Several centers have investigated the potential benefit of neoadjuvant chemoradiation delivered before surgery in patients with resectable pancreatic cancer. This neoadjuvant strategy provides several theoretical advantages.[46,47] First, radiation may be most effective when delivered to well-oxygenated tissues not devascularized by surgery. Second, chemoradiation delivered before surgery may prevent dissemination and implantation of tumor cells at laparotomy. Third, downsizing tumors with preoperative chemoradiation may increase the likelihood of margin-negative resection. Fourth, patients whose previously occult metastatic disease progresses during preoperative chemoradiation will be spared a nontherapeutic laparotomy. Finally, reserving surgery as the final component of multimodality therapy allows the most toxic constituent of the multimodality regimen to be delivered last, thereby maximizing the likelihood that all patients will receive the fullest treatment in a timely manner.
One of the initial hopes regarding neoadjuvant chemoradiation in pancreatic cancer was that chemoradiation-induced tumor downsizing might render unresectable disease resectable. Several reports have evaluated this issue in patients with initially unresectable pancreatic cancer. Pilepich and Miller provided one of the earliest reports. In this series, 17 patients with adenocarcinoma of the pancreas underwent an initial laparotomy and were deemed to have localized tumors that were either unresectable or of borderline resectability. These patients were subsequently treated with external-beam radiation alone to a total dose of 40 to 50 Gy. Eleven patients continued to show no evidence of metastatic disease at restaging and were re-explored; six of the 11 patients underwent pancreaticoduodenectomy, and two of the six remained disease-free 5 years following surgery.
A similar small series of patients treated with neoadjuvant chemoradiation in the setting of initially unresectable disease was reported from the New England Deaconess Hospital (Boston, Mass). External-beam radiation (45 Gy) was delivered in conjunction with continuous infusion of 5-FU at a dose of 225 mg/m²/day. In this series, neoadjuvant therapy was well tolerated but not associated with a significant rate of tumor downsizing (assessed by the tumor diameter as measured by computed tomography). Ten patients completed neoadjuvant therapy without evidence of tumor progression and underwent laparotomy. Two who were found to have resectable disease amenable to pancreaticoduodenectomy remained disease-free at 20 and 22.5 months postresection.
While these small series involving patients with initially unresectable disease provide evidence that pancreatic resection is safe after neoadjuvant chemoradiation, it is clear that 5-FUbased neoadjuvant chemoradiation rarely converts an unresectable to a resectable tumor. Even in the face of significant tumor cell kill, the tumor may remain unresectable. For example, desmoplastic infiltration of the superior mesenteric vessels is likely to persist even in the face of a tumoricidal response to treatment. In a recent update of a pilot program initiated at Brown University (Providence, RI), 14 patients with locally advanced, initially unresectable adenocarcinoma were treated with an aggressive regimen of neoadjuvant chemoradiation involving 45 Gy with continuous infusion of cisplatin(Drug information on cisplatin) (25 mg/m²/day ×3 days) and bolus 5-FU (400 mg/m²/day × 3 days) on the first and fourth weeks. Nine patients eventually underwent surgical exploration following chemoradiation, and the tumors of eight patients were resected. Three of the eight resected specimens demonstrated a complete pathologic response to neoadjuvant therapy with no tumor in the resected specimen. However, five patients required en-bloc resection of the portal vein because of perceived tumor infiltration. These results underscore the fact that apparent conversion from unresectable to resectable disease is often a function of the experience and aggressiveness of the operating surgeon, and that dramatic cellular responses to therapy may not correlate with a reduction in the required extent of resection.
The Fox Chase Cancer Center (Philadelphia, Pa) has reported substantial experience using a neoadjuvant chemoradiation protocol involving 50.4 Gy radiation in conjunction with two cycles of chemotherapy.[51-54] In this program, chemotherapy consisted of continuous-infusion 5-FU (1000 mg/m²/day on days 2 through 5 and 28 through 32) and bolus mitomycin(Drug information on mitomycin) C (Mutamycin) (10 mg/m²on day 2). Over a 6-year period, 34 patients with localized pancreatic cancer were treated using this regimen. Of note, locally advanced lesions that would have been considered unresectable in many centers predominated in this cohort; only 13% of patients had evidence of an uninvolved superior mesenteric vein as assessed by venous angiography. Thirteen patients had undergone previous laparotomy, where their disease had been classified as unresectable. Following chemoradiation, radiographic determination of tumor volume suggested a single partial response and three minor responses. At restaging, 21 patients were eventually confirmed to have metastatic or unresectable disease, while 11 patients underwent a potentially curative resection. Margin-negative tumor excision was accomplished in 10 of these 11 patients (91%). Among the small number of patients whose disease was resected with curative intent, actuarial 5-year survival was 40%, and local tumor recurrence was documented in only one of 11 patients.
In contrast to these series involving primarily patients with locally advanced disease, two centers have evaluated the utility of chemoradiation delivered before surgery to patients whose tumors are judged to be resectable at the initial evaluation. The largest such series has been reported from the M. D. Anderson Cancer Center.[46,47,55] In this series, careful attention was paid to preoperative patient selection using contrast-enhanced, thin-section computed tomography and staging laparoscopy to limit accrual to patients with localized adenocarcinoma of the pancreatic head with or without involvement of the superior mesenteric vein (American Joint Committee on Cancer stage I, II, or III disease; T1-3, N0-1, M0). Patients were treated with neoadjuvant chemoradiation delivered to a total dose of 50.4 Gy, with concurrent continuous-infusion 5-FU at a dose of 300 mg/m²/day, 5 days/week. The outcome of 39 patients who underwent pancreaticoduodenectomy following this protocol has recently been reported. Intraoperative radiation to a dose of 10 Gy was used in 33 of these 39 patients, and 13 required segmental resection of the superior mesenteric-portal vein confluence. Margin-negative resection was accomplished in 32 of 39 patients (82%). At a median follow-up of 19 months (range, 4 to 56), median survival was 19 months, and actuarial 4-year survival was 19%. This regimen appeared to significantly improve local tumor control and alter the pattern of disease recurrence; local or peritoneal recurrence was reported in only four patients (10%). The liver became the most frequent site of failure, with 53% of patients eventually developing liver metastases. A separate pilot protocol investigated the ability of additional low-dose (23.4 Gy) hepatic irradiation to further reduce the incidence of hepatic recurrence; this protocol was terminated prematurely due to excessive liver-specific toxicity.
The M. D. Anderson Cancer Center data suggest that 5-FUbased neoadjuvant chemoradiation offers a long-term survival benefit equivalent to that offered by postoperative radiation as documented by the GITSG. In addition, the low rate of local recurrence documented following neoadjuvant chemoradiation in patients with resectable disease compares favorably with the 43% rate of local recurrence reported following postoperative chemoradiation. To date, no randomized trial has documented these apparent benefits of neoadjuvant chemoradiation. However, a retrospective analysis of patients with resectable pancreatic cancer treated with neoadjuvant chemoradiation vs traditional postoperative chemoradiation has recently been reported. This report reviewed the outcome of 60 patients undergoing 5-FUbased chemoradiation and resection; 41 patients received preoperative chemoradiation followed by pancreaticoduodenectomy, while 19 patients received surgery first followed by traditional postoperative chemoradiation. The median tumor diameter was larger in the preoperative chemoradiation group, reflecting the need for successful preoperative tissue diagnosis by computed tomography-guided percutaneous needle biopsy. Trends toward a higher rate of margin-negative resection (88% vs 74%) and a lower rate of locoregional tumor recurrence (10% vs 21%) were observed in the preoperative group.
A retrospective comparison of patients treated by neoadjuvant radiation followed by surgery vs surgery alone has been reported by Ishikawa and colleagues from Osaka, Japan. In this series, 54 consecutive patients with apparently resectable cancer of the pancreatic head were selected to receive neoadjuvant radiation (n = 23) vs immediate surgery (n = 31) based on patient preference. These two groups were similar in terms of tumor size, extent, histologic differentiation, and incidence of nodal metastasis. The group receiving neoadjuvant therapy was treated to a dose of 50 Gy targeted to the pancreatic head as well as surrounding nodal and retroperitoneal tissues. No radiation-sensitizing chemotherapy was employed. At the time of surgical exploration, seven patients who had received neoadjuvant therapy and 19 who had not were found to have disease amenable to surgical resection. Data regarding the status of the surgical margins in each group were not provided. Among the patients who underwent surgery, the group treated with neoadjuvant radiation demonstrated a reduced incidence of death with regional recurrence (12% vs 35%; P < .05), improved median survival (15 vs 11 months; P = NS), and improved 1-year survival (75% vs 43%; P < .05). However, the 3-year and 5-year survival rates were not different between the two groups, indicating an identical rate of hepatic metastases.
For patients with unresectable gastric cancer, limited but nonetheless convincing data demonstrate that 5-FUbased chemoradiation, when given at sufficient doses, can provide effective palliation. The indications for such treatment are limited, however, as most patients have coexistent distant disease that is more effectively treated with systemic chemotherapy.
Currently, the use of chemoradiation in treating resectable gastric cancer remains investigational. When available, the results of the Intergroup 0016 trial will provide additional information about the utility of a 5-FUbased regimen in the adjuvant setting. Future studies will no doubt include the use of other radiation sensitizers like paclitaxel(Drug information on paclitaxel) and gemcitabine(Drug information on gemcitabine), alone or in combination with 5-FU. Further neoadjuvant trials are necessary to assess whether the theoretical benefits of this approach are realized.
For surgeons, the primary goal in managing patients with resectable pancreatic cancer should be to perform a margin-negative pancreaticoduodenectomy safely. Together with careful preoperative staging and the selective application of extended resection techniques,[59,60] the use of neoadjuvant chemoradiation may increase the likelihood of achieving this goal. Future studies will continue to provide important data regarding optimal neoadjuvant treatment regimens and will determine which subgroups of patients are most likely to benefit from this approach.