FIGURE 3 Hematoxylin & Eosin staining of PD specimen reveals poorly ductal adenocarcinoma with signet ring appearance at low (A) and high power (B). There is lack of glandular formation in addition to severe cellular atypia. |
Surgical Management of Pancreatic Cancer
DR. ARVIND DASARI: The patient had pre-operative biliary stenting that was unfortunately complicated by stent blockage a couple of weeks later and required stent exchange. When is pre-operative biliary drainage indicated?
DR. MARTIN McCARTER: Biliary drainage can be performed surgically, percutaneously, or endoscopically with either plastic or metal stents. A recent well-designed multicenter, randomized trial compared pre-operative biliary drainage with plastic stents followed by surgery four to six weeks later (n = 106) with immediate surgery within a week of diagnosis (n = 96) in patients with cancer of the pancreatic head and elevated total bilirubin (2.3 – 14.6 mg/dl). In line with previous studies, this study failed to show any mortality benefit from pre-operative biliary drainage, and in fact showed that the group with pre-operative stenting had significantly higher morbidity within four months of randomization—mostly from cholangitis and stent occlusion.[9] However, this study used plastic stents and not self-expanding metal stents (SEMS), which have a longer patency and fewer stent-related problems than plastic stents because of their larger diameter.[10] Therefore, pre-operative biliary stenting is best reserved for patients with acute cholangitis and severe cholestatic symptoms, in whom surgery will be delayed because of neoadjuvant therapy or for other reasons. In such cases, the preferred approach is endoscopic.
DR. ARVIND DASARI: Could you review the surgical management of this patient and your selection of this approach?
DR. MARTIN McCARTER: The patient underwent a Whipple’s procedure; i.e., the classical pancreaticoduodenectomy (PD) that’s done for tumors of the pancreatic head or uncinate process. It includes resection of the head of the pancreas; lower part of the body and pylorus of the stomach; proximal small intestine including the duodenum; the first 10-15 cm of the jejunum; and common bile duct (CBD), along with cholecystectomy. The jejunum is then anastamosed to the pancreas, bile duct and gastric stump.[11] Since its original description, there have been a number of modifications made to the Whipple’s procedure in an effort to improve outcomes and decrease morbidity. These improvements include preservation of the pylorus, total pancreatectomy, extended lymphadenectomy, and variations in pancreatic anastamoses. However, none of these variations have been shown to improve outcomes significantly. In contrast, resection and reconstruction of the portal vein (PV) or superior mesenteric-portal vein confluence (SMPV) can be performed with acceptable added morbidity and several retrospective studies have found that the median survival time remains similar (20-26 months) to those who undergo PD without SMPV resection.[12-16] On evaluation intra-operatively, our patient did not have any evidence of vascular invasion—however several peri-pancreatic lymph nodes were enlarged and were removed with the operation. Not surprisingly, one factor known to have a significantly favorable impact on outcome is the surgeon and hospital volume.[17-19] The NCCN guidelines recommend that pancreatic resections be done at hospitals that perform at least 15 – 20 resections per year.
Pathological Evaluation
DR. ARVIND DASARI: Dr. McManus, could you discuss the pathological findings here?
DR. MARTINE McMANUS: We received a PD specimen that included the distal stomach, duodenum, head of the pancreas, and multiple lymph nodes fixed in formalin (Figure 3). The tumor was in the head of the pancreas with a maximum dimension of 3.1 cm and with extension into the adjacent duodenum. Lymphovascular and perineural invasion were present. The common bile duct margin was positive and pancreatic neck surgical margin was close (0.3 mm). 12 out of 20 regional lymph nodes examined had metastatic tumor. Histological examination revealed poorly differentiated ductal adenocarcinoma with signet ring cell features invasive into the peripancreatic soft tissue.
DR. ARVIND DASARI: Could the positive margin explain the immediate recurrence after resection in our patient?
DR. MARTINE McMANUS: The majority of pancreatic ductal adenocarcinomas invade the peri-pancreatic adipose tissue and can involve the intrapancreatic bile duct as well as the duodenal wall. It is therefore very important to have standardized criteria for surgical staging. Approximately 90% of PD specimens have regional lymph node involvement. Only 15% have R0 resection, defined as a negative margin of at least 1 mm. About 43% have multiple positive margins, most commonly with involvement of the retroperitoneal margin (soft tissue that contains perineural tissue adjacent to the superior mesenteric artery).[20] Most recurrences arise in the pancreatic bed along this critical margin. Other than positive margins, pathologic factors associated with decreased survival duration include metastatic disease in regional lymph nodes, poorly differentiated histology, and increased size of primary tumor. Another important feature is the presence of pancreatic cancer stem cells, which compose just 1-5% of the tumor but are capable of unlimited self-renewal and are also resistant to chemotherapy and radiation, which makes them difficult to eradicate.[21] All of these factors probably played a role in the recurrence seen in our patient.
Medical Management
DR. ARVIND DASARI: Dr. Messersmith, our patient has unfortunately been diagnosed with metastatic pancreatic cancer. Gemcitabine(Drug information on gemcitabine) is the mainstay of treatment in this setting. Could you review the benefit of palliative gemcitabine chemotherapy?
DR. WELLS MESSERSMITH: The trial that established gemcitabine as the mainstay of treatment in this setting was conducted in 1997 by Burris and colleagues on 126 patients randomized to either gemcitabine (1000 mg/m2 bolus weekly × seven weeks followed by one week of rest then weekly × three weeks every four weeks thereafter) or 5-FU (600 mg/m2 weekly).[22] The primary endpoint was clinical benefit response, defined as a sustained improvement in cancer-related symptoms. Significantly higher clinical benefit response was seen in 23.8 % of gemcitabine-treated patients vs 4.8% of 5-FU-treated patients (P = .0022). Significant benefit with gemcitabine was also seen in the secondary end points, including median survival (5.6 months vs 4.4 months,
P = .0025) and one-year survival rate (18% vs 2%).
DR. ARVIND DASARI: Is there evidence that shows a benefit from adding another cytotoxic chemotherapeutic agent to gemcitabine in metastatic pancreatic cancer?
DR. WELLS MESSERSMITH: Several large phase III trials have been conducted combining gemcitabine with conventional cytotoxic agents such as exatecan, cisplatin, oxaliplatin, 5-FU, irinotecan, capecitabine(Drug information on capecitabine), and pemetrexed(Drug information on pemetrexed) but the primary end point of improved overall survival (OS) was not met in any of these trials. Some trials did show improvement in response rate or progression-free survival (PFS).[23] However, some investigators believe that combination therapy with a platinum or capecitabine may improve outcomes in patients with good performance status. Combination therapy with capecitabine or a platinum compound has been evaluated in several phase II and two phase III trials and, overall, failed to show a convincing improvement in outcomes. However, meta-analyses of these trials have consistently shown a slight survival benefit in favor of combination therapy. These benefits were even more marked in patients with excellent PS.[24]
These results are supported by a meta-analysis of 15 trials with over 4,000 patients comparing gemcitabine to gemcitabine combination therapy. [25] This study showed a modest OS benefit (hazard ratio, HR of 0.91, 95% CI, 0.85 – 0.97, P = .004) that was limited to trials with platinum or fluoropyrimidine-based regimens with no benefit from other combinations. However, within this study, a meta-analysis of five trials with 1,682 patients with information on baseline PS showed a more substantial survival benefit with combination therapy (HR = 0.76; 95% CI: 0.67 – 0.87, P < .0001) in patients with a good performance status (ECOG PS 0-1) in contrast to patients with a poor performance status (ECOG 2) who did not appear to benefit from combination therapy (HR = 1.08,
P = 0.40). Thus, these meta-analyses suggest that gemcitabine-based cytotoxic combination chemotherapy is beneficial in patients with good PS. However, most individual phase III trials are negative – suggesting that any absolute benefit is modest. Moreover, in a setting where the median survival is only six months, and the primary objective is symptomatic relief, the slight survival benefit may be outweighed by increased toxicity and complications with combination therapy. I would consider combination therapy in patients with excellent PS (ECOG 0-1) after a detailed discussion with them.
DR. ARVIND DASARI: How about combining gemcitabine with a targeted agent?
DR. WELLS MESSERSMITH: Overall, the evidence collected over the last decade has shown that adding cytotoxic chemotherapy to gemcitabine will probably not provide any meaningful benefit in OS. As a result, multiple attempts were made to improve outcomes by adding targeted agents to gemcitabine, including several in large phase III trials. Of these, the only agent shown to have any benefit was erlotinib, in a phase III study of 569 patients conducted by the National Cancer Institute of Canada in 2007.[26] Patients were randomized to gemcitabine monotherapy (administered as in the Burris trial) or in combination with erlotinib (100 mg po daily). Median survival (6.2 months vs 5.9 months, P = .038), 1-year survival (32% vs 17%) and PFS (HR 0.77, 95% CI, 0.64 – 0.92, P = .004) were significantly better in the combination arm. There is considerable doubt as to whether such a marginal benefit is clinically significant.
Current NCCN guidelines recommend single agent gemcitabine or, in patients with good PS, combination of gemcitabine with cisplatin(Drug information on cisplatin), erlotinib or a fluropyrimidine. These guidelines also recommend fixed dose rate gemcitabine as a category 2B recommendation—however, based on the phase III trial that was published since, I do not believe that there is any benefit with this approach.
DR. ARVIND DASARI: If our patient progresses on first-line gemcitabine, what options are available for second-line therapy?
DR. WELLS MESSERSMITH: In contrast to the first-line setting, there are no recommendations for the second-line therapy of advanced pancreatic adenocarcinoma. However, current data suggests that a fluoropyrimidine in combination with oxaliplatin(Drug information on oxaliplatin) may have some activity. The phase II CONKO-003 trial randomized patients with gemcitabine-resistant pancreatic cancer to 5-FU (2 gm/m2 over 24 hours) and leucovorin (200 mg/m2 over 30 minutes) on days 1,8,15 and 22 with (OFF) or without oxaliplatin (85 mg/m2 days 8 and 22) over a 42 day cycle. In a preliminary report, OFF was associated with longer OS (26 vs 13 weeks, P = .014) and PFS (13 vs 9 weeks, P= .012).[27] In another phase II study, oxaliplatin (130 mg/m2 on day 1) was evaluated in combination with capecitabine (1000 mg/m2 twice daily for 14 days). The doses of oxaliplatin and capecitabine were reduced in patients with age > 65 years or ECOG PS 2 (to 110mg/m2 and 750 mg/m2 respectively). Of the 39 evaluable patients, 1 patient had a partial response and 10 had stable disease. Median OS was 23 weeks (95% CI 17 – 31 weeks) and PFS was 9.9 weeks (95% CI 9.6 – 14.5 weeks). The 6-month and 1-year survivals were 44% (95% CI, 31%-62%) and 21% (95% CI, 11%-38%), respectively.[28] Based on these small studies, although it would be difficult to make definitive recommendations, a regimen containing oxaliplatin and a fluoropyrimidine (typically capecitabine for convenience) is a reasonable second-line choice in patients with good performance status. Insurance coverage, particularly in Medicare patients, can be an issue since oxaliplatin is not listed on all compendia for pancreatic cancer.
To summarize, this is a patient who presented with pancreatic cancer and who was found to have more extensive disease than was initially determined per peri-operative staging. She underwent pre-operative biliary stenting followed by an uncomplicated PD and presented to the clinic for discussion regarding adjuvant therapy, but was found to have new metastatic disease. She subsequently started palliative chemotherapy with single agent gemcitabine which she tolerated well, and had stable disease for three months before developing progressive disease. At that time, after discussion regarding second-line therapy with cytotoxic therapy vs enrollment in a clinical trial, she was started on a phase I clinical trial, per her wishes. She progressed rapidly on this trial and finally elected palliative care.
Concluding Remarks
DR. ARVIND DASARI: During evaluation of pancreas cancer, it is important to remember that CT scans have high sensitivity and specificity in primary tumor diagnosis and adjacent vessel invasion. They also have high sensitivity and specificity in detecting metastatic lesions greater than 1 cm. MRIs can be useful to evaluate subcentimeter metastatic lesions in the liver. CT scans also have limited utility in detecting lymph node and peritoneal spread. In cases where there is a high probability of occult metastatic disease, diagnostic laparoscopy is a reasonable approach.
In spite of modern treatment modalities, resectable pancreatic cancer continues to have a high rate of recurrence and mortality. These recurrences are secondary to multiple factors including high rates of positive surgical margins, lymph node involvement and tumor biology. In the metastatic setting, gemcitabine continues to be the mainstay of first-line treatment. Multiple attempts have been made to improve outcomes by adding a second cytotoxic or targeted agent, however, the only agents shown to have significant benefit are capecitabine and erlotinib, and even then, the benefits are slight and at the cost of side effects. Therefore, these agents should be reserved for patients with good performance status. In the second-line setting, a combination of oxaliplatin and a fluoropyrimdine is a reasonable option for patients with preserved performance status.
Financial Disclosure: The authors have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.
