Introduction

Pancreatic ductal adenocarcinoma (PDAC) is often perceived as an incurable disease by both patients and their physicians. While the outcome of most solid tumors has improved over the last 20 years, the median survival for patients with PDAC has steadily remained less than 12 months, even with the best therapeutic regimens currently available. Despite the discouraging statistics, approximately 20% of all patients who undergo surgical resection for PDAC survive 5 years.

In contrast to localized tumor stages, the overall survival (OS) of patients with metastatic disease (M1) is even worse [1–3]. Once distant metastases are diagnosed, the guidelines do not recommend resection of the primary tumor [4, 5].

In such palliative settings, therapeutic regimes, such as FOLFIRNOX or gemcitabine and nab-paclitaxel, have been established recently, showing increased OS with a median of 11 and 8.5 months, respectively, compared to 7 or 6.7 months with single agent gemcitabine [2].

The increasing safety of operations in the recent decades with mortality rates of <5% has led to an extension of localized approaches in pancreatic operations, partially combined with neoadjuvant treatments [6, 7]. As a result, macroscopically complete resection seems to be one of the most important prognostic factors, but the performance of additional vessel resections and/or multivisceral resections (MVRs) must be weighed carefully against the increased morbidity and mortality of these procedures [8–11]. Not only locally advanced PDACs are operatively treated; indeed, small studies and case reports have described select patients with hepatic metastases undergoing “curative” resections of the pancreas and the hepatic metastases [12–14]. A decision for an intentional resection in a patient with PDAC metastatic to the liver has been made on a highly individual basis and is multifactorial, including the patient’s wishes, age, clinical status, local resectability, and the individual risk of complications. In this subset of patients, synchronous resection of both the primary tumor, as well as the liver metastases may be beneficial and improve the outcome.

Palliative chemotherapy

Gemcitabine in a conventional dosage (1000 mg/m2 in a 30-min infusion) is standard for the treatment of locally advanced and/or metastastic pancreatic carcinoma. The 1-year survival rate is 18–20% [15]. There is only a grade B recommendation for gemcitabin, as two more effective treatments exist for defined patient groups, namely, folfirinox and a combination of gemcitabine with nab-paclitaxel.

As an alternative to gemcitabine monotherapy, combination therapy with gemcitabine and the EGF-receptor tyrosine kinase inhibitor erlotinib can be used for patients with metastatic pancreatic carcinoma [16].

A combination of 5-FU/folinic acid, irinotecan, and oxaliplatin (the folfrinox protocol) can be used in patients with metastatic pancreatic carcinoma who have a favorable risk profile (ECOG status 0–1, bilirubin value <1.5 times the upper limit of normal, age up to 75 years). This protocol significantly prolongs the overall survival compared to gemcitabine [median survival 11.1 versus 6.8 months, p < 0.0001, hazard ratio (HR) 0.57] [2]. It likewise prolongs progression-free survival from 3.3 to 6.4 months and has a higher response rate than gemcitabine (31.6%, compared to 9.4%). It is, however, also more toxic (grade III/IV neutropenia, 45.7 versus 18.7%; febrile neutropenia, 5.4 versus 0.6%; grade III/IV diarrhea, 12.7 versus 1.2%).

Gemcitabine combined with nab-paclitaxel (G + nab-P) is a further new treatment option for metastatic pancreatic carcinoma that was just approved for use in the European Union. In a phase III trial, this combination was found to improve median progression-free survival (PFS), overall survival (OS), and tumor response rates significantly in comparison with gemcitabine alone [PFS: G + nab-P: 5.5 months versus G: 37 months, HR = 0.69 (95% CI 0.58–0.82); p < 0.001; OS: G + nab-P: 8.5 months; G: 6.7 months, HR = 0.72 (95% CI 0.62–0.83); p < 0.001; response rate (RR) G + nab-P: 23%, G: 7%] [17]. Grade III/IV hematotoxicity, fatigue, neuropathy, and diarrhea were all more common in the combination group.

Combinations of gemcitabine with oxaliplatin, cisplatin, or capecitabin should not be used as the standard first-line therapy for metastatic or locally advanced pancreatic carcinoma, as a significant prolongation of survival with such combinations in comparison with gemcitabine monotherapy has been documented only in meta-analyses [15, 18] (Table 1).

Table 1 Recent randomized and controlled trials on combination chemotherapy showing a survival advantage over gemcitabine monotherapy in the palliative setting
Full size table

Synchrounous resection of metastatic pancreatic cancer

Currently, complete operative resection of localized PDAC remains the only available treatment option with curative intent in this aggressive disease. Because the majority of patients are diagnosed after the disease has become unresectable, curatively aimed resection is performed in only a small subset of patients [19]. Nonetheless, as shown in the literature [20], macroscopically tumor-free resection is a powerful tool to prolong the overall survival of patients with PDAC.

In consideration of the increasing safety of liver and pancreatic resections, the impact on OS, symptom control, and quality of life of synchronous resections in selected cases of oligometastatic disease is being revisited. In other malignancies, such as colorectal cancer, sarcoma, and even gastric cancer, evidence is increasing that metastasectomies can increase survival of selected patients if performed safely and in appropriate candidates [13, 21, 22].

Published data dealing with this issue are rare and restricted to small case series and a number of case reports. The resection of the primary tumor and synchronous liver metastases is not recommended under current national and international guidelines for the treatment of PDAC, and therefore, this particular treatment is performed only in highly selected patients [4, 5]. Eight reports, including ≥9 patients, found median OS between 5.9 and 11.4 months after resection [13, 14, 23–28]. Five of these studies present comparative groups of patients with M1 disease (TNM classification) without resection of the primary PDAC, 4 of them with less OS compared to the resection group (median OS 3–5.9 months versus 5.9–14.5 months, respectively, Table 2 [14, 23, 24, 26, 28]

Table 2 Survival data from studies with combined synchronous hepatic and pancreatic resections and a palliative bypass group (patients n ≥ 9)
Full size table

In accordance with these studies, the data in an analysis of patients from six European pancreas centers suggested a significant survival benefit with acceptable morbidity and mortality for patients receiving a combined liver and pancreas resection compared with patients with liver metastases who did not undergo resection of the primary PDAC (14.5 versus 7.5 months, p < 0.001) [29]

Despite the expected greater perioperative morbidity, mortality was not increased in the resection group. Only a small percentage of patients received neoadjuvant therapy and primarily in the resected patient group. There was no impact of neoadjuvant therapy on OS. The primary adjuvant therapy after the operation was gemcitabine-based and did not differ between the two groups; indeed, during the timespan of this study, gemcitabine was the standard of care in both the adjuvant and palliative setting [30, 31]. Only a few patients were given FOLFIRINOX as the first- or second-line treatment due to the timeframe of the study, >15 years before publication of the FOLFIRINOX trial [2].

The median OS of patients after resection of the primary PDAC and synchronous liver metastases tended to be greater compared with patients in the non-resection group [14.5 months, 95% confidence interval (CI) 10.8–18.2 months and median 7.5 months, 95% CI 4.9–10.2 months, p < 0.001]. The 5-year survival was 0% in the non-resection group, while 4 of 69 evaluable patients (5.8%) after combined resection were still alive after 5 years.

Despite the collaboration of eight high-volume centers in Europe, the total sample size was rather small over this 20-year interval (n = 69); therefore, the subgroup analyses of either head or body/tail resections resulted in even smaller group sizes, limiting the interpretation and conclusions. While the benefit in OS appears to be in the group of PDACs to the right of the portal axis, no benefit was apparent with left-sided PDACs, especially when the increase in morbidity is taken into consideration. In general, localized pancreatic body/tail PDACs have a better OS than PDACs in the head, but this effect is not present in stage IV PDACs. The limited prognosis is generally considered to be driven by distant metastasis and not by the site of the primary tumor [32] (Fig. 1).

Fig. 1
figure 1

Kaplan–Meier survival curves showing the overall survival of the resection and non-resection groups in the complete cohort (left), pancreatic head tumors (middle), and pancreatic body and tail tumors (right)

Full size image

Conclusion

As recently promoted by Conroy and colleagues [2], an intensified chemotherapeutic regimen (FOLFIRINOX) in the palliative setting (including peritoneal, pulmonary, and hepatic metastases as well as local disease) showed promising results with an observed OS of 11.1 months, but this regimen is toxic and results in greatly increased rates of grades III and IV toxicity compared with gemcitabine monotherapy.

Therefore, this chemotherapeutic option may be appropriate only for a small subset of patients, especially when taking into consideration that even the adjuvant standard regimen with gemcitabine is not always well tolerated and cannot be completed in a substantial percentage of patients. Despite increased toxicity rates, the overall quality of life of patients receiving FOLFIRINOX as palliative therapy is increased compared to gemcitabine mono- therapy, likely due to an increase in the time to deterioration, as assumed by the authors. This aspect further highlights the need for quality-of-life assessments when such palliative trials are performed. One might suggest that the outcome of patients in the resectable group could be further improved by combining both treatment approaches in a perioperative neoadjuvant setting, either with FOLFIXRINOX or gemcitabine/nab-paclitaxel therapy. The crucial question is, which patients might benefit from such an approach? Should only patients with a stable disease or a disease that appeared to regress after neoadjuvant therapy be offered aggressive combined resection, or should resection be performed in chemonaive patients with a small tumor burden, considering the number of metastases correlates with patient survival?

Future prospective trials should compare the impact on OS, morbidity, and quality of life of systemic chemotherapy with FOLFIRINOX or combined resections in the case of a preoperative therapeutic response.