweight/obesity,
accounting for 15.1% and 6.3%, respectively.[3] Obesity-related cancers accounted for nearly 43.5% of total direct cancer care expenditures,
estimated at $35.9 billion in 2015 in USA alone.[4] The trend in increasing obesity is more marked in Saudi Arabia than in India. This
has resulted in a disproportionately higher level of JRCs in Saudi Arabia (4%–9%)
as compared to a more modest 0.2%–1.2% in India.[1]
Increasing childhood obesity is a matter of grave concern as it has shifted the burden
of cancer to younger age groups.[5] In addition, being overweight before the age of 40 increases the risks of various
JRCs by 15%. The study from Bergen (Norway) showed increased risk of cancers of the
endometrium (by 70%), renal cancer in males (by 58%), and colon cancer in male (by
29%).[6]
Defining Obesity
The cause (s) of obesity and the current epidemic is a matter of controversy. The
classical energy imbalance (“calorie in, calorie out”) model attributes obesity to
eating more and moving less (“gluttony and sloth”). This has been challenged by the
“carbohydrate-insulin” model which suggests that the components of the Western diet
such as highly refined carbohydrates (sugar and fructose) and processed food (including
some seed oils and artificial sweeteners) spike insulin levels, which leads to fat
storage and continued hunger [Figure 1].[7],[8] Some researchers blame the governmental advice in the seventies to cut down on fat
and eat more carbohydrates for this epidemic. The field is further clouded by difficulties
in defining and quantifying “unhealthy obesity” as it appears that not all obese adults
have metabolic complications.
Figure 1: The carbohydrate-insulin model of obesity
The standard method of quantifying obesity is by the body mass index (BMI) (also known
as the Quetelet index), which is weight (in kilograms) divided by the height (in meters)
squared. Healthy BMI has been defined as a value between 18 and 25; overweight is
more than 25, and obese more than 30. Although a convenient method of measurement,
this index suffers from serious deficiencies. This index cannot, for instance, distinguish
between fat and lean weight. A muscular man will be classified as overweight or worse
(for instance, Dwyane Johnson <xref>The Rock</xref> who is 188 cm tall and weighs
around 119 kg has a BMI of 33 and is clearly obese by this criteria). Again, a proportion
of the population who are apparently obese by BMI remain healthy (MHO or metabolically
healthy obese);[9],[10] this appears to be related to the distribution of body fat. A pear-shaped (gynecoid)
body with gluteofemoral distribution of fat (and a low waist–hip ratio (WHR)) is healthier
than an apple shape (android) with fat stored in visceral adipose tissue (VAT). This
difference in fat distribution between men and women is due to sex hormones, and I
believe that this is related to the evolutionary need to leave space in the abdominal
cavity in women for the growing fetus.
It has been shown that postmenopausal women with a normal BMI but with higher body
fat levels (as measured by the gold standard of dual-energy X-ray absorptiometry or
DXA; other methods include hydrostatic weighing; bioelectrical impedance analysis;
air displacement plethysmography; and bioimpedance spectroscopy; these can be combined
together to generate multicompartment models) are at elevated risk for breast cancer.[11] These are hold the middle ground – “the metabolic obesity in normal weight”. On
the other end of the spectrum, there are people who are thin but diabetic (“TOFI”,
thin outside, fat inside) [Figure 2].[12] This variation is partly explained by the idea that fat storage in subcutaneous
tissue is essentially safe (to a limit), but when it spills over and stored ectopically
(in muscle, and especially in the liver), leads to insulin resistance, hyperinsulinemia
(HI), hyperglycemia, and diseases associated with the metabolic syndrome.[13] WHR is one way to measure the ectopic fat stored in VAT and correlates better with
metabolic syndrome than BMI;[14] other methods such as relative fat mass (RFM) have been proposed to overcome limitations
of BMI.[15] RFM correlates closely with the gold standard DXA scan. However, some studies suggest
that ectopic fat stored in liver (as in NAFLD) poses more risk than the fat in other
sites.[16] As of now, there is no answer to the pressing question, “Which is the 'real' obesity?.”
Without a standard method of defining and quantifying “unhealthy obesity,” accurately
identifying cancers that are distinctly and specifically obesity related will remain
imprecise.
Figure 2: The spectrum of obesity as per body mass index
Nevertheless, the International Agency for Research on Cancer has come up with a list
of 13 cancers associated with obesity[17] (as defined by BMI) [Table 1] including common ones such as those of colon and breast. Since then, other associations
have been reported,[6],[18] including new possibilities such as of the prostate,[19] neuroendocrine tumors,[20] and of the urinary bladder.[21] Theoretically, cancers with similar etiology should have similar mutation spectra,[22] but since cancers are rarely caused by a single factor, defining a homogenous population
of JRCs and generating a universal “molecular signature” that could identify a cancer
as a member of JRC remains difficult. Overexpression of genes such as the fatty acid
synthase (FASN)[23] and fat mass and obesity-associated (FTO)[24],[25] have been identified in JRCs.
Table 1
Obesity related cancer as per International Agency for Research on Cancer
|
Cancer
|
|
Colon cancer
|
|
Breast cancer
|
|
Thyroid cancer
|
|
Liver cancer
|
|
Endometrial cancer
|
|
Esophageal cancer
|
|
Renal cancer
|
|
Gall bladder cancer
|
|
Pancreatic cancer
|
|
Ovarian cancer
|
|
Gastric cardia cancer
|
|
Multiple myeloma
|
|
Meningioma
|
Association Versus Causation and Hill's Criteria
Association Versus Causation and Hill's Criteria
Bradford Hill's criteria is one attempt to demonstrate causality,[26] and as per this criteria, obesity is a plausible cause of cancer [Table 2]. Way back in 2014, a BMJ editorial titled “Obesity: a certain and avoidable cause
of cancer” acknowledged that “obesity is an important cause of unnecessary suffering
and death from many forms of cancer.”[27] Unfortunately, the exact molecular mechanism (s) and pathways are yet to be worked
out.[28]
Table 2
Postulated mechanisms of obesity related cancers
|
Cancer
|
Postulated mechanism
|
|
NAFLD – Nonalcoholic fatty liver disease
|
|
Breast cancer (postmenopausal)
|
Estrogen produced by adipose tissue
|
|
Endometrial cancer
|
High estrogen levels
|
|
Gastro-esophageal cancer
|
Increased gastro-esophageal reflux due to high visceral fat
|
|
Gall bladder cancer
|
Cholesterol gall stones
|
|
Liver cancer
|
Fatty liver, NAFLD
|
Mechanisms of Cancer Causation
Mechanisms of Cancer Causation
The association of some cancers with obesity can be better understood than the others
[Table 3]. The weight loss drug lorcaserin was recently recalled by the FDA for slight excess
of cancers in the study arm. However, for the vast majority, the links are less well
understood.[29] Suggested possibilities are as follows.
Table 3
Causation versus association: Hill’s criteria
|
Number
|
Criterion
|
Explanation
|
Obesity and cancer
|
|
1
|
Strength
|
Difference between exposed versus nonexposed
|
Significant for select subsets
|
|
2
|
Consistency
|
Observed by different people at different places
|
Yes
|
|
3
|
Specificity
|
Linked to specific outcome
|
Yes but inconsistent
|
|
4
|
Temporality
|
Exposure precede the disease
|
Yes
|
|
5
|
Biological gradient
|
Dose response curve
|
Yes, with breast cancer
|
|
6
|
Plausibility
|
Biologically plausible
|
Yes
|
|
7
|
Coherence
|
Cause-effect consistent with known natural history
|
Yes but further investigations needed
|
|
8
|
Experiment
|
Intervention change outcome?
|
Yes
|
|
9
|
Analogy
|
Similar agents cause similar disease?
|
Unique experience
|
Hyperinsulinemia
High insulin levels precede metabolic syndrome by at least a decade. The modern man
has relatively high insulin levels partly in response to highly refined carbohydrate
and processed food diet and partly due to the tendency to snack between meals.[30] Insulin is both anabolic (leading to fat storage) and proliferative (stimulating
at least two pathways known to be involved in carcinogenesis – the PI3K/AKT/mTJR pathway
and the MAPK pathway).[31] Single-nucleotide polymorphisms (SNPs) in the insulin receptor gene have been associated
with JRCs.[32] A Japanese study showed that HI was independently associated with higher cancer
risk irrespective of BMI.[33] Foods high in glycemic index have been implicated in some types of JRCs.[34] Other hormones released by adipose tissue such as leptin[35] and adiponectin have also been implicated.[36] Overexpression of gastric leptin has been linked to stomach cancer.[37] Newer candidates incriminated include adipose fatty acid-binding protein (A-FBP)
in breast cancer.[38]
Chronic inflammation
Chronic inflammation and antigenic stimuli, whether due to autoimmune disorders or
infection, are linked to cancer. About 10%–15% of cancers are due to infections. Chronic
infections by viruses and bacteria are associated with lymphomas and cancers of the
gastrointestinal tract; autoimmune disorders of thyroid and gastrointestinal tract
(celiac disease, ulcerative colitis) are also known risk factors. Obesity is an inflammatory
state; it is thought that excess fat storage leads to rupture of adipocytes, leading
to infiltration by immune cells and secretion of cytokines such as interleukin-6 and
tumor necrosis factor, and results in chronic low-grade inflammation.[39] Metabolic and inflammatory changes related to the obese adipose tissue microenvironment
are thought to contribute to cancer development and progression.[40] Obesity-related inflammation can also lead to DNA damage and thus cancer.[41]
Gut microbiome
The influence of gut organisms on obesity (“obesogenic bacteria”) and on immunity
is a matter of ongoing research.[42] Obese people have altered gut bacteria that has been linked to increased risk of
gastrointestinal cancers such as of colon, liver, and pancreas.[43],[44]
Implications
Obesity is the cause of a significant and growing subset of cancers, and this carries
several implications.
Prevention
There is sufficient evidence that weight reduction cuts the risk of cancer.[45] This is best seen in patients of bariatric surgery who undergo significant weight
loss and have reduced risk of various cancers, such as that of breast, uterus,[46] colon, and even skin cancer.[47] Intermittent fasting has been suggested as another method of weight control and
increasing life span; however, the evidence in humans is limited. The role of insulin
in JRCs was mentioned earlier; blocking the insulin pathway is a specific option to
treat JRCs. Drugs such as ceritinib, alpelisib, capivasertib, everolimus, and rapamycin
(rapalogs) that block proteins along the insulin pathway are effective anticancer
drugs; it is unknown if they are more effective in JRCs. Epidemiological studies suggest
that metformin may have a preventive role in cancers such as NAFLD-associated HCC,
but clinical data are lacking. Clearly, the best preventive method is control of the
obesity epidemic, but this requires long-term solutions with social and legal will
to enforce measures such as the sugar tax. These measures did work, for instance,
in control of tobacco abuse and there is no reason why they should not work to modify
unhealthy diets.[48]
Screening
The obese population presents a unique opportunity to screen for specific cancers
with the certainty of higher yield. Patients of NAFLD have not only higher risk of
liver cancer but also of stomach, pancreas, uterus, and colon;[16] endometrial and thyroid cancers are other high-yield JRCs detectable by screening.
The evidence that HI is related to JRCs[49] suggests that insulin levels can be useful as a screening tool; much work needs
to be done in this respect. On the flip side, obesity could make screening procedures
problematic. In Spanish women, caloric intake above predicted levels seems to increase
mammographic density, such that for every 20% increase in relative energy intake,
mammographic density increased by 5%.[50]
Second malignant neoplasms
Survivors of childhood cancers have higher risk of second cancers if they put on weight;[51] this presents a challenge as they are prone to weight gain.
Prognosis
As a general rule, obesity predicts poorer survival from various cancers, even in
the early stages,[52] and correlates with more visceral metastases.[53] Hyperglycemia itself exerts a negative influence on survival from cancers of breast,
liver, and colon.[54] Curiously, some patients of JRCs have better survival which has been termed the
“obesity paradox.”[55],[56],[57] Several explanations have been advanced such as better nutrition, unreliability
of BMI, and statistical issues such as reverse causality and collider bias.[58] There is a possibility that obesity gives an advantage to low-grade tumors with
inbuilt survival advantage.[59] In renal cell cancer, altered microenvironment of the peritumoral adipose tissue
has been suggested as an explanation of this paradox.[60]
Treatment considerations
In addition to its impact on surgery (for example, postoperative infections, anesthetic
complications) and radiation (postradiation fibrosis), obesity influences chemotherapy
administration.[61] There is a tendency to cap chemotherapy drug dose leading to suboptimal dosage and
reduced survival; guidelines recommend full dose as per the individual's actual weight.[62] Despite this, survival can be compromised as drug metabolism may be different in
the obese. For instance, adipocytes have been shown to promote doxorubicin resistance
by upregulating a drug efflux protein MVP.[63] In addition, adipocytes can sequester chemotherapy drugs and protect cancer cells.[64] Side effects can be more; cardiotoxicity of trastuzumab is increased in obese, dyslipidemic
patients.[65] A recent study showed that excess adiposity, detected on usual computed tomography
scans as larger visceral and intramuscular fat deposition, was related to reduced
relative dose intensity and worse breast cancer-specific survival.[66] Paradoxically, retrospective studies show that immunotherapy may work better in
the obese.[67]
Targeted therapy
Upregulation of specific genes related to JRCs present a unique opportunity for targeted
therapy; studies are currently ongoing with FASN[68] and FTO[69] inhibitors.
Sarcopenic obesity
About 15% of obese people have limited muscle mass (sarcopenic obesity) and this is
further aggravated in patients with cancer cachexia. Sarcopenic obesity confers poorer
outcomes in cancer patients including reduced survival.[70] Sarcopenic obesity cannot be detected by clinical examination, but standard imaging
done routinely as part of cancer treatment can be used to specifically measure muscle
mass and alter management.[71]
Conclusions
The world, and especially the Gulf countries, is in the grip of an obesity pandemic.
Metabolic disorders such as NAFLD and T2DM are on the rise and will cost the world
economy, billions of dollars. As far as cancers are concerned, obesity is the new
smoking and requires equally careful management measures. Several subtypes of cancers
will increase and impact the lives of many, unless the importance of JRCs is acknowledged
and proactively managed.[72] Oncologists must discuss with the patient importance of weight management as an
essential part of cancer treatment, and the need for good glycemic control for better
outcomes; these are measures that can be implemented at minimal or no cost. A shared
decision on weight and glycemic management and an “exercise prescription” is good
clinical practice. Unfortunately, awareness of the link between obesity and cancer
is limited, and education, starting at school level, should be an important measure
in future programs.[73] Research on the molecular mechanisms of JRCs is crucial.[74] It is well known that Indians develop cardiometabolic complications of obesity at
much lower levels of BMI (for which modified criteria have been suggested[75]); whether similar risk exists for JRCs is a potential area for study.
