Pancreatic cancer is the fourth leading cause of cancer death in Europe and is expected to rank second in cancer mortality by 2030 [1]. The only potentially curative treatment option is surgical resection. The overall 5-year survival rate is only 10% [2]. The aggressive tumor biology has led to the introduction of new, more effective chemotherapeutic regimens over the past 10 years, both adjuvant and neoadjuvant, leading to the establishment of multimodal therapy concepts.
Indication for Surgery
On the initiative of the German Society for General and Visceral Surgery (DGAV), evidence-based recommendations for the indication for surgery of pancreatic cancer have been defined, with the indication to be made by a tumor board of experienced pancreatic surgeons in accordance with guidelines, taking into account individual patient characteristics [3]. According to the recommendations, which are based on a systematic analysis of 58 original papers and 10 guidelines, there is an indication for surgery in the case of histologically confirmed pancreatic cancer as well as in the case of a high suspicion of resectable pancreatic cancer [3, 4].
Resectability
The greatest chance of survival is with resection in healthy tissue, the R0 resection [5, 6]. Current guidelines now divide the R0 classification into "R0 narrow" (≤ 1 mm) and "R0 wide" (> 1 mm), depending on whether the carcinoma is less than or more than one millimeter from the resection margin [7]. In addition to anatomical resectability (relationship between tumor and major visceral vessels), since 2017, tumor biology and the general condition of the patient have been considered as co-determining resectability criteria and have been included in the current S3 guidelines as the ABC consensus classification of resectability [8].
Currently, resectability is assessed according to the so-called ABC criteria of resectability according to the International Association of Pancreatology (IAP) consensus:
Resectability | A (anatomical) | B (biological) | C (conditional) |
Resectable | R-Type A | Neg: R-Type A | Neg: R-Type A |
Pos: BR-Type B | Pos: BR-Type C | ||
Borderline resectable (BR, borderline resectable) | BR-Type A | Neg: BR-Type A | Neg: BR-Type A |
Pos: BR-Type AB | Pos: BR-Type AC | ||
Locally advanced (LA, locally advanced) | LA-Type A | Neg: LA-Type A | Neg: LA-Type A |
Pos: LA-Type AB | Pos: LA-Type AC | ||
Abbreviations: A: "anatomical": relationship to vessels B: "biological": CA19-9 > 500 IU/ml or regional lymph nodes involved (PET-CT or biopsy) C: "conditional": ECOG Performance Status 2 or higher Neg: negative for the above parameters Pos: positive for the above parameters Other combinations possible: e.g. BR-BC, BR-ABC, LA-ABC etc. |
To assess anatomical resectability, the S3 guidelines recommend a contrast-enhanced multiphasic computed tomography [7]. Based on the anatomical resectability criteria, a tumor can be classified as primarily resectable, borderline resectable, and non-resectable or locally advanced [7].
The assessment of biological resectability is most often based on the tumor marker CA 19-9. The threshold was defined as > 500 IU/ml, as above this value, resectability is given in only less than 70% of cases, and survival of less than 20 months is expected [8, 9].
As another criterion, the ECOG Performance Status is used as conditional resectability, with patients with a status ≥ 2 having a poor prognosis [8].
Mesopancreas
The mesopancreas, the connective tissue region around the major vessels of the pancreatic region, which is densely traversed by blood and lymph vessels as well as nerve plexuses, has been discussed for several years [10]. Meta-analyses suggest that total mesopancreatic resection allows for better oncological outcomes [11]. In pancreatic head resection, the complete removal of mesopancreatic tissue between the portal vein, hepatic artery, base of the celiac trunk, and superior mesenteric artery (Triangle operation [12, 13]) is performed, while in left pancreatic resections (body, tail carcinomas), radical antegrade modular pancreatosplenectomy (RAMPS [14]) is performed.
[RAMPS: Depending on the extent of the tumor, an anterior or posterior RAMPS procedure is distinguished, in which essentially more radical resection is performed dorsally. In anterior RAMPS, resection is performed with removal of Gerota's fascia and perirenal fat on the left side. In contrast, in posterior RAMPS, in addition to Gerota's fascia and perirenal fat, the left adrenal gland is also resected.]
Vascular Resection
In centers, venous resections have minimally increased morbidity and mortality, and adequate overall survival is achieved [15, 16]. According to current S3 guidelines, vascular resection of the portal vein can be performed in the case of tumor infiltration ≤ 180° or in complex situations such as cavernous transformation with reconstruction [17]. Arterial resections, on the other hand, are very risky, often complex, and frequently require simultaneous venous reconstructions. Patients often do not benefit oncologically from extensive procedures and often show worse survival data than patients without vascular resection [18]. Therefore, arterial resections should be avoided outside of centers.
Unexpected arterial resections can be avoided by checking the tumor-free status of the superior mesenteric artery and celiac trunk early during a curative-intended pancreatic resection. The "Artery-first" strategy helps avoid futile procedures, allows for better planning of vascular resections and reconstructions, and improves long-term survival for selected patients in centers with appropriate expertise [19].
Oligometastasis
The term oligometastasis appears for the first time in the current S3 guidelines and describes the presence of ≤ 3 metastases, which should only be resected within studies as part of a multimodal treatment concept [7]. There are no randomized studies so far, but resection of oligometastases seems to improve patient survival data compared to palliative chemotherapy, especially after neoadjuvant therapy [20 - 23]. In Germany, the HOLIPANC and METAPANC studies are currently addressing the topic [24].
Neoadjuvant Therapy Concepts
For patients with borderline resectable pancreatic cancer, the current guideline recommends preoperative chemotherapy or radiochemotherapy, while for resectable carcinomas, it should not be performed outside of studies [7]. The recommendations are based on data from a meta-analysis as well as currently published study data [25, 26]. Since after neoadjuvant therapy, resectability in initially borderline resectable and locally advanced pancreatic carcinomas is difficult to assess morphologically, the guideline recommends surgical exploration in the case of stable disease to assess secondary resectability [7, 27]. A decrease in CA 19-9 levels can also help in assessing secondary resectability [28, 29].
Surgical Technique
The open partial pancreatoduodenectomy (OPD) is currently the gold standard for the surgical treatment of a variety of pancreatic head diseases, but is associated with morbidity in about 40% of cases. Therefore, efforts have been made in recent years to reduce this through the use of minimally invasive procedures.
Laparoscopic Techniques in Pancreatic Cancer
Left pancreatic and pancreatic head resections must be considered separately. For left resections in laparoscopic technique, the randomized controlled LEOPARD study showed faster recovery, less blood loss, and no higher complication rate compared to the open technique [30]. The combined analysis of the LEOPARD and LAPOPS studies confirmed the data [31]. Long-term quality of life remains unchanged by the laparoscopic technique [32]. A meta-analysis of the available data showed comparable results for the R0 resection rate and the rate of adjuvant chemotherapy [33]. The median overall survival was the same at 28 and 31 months for laparoscopic and open left pancreatic resections [34].
For pancreatic head resections, the 2019 published randomized and controlled LEOPARD-2 study showed higher mortality (90-day mortality 10%) in the laparoscopic group, which showed no advantages over the open group in terms of postoperative pain, recovery, hospital stay, and quality of life [36]. A recent Chinese randomized study showed comparable mortality in laparoscopic pancreatic head resection with only slight advantage of the laparoscopic technique [37].
Robotics in Pancreatic Cancer
Robotics has also established itself in pancreatic surgery over the past 10 years. In addition to the technically simpler left resection, pancreatoduodenectomy is increasingly being performed. However, a long learning curve is required [37], and a final evaluation regarding oncological outcomes is not yet possible. In recent years, only observational studies have initially been conducted, demonstrating feasibility and potential advantages of the minimally invasive technique, so the evidence to support this practice is still low [38, 39, 40]. According to international guidelines, a malignant indication is not a fundamental contraindication for robotics. In the coming years, results from randomized controlled trials and thus high-quality results are increasingly expected [41].
Very recently, the results of the EUROPA trial from Heidelberg were published [42]. This is an exploratory RCT with two parallel study arms. Patients scheduled for elective partial pancreatoduodenectomy (PD) for any indication were randomized using a centralized web-based tool (1:1) to robotic PD (RPD) or open PD (OPD). The primary endpoint was postoperative cumulative morbidity within 90 days, determined by the Comprehensive Complication Index (CCI). Between 2020 and 2022, 81 patients were assigned to RPD (n = 41) or OPD (n = 40), of whom 62 patients (RPD: n = 29, OPD: n = 33) were analyzed in the modified intention-to-treat analysis.
The CCI after 90 days was comparable between the groups (RPD: 34.02 ± 23.48 versus OPD: 36.45 ± 27.65, mean difference [95% CI]: -2.42 [-15.55; 10.71], p = 0.713). More pancreas-specific complications of grade B/C occurred in the RPD group than in the OPD group (17 (58.6%) versus 11 (33.3%); rate difference [95% CI]: 25.3% [1.2%; 49.4%], p = 0.046). The only complication that occurred significantly more frequently in the RPD group than in the OPD group was clinically relevant delayed gastric emptying. The procedure-related costs and total hospital costs were significantly higher, and the duration of the procedure was longer in the RPD group. Blood loss did not differ significantly between the groups. The intraoperative conversion rate of the RPD was 23%. The 90-day overall mortality was 4.8% with no significant differences between RPD and OPD.
In interpreting the data, the authors conclude that in a center with a very high patient volume, both RPD and OPD can be considered safe techniques. However, no advantage for robotic surgery was demonstrably shown here. The available data do not yet allow a final evaluation.
Centralization of Pancreatic Surgery
In high-volume centers for pancreatic surgery, postoperative mortality can be reduced and survival increased [43, 44, 45]. Against this background, the minimum volumes for complex pancreatic procedures in Germany will be increased from the current 10 to 20 resections per year from 2024, as decided by the Joint Federal Committee.
Whipple Procedure versus Pylorus-preserving Pancreatoduodenectomy (PPPD)
Two surgical procedures are available for the resection of pancreatic head and periampullary carcinomas, the classic Kausch-Whipple resection and the pylorus-preserving pancreatoduodenectomy. The latter has the advantage of preserved physiological food passage and reduction of dumping syndromes, postoperative weight loss, and reflux [46-53].
More recent studies [50, 52, 53] have shown a lower transfusion rate and hospital stay for PPPD patients compared to the Whipple group. Postoperative morbidity did not differ significantly between the two groups. The occurrence of gastric emptying disorders was comparable in both groups (Whipple 23% vs. PPPD 22%). There was also no significant difference in surgical radicality (R0-Whipple 82.6% vs. R0-PPPD 73.6%). Long-term follow-up showed comparable overall survival rates.