Use of endoscopic ultrasonography in screening patients at high-risk for pancreatic cancer

Reviewer: Christine Hill, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: February 7, 2009

Presenter: J. Poley, University Medical Center, Rotterdam, Netherlands


  • Screening of the general population for certain malignancies, including breast cancer, prostate cancer, colon cancer, and cervix cancer has become standard of care within the United States and in many countries abroad.
  • The feasibility and cost-effectiveness of screening tests for cancer depends on several factors, including the disease incidence, the test sensitivity, level of invasiveness, and cost.
  • Pancreatic cancer continues to be associated with extremely high rates of mortality, with long-term survival rates remaining dismal; however, disease incidence remains relatively low.
    • The poor prognosis associated with pancreatic cancer is felt to be due in part to lack of symptomatology during early disease stages, often resulting in the diagnosis being made only once disease has progressed locally or metastasized.
    • Endoscopic ultrasound has been demonstrated to be highly sensitive for tumors of the pancreas (Helmstaedter, Langenbecks Arch Surg, 2008); however, the test is invasive and expensive, and not appropriate for screening of the general population.
  • Approximately 10-15% of pancreatic adenocarcinomas are familial or hereditary in origin. Known syndromes associated with development of pancreatic cancer include familial breast cancer syndromes, particularly BRCA 2 mutations, Peutz-Jeghers syndrome, and familial atypical multiple mole melanoma (FAMMM) syndrome.
  • Early detection of pancreatic cancers is expected to improve prognosis associated with the disease; this study was undertaken to evaluate the role of endoscopic ultrasonography for screening of high-risk populations for pancreatic cancer.


  • Patients considered to be at-risk for familial or hereditary pancreatic cancer were included in this study. Eligibility requirements included the following:
    • Individuals with one or more first-degree relatives with diagnosis of pancreatic cancer.
    • Individuals known to carry CDKN2A or PRSS1 mutations, signifying known pancreatic cancer-prone hereditary syndromes.
    • Individuals with Peutz-Jeghers’ syndrome or FAMMM.
    • Those with familial clustering of pancreatic cancer (in two or more non-first-degree relatives) and known mutations in BRCA1 or 2, APC, p53, or MMR.
    • All patients were asymptomatic and had never previously undergone endoscopic ultrasound.
  • All patients underwent endoscopic ultrasound on one occasion at one of two academic medical centers within the Netherlands.


  • 43 patients were identified as eligible for this study.
    • 18 patients were male and 25 female.
    • Age range 32 – 75 years, with median age 51 years.
  • The patients’ genetic background was diverse within the eligibility requirements:
    • 13 patients were from families with FAMMM, 8 of whom were carriers of the CDKN2A mutation.
    • 21 had first-degree relatives with diagnosis of pancreatic cancer.
    • 3 had proven mutations in the PRSS1 gene.
    • 2 had Peutz-Jeghers’ syndrome.
    • 6 had familial clustering of pancreatic cancer in non-first degree relatives, 3 of whom carried BRCA1 mutations, 2 BRCA2 mutations, and 1 a p53 mutation.
  • Asymptomatic mass lesions were identified in the pancreatic body or tail in three screened patients.
    • 2 had proven FAMMM syndrome, and 1 had familial pancreatic cancer clustering and a BRCA1 mutation.
    • Lesions were 12 mm, 27 mm, and 50 mm.
    • The larger two lesions were subsequently visible on CT and MRI of the abdomen; the smallest lesion could not be visualized on these studies.
  • All three lesions were completely removed surgically. Pathology demonstrated moderately differentiated adenocarcinoma with negative surgical margins in all three cases.
    • The first patient (FAMMM, 12 mm lesion) had pathologic N0 status, with 10 lymph nodes negative for cancer.
    • The second (BRCA1, 27 mm lesion) had pathologic N1 status, with metastatic carcinoma identified in 5 of 11 lymph nodes.
    • The third (FAMMM, 50 mm lesion) had pathologic N1 status, with metastatic carcinoma identified in 1 of 9 lymph nodes.
  • Branch-type intraductal papillary mucinous neoplasias were identified in 7 patients:
    • 3 had FAMMM syndrome.
    • 3 had familial pancreatic cancer (diagnosed in at least one first-degree relative).
    • 1 had known BRCA1 mutation.
  • No complications associated with endoscopic ultrasound occurred.

Authors’ conclusions

  • The authors conclude that screening patients at high-risk for pancreatic cancer with endoscopic ultrasound is feasible and safe.
  • They describe the incidence of clinically relevant findings to be high, with 7% asymptomatic carriers and 16% premalignant lesions identified.
  • They note that effect of screening on survival remains to be investigated.


  • Pancreatic cancer remains one of the most feared oncologic diagnoses, and is associated with extremely high risk of mortality.
  • Patients with surgically resectable disease are recognized to fare better than those with locally advanced, unresectable tumors, and certainly better than those with metastatic disease.
  • Earlier diagnosis of pancreatic cancer would be expected to result in improved outcomes. The relatively low incidence of pancreatic cancer makes screening of the general population infeasible; however, the data presented here are very interesting with regard to screening a particularly high-risk population.
  • As the authors point out, endoscopic ultrasound performed in this study resulted in identification of three resectable pancreatic carcinomas; prognosis for these patients is certainly expected to be improved as a result of this early detection.
  • The authors note, however, that the true impact of screening on overall mortality has yet to be determined. Long-term studies in the setting of randomized clinical trials might answer the question of impact on mortality more definitively; however, such studies might be quite difficult to carry out. Once patients are identified as being at high-risk for development of pancreatic cancer, randomizing them to “no screening” would be ethically questionable. The question of true, quantitative, reduction in mortality with screening of high-risk patients may never be answered in a randomized fashion; however, long-term observation of these populations will certainly provide further information.
  • The authors do not comment on cost of the screening tests carried out. As healthcare resources continue to shrink, attention to cost-effectiveness of any screening test remains important.
  • Having said this, the data presented here are interesting and important, and certainly supports possible implementation of screening of high-risk patients for pancreatic cancers.