Colonoscopic Screening and Follow-Up in Patients with Acromegaly: A Multicenter Study in Italy
http://www.100md.com
《临床内分泌与代谢杂志》
Abstract
Acromegaly is an infrequent disease attributable to endogenous excess of GH and IGF-I. Human studies have associated the GH-IGF-I axis with the development of colorectal cancer; however, the question of whether colorectal cancer is a problem in acromegaly is currently unresolved. We performed a cross-sectional study to assess the risk of colonic neoplasia in patients with acromegaly. Colonoscopic screening was performed in 235 patients with acromegaly at five tertiary care hospitals in Italy between January 1, 1996, and December 31, 2001. A repeat colonoscopy was performed in 121 patients after a mean interval of 32.1 months. Colonoscopic findings in patients with acromegaly were compared with those of 233 patients with nonspecific abdominal complaints who were referred for endoscopy during the study period. A total of 65 patients (27.7%) and 36 controls (15.5%) had colonic neoplasia. In 55 patients (23.4%) and 34 control subjects (14.6%), the most important findings were adenomas (odds ratio, 1.7; range, 1.1–2.5), whereas 10 patients (4.3%) and two control subjects (0.9%) had carcinoma (odds ratio, 4.9; range, 1.1–22.4). The risk of colonic neoplasia was higher for younger patients with acromegaly compared with age-matched controls. Patients with acromegaly with or without colonic neoplasia did not differ significantly for IGF-I levels or duration of disease. A neoplastic recurrence was found in 16.5% of patients who underwent follow-up; 90% of them had had a neoplasm removed at the first colonoscopy. Acromegaly carries with it a moderate, but definitive, increase in the risk of colonic neoplasia that occurs at a younger age than in the general population. Patients who are found to harbor a colonic neoplasia are at risk for recurrence.
Introduction
ACROMEGALY IS AN infrequent disease attributable to endogenous excess of GH and IGF-I (1, 2). Acromegaly is associated with considerable morbidity and reduced life expectancy. Excess mortality is mostly due to cardiovascular disease and can be reduced when serum GH and IGF-I are effectively lowered by treatment (3, 4).
Human studies have associated the GH-IGF-I axis with the development of some of the most common cancers in the western world (5). Several population-based studies indicate that individuals with higher IGF-I levels within the normal range have an increased risk for colorectal cancer (6), breast cancer (7), and prostate cancer (8, 9), even if causality has not yet been established (10). Of particular interest in this context is a recent report that a common polymorphism in the human GH gene, which is associated with reduced circulating concentrations of GH and IGF-I, is inversely associated with colorectal cancer risk (11). Moreover, Swerdlow et al. (12) have very recently reported an increased risk of colorectal cancer in a cohort of GH-deficient patients who were treated with human pituitary GH at young ages between 1959 and 1985 in the United Kingdom. However, these findings cannot be considered conclusive because of the small number of patients with colorectal cancer.
The study of the association between acromegaly and cancer has attracted many researchers, but debate continues as to whether patients with acromegaly are at increased cancer risk (13, 14, 15, 16, 17, 18, 19, 20). There are several reasons that may explain why epidemiological studies of the cancer risk of patients with acromegaly have produced mixed results. First, all the studies are retrospective in nature, and many patients in the studied cohorts were enrolled in the early years, when the life expectancy of patients with acromegaly was much shorter than what it is currently, thanks to recent treatment improvement (4). In such patients, premature mortality because of cardiovascular or metabolic diseases could have prevented the development of malignancy, which appears to be an age-related phenomenon. It is only over the last decade that patients with acromegaly regularly entered the age when the incidence of cancer is remarkably increased; thus, competitive mortality may have represented a confounding factor, leading to underestimation of cancer risk (21). Second, the rarity of acromegaly itself makes it difficult to gather cohorts of patients large enough for meaningful statistical analysis. Third, these epidemiological studies are rather inhomogeneous concerning the demographic characteristics of the patients and the methodology of collection and analysis of data.
The association between acromegaly and colorectal cancer has been the most extensively investigated, and evidence for an increased risk of colonic cancer in acromegaly is stronger than that for any other tumor (16, 22, 23, 24, 25), although data are not entirely consistent (20). Most of the cohort studies using colonoscopy screening have reported an increased prevalence of either benign or malignant colonic tumors in patients with acromegaly (22, 23, 26, 27, 28, 29, 30, 31), but there are also reports at variance with this finding (32, 33, 34). There is only a cohort study on the colonoscopy follow-up of patients with acromegaly, suggesting that the occurrence of newly developed colonic adenomas is significantly related to elevated serum GH and IGF-I levels (35).
We therefore examined the risk of benign and malignant colorectal tumors in a cohort of patients with acromegaly who were enrolled in a program of screening colonoscopy at five different centers in Italy; a subgroup of patients underwent a repeat colonoscopy at varying intervals after their first exam.
Subjects and Methods
Inclusion criteria
The diagnosis of acromegaly was established clinically and confirmed by the following data: 1) high serum GH concentrations (>2.5 μg/liter as the mean of at least five samplings), 2) GH concentrations not suppressed below 1 μg/liter after administration of an oral glucose load (75 mg), 3) circulating IGF-I levels above the upper limit of the age-related reference range developed by the local laboratory of each center, and 4) demonstration of a pituitary tumor at neuroradiological imaging by computed tomographic scan or magnetic resonance. Patients who underwent pituitary surgery had histological confirmation of pituitary adenoma with positive tissue immunostaining for GH. The presumed duration of acromegaly was assessed by comparison of photographs and by patient interviews to date the onset of acral enlargement. The interval between the clinical onset of acromegaly and the time of definitive treatment, or the time when colonoscopy was performed, was calculated and recorded as the duration of disease. Hormones were measured in-house at each participating center using commercially available reagents. Patients were considered to have active disease when their IGF-I concentrations were higher than the upper limit of the age-related reference range, as it was defined in the local laboratory, and nadir GH concentrations after 75 mg oral glucose were greater than 1 μg/liter. Conversely, they were considered to have controlled disease when IGF-I concentrations were normal, and nadir GH concentrations were less than 1 μg/liter, during ongoing medical treatment. When such data were attained after a definitive treatment (surgery, radiotherapy, or both) without the need of concomitant medication, the patients were considered to be cured. Due to the multicentric nature of the study, IGF-I concentrations were expressed as percentages of the upper limit of the reference range. Specific gender- and age-adjusted reference ranges of IGF-I concentrations were defined at each center in large groups of healthy subjects. We chose not to use GH concentrations in the statistical analysis because absolute levels must be treated with circumspection when different assays are used (36). However, GH concentrations were used to assess disease activity. A comprehensive clinical history was taken by medical staff to record presenting symptoms and to identify risk factors for colorectal cancer, especially a family history of colorectal cancer, colonic polyps, or inflammatory bowel disease.
Subjects
We performed a retrospective cross-sectional analysis of patients with acromegaly who underwent colonoscopy screening for the first time from January 1, 1996, through December 31, 2001. The patients were recruited from five different centers in Italy, two of them were located in northern Italy (Turin and Milan), one was in central Italy (Pisa), and the other two were in southern Italy (Naples and San Giovanni Rotondo). The institutional ethics committee at each clinical center approved the study, and patients provided written informed consent. Since 1996, screening colonoscopy has been recommended to all patients with acromegaly referred to any of the five centers involved in the study. During the study period, 338 patients with acromegaly were newly diagnosed at the different centers: 35 patients were excluded for serious medical disorders that could increase the risk associated with colonoscopy, and 68 eligible patients declined participation. The remaining 235 subjects (77.5% of the eligible patients) underwent colonoscopy. A repeat colonoscopy was recommended to all patients, but 68 declined to participate, and 46 have not yet completed the interval period after the screening exam; a total of 121 patients underwent a second colonoscopy after a mean interval of 32.1 months.
Because of ethical and practical problems in performing colonoscopy in healthy subjects, consecutive patients presenting with nonspecific abdominal symptoms (intermittent abdominal pain, diarrhea, or constipation) who were referred for endoscopic assessment were chosen as controls. They gave their informed consent to total colonoscopy after the nature of the study was described. Patients who presented with high risk symptoms and signs (37), or patients who were referred for a suspected or a previously known colonic lesion, were excluded. A comprehensive clinical history was taken to identify risk factors for colorectal cancer, as it was for patients with acromegaly.
Endoscopic procedures
After careful bowel preparation using a polyethylene glycol-based electrolyte solution, total colonoscopy was performed with a video colonoscope while the patients were under conscious sedation. All colonoscopy examinations in the patients and controls were performed at the participating centers by the same board-certified endoscopists who had extensive experience with colonoscopy. During colonoscopy, the location and size of all polyps were determined before they were removed. Visible polyps were removed by snare and recovered. The size of the polyp was estimated with the use of open biopsy forceps. Small polyps (5 mm) were removed by cold snare transsection to avoid thermal destruction of the small lesions and provide appropriate and complete material to the pathologists. Polyps larger than 5 mm underwent classic diatherm resection. Endoscopists were required to provide photographic documentation of cecal landmarks and all important lesions. If the colonoscopic examination was incomplete because of problems with bowel preparation or failure to reach the cecum, the patient was asked to return for a second attempt. If a complete examination was performed within 1 month after the first attempt, a complete examination was reported, and the combined results of the two examinations were included in the analysis. For the purpose of our analysis, the boundary between the proximal colon and the distal colon was defined as the junction of the splenic flexure and the descending colon, as assessed by the endoscopist.
All retrieved polypoid lesions were sent to local pathology laboratories for histological evaluation. Pathological specimens were evaluated at each center by board-certified pathologists, who classified polyps according to the criteria established by the World Health Organization (38). Colonoscopic findings were categorized as indicating normal mucosa (no polyps), hyperplastic polyps, tubular adenomas, or cancer. Findings such as lipomas, lymphoid aggregates, or inflammatory polyps were categorized as indicating normal mucosa.
Statistical analysis
In patients with acromegaly, colonoscopic findings were evaluated in relation to age, gender, family history of colorectal cancer, duration and status of disease (active, controlled by medical treatment, or cured), and IGF-I concentrations. Data concerning colonoscopic findings at the first exam were compared with corresponding findings in controls in relation to demographic characteristics.
Database management and all statistical analyses were performed using the SPSS 11.0.1 software package (SPSS, Inc., Chicago, IL). Rates and proportions were calculated for categorical data, and means and SDs were determined for continuous data; 95% confidence intervals were always provided. The normality of data was assessed using the Kolmogorov-Smirnov test. For continuous variables, differences were analyzed by two-tailed t test when data were normally distributed and by Mann-Whitney U test for nonparametric data. For categorical variables, differences were analyzed by the 2 test and Fisher’s exact test. In addition, standard logistic regression methods were used to calculate odds ratio for benign and malignant colorectal tumors. Logistic regression analysis was used to investigate the relationship between the binary outcome cancer/adenoma vs. no colonic neoplasia and the predictor variables age, gender, family history of colorectal cancer, status and duration of acromegaly, and hormonal data. The level of statistical significance was set at P < 0.05.
Results
Screening colonoscopy
Selected demographic characteristics of the patient group are shown in Table 1. Demographic and clinical characteristics of the eligible patients who did not participate in the study [31 men and 37 women; aged 52.0 ± 14.3 yr; 53 of them (77.9%) with active disease] did not differ from those of the patients who were included. There were no major differences in the demographic characteristics of the patients across different centers. A total of 194 patients (82.5%) had active disease, 19 (8.1%) had controlled disease, and 22 (9.4%) had been cured by previous pituitary surgery and/or radiotherapy. None of the patients had a personal history of colonic disease (colitis, polyps, or cancer) or colonic surgery, or a colonic examination within the previous 10 yr (sigmoidoscopy, colonoscopy, or barium enema), whereas eight patients (3.4%) had a family history of colorectal cancer. Nineteen of the patients (8.1%) reported bleeding from the rectum or occasional episodes of abdominal pain or perianal symptoms.
A total of 233 consecutive subjects with nonspecific abdominal symptoms who were referred by their general practitioners for endoscopic assessment were enrolled as controls from January 1, 1999, through December 31, 2001. Selected demographic characteristics of the control group are shown in Table 1. None of the subjects had a personal history of colonic disease or colonic surgery, or colonic examination within the previous 10 yr, whereas four subjects (1.7%) had a family history of colorectal cancer.
The colonoscope was successfully inserted to the cecum with visualization of the ileocecal valve in all patients and controls. In 35 patients (14.7%), more than one procedure was required to complete the examination. There were no serious adverse events related to colonoscopy. Patients and control subjects were classified on the basis of the most advanced lesion found in the colon (Table 2). A total of 125 patients (53.2%) and 175 control subjects (75.1%) had no polypoid lesions. In 45 patients (19.1%) and 22 control subjects (9.4%), the most advanced lesions were hyperplastic polyps. In all, 72.3% of the patients and 84.5% of the controls had no evidence of neoplasia. A total of 65 patients (27.7%) and 36 controls (15.5%) had one or more adenomas of any type or invasive cancer. In 55 patients (23.4%) and 34 control subjects (14.6%), the most important findings were adenomas, whereas 10 patients (4.3%) and two control subjects (0.9%) had invasive cancer. Colonic adenomas were more frequently found in the distal colon in both patients (61.5%) and control subjects (74%). Among the patients with cancer, the stage was T1N0M0 in four patients, T1N1M0 in one, T2N0M0 in two, T2N1M0 in one, T3N0M0 in one, and T4N0 in one patient; five cancers (50%) were in the distal colon. Colonic cancers were detected in four men and six women with acromegaly; their mean age was 59.9 ± 10.7 yr. The patients with acromegaly had a significantly increased risk of colonic neoplasia compared with the control subjects [odds ratio for adenoma, 1.7 (95% confidence interval, 1.1–2.5); odds ratio for carcinoma, 4.9 (1.1–22.4)].
Patients with or without colonic neoplasia did not differ for age, gender, family history of colorectal cancer, status and duration of acromegaly, and IGF-I values, whereas control subjects with colonic neoplasia were older than subjects without neoplasia (Table 3). The different effect of age on the prevalence of colonic neoplasia between patients and controls is shown in Table 4. Multiple logistic regression analysis did not disclose any predictor variable to be significantly related to the outcome of colonoscopy.
Repeat colonoscopy
A repeat colonoscopy was performed in 121 patients (60 men and 61 women; mean age ± SD, 51.6 ± 12.7 yr) after a mean interval of 32.1 months (range, 12–132 months). The interval between the first and second colonoscopies was determined according to the suggestions of the endoscopist at the first evaluation. At the second colonoscopy, the colon was visualized to the cecum in all patients. Of the 121 patients who underwent repeat colonoscopy, 44 (36.4%) had had at least one neoplasm removed at the screening exam (six had cancer). When comparing the patients who underwent follow-up to the patients who did not, we found that the first group was selected for the presence of a colonic neoplasia at the first colonoscopy (Table 5). Eighteen of the 121 patients (14.9%) had at least one new adenoma, and two patients (1.7%) had a new colonic cancer. A new T1N0M0 cancer was found in a patient who previously had a T1N0M0 cancer, and a T2N0M0 cancer was found in a patient who previously had an adenoma; both cancers were found in the cecum. Eighteen of 44 patients (40.9%) with a previous neoplasm at the screening colonoscopy had a new neoplasm at the second colonoscopy, whereas only two of 77 patients (2.6%) with no previous neoplasm had a new adenoma (P < 0.0001). Overall, 90% of patients with a neoplastic recurrence had had a neoplasm removed at the first colonoscopy. At the second colonoscopy, the patients with colonic neoplasia had increased IGF-I levels and were older than patients without colonic neoplasia (Table 6). The presence of a colonic neoplasm at the screening colonoscopy was a predictive variable of neoplastic recurrence at the repeat exam (P = 0.0001).
Discussion
Concern about cancer risk in patients with acromegaly has focused primarily on colorectal cancer, yet the findings of previous studies are far from conclusive; thus, the question of whether colorectal cancer is a problem in acromegaly is currently passionately debated (21, 39, 40, 41, 42).
Increased risk of colorectal cancer could be assumed indirectly from the evidence of increased prevalence of adenomatous polyps in acromegaly. The data, however, are not entirely consistent, with rates ranging between 9% and 38% across different colonoscopic series (22, 23, 26, 27, 28, 29, 30, 31, 32, 33, 34). The reported figures for colorectal cancer ranged between 0% and 20% in such studies. Several reasons account for this discrepancy. The first reason is the small sample size of most studies, because only three of them included 100 patients or more (28, 31, 34). However, these studies produced conflicting results, because the prevalence of adenomatous polyps was found to be 9% (34), 22% (31), or 26% (28), respectively, whereas the prevalence of colorectal cancer was 0% (31), 2.6% (34), or 6.5% (28), respectively. Second, there was a rather low colonoscopy completion rate in some studies that reported a lower prevalence, because the cecum was reached in only about 70% of patients (27, 32, 33). Third, the absence of contemporary control groups in several studies prevented the assessment of relative risk (22, 23, 26, 28, 32, 33, 34). The studies with matched control groups reported significantly increased prevalence rates, because grouping the results of these studies gives an overall prevalence rate of 21% in acromegaly compared with 9% in controls for adenomatous polyps (relative risk, 2.36), and a prevalence of 3.7% for colorectal cancer in acromegaly compared with 0.5% in controls (relative risk, 13.4) (21).
In our study, the risk of either benign or malignant colonic neoplasia was increased in a cohort of 235 patients with acromegaly recruited from different centers in Italy between 1996 and 2001 compared with a control group of consecutive patients with nonspecific abdominal complaints. The prevalence of colonic cancer was increased in patients with acromegaly compared with control subjects; the relative risk was 4.9 (95% confidence interval, 1.1–22.4). The confidence interval was wide because of the small number of cancers observed in patients or controls; thus, our results should be considered cautiously. Also, the prevalence of adenomatous polyps was increased in acromegaly with a relative risk of 1.7 (95% confidence interval, 1.1–2.5). This last estimate seems more robust and is likely to give a more correct interpretation of the overall risk of colonic neoplasia in acromegaly. This estimate accords with the results of the three larger epidemiological studies assessing the risk of colonic cancer in acromegaly that consistently show a risk 2- to 3-fold higher than that in nonacromegalic subjects (16, 20, 25).
That the risk of colonic cancer might be increased in acromegaly is plausible for several reasons. There is an increasing body of literature suggesting that adults with circulating concentrations of IGF-I at the high end of the normal range are at increased risk of colorectal cancer (6, 43, 44). In some of these epidemiological studies, high IGF binding protein-3 (IGFBP-3) levels were associated with a lower risk of cancer (6, 43). In acromegaly, GH excess increases serum IGF-I and, to a lesser extent, IGFBP-3 concentrations, with the IGF-I to IGFBP-3 ratio being greater as GH concentrations increase (45, 46). Thus, an elevated IGF-I to IGFBP-3 ratio is expected to increase cancer risk in acromegaly (6, 47). IGF-I receptors and mRNAs for IGF-I have been identified on human colorectal cells (48). IGF-I is a known mitogen that may stimulate, by autocrine and paracrine actions, the proliferation of intestinal epithelial cells and their migration (49). In fact, increased proliferation of colonic epithelium, proportional to circulating IGF-I levels, has been demonstrated in acromegaly (50). Moreover, IGF-I is able to stimulate the growth of colorectal cancer cells in vitro, whereas blockade of IGF-I receptors inhibits cell growth in the same model (48, 51, 52).
It should be emphasized, however, that although GH and IGF-I are potential mitogens, this does not equate with transforming (tumorigenic) ability. Thus, it could be designated an alternative hypothesis, assuming that patients with pituitary adenomas, including those with acromegaly, may share common genetic defects that predispose them to develop either pituitary or extrapituitary neoplasms. Argument in favor of this theory of tumorigenesis in acromegaly may be found in the results of two previous retrospective studies. Popovic et al. (53) found an increased incidence of malignant tumors in patients with nonfunctioning pituitary adenomas, but not in patients with prolactinomas, compared with data generated by a local tumor registry. Barzilay et al. (17) also found a slight increase in the standardized incidence rate of malignancy in patients with nonfunctioning pituitary adenomas or prolactinomas. Future research assessing the prevalence of colonic tumors by colonoscopy in patients with non-GH-secreting pituitary adenomas is warranted to clarify the mechanisms of colon tumorigenesis in acromegaly and to elucidate the merits of the genetic theory, assuming that patients with pituitary adenomas are at increased risk independent of GH status, because genetic factors underlie either pituitary or colon neoplasia, as opposed to the endocrine theory, maintaining a pivotal role for IGF-I in the process of colon neoplasia formation in states of sustained GH excess.
We are aware of the limit of considering patients with nonspecific gastrointestinal symptoms as controls, but there is evidence that these patients are not at lower risk for serious colonic pathology than asymptomatic patients (54). Our control group was highly comparable for age, whereas the preponderance of men, compared with that in the patient group, did not bias the finding of increased prevalence of colonic neoplasia in acromegaly, because men have higher prevalence rates than women (55, 56). The prevalence rate of colorectal cancer in our controls (0.9%) accords with the 0.6–1.0% found in recently published series of colonoscopic screening in asymptomatic subjects (56, 57, 58). The rate of adenomatous polyps in our controls (14.6%) accords with that found in the study by Imperiale et al. (55) (11%), but it is lower than that found by Lieberman et al. (56) (36.5%). In their study population, however, about 97% of the enrolled subjects were men, and as many as 13.9% had a family history of colorectal cancer (56); thus, the results of that study are not fully applicable to the general population. Moreover, the mean age of the subjects was 62.9 yr in that series compared with 50.8 yr in the present one. Therefore, we believe that our control subjects are not a selected group at a lower than average risk for colonic neoplasia.
We also disclose the limit of the multicentric and partially retrospective nature of our study. Notwithstanding the fact that different operators were involved, the polyp detection rate was highly comparable among the five centers, because every endoscopist had had extensive experience with colonoscopy in patients with acromegaly, in whom the procedure may be particularly demanding because of increased colonic length (40). The expertise of the endoscopists is reflected by the high rate of successful cecal intubation at the first attempt. Since 1996, screening colonoscopy has been recommended to patients with acromegaly referred to each center involved in the study, and 77.5% of the eligible patients observed during the study period underwent colonoscopy. The other patients did not consent to undergo colonoscopy; thus, we believe that our cohort is representative and not intentionally selected. Endoscopic follow-up was recommended to all patients, but only 51.4% of the whole series have up completed the surveillance program to date. As expected, the patients who underwent follow-up were selected by the presence of colonic neoplasia at the first colonoscopy.
In our study the risk of colonic neoplasia in acromegaly was not influenced by male gender and increasing age, in striking contrast with what has been demonstrated in the general population (55, 56, 57). It is conceivable that the effect of GH and IGF-I excess on cancer risk may be more important than that of these demographic factors. The fact that younger patients usually display a more severe course of acromegaly, whereas patients who have milder disease are diagnosed later in life (58), is consistent with the present finding of an increased prevalence of colonic neoplasia in patients under 50 yr of age. The difference with age-matched controls was less remarkable in older age groups. Previous colonoscopic studies in acromegaly have provided conflicting results on whether male gender and older age carry an increased risk of colonic neoplasia (22, 23, 26, 27, 28, 29, 30, 31, 32, 33, 34), but most of these studies were too limited in size to adequately assess the roles of these factors.
We failed to demonstrate any difference in IGF-I levels or in the share of patients with active disease between patients with or without neoplasia at screening colonoscopy, in agreement with all but one previous series (34). This finding does not discredit the hypothesis that GH and IGF-I are implicated in colorectal tumorigenesis. A single hormonal evaluation at a random moment in the course of a long-lasting disease such as acromegaly cannot fully reflect the degree of GH and IGF-I excess to which patients have been exposed over the preceding years. Moreover, the estimate of disease duration is admittedly imprecise and largely operator dependent. We have indeed observed a great variability among the five centers in the reporting of disease duration that may have compounded appreciation of its pathogenetic role, if any. Not surprisingly, previous studies have produced contradictory results on this topic (22, 23, 26, 27, 28, 29, 30, 31, 32, 33, 34). It has been recently suggested that the role of IGF-I should be better appreciated in a follow-up study looking for the development of new colonic neoplasms rather than in a transversal study, where the lesions detected have been present for a variable, possibly long, time (35).
In a recent study, 66 patients underwent a second colonoscopic evaluation after a mean interval of 33 months from the original screening examination, when all visible lesions had been removed. The occurrence of newly developed colonic adenomas was found in 14% of patients and was significantly associated with elevated circulating levels of GH and IGF-I and with a previous adenoma at the original colonoscopy (35). The present study of a cohort of 121 patients, who underwent colonoscopic follow-up after a mean interval of 32.1 months, confirms that high IGF-I concentrations may be implicated in the development of recurrent colonic adenoma and cancer in acromegaly. However, the strongest predictor of recurrence was the previous detection of a colonic neoplasm. The rate of recurrence of colonic neoplasia was 16.5% in the overall series and 40.9% among patients who had a previous colonic neoplasm. Although it cannot be determined whether some lesions were missed at the initial colonoscopy and subsequently detected at the second, the new neoplasias were found by the same endoscopists who had adequately examined the full colon at the initial exam. Moreover, such a bias would have occurred equally among all patients; thus, our results give a representative estimate of the factors associated with neoplasm recurrence. The aim of our study was not to evaluate whether patients with acromegaly and colonic adenomas have a higher recurrence rate than general population, because we did not repeat colonoscopy in our control subjects. In a study of surveillance colonoscopy, however, the percentage of recurrent adenomas ranged between 32–41.7% (59) among patients who were subjected to a second colonoscopy 3 yr after colonoscopic removal of adenomatous polyps. This figure is comparable to that found in our patients with acromegaly.
In conclusion, the present results suggest that acromegaly carries with it a moderate, but definitive, increase in the risk of colonic neoplasia that occurs at a younger age in patients with acromegaly than in general population. The moderate increase in the relative risk may explain why some previous small studies have failed to demonstrate it. Our data are biologically plausible and consistent with the larger epidemiological observations in acromegaly (16, 20, 25). Based on our findings, it is advisable to include colonoscopy in the management strategy of patients with acromegaly and to submit to follow-up patients who have been found to harbor colonic neoplasia. Patients with more severe disease may be at increased risk for the development and recurrence of colonic neoplasm. However, our results should be considered cautiously with respect to the factors associated with increased risk for the limited numbers in some subgroup analyses and the possibility of selection for patients who underwent follow-up.
Therefore, we emphasize the need for more data from other sources to allow firm identification of groups of patients at higher risk that may have important implications for developing cost-effective strategies of screening and surveillance for colonic cancer in patients with acromegaly.
Footnotes
This work was supported by the Ministero dell’Università e della Ricerca Scientifica e Tecnologica (Grant 2002068252). The funding source had no role in the collection, analysis, and interpretation of the data or in the decision to submit the results for publication.
First Published Online October 26, 2004
Abbreviation: IGFBP-3, IGF binding protein-3.
Received February 10, 2004.
Accepted October 15, 2004.
References
Melmed S 1990 Acromegaly. N Engl J Med 322:966–977
Colao A, Lombardi G 1998 Growth-hormone and prolactin excess. Lancet 352:1455–1461
Giustina A, Barkan A, Casanueva FF, Cavagnini F, Frohman L, Ho K, Veldhuis J, Wass J, Von Werder K, Melmed S 2000 Criteria for cure of acromegaly: a consensus statement. J Clin Endocrinol Metab 85:526–529
Melmed S, Jackson I, Kleinberg D, Klibanski A 1998 Current treatment guidelines for acromegaly. J Clin Endocrinol Metab 83:2646–2652
Khandwala HM, McCutcheon IE, Flyvbjerg A, Friend KE 2000 The effects of insulin-like growth factors on tumorigenesis and neoplastic growth. Endocr Rev 21:215–244
Ma J, Pollak MN, Giovannucci E, Chan JM, Tao Y, Hennekens CH, Stampfer MJ 1999 Prospective study of colorectal cancer risk in men and plasma levels of insulin-like growth factor (IGF)-I and IGF-binding protein-3. J Natl Cancer Inst 91:620–625
Hankinson SE, Willett WC, Colditz GA, Hunter DJ, Michaud DS, Deroo DB, Rosner B, Speizer FE, Pollak M 1998 Circulating concentrations of insulin-like growth factor-I and risk of breast cancer. Lancet 351:1393–1396
Wolk A, Mantzoros CS, Andersson SO, Bergstrom R, Signorello LB, Lagiou B, Adami HO, Trichopoulos D 1998 Insulin-like growth factor 1 and prostate cancer risk: a population-based, case-control study. J Natl Cancer Inst 90:911–915
Chan JM, Stampfer MJ, Giovannucci E, Gann PH, Ma J, Wilkinson P, Hennekens CH, Pollak M 1998 Plasma insulin-like growth factor-I and prostate cancer risk: a prospective study. Science 279:563–566
Bustin SA, Jenkins PJ 2001 The growth hormone-insulin-like growth factor-I axis and colorectal cancer. Trends Mol Med 7:447–454
Le Marchand L, Donlon T, Seifried A, Kaaks R, Rinaldi S, Wilkens LR 2002 Association of a common polymorphism in the human GH1 gene with colorectal neoplasia. J Natl Cancer Inst 94:454–460
Swerdlow AJ, Higgins CD, Adlard P, Preece MA 2002 Risk of cancer in patients treated with human pituitary growth hormone in the UK, 1959–85: a cohort study. Lancet 360:273–277
Wright AD, Hill DM, Lowy C, Fraser TR 1970 Mortality in acromegaly. Q J Med 39:1–16
Alexander L, Appleton D, Hall R, Ross WM, Wilkinson R 1980 Epidemiology of acromegaly in the Newcastle region. Clin Endocrinol (Oxf) 12:71–79
Bengtsson BA, Eden S, Ernest I, Oden A, Sjogren B 1988 Epidemiology and long-term survival in acromegaly. A study of 166 cases diagnosed between 1955 and 1984. Acta Med Scand 223:327–335
Ron E, Gridley G, Hrubec Z, Page W, Arora S, Fraumeni Jr JF 1991 Acromegaly and gastrointestinal cancer. Cancer 68:1673–1677
Barzilay J, Heatley GJ, Cushing GW 1991 Benign and malignant tumors in patiens with acromegaly. Arch Intern Med 151:1629–1632
Nabarro JD 1987 Acromegaly. Clin Endocrinol (Oxf) 26:481–512
Cheung NW, Boyages SC 1997 Increased incidence of neoplasia in females with acromegaly. Clin Endocrinol (Oxf) 47:323–327
Orme SM, McNally RJ, Cartwright RA, Belchetz PE 1998 Mortality and cancer incidence in acromegaly: a retrospective cohort study. United Kingdom Acromegaly Study Group. J Clin Endocrinol Metab 83:2730–2734
Jenkins PJ, Besser M 2001 Acromegaly and cancer: a problem. J Clin Endocrinol Metab 86:2935–2941
Klein I, Parveen G, Gavaler JS, Vanthiel DH 1982 Colonic polyps in patients with acromegaly. Ann Intern Med 97:27–30
Ituarte EM, Petrini J, Hershaman JM 1984 Acromegaly and colon cancer. Ann Intern Med 101:627–628[Medline]
Pines A, Rozen P, Ron E, Gilat T 1985 Gastrointestinal tumors in acromegalic patients. Am J Gastroenterol 80:266–269
Baris D, Gridley G, Ron E, Weiderpass E, Mellemkjaer L, Ekbom A, Olsen JH, Baron JA, Fraumeni Jr JF 2002 Acromegaly and cancer risk: a cohort study in Sweden and Denmark. Cancer Causes Control 13:395–400
Ezzat S, Strom C, Melmed S 1991 Colon polyps in acromegaly. Ann Intern Med 114:754–755
Vasen HF, van Erpecum KJ, Roelfsema F, Raue F, Koppeschaar H, Griffioen G, van Berge Henegouwen GP 1994 Increased prevalence of colonic adenomas in patients with acromegaly. Eur J Endocrinol 131:235–237
Jenkins PJ, Fairclough PD, Richards T, Lowe DG, Monson J, Grossman A, Wass JA, Besser M 1997 Acromegaly, colonic polyps and carcinoma. Clin Endocrinol (Oxf) 47:17–22
Terzolo M, Tappero G, Borretta G, Asnaghi G, Pia A, Reimondo G, Boccuzzi A, Cesario F, Rovero E, Paccotti P, Angeli A 1994 High prevalence of colonic polyps in patients with acromegaly. Influence of sex and age. Arch Intern Med 154:1272–1276
Colao A, Balzano A, Ferone D, Panza N, Grande G, Marzullo P, Bove A, Iodice G, Merola B, Lombardi G 1997 Increased prevalence of colonic polyps and altered lymphocyte subset pattern in the colonic lamina propria in acromegaly. Clin Endocrinol (Oxf) 47:23–28
Delhougne B, Deneux C, Abs R, Chanson P, Fierens H, Laurent-Puig P, Duysburgh I, Stevenaert A, Tabarin A, Delwaide J 1995 The prevalence of colonic polyps in acromegaly: a colonoscopic and pathological study in 103 patients. J Natl Cancer Inst 80:3223–3226
Ortego J, Vega B, Sampedro J, Escalada J, Boixeda D, Varela C 1994 Neoplastic colonic polyps in acromegaly. Horm Metab Res 26:609–610
Ladas SD, Thalassinos NC, Ioannides G, Raptis SA 1994 Does acromegaly really predispose to an increased prevalence of gastrointestinal tumours? Clin Endocrinol (Oxf) 41:597–601
Renehan AG, Bhaskar P, Painter JE, O’Dwyer ST, Haboubi N, Varma J, Ball SG, Shalet SM 2000 The prevalence and characteristics of colorectal neoplasia in acromegaly. J Clin Endocrinol Metab 85:3417–3424
Jenkins PJ, Frajese V, Jones AM, Camacho-Hubner C, Lowe DG, Fairclough PD, Chew SL, Grossman AB, Monson JP, Besser GM 2000 Insulin-like growth factor I and the development of colorectal neoplasia in acromegaly. J Clin Endocrinol Metab 85:3218–3221
Wass JAH 2003 Dynamic testing in the diagnosis and follow-up of patients with acromegaly. J Endocrinol Invest 26(Suppl 7):48–53
Talley NJ, Spiller R 2002 Irritable bowel syndrome: a little understood organic bowel disease? Lancet 360:555–564
Konishi F, Morson BC 1982 Pathology of colorectal adenomas: a colonoscopic survey. J Clin Pathol 35:830–841
Melmed S 2001 Acromegaly and cancer: not a problem? J Clin Endocrinol Metab 86:2929–2934
Jenkins PJ, Fairclough PD 2001 Colorectal neoplasia in acromegaly. Clin Endocrinol (Oxf) 55:727–729
Atkin WS 2001 Risk of colorectal neoplasia in acromegaly: an independent view. Clin Endocrinol (Oxf) 55:723–725
Renehan AG, Odwyer ST, Shalet SM 2001 Screening colonoscopy for acromegaly in perspective. Clin Endocrinol (Oxf) 55:731–733
Giovannucci E, Pollak MN, Platz EA, Willett WC, Stampfer MJ, Majeed N, Colditz GA, Speizer FE, Hankinson SE 2000 A prospective study of plasma insulin-like growth factor-1 and binding protein-3 and risk of colorectal neoplasia in women. Cancer Epidemiol Biomarkers Prev 9:345–349
Kaaks R, Toniolo P, Akhmedkhanov A, Lukanova A, Biessy C, Dechaud H, Rinaldi S, Zeleniuch-Jacquotte A, Shore RE, Riboli E 2000 Serum C-peptide, insulin-like growth factor (IGF)-I, IGF-binding proteins, and colorectal cancer risk in women. J Natl Cancer Inst 92:1592–1600
Juul A, Pedersen SA, Sorensen S, Winkler K, Jorgensen JO, Christiansen JS, Skakkebaek NE 1994 Growth hormone (GH) treatment increases serum insulin-like growth factor binding protein-3, bone isoenzyme alkaline phosphatase and forearm bone mineral content in young adults with GH deficiency of childhood onset. Eur J Endocrinol 131:41–49
Ghigo E, Aimaretti G, Maccario M, Fanciulli G, Arvat E, Minuto F, Giordano G, Delitala G, Camanni F 1999 Dose-response study of GH effects on circulating IGF-I and IGFBP-3 levels in healthy young men and women. Am J Physiol 276:E1009–E1013
Giovannucci E, Pollak M 2002 Risk of cancer after growth-hormone treatment. Lancet 360:268–269
Lahm H, Amstad P, Wyniger J, Yilmaz A, Fischer JR, Schreyer M, Givel JC 1994 Blockade of the insulin-like growth-factor-I receptor inhibits growth of human colorectal cancer cells: evidence of a functional IGF-II-mediated autocrine loop. Int J Cancer 58:452–459
Simmons JG, Pucilowska JB, Lund PK 1999 Autocrine and paracrine actions of intestinal fibroblast-derived insulin-like growth factors. Am J Physiol 276:G817–G827
Cats A, Dullaart RP, Kleibeuker JH, Kuipers F, Sluiter WJ, Hardonk MJ, de Vries EG 1996 Increased epithelial cell proliferation in the colon of patients with acromegaly. Cancer Res 56:523–526
Durrant LG, Watson SA, Hall A, Morris DL 1991 Co-stimulation of gastrointestinal tumour cell growth by gastrin, transforming growth factor and insulin like growth factor-I. Br J Cancer 63:67–70
Lahm H, Suardet L, Laurent PL, Fischer JR, Ceyhan A, Givel JC, Odartchenko N 1992 Growth regulation and co-stimulation of human colorectal cancer cell lines by insulin-like growth factor I, II and transforming growth factor . Br J Cancer 65:341–346
Popovic V, Damjanovic S, Micic D, Nesovic M, Djurovic M, Petakov M, Obradovic S, Zoric S, Simic M, Penezic Z, Marinkovic J; Pituitary Study Group 1998 Increased incidence of neoplasia in patients with pituitary adenomas. Clin Endocrinol (Oxf) 49:441–445
Lieberman DA, de Garmo PL, Fleischer DE, Eisen GM, Chan BK, Helfand M 2000 Colonic neoplasia in patients with nonspecific GI symptoms. Gastrointest Endosc 51:647–651
Imperiale TF, Wagner DR, Lin CY, Larkin GN, Rogge JD, Ransohoff DF 2000 Risk of advanced proximal neoplasms in asymptomatic adults according to the distal colorectal findings. N Engl J Med 343:169–174
Lieberman DA, Weiss DG, Bond JH, Ahnen DJ, Garewal H, Chejfec G 2000 Use of colonoscopy to screen asymptomatic adults for colorectal cancer. Veterans Affairs Cooperative Study Group 380. N Engl J Med 343:162–168
Rex DK, Lehman GA, Ulbright TM, Smith JJ, Pound DC, Hawes RH, Helper DJ, Wiersema MJ, Langefeld CD, Li W 1993 Colonic neoplasia in asymptomatic persons with negative fecal occult blood tests: influence of age, gender, and family history. Am J Gastroenterol 88:825–831
van der Lely AJ, Harris AG, Lamberts SW 1992 The sensitivity of growth hormone secretion to medical treatment in acromegalic patients: influence of age and sex. Clin Endocrinol (Oxf) 37:181–185
Winawer SJ, Zauber AG, O’Brien MJ, Ho MN, Gottlieb L, Sternberg SS, Waye JD, Bond J, Schapiro M, Stewart ET, Panish J, Ackroyd F, Kurtz R, Shike M 1993 Randomized comparison of surveillance intervals after colonoscopic removal of newly diagnosed adenomatous polyps. The National Polyp Study Workgroup. N Engl J Med 328:901–906(Vitale, Alfredo Scillitan)
Acromegaly is an infrequent disease attributable to endogenous excess of GH and IGF-I. Human studies have associated the GH-IGF-I axis with the development of colorectal cancer; however, the question of whether colorectal cancer is a problem in acromegaly is currently unresolved. We performed a cross-sectional study to assess the risk of colonic neoplasia in patients with acromegaly. Colonoscopic screening was performed in 235 patients with acromegaly at five tertiary care hospitals in Italy between January 1, 1996, and December 31, 2001. A repeat colonoscopy was performed in 121 patients after a mean interval of 32.1 months. Colonoscopic findings in patients with acromegaly were compared with those of 233 patients with nonspecific abdominal complaints who were referred for endoscopy during the study period. A total of 65 patients (27.7%) and 36 controls (15.5%) had colonic neoplasia. In 55 patients (23.4%) and 34 control subjects (14.6%), the most important findings were adenomas (odds ratio, 1.7; range, 1.1–2.5), whereas 10 patients (4.3%) and two control subjects (0.9%) had carcinoma (odds ratio, 4.9; range, 1.1–22.4). The risk of colonic neoplasia was higher for younger patients with acromegaly compared with age-matched controls. Patients with acromegaly with or without colonic neoplasia did not differ significantly for IGF-I levels or duration of disease. A neoplastic recurrence was found in 16.5% of patients who underwent follow-up; 90% of them had had a neoplasm removed at the first colonoscopy. Acromegaly carries with it a moderate, but definitive, increase in the risk of colonic neoplasia that occurs at a younger age than in the general population. Patients who are found to harbor a colonic neoplasia are at risk for recurrence.
Introduction
ACROMEGALY IS AN infrequent disease attributable to endogenous excess of GH and IGF-I (1, 2). Acromegaly is associated with considerable morbidity and reduced life expectancy. Excess mortality is mostly due to cardiovascular disease and can be reduced when serum GH and IGF-I are effectively lowered by treatment (3, 4).
Human studies have associated the GH-IGF-I axis with the development of some of the most common cancers in the western world (5). Several population-based studies indicate that individuals with higher IGF-I levels within the normal range have an increased risk for colorectal cancer (6), breast cancer (7), and prostate cancer (8, 9), even if causality has not yet been established (10). Of particular interest in this context is a recent report that a common polymorphism in the human GH gene, which is associated with reduced circulating concentrations of GH and IGF-I, is inversely associated with colorectal cancer risk (11). Moreover, Swerdlow et al. (12) have very recently reported an increased risk of colorectal cancer in a cohort of GH-deficient patients who were treated with human pituitary GH at young ages between 1959 and 1985 in the United Kingdom. However, these findings cannot be considered conclusive because of the small number of patients with colorectal cancer.
The study of the association between acromegaly and cancer has attracted many researchers, but debate continues as to whether patients with acromegaly are at increased cancer risk (13, 14, 15, 16, 17, 18, 19, 20). There are several reasons that may explain why epidemiological studies of the cancer risk of patients with acromegaly have produced mixed results. First, all the studies are retrospective in nature, and many patients in the studied cohorts were enrolled in the early years, when the life expectancy of patients with acromegaly was much shorter than what it is currently, thanks to recent treatment improvement (4). In such patients, premature mortality because of cardiovascular or metabolic diseases could have prevented the development of malignancy, which appears to be an age-related phenomenon. It is only over the last decade that patients with acromegaly regularly entered the age when the incidence of cancer is remarkably increased; thus, competitive mortality may have represented a confounding factor, leading to underestimation of cancer risk (21). Second, the rarity of acromegaly itself makes it difficult to gather cohorts of patients large enough for meaningful statistical analysis. Third, these epidemiological studies are rather inhomogeneous concerning the demographic characteristics of the patients and the methodology of collection and analysis of data.
The association between acromegaly and colorectal cancer has been the most extensively investigated, and evidence for an increased risk of colonic cancer in acromegaly is stronger than that for any other tumor (16, 22, 23, 24, 25), although data are not entirely consistent (20). Most of the cohort studies using colonoscopy screening have reported an increased prevalence of either benign or malignant colonic tumors in patients with acromegaly (22, 23, 26, 27, 28, 29, 30, 31), but there are also reports at variance with this finding (32, 33, 34). There is only a cohort study on the colonoscopy follow-up of patients with acromegaly, suggesting that the occurrence of newly developed colonic adenomas is significantly related to elevated serum GH and IGF-I levels (35).
We therefore examined the risk of benign and malignant colorectal tumors in a cohort of patients with acromegaly who were enrolled in a program of screening colonoscopy at five different centers in Italy; a subgroup of patients underwent a repeat colonoscopy at varying intervals after their first exam.
Subjects and Methods
Inclusion criteria
The diagnosis of acromegaly was established clinically and confirmed by the following data: 1) high serum GH concentrations (>2.5 μg/liter as the mean of at least five samplings), 2) GH concentrations not suppressed below 1 μg/liter after administration of an oral glucose load (75 mg), 3) circulating IGF-I levels above the upper limit of the age-related reference range developed by the local laboratory of each center, and 4) demonstration of a pituitary tumor at neuroradiological imaging by computed tomographic scan or magnetic resonance. Patients who underwent pituitary surgery had histological confirmation of pituitary adenoma with positive tissue immunostaining for GH. The presumed duration of acromegaly was assessed by comparison of photographs and by patient interviews to date the onset of acral enlargement. The interval between the clinical onset of acromegaly and the time of definitive treatment, or the time when colonoscopy was performed, was calculated and recorded as the duration of disease. Hormones were measured in-house at each participating center using commercially available reagents. Patients were considered to have active disease when their IGF-I concentrations were higher than the upper limit of the age-related reference range, as it was defined in the local laboratory, and nadir GH concentrations after 75 mg oral glucose were greater than 1 μg/liter. Conversely, they were considered to have controlled disease when IGF-I concentrations were normal, and nadir GH concentrations were less than 1 μg/liter, during ongoing medical treatment. When such data were attained after a definitive treatment (surgery, radiotherapy, or both) without the need of concomitant medication, the patients were considered to be cured. Due to the multicentric nature of the study, IGF-I concentrations were expressed as percentages of the upper limit of the reference range. Specific gender- and age-adjusted reference ranges of IGF-I concentrations were defined at each center in large groups of healthy subjects. We chose not to use GH concentrations in the statistical analysis because absolute levels must be treated with circumspection when different assays are used (36). However, GH concentrations were used to assess disease activity. A comprehensive clinical history was taken by medical staff to record presenting symptoms and to identify risk factors for colorectal cancer, especially a family history of colorectal cancer, colonic polyps, or inflammatory bowel disease.
Subjects
We performed a retrospective cross-sectional analysis of patients with acromegaly who underwent colonoscopy screening for the first time from January 1, 1996, through December 31, 2001. The patients were recruited from five different centers in Italy, two of them were located in northern Italy (Turin and Milan), one was in central Italy (Pisa), and the other two were in southern Italy (Naples and San Giovanni Rotondo). The institutional ethics committee at each clinical center approved the study, and patients provided written informed consent. Since 1996, screening colonoscopy has been recommended to all patients with acromegaly referred to any of the five centers involved in the study. During the study period, 338 patients with acromegaly were newly diagnosed at the different centers: 35 patients were excluded for serious medical disorders that could increase the risk associated with colonoscopy, and 68 eligible patients declined participation. The remaining 235 subjects (77.5% of the eligible patients) underwent colonoscopy. A repeat colonoscopy was recommended to all patients, but 68 declined to participate, and 46 have not yet completed the interval period after the screening exam; a total of 121 patients underwent a second colonoscopy after a mean interval of 32.1 months.
Because of ethical and practical problems in performing colonoscopy in healthy subjects, consecutive patients presenting with nonspecific abdominal symptoms (intermittent abdominal pain, diarrhea, or constipation) who were referred for endoscopic assessment were chosen as controls. They gave their informed consent to total colonoscopy after the nature of the study was described. Patients who presented with high risk symptoms and signs (37), or patients who were referred for a suspected or a previously known colonic lesion, were excluded. A comprehensive clinical history was taken to identify risk factors for colorectal cancer, as it was for patients with acromegaly.
Endoscopic procedures
After careful bowel preparation using a polyethylene glycol-based electrolyte solution, total colonoscopy was performed with a video colonoscope while the patients were under conscious sedation. All colonoscopy examinations in the patients and controls were performed at the participating centers by the same board-certified endoscopists who had extensive experience with colonoscopy. During colonoscopy, the location and size of all polyps were determined before they were removed. Visible polyps were removed by snare and recovered. The size of the polyp was estimated with the use of open biopsy forceps. Small polyps (5 mm) were removed by cold snare transsection to avoid thermal destruction of the small lesions and provide appropriate and complete material to the pathologists. Polyps larger than 5 mm underwent classic diatherm resection. Endoscopists were required to provide photographic documentation of cecal landmarks and all important lesions. If the colonoscopic examination was incomplete because of problems with bowel preparation or failure to reach the cecum, the patient was asked to return for a second attempt. If a complete examination was performed within 1 month after the first attempt, a complete examination was reported, and the combined results of the two examinations were included in the analysis. For the purpose of our analysis, the boundary between the proximal colon and the distal colon was defined as the junction of the splenic flexure and the descending colon, as assessed by the endoscopist.
All retrieved polypoid lesions were sent to local pathology laboratories for histological evaluation. Pathological specimens were evaluated at each center by board-certified pathologists, who classified polyps according to the criteria established by the World Health Organization (38). Colonoscopic findings were categorized as indicating normal mucosa (no polyps), hyperplastic polyps, tubular adenomas, or cancer. Findings such as lipomas, lymphoid aggregates, or inflammatory polyps were categorized as indicating normal mucosa.
Statistical analysis
In patients with acromegaly, colonoscopic findings were evaluated in relation to age, gender, family history of colorectal cancer, duration and status of disease (active, controlled by medical treatment, or cured), and IGF-I concentrations. Data concerning colonoscopic findings at the first exam were compared with corresponding findings in controls in relation to demographic characteristics.
Database management and all statistical analyses were performed using the SPSS 11.0.1 software package (SPSS, Inc., Chicago, IL). Rates and proportions were calculated for categorical data, and means and SDs were determined for continuous data; 95% confidence intervals were always provided. The normality of data was assessed using the Kolmogorov-Smirnov test. For continuous variables, differences were analyzed by two-tailed t test when data were normally distributed and by Mann-Whitney U test for nonparametric data. For categorical variables, differences were analyzed by the 2 test and Fisher’s exact test. In addition, standard logistic regression methods were used to calculate odds ratio for benign and malignant colorectal tumors. Logistic regression analysis was used to investigate the relationship between the binary outcome cancer/adenoma vs. no colonic neoplasia and the predictor variables age, gender, family history of colorectal cancer, status and duration of acromegaly, and hormonal data. The level of statistical significance was set at P < 0.05.
Results
Screening colonoscopy
Selected demographic characteristics of the patient group are shown in Table 1. Demographic and clinical characteristics of the eligible patients who did not participate in the study [31 men and 37 women; aged 52.0 ± 14.3 yr; 53 of them (77.9%) with active disease] did not differ from those of the patients who were included. There were no major differences in the demographic characteristics of the patients across different centers. A total of 194 patients (82.5%) had active disease, 19 (8.1%) had controlled disease, and 22 (9.4%) had been cured by previous pituitary surgery and/or radiotherapy. None of the patients had a personal history of colonic disease (colitis, polyps, or cancer) or colonic surgery, or a colonic examination within the previous 10 yr (sigmoidoscopy, colonoscopy, or barium enema), whereas eight patients (3.4%) had a family history of colorectal cancer. Nineteen of the patients (8.1%) reported bleeding from the rectum or occasional episodes of abdominal pain or perianal symptoms.
A total of 233 consecutive subjects with nonspecific abdominal symptoms who were referred by their general practitioners for endoscopic assessment were enrolled as controls from January 1, 1999, through December 31, 2001. Selected demographic characteristics of the control group are shown in Table 1. None of the subjects had a personal history of colonic disease or colonic surgery, or colonic examination within the previous 10 yr, whereas four subjects (1.7%) had a family history of colorectal cancer.
The colonoscope was successfully inserted to the cecum with visualization of the ileocecal valve in all patients and controls. In 35 patients (14.7%), more than one procedure was required to complete the examination. There were no serious adverse events related to colonoscopy. Patients and control subjects were classified on the basis of the most advanced lesion found in the colon (Table 2). A total of 125 patients (53.2%) and 175 control subjects (75.1%) had no polypoid lesions. In 45 patients (19.1%) and 22 control subjects (9.4%), the most advanced lesions were hyperplastic polyps. In all, 72.3% of the patients and 84.5% of the controls had no evidence of neoplasia. A total of 65 patients (27.7%) and 36 controls (15.5%) had one or more adenomas of any type or invasive cancer. In 55 patients (23.4%) and 34 control subjects (14.6%), the most important findings were adenomas, whereas 10 patients (4.3%) and two control subjects (0.9%) had invasive cancer. Colonic adenomas were more frequently found in the distal colon in both patients (61.5%) and control subjects (74%). Among the patients with cancer, the stage was T1N0M0 in four patients, T1N1M0 in one, T2N0M0 in two, T2N1M0 in one, T3N0M0 in one, and T4N0 in one patient; five cancers (50%) were in the distal colon. Colonic cancers were detected in four men and six women with acromegaly; their mean age was 59.9 ± 10.7 yr. The patients with acromegaly had a significantly increased risk of colonic neoplasia compared with the control subjects [odds ratio for adenoma, 1.7 (95% confidence interval, 1.1–2.5); odds ratio for carcinoma, 4.9 (1.1–22.4)].
Patients with or without colonic neoplasia did not differ for age, gender, family history of colorectal cancer, status and duration of acromegaly, and IGF-I values, whereas control subjects with colonic neoplasia were older than subjects without neoplasia (Table 3). The different effect of age on the prevalence of colonic neoplasia between patients and controls is shown in Table 4. Multiple logistic regression analysis did not disclose any predictor variable to be significantly related to the outcome of colonoscopy.
Repeat colonoscopy
A repeat colonoscopy was performed in 121 patients (60 men and 61 women; mean age ± SD, 51.6 ± 12.7 yr) after a mean interval of 32.1 months (range, 12–132 months). The interval between the first and second colonoscopies was determined according to the suggestions of the endoscopist at the first evaluation. At the second colonoscopy, the colon was visualized to the cecum in all patients. Of the 121 patients who underwent repeat colonoscopy, 44 (36.4%) had had at least one neoplasm removed at the screening exam (six had cancer). When comparing the patients who underwent follow-up to the patients who did not, we found that the first group was selected for the presence of a colonic neoplasia at the first colonoscopy (Table 5). Eighteen of the 121 patients (14.9%) had at least one new adenoma, and two patients (1.7%) had a new colonic cancer. A new T1N0M0 cancer was found in a patient who previously had a T1N0M0 cancer, and a T2N0M0 cancer was found in a patient who previously had an adenoma; both cancers were found in the cecum. Eighteen of 44 patients (40.9%) with a previous neoplasm at the screening colonoscopy had a new neoplasm at the second colonoscopy, whereas only two of 77 patients (2.6%) with no previous neoplasm had a new adenoma (P < 0.0001). Overall, 90% of patients with a neoplastic recurrence had had a neoplasm removed at the first colonoscopy. At the second colonoscopy, the patients with colonic neoplasia had increased IGF-I levels and were older than patients without colonic neoplasia (Table 6). The presence of a colonic neoplasm at the screening colonoscopy was a predictive variable of neoplastic recurrence at the repeat exam (P = 0.0001).
Discussion
Concern about cancer risk in patients with acromegaly has focused primarily on colorectal cancer, yet the findings of previous studies are far from conclusive; thus, the question of whether colorectal cancer is a problem in acromegaly is currently passionately debated (21, 39, 40, 41, 42).
Increased risk of colorectal cancer could be assumed indirectly from the evidence of increased prevalence of adenomatous polyps in acromegaly. The data, however, are not entirely consistent, with rates ranging between 9% and 38% across different colonoscopic series (22, 23, 26, 27, 28, 29, 30, 31, 32, 33, 34). The reported figures for colorectal cancer ranged between 0% and 20% in such studies. Several reasons account for this discrepancy. The first reason is the small sample size of most studies, because only three of them included 100 patients or more (28, 31, 34). However, these studies produced conflicting results, because the prevalence of adenomatous polyps was found to be 9% (34), 22% (31), or 26% (28), respectively, whereas the prevalence of colorectal cancer was 0% (31), 2.6% (34), or 6.5% (28), respectively. Second, there was a rather low colonoscopy completion rate in some studies that reported a lower prevalence, because the cecum was reached in only about 70% of patients (27, 32, 33). Third, the absence of contemporary control groups in several studies prevented the assessment of relative risk (22, 23, 26, 28, 32, 33, 34). The studies with matched control groups reported significantly increased prevalence rates, because grouping the results of these studies gives an overall prevalence rate of 21% in acromegaly compared with 9% in controls for adenomatous polyps (relative risk, 2.36), and a prevalence of 3.7% for colorectal cancer in acromegaly compared with 0.5% in controls (relative risk, 13.4) (21).
In our study, the risk of either benign or malignant colonic neoplasia was increased in a cohort of 235 patients with acromegaly recruited from different centers in Italy between 1996 and 2001 compared with a control group of consecutive patients with nonspecific abdominal complaints. The prevalence of colonic cancer was increased in patients with acromegaly compared with control subjects; the relative risk was 4.9 (95% confidence interval, 1.1–22.4). The confidence interval was wide because of the small number of cancers observed in patients or controls; thus, our results should be considered cautiously. Also, the prevalence of adenomatous polyps was increased in acromegaly with a relative risk of 1.7 (95% confidence interval, 1.1–2.5). This last estimate seems more robust and is likely to give a more correct interpretation of the overall risk of colonic neoplasia in acromegaly. This estimate accords with the results of the three larger epidemiological studies assessing the risk of colonic cancer in acromegaly that consistently show a risk 2- to 3-fold higher than that in nonacromegalic subjects (16, 20, 25).
That the risk of colonic cancer might be increased in acromegaly is plausible for several reasons. There is an increasing body of literature suggesting that adults with circulating concentrations of IGF-I at the high end of the normal range are at increased risk of colorectal cancer (6, 43, 44). In some of these epidemiological studies, high IGF binding protein-3 (IGFBP-3) levels were associated with a lower risk of cancer (6, 43). In acromegaly, GH excess increases serum IGF-I and, to a lesser extent, IGFBP-3 concentrations, with the IGF-I to IGFBP-3 ratio being greater as GH concentrations increase (45, 46). Thus, an elevated IGF-I to IGFBP-3 ratio is expected to increase cancer risk in acromegaly (6, 47). IGF-I receptors and mRNAs for IGF-I have been identified on human colorectal cells (48). IGF-I is a known mitogen that may stimulate, by autocrine and paracrine actions, the proliferation of intestinal epithelial cells and their migration (49). In fact, increased proliferation of colonic epithelium, proportional to circulating IGF-I levels, has been demonstrated in acromegaly (50). Moreover, IGF-I is able to stimulate the growth of colorectal cancer cells in vitro, whereas blockade of IGF-I receptors inhibits cell growth in the same model (48, 51, 52).
It should be emphasized, however, that although GH and IGF-I are potential mitogens, this does not equate with transforming (tumorigenic) ability. Thus, it could be designated an alternative hypothesis, assuming that patients with pituitary adenomas, including those with acromegaly, may share common genetic defects that predispose them to develop either pituitary or extrapituitary neoplasms. Argument in favor of this theory of tumorigenesis in acromegaly may be found in the results of two previous retrospective studies. Popovic et al. (53) found an increased incidence of malignant tumors in patients with nonfunctioning pituitary adenomas, but not in patients with prolactinomas, compared with data generated by a local tumor registry. Barzilay et al. (17) also found a slight increase in the standardized incidence rate of malignancy in patients with nonfunctioning pituitary adenomas or prolactinomas. Future research assessing the prevalence of colonic tumors by colonoscopy in patients with non-GH-secreting pituitary adenomas is warranted to clarify the mechanisms of colon tumorigenesis in acromegaly and to elucidate the merits of the genetic theory, assuming that patients with pituitary adenomas are at increased risk independent of GH status, because genetic factors underlie either pituitary or colon neoplasia, as opposed to the endocrine theory, maintaining a pivotal role for IGF-I in the process of colon neoplasia formation in states of sustained GH excess.
We are aware of the limit of considering patients with nonspecific gastrointestinal symptoms as controls, but there is evidence that these patients are not at lower risk for serious colonic pathology than asymptomatic patients (54). Our control group was highly comparable for age, whereas the preponderance of men, compared with that in the patient group, did not bias the finding of increased prevalence of colonic neoplasia in acromegaly, because men have higher prevalence rates than women (55, 56). The prevalence rate of colorectal cancer in our controls (0.9%) accords with the 0.6–1.0% found in recently published series of colonoscopic screening in asymptomatic subjects (56, 57, 58). The rate of adenomatous polyps in our controls (14.6%) accords with that found in the study by Imperiale et al. (55) (11%), but it is lower than that found by Lieberman et al. (56) (36.5%). In their study population, however, about 97% of the enrolled subjects were men, and as many as 13.9% had a family history of colorectal cancer (56); thus, the results of that study are not fully applicable to the general population. Moreover, the mean age of the subjects was 62.9 yr in that series compared with 50.8 yr in the present one. Therefore, we believe that our control subjects are not a selected group at a lower than average risk for colonic neoplasia.
We also disclose the limit of the multicentric and partially retrospective nature of our study. Notwithstanding the fact that different operators were involved, the polyp detection rate was highly comparable among the five centers, because every endoscopist had had extensive experience with colonoscopy in patients with acromegaly, in whom the procedure may be particularly demanding because of increased colonic length (40). The expertise of the endoscopists is reflected by the high rate of successful cecal intubation at the first attempt. Since 1996, screening colonoscopy has been recommended to patients with acromegaly referred to each center involved in the study, and 77.5% of the eligible patients observed during the study period underwent colonoscopy. The other patients did not consent to undergo colonoscopy; thus, we believe that our cohort is representative and not intentionally selected. Endoscopic follow-up was recommended to all patients, but only 51.4% of the whole series have up completed the surveillance program to date. As expected, the patients who underwent follow-up were selected by the presence of colonic neoplasia at the first colonoscopy.
In our study the risk of colonic neoplasia in acromegaly was not influenced by male gender and increasing age, in striking contrast with what has been demonstrated in the general population (55, 56, 57). It is conceivable that the effect of GH and IGF-I excess on cancer risk may be more important than that of these demographic factors. The fact that younger patients usually display a more severe course of acromegaly, whereas patients who have milder disease are diagnosed later in life (58), is consistent with the present finding of an increased prevalence of colonic neoplasia in patients under 50 yr of age. The difference with age-matched controls was less remarkable in older age groups. Previous colonoscopic studies in acromegaly have provided conflicting results on whether male gender and older age carry an increased risk of colonic neoplasia (22, 23, 26, 27, 28, 29, 30, 31, 32, 33, 34), but most of these studies were too limited in size to adequately assess the roles of these factors.
We failed to demonstrate any difference in IGF-I levels or in the share of patients with active disease between patients with or without neoplasia at screening colonoscopy, in agreement with all but one previous series (34). This finding does not discredit the hypothesis that GH and IGF-I are implicated in colorectal tumorigenesis. A single hormonal evaluation at a random moment in the course of a long-lasting disease such as acromegaly cannot fully reflect the degree of GH and IGF-I excess to which patients have been exposed over the preceding years. Moreover, the estimate of disease duration is admittedly imprecise and largely operator dependent. We have indeed observed a great variability among the five centers in the reporting of disease duration that may have compounded appreciation of its pathogenetic role, if any. Not surprisingly, previous studies have produced contradictory results on this topic (22, 23, 26, 27, 28, 29, 30, 31, 32, 33, 34). It has been recently suggested that the role of IGF-I should be better appreciated in a follow-up study looking for the development of new colonic neoplasms rather than in a transversal study, where the lesions detected have been present for a variable, possibly long, time (35).
In a recent study, 66 patients underwent a second colonoscopic evaluation after a mean interval of 33 months from the original screening examination, when all visible lesions had been removed. The occurrence of newly developed colonic adenomas was found in 14% of patients and was significantly associated with elevated circulating levels of GH and IGF-I and with a previous adenoma at the original colonoscopy (35). The present study of a cohort of 121 patients, who underwent colonoscopic follow-up after a mean interval of 32.1 months, confirms that high IGF-I concentrations may be implicated in the development of recurrent colonic adenoma and cancer in acromegaly. However, the strongest predictor of recurrence was the previous detection of a colonic neoplasm. The rate of recurrence of colonic neoplasia was 16.5% in the overall series and 40.9% among patients who had a previous colonic neoplasm. Although it cannot be determined whether some lesions were missed at the initial colonoscopy and subsequently detected at the second, the new neoplasias were found by the same endoscopists who had adequately examined the full colon at the initial exam. Moreover, such a bias would have occurred equally among all patients; thus, our results give a representative estimate of the factors associated with neoplasm recurrence. The aim of our study was not to evaluate whether patients with acromegaly and colonic adenomas have a higher recurrence rate than general population, because we did not repeat colonoscopy in our control subjects. In a study of surveillance colonoscopy, however, the percentage of recurrent adenomas ranged between 32–41.7% (59) among patients who were subjected to a second colonoscopy 3 yr after colonoscopic removal of adenomatous polyps. This figure is comparable to that found in our patients with acromegaly.
In conclusion, the present results suggest that acromegaly carries with it a moderate, but definitive, increase in the risk of colonic neoplasia that occurs at a younger age in patients with acromegaly than in general population. The moderate increase in the relative risk may explain why some previous small studies have failed to demonstrate it. Our data are biologically plausible and consistent with the larger epidemiological observations in acromegaly (16, 20, 25). Based on our findings, it is advisable to include colonoscopy in the management strategy of patients with acromegaly and to submit to follow-up patients who have been found to harbor colonic neoplasia. Patients with more severe disease may be at increased risk for the development and recurrence of colonic neoplasm. However, our results should be considered cautiously with respect to the factors associated with increased risk for the limited numbers in some subgroup analyses and the possibility of selection for patients who underwent follow-up.
Therefore, we emphasize the need for more data from other sources to allow firm identification of groups of patients at higher risk that may have important implications for developing cost-effective strategies of screening and surveillance for colonic cancer in patients with acromegaly.
Footnotes
This work was supported by the Ministero dell’Università e della Ricerca Scientifica e Tecnologica (Grant 2002068252). The funding source had no role in the collection, analysis, and interpretation of the data or in the decision to submit the results for publication.
First Published Online October 26, 2004
Abbreviation: IGFBP-3, IGF binding protein-3.
Received February 10, 2004.
Accepted October 15, 2004.
References
Melmed S 1990 Acromegaly. N Engl J Med 322:966–977
Colao A, Lombardi G 1998 Growth-hormone and prolactin excess. Lancet 352:1455–1461
Giustina A, Barkan A, Casanueva FF, Cavagnini F, Frohman L, Ho K, Veldhuis J, Wass J, Von Werder K, Melmed S 2000 Criteria for cure of acromegaly: a consensus statement. J Clin Endocrinol Metab 85:526–529
Melmed S, Jackson I, Kleinberg D, Klibanski A 1998 Current treatment guidelines for acromegaly. J Clin Endocrinol Metab 83:2646–2652
Khandwala HM, McCutcheon IE, Flyvbjerg A, Friend KE 2000 The effects of insulin-like growth factors on tumorigenesis and neoplastic growth. Endocr Rev 21:215–244
Ma J, Pollak MN, Giovannucci E, Chan JM, Tao Y, Hennekens CH, Stampfer MJ 1999 Prospective study of colorectal cancer risk in men and plasma levels of insulin-like growth factor (IGF)-I and IGF-binding protein-3. J Natl Cancer Inst 91:620–625
Hankinson SE, Willett WC, Colditz GA, Hunter DJ, Michaud DS, Deroo DB, Rosner B, Speizer FE, Pollak M 1998 Circulating concentrations of insulin-like growth factor-I and risk of breast cancer. Lancet 351:1393–1396
Wolk A, Mantzoros CS, Andersson SO, Bergstrom R, Signorello LB, Lagiou B, Adami HO, Trichopoulos D 1998 Insulin-like growth factor 1 and prostate cancer risk: a population-based, case-control study. J Natl Cancer Inst 90:911–915
Chan JM, Stampfer MJ, Giovannucci E, Gann PH, Ma J, Wilkinson P, Hennekens CH, Pollak M 1998 Plasma insulin-like growth factor-I and prostate cancer risk: a prospective study. Science 279:563–566
Bustin SA, Jenkins PJ 2001 The growth hormone-insulin-like growth factor-I axis and colorectal cancer. Trends Mol Med 7:447–454
Le Marchand L, Donlon T, Seifried A, Kaaks R, Rinaldi S, Wilkens LR 2002 Association of a common polymorphism in the human GH1 gene with colorectal neoplasia. J Natl Cancer Inst 94:454–460
Swerdlow AJ, Higgins CD, Adlard P, Preece MA 2002 Risk of cancer in patients treated with human pituitary growth hormone in the UK, 1959–85: a cohort study. Lancet 360:273–277
Wright AD, Hill DM, Lowy C, Fraser TR 1970 Mortality in acromegaly. Q J Med 39:1–16
Alexander L, Appleton D, Hall R, Ross WM, Wilkinson R 1980 Epidemiology of acromegaly in the Newcastle region. Clin Endocrinol (Oxf) 12:71–79
Bengtsson BA, Eden S, Ernest I, Oden A, Sjogren B 1988 Epidemiology and long-term survival in acromegaly. A study of 166 cases diagnosed between 1955 and 1984. Acta Med Scand 223:327–335
Ron E, Gridley G, Hrubec Z, Page W, Arora S, Fraumeni Jr JF 1991 Acromegaly and gastrointestinal cancer. Cancer 68:1673–1677
Barzilay J, Heatley GJ, Cushing GW 1991 Benign and malignant tumors in patiens with acromegaly. Arch Intern Med 151:1629–1632
Nabarro JD 1987 Acromegaly. Clin Endocrinol (Oxf) 26:481–512
Cheung NW, Boyages SC 1997 Increased incidence of neoplasia in females with acromegaly. Clin Endocrinol (Oxf) 47:323–327
Orme SM, McNally RJ, Cartwright RA, Belchetz PE 1998 Mortality and cancer incidence in acromegaly: a retrospective cohort study. United Kingdom Acromegaly Study Group. J Clin Endocrinol Metab 83:2730–2734
Jenkins PJ, Besser M 2001 Acromegaly and cancer: a problem. J Clin Endocrinol Metab 86:2935–2941
Klein I, Parveen G, Gavaler JS, Vanthiel DH 1982 Colonic polyps in patients with acromegaly. Ann Intern Med 97:27–30
Ituarte EM, Petrini J, Hershaman JM 1984 Acromegaly and colon cancer. Ann Intern Med 101:627–628[Medline]
Pines A, Rozen P, Ron E, Gilat T 1985 Gastrointestinal tumors in acromegalic patients. Am J Gastroenterol 80:266–269
Baris D, Gridley G, Ron E, Weiderpass E, Mellemkjaer L, Ekbom A, Olsen JH, Baron JA, Fraumeni Jr JF 2002 Acromegaly and cancer risk: a cohort study in Sweden and Denmark. Cancer Causes Control 13:395–400
Ezzat S, Strom C, Melmed S 1991 Colon polyps in acromegaly. Ann Intern Med 114:754–755
Vasen HF, van Erpecum KJ, Roelfsema F, Raue F, Koppeschaar H, Griffioen G, van Berge Henegouwen GP 1994 Increased prevalence of colonic adenomas in patients with acromegaly. Eur J Endocrinol 131:235–237
Jenkins PJ, Fairclough PD, Richards T, Lowe DG, Monson J, Grossman A, Wass JA, Besser M 1997 Acromegaly, colonic polyps and carcinoma. Clin Endocrinol (Oxf) 47:17–22
Terzolo M, Tappero G, Borretta G, Asnaghi G, Pia A, Reimondo G, Boccuzzi A, Cesario F, Rovero E, Paccotti P, Angeli A 1994 High prevalence of colonic polyps in patients with acromegaly. Influence of sex and age. Arch Intern Med 154:1272–1276
Colao A, Balzano A, Ferone D, Panza N, Grande G, Marzullo P, Bove A, Iodice G, Merola B, Lombardi G 1997 Increased prevalence of colonic polyps and altered lymphocyte subset pattern in the colonic lamina propria in acromegaly. Clin Endocrinol (Oxf) 47:23–28
Delhougne B, Deneux C, Abs R, Chanson P, Fierens H, Laurent-Puig P, Duysburgh I, Stevenaert A, Tabarin A, Delwaide J 1995 The prevalence of colonic polyps in acromegaly: a colonoscopic and pathological study in 103 patients. J Natl Cancer Inst 80:3223–3226
Ortego J, Vega B, Sampedro J, Escalada J, Boixeda D, Varela C 1994 Neoplastic colonic polyps in acromegaly. Horm Metab Res 26:609–610
Ladas SD, Thalassinos NC, Ioannides G, Raptis SA 1994 Does acromegaly really predispose to an increased prevalence of gastrointestinal tumours? Clin Endocrinol (Oxf) 41:597–601
Renehan AG, Bhaskar P, Painter JE, O’Dwyer ST, Haboubi N, Varma J, Ball SG, Shalet SM 2000 The prevalence and characteristics of colorectal neoplasia in acromegaly. J Clin Endocrinol Metab 85:3417–3424
Jenkins PJ, Frajese V, Jones AM, Camacho-Hubner C, Lowe DG, Fairclough PD, Chew SL, Grossman AB, Monson JP, Besser GM 2000 Insulin-like growth factor I and the development of colorectal neoplasia in acromegaly. J Clin Endocrinol Metab 85:3218–3221
Wass JAH 2003 Dynamic testing in the diagnosis and follow-up of patients with acromegaly. J Endocrinol Invest 26(Suppl 7):48–53
Talley NJ, Spiller R 2002 Irritable bowel syndrome: a little understood organic bowel disease? Lancet 360:555–564
Konishi F, Morson BC 1982 Pathology of colorectal adenomas: a colonoscopic survey. J Clin Pathol 35:830–841
Melmed S 2001 Acromegaly and cancer: not a problem? J Clin Endocrinol Metab 86:2929–2934
Jenkins PJ, Fairclough PD 2001 Colorectal neoplasia in acromegaly. Clin Endocrinol (Oxf) 55:727–729
Atkin WS 2001 Risk of colorectal neoplasia in acromegaly: an independent view. Clin Endocrinol (Oxf) 55:723–725
Renehan AG, Odwyer ST, Shalet SM 2001 Screening colonoscopy for acromegaly in perspective. Clin Endocrinol (Oxf) 55:731–733
Giovannucci E, Pollak MN, Platz EA, Willett WC, Stampfer MJ, Majeed N, Colditz GA, Speizer FE, Hankinson SE 2000 A prospective study of plasma insulin-like growth factor-1 and binding protein-3 and risk of colorectal neoplasia in women. Cancer Epidemiol Biomarkers Prev 9:345–349
Kaaks R, Toniolo P, Akhmedkhanov A, Lukanova A, Biessy C, Dechaud H, Rinaldi S, Zeleniuch-Jacquotte A, Shore RE, Riboli E 2000 Serum C-peptide, insulin-like growth factor (IGF)-I, IGF-binding proteins, and colorectal cancer risk in women. J Natl Cancer Inst 92:1592–1600
Juul A, Pedersen SA, Sorensen S, Winkler K, Jorgensen JO, Christiansen JS, Skakkebaek NE 1994 Growth hormone (GH) treatment increases serum insulin-like growth factor binding protein-3, bone isoenzyme alkaline phosphatase and forearm bone mineral content in young adults with GH deficiency of childhood onset. Eur J Endocrinol 131:41–49
Ghigo E, Aimaretti G, Maccario M, Fanciulli G, Arvat E, Minuto F, Giordano G, Delitala G, Camanni F 1999 Dose-response study of GH effects on circulating IGF-I and IGFBP-3 levels in healthy young men and women. Am J Physiol 276:E1009–E1013
Giovannucci E, Pollak M 2002 Risk of cancer after growth-hormone treatment. Lancet 360:268–269
Lahm H, Amstad P, Wyniger J, Yilmaz A, Fischer JR, Schreyer M, Givel JC 1994 Blockade of the insulin-like growth-factor-I receptor inhibits growth of human colorectal cancer cells: evidence of a functional IGF-II-mediated autocrine loop. Int J Cancer 58:452–459
Simmons JG, Pucilowska JB, Lund PK 1999 Autocrine and paracrine actions of intestinal fibroblast-derived insulin-like growth factors. Am J Physiol 276:G817–G827
Cats A, Dullaart RP, Kleibeuker JH, Kuipers F, Sluiter WJ, Hardonk MJ, de Vries EG 1996 Increased epithelial cell proliferation in the colon of patients with acromegaly. Cancer Res 56:523–526
Durrant LG, Watson SA, Hall A, Morris DL 1991 Co-stimulation of gastrointestinal tumour cell growth by gastrin, transforming growth factor and insulin like growth factor-I. Br J Cancer 63:67–70
Lahm H, Suardet L, Laurent PL, Fischer JR, Ceyhan A, Givel JC, Odartchenko N 1992 Growth regulation and co-stimulation of human colorectal cancer cell lines by insulin-like growth factor I, II and transforming growth factor . Br J Cancer 65:341–346
Popovic V, Damjanovic S, Micic D, Nesovic M, Djurovic M, Petakov M, Obradovic S, Zoric S, Simic M, Penezic Z, Marinkovic J; Pituitary Study Group 1998 Increased incidence of neoplasia in patients with pituitary adenomas. Clin Endocrinol (Oxf) 49:441–445
Lieberman DA, de Garmo PL, Fleischer DE, Eisen GM, Chan BK, Helfand M 2000 Colonic neoplasia in patients with nonspecific GI symptoms. Gastrointest Endosc 51:647–651
Imperiale TF, Wagner DR, Lin CY, Larkin GN, Rogge JD, Ransohoff DF 2000 Risk of advanced proximal neoplasms in asymptomatic adults according to the distal colorectal findings. N Engl J Med 343:169–174
Lieberman DA, Weiss DG, Bond JH, Ahnen DJ, Garewal H, Chejfec G 2000 Use of colonoscopy to screen asymptomatic adults for colorectal cancer. Veterans Affairs Cooperative Study Group 380. N Engl J Med 343:162–168
Rex DK, Lehman GA, Ulbright TM, Smith JJ, Pound DC, Hawes RH, Helper DJ, Wiersema MJ, Langefeld CD, Li W 1993 Colonic neoplasia in asymptomatic persons with negative fecal occult blood tests: influence of age, gender, and family history. Am J Gastroenterol 88:825–831
van der Lely AJ, Harris AG, Lamberts SW 1992 The sensitivity of growth hormone secretion to medical treatment in acromegalic patients: influence of age and sex. Clin Endocrinol (Oxf) 37:181–185
Winawer SJ, Zauber AG, O’Brien MJ, Ho MN, Gottlieb L, Sternberg SS, Waye JD, Bond J, Schapiro M, Stewart ET, Panish J, Ackroyd F, Kurtz R, Shike M 1993 Randomized comparison of surveillance intervals after colonoscopic removal of newly diagnosed adenomatous polyps. The National Polyp Study Workgroup. N Engl J Med 328:901–906(Vitale, Alfredo Scillitan)