Chirurgie.ro

Obesity Surgery, 15, 1171-1176
Prevalence of Cancer in Italian Obese Patients
Referred for Bariatric Surgery
Cristian Boru, MD; Gianfranco Silecchia, MD, PhD; Alessandro Pecchia,
MD; Gianluca Iacobellis, MD1; Francesco Greco, MD; Mario Rizzello, MD;
Nicola Basso, MD, Professor of Surgery
Department of Surgery “Paride Stefanini”, Policlinico Umberto I, University of Rome “LaSapienza”, Rome, Italy; 1Center for Human Nutrition, The University of Texas SouthwesternMedical Center, Dallas, TX, USA Background: An association between obesity and
Introduction
cancer has been shown in large epidemiological stud-
ies. The aim of this study was to evaluate the preva-
lence and types of malignancies in an Italian cohort of
An association between obesity and cancer has been obese patients referred to a bariatric center.
found in large epidemiological studies.1-3 An Methods: A retrospective, observational study was
increased prevalence of cancer of the endometrium, conducted. Between Jan 1996 and Dec 2004, 1,333
kidney, gallbladder and postmenopausal breast in obese patients (M=369, F=964) were seen in the cen-
ter for minimally invasive treatment of morbid obesity.
women, and of the colon, esophagus, gallbladder Morbid obesity were considered as BMI >40 kg/m2 or
and prostate in men, has been reported in obese BMI >35 kg/m2 with at least one co-morbidity. Obese
patients.3-9 For other sites (liver, pancreas, stomach, and morbidly obese patients who suffered any form of
thyroid, lung, hematopoietic cancers), an associa- cancer were reviewed.
tion has been indicated, although the risk of higher Results: 43 patients (3.2%) presented various
malignancies, with 88.3% in females. The prevalence
prevalence is still under investigation.2-3,10-13 of cancer in the younger group (21-46 years) was
The explanation for the increased risk of cancer in higher than in the older group (47-70 years), 2.1% vs
the obese population is not completely known.
1.1%. 26 obese patients out of the 43 (60.5%) (age
Elevated plasma estrogen levels and a relative hyper- 41±7.9 years, BMI 38.2±9.9) presented hormone-
androgenism are hypothesized to be the reason for an related tumors. The most frequent site of cancer was
breast (20.9%), followed closely by thyroid.
increased risk for hormone-related tumors, as well as Conclusion: This is the first Italian report on preva-
increased exposure to insulin and insulin-like growth lence of cancer in a homogeneous obese population
factor I (IGF-I). The influence of excess body weight attending an academic bariatric center. The morbidly
on the risk of death from cancer has been studied in obese patients appear to have a higher risk of devel-
a population of more than 900,000 U.S. adults oping cancer, with a higher prevalence of hormone-
related tumors. The predominant gender affected by
(404,576 men and 495,477 women); in both sexes, both obesity and cancer was female. Thus, a preoper-
the BMI was directly associated with higher rates of ative work-up for cancer screening is indicated in this
death from cancer of the esophagus, colon and rec- group of patients.
tum, liver, gallbladder, pancreas and kidney.2 The prevalence of obesity increases with age to a Key words: Morbid obesity, obesity, cancer, hormone- maximum peak between 45-64 years, with women having higher levels of obesity and morbid obesity.9 Reprint requests to: Dr. Cristian Boru, U.O.C. Laparoscopic In Italy, obesity has shown a continuous positive Surgery, Dipartmento di Chirurgia Generale “Paride Stefanini”, trend in the last decades. Currently, obese Italians Policlinico Umberto I, University of Rome “La Sapienza”, Viale represent 9.1% of the total population,14 with an Policlinico, 155, 00161, Rome, Italy. Fax: 0039/064465513;e-mail: norinacristian@libero.it equal distribution between genders. The highest per- Obesity Surgery, 15, 2005 1171
centile of the obese population is 55-64 years of age ultrasound, and electrocardiogram. A total of 964 (15%), closely followed by the 45-54 years age- morbidly obese patients received different mini- group. Mortality from cancer in Italy is 9.4 per mally invasive bariatric procedures (Table 2). All 1,000 inhabitants, responsible for 154,618 deaths in treated patients attended a follow-up program.
2000.15 Breast is the most frequent cause of death Clinic visits were carried out at 1, 3, 6 and 12 months after surgery in the first year and every 6 cases/year.16 The aim of this study was to ascertain months afterwards. The follow-up included clinical the prevalence and types of malignancies in an evaluations (surgeon, psychiatrist, nutritionist), rou- Italian morbidly obese population referred to a tine laboratory tests, barium swallow (every 6 months for the first year, once per year thereafter),and upper GI endoscopy (once per year). Patients with a cancer history 5 years before their first bariatric visit or suspected of malignancy dur-ing preoperative work-up completed their investiga-tions with more specific tests according to the A retrospective study was conducted based on a malignancy site: tumor markers, mammogram, thy- review of the Center of Minimally Invasive roid fine-needle aspiration, imaging studies (ultra- Treatment of Morbid Obesity database, Policlinico sound examination, computed-tomography scan, Umberto I, Rome. From Jan 1996 to Dec 2004, 1,333 consecutive obese patients (M=369, F=964)were observed. Subjects were classified according to the FAO/WHO/UNU criteria16 into overweight(BMI 25-29.9 kg/m2), grade I obese (BMI 30-34.9 Data are expressed as percent (%) and mean ± stan- kg/m2) and grade II obese subjects (BMI 35-39.9 dard deviation (SD). The database of the Center of kg/m2). Morbidly obese patients were those with Minimally Invasive Treatment of Morbid Obesity BMI >40 kg/m2 or >35 with at least one of the co- was reviewed. Data had been collected and recorded using the FileMaker Pro 6 database program All candidates for bariatric surgery underwent (FileMaker Inc, Santa Clara, CA, USA). The preva- preoperative work-up which included physical lence of cancer in the 1,333 obese subjects was cal- examination, routine laboratory tests, endocrino- culated. Because of the small number of patients logic, nutritional and psychiatric evaluation, chest with the diagnosis of cancer, the prevalence by age X-ray, upper GI endoscopy, abdominal and thyroid was calculated in two age-groups: 21-46 and 47-70years. Unpaired t-test with 95% confidence interval Table 1. Co-morbidities presented in 1,333 obese
(CI) was applied to evaluate the differences between patients
Table 2. Minimally invasive bariatric treatment in
964 morbidly obese patients
LAGB = laparoscopic adjustable gastric banding; LAP-GBP = Hyperglycemia=impaired fasting glucose 110-126 mg/dL; laparoscopic gastric bypass; LAP-BPD-DS = laparoscopic bil- hyperinsulinemia=fasting insulin >25 µU/L; OSAS=obstructive iopancreatic diversion with duodenal switch; BIB = BioEnterics 1172 Obesity Surgery, 15, 2005
Prevalence of Cancer in Obese Patients variables. A two-tailed P<0.05 indicated statistical during the preoperative work-up for bariatric sur- significance. The analysis was done using GraphPad gery; and group C, patients diagnosed with cancer during the follow-up after bariatric surgery.
Group A: 31 patients (72.8%) were diagnosed and treated for their malignancies before the bariatric sur-geon visit, and 27 (87.1%) were female. Mean age, BMI and waist-hip ratio (WHR) at cancer diagnosis had been40.6 ± 8.8 years, 35.6 ± 7.8 kg/m2 and 0.91 ± 0,065, The BMI distribution in the 1,333 patients is shown respectively. Mean interval between surgery for cancer in Figure 1. BMI ranged from 30 to 82 kg/m2, and age and the first visit for bariatric surgery was 9.8 years.
ranged from 18 to 70 years. Morbid obesity was pres- Group B: Nine malignancies (20.1%) were dis- ent in 1,045 of the obese patients. Forty-three (3.2%) covered during the preoperative work-up or intraop- of the 1,333 obese patients (mean age 48.3±10.1 eratively. Mean age, BMI and WHR at cancer diag- years and mean BMI 43.9±6.8 at first bariatric visit) nosis was 44.9 ± 12.7 years, 47.1 ± 7.6 kg/m2 and presented different forms of malignant tumors. The predominant gender was female (88.3%). The site- Once the diagnosis was made, these patients under- specific prevalence of malignant disease in this went more specific investigations as indicated: blood tests, fine-needle aspiration for suspected thyroid can- The prevalence of cancer in the two age-groups cer, and abdominal CT scan for gastric adenocarci- (21-46 and 47-70 years) in the 1,333 obese patients noma, gastric GIST (GI stromal tumors) and adrenal adenocarcinoma, hematological consultation and spe- The younger group showed a higher prevalence of cific blood tests for multiple myeloma. Afterwards, cancer (2.1% vs 1.1%) and lower BMI than the these patients received appropriate treatment (Table 4).
older group (35.28±7.8 vs 42.9±8.48, P< 0.05, 95% Group C: Three of the 43 “cancer” patients (7%) developed malignant disease at a mean of 4.6 years We classified the 43 patients with cancer into 3 after bariatric surgery, with the diagnosis established subgroups: group A, patients treated for cancer at the scheduled follow-up visit. Mean age, BMI and before requiring treatment for morbid obesity; WHR at bariatric surgery were 39.6 ± 10.2 years, group B, patients diagnosed with malignant disease 48.2 ± 5.0 kg/m2 and 0.94 ± 0.07, respectively, whileage and BMI at cancer diagnosis were 44.3 ± 8.1 years and 40.4 ± 6.7 kg/m2, respectively. Patientsreceived indicated surgical treatment: left quadran-tectomy and axillary dissection with postoperative radiotherapy for infiltrating ductal breast carcinoma,ileal resection for carcinoid and total thyroidectomy and lymphadenectomy for papillary carcinoma.
Of the 43 “cancer” patients, 24 were morbidly obese at time of cancer diagnosis, 11 were obese(grade I or II), 7 were overweight and one was nor-mal weight. All 43 had been morbidly obese at thetime of their first bariatric visit.
Twenty-three morbidly obese patients (53.4%) received various bariatric operations: 12 laparo-scopic adjustable gastric banding, 3 laparoscopic biliopancreatic diversion with duodenal-switch, 1biliopancreatic diversion (Scopinaro), 1 laparo-scopic gastric bypass, 1 vertical banded gastro-plasty (Mason), and 5 BioEnterics intragastric bal- Figure 1. Distribution of BMI in the 1,333 obese patients.
loon. In one patient, bariatric surgery was con- Obesity Surgery, 15, 2005 1173
Table 3. Site of malignancy in this obese population
(1) Mean age at time of cancer diagnosis; (2) Mean BMI at cancer diagnosis; (3) Moment of cancer diagnosis reported to bariatricvisit / bariatric operation: pre=before, intraop=intraoperative, post=postoperative.
*GIST = GI stromal tumor.
traindicated for anesthesiologic reasons (definitive cer in our obese patients was breast (20.1%). An tracheotomy after operation for laryngeal cancer).
inverse relationship between BMI and breast cancerin premenopausal6,21,24,25 and a direct relationshipbetween postmenopausal women6 has been sug-gested. The increased serum concentration of bio- Discussion
available estradiol, from increased estrogen produc-tion by aromatase activity in excessive adipose tis- The results of this study showed that hormone- sue and from decreased serum concentration of sex- related tumors are the most prevalent in the mor- hormone-binding globulin, have been evoked as bidly obese population, consistent with previousreports.8,18-21 No clear biologic mechanism has been Table 4. Treatment in 9 morbidly obese patients
proved showing the connection between obesity and diagnosed with cancer/borderline tumors during
non-endocrine component cancers.21 Our obese sub- preoperative work-up or intraoperatively
jects presenting malignant neoplasms were younger than normal weight subjects, previously reported.22 More than 62% of the patients presented hor- mone-related tumors, in agreement with larger epi- demiological studies.18,19 The most frequent for can- prevalence (%)
Lap-BPD-DS=laparoscopic biliopancreatic diversion with duo- age groups
denal switch; LAGB=laparoscopic adjustable gastric banding.
Figure 2. Prevalence of cancer by age groups in 1,333
1174 Obesity Surgery, 15, 2005
Prevalence of Cancer in Obese Patients potential mechanisms.26-28 Thyroid cancer was the This study has limitations. First, it was retrospec- second most prevalent tumor (18.6%) in our data- tive, based on a database review, and no conclusion base and all these patients were female. Obesity- on risk of developing malignant neoplasm in obese related thyroid cancer has been previously subjects can be drawn. Second, the small number of described,12,13,29,30 with a higher prevalence in patients with cancer did not permit calculation of the men.11 As in the breast cancer group, thyroid cancer possible correlations among cancer and anthropomet- occurred at a relatively young age (mean age 37.8 ric and clinical variables. Nevertheless, this study may provide information on the prevalence of cancer Tumors of uterus, skin melanoma and gastric in an obese Italian population. In fact, we believe that leiomyomas have been also found in our patients.
the homogeneous population characteristics, the fact Two patients were found to have asymptomatic gas- that all patients were seen in a single academic tric GIST (GI stromal tumors), reported as stromal bariatric institution, the availability of demographic tumor with borderline malignancy. One gastric car- measurements (weight, height, BMI, WHR, co-mor- cinoma was discovered during the preoperative bidities), and the multidisciplinary surveillance work-up; no bariatric surgery was indicated in this (bariatric surgeon, endocrinologist, nutritionist, psy- patient, who underwent total gastrectomy.
chiatrist), giving a very high likelihood of detection, Although the prevalence of cancer in the Italian could represent advantages of this database study.
general population has been widely reported,31-33 Furthermore, this report underlines that accurate there has been no analysis from an Italian obese work-up and rigorous follow-up for the more fre- subjects database. Because different statistical quent malignancies should be beneficial in the man- methods were applied and there were larger cohorts of subjects in the general population studies, a com-parison with our study would be considered withcaution. The prevalence of cancer in our obese pop- Refferences
ulation was higher than the prevalence reported inthe ITAPREVAL study of the general populationfrom Southern Italy (1.3%) and similar to that in Northern Italy (3.6%).31 However, the prevalence of malignant tumors among our obese subjects was 2. Calle EE, Rodriguez C, Walker-Thurmond K et al.
higher than the average (2.7%) observed from all Overweight, obesity and mortality from cancer in a the areas covered by the Italian cancer registries.31 That analysis showed also that the prevalence of all malignancies combined was 3.1% for females and 3. Bergstrom A, Pisani P, Tent V et al. Overweight as an 2.2% for males.31 The majority of our obese sub- jects were females, so that their data are substan- tially in agreement with our finding (3.2%). Unlike 4. Hill HA, Austin H. Nutrition and endometrial cancer.
the data from Italian cancer registries on the general population (obese and non-obese),32,33 we found a 5. Chow W, Gridley G, Fraumeni J et al. Obesity, hyper- higher prevalence of cancer in the younger obese group compared with the older one. This finding may suggest an independent role of obesity in 6. Carmichael AR, Bates T. Obesity and breast cancer: a increasing the prevalence of cancer among younger obese subjects. However, our observations need to 7. Murphy TK, Calle E, Rodriguez C et al. Body mass be confirmed in a larger population of obese patients index and colon cancer mortality in a large prospec- In summary, this analysis found that morbidly 8. Wolk A, Gridley G, Svensson M et al. A prospective obese patients from this Italian cohort have various tumors with a higher prevalence of hormone-related site-specific, female gender and younger age. 9. IARC Handbooks of Cancer prevention. Weight con- Obesity Surgery, 15, 2005 1175
trol and physical activity. International Agency for 22. Calle EE, Thun MJ. Obesity and cancer 10. Berrington De Gonzalez A, Sweetland S et al. A meta- analysis of obesity and the risk of pancreatic cancer.
23. Cook PJ, Doll R, Fellingham SA. A mathematical 11. Samanic C, Gridley G, Chow WH et al. Obesity and cancer risk among white and black United States vet- 24. London SJ, Colditz GA, Stampfer MJ et al.
Prospective study of relative weight, height and risk 12. Kolonel LN, Hankin JH, Wilkens LR et al. An epi- of breast cancer. JAMA 1989; 262: 2853-8.
demiological study of thyroid cancer in Hawaii.
25. Vatten LJ, Kvinnsland S. Prospective study of relative height, body mass index and risk of breast cancer.
13. Ron E, Kleinerman RA, Boice JD et al. A population- based case-control study of thyroid cancer 26. Endogenous Hormones and Breast Cancer Collaborative Group. Body mass index, serum sex 14. Sistema statistico nazionale. Istituto nazionale di sta- hormones, and breast cancer risk in postmenopausal tistica. Le condizioni di salute della populazione.
Indagine multiscopo sulle famiglie. Condizioni della 27. Siiteri PK, Hammond GL, Nisker JA. Increased avail- salute e ricorso ai servizi sanitari. Anni 1999-2000.
ability of serum estrogens in breast cancer: a new Collana Informazioni, ed. 2001: 10-20.
hypothesis. In: Pike MC, Siiteri PK, Welsch CW, eds.
15. Ferlay J, Bray F, Pisani P et al. Cancer incidence, Banbury Report 8. Hormones and breast cancer. ColdSpring Harbor Laboratory, 1981: 87-106. mortality and prevalence worldwide. Version 1.0 28. Judd HL, Shamonki IM, Frumar AM et al. Origin of IARC CancerBases No. 5. Lyon, France: International 16. Rosso S, Spitale A, Balzi D et al. Estimate of cancer 29. Romiti A, Buzzetti R, Salandri A et al. Thyroid tumors incidence in Italian regions, 2001. Epidemiol Prev 30. Festuccia F, Romiti A, Buzzetti R et al. Latent thyroid 17. Orii T, Ohkohchi N, Sasaki K et al. Cholangio- carcinoma arising from preexisting biliary hamartoma 31. Micheli A, Francisci S, Krogh V et al. Cancer preva- lence in Italian cancer registry areas: the ITAPREVAL 18. Joint FAO/WHO/UNU Export Consultation. Energy and protein requirements. Geneva: WHO. Technical 32. Verdecchia A, Mariotto A, Capocaccia R et al.
Incidence and prevalence of all cancerous diseases in 19. Jonsson F, Wolk A, Pedersen NL et al. Obesity and Italy: trends and implications. Eur J Cancer 2001; 37: hormone-dependent tumors: cohort and co-twin con- trol studies based on the Swedish twin registry 33. Crocetti E, Capocaccia R, Casella C et al. Network of the Italian Cancer Registries (AIRT). Population-based 20. Gann PH, Hennekens CH, Ma J et al. Prospective incidence and mortality cancer trends (1986-1997) study of sex hormone levels and risk of prostate can- 21. Huang Z, Hankinson SE, Colditz GA et al. Dual effects of weight and weight gain on breast cancer (Received April 6, 2005; accepted May 31,2005) 1176 Obesity Surgery, 15, 2005

Source: http://chirurgie.ro/surgery.ro/BoruObesSurg2005.pdf