Diagnosing Early Pancreatic Cancer

Abstract & Commentary

By William B. Ershler, MD, Editor

Synopsis: Although pancreatic cancer growth is considered rapid, early recognition of resectable disease remains the best chance for long-term survival. It is possible that an early sign of evolving pancreatic neoplasm is glucose intolerance. In a series of 30 pancreatic cancer patients evaluated at the Mayo Clinic, CT scans obtained 6 months or more before the diagnosis revealed potentially resectable lesions in some, and this was notably true for those who had CT scans and new-onset diabetes several months before the diagnosis of pancreatic cancer. Thus, physicians evaluating adults with newly diagnosed diabetes should consider the possibility that the glucose intolerance is an accompaniment of early pancreatic neoplasia.

Source: Pelaez-Luna M, et al. Resectability of presymptomatic pancreatic cancer and its relationship to onset of diabetes: a retrospective review of CT scans and fasting glucose values prior to diagnosis. Am J Gastroenterol. 2007;102:2157-2163.

Pancreatic cancer remains both a challenge to diagnose and an even greater challenge to effectively treat. In fact, only patients discovered early and with resectable disease have a chance for long-term survival. Unfortunately, the majority of patients (greater than 85%) have unresectable disease by the time disease-associated symptoms occur and a diagnosis is made.1 Patients who have the greatest chance for curative resection are those who have their tumors diagnosed when under evaluation for other problems and the pancreatic mass is discovered before symptoms occur. The timeline for progression of pancreatic cancer from resectable to unresectable is unknown. Evidence for glucose intolerance is known to occur in a substantial percentage of pancreatic cancer patients and it may occur earlier than other signs or symptoms of disease. In an effort to determine whether the investigation of new-onset diabetes for pancreatic cancer or the serendipitous discovery of a CT detected pancreatic mass offers a sufficiently early diagnosis to improve cure rates, Pelaez-Luna and colleagues at the Mayo Clinic reviewed 30 patients with pancreatic cancer who either had abdominal CT scans performed months or years prior to the diagnosis of pancreatic cancer and/or had developed diabetes prior to or concurrent with the diagnosis of pancreatic cancer.

All CT scans including those done at the time of diagnosis and those performed at earlier dates were reviewed and classified as either normal, potentially resectable, or unresectable pancreatic cancer. Fasting blood glucose levels were obtained at the time of diagnosis and also prior to diagnosis in 18 cases of the 30 patients.

Of the 30 patients, 28 had a total of 38 CT scans done at a median of 18 months (range 1 to 41 months) before the cancer diagnosis. At cancer diagnosis, only 7 of 30 patients could undergo margin negative surgical resection. CT scans done at six months before diagnosis or earlier revealed either a normal pancreas (n = 20) or a resectable mass (n = 6). None of the CT scans that were obtained earlier than six months before diagnosis revealed an unresectable mass.

With regard to diabetes, the mean interval between the onset of laboratory confirmed glucose intolerance and pancreatic cancer was 10 months (range 5 to 29 months). At the time of diagnosis of diabetes, 13 patients had CT scans. Of these, three had a normal appearing pancreas, six had a mass that was, even at that time, considered resectable, and four had a mass that was, even at that time considered unresectable.

Thus, the authors conclude that undetectable or resectable pancreatic cancer was apparent on CT scans obtained greater than six months prior to clinical diagnosis. At the onset of diabetes, pancreatic cancers were, in this series, generally resectable.


This series highlights the frustrating aspect of early diagnosis in pancreatic cancer. Unlike colon cancer, for example, where the development of a neoplastic lesion occurs over years and for which surveillance initiatives have demonstrated the capacity for early recognition allowing curative resection, such has not been the case for pancreatic cancer. This, no doubt results from what must be a rapid transition from resectable to unresectable disease and the lack of an effective and feasible screening device. Most of the patients in this series had become diabetic, and for those who had CT scans obtained at or near the time of the newly diagnosed diabetes, at least half were shown to have a pancreatic mass that was considered resectable. In fact, patients with or without diabetes who had CT scans obtained six months or more prior to the onset of pancreatic cancer for unrelated conditions were likely to have resectable tumors as well. The fact that the masses discovered by retrospective analysis were not actively pursued in a timely fashion highlight the uncertainty of pancreatic imaging (by CT scans utilized during the years of this analysis) and the rapidity with which pancreatic cancers grow. Thus, although some might consider CT scanning for pancreatic cancer screening, the applicability of this expensive approach would likely be seriously hampered by measures of both sensitivity and specificity.

However, if a high-risk category were identified, CT scanning might prove reasonable. One such category of high-risk individuals would be those with newly-discovered diabetes. In a population-based study performed by this same group of Mayo Clinic investigators, those with new onset diabetes were shown to have eight times the likelihood of being diagnosed with pancreatic cancer within three years than the general population.2 Thus, it would seem reasonable to investigate the role of CT scan screening for patients with new onset diabetes. However, prior studies that have addressed this question (screening for pancreatic cancer in those with new onset diabetes and cancer-related symptoms) have identified mostly unresectable pancreatic cancer.2,3 However, those studies relied on the presence of cancer-related symptoms in newly diagnosed diabetics, and thus, the lack of discovering early (or small) pancreatic lesions is not surprising. A similar analysis of asymptomatic patients with newly discovered diabetes might be more successful.

As imaging studies become more precise and less intrusive, prospective studies may result in an improved understanding of the relationship between diabetes development, pancreatic mass, and pancreatic cancer. More importantly, these studies may identify individuals who would benefit from screening and heightened surveillance such that pancreatic cancers could be discovered in a resectable stage.


1. Jemal A, et al. Cancer statistics, 2006. CA Cancer J Clin. 2006;56(2):106-130.

2. Chari ST, et al. Probability of pancreatic cancer following diabetes: a population-based study. Gastroenterology. 2005;129(2):504-511.

3. Damiano J, et al. Should pancreas imaging be recommended in patients over 50 years when diabetes is discovered because of acute symptoms? Diabetes Metab. 2004;30(2):203-207.