Article Text
Abstract
Background and aims Intraductal papillary mucinous neoplasms (IPMNs) of the pancreas have been reported to be associated with extrapancreatic malignancies, but there have been no prospective studies evaluating the incidence of extrapancreatic cancers in patients with IPMNs. In this study, the frequency of patients with IPMNs developing extrapancreatic cancers during follow-up was examined.
Methods 642 patients with IPMNs were prospectively followed up for 4.8 years on average. They underwent clinical examination at least twice a year. The incidence of the observed extrapancreatic malignancies was compared with the expected incidence of the age- and gender-matched general Japanese population based on the Vital Statistics of Japan.
Results 40 extrapancreatic cancers developed in 39 patients during follow-up (1.3% per year). The most common malignancies were hepatocellular (n=7), colorectal (n=6), gastric (n=6), lung (n=5) and prostate cancers (n=4). The calculated standardised incidence rate of each of the malignancies was not significant, as follows: hepatocellular 2.17 (95% CI 0.87 to 4.47), colorectal 1.02 (0.37 to 2.21), gastric 0.76 (0.28 to 1.66), lung 0.75 (0.24 to 1.76) and prostate 1.00 (0.71 to 1.29). Pancreatic cancer was observed in 17 patients during the same period, with a standardised incidence rate of 10.7 (6.2 to 17.1).
Conclusions The incidence of extrapancreatic malignancies in patients with IPMNs was similar to that in the general population, whereas the incidence of pancreatic cancers was significantly high. Intraductal papillary mucinous neoplasms are not associated with systemic carcinogenesis except for pancreatic cancer.
- Intraductal papillary mucinous neoplasm
- standardised incidence rate
- extrapancreatic malignancy
- carcinogenesis
- pancreatic disease
- pancreatic epidemiology
Statistics from Altmetric.com
- Intraductal papillary mucinous neoplasm
- standardised incidence rate
- extrapancreatic malignancy
- carcinogenesis
- pancreatic disease
- pancreatic epidemiology
Significance of this study
What is already known about this subject?
The prevalence of extrapancreatic malignancies in patients with intraductal papillary mucinous neoplasms (IPMNs) was 10–38% according to previous reports.
What are the new findings?
The incidence of extrapancreatic malignancies in patients with IPMNs was similar to that in the general population, in contrast to the high incidence of pancreatic cancers.
How might it impact on clinical practice in the foreseeable future?
Intensive screening for extrapancreatic cancers is not necessary in patients with IPMNs during follow-up.
Introduction
The widespread availability of non-invasive imaging modalities as well as growing awareness among clinicians have led to increasing rates of identification of intraductal papillary mucinous neoplasms (IPMNs) of the pancreas.1–4 The management of IPMNs, surgically or conservatively, is challenging because of their wide-ranging histological variety from adenoma to invasive carcinoma.5–9 Furthermore, long-term observation is needed even with a diagnosis of benign adenoma, because their potential to develop into pancreatic cancers is distinct.3 4 10
In addition to pancreatic cancer,11–13 several previous reports have suggested that patients with IPMNs have a greater risk of harbouring extrapancreatic malignancies compared with patients with pancreatic ductal adenocarcinoma or other pancreatic cystic neoplasms, or referral patients.14–20 According to these studies, the frequency of extrapancreatic malignancies in patients with IPMNs was 10–38%. However, these reports were retrospective studies or included only operatively managed IPMNs, leading to significant recruitment bias. IPMNs were often diagnosed at advanced age, making it likely for the frequency of extrapancreatic malignancies to be high.
There have been no prospective studies that estimated the actual incidence of extrapancreatic malignancies in all patients with both operatively and non-operatively managed IPMNs. In this study, we prospectively evaluated the incidence of extrapancreatic malignancies in patients with IPMNs during follow-up, comparing it with that of the general Japanese population.
Patients and methods
Eight hundred and three consecutive patients diagnosed with IPMNs at the Department of Gastroenterology of Tokyo University Hospital from January 1995 to December 2008 were included in this study. Patients with IPMNs underwent physical examination, medical history evaluation and blood tests that included tumour markers, glycaemia and amylase levels. We also performed imaging modalities including abdominal ultrasound (US), endoscopic ultrasound (EUS), contrast-enhanced CT, or magnetic resonance cholangiopancreatography (MRCP). Endoscopic retrograde cholangiopancreatography (ERCP) and EUS fine-needle aspiration (EUS-FNA) were carried out if necessary. Main-duct IPMNs and branch-duct IPMNs were radiologically defined according to the consensus guidelines.5 Diagnosis of branch-duct IPMNs was made by imaging criteria based on the presence of pancreatic cystic lesions obviously communicating with the main pancreatic duct.21 Pancreatic cystic lesions without communication with the main pancreatic duct were not included in this study. Once diagnosed as IPMNs, surgical resection was proposed when the following features were observed before 2006: presence of mural nodules, main pancreatic duct dilation of >7 mm, diameter of IPMNs >40 mm or positive cytology. After the consensus criteria in 2006, we recommended surgical resection according to those criteria.5 Patients who had suffered from pancreatitis were also recommended surgery. At the time of diagnosis, 45 patients underwent surgical resections of IPMNs according to the criteria described above, while the remaining patients were non-operatively managed. Patients with contraindication to anaesthesia (n=15) or who refused surgery (n=95) did not undergo surgical resection and were also included in this study. Follow-up of <1 year led to the exclusion of 161 others. The remaining 642 patients were followed up prospectively.
Once included, the patients' follow-up consisted of twice-yearly clinical examinations, laboratory tests including fasting glucose levels, amylase levels and tumour markers, plus one of the imaging examinations. Imaging examinations consisted of US, CT, MRCP or EUS every 6 months. During the observation period, we performed ERCP or EUS-FNA to obtain cytological confirmation when any sign suspicious of pancreatic cancer development was detected while undergoing routine clinical examinations.
When patients without pancreatic cancer showed any signs suggestive of malignancies, such as weight loss, anaemia or tumour marker elevation without pancreatic cancer, they underwent additional examinations such as colonoscopy, gastroduodenoscopy, chest x-ray or positron emission tomography according to the suspected malignancy. In patients without any symptoms or signs of extrapancreatic malignancies, we recommend a screening test once a year during follow-up, and especially for cancers commonly found in the Japanese. As a rule, pathological confirmation was obtained for the diagnosis of extrapancreatic malignancies.
To estimate the actual incidence of the respective extrapancreatic malignancies, we analysed the patients who were free of them before and at the time of the diagnosis of IPMNs. The expected number of extrapancreatic malignancies was determined using age-stratified and sex-specific data on the incidence of cancer in Japan, provided by the Center for Cancer Control and Information Services, National Cancer Center, Japan.22 The standardised incidence ratio (SIR) was calculated as the ratio of the observed to the expected number of patients developing extrapancreatic malignancies. The 95% CI of the SIR was estimated assuming a Poisson distribution.
All statistical analyses were performed with JMP 7.0.1 software (SAS Institute). The cumulative incidence of developing extrapancreatic malignancies was analysed by the Kaplan–Meier method. Unconditional logistic regression analysis was used to estimate the HR for developing extrapancreatic malignancies. All of the statistical tests were two-sided, and a p value of <0.05 was considered statistically significant.
Results
Patient characteristics
Six hundred and forty-two patients with IPMNs (361 men and 281 women, mean age 68 years at the time of diagnosis), were prospectively followed for >1 year. The median follow-up period was 50 months (range 12–189). Clinical characteristics are summarised in table 1. In 618 asymptomatic patients, 135 were diagnosed by medical checkups, 409 by chance during follow-up of non-pancreatic diseases and 68 during assessment of ambiguous abdominal discomfort. The median size of IPMNs was 16 mm (range 2–80), the median size of the main pancreatic duct was 2 mm (range 1–25) and a mural nodule was observed in 54 (8%) patients. One hundred and forty-three patients met the consensus criteria for resection.
Extrapancreatic malignancies at the time of IPMNs diagnosis
One hundred and sixty-seven (26%) patients had extrapancreatic malignancies before follow-up—101 (16%) patients before diagnosis and 66 (10%) patients at the time of diagnosis. The primary sites of extrapancreatic malignancies at the time of IPMNs diagnosis are summarised in table 2. The most common sites of extrapancreatic malignancies before diagnosis were colorectal 25, gastric 23, hepatocellular 13, breast 10, uterine 9 and prostate 9, whereas the most common sites of synchronously diagnosed extrapancreatic malignancies were colorectal 16, hepatocellular 14, biliary tract 7, gastric 7 and prostate 6.
Development of pancreatic cancer during follow-up
Pancreatic cancer was discovered in 17 patients (one of 28 patients with main-duct IPMN and 16 of 614 patients with branch-duct IPMN) during the same follow-up periods with an annual incidence of 0.6%. The annual incidence in patients who met consensus criteria for resection was 0.8% and that in patients who had main-duct IPMN was 1.5%. Eight patients were managed operatively (ductal carcinoma in six and invasive intraductal papillary mucinous carcinoma in two), and the other nine were unresectable. In patients with unresectable pancreatic cancer, in eight of them surgery was contraindicated due to co-morbidity. The SIR of pancreatic cancer was 10.7 (95% CI 6.2 to 17.1).
Development of extrapancreatic malignancies during follow-up
Forty extrapancreatic malignancies developed in 39 patients during follow-up. They were diagnosed between 8 and 124 months (48 on average) after the diagnosis of IPMNs. One patient developed both gastric cancer 8 months and colorectal cancer 27 months after the diagnosis metachronously. The observed number and SIR of each of the extrapancreatic malignancies are listed in table 3. The most common sites were hepatocellular (n=7), colorectal (n=6), gastric (n=6), lung (n=5) and prostate (n=4). Their calculated SIR and 95% CI were as follows: hepatocellular 2.17 (0.87 to 4.47), colorectal 1.02 (0.37 to 2.21), gastric 0.76 (0.28 to 1.66), lung 0.75 (0.24 to 1.76) and prostate 1.00 (0.71 to 1.29). All of the 95% CIs for SIR included unity.
The incidence of all extrapancreatic malignancies was 1.3% (95% CI 1.0% to 1.8%) per year, and the SIR was 0.94 (95% CI 0.67 to 1.29). The 5 and 10 year rates of developing all extrapancreatic malignancies analysed by the Kaplan–Meier method were 5.7% and 12.5% (figure 1). Subgroup analyses showed no significant differences in sex, body mass index, diabetes, smoking history and size of IPMNs between patients developing extrapancreatic malignancies and the other patients, except in age, by Cox proportional hazard model analysis (table 4).
Discussion
This is the first report of a prospective estimation of the actual incidence of extrapancreatic malignancies in patients with IPMNs by long-term follow-up. In this study, we followed up 642 patients with IPMNs for an average period of 4.8 years and observed 40 extrapancreatic malignancies in 39 patients, with an annual incidence rate of 1.3%. All of the calculated CIs for the SIR of each extrapancreatic malignancy included unity, indicating that the incidence of each extrapancreatic malignancy in patients with IPMNs was similar to that in age- and gender-matched Japanese controls.
There have been several investigations that examined the frequency of extrapancreatic malignancies in patients with IPMNs. They reported that patients with surgically resected IPMNs were at higher risk of extrapancreatic malignancies than patients with ductal carcinoma of the pancreas or other pancreatic cystic neoplasms, or general referrals.14–16 19 The observed frequency of preoperative extrapancreatic malignancies was 10–38%. These were case–control studies that estimated the prevalence, and therefore the actual incidence of extrapancreatic malignancies in patients with IPMNs was not known. Eguchi et al reported that the calculated observed/expected ratio of preoperative extrapancreatic malignancies using the correlative data file of the cancer database was 2.41-fold for patients with IPMNs, which was statistically significant.17 However, there may be a selection bias that leads to overestimation of the frequency of extrapancreatic malignancies in patients with IPMNs. Recently, Reid-Lombardo et al examined the frequency of both benign and malignant extrapancreatic neoplasms in patients with both operatively and non-operatively managed IPMNs.20 Although they concluded that patients with IPMNs have increased risk of harbouring extrapancreatic neoplasms compared with patients with ductal carcinoma of the pancreas or general referrals, their study was also retrospective, not estimating the incidence of extrapancreatic malignancies after the diagnosis. In the present study, the prevalence of extrapancreatic malignancies in patients with IPMNs was 25.9%, which was consistent with previous reports. Therefore, the clinical characteristics of patients in our study were similar to those in other studies.
Some reports mentioned the postoperative extrapancreatic malignancies. Eguchi et al found 10 postoperative extrapancreatic cancers in 69 patients with operated IPMNs, a number too small to allow estimation of the incidence.17 Rial et al reported 282 additional cancers after the diagnosis in 19 647 patients with pancreatic cancer including IPMNs.19 IPMNs accounted for only 5% of all pancreatic cancers in their report, meaning that the incidence of extrapancreatic malignancies in patients with IPMNs could not be evaluated. Sugiyama et al reported that the incidence of postoperative extrapancreatic malignancies in patients with IPMNs was higher than that in patients with ductal carcinoma of the pancreas,16 but this was because the prognosis of patients with pancreatic ductal carcinoma was considerably poorer than that of patients with IPMNs, and the follow-up periods in patients with IPMNs and with pancreatic ductal carcinoma were different.
The pattern of extrapancreatic malignancies at the diagnosis of IPMNs correlated with the most prevalent cancers in Japan (colorectal, liver, gastric, breast and prostate cancers), which was identical to the other studies except for frequency.16 17 The most and second most frequently observed sites of extrapancreatic malignancies were gastric and colorectal in Korea,14 breast and prostate in France,15 and colorectal and breast in the USA.19 20 The repartition of observed malignancies correlated with the cancers found commonly in each country. Some investigations reported that patients with IPMNs were at higher risk of colorectal cancers.17 19 20 However, this is probably because the prevalence of colorectal cancers is high and increasing worldwide, especially in Asia.23 Although we did not conduct screening colonoscopy for all of the patients in this study, the incidence of colorectal cancers in patients with IPMNs was not elevated compared with that in the general Japanese population. Specific cancers could develop in genetically susceptible patients, but there was no tendency to develop specific cancers in patients with IPMNs, indicating that IPMNs were not genetically associated with specific carcinogenesis.
The common primary site of observed extrapancreatic malignancies during follow-up was slightly different from that of extrapancreatic malignancies at the diagnosis of IPMNs. Lung cancers were observed more frequently during follow-up than at diagnosis, a pattern similar to that of previous reports, probably because the peak incidence of lung cancers was at an older age than that of IPMNs.17 19 Hepatocellular carcinoma was observed more frequently than reported in other studies probably because 20% of our patients with IPMNs were infected with hepatitis C virus, a significantly higher rate than in the general Japanese population. In fact, all the patients developing hepatocellular carcinoma were infected with hepatitis C virus.24
We analysed the clinical features of patients with IPMNs who developed extrapancreatic malignancies. In this study, older patients were at higher risk for extrapancreatic malignancies. There was no significant difference in frequency of extrapancreatic malignancies according to the size of IPMNs. Although smoking, obesity and diabetes mellitus are well known risk factors for cancers,25–27 we did not find any significant predictor of developing extrapancreatic malignancies. This can probably be attributed to the fact that the number of patients in our study was small.
In this study, although we recommended a screening programme for extrapancreatic malignancies, surveillance for extrapancreatic malignancies was not provided for all of the patients. Therefore, we cannot rule out the possibility that we underestimated the incidence of extrapancreatic malignancies. However, in the follow-up, they underwent clinical examinations, laboratory tests and imaging tests at least twice a year, so advanced cancers could not have been overlooked.
In summary, the incidence of extrapancreatic malignancies in patients with IPMNs was similar to that in the general Japanese population, in contrast to the high incidence of pancreatic cancers. IPMNs are associated with pancreatic cancers but not with systemic carcinogenesis.
References
Footnotes
Competing interests None.
Patient consent Obtained.
Provenance and peer review Not commissioned; externally peer reviewed.