SEARCH

SEARCH BY CITATION

Keywords:

  • gallbladder cancer;
  • risk factors;
  • gallstones

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Risk factors for gallbladder cancer (GBC) except gallstones are not well known. The objective was to study the risk factors for GBC. In a case–control study, 200 patients with GBC, 200 healthy controls and 200 gallstones patients as diseased controls were included prospectively. The risk factors studied were related to socioeconomic profile, life style, reproduction, diet and bile acids. On comparing GBC patients (mean age 51.7 years; 130 females) with healthy controls, risk factors were chemical exposure [odd ratios (OR): 7.0 (2.7–18.2); p < 0.001)], family history of gallstones [OR: 5.3 (1.5–18.9); p < 0.01)], tobacco [OR: 4.1 (1.8–9.7); p < 0.001)], fried foods [OR: 3.1 (1.7–5.6); p < 0.001], joint family [OR: 3.2 (1.7–6.2); p < 0.001], long interval between meals [OR: 1.4 (1.2–1.6); p < 0.001] and residence in Gangetic belt [OR: 3.3 (1.8–6.2); p < 0.001]. On comparing GBC cases with gallstone controls, risk factors were female gender [OR: 2.4 (1.3–4.3); p = 0.004], residence in Gangetic belt [OR: 2.3 (1.2–4.4); p = 0.012], fried foods [OR: 2.5 (1.4–4.4); p < 0.001], diabetes [OR: 2.7 (1.2–6.4); p = 0.02)], tobacco [OR 3.8 (1.7–8.1); p < 0.001)] and joint family [OR: 2.1 (1.2–3.4); p = 0.004]. The ratio of secondary to primary bile acids was significantly higher in GBC cases than gallstone controls (20.8 vs. 0.44). Fried foods, tobacco, chemical exposure, family history of gallstones, residence in Gangetic belt and secondary bile acids were significant risk factors for GBC.

Gallbladder cancer (GBC) is one of the commonest gastrointestinal malignancies, especially among females. The incidence of GBC is high in many parts of the world including Chile, Peru, Bolivia, Korea, Japan, Czech Republic, Slovakia, Spain and India.1 The incidence of GBC among women in northern India is one of the highest in the world, and the incidence of GBC is steadily increasing from 10.1/100,000 population in women in 1993 to 19.6/100,000 population in 2006.2, 3 The prognosis of GBC is poor with <10% of patients suitable for curative resection and an overall 5-year survival of <5%.4 The etiopathogenesis of GBC is not well understood. The study of the risk factors for GBC is important not only for understanding the etiopathogenesis but also for preventive strategies. The putative risk factors for GBC include female sex, gallstones, chronic Salmonella typhi carrier status, dietary factors and environmental exposure to specific chemicals.5–13 However, the current knowledge about the risk factors for GBC is limited, and there is inconsistency about the role of risk factors in various studies from different centers. For example, the role of dietary factors and female sex hormones in the pathogenesis of GBC is ambiguous.14–16 Furthermore, one of the major issues in many previous studies was the confounding effect of gallstones while determining the risk associated with other factors. Although the causality is not established, it is well known that gallstones are strongly associated with GBC. It is quite possible that given their similar profile, patients with GBC and those with gallstones share many of the risk factors. Thus, it is important that a study to assess the risk factors for GBC should take into account the confounding effect of gallstones.

The present study was conducted in northern India, a high-incidence area for GBC, with the objective to find out various risk factors related to sociodemographic profile, diet, lifestyle, biliary bile acids, biliary infection, occupational factors, etc., that might be associated with GBC while adjusting for the confounding effect of gallstones.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Cases

All consecutive newly diagnosed patients with GBC were included as cases. They were enrolled from the out-patient Departments of Gastroenterology and Surgery and indoor admissions at the All India Institute of Medical Sciences, New Delhi.

Diagnosis of GBC

The diagnosis of gallbladder cancer (GBC) was made on the basis of imaging features and was confirmed by histology. Imaging studies to diagnose GBC included an abdominal ultrasonography (USG) and a contrast-enhanced computed tomography (CECT) scan of the abdomen. For the confirmation of the diagnosis, either a fine-needle aspiration cytology or percutaneous biopsy or histopathology of the surgically resected specimen was performed. In patients in whom the cyto/histopathology was not available, a combination of clinical and imaging features and a progressive downhill clinical course of the disease was substituted for histology.

Inclusion criteria

  • Consecutive patients with GBC who were in sufficiently good physical and mental health to give reliable answers to the questionnaire were included in the study.

Exclusion criteria

  • Patients with any other major comorbid illness.

  • Refusal to participate in the study.

Controls

Two groups of controls were included:

  • 1
    Healthy controls: Healthy controls were included in 1:1 ratio. Subjects in the control group were similar to patients for their age (±5 years) and sex. Healthy controls were enrolled from amongst the relatives and attendants of patients (other than those with gallbladder or any other cancer) attending the out-patient department during the same time period when cases were being recruited.
  • 2
    Diseased controls: Since gallstones disease is considered a risk factor for GBC and 60–80% of patients with GBC might contain gallstones,4 a second group of patients with gallstone disease but without malignancy was included as diseased control during the same time period to study the risk factors other than gallstones. Subjects in this group were also similar to patients for age (±5 years) and sex. Gallstones were diagnosed on abdominal sonography.

Diagnostic work-up of study participants

All the patients and controls underwent a detailed clinical evaluation and diagnostic work-up. Their age, sex, body weight, body mass index (BMI), family history, personal history and physical examination findings were duly recorded. In the diagnostic work-up, hematological and biochemical parameters were measured. Imaging studies included abdominal USG and a CECT scan of the abdomen. A magnetic resonance imaging was done if indicated clinically. Patients with GBC and gallstones were managed appropriately as per the standard protocol.

Study of risk factors

A pretested, semistructured questionnaire was administered to each subject to elicit information on the following variables.

Sociodemographic factors

Details regarding occupation, income, religion, education and residence were noted. Socioeconomic status of the study subjects was assessed using the modified Kuppuswamy classification.17 The place of primary residence for each subject was enquired. The place of residence was divided into zones, i.e., north, central and other zones (details in Supporting Information). A few studies have shown that the incidence of GBC is higher in some of the states in India that are located along the river “Ganga.”18 The Gangetic belt comprised the states of Bihar, Uttar Pradesh and Uttarakhand.

Lifestyle factors

A history of alcohol consumption, smoking and/or tobacco consumption and chemical exposure was recorded. The frequency, quantity and duration of consumption of these items were noted. Chemical exposure was defined as exposure to coal dust, wood dust, metal dust, paint dust, diesel/kerosene fumes or any other occupation/industry-related chemical substance or toxic fumes.

Reproductive history was taken in females. Information on age at menarche, menopausal status, age at first child birth, parity and a history of the use of oral contraceptives was obtained.

Family history

A history of any cancer in the family especially digestive tract cancer was recorded.

Past medical illness

A past history of any major illness, in particular typhoid fever and any gallbladder disease was also obtained.

Dietary assessment

Dietary consumption pattern was assessed using standard Food Frequency Questionnaire method.19 The commonly consumed food items included cereals (wheat/rice), pulses/legumes, green leafy vegetables, roots and tubers, other vegetables, fruits, milk, milk products, flesh foods (red meat/fish/poultry), eggs, tea, coffee, green leafy vegetables in winter, roots and tubers in winter, fruits in winter, flesh foods in winter and eggs in winter. These were categorized under the major food groups, i.e., cereals, pulses and legumes, green leafy vegetables, other vegetables, fruits, milk, milk products, flesh foods and eggs. The enquiry was for adult life preceding the symptoms of disease in cases and immediate past in controls. For each food group, the study subjects were asked how frequently they were consuming various food items belonging to a major food group (before the diagnosis of disease). Frequency of consumption was assessed in (i) number of days per week (i.e., 1–7), (ii) once per fortnight, (iii) once per month, (iv) occasionally or (v) never. Seasonal variation in consumption of certain products was also taken into account. Intake of tea, coffee, the type of oil such as mustard oil and the use of fried or spicy foods was enquired. The longest interval between the meals in a day was noted.

Physical performance status

The physical performance status of GBC patients was measured as per the Karnofsky scale.

Nutritional status

Height and weight of all the study subjects were measured using standard procedures. Nutritional status was assessed by BMI, where BMI = weight in kilograms/(height in meters)2.

Bile acids

Because secondary bile acids have been implicated in the pathogenesis of GBC, we measured primary and secondary bile acids in the bile samples. The methods of bile collection and measurement of bile acids are given in Supporting Information and Table S1. The concentration of primary and secondary bile acids was estimated through high-performance liquid chromatography on Thermo Electron Corporation Finnigan Surveyor.

Bile culture

Aerobic and anaerobic cultures of bile samples were done to find out the presence of bacterial infection in the bile.

Sample size calculation

We compared GBC patients with healthy controls and gallstones controls for a number of variables such as dietary, lifestyle, occupational, geography and reproductive. In view of multiple risk factors being studied, we fixed the sample size as 200 for each group to give an 80% power in a two-sided test with 5% α error for any factor whose prevalence is at least 15% and assuming the associated odd ratios (OR) to be 2.0.

Statistical analysis

Data were recorded on a predesigned proforma. Before entering the data on an excel spreadsheet, the proforma was reviewed for any incomplete information. An exploratory analysis was done to finalize the categorization of some of the variables into appropriate forms for the study and then subsequently coded. After the categorization of data, the descriptive statistics, i.e., frequency distribution and percentages of demographic profile and various risk covariates were calculated. The frequency of consumption of various food groups was categorized into two categories: <3 or ≥ 3 times per week.

The data are expressed as mean and standard deviation (SD). Descriptive statistics, i.e., mean, SD and frequency distribution were calculated for each variable. Student's t test was used for comparison of continuous data, and chi-square test was applied for comparison of the categorical data. Analysis was carried between GBC and healthy controls. Analysis was also done between diseased controls (with gallstones) and GBC cases having associated gallstones. Logistic regression was used for producing OR. Step-wise multiple logistic regression analysis was carried out with insertion of all the variables found to be significant on univariate analysis. Analysis was also done separately for females, which included reproductive factors and the other factors. A p-value of <0.05 was considered as statistically significant. Statistical software Stata version 9.0 was used for statistical analysis.

Ethical clearance

Ethical clearance was obtained from the Ethics Committee of the All India Institute of Medical Sciences, New Delhi. All the subjects were included after an informed written consent.

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

A total of 200 cases with GBC, 200 healthy controls and 200 diseased controls (with gallstones) were included in the study. The mean age of patients with GBC was 51.7 ± 12.2 years (range, 21–86 years); 17% of the patients were <40 years of age. Of the 200 patients, 130 were females.

Clinical features

Jaundice was the most common symptom among GBC patients. Diabetes was present in 11.5% of the patients being higher than that in controls.

Nutritional status

The mean BMI of patients was found to be 20.3 ± 3.7 kg/m2, which was significantly lower than that of gallstone controls (22.8 ± 4.6) and healthy controls (23.4 ± 4 kg/m2). The BMI was <18.5 kg/m2 in 36% of GBC patients when compared with 16% in gallstone control group and 7.5% in healthy control group. But when the usual BMI (premorbid BMI) was compared between cases and controls, no significant difference was found.

The performance status of cases was poor with 15% of the GBC patients having Karnofsky's scale <40 and 64.5% had scale from 50 to 70.

Risk factors for GBC

Supporting Information Tables S2 shows the univariate analysis between gallbladder cancer (GBC) and healthy controls. The significant variables were residence in central zone [2.3 (1.5–3.4); p < 0.001], Gangetic belt [7.3 (4.6–11.5); p < 0.001], joint family [3.1 (2.0–4.8); p < 0.001], exposure to coal or wood dust [10.9 (5.2–22.6); p < 0.001], ever tobacco use [2.7 (1.5–4.7); p < 0.01], nonvegetarian diet [1.6 (1.1–2.4); p = 0.017], pulses consumption >3 times per day [0.4 (0.2–0.7); p = 0.002], green leafy vegetables consumption >3 times per week [0.3 (0.2–0.5); p < 0.001], fruits consumption >3 times per week [0.5 (0.3–0.7); p < 0.001], milk consumption >3 times per week [2.6 (1.2–5.8); p = 0.018], flesh foods consumption >3 times per week [2.4 (1.0–5.3); p = 0.04] and fried foods [5.0 (3.3–7.7); p < 0.001]. Tea and coffee (both) were consumed more by the healthy controls when compared with gallstone controls and GBC patients. Around 46% of the GBC patients had >12 hr gap between their two meals in a day. Family history of gallstone disease [5.1 (2.2–11.8); p < 0.001], family history of cancer [3.5 (1.3–9.6); p = 0.01] and diabetes [2.8 (1.2–6.1); p = 0.01] were also found to be significantly different between the two groups on univariate analysis.

Results of multivariable analysis to identify risk factors for GBC when compared with healthy controls (Table 1)

Multivariable analysis showed residence in Gangetic belt, fried foods, long interval between the meals, family history of gallstone disease, tobacco consumption, joint family and exposure to wood dust or coal dust, as significant risk factors for GBC when compared with healthy controls whereas tea and coffee consumption were found to be protective.

Table 1. Multivariable analysis to identify risk factors for patients with gallbladder cancer (gallbladder cancer vs. healthy controls)
inline image

Results of multivariable analysis to identify risk factors for GBC (with gallstones) as compared to gallstone controls (Table 2)

The results of univariate analysis between GBC cases (with gallstones) and gallstones controls are given in Supporting Information (Supporting Information Table S3). Multivariable analysis showed female sex, residence in central zone or in Gangetic belt, joint family, fried foods, consumption of green leafy vegetables ≥3 times per week, diabetes and tobacco to be significant risk factors for GBC cases (with stones) when compared with gallstone controls.

Table 2. Multivariable analysis to identify risk factors for patients with gallbladder cancer having gallstones (Gallstones vs. gallbladder cancer with gallstones)
inline image

Risk factors among females only

On comparison with various factors in case of females only, it was found that number of live births and lower age at marriage were found to be the risk factors in addition to the other risk factors (Supporting Information Tables S4–S7).

Biliary bile acid analysis

Biliary bile acid analysis was done in 41 patients with GBC and 41 controls with gallstones. The mean age of the patients with GBC (50.9 ± 9.3 years) and gender distribution (29 females, 12 males) was comparable to those of controls (mean age: 49.6 ± 12.1; 28 females and 13 males). The clinical profile and biochemical parameters of the patients and controls are given in Supporting Information Tables S8 and S9, respectively. The total bilirubin level of the patients was statistically higher when compared with the controls.

Bile acid concentrations

The results of bile acid analysis of GBC patients and gallstone controls are given in Table 3. The total bile acids and the individual bile acids (glycocholic acid, taurocholic acid, glycochenodeoxycholic acid, taurochenodeoxycholic acid, glycodeoxy cholic acid, glycoursodeoxycholic acid and tauroursodeoxycholic acid) were found to be significantly higher in gallstone controls when compared with the patients with GBC. However, the ratio of total secondary bile acids (glycodeoxycholic acid + taurodeoxycholic acid + glycoursodeoxycholic acid + tauroursodeoxycholic acid) to the total primary bile acids (glycocholic acid + taurocholic acid + glycochenodeoxycholic acid + taurochenodeoxycholic acid) in GBC cases was found to be 20.8 compared to 0.44 in controls (Table 3). Thus, the relative concentration of secondary bile acids was significantly higher in patients with GBC compared to controls.

Table 3. Concentration (median values, interquartile range) and ratio of primary and secondary bile acids in bile samples of gallbladder cancer (GBC) and gallstone (GS) disease patients
inline image

Biliary infections

In aerobic cultures, micro-organisms like Escheria coli, Klebsiella pneumoniae, Acinetobacter spp., Enterobacter spp., Staphylococcus aureus, Streptococcus vividans, pseudomonas spp., Vividans streptococci and Citrobacter spp. were grown both in patients with GBC and controls. There was no major difference (39% in GBC and 48.8% in controls) in the culture positivity of bile between patients and controls. These infections were attributed to biliary obstruction. No anaerobes were grown in any of the samples.

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

GBC is a highly malignant tumor and often diagnosed at an advanced stage. To decrease mortality from GBC, we need to implement preventive and surveillance strategies. Identifying significant risk factors is an important step in that direction. In the present case–control study, we investigated many risk factors for GBC including those related to environmental influences, such as diet, alcohol, smoking, geographical area, and host factors such as bile acids, family history and gallstones in a comprehensive manner. We included two groups of controls—one healthy and the other with gallstones. Because patients with GBC often have associated gallstones,1, 8, 20 patients with GBC and those with gallstone might share certain risk factors. Hence, a control group of gallstones patients was included to take care of the confounding effect of gallstones and study risk factors other than gallstones.

The risk factors that turned out to be significant in the present study were fried foods, tobacco use, exposure to wood and coal dust, long interval between the meals, family history of gallstone disease, joint family, and residence in Gangetic belt. The significant risk factors that turned out to be common between different analyses, i.e., comparing GBC with healthy controls, and GBC with GS controls for all patients and for females, were preference to fried food, joint family and residence in Gangetic belt.

Among the dietary factors, fried food was found to be a risk factor for GBC. A study from Japan had shown that increased risk of the cancer was associated with increased consumption of oily food.21 A longer interval between the meals was found to be a risk factor. Long interval between meals increases the contact time of bile with the gallbladder mucosa. Frequent consumption of green leafy vegetables was found to be a risk factor in the present study when GBC patients were compared to gallstone controls. This is contradictory to the results of earlier studies.22, 23 The probable reason could be less frequent consumption of green leafy vegetables by patient with gallstone as there is a belief that these might increase the risk of stone formation. Thus, the association is likely to be spurious.

No association was seen with alcohol and smoking in the present study although tobacco came out to be a significant risk factor for GBC when compared with healthy controls and gallstones. Earlier studies were contradictory with regard to the effect of alcohol, smoking or tobacco on GBC.24–27

Tea and coffee were found to be protective factors in the present study. Tyagi et al.28 showed tea to be a protective factor, but their results showed coffee to be a risk factor for GBC. However, a study from Japan reported a low risk of GBC among coffee-drinkers.21

A family history of gallstone disease was found to be a significant risk factor and might suggest a genetic predisposition to develop gallbladder diseases including gallstones and GBC. Hsing et al.29 also showed the association of family history of gallstone disease with increased risk of biliary stones, GBC and bile duct cancer.

The risk of GBC was found to be 3.2 times higher in subjects living in joint families when compared with those living in nuclear families in the present study. In India, joint family system is quite prevalent especially in rural areas. Joint family is a proxy for other factors such as poor sanitation and increased chances of food and water borne infections.

Exposure to coal or wood dust was found to be an independent risk factor in the present study. High cadmium, chromium and lead contents have been suggested to be risk factor for GBC along the Gangetic belt.18 Occupational risk factors leading to high-biliary metal content have also been reported to be associated with GBC.1, 30, 31

The majority (82%) of the patients belonged to the Gangetic belt (i.e., Bihar, Uttar Pradesh and Uttarakhand) in the present study. A few earlier studies had also shown that the incidence of GBC was relatively high in these states of India.18, 32 These studies suggest that water borne infections, soil contents or predominantly rural cooking practices such as wood and coal burning might be some of the important contributing factors for such a geographical proclivity to GBC within the same country.

We investigated reproductive variables as risk factors, because GBC is more common among females. In the present study, 65% of the patients were females. An earlier study by Dutta et al.8 showed that the percentage of females was 76.5%. Other studies also support that GBC is more commonly seen in females.33, 34 Among the reproductive factors, the number of live births and a younger age at marriage were found to be important risk factors on comparison of patients with GBC and healthy controls (females) in the present study. Similar results were obtained by Shukla et al.35 They showed that a younger age at menarche (<13 years), higher number of child births (>4), higher number of pregnancies (>4) and a higher age at last childbirth (>25 years) were factors responsible for a relatively increased risk of GBC. Singh et al.36 reported a high risk of GBC in women with early menarche, late marriage, late pregnancy and prolonged reproductive phase. Andreotti et al.37 also showed that high parity led to an increased risk of GBC in Chinese patients with biliary tract cancers. Age at menarche has been controversial in this regard. Some authors have shown a late age, and others have shown an early age at menarche to be a risk factor.37, 38 These observations from various studies suggest a possible role of female hormonal factors in the pathogenesis of GBC. Another possibility is that there is stasis of bile during pregnancy, which may be toxic to the gallbladder mucosa.

Bile acids have been implicated in the pathogenesis of GBC. Only a few studies are available in which the individual bile acids were measured in GBC patients. Furthermore, the sample size was small in these studies.39, 40 Park et al.40 showed that the GBC patients had both significantly lower total bile acid concentration and deoxycholic acid (DCA) concentration. But they included only six GBC patients. The decrease in biliary bile acids was associated with bile acid accumulation in the liver. Earlier studies have shown that DCA acid is cytotoxic, and the carcinogenic potential is due to high upregulation of COX-2 and CDX-2 and downregulation of DNA repair enzymes.41 One of the important findings of the present study was a very high relative concentration of secondary bile acids in GBC patients. The ratio of secondary to primary bile acids was completely reversed in patients with GBC. An earlier study had also shown that the concentrations of secondary bile acids were much higher in patients with GBC when compared with gallstone patients.42 Bacterial degradation of primary bile acids in the gallbladder itself could be responsible for gallbladder carcinogenesis although GBC patients with negative bile culture also had significantly high secondary bile acids.

It is still not clear whether this increase in the secondary bile acids is a cause or effect and whether the increase in the percentage of secondary bile acids is due to increased conversion of primary bile acids to secondary bile acids or due to decreased rate of synthesis of primary bile acids. It might be useful to study hepatic or gallbladder bile without biliary obstruction. But this is difficult due to the practical reasons that include (i) most GBC patients present with obstruction and (ii) GB is generally replaced by a mass and there is hardly any bile in the GB.

There are a few potential limitations of the present study. In particular, the case–control design may render some results difficult to interpret due to a possibility of recall bias and reverse causation. For example, recall bias might explain the very high OR for exposure to coal and dust and reverse causation the inverse association with coffee.

We conclude that the important risk factors identified for GBC were fried foods, tobacco use, a long interval between the meals, chemical exposure, family history of gallstone disease, residence in Gangetic belt and relatively high concentration of secondary bile acids. Increased number of live birth and a younger age at marriage and at first child birth were found to be additional risk factors for GBC in females.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The authors thank all the patients who gave their consent to participate in the study. The funding agencies had no role in study design, in the collection, analysis and interpretation of data, in the writing of the report or in the decision to submit the work for publication.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
IJC_27777_sm_SuppInfo.doc339KSupporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.