Clinical presentation and waiting time targets do not affect prognosis in patients with pancreatic cancer

Surgeon. 2010 Oct;8(5):239-46. doi: 10.1016/j.surge.2010.03.001. Epub 2010 Apr 2.


Introduction: The prognosis of patients with pancreatic cancer remains poor despite recent advances in treatment. It is not known whether delays in referring, diagnosing and treating these patients and the way they present can affect their survival.

Aims: In our study we investigated the impact of clinical presentation (jaundice, abdominal pain, weight loss) and delays in management of these patients on their treatment and survival.

Methods: Data on all patients with pancreatic cancer referred to the Pancreatic Unit (1997-2002) were collected prospectively and analysed using SPSS 16((R).) The delay in diagnosis and treatment for each patient was measured by estimating the time from the beginning of symptoms to the date of the referral letter (T1), the time from the referral date to the date of first review at the Unit (T2) and the time from date of review to the date of diagnosis/treatment (T3). Treatments were defined as 1) pancreatic resections, 2) gastric and biliary bypass and 3) biliary stents. The term 'operability' was used to describe patients thought to have a potentially removable tumour and had an operation and 'resectability' applied to the patients whose tumour was actually removed at the operation. Follow-up time and survival were recorded by reviewing the patient's notes, hospital electronic databases and by contacting patients General Practitioners.

Results: There were a total of 355 patients with pancreatic cancer. Median age at diagnosis was 64 (i.q.r. 56-71) years and median follow-up was 8 (i.q.r. 4-14) months. The overall 1, 3 and 5 years patient's survival was 26%, 5% and 4% respectively. 1, 3 and 5 years survival of inoperable patients was 24%, 2% and 0% and for operable patients was 35%, 13% and 9% respectively. The median survival time for those patients that underwent operation was significantly higher than those that did not (12 vs 6 months, p < 0.001). The overall median time from initial symptoms to diagnosis/treatment (T1 + T2 + T3) was 102 (i.q.r. 56-182) days, T1 was 65 (i.q.r. 31-143), T2 17 (i.q.r. 8-28) and T3 11 (i.q.r. 6-21) days. The time delay from symptoms to referral (T1) had minimal clinical relevance to survival, with a hazard ratio of only 1.001 (95% CI 0.001-0.002, p = 0.043) per day. Of all 355 patients, 305 (86%) were reviewed and treated within 62 days from the GP referral (T2 + T3). There was no significant difference in operability, resectability and survival of patients that were diagnosed/treated before or after 62 days from referral (T2 + T3) (median months 6.5 and 7.9 respectively, p = 0.134). Patients presenting with jaundice were referred (T1, median 56 vs 103) and diagnosed/treated (T2 + T3, median 96 vs 130) days (p < 0.001) sooner, had a higher operability rate (33% vs 21%, p = 0.035) but not a significantly higher resectability rate of (37% vs 29%) (p = 0.608). Isolated or combined mode of clinical presentation had no significant effect on survival (p = 0.965). On multivariate regression analysis, prognostic factors of survival were a resectable tumour and the time from the beginning of symptoms to referral.

Conclusion: This study showed that pre-hospital delays in referring patients to a specialist unit, but not hospital related 62 days target, had an no impact on operability, resectability and survival. Clinical presentation also had no impact on the survival. We confirmed that pancreatic resection is the most important factor in determining the length of survival in patients with pancreatic cancer. Our study implies that the successful implementation of the 62 days National Cancer Waits Target across the UK is unlikely to have an impact on prognosis in patients with pancreatic cancer. Focusing on early referral to specialist Pancreatic Units might be more effective.

MeSH terms

  • Abdominal Pain / etiology
  • Aged
  • Carcinoma, Pancreatic Ductal / diagnosis
  • Carcinoma, Pancreatic Ductal / mortality*
  • Female
  • Humans
  • London / epidemiology
  • Male
  • Middle Aged
  • Outcome Assessment, Health Care*
  • Pancreatic Neoplasms / diagnosis
  • Pancreatic Neoplasms / mortality*
  • Pancreatic Neoplasms / surgery
  • Prognosis
  • Referral and Consultation / statistics & numerical data
  • Survival Analysis
  • Waiting Lists*
  • Weight Loss