Is cytoreductive surgery and hyperthermic intraperitoneal chemotherapy indicated in hepatobiliary malignancies?

CRS/HIPEC was performed in a standard fashion as previously described [15]. Diagnostic laparoscopy was performed in all cases to assess the feasibility of cytoreduction prior to HIPEC. The procedure was aborted at the discretion of the operating surgeon if the tumor burden was deemed too bulky to attempt cytoreduction. The peritoneal cancer index (PCI) was calculated prior to operative debulking [16], and the completeness of cytoreduction (CC) score recorded at the conclusion of the procedure. Complete cytoreduction was defined as CC-0 (no macroscopic disease) or CC-1 (residual tumor implants < 2.5 mm). Incomplete cytoreduction was defined as CC-2 (residual tumor implants between 2.5 and 2.5 cm in diameter) or CC-3 (residual tumor implants greater than 2.5 cm in diameter). All patients who underwent HIPEC received 40 mg of mitomycin C at 42 °C for 90 min (30 mg for 60 min followed by an additional 10 mg for the final 30 min). Creation of anastomoses was performed after the completion of HIPEC. Major perioperative complications were graded according to the Clavien-Dindo classification system (III–V) [17] and defined as occurring within 30 days of CRS/HIPEC.

Detailed clinicopathologic data encompassing the preoperative, intraoperative, and postoperative course was collected retrospectively and maintained within a prospectively managed database.

Primary endpoints were progression-free survival (PFS) and overall survival (OS). OS was defined as time from surgery to death or last follow-up. PFS was defined as time from surgery to disease progression or relapse (diagnosed on imaging or re-operation). Postoperatively, patients were followed with serial contrast-enhanced cross-sectional imaging, although there were no strict institutional protocols in place. Secondary endpoints were rate of complete cytoreduction, estimated blood loss (EBL), operative time, length of stay (LOS), and major perioperative morbidity.

A total of 17 patients with PC secondary to primary HPB malignancies underwent attempted CRS/HIPEC. There were 9 HCC primaries and 8 pancreaticobiliary primaries (4 CCa (3 intrahepatic, 1 hilar), 3 GBC (adenocarcinoma), and 1 PCa (mucin-producing ductal adenocarcinoma)). Clinicopathologic characteristics are reported in Table 1. The cohorts were similar in all preoperative factors, except that significantly more pancreaticobiliary patients underwent neoadjuvant chemotherapy (CCa n = 2, 50%; GBC n = 3, 100%; PCa n = 1, 100% versus HCC n = 1, 11%, p = 0.008). The most frequent regimens used were gemcitabine for CCa and GBC. The PCa patient received gemcitabine. The regimen was unknown for the one HCC patient who received neoadjuvant chemotherapy. Overall, the majority of patients underwent CRS/HIPEC for metachronous lesions (HCC n = 8, 89%; CCa n = 2, 50%; GBC n = 3, 100%), although the PCa patient had synchronous PC.

All 17 patients underwent both CRS and HIPEC (Table 2). For the entire cohort, the median PCI was 10 (IQR 6–18). Complete cytoreduction was achieved in the majority of patients in the HCC (n = 7, 78%), CCa (n = 4, 100%), and PCa (n = 1, 100%) cohorts. The GBC cohort only achieved complete cytoreduction in one third of patients (n = 1). There was no significant difference between the cohorts in terms of PCI, number of organs resected, CC score, EBL, or OR time. Perioperative outcomes were also similar (LOS, ICU stay, perioperative major morbidity, perioperative mortality). The use of adjuvant chemotherapy was similar in both cohorts. The most frequent regimens used were gemcitabine and FOLFIRI for CCa, and gemcitabine-oxaliplatin for GBC. Sorafenib was the agent in all 7 of the HCC patients who received adjuvant therapy. The regimen was unknown for the one PCa patient who received adjuvant chemotherapy. Postoperative serial imaging for HCC patients included a CT with or without MRI every 3–4 months for up to 5 years. The pancreatic patient had no follow-up imaging recorded. Two of the patients with gallbladder cancer underwent serial CT scans every 2–3 months; one patient had no postoperative imaging. All 4 of the CCa patients had serial CT scans every 3–4 months; 2 of those patients had one or more PET/CT scans.

Table 3 displays survival outcomes. Median follow-up time was 15 months. Median OS for the entire cohort was 23 months with 1-year and 3-year OS rates of 73% and 41%, respectively. There was a notable trend towards longer OS in the HCC cohort (median 42 months) compared to the other cohorts, though statistical significance was not reached, p = 0.188 (Fig. 1a). Of the pancreaticobiliary malignancies, the longest median OS was seen in CCa (19 months), though this was still considerably shorter than in HCC. Of note, the 3 patients with intrahepatic CCa had a much longer median OS (19 months) than the 1 patient with hilar CCa (11 months). Median PFS for the entire cohort was 8 months, and there was no statistically significant difference between the cohorts, p = 0.550 (Fig. 1b). All patients experienced tumor recurrence by 3 years postoperatively. The shortest median PFS was in GBC (2 months), and the longest was in PCa (15 months). On multivariate analysis, age at surgery (HR 1.13, CI 1.01–1.26, p = 0.027) and PCI (HR 1.24, CI 1.05–1.47, p = 0.011) were independent predictors of OS. There were no independent predictors of PFS.

Compared to pancreaticobiliary malignancies, relatively fewer patients with HCC (13–42%) have evidence of metastatic disease at diagnosis [37]. In unresectable HCC, the only effective systemic therapy is sorafenib, which offers significant OS benefits of up to 10.7 months, 2–3 months longer than with best supportive care [38]. However, the role of sorafenib in isolated PC without distant metastases is unclear. As such, there have been many studies examining the role of aggressive surgical therapy in patients with PC, with some promising results. Ding et al. conducted a literature review of 24 patients with PC from HCC treated with CRS and found 1-year and 2-year OS of 83% and 71%, respectively [39]. These findings were echoed by Lin et al. who reviewed 53 patients with HCC and PC and demonstrated significantly improved OS for CRS compared to best supportive care (12.5 versus 2.1 months, p = 0.0013) [40]. As an extension of this concept, CRS and HIPEC have been evaluated by a few authors [7, 10, 41]. Spiliotis et al. [41] reviewed 4 patients with localized PC who underwent CRS and HIPEC, with a mean PCI of 10.2. All patients received adjuvant sorafenib. The median PFS and OS were 13.5 and 30 months, respectively. In 2018, Mehta et al. published results from a multicenter study of 21 patients with HCC [11]. At the time of surgery, the median PCI was 14 and complete cytoreduction was achieved in 16 patients. The most frequent HIPEC agent used was cisplatin, followed by doxorubicin and mitomycin C. They did not comment on concomitant use of sorafenib. The median PFS and OS were 26.3 months and 46.7 months. Those patients who achieved complete cytoreduction had a significantly longer median OS (not reached) than those with incomplete cytoreduction (5.9 months), considerably longer than with sorafenib alone. A previous report from our institution by Tabrizian et al. described 14 patients with HCC and PC who were considered for CRS, of whom 7 also underwent HIPEC [10]. At the time of surgery, 6 were for metachronous disease and the median PCI was 12. Complete cytoreduction was possible in 6 patients, and the HIPEC agent used was mitomycin C in all cases. The median OS for all patients was 35.6 months, and they noted that the cohort with HIPEC had a higher OS (42.1 months) despite having higher PCI scores. This study reports on nine patients who underwent both CRS and HIPEC; 8 were for metachronous lesions. The median PCI was 12, and complete cytoreduction was achieved in 78% of patients. Two patients experienced major perioperative morbidity; one developed sepsis from pneumonia and required ICU stay, and the other died. Seven patients received adjuvant sorafenib. Median PFS and OS were 7 and 42 months, respectively.

When considering the role of CRS/HIPEC for HPB malignancies with PC, it is evident that these cancers behave very differently depending upon their primary tumor origin. However, current literature is very sparse. This study, similar to others, is underpowered due to the small sample size. The short follow-up period is also a limitation. Despite these drawbacks, there does appear to be evidence suggesting the utility of CRS in HCC with PC with an acceptable morbidity profile, provided complete cytoreduction can be achieved. Further survival benefit may be gained with the addition of HIPEC; more studies need to be conducted to better evaluate this. Perhaps, the most likely subset of patients to benefit from CRS/HIPEC would be those with ruptured HCC and presumed peritoneal dissemination, in whom primary hepatic resection may be otherwise contraindicated. The role of CRS/HIPEC in pancreaticobiliary malignancies is still unclear, though it could be considered in patients with limited PCI, resectable disease, and metachronous metastases with a longer disease-free interval. Future studies with longitudinal evaluation need to be conducted in order to better define its value in these settings.