Abstract
Background/Aim: Pancreatic adenocarcinoma (PDAC) with synchronous oligometastases may indicate a surgical benefit after chemotherapy. We investigated whether primary and metastatic resection of PDAC with oligometastases can improve the survival and then explored prognostic factors to identify indications for conversion surgery. Patients and Methods: We reviewed 425 patients with PDAC who underwent pancreatic resection from 2005 to 2019. Clinical characteristics and outcomes were analyzed. Two-stage resection was defined as preceding metastasectomy and subsequent primary resection after chemotherapy. Results: Fifteen patients (3.5%) had synchronous oligometastases. We evaluated the overall survival of the patients with oligometastases and those without metastases. The survival curves almost completely overlapped (median survival time: 35.9 vs. 32.1 months). The univariate Cox regression analysis revealed a normal level of preoperative CA19-9 (p=0.075), two-stage resection (p=0.072), and R0 resection (p=0.064) were likely promising prognostic factors. The combination of a normal level of preoperative CA19-9 with two-stage resection was a significant prognostic factor (p=0.038). In addition, patients with a normal preoperative CA19-9 level and two-stage resection had better survival (46.1 vs. 28.1 months, p=0.026). Conclusion: The combination of normal preoperative CA19-9 with two-stage resection can be a useful way to identify patients with PDAC and oligometastases for surgical indication.
Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest gastrointestinal malignant diseases and its incidence is increasing worldwide. In the United States, it is the fourth-leading cause of cancer-related mortality (1). Although multi-diagnostic modalities have been developed, identifying it at an early stage is still difficult. Unfortunately, at the time of diagnosis, 80% of such patients are in advanced stage and 50% of them have distant metastasis (2). Although new chemotherapy regimens, such as the combination of gemcitabine and nab-paclitaxel (GnP) and oxaliplatin, 5-fluorouracil, leucovorin, and irinotecan (FOLFIRINOX) have been introduced (3, 4), the prognosis of PDAC is still poor and the 5-year survival is less than 10% (1).
Oligometastases is defined as a limited number of distant metastases. It has been proposed that oligometastases is an intermediate state between localized and polymetastatic disease, and represents a relatively better survival compared to polymetastatic disease (5, 6). In other types of cancers, such as colorectal cancer, breast cancer, and gastric cancer, a benefit of primary tumor and metastasis resection has been reported (7-10). Recently, two retrospective studies showed that primary tumor and metastasis resection improved survival in patients with PDAC compared with patients who received palliative treatment (11, 12). However, the usefulness of surgery for oligometastases in PDAC remains controversial due to limited data. Therefore, current guidelines do not recommend primary tumor and metastasis resection for patients with PDAC without perioperative chemotherapy (13, 14).
Since more effective chemotherapy regimens for PDAC have been developed, a new approach to PDAC with oligometastases currently has come into focus. A few patients can receive a remarkable response to chemotherapy and be converted to resection, which is called conversion surgery (CS) (15, 16). Previous studies have mentioned that highly selective patients with better tumor biology can receive the survival benefit of CS (17-19). However, how to identify the patients with better tumor biology has not been determined. In this study, we first investigated whether primary and metastatic resection for PDAC with oligometastases can improve the survival, and then, explored prognostic factors to stratify patients in order to identify promising indications for CS.
Patients and Methods
Patients. We reviewed 425 patients with PDAC who underwent pancreatic resection from 2005 to 2019 (Figure 1). All the patients enrolled were diagnosed with PDAC histologically. There were no specific exclusion or inclusion criteria. The study protocol was approved and reviewed by the Ethics Committee of Tohoku University Hospital (approval number: 2022-1-245). As this was a retrospective study, informed consent was waived.
Study flow chart. PDAC: Pancreatic adenocarcinoma; OS: overall survival; RFS: recurrence-free survival.
Assessment and definitions. From an electronic database collected retrospectively in our institution, the preoperative characteristics and postoperative outcomes including age, sex, Eastern Cooperative Oncology Group (ECOG) performance status, tumor variables, chemotherapy regimens received, recurrence-free survival (RFS: time from surgery to tumor recurrence) and overall survival time (OS: time from initial diagnosis to death) were retrospectively analyzed. Oligometastases were defined as ≤3 metastatic lesions based on a previous study (20). The diagnosis of oligometastases was made by thin-slice contrast-enhanced computed tomography (CT), magnetic resonance imaging (MRI) or (18F)-2-deoxy-D-glucose positron emission tomography/CT (PET/CT) at the time of the initial diagnosis. As well as imaging diagnosis, we diagnosed the patients as radiographically undiagnosed for oligometastases during surgery, which is defined as occult oligometastases. The assessment of efficacy for chemotherapy was carried out with CT scan based on the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 (21). Two-stage resection was defined as that of a patient who received metastasis resection and primary resection at different times. In such patients, metastatic resection was performed firstly and, if the disease was under control and no new metastatic lesion was detected, pancreatic resection was performed as the second surgery. We performed two-stage resection for the patients with occult oligometastases. CS was defined as surgical resection after chemotherapy in PDAC, which was initially diagnosed as unresectable. The pretreatment CA19-9 was usually assessed at the first contact. If the patient had obstructive jaundice, we evaluated CA19-9 after biliary drainage and within 2.0 mg/dl of total bilirubin value. Patients with CA19-9 under 37 U/ml were considered as CA19-9 normal. In the patients with two-stage resection, the preoperative CA19-9 was assessed before the curative resection of the primary tumor. The evaluation of the histological response to chemotherapy was performed by the Evans grade system (I: less than 10% or no tumor cells destroyed; IIa: destruction of 10-50% tumor cells; IIb: destruction of 51-90% tumor cells; more than 90% cells destroyed; IV: no viable tumor cells).
Treatment pattern for patients with PDAC and oligometastases. In most cases, preoperative chemotherapy was administered, but a few cases who had been treated before GnP or FOLFIRINOX was developed, underwent simultaneous resection of the primary lesion and metastases without preoperative chemotherapy. In cases with radiographically diagnosed oligometastases at the initial diagnosis, systemic chemotherapy was administered. CS was performed according to the criteria as follows: 1. Stable disease or tumor shrinkage after chemotherapy. 2. No new metastasis. 3. Good performance status [Eastern Cooperative Oncology Group (ECOG) performance status: 0 or 1]. In cases with occult oligometastases detected during open laparotomy or staging laparoscopy, all metastatic lesions, except deep portions of the liver, were removed as total biopsy. Then, according to the criteria for CS, two-stage resection was performed. The final decision for CS was determined by the surgical team. In principle, adjuvant chemotherapy was given, except for patients with a poor physical condition due to postoperative complications.
Statistical analysis. Univariate Cox regression analysis was employed to evaluate individual prognostic factors. For comparison of the survival curves using the Kaplan–Meier method, a log-rank test was performed. When the data of death were not recorded, patients were included but censored at the last contact. We followed up the data until December 31, 2022. Statistical significance was defined as p<0.05. Statistical analyses were carried out with JMP software (SAS Institute Inc., Cary, NC, USA).
Results
Patient characteristics. The clinicopathological characteristics of the patients who underwent metastasectomy for oligometastases are shown in Table I and Table II. A total of 15 patients (3.5%) had oligometastases and metastatic sites were as follows. Liver: 7 (46.7%); Peritoneal: 8 (53.3%). Of the 15 patients, 9 patients (60.0%) had solitary metastasis, five patients (33.3%) had two metastases, and one patient (6.7%) had three metastases. Six patients (40.0%) had a tumor in the pancreas head region, 9 patients (60.0%) in the pancreas body and tail. Four patients (26.7%) were diagnosed by image findings, while 11 patients had (73.3%) occult oligometastases. Preoperative chemotherapy was conducted in 12 patients (80%) and the median duration was 5.1 months. Eleven patients (73.3%) received systemic chemotherapy and one patient (6.7%) was given chemoradiotherapy. Detailed information of the preoperative treatments are shown in Table II. Nine patients received only first line regimen, whereas three patients received two or three kinds of regimens. In terms of the radiological response, a partial response (PR) and stable disease (SD) were found in seven (58.3%) and five (41.7%), respectively. No patient showed a complete response (CR). The median CA19-9 level at the initial diagnosis was 94.3 U/ml (0.6-4,704) and the median CA19-9 level before surgery was 30 U/ml (0.6-2,242). Six patients (40.0%) showed a normal level of CA19-9 at the initial diagnosis. A normal level of CA19-9 after chemotherapy was observed in seven patients (46.7%).
Patients’ characteristics with oligometastases.
Individual outcomes of the patients with oligometastases who underwent pancreatic resection.
Surgical and pathological outcomes. Subtotal stomach-preserving pancreaticoduodenectomy (SSPPD) and distal pancreatectomy (DP) were performed in six patients (40.0%) and nine patients (60.0%), respectively. Portal vein resection (PVR) was performed in four patients (26.7%), who included three with SSPPD and one with DP. DP with en bloc celiac axis resection (DP-CAR) was performed in four patients (26.7%). Two-stage resection was conducted in seven patients (46.7%). The median operative time was 507 min (322-708 min) and the median blood loss was 1,026 ml (348-5,980 ml). Although postoperative complications of the Clavien-Dindo (C-D) classification ≥IIIa were observed in seven patients (46.7%), none had died. The median duration of the postoperative hospital stay was 23 days (15-137 days).
R0 resection was histologically achieved in nine patients (60.0%). Evans grade IIa was observed in eight patients (66.7%), IIb in two patients (16.7%), and III in two patients (16.7%), respectively. No patient showed grade IV in this study. In terms of the pathological T status based on UICC 8th, T1 was found in two patients (13.3%) and T3 in 13 patients (86.7%). Pathological lymph node metastasis was observed in 11 patients (73.3%). Thirteen patients (86.7%) received adjuvant chemotherapy, the regimen of which was gemcitabine, S-1 and FFX in two, ten and one patient, respectively.
Comparison of OS between the patients with oligometastases and those without metastases. To evaluate the usefulness of surgical resection for oligometastases, we first evaluated the OS from the initial diagnosis between the patients with oligometastases and those without metastasis (Figure 2). The clinicopathological characteristics of the patients without metastasis are shown in Table III. The median observation time was 35.4 months (2.6-136 months). The survival curve almost completely overlapped between that with and that without metastasis. The median OS of patients without metastasis was 35.9 months and of those with oligometastases was 32.1 months.
Comparison of overall survival (OS) between the pancreatic adenocarcinoma (PDAC) patients with and without oligometastases. The median OS in two groups were 35.9 and 32.1 months, respectively.
Patients’ characteristics without metastasis.
Cox regression model analysis for the OS. To investigate the prognostic factors in patients with oligometastases who underwent surgical resection, univariate analysis with cox regression model was performed in various clinicopathological factors (Table IV). The normal level of preoperative CA19-9 [hazard ratio (HR)=0.35, 95%confidence interval (CI)=0.11-1.11, p=0.075], two-stage resection (HR=0.34, 95%CI=0.10-1.15, p=0.072), and R0 resection (HR=0.22, 95%CI=0.04-1.09, p=0.064) were likely promising prognostic factors. In terms of RFS, R0 resection was a significant prognostic factor (HR=0.10, 95%CI=0.02-0.60, p=0.012). Two-stage resection and a normal level of preoperative CA19-9 were likely to improve RFS as well as OS. We next compared the OS based on preoperative CA19-9, two-stage resection, and R0 resection (Figure 3). Comparison of the median OS between the patients with and those without normal preoperative CA19-9, two-stage resection and R0 resection were as follows. preoperative CA19-9: 46.1 vs. 22.6 months, p=0.064; two-stage resection: 46.1 vs. 29.9 months, p=0.069; R0 resection: 46.1 vs. 28.0 months, p=0.056.
Cox regression analysis for OS and RFS in pancreatic cancer patients with oligometastases.
Comparison of overall survival (OS) and recurrence-free survival (RFS) based on normal CA19-9, two-stage resection, and R0 resection. (A) OS comparison, normal CA19-9 vs. CA19-9 over 37 U/ml, median OS: 46.1 vs. 22.6 months. In the patients with two-stage resection, the preoperative CA19-9 was assessed before the curative resection of the primary tumor. (B) RFS comparison; normal CA19-9 vs. CA19-9 over 37 U/ml, median RFS: 21.8 vs. 6.6 months. In the patients with two-stage resection, the preoperative CA19-9 was assessed before the curative resection of the primary tumor. (C) OS comparison, two-stage resection vs. simultaneous resection, median OS: 46.1 vs. 29.9 months. (D) RFS comparison, two-stage resection vs. simultaneous resection, median RFS: 22.9 vs. 8.1 months. (E) OS comparison, R0 vs. R1 or R2, median OS: 45.0 vs. 27.8 months. (F) RFS comparison, R0 vs. R1 or R2, median RFS: 16.9 vs. 2.7 months.
OS and RFS in the patients with both preoperative normal CA19-9 and two-stage resection. Among these potential prognostic factors, we focused on preoperative CA19-9 and two-stage resection since they are important as a key factor for the selection of patients with oligometastases who can receive a surgical benefit. To this end, we investigated whether patients with a normal preoperative CA19-9 level and two-stage resection had better survival. Cox regression analysis showed that a normal preoperative CA19-9 level with two-stage resection was a significant prognostic factor (HR=0.24, 95%CI=0.06-0.92, p=0.038) (Table V). In addition, patients with a normal preoperative CA19-9 level and two-stage resection had better survival (p=0.026) (Figure 4). The median OS of the patients with normal preoperative CA19-9 level and two-stage resection was 46.1 months and the median OS of other patients was 28.1 months. Furthermore, the RFS of the patients with a normal preoperative CA19-9 level and two-stage resection was better than that of other patients (27.8 vs. 7.8 months, p=0.028).
Cox regression analysis for OS and RFS in pancreatic cancer patients with and without two-stage resection and normal preoperative CA19-9.
Comparison of overall survival (OS) and recurrence-free survival (RFS) between the patients with and without normal CA19-9 and two-stage resection. (A) OS comparison, normal CA19-9 vs. CA19-9 over 37 U/ml, median OS: 75.0 vs. 28.1 months. (B) RFS comparison, normal CA19-9 vs. CA19-9 over 37 U/ml, median RFS: 27.8 vs. 7.8 months. In the patients with two-stage resection, the preoperative CA19-9 was assessed before the curative resection of the primary tumor.
Discussion
In the current study, we first investigated whether surgery for patients with PDAC and oligometastases can improve survival. We found that OS of surgical resection for PDAC with oligometastases is equivalent to that of surgical resection for PDAC without oligometastases. Second, we revealed that a normal preoperative CA19-9, two-stage resection, and R0 resection were promising prognostic factors. Among these factors, we further focused on a normal preoperative CA19-9 level and two-stage resection to reveal good preoperative indicators to determine the surgical indication of CS for PDAC with oligometastases. To this end, we found that the patients with a normal preoperative CA19-9 level and two-stage resection showed a significantly better OS. Therefore, a normalized preoperative CA19-9 level and two-stage resection can be a useful CS indication. To the best of our knowledge, this is the first study showing the usefulness of two-stage resection in the field of PDAC with oligometastases.
Recently, two retrospective studies about CS in metastasized PDAC have been reported. Wright et al. analyzed 1,147 patients with metastatic PDAC treated in the Johns Hopkin Hospital and the University of Pittsburg (22). Twenty-three cases (2.0%) who showed a favorable response to chemotherapy received CS and the median OS from diagnosis was 34.1 months. In their study, FOLFIRINOX or a gemcitabine-based regimen was given and the median cycles of chemotherapy were nine. Frigerio et al. also reported the CS for the patients with PDAC and liver metastases (12). In their study, the indications of CS were the disappearance of liver metastases and a normalized or remarkable decrease in CA19-9. Mainly, FOLFIRINOX was given and the median interval of chemotherapy was 9.7 months. The patients who could meet the indication above were 24 of 535 patients with liver metastases (4.5%). They conducted resection of the primary lesion and liver resection, if the liver metastasis was still suspicious during the operation. The median OS from the diagnosis was 56 months. Since the median survival of patients with metastatic PDAC who received chemotherapy has been reported to range from 8.5 to 11.2 months (4, 23, 24), CS is suggested to improve survival. In our study, the median OS was 32.1 months, which is equivalent to the OS reported before, and the survival curve almost completely overlapped between that with and that without metastasis. Accordingly, selected patients with PDAC and oligometastases can benefit from CS.
Although the OS of the patients who underwent CS seemed to be improved, the population who can undergo CS in the metastasized PDAC is less than 5% (18). The challenge right now is to identify the best indicator for CS and how to select the patients with better tumor biology. So far, studies focusing on CS for metastasized PDAC are limited and no prognostic factors have been identified. Therefore, we next accessed prognostic factors for patients with PDAC and oligometastases to stratify the patients and identify the patient groups who can benefit from multi-modal treatment. From the univariate cox regression model analysis, we found that a normal level of CA19-9 before surgery, two-stage resection, and R0 resection were promising prognostic factors. Although a normal level of CA19-9 was not significant in our study, it is well known that CA19-9 is a prognostic factor in PDAC (25). In addition, the usefulness of CA19-9 for the indication of CS has been reported in locally advanced PDAC (26, 27).
This is the first study to report the usefulness of two-stage resection in PDAC. In two-stage resection, we can conduct chemotherapy after metastasis resection and evaluate whether the disease is under control. Then, we can consider pancreatic resection. Therefore, two-stage resection can be useful for the decision to perform CS. Generally, two-stage resection is employed to reduce the invasiveness of surgery in high risk patients and complicated procedures such as multiple primary cancers (28, 29). In addition, two-stage hepatectomy for bilateral colorectal liver metastases has been reported to be useful for the reduction of surgical invasiveness (30). On the other hand, our two-stage resection strategy in CS for PDAC with oligometastases is a way to select candidates for pancreatic resection. Furthermore, the advantage of the two-stage resection strategy is to remove metastatic lesions before chemotherapy.
Given that it is difficult to recognize metastatic lesions after tumor shrinkage due to chemotherapy, metastasis resection before chemotherapy would be reasonable. Patients with PDAC often have occult metastases. Staging laparoscopy (SL) is reportedly helpful for the diagnosis of occult metastasis in patients with PDAC (31). When we detect occult oligometastases during SL, we try to resect all of the metastatic lesions, except deep lesions of the liver, to verify the malignancy histologically. In these cases, we plan the pancreatic resection for a two-stage resection strategy if chemotherapy can control the disease.
Regarding the chemotherapy regimens for PDAC with oligometastases, FFX or GnP are favorable regimens so far, but, the latest study reported that liposomal irinotecan, 5-fluorouracil, leucovorin, and oxaliplatin (NALIRIFOX) showed significant improvement in OS and RFS compared to GnP in treatment-naïve metastasized PDAC (32). Therefore, NALIRIFOX could be added in favorable regimens. The optimal duration for chemotherapy in CS for PDAC with oligometastases is under discussion. Satoi et al. showed that patients with locally advanced PDAC who received chemotherapy more than eight months before CS had better survival (33). Based on this result, several institutions in Japan have tended to conduct chemotherapy for at least eight months (34, 35). Gemenetzis et al. reported favorable survival in locally advanced PDAC resected after preoperative chemotherapy (36). In this study, they considered CS after more than four months of chemotherapy. In contrast, Michelakos et al. analyzed 110 patients with borderline resectable and unresectable PDAC who received FOLFIRINOX before surgery, and they reported that patients who had more than an eight month interval between diagnosis and surgery had a worse prognosis (37). On the other hand, the median duration of preoperative chemotherapy in patients who underwent CS for metastasized PDAC has been reported to be 6.2 to 10 months (6, 11, 12, 22, 38). Therefore, the optimal regimen and duration of chemotherapy are still unclear and further investigations are needed.
This study has some limitations. First, our study was in a single institute, retrospective, and had a small sample size. Secondly, regarding that only selected patients underwent CS, a selection bias existed. In addition, the conversion rate is unclear. As we do not have the survival data of all the patients diagnosed as PDAC with oligometastases, we could not evaluate the CS rate. Thirdly, as chemotherapeutic strategy for PDAC has advanced in the last decade, the patients of this study had received different chemotherapy regimens. This difference due to historical background may affect the prognosis. Although high level evidence has not been made in PDAC with oligometastases, a few prospective randomized clinical trials to verify whether CS can improve the OS of patients with PDAC and liver metastasis are ongoing (20, 39).
Conclusion
Our results suggest that the combination of normal CA19-9 with two-stage resection can be a useful way to select patients with better tumor biology and is associated with a better prognosis in PDAC oligometastases.
Acknowledgements
The Authors would like to thank Brent Bell for the English language review.
Footnotes
Authors’ Contributions
KI: Conceptualization, Data curation; MM, FM: study design, data collection, Supervision; TK, HS, AK, SA MI, DD, TM, SM, MI, HO, KN, TK, MI: Conceptualization, Data curation.
Conflicts of Interest
Michiaki Unno receives honoraria and research funding from Taiho Pharmaceutical Co., Ltd. Fuyuhiko Motoi receives research funding from Taiho Pharmaceutical Co., Ltd. Other co-authors have no conflicts of interest.
Funding
This work was supported by Japan Society for the Promotion of Science Grant-in-Aid for Early-Career Scientists (22K16503) (KI).
- Received September 10, 2023.
- Revision received October 2, 2023.
- Accepted October 3, 2023.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.