Safety of Resection for Pancreatic Metastases

In a recent review of 19 series, a mortality rate of 1.2% and a morbidity rate of 38.3% were reported for resection of isolated pancreatic metastases (Reddy & Wolfgang, 2009). Resections for metastatic disease may be more difficult than resections for primary pancreatic tumors because of the necessity of extensive adhesiolysis after previous surgery in approximately 80% of patients and because of the necessity of multivisceral resections. Our series, which included 41% multivisceral resections, reported a mortality rate of 4.4% and a morbidity rate of 33% (Strobel et al, 2009). These findings are comparable to the data reported for primary pancreatic malignancies and show that resection for pancreatic metastases can be performed safely by specialized surgeons.

Oncologic Outcome of Resection for Pancreatic Metastases

The follow-up results of the largest studies of resection for pancreatic metastases are summarized in Table 60.1 and show a very favorable outcome. However, most series include several primary tumor entities, and conclusions about individual tumor types must be made with great caution. RCC is, by far, the most frequent primary tumor associated with resectable, isolated pancreatic metastases. The frequency of various primary tumors derived from surgical series is given in Table 60.2. The data for resection of RCC metastatic to the pancreas are summarized in Table 60.3; they show considerably longer survival rates compared with the mixed group of other histologies (see Table 60.1). Two systematic reviews of 236 and 311 published cases of surgery for RCC pancreatic metastases (Sellner et al, 2006; Tanis et al, 2009) report 5-year survival rates of about 75% after radical resection. However, because both reviews included case reports and smaller series, they may overestimate survival rates because of publication bias. However, RCC histology was also associated with improved outcome in several of the published series, including that of the authors (Eidt et al, 2007; Hiotis et al, 2002; Reddy et al, 2008; Strobel et al, 2009).

Table 60.2 Pathologic Diagnosis in Patients with Pancreatic Resections for Metastases

Based on combined data from Reddy & Wolfgang, 2009, and Strobel et al, 2009.

Although 57% of cases in our series had extrapancreatic metastatic disease, the overall 5-year survival rate was 56.8%, and the 5-year survival rate was 67.3% after resection of RCC metastases. The 5-year survival rate after resection of isolated pancreatic metastases was as high as 75% (Strobel et al, 2009).

Efficacy of Resection for Pancreatic Metastases

Although comparison of the survival data after metastasectomy with the historic results of conservative treatment suggests a benefit of surgery, the efficacy of surgical resection and its superiority over conservative treatment alone cannot be formally proven because of the lack of controlled trials. However, for RCC, three studies observed longer survival in patients after resection compared with noncontrolled groups with irresectable disease (Sellner et al, 2006; Tanis et al, 2009; Zerbi et al, 2008). Furthermore, the survival data after pancreatic resection for RCC metastases compare very favorably to the survival reported for resection of RCC metastases in other organs (Kavolius et al, 1998). Another point supporting resection for pancreatic metastases is good local tumor control. In our series, local tumor control after 5 years was 87%, but disease-free survival was only 33% because of distant metastases (Strobel et al, 2009). This underlines the importance of interdisciplinary therapy with consideration of adjuvant treatment in these patients.

A major limitation of the literature concerning pancreatectomy for metastatic disease is the heterogeneity of patients, with most series including metastases from several different primary tumors. Except for RCC, the number of patients with other diagnoses is too small to draw reliable conclusions about the outcome and value of pancreatic resection for metastases. Thus, surgical decision making is generally based on the overall experience with resection for pancreatic metastases and on the reported outcome for each tumor type.

Presentation of Patients with Pancreatic Metastases

Patients with pancreatic metastases can be divided into two groups: those with symptoms and those without. Patients in the first group experience typical symptoms caused by a pancreatic tumor, and the diagnostic workup reveals a pancreatic mass that often has features atypical for primary pancreatic cancer (Palmowski et al, 2008). In the second group of patients, the lesion is found during standardized follow-up exams for the primary tumor or by imaging for other reasons. Within the past few decades, the distribution has shifted from symptomatic to asymptomatic patients, most likely as a result of standardized follow-up programs and broad use of cross-sectional imaging techniques. In studies that include patients from previous decades, 60% to 90% of patients were seen because of symptoms (Reddy et al, 2008; Sellner et al, 2006; Tanis et al, 2009); in the authors’ recent series, only 18% of metastases were diagnosed because of symptoms, and 82% were detected during oncologic follow-up (Strobel et al, 2009). Patients with symptoms appear to have a worse prognosis, as has been shown for RCC metastases (Tanis et al, 2009). These findings suggest that symptoms eventually develop as local complications during progression of pancreatic metastases and present another point in favor of surgery for resectable metastatic disease.

Interestingly, pancreatic metastases frequently are identified long after resection of the primary tumor. This is best documented for RCC metastases, which are diagnosed at a median interval of 10 years after resection of the primary tumor (Sellner et al, 2006; Strobel et al, 2009). This long disease interval must be taken in account in the design of oncologic follow-up programs.

Diagnostic Workup for Suspected Pancreatic Metastases

Usually, a pancreatic mass is first detected by contrast-enhanced computed tomography or magnetic resonance imaging. Pancreatic metastases have features different from primary pancreatic malignancies (see Chapters 16 and 17). In contrast to pancreatic adenocarcinoma, most metastatic entities exhibit contrast enhancement, as shown in Figure 60.1 (Palmowski et al, 2008), particularly mestastatic RCC. Another feature of metastases is their multifocality within the pancreas: 40% in pathologic workup in our series. Perhaps because of the long interval between the primary tumor resection and the pancreatic metastases, these lesions are often misdiagnosed as primary pancreatic malignancy.

FIGURE 60.1 Typical radiologic features of pancreatic metastases. On contrast-enhanced computed tomography, pancreatic metastases typically appear as lesions with contrast enhancement in the arterial phase. A, Metastases of renal cell cancer to the pancreatic head and tail (arrows) exhibit intense contrast enhancement in the arterial phase. B, Metastases of soft tissue sarcoma to the pancreatic head (arrow) with contrast enhancement and central hypodensity.

Clearly, if a pancreatic lesion is detected in a patient with a history of cancer, the serum tumor markers for the previous malignancy, if relevant, should be determined, and state-of-the-art restaging should be performed to rule out local recurrence and extrapancreatic systemic disease. The importance of a thorough workup is highlighted by the high frequency of additional, extrapancreatic lesions in patients with pancreatic metastases. If additional extrapancreatic disease is suspected, positron emission tomography is recommended to identify patients with diffuse spread who will not benefit from surgery. In contrast, preoperative tissue diagnostics are usually not needed, especially in patients with disease confined to the pancreas, because it does not alter the indication for operation, which would also be offered for a primary pancreatic malignancy without evidence of spread. Only in cases with extrapancreatic spread or features that would otherwise contraindicate resection should a tissue diagnosis be pursued because it may be helpful to rule out metastatic pancreatic cancer.

Patient Selection for Surgery

As discussed, solid data regarding the outcome of resection for pancreatic metastases are lacking for tumor types other than RCC. Decision making and selection of patients for surgery must therefore be considered on an individual basis and by using several clinical parameters, including type and history of malignant disease, pattern of metastatic spread, patient comorbidities, and operative risk. Ideally, the decision process should involve both medical oncologists and experienced pancreatic surgeons and does not differ much from decision making for resection of other metastatic sites, such as liver and lung.

Technique of Resection for Pancreatic Metastases (See Chapter 62A)

No studies have compared different techniques of pancreatic resection, and most larger studies only describe radical resections. However, in one report, a high rate of pancreatic recurrence was observed after limited resections (Bassi et al, 2003). In our series, we performed standard partial resections but avoided total pancreatectomy whenever possible and observed only two pancreatic recurrences after 30 partial (6.7%) pancreatectomies (Strobel et al, 2009). Because pancreatic metastases are frequently multifocal (Fig. 60.2), a thorough exploration of the pancreatic remnant by palpation and ultrasound is mandatory whenever a partial pancreatectomy is performed (Zerbi et al, 2008). Because the two largest series showed that pancreatic metastases are associated with peripancreatic lymph node metastases in 26% to 30% of patients (Reddy et al, 2008; Strobel et al, 2009), a standard peripancreatic lymphadenectomy should be performed during pancreatic resection for metastases. To achieve curative resection, multivisceral resections should be performed if the operative risk and potential benefit are acceptable.

FIGURE 60.2 Pancreatic metastases of sarcoma. Cross-sections through the pancreas revealed two independent lesions located in different parts of the pancreas.