my personal edition > pancreatic cancer > news


  E-Mail this DGDispatch to a colleague

DGDispatch

Surgery Can Be Effective Monotherapy for Pancreatic Metastases: Presented at SSO

By Mary Beth Nierengarten

CHICAGO -- March 17, 2008 -- Pancreatic metastectomy (PM) is safe and effective, particularly in patients with primary renal cell carcinoma, reports a study presented here at the Society of Surgical Oncology (SSO) 61st Annual Cancer Symposium.

Billed as the largest single-institution experience to date on the safety and efficacy of PM, the study was conducted by investigators at Johns Hopkins Hospital to assess the safety and efficacy of PM given the dearth of data currently available.

Although surgical removal of isolated sites of metastases has become more common in the last 10 to 15 years and is becoming standard of care, few patients with pancreatic metastases are being treated this way, according to lead author Sushanth Reddy, MD, Resident, Department of Surgery, Johns Hopkins Hospital, Baltimore, Maryland.

"[Pancreatic resection] is not undertaken lightly because it is associated with high morbidity," Dr. Reddy said in a presentation on March 14.

In the retrospective study, Dr. Reddy and colleagues collected demographic, clinical, and pathological data on all patients who underwent PM at Johns Hopkins Hospital. Data were retrospectively analysed from a pancreaticobiliary database compiled at Johns Hopkins from 1970 to 2006.

Of the 3,830 patients who underwent a pancreatic resection during that time, 60 (1.6%) had resections of the pancreatic or periampullary lesion from nonpancreatic primary tumours. Of these, 11 patients had lymphoma and were excluded from the analysis, leaving 49 evaluable patients.

Pathology analysis showed that sites of primary tumour included renal cell carcinoma (RCC, n = 21), gallbladder cancer (n = 6), ovarian (n = 4), lung (n = 4), sarcoma (n = 4), melanoma (n = 3), colon (n = 2), breast (n = 1), hepatocellular carcinoma (n = 1), seminoma (n = 1), and Langerhans cell histiocytosis (n = 1).

Most patients underwent pancreaticoduodenectomies (63%), followed by distal pancreatectomies (29%), and total pancreatectomies (8%).

At a median follow-up of 31 months, the overall median survival for all cancer types was 35 months.

Predictors of better outcome on univariate analysis were patients with RCC, women, and patients who did not have complications.

Patients with RCC had significantly better outcomes than any other given tumour type (Hazard ratio [HR] 0.01, P = .04), whereas patients with melanoma had significantly worse outcomes compared with all other pathologies together (HR 8.43, P = .002).

Better outcomes after PC were also found in women (HR 0.51, P = .07), and in patients who did not have a postoperative complication (HR 0.47, P = .05). Patients who had a postoperative complication did significantly worse after PC (HR 2.11, P = .05).

On multivariate analysis, only melanoma (HR 13.5, P < .001) and male gender (HR 2.6, P = .02) remained poor prognostic indicators. However, Dr. Reddy cautioned against interpreting the multivariate numbers because of the small sample size.

A subset analysis of only the patients with RCC found a median length of 9.4 years between the initial operation and metastectomy, with a median survival of 4.8 years. Most patients had metachronous lesions (86%), but no difference in survival was seen between these patients and those with synchronous lesions.

"These [RCC] cancers have a much more indolent behaviour pattern than, say, melanoma," said Dr. Reddy. "They tend to grow slow and generally present much earlier than other abnormal malignancies. If these cancers were to metastasise, having a single metastasis to the pancreas would be usual rather than widespread disease,"

Based on these data, Dr. Reddy said that if kidney cancer metastasises to the pancreas, "you shouldn't write these patients off anymore. You should surgically remove [the metastasis]."

Along with high efficacy, particularly in RCC patients, the study found PC to be safe, with an overall morbidity of 48%. "This may seem high," said Dr. Reddy, "but most of the morbidity was minor."

He also said the classification system used to calculate complications was a fairly stringent one and may account for why the numbers look so high. For example, the classification system considers blood transfusions to be a grade 2 complication, which, he said, many surgeons would argue they are not.

Overall, all patients had only grade 1 complications that required medical treatment. Only a small number of patients had pancreatic leaks (4%) and hepaticojejunotomy leaks (4%).

One caveat to these findings may be the enormous selection bias in the study. "All the patients in the study were highly favourable in that they only had 1 site of metastasis after 5 to 10 years," said Dr. Reddy.


[Presentation title: The Role of Pancreatic Metastectomy: A Retrospective Analysis of the Largest Single Institutional Series. Abstract P52]

E-Mail this DGDispatch to a colleague

To print, use this version