Introduction
Pancreatic ductal adenocarcinoma (PDAC) is a lethal disease with an extremely low (<6%) overall survival (OS) rate (1,2). Recently, 5-fluorouracil, leucovorin, irinotecan, and oxaliplatin (FOLFIRINOX) therapy and gemcitabine (GEM) plus nab-paclitaxel (PTX; GnP) therapy were revealed to improve treatment outcomes among patients with unresectable PDAC (3,4); however, surgery remains the only method for achieving cure or long-term survival. Previously, we attempted radical surgery for patients with localized disease, including wide lymph node dissection and complete removal of the extrapancreatic nerve plexus around the superior mesenteric artery or celiac axis, in an effort to extend survival (5,6). Moreover, in some cases, en bloc resection of the pancreatic head with the superior mesenteric artery was performed (7). However, even in patients who underwent R0 resection, the 5-year survival rate was poor (7-24%), and the median survival was approximately 1 year in most series. These unfortunate results indicating that surgery alone is inadequate are likely attributable to early distant metastasis before surgery (8). Therefore, postoperative adjuvant chemotherapy is important (9). Recently, it was demonstrated that adjuvant chemotherapy for PDAC with S-1, an oral fluoropyrimidine analog, significantly improved OS and recurrence-free survival after PDAC resection compared with the effects of GEM (10).
An important limitation of adjuvant chemotherapy for PDAC is that a certain proportion of patients cannot receive the designated therapy because of postoperative complications or early disease recurrence (11-13). For this reason, preoperative neoadjuvant chemotherapy (NAC) has attracted attention. The theoretical advantages of NAC include the early treatment of occult metastases, reduction of the risk of tumor seeding during surgery, and improved tolerance compared with postoperative therapy (14). Meanwhile, the potential disadvantages include problems related to biliary drainage during chemotherapy, potential progression to an unresectable stage in patients whose disease does not respond to therapy, and an increasing risk of postoperative complications (14).
Recently, several important studies of preoperative chemotherapy for unresectable or borderline resectable PDAC have been reported (15-18). However, few studies have assessed NAC in the treatment of resectable PDAC. We previously used NAC for resectable PDAC and found that NAC with GEM-based regimens is useful for pancreatic head cancer with lymph node metastases (19). However, the effectiveness of GnP therapy for patients with resectable PDAC has not been investigated. Therefore, we conducted a phase I study of NAC using GnP for resectable PDAC to determine the maximum tolerated dose (MTD) of each drug.
Materials and methods
Patient selection. Patients with radiologically proven resectable pancreatic cancer were eligible for this study. Other eligible criteria included an age of 20-79 years; Eastern Cooperative Oncology Group performance status of 0 or 1; and adequate renal function (normal serum creatinine and blood urea nitrogen levels), liver function [total bilirubin level, <2.5x the upper normal limit (UNL) or <3x the UNL after biliary drainage if the patient had jaundice and serum transaminase (GOT, GPT) levels, and <2.5x the UNL or <3x the UNL after biliary drainage if the patient had jaundice], bone marrow reserve (white blood cell count, 4,000-12,000 mm3; neutrophil count, >2,000 mm3; platelet count, >100,000 mm3; and hemoglobin level, >9.5 g/dl), and pulmonary function (PaO2, >70 mmHg). All enrolled patients were required to have completed or discontinued prior treatment (tumor resection, chemotherapy, immunotherapy, or radiotherapy) at least 4 weeks prior to study enrollment.
The exclusion criteria were as follows: Pulmonary fibrosis or interstitial pneumonia, marked pleural or pericardial effusion or marked peripheral edema, severe heart disease, difficult-to-control diabetes mellitus, active infection, pregnancy or lactation, childbearing age in women who did not use effective contraception, severe drug hypersensitivity, appearance of distant metastases during preoperative chemotherapy, severe neurological impairment or mental disorder, active concomitant malignancy, and other serious medical conditions.
This clinical trial was approved by the Institutional Review Board of Kanazawa University Hospital (no. 5849) and registered with the UMIN Clinical Trials Registry (ID: UMIN000011062).
Study design
This was an open-label, single-center, nonrandomized, dose-escalation phase I study. The primary endpoints of this study were the optimal dose and safety of GnP therapy as preoperative chemotherapy. The secondary endpoints were disease-free survival (DFS), treatment response, and adverse events including postoperative complications. All laboratory tests to assess eligibility were completed within 7 days prior to the start of treatment. Nab-PTX was administered for 60 min followed by an infusion of GEM for 30 min intravenously on days 1, 8 and 15 in two 28-day cycles. Surgery was performed >14 days after chemotherapy ended. The dose of each drug in this study was planned as follows: Level 1, 75 mg/m2/day for nab-PTX and 600 mg/m2/day for GEM; level 2, 100 mg/m2/day for nab-PTX and 800 mg/m2/day for GEM; and level 3, 125 mg/m2/day for nab-PTX and 1,000 mg/m2/day for GEM.
Definition of dose-limiting toxicities (DLTs) and MTD
DLTs were assessed during both treatment cycles. DLT was defined using the National Cancer Institute Common Toxicity Criteria for Adverse Events version 4.0(20) as one or more of the following effects attributable to the study drug: i) Grade 3/4 neutropenia complicated by fever; ii) grade 4 neutropenia lasting longer than 4 days; iii) grade 4 thrombocytopenia; iv) any other grade 3/4 non-hematologic toxicity excluding anorexia, nausea, and vomiting in the absence of an appropriate antiemetic; and v) delayed recovery from treatment-related toxicity for more than 2 weeks. At least 10 patients were enrolled at each dose level. If DLTs were observed or treatment could not be completed in <50% of patients, dose escalation was continued. If DLTs were observed or treatment could not be completed in >50% of patients, treatment at that dose was discontinued. The MTD was then set at the preceding dose level.
Assessment of efficacy
Treatment responses were evaluated according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1(21). Complete response was defined as the disappearance of all clinical evidence of a measurable tumor. Partial response was defined as a ≥30% reduction in the sum of the products of two perpendicular diameters of all measurable lesions compared with the baseline values with no evidence of new lesions. Stable disease (SD) was defined as a <30% reduction or <20% increase in the sum of the products of two perpendicular diameters of all measurable lesions compared with the baseline values with no evidence of new lesions. Progressive disease (PD) was defined as an increase of ≥20% in the sum of the products of two perpendicular diameters of all measurable lesions compared with the baseline values, the appearance of any new lesion, or deterioration of clinical status consistent with disease progression. To assess objective responses, patients were evaluated after two cycles of preoperative chemotherapy.
Postoperative complications
The global morbidity rate and types of complications were evaluated according to the Clavien-Dindo classification (22). Mortality was defined as any deaths related to surgery.
Pathological diagnosis
All surgically resected specimens were immediately fixed in 10% neutral-buffered formaldehyde solution. As described previously, after the specimens had been cut horizontally into 5-mm tissue blocks (23), they were dehydrated and embedded in paraffin. Finally, 5-mm sections were cut and stained with hematoxylin and eosin. Each section was examined by pathologists using light microscopy. The tumors, including the effect of preoperative chemotherapy, were evaluated according to the General Rules for the Clinical and Pathological Study of Pancreatic Cancer proposed by the Japanese Pancreatic Cancer Group (24). The degree of tumor destruction was graded on a scale of 1 to 4 as follows: Grade 1, poor or no response; grade 1a, estimated residual tumor rate ≥90%; grade 1b, residual tumor rate of 50-90%; grade 2 (moderate response), residual tumor rate of 10-50%; grade 3 (marked response), estimated residual tumor rate of <10%; and grade 4 (complete response), presence of no viable tumor cells.
Statistical analyses
Categorical variables were compared using the Chi-squared test, Student's t-test, and the paired t-test. P<0.05 indicated statistical significance. All analyses were performed using SPSS II statistical software (version 23.0; SPSS, Inc.).
Discussion
The National Comprehensive Cancer Network guidelines recommend that NAC with GnP or (modified) FOLFIRINOX should be considered for patients with borderline resectable PDAC (26). Conversely, for resectable PDAC, frontline surgery is recommended for all but high-risk patients. However, patients with pancreatic head cancer in particular require a long postoperative recovery period before adjuvant chemotherapy because of surgical stress. Therefore, preoperative therapy could be expected to reduce the risk of distant metastasis. We introduced NAC for resectable PDAC using GEM plus S-1 in 2007 (14,27). Recently, the effectiveness of NAC with GEM plus S-1 for resectable PDAC was demonstrated in a randomized phase II/III clinical trial (28), and this regimen has become the standard treatment in Japan.
GnP and FOLFIRINOX are currently in use as NAC for patients with borderline resectable or locally advanced PDAC (18,29,30). Okada et al (31) reported a phase I study of GnP as NAC for borderline resectable PDAC at doses equivalent to level 3 in the current study (nab-PTX 125 mg/m2/day and GEM 1,000 mg/m2). The mean frequency of administration in their study was 4.6 of 6 possible doses during two treatment cycles, compared with 5.4 at level 1 and 4.8 at level 2 in our series. From these results, our results are considered valid, but because <50% of planned doses were administered at level 2, we concluded that the MTD was dose level 1. The most frequent DLT was neutropenia, as described previously (4,31). Whether the dose of the drug should be increased in combination with granulocyte colony-stimulating factor to prevent neutropenia is a future issue. There were no serious adverse events such as sepsis, febrile neutropenia, interstitial pneumonia, and severe peripheral sensory neuropathy in the current study, probably because of the short-term treatment duration. Skin rash caused by nab-PTX, occurring in four patients (10.0%), was not a severe adverse event, but this event precluded further protocol treatment. Fortunately, these four patients were able to continue chemotherapy with another regimen (GEM alone or modified FOLFIRINOX therapy) and finally undergo surgical resection.
Postoperative complications attributable to NAC were not recognized. Although infectious complications were observed in patients treated at dose level 2, these events were considered accidental phenomena. Acute appendicitis developed immediately before discharge, and infectious spondylitis occurred in a patient with steroid-treated autoimmune disease. The most common complication was grade B pancreatic fistula. There was no severe complication classified as grade IV or V in this study. Two cycles of GnP therapy as NAC are not considered to have safety issues based on these findings.
The most important purposes of preoperative chemotherapy are preventing metastasis from the primary site and treating occult metastasis. Recently, it was reported that epithelial-mesenchymal transition (EMT) of cancer stem cells plays an important role in tumor invasion, metastasis, and chemoradioresistance in PDAC (32). PTX suppresses the EMT of cancer stem cells and activation of cancer-associated fibroblasts (33-36). We previously reported that NAC with GnP reduced the numbers of cancer-associated fibroblasts (37). Therefore, we determined that GnP therapy is most suitable for preoperative chemotherapy theoretically (19).
However, it was recently demonstrated that FOLFIRINOX was associated with longer OS than GnP in patients with unresectable locally advanced or metastatic PDAC (38). Moreover, FOLFIRINOX treatment was reported to decrease the frequency of lymph node metastasis (30,39), consistent with our results supporting the effectiveness of NAC in patients with node-positive pancreatic head cancer (19). In addition, in a study limited to NAC for resectable and borderline resectable pancreatic head cancer, the OS of the FOLFIRINOX group was slightly higher than that of the GnP group, although the difference was not significant because of the limited frequency of treatment in the preoperative period (39).
Recently, many studies on preoperative chemoradiotherapy (CRT) for locally advanced or borderline PDAC have been published and these data demonstrate that CRT improves resectability, R0 resection rate, OS and DFS rate, and reveal a survival advantage compared with NAC cases (40,41). On the other hand, few authors have prospectively evaluated the role of CRT for preoperative treatment for resectable PDAC (42). Theoretically, radiotherapy may have a marked impact on microscopic tumor spread and local disease management. However, it is reported that radiotherapy could induce EMT in cancer cells (43) and increase distant metastasis in patients with locally advanced PDAC (44). In recent years, progression of image diagnosis made possible accurate understand of the tumor spreading (45), for that reason we have adopted optimal resection rather than radiation for local spread of resectable PDAC.
In pancreatic head cancer treatment in particular, postoperative chemotherapy may often be limited by surgical stress, and thus, the concept that more powerful regimens should be used before surgery was also affirmed. Uesaka et al (10) obtained good results for postoperative adjuvant chemotherapy with oral S-1, which has become the standard treatment in Japan, in patients with PDAC. In this study, adjuvant chemotherapy with S-1 was administered to approximately 80% of patients with PDAC who underwent resection. In our previous study (19), it was revealed that NAC with GEM-based regimens improve prognosis of pancreatic head cancer with lymph node metastases compared with surgery first cases. That study included three regimens of preoperative chemotherapy, GEM plus S-1, GEM alone and GnP therapy. Therefore, we will clarify the long term outcomes of NAC with single regimen of GnP in the future.
In conclusion, NAC with two cycles of GnP for resectable PDAC was safe and feasible at dose level 1 (nab-PTX 75 mg/m2/day and GEM 600 mg/m2/day). In the future, it will be necessary to verify the long-term results of this regimen, and a phase II clinical trial based on the results of this study is anticipated.