Abstract
Background/Aim: To identify risk factors of early recurrence after neoadjuvant chemoradiation therapy (NACRT) and curative pancreatectomy in patients with borderline resectable (BR) pancreatic ductal adenocarcinoma (PDAC). Patients and Methods: Sixty-one patients with BR-PDAC who underwent curative resection after NACRT during July 2009-June 2014 were included. Patients were divided into early recurrence (i.e., developed recurrence within 1 year after pancreatectomy; n=30) and late/non-recurrence groups (n=31). The patient characteristics, clinicopathological factors of early recurrence, and survival time were retrospectively compared between groups. Results: In the univariate analysis, the maximum standardized uptake value (SUVmax), microvascular invasion, and lymph node metastasis were associated with early recurrence. In the multivariate analysis, the pre-NACRT SUVmax and microvascular invasion in the early recurrence group were significantly different from that in the late/non-recurrence group. A pre-NACRT SUVmax >4.1 was an independent predictor of poor recurrence-free and overall survival. Conclusion: SUVmax and microvascular invasion are independent predictors of poor recurrence-free and overall survival after NACRT for BR-PDAC. Although complete pancreatectomy after NACRT was performed, approximately half of the patients had recurrence within 1 year.
Despite significant advances in cancer diagnosis and treatment, pancreatic ductal adenocarcinoma (PDAC) still has a very poor prognosis. It is estimated that there will be 55,440 new cases and 44,330 deaths attributable to pancreatic cancer by 2018; approximately, 96% of these will be PDAC (1). Although margin-negative complete surgical resection remains the only treatment with curative intent with a 5-year survival rate of 20-25%, only 15% to 20% of patients have resectable disease at the time of diagnosis (2). Several advances in the management of PDAC have recently been reported. For instance, based on an expert consensus report, PDAC can be broadly categorized as resectable, borderline resectable (BR), and unresectable based on tumor encasement of arterial (celiac axis, superior mesenteric artery [SMA], and common hepatic artery [CHA]), or venous (superior mesenteric vein [SMV] and portal vein) structures (3). These definitions have facilitated a more uniform anatomical classification of PDAC and reliable cross-study comparisons.
BR-PDAC is frequently associated with positive surgical margins and poor prognosis after resection compared to resectable-PDAC because frequent positive surgical margins are expected with upfront resection (4). Therefore, neoadjuvant chemoradiation therapy (NACRT) strategies are increasingly being employed for BR-PDAC (5, 6). These strategies include early initiation of systemic therapy in these patients in contrast to a surgery-first approach where up to half of the patients may not receive adjuvant therapy due to postoperative complications or a decline in functional status. Theoretically, the neoadjuvant approach downstages nodal disease, increases the rate of margin negative resection, and helps identify patients at risk of early disease progression. Several researchers have reported favorable outcomes with NACRT followed by complete surgical resection for BR-PDAC compared to patients who did not undergo surgery (6, 7). However, 60% of patients experience local and systemic recurrence within the first year after curative surgery and survival outcomes are negatively influenced by early recurrence following pancreatic resection (8). Additionally, few studies have described the risk factors associated with early recurrence in patients with BR-PDAC after NACRT. Therefore, identification of prognostic factors associated with early recurrence in patients with BR-PDAC after NACRT is of paramount importance in order to improve patient outcomes. The present study aimed to identify risk factors of early recurrence and characterize treatment and outcomes of patients with BR-PDAC after NACRT and surgical resection.
Materials and Methods
Patients. The medical records of patients with BR-PDAC who underwent NACRT between July 2009 and June 2014 at the Department of Gastroenterological Surgery, Yokohama City University Hospital, Japan, were retrospectively reviewed. Clinical staging was determined by thin slice contrast enhanced computed tomography (CT), magnetic resonance imaging, endoscopic retrograde cholangiopancreatography, endoscopic ultrasound, and [18F]-2-deoxy-D-glucose positron emission tomography/CT (18F-FDG-PET/CT). According to the National Comprehensive Cancer Network (NCCN) Clinical Practice Guidelines 2009, BR-PDAC has been defined as “involvement of <180 degrees of the celiac trunk or SMA, involving only a short segment of the hepatic artery, and occluding only a short segment of the SMV, portal vein, or confluence”(3). Onodera's prognostic nutritional index (PNI) value was calculated using the following formula: 10×serum albumin (g/dl)+0.005×total lymphocyte count (per mm3). Modified Glasgow Prognostic Score (mGPS) was constructed as follows: patients with elevated CRP (>1.0 mg/dl) and decreased serum albumin (<3.5 g/dl) levels were assigned a score of 2. Patients with elevated CRP (>1.0 mg/dl) and albumin (≥3.5 g/dl) levels were assigned a score of 1 and those with a normal CRP level (≤1.0 mg/dl) were assigned a score of 0.
This study was approved by the Ethical Review Board Committee (B131107020) of Yokohama City University Hospital, Yokohama, Japan. Surgery was performed after all possible alternative procedures or treatments had been explained to the patients and they had provided their informed consent. The treatments were performed in accordance with the ethical standards described by the Declaration of Helsinki.
Procedures. In this study, the dosage of tegafur/gimeracil/oteracil (S-1) administered for NACRT was based on the results of a phase II study of gemcitabine+S-1 in which 1000 mg/m2 of gemcitabine was combined with a daily dose of 60 mg/m2 of S-1 (9). Patients received 3 cycles of gemcitabine intravenously, on days 8 and 15 of a 21-day cycle, and oral S-1, twice daily on days 1-14. Radiotherapy was administered concurrently with S-1 daily for 14 days; a total dose of 30 Gy in 10 fractions. Following NACRT, patients had their disease re-staged by thin slice contrast CT and 18F-FDG-PET/CT. Patients were examined by our multidisciplinary pancreatic cancer treatment team. If locoregional disease was stable or there was less or no distant metastatic disease, the patients underwent pancreatectomy with standard lymph node dissection within 6 weeks of NACRT. The perineural tissue around the major arteries was dissected by half of the circumference to clear the margin. All patients who underwent resection received gemcitabine or S-1 treatment for 6 months. If tumor progression was apparent after NACRT, the need for additional chemotherapy or chemoradiotherapy was determined on an individual basis. Pathology and operative reports were reviewed to evaluate the margin status and details about the resection. Curative resection was defined as resection with a tumor-free margin and no residual tumors in other organs (R0). Resection margins were defined as positive (R1) if malignant cells were found within 1 mm of the pancreatic resection margin, the plexus around SMA or CHA, bile duct, duodenum, or retroperitoneal tissue. If vein resection was performed, the vein margin was examined by the pathologist. The histologic grade of NACRT response was evaluated by the Evans grading system (I, 0-10% tumor cells destroyed; IIa, 10-50% of tumor cells destroyed; IIb, 50-90% of tumor cells destroyed; III, 90% of tumor cells destroyed; and IV, no viable tumor cells) (10).
Early recurrence was defined as recurrence within 1 year after operation. The early recurrence patients were distinguished from the late/non-recurrence patients, which included patients who developed recurrence more than 1 year after resection, and patients who did not develop recurrence. The medical records of all patients were reviewed retrospectively and follow-up data for contrast CT were evaluated every 3 months. The laboratory results, including tumor markers, tumor pathology, and recurrence, were collected for each patient. The first site of disease recurrence was defined as follows: locoregional recurrence, development of a new low-density mass in the region of the pancreatic bed and root of the mesentery; distant metastasis, new low-density region in the liver or lung; and peritoneal dissemination, new ascites on CT and subsequently confirmed by cytological examination. Recurrence-free survival (RFS) was calculated as the time from the date of surgery to the date of initial recurrence. Overall survival (OS) was calculated as the time from the date of initial treatment to the date of death. The length of the tumor was estimated based on the contrast CT image before treatment and on the resected specimen.
Statistics. Continuous variables were expressed as means with standard deviations and compared using the independent Student's t-test. Categorical variables were expressed as numbers and percentages and compared using the Chi-squared or Fisher's exact test. A receiver operating characteristics (ROC) curve was constructed to determine the optimal cutoff value of tumor markers and median maximum standardized uptake value (SUVmax). ROC curve analyses were conducted for RFS. Univariate and multivariate logistic regression models were used to determine factors associated with early recurrence following initial surgery. Relative risks were expressed as hazard ratios (HR) and 95% confidence intervals (CI). Variables with a p-value <0.1 by univariate analysis were included in the multivariate model. Two-tailed p-values <0.05 were considered statistically significant. Statistical analyses were performed using SPSS 22.0 (Chicago, IL, USA).
Results
Patient characteristics. Of the 85 patients with BR-PDAC who underwent NACRT, 84 patients completed NACRT (Figure 1). The median follow-up times for all patients were 19.0 months (range=3.2-85.5 months). Approximately, 24.7% (21/85) of patients did not undergo pancreatectomy because they had progressive disease (PD) or intraoperative detection of distant metastasis, or that was the patient's request. The adverse events related to NACRT did not prevent any of the patients from completing treatment or lead to any subsequent surgical morbidity. There were no cases of liver or renal failure and there were no deaths related to NACRT or surgery. Approximately 76% (64/85) of patients underwent pancreatectomy, and the R0 resection rate was 95.3% (61/64). All patients who were admitted for surgery were discharged home. During follow-up, early recurrence was noted in 30 (49.2%) patients. Of the remaining patients, 7 (11.5%) had late recurrence and 24 (39.3%) patients did not have any recurrence. There were 35 (57.4%) patients who died of disease progression, 2 (3.3%) patients who were alive with disease; and 16 (26.2%) patients who were alive with no evidence of disease at the end of the study. The median OS was 19 months, and the cumulative incidence of recurrence at 1, 3, and 5 years was 50.7%, 60.4%, and 75.1%, respectively (Figure 2A and B). Furthermore, the mortality rate of patients who were recognized as having both an SUVmax >4.1 and microvascular invasion was significantly higher than that of patients without these 2 factors (91.3% vs. 30.0%; p=0.01).
Flow-chart of the selection of patients included in the analysis.
Kaplan–Meier curves for recurrence-free survival (A) and overall survival (B). The median recurrence-free survival and overall survival of all patients were 12 months and 19 months, respectively.
Patient characteristics.
Risk factors associated with early recurrence. Regarding preoperative risk factors, there were no significant differences in NCCN resectability, dose intensity, and tumor markers, including carbohydrate antigen 19-9 (CA19-9), between the 2 groups (Table I). Early recurrence occurred more commonly in patients with a significantly higher pre- and post-NACRT SUVmax than those who were recurrence-free for 1 year after curative pancreatectomy followed by NACRT (6.84 vs. 5.14, p=0.039 and 5.01 vs. 3.97, p=0.021, respectively). However, the reduction in the SUVmax was not different between the early and late recurrence groups. Having microvascular invasion (p=0.003), portal vein invasion (p=0.043), and positive lymph nodes (p=0.015) were significant early recurrence risk factors (Table II). Patients who were recurrence-free for 1 year were likely to benefit more from NACRT, as evaluated by the Evans grading system, than those who had early recurrence. The adjuvant chemotherapy induction rates were almost similar in both groups (73.3% and 80.6% of early and late/non-recurrence patients, respectively; Table III). The distribution of all anatomic sites and locations of tumor recurrence in the early recurrence group were as follows: liver (43.3%), peritoneal carcinomatosis (20%), lungs (20%), locoregional (10%), and lymph nodes (6.7%). Liver metastasis was likely to occur earlier than metastasis at other sites (Table IV). Based on the ROC curves for RFS for 1 year, with an area under the curve of 0.662, we identified an SUVmax of 4.1 pre-NACRT as the optimal cutoff value (sensibility, 86.7%; specificity, 43.3%). On multivariate analysis, pre-NACRT SUVmax >4.1 (HR=2.8; 95%CI=1.2-6.4; p=0.017) and microvascular invasion (HR=3.0; 95%CI=1.5-5.8; p=0.002) were independent risk factors associated with early recurrence following curative pancreatectomy after NACRT for BR-PDAC (Table V). Furthermore, in the Kaplan–Meier analysis, having an SUVmax >4.1 (Figure 3A, B; RFS, 8 vs. 41 months, p=0.004 and OS, 15 months vs. not reached, p=0.006) and microvascular invasion (Figure 3C, D; RFS, 8 months vs. not reached, p<0.001 and OS, 15 vs. 33 months, p=0.011) were significant factors for poor RFS and OS.
Histological outcome.
Adjuvant chemotherapy
Recurrence pattern.
Discussion
Although surgical resection of BR-PDAC represents a potentially curative treatment, the high recurrence rate after pancreatectomy results in a poor prognosis (4-6). Therefore, understanding and predicting recurrence is important for improving the prognosis of BR-PDAC. Based on our findings, the 18F-FDG-PET/CT SUVmax and microvascular invasion could be utilized in predicting early tumor recurrence for BR-PDAC patients after NACRT. The potential role of FDG uptake values in the prediction of prognosis has been recently reported in several meta-analyses. High SUV values at diagnosis are more highly associated with poor survival than low SUV values in a variety of cancers, such as head and neck cancer, hepatocellular carcinoma, and bone and soft tissue sarcoma (11-13). To date, 18F-FDG-PET/CT has been considered the imaging modality that is most commonly used for diagnosis, staging, evaluating response to treatment, and detecting postoperative recurrence and metastasis in PDAC. Furthermore, some researchers have also reported the prognostic value of preoperative 18F-FDG-PET/CT patients with SUVmax cutoff values of 3.3-7.0 after resection, including locally advanced PDAC (14-16). Other 18F-FDG-PET/CT parameters, such as metabolic tumor volume and total lesion glycolysis, which reflect both metabolic activity and tumor volume, have been shown to be independent preoperative risk factors for OS and RFS (16). However, the predictive value of 18F-FDG-PET/CT in patients with BR-PDAC after NACRT has not been previously studied. In evaluating the predictive value of SUVmax in these patients, we found that a cutoff value of 4.1 was significantly associated with 1-year recurrence and survival rates after tumor recurrence, with a SUVmax >4.1 being a predictor of poor survival after recurrence. These findings suggest that 18F-FDG-PET/CT might be a tool to help determine which patients require more aggressive neoadjuvant therapy before surgery.
Univariate and multivariate analyses of predictors of RFS
Although an elevation of preoperative or postoperative tumor markers has been reported to be a predictor of poor survival, there were no significant preoperative and postoperative differences in tumor marker levels, including CA19-9, between the early recurrence and late/non-recurrence groups in our study (17-19). Nevertheless, a significant reduction in CA19-9 was evident after NACRT, and there was no improvement in the disease stage. Moreover, upon evaluation of the effect of NACRT according to the Evans grading system, the percentage of patients with a grade >IIa was 90.2% after NACRT (Evans I, 6 patients; IIa, 25 patients; IIb, 24 patients; and III, 6 patients), which is consistent with several studies that reported an NACRT response rate of approximately 89.4% (20). Although the patients in the early-recurrence group were likely to have fewer anticancer effects of NACRT than those in the late/non-recurrence group, as evaluated by the Evans grading system, there was no difference in the reduction of CA19-9 levels between these 2 groups. Therefore, there is a discrepancy between the significant reduction in CA19-9 levels and tumor regression rates.
Kaplan–Meier curves for recurrence-free survival (RFS) and overall survival (OS) based on the maximum standard uptake value (SUVmax) and microvascular invasion. Patients with an SUVmax >4.1 had a significantly poorer median RFS and OS compared to those with an SUVmax ≤4.1 (A and B; RFS, 8 vs. 41 months, p=0.004; OS, 15 months vs. not reached, p=0.006). Patients with microvascular invasion (v+) had a significantly poorer median RFS and OS compared to those with no microvascular invasion (v-) (C and D; RFS, 8 months vs. not reached, p<0.001; OS, 15 vs. 33 months, p=0.011).
Currently, the standard of care for BR-PDAC is NACRT followed by surgery. NACRT results in greater survival and time to local recurrence compared with upfront resection. Moreover, there was a lower rate of lymphovascular invasion and smaller tumor size in these patients. Given the potential benefits, NACRT can be useful in patients with unfavorable prognostic features during the preoperative stage. However, PDAC maintains a poor prognosis due to an early appearance of metastases (21). These outcomes support a systemic approach of PDAC from an early stage. Therefore, accurate preoperative diagnosis and staging for PDAC are crucial not only to detect local invasion and distant metastases when assessing resectability, but also to select patients who might benefit more from NACRT instead of upfront surgery (i.e., those with more aggressive tumor biology and at higher risk of early recurrence after surgery) (21).
While the main goal of any neoadjuvant strategy is to prolong survival, another goal is to improve R0 resection because positive resection margins have been shown to be associated with poor survival (22) (23). Based on previous trials, the R1 resection rate in adjuvant therapy has been reported as 24-42% (24, 25), whereas the margin positive rate in neoadjuvant therapy has been <10% in selected centers (5). In the current study, >90% of patients had R0 resection after NACRT, which suggests that neoadjuvant therapy increases the likelihood of R0 resection among BR-PDAC.
Although R0 resection for BR-PDAC after NACRT was performed, approximately half of the patients had recurrence within 1 year. Liver metastases were the most common type of tumor relapse in these patients. These findings can be influenced by the possible presence of occult metastases at the time of operation, which could be interpreted as early recurrence. Some researchers report that <58% of patients with locally advanced PDAC have occult distant metastases at the time of operation (26, 27). In this study, almost all patients with early recurrence were diagnosed with distant metastases. Therefore, there is a chance that some of these “recurrences” could have already been present at the time of surgery. Additionally, an increased R0 resection rate and local control in advanced PDAC patients did not translate into significantly improved survival in recent meta-analyses (28). The MD Anderson Cancer Center group found that 45% of patients had recurrence in distant organs such as lung, liver, or bone (6), indicating the need for effective systemic therapy. These results and our current findings suggest that in addition to improved R0 rates, effective systemic neoadjuvant and adjuvant treatment is also important for improving survival. In this sense, it is possible that newer neoadjuvant chemotherapeutic regimens (gemcitabine-nab-paclitaxel and FOLFIRINOX), which are more potent than gemcitabine+S-1, could improve overall oncologic outcomes. In fact, several phase II clinical trials have been initiated to evaluate the potential efficacy of neoadjuvant treatment with gemcitabine-nab-paclitaxel and FOLFIRINOX in BR-PDAC patients (29, 30). No improvement in RFS with adjuvant therapy was detected in BR-PDAC patients after NACRT. This could be the result of administration of S-1 to almost all patients who received adjuvant therapy, which may have reduced the efficacy of this treatment. Additionally, predictive markers may be useful when deciding which BR-PDAC patients may benefit from adjuvant chemotherapy following complete resection and which chemotherapeutic agents may be used in individual patients to maximize survival. Both SUVmax >4.1 and microvascular invasion were strongly correlated with poor prognosis in this study. Therefore, a more powerful regimen, such as gemcitabine-nab-paclitaxel and modified FOLFIRINOX, should be provided to these patients to improve their prognosis.
This study has several limitations. Certainly, the retrospective nature of our cohort study and the small sample size limited the possibility of inferring robust conclusions. However, patient selection, diagnostic workup, and surgical treatment were conducted in the same institution by an established multidisciplinary team. Moreover, although ROC curves showed a SUVmax ≥4.1 as the best cutoff value to predict RFS and OS in the present study, no consensus has yet been established in the literature about the optimal SUVmax cutoff (15). This is a reasonable concern for the use of 18F-FDG-PET/CT, and further investigations in a larger sample size are needed to achieve a standardization of the SUVmax cutoff value.
In conclusion, an SUVmax >4.1 and microvascular invasion are independent predictors of poor RFS and OS in patients with BR-PDAC after NACRT. NACRT for BR-PDAC can help achieve a clear margin with good local control. Although R0 resection for BR-PC after NACRT was performed, approximately half of the patients had recurrence within 1 year. Therefore, more powerful and effective chemotherapy is needed to reduce distant metastasis, that would improve patient outcomes and reduce healthcare expenses.
Footnotes
Authors' Contributions
Nobuhiro Tsuchiya designed the study, and wrote the initial draft of the manuscript with support from Itaru Endo. All other authors have contributed to data collection and approved the final version of the manuscript, and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Conflicts of Interest
The Authors declare no conflict of interest regarding this study.
- Received June 24, 2019.
- Revision received July 1, 2019.
- Accepted July 2, 2019.
- Copyright© 2019, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
