Article Text

Download PDFPDF

Pancreatic cancer risk in Peutz-Jeghers syndrome patients: a large cohort study and implications for surveillance
  1. Susanne E Korsse1,
  2. Femme Harinck1,
  3. Margot G F van Lier1,
  4. Katharina Biermann2,
  5. G Johan A Offerhaus3,4,
  6. Nanda Krak5,
  7. Caspar W N Looman6,
  8. Wendy van Veelen1,
  9. Ernst J Kuipers1,7,
  10. Anja Wagner8,
  11. Evelien Dekker9,
  12. Elisabeth M H Mathus-Vliegen9,
  13. Paul Fockens9,
  14. Monique E van Leerdam1,
  15. Marco J Bruno1
  1. 1Department of Gastroenterology & Hepatology, Erasmus MC University Medical Centre, Rotterdam, The Netherlands
  2. 2Department of Pathology, Erasmus MC University Medical Centre, Rotterdam, The Netherlands
  3. 3Department of Pathology, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands
  4. 4Department of Pathology, University Medical Centre Utrecht, Utrecht, The Netherlands
  5. 5Department of Radiology, Erasmus MC University Medical Centre, Rotterdam, The Netherlands
  6. 6Department of Public Health, Erasmus MC University Medical Centre, Rotterdam, The Netherlands
  7. 7Department of Internal Medicine, Erasmus MC University Medical Centre, Rotterdam, The Netherlands
  8. 8Department of Clinical Genetics, Erasmus MC University Medical Centre, Rotterdam, The Netherlands
  9. 9Department of Gastroenterology & Hepatology, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands
  1. Correspondence to Susanne E Korsse, Department of Gastroenterology & Hepatology, Erasmus MC University Medical Centre, ‘s Gravendijkwal 230, Room Hs-306, 3015 CE Rotterdam, The Netherlands; s.korsse{at}erasmusmc.nl

Abstract

Background Although Peutz-Jeghers syndrome (PJS) is known to be associated with pancreatic cancer (PC), estimates of this risk differ widely. This hampers counselling of patients and implementation of surveillance strategies. We therefore aimed to determine the PC risk in a large cohort of Dutch PJS patients.

Methods PJS was defined by diagnostic criteria recommended by the WHO, a proven LKB1 mutation, or both. All patients with a presumptive diagnosis of pancreatic, ampullary or distal bile duct cancer were identified. Cases were reviewed clinically, radiologically and immunohistochemically. Cumulative PC risks were calculated by Kaplan-Meier analysis and relative risks by Poisson regression analysis.

Results We included 144 PJS patients (49% male) from 61 families (5640 person years follow-up). Seven (5%) patients developed PC at a median age of 54 years. Four patients (3%) were diagnosed with distal bile duct (n=2) or ampullary cancer (n=2) at a median age of 55 years. The cumulative risk for PC was 26% (95% CI 4% to 47%) at age 70 years and relative risk was 76 (95% CI 36 to 160; p<0.001). The cumulative risk for pancreatico-biliary cancer was 32% (95% CI 11% to 52%) at age 70 years, with a relative risk of 96 (95% CI 53 to 174; p<0.001).

Conclusions PJS patients have a highly increased risk for pancreatico-biliary cancer. Therefore, patients are eligible for surveillance within well defined research programmes to establish the benefit of such surveillance.

  • ampullary cancer
  • distal bile duct cancer
  • inherited cancer syndrome
  • Peutz-Jeghers syndrome
  • pancreactic cancer

Statistics from Altmetric.com

Introduction

Despite the relative low incidence of pancreatic cancer (PC) (8–10 per 100 000 per year, with an approximate 1% life time risk in western populations1), PC is among the top five causes of cancer related deaths in both the USA and Europe.2 ,3 The mean survival after diagnosis is <6 months and the overall 5-year survival is <5%.4 This poor prognosis is mainly due to the late onset of symptoms and anatomic location of the disease. Consequently, <20% of all patients present with localised disease and are therefore eligible for curative treatment. Unfortunately, this intended curative treatment proves only to be effective for the minority of patients with an overall 5-year survival after surgical resection of <10%.5 Despite recent advantages in the field of surgery and oncology, this dismal prognosis has not significantly changed over the past decades.6

Detection of precursor lesions or malignancies at an early asymptomatic stage by surveillance with endoscopic ultrasound (EUS) and/or MRI could offer a way to improve the prognosis.7 In particular, when surveillance is directed towards populations of individuals that carry a high risk for developing PC, the potential health gain could be substantial.

One such high risk population consists of patients with Peutz-Jeghers syndrome (PJS). PJS is an autosomal dominant inherited disorder, caused by germline mutations in the LKB1 tumour suppressor gene (also known as STK11).8 It is characterised by gastrointestinal hamartomas and mucocutaneous pigmentations. Furthermore, patients with PJS are at risk for developing various types of cancer, including PC.9–11 The actual risk of developing PC for PJS patients is currently unclear. Previous studies reported relative risks ranging from 0–132-fold increase and an average age of PC onset ranging from 41–60 years of age.10–14 Consequently, this hampers counselling of PJS patients and implementation of surveillance strategies.

These disparate risk estimates were mainly derived from heterogeneous multicentre populations, small single centre cohort studies, and meta-analyses of these same studies. It is therefore key to perform such a study in a large homogenous population. In 2011, our research group reported on the high overall cancer risk in a unique, large pedigree based homogenous cohort of Dutch PJS patients with a substantial prospective period of follow-up.9 For the present study, we performed a thorough re-evaluation of all reported cancers in the pancreatico-biliary region in this patient cohort, including 2 years of extended follow-up. Thus, we aimed to conclude the ongoing debate regarding the true PC risk in PJS, and to provide a more scientific rationale for the implementation of surveillance strategies.

Methods

PJS database

This nationwide cohort study was initiated by two Dutch academic hospitals. Between 1995 and July 2011, PJS patients throughout the Netherlands were included without selection for medical history. All patients had a definite diagnosis of PJS, defined by diagnostic criteria recommended by the WHO (box 1), a proven LKB1 mutation, or both. Informed consent was obtained from all patients and the study was approved by the Institutional Review Board of both participating hospitals. Patients were followed prospectively between January 1995 and July 2011. Patient information at baseline and during follow-up was obtained by interview and chart review. Clinical data from the period before 1995 as well as data of deceased family members fulfilling the diagnostic criteria for PJS were collected retrospectively.

Box 1

Diagnostic criteria for Peutz-Jeghers syndrome (PJS) recommended by the WHO (2010)

  • Positive family history of PJS, and

    • Any number of histologically confirmed PJS polyps,* or

    • Characteristic, prominent, mucocutaneous pigmentations.

  • Negative family history of PJS, and

    • Three or more histologically confirmed PJS polyps, or

    • Any number of histologically confirmed PJS polyps and characteristic, prominent, mucocutaneous pigmentations.

*Histology of PJS polyps: a central core of smooth muscle that shows tree-like branching, covered with normal epithelium.

Case selection and data collection

PC cases were identified from the PJS database. PC was defined according to the most recent WHO classification of tumours of the digestive system.4 In addition, patients with a diagnosis of distal bile duct cancer or ampullary cancer were included. Surveillance of the pancreas might also detect these malignancies and accurate distinction between these three tumours often proves difficult. From all selected cases, medical records were reviewed by two MDs (SEK and FH). The following data were collected: gender, date of birth, cancer diagnosis and death, mutation status and type of mutation, family history of PJS, and family history of PC. The recorded cancer characteristics included tumour type and origin, tumour invasion, data on confirmation (medical record or histology), and presentation (surveillance, accidentally or symptomatic). Radiological images were reviewed by an expert abdominal radiologist (NK). Available formalin-fixed and paraffin-embedded tissue was reviewed by two expert pathologists independently (KB and GJAO). Immunohistochemical staining for SMAD4, CDX2 and cytokeratins was performed to ascertain the diagnosis.4 Eventually, all available information was re-assessed by an expert panel (KB, GJAO, NK, ED, EMHMV, MEvL and MJB) to determine a definite diagnosis of PC, distal bile duct cancer or ampullary cancer.

Statistical analysis

Data were analysed using the SPSS V.17.0 statistical software for Windows (IBM, Somers, New York, USA). All risks were calculated for two groups: (1) PC cases; (2) cases of cancer in the pancreatico-biliary region, including cases with PC, distal bile duct cancer or ampullary cancer. Cumulative risks were estimated as a function of time using the Kaplan-Meier method and the Cox regression model. For these cumulative risk analyses, all subjects of the cohort were included. For relative cancer risk calculation, the tumour specific cancer incidence observed in our study population was compared to the age specific and gender specific incidence rates of the Dutch general population from 1960 to 2011 by Poisson regression analysis (log linear analysis) using the package R.15 Subjects were studied with respect to their risk of developing cancer from birth until the date of death, date of last contact or the closing date of the study (1 July 2011). Sociodemographic data and incidence rates of the Dutch general population were derived from the Eindhoven Cancer Registry (1960–2009). These data are representative for the Netherlands. Incidence rates for 2009 were assumed to be representative for 2010 and 2011.

Results

Study population

In total, 144 PJS patients from 61 families were included in the cohort with a total of 5640 person-years of follow-up (including 1757 person-years of prospective follow-up). Forty-nine per cent were male (3050 person-years). At the closing date of the study six patients had been lost to follow-up (4%), and 48 (33%) had died at a median age of 46 years (IQR 32–58 years); the median age of the 90 patients still alive (63%) was 37 years (IQR 21–52 years). The baseline characteristics of the cohort are shown in table 1.

Table 1

Baseline characteristics of the Dutch PJS cohort

PC cases

The case selection process is shown in figure 1. During follow-up, seven (4.9%) PJS patients from seven families developed PC, six male and one female. None of the cases was detected within the framework of a PC screening/surveillance programme. Adenocarcinoma of the pancreas was found in six patients and acinar cell carcinoma in one. Six cases were confirmed by revision of histology, and no histological material was available for the seventh case. Median age at diagnosis was 54 years (IQR 37–62 years). Six patients presented with symptoms. In one patient a tumour mass was incidentally found during a laparotomy because of small bowel polyps. None of the patients could be treated curatively. Median survival of patients after diagnosis was 6 months (IQR 4–17 months). Mutation analysis for the LKB1 gene was performed in five patients, detecting a pathogenic mutation in four of them. For the other two cases, a pathogenic mutation in the LKB1 gene was detected in affected family members, but mutation analysis was not performed in the PC patients. Two of the patients had been treated curatively for another malignancy (colorectal cancer and liposarcoma) before the development of PC. Individual patient characteristics are shown in table 2A.

Table 2

Characteristics of Peutz-Jeghers syndrome patients with pancreatic cancer (A) and with distal bile duct cancer and ampullary cancer (B)

Ampullary cancer and distal bile duct cancer cases

In addition to the patients with PC, distal bile duct cancer was diagnosed in a male patient at the age of 57 and in a female patient at the age of 73 years. Both patients only underwent palliative treatment; survival after diagnosis was 3 and 8 months, respectively. Furthermore, ampullary cancer/cancer involving the ampulla was detected in two male patients at the age of 41 and 53 years. Both patients underwent a pylorus preserving pancreaticoduodenectomy as curative treatment. One patient died of metastasised disease 5 years after diagnosis; the other patient is still alive 11 years after diagnosis. Patient characteristics are shown in table 2B. The median age at diagnosis for the group of patients with pancreatic, distal bile duct or ampullary cancer (n=11) was 54 years (IQR 42–62).

Cumulative cancer risk

PC (n=7)

The Kaplan-Meier estimate for the cumulative PC risk was 2.4% (SE 1.7%, 95% CI −0.9% to 5.7%) at age 40; 3.9% (SE 2.2%, 95% CI −0.4% to 8.2%) at age 50; 11.1% (SE 5.3%, 95% CI 0.7% to 21.5%) at age 60; and 25.6% (SE 10.8%, 95% CI 4.4% to 46.8%) at age 70 (figure 2, left). There was no significant difference in risk between males and females (p=0.272).

Figure 2

Cumulative cancer risk in Peutz-Jeghers syndrome patients according to age.

Pancreatic, ampullary or distal bile duct cancer (n=11)

For pancreatic, distal bile duct or ampullary cancer the cumulative risk was 2.4% (SE 1.7%, 95% CI −0.9% to 5.7%) at age 40; 5.2% (SE 2.6%, 95% CI 0.1% to 10.3%) at age 50; 18.2% (SE 6.5%, 95% CI 5.5% to 30.9%) at age 60; and 31.6% (SE 10.6%, 95% CI 10.8% to 52.4%) at age 70 (figure 2, right). There was no significant difference in risk for these malignancies between males and females (p=0.248).

Relative cancer risk

From 1960, 131 patients contributed to 4430 person-years at risk (males 2259 person-years, females 2171 person-years). Poisson regression analysis showed that the relative risk for PC (HR 76.2, 95% CI 36.3 to 160.0) as well as for pancreatic, ampullary or distal bile duct cancer (HR 95.8, 95% CI 52.8 to 173.7) was significantly higher in PJS patients than in the general population (p<0.001).

Discussion

This nationwide, long term follow-up cohort study shows that patients with PJS have a highly increased PC risk. We found a cumulative cancer risk of >25% at the age of 70 years and a 76-fold increased risk compared to the general population. The cumulative risk for developing any type of malignancy in the pancreatico-biliary region, including pancreatic, distal bile duct, or ampullary cancer, is as high as 29% at the age of 70 years and the relative risk for these cancers is 96. These data emphasise the relevance and clinical potential of surveillance of the pancreas for PJS patients, provided a suitable and effective surveillance programme is available.

Estimates on the risk of PC in patients with PJS vary widely within the literature. Our data are most in line with those from Giardiello et al.11 In a meta-analysis in which 210 PJS patients from six American and European studies were included, six cases of PC were identified which amounted to a cumulative risk of 36% by the age of 64 years and a relative risk of 132. The mean age at PC onset was 40.8 years (SD 16.2). Because the source data were contributed by multiple centres worldwide, the authors were not able to give extensive information about the intricacies of case selection or confirmation of the cancer diagnosis, including revision of pathology specimens and potential confounding issues relating to the problematic distinction between pancreatic, distal bile duct, and ampullary cancer. Furthermore, incidence rates of the US population were used for relative risk calculation, while the study population consisted of both US and European (UK and Dutch) patients. This might have led to biased relative cancer risks, as differences in cancer risk between PJS and control populations could exist due to variations in geography, race, culture and diet.

The extremely elevated risk found by Giardiello et al has not been reproduced by more recent studies. An international collaborative study concerning 419 PJS patients from eight centres worldwide found a cumulative risk for developing PC of 11% by the age of 70.12 Relative risk was not reported. Another collaborative study found no PC in a total of 149 PJS cases.13 Interpretation of the results of these two series is difficult, since no information is provided on the average age of the cohort or follow-up period. Albeit speculative, the lower risk found in these series could be the result of the participants being too young or the lack of sufficient follow-up time. The same data on age and follow-up period is missing in most current nationwide or single centre studies, which are often limited by a small sample size.

To our knowledge this is the first study to investigate the PC risk within a large, nationwide PJS patient cohort with long term follow up. This cohort goes back to the original family described by Jan Peutz in 1921,16 and encompasses a substantial period of prospective follow-up time amounting to 5640 person-years including 1757 person-years of prospective follow-up. Because it is well known that differentiation between pancreatic, distal bile duct and ampullary cancer is a diagnostic challenge, we attempted to address the issue of case selection by careful expert revision of clinical, radiological, and histological materials. This enabled us to provide reliable risk estimates for PC alone and for cancers of the pancreas and pancreatico-biliary region including distal bile duct and ampullary cancer. The latter are sometimes misclassified or impossible to differentiate from PC. As such, these numbers could be looked upon as absolute minimum and maximum risk estimates. Another important clinical consideration when making such a separate risk assessment is that distal bile duct and ampullary cancer, just like PC, have a potential for early detection in surveillance programmes of the pancreas.

A few limitations of our study warrant consideration. Firstly, because this PJS patient cohort was initiated by two tertiary referral centres, selection bias could potentially have led to an overestimated incidence of PC. Secondly, we were unable to gather reliable information about the smoking behaviour of our patients. This is unfortunate because smoking is one of the most important risk factors for the development of PC17 ,18 and therefore a probable confounding factor between different PJS populations.

The evidence is slowly accumulating that surveillance of high risk individuals leads to the detection of high grade precursor lesions and asymptomatic early stage PC.19–27 However, we currently still lack definite evidence that surveillance has a net benefit in terms of mortality reduction of PC related mortality and gain in life years, and whether this benefit outweighs the potential negative side effects of overtreatment, including associated complications and costs. We and others therefore suggest that surveillance of PJS patients should only be performed within the framework of well established research protocols.28–30 Results of the international Cancer of the Pancreas screening summit meeting in 2011 indicate that surveillance in high risk individuals should be regarded as a promising development, though more evidence is needed to address its real value.30 During this meeting, 49 experts in the field of PC voted on statements with respect to PC surveillance. This resulted in a number of outstanding questions that still need to be addressed, including questions with respect to who to screen, when to start screening, the optimal frequency of screening, and particularly the optimal management of the asymptomatic pancreatic lesions detected.

Based on the results of our current study, we recommend that PJS patients should be offered surveillance regardless of family history for PC, since all subjects with PC in our series had a negative family history of PC. Although the median age of PC onset in our cohort was 54 years, we propose that surveillance starts at the age of 30 years. This suggestion is based on the fact that two patients in our series developed cancer in the pancreatico-biliary region at a very young age. If screening had started 10 years earlier than the median age of PC onset, these cases would have been missed.

It has been noted that some patients with PJS develop intestinal-type intraductal papillary mucinous neoplasms (IPMNs).31 IPMNs are well defined premalignant lesions of PC. One pancreatic adenocarcinoma in our study showed histological indication for development out of an IPMN lesion. Future research should be directed towards unravelling the molecular pathway of PC development in PJS patients. Such knowledge may tailor surveillance recommendations even more. Furthermore, the efficacy and cost effectiveness of PC surveillance must be further studied.

In conclusion, absolute and relative risks of developing pancreatic, distal bile duct and ampullary cancer are very high in patients with PJS. This observation, and the prospect that detection of these malignancies, or preferably their precursor lesions, might be possible at an early and potentially curable point in time, render PJS patients eligible for surveillance by yearly EUS and/or MRI within well defined research protocols.

Acknowledgments

We would like to thank the Dutch Eindhoven Cancer Registry maintained by the Comprehensive Cancer Centre South for kindly providing us the sociodemographic data and incidence rates of the Dutch general population.

References

Footnotes

  • SEK and FH contributed equally

  • Contributors MJB and MEvL supervised the study. SEK and FH were responsible for acquisition of the data, the analysis and interpretation of the data and were responsible for drafting of the manuscript. KB, GJAO, NK, ED, EMHMV, MEvL and MJB formed the expert panel and were responsible for the determination of the definite diagnosis. Furthermore, they helped with the critical revision of the manuscript for important intellectual content. CWNL performed the statistical analyses and helped with the critical revision of the manuscript for important intellectual content. MGFvL, WvV, AW, EJK and PF helped with the critical revision of the manuscript for important intellectual content. All authors have approved the final draft submitted.

  • Funding None.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Institutional Review Board Erasmus MC University Medical Centre Rotterdam & Academic Medical Centre Amsterdam, the Netherlands.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.