Cytoreductive Surgery with Hyperthermic Intraperitoneal Chemotherapy for Colorectal Peritoneal Metastases Offers 60% Five-Year Overall Survival for Low-Volume Disease

Abstract

Introduction: Cytoreductive surgery (CRS) with hyperthermic intraperitoneal early chemotherapy (HIPEC) has gained traction as a viable treatment option for patients with colorectal cancer peritoneal metastases (CRC-PM). Refinements have been made to patient selection and choice of HIPEC agent. We report outcomes with respect to peritoneal disease volume (peritoneal cancer index, PCI) and HIPEC agent for patients treated at the Western Australian Peritonectomy Service (WAPS) in the ten years from December 2013. Methods: A retrospective statistical analysis assessing the factors affecting survival outcomes of patients with CRC-PM who received CRS with HIPEC was performed, with particular focus on disease volume and HIPEC agent (Mitomycin C and Oxaliplatin). Results: 89 patients with CRC-PM were treated with CRS-HIPEC with a median overall survival (OS) of 58 months, 5-year OS of 48% and disease-free survival (DFS) of 20%. PCI <10 (n = 57) had OS and DFS of 60% and 29%, compared to 23% and 0% for PCI ≥ 10 (n = 32); HR = 2.9, p = 0.002. Three-year OS and DFS for treatment with Oxaliplatin HIPEC (n = 40) were 61% and 41%, which was not significantly different from 71% and 34% with Mitomycin C HIPEC (n = 49); HR = 1.5, p = 0.3. Conclusions: CRS/HIPEC should continue to evolve into the standard of care for carefully selected patients with CRC-PM as almost half of all selected patients survive to at least five years; in particular patients with low-volume disease (PCI < 10) can benefit greatly with a 60% five-year OS and 29% five-year DFS with low morbidity. The choice of HIPEC agent, Oxaliplatin or Mitomycin C, remains uncertain.

1. Introduction

Colorectal cancer (CRC) is one of the most common solid tumours and the fourth most common cause of cancer-related death worldwide [1]. The peritoneum is the third most common site for CRC metastases and represents a most challenging problem as this condition confers the worst prognosis of all known metastatic sites. Metachronous disease with peritoneal metastases (PM) can occur in up to 19% of all CRC and synchronous PM are found in up to 7% of cases. Historically, the management of CRC-PM was palliative chemotherapy with a poor median survival of 3–12 months.

Over the past decade surgical oncology units worldwide have begun to consider selected cases of CRC-PM as locoregional disease rather than systemic disease. As a consequence more aggressive treatment of CRC-PM in the form of cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) has become more widespread, representing an evolving paradigm shift in management [2].

As with all emerging, and indeed established, surgical treatments modifications to their regimen are part of their evolution. Initially the upper limit of PCI for CRC-PM patients was 20. This was reduced to 15 and subsequently to 10 for redo procedures.

With regard to HIPEC, Oxaliplatin had been the mainstay for CRC-PM patients. However in 2018 the PRODIGE trial [3] suggested that 30 min Oxaliplatin HIPEC did not contribute to survival over systemic therapy in combination with CRS, hence raising questions over not only Oxaliplatin but also over the use of HIPEC at all. HIPEC remains in routine use, however many units have commenced using 90 min Mitomycin C HIPEC in preference to Oxaliplatin.

This study aims to assess the outcomes of patients with CRC-PM who have undergone CRS/HIPEC at the WAPS in light of recent world literature. In particular, to compare outcomes between patients who had low disease volumes (PCI of less than 10) with higher disease volumes, and to compare those who received Oxaliplatin HIPEC with those who received Mitomycin C HIPEC.

2. Methods

2.1. Patients and Inclusion Criteria

Patients referred to the WAPS for management of CRC-PM were assessed clinically, with imaging and staging laparoscopy. Patients planned to undergo CRS were worked up and allocated a Peri·1toneal Cancer Index (PCI) score. PCI is a score out of 39 that reflects disease volume and correlates with prognosis and risk of disease recurrence and is used to determine suitability for CRS. CRC patients were offered CRS with hyperthermic intraperitoneal chemotherapy (HIPEC) if PCI was ≤15, or ≤10 in the case of previous CRS with recurrent peritoneal disease. Each patient was staged with a peritoneal cancer index (PCI) score and discussed in a multidisciplinary team meeting. CRCs with preoperative radiologic or laparoscopic PCI ≤15 were considered suitable for surgery.

2.2. CRS and HIPEC

As per the principles of cytoreductive surgery described by Sugarbaker [4], selected patients underwent CRS with curative intent. In the first 5 years of operation, patients with CRC-PM were treated with intravenous 5-fluorouracil (400 mg/m²) and 50 mg of intravenous folinic acid at anaesthetic induction followed by HIPEC consisting of 30 min of Oxaliplatin (350 mg/m²) at 42C. This regimen was changed to 90 min of Mitomycin C (12.5 mg/m²) at 42C from December 2018 following the results of the PRODIGE 7 trial [3] which failed to show any advantage of Oxaliplatin HIPEC over CRS and systemic treatments alone. HIPEC was administered via the open “colosseum” technique. Patients with CRC received IV 5-fluorouracil (400 mg per metre squared) and 50 mg of IV Folinic Acid (Folinic Acid Injection, Phebra, Sydney, Australia) at anaesthetic induction.

2.3. Data and Statistical Analysis

Data collected included: patient demographics and disease and surgical factors including PCI, CC score, operative time (hours) stoma formation and requirement of packed red blood cell transfusion (pRBC). Postoperative and longer term outcomes analysed included length of ICU (intensive care unit) stay, HDU (lower acuity high dependency unit) stay and overall hospital stay, requirement for total parenteral nutrition (TPN), reoperation rate, readmission, postoperative morbidity, including categorising complications into Clavien Dindo minor (I/II) and major (III/IV) and mortality were recorded as well as long term survival follow up (overall and disease free).

Patient outcomes were categorised and compared. The Kaplan-Meir method was used to determine the proportion of OS and DFS and 95% CI. The log-rank (Mantel–Cox) test was used to determine the test of significance between single groups. The Cox proportional hazards model was used to develop univariate and multivariate models and provide the proportional hazards ratio, 95% CI and p values for the test of significance. Multiple models were generated with continuous and categorical predictor variables, but the most clinically relevant categorical data are presented for simplicity. All tests of significance were two sided with p < 0.05 considered significant. All data was analysed using customised scripts in MATLAB (Release 2024a; MathWorks Inc., Natick, MA, USA).

3. Results

3.1. Demographic and Perioperative Data

In 8.5 years WAPS has treated 89 patients with CRC-PM with CRS/HIPEC. Demographic and perioperative data is summarized in

Table 1. Demographic and perioperative data by disease volume. Range or proportion of patients in brackets.

	PCI < 10	PCI ≥ 10	All Patients
Number	57	32	89
Age	58 (23–78)	57 (34–74)	58 (23–78)
Male/Female	33:34	16:16	49:40
PCI	6 (1–9)	14 (10–24)	7 (1–24)
Synchronous/Metachronous	12/45	18/14	30/59
Op time (hours)	9 (6–16)	11 (8–14)	10 (6–16)
Inpatient days	13 (7–37)	13 (9–74)	13 (7–74)
Transfusion	2 (3.5%)	1 (3%)	3 (3%)
Stoma	16 (28%)	16 (52%)	32 (36%)
TPN	12 (21%)	11 (34%)	23 (26%)
CD I/II	15 (26%)	9 (28%)	24 (27%)
CD III/IV	3 (5%)	6 (19%)	9 (10%)
Return theatre	3 (5%)	6 (19%)	9 10%
Readmission	8 (14%)	6 (19%)	14 16%

below. The median PCI was 7 (range 1–24); 14 cases had a PCI between 16 and 24 which was only determined intraoperatively. Seventy patients (84%) had received preoperative systemic chemotherapy and 68 patients (82%) had colorectal surgery a priori. HIPEC was administered to all patients including 8 cases where complete cytoreduction was not achieved.

There was no 30-day perioperative mortality. The major (Clavien-Dindo III/IV) complication rate was 10%; unsurprisingly greater in cases where PCI ≥ 10. Such cases also had greater need for TPN, stoma (three of which were end colostomy, the remainder were loop ileostomy) and return to theatre.

The most common significant complications were anastomotic leak (3), postoperative haemorrhage (3), deep wound dehiscence (1), pneumothorax (3), prolonged ileus (18) and PE (10), DVT (9).

3.2. Survival Data

The Kaplan–Meier (KM) OS and DFS for the whole cohort are shown in Figure 1. The median OS was 58 months and median DFS was 14 months.

Cox univariate regression identified PCI as the only statistically significant factor affecting both OS and DFS (

Table 2. Cox univariate regression model for overall survival.

	HR	CI (95%)		p Value
Male	1.46	0.72	2.94	0.29
Age > 40	2.42	0.57	10	0.23
PCI ≥ 10	2.90	1.43	5.9	<0.01

and

Table 3. Cox univariate regression model for DFS.

	HR	CI (95%)		p Value
Male	1.43	0.82	2.48	0.20
Age > 40	1.52	0.65	3.5	0.33
PCI ≥ 10	3.25	1.75	6.0	<0.01

, respectively). Age and Sex were not found to have a significant effect. The univariate proportional hazard ratios are listed in Table 2 and Table 3. A multivariate model was constructed and not shown to significantly alter the results. When age and PCI are modelled as continuous variables the results are similar though the categorical groups are only presented for simplicity.

Given the significant effect of PCI on survival, Figure 2 and Figure 3 compare the OS and DFS between patients with a PCI < 10 and those with a PCI ≥ 10; both survivals are strongly in favour of patients with PCI < 10.

This finding still holds if the PCI < 10 group is compared to patients with a PCI from 10 to 15 (n = 19 patients). The latter group has 5-year overall survival of 18% (95 CI 2.6–68%) and 2-year disease free survival of 5.3% (95%CI 0–15%). Only one patient in with a PCI ≥ 10 was disease free at 2.2 years. The univariate Cox-regression HR for overall survival was 1.15 (95%CI 1.03–1.27, p < 0.01) and for disease free survival 1.25 (95%CI 1.14–1.35, p < 0.01).

Figure 4 shows the overall survival for patients who were treated with Oxaliplatin HIPEC compared to those treated with Mitomycin C; no significant difference between the two groups was found. The Mitomycin group was right censored at 4.8 years and so the 3-year survival is reported in Figure 4.

The presence of synchronous peritoneal disease (SD, n = 30), at the time of initial diagnosis of colorectal cancer, or the development of peritoneal metachronous disease (MD, n = 59) did not impact survival. There was no difference in survival outcomes between the two groups with median and five-year OS for SD vs. MD of 43 vs. 47 months; 41% vs. 45%, respectively. Median DFS for SD v MD were 15 vs. 18 months, respectively.

4. Discussion

Our results demonstrate five-year OS and DFS of 48% and 20%, respectively, and a median survival of 58 months. Thirty-day perioperative mortality was zero and morbidity was low with CD I/II and CD III/IV of 35% and 10%, respectively.

Recent publications from world CRS/HIPEC centres have reported five-year OS of 25–52% and median OS of 21–41 months on a total of 1311 patients [3,4,5,6,7,8,9,10,11,12,13,14,15,16,17], which are comparable with our results and include the PRODIGE trial which reported a median survival of 41.2 months and 5-year OS of 36.7% [3]. Survival outcomes following CRS/HIPEC are certainly stronger than median survivals for patients with CR-PM treated with systemic chemotherapy only, which are in the order of 12.6 months [12].

Multiple studies reported various independent prognostic factors for overall survival, with PCI and cytoreduction completeness the most prominent [9]. Our study found a strong negative impact of increasing disease volume with survival. Our five-year OS and DFS for patients with a PCI < 10 were 60% and 29%. For patients with PCI ≥ 10 this was 23% and censored to 0%. Although this is a marked deterioration, we consider CRS/HIPEC is still supported for patients with a PCI ≥ 10 by the finding that three-year OS was still 53% and that these patients experienced no perioperative mortality and an acceptable CD III/IV of 19%. Our survival outcomes for PCI < 10 are consistent with recent studies reporting five-year overall survivals of 65.7%, 48% and 61%; with latter survival based on PCI < 6 [18,19,20].

The striking drop-off in survival for patients with PCI ≥ 10 (n = 32) is partly attributable to 11 cases who ultimately had PCI > 15. However even when the data was reanalysed following the removal of these 11 cases, the survival for patients with PCI between 10 and 15 was still greatly inferior to those with lower disease volume.

The findings of this study support permissive but cautious approach to patient selection for PCI ≥ 10, with care taken to avoid surgery if PCI > 15. Thus, accuracy of pre-operative PCI assessment is imperative. Staging laparoscopy offers reliable PCI assessment, with high inter-observer consistency in those without previous abdominal surgery [21]. However, some studies report laparoscopy under-stages PCI by up to 2 points when compared to PCI at time of CRS [22]. While limitations of conventional CT-based PCI assessment are recognised [23], ongoing debate exists regarding other imaging modalities. This includes ⁶⁸Ga-FAPI-04 PET/CT, which shows promise in more accurate detection of peritoneal carcinomatosis in comparison to FDG-PET [24]. Thus, improved pre-operative PCI predictive tools may improve patient selection further.

This study found no difference in survival between patients who had synchronous disease compared to those who had metachronous disease. A study on 310 patients that compared SD outcomes with MD outcomes reported similar results to this study with five-year OS for SD vs. MD of 38 vs. 41 months and median survivals of 12.5 vs. 9.4 months, respectively; however the finding with respect to DFS was significant in favour of SD, this is possibly due to our study being under powered given relatively low number of subjects [25].

Recently the Chicago Consensus [26] suggested that up front systemic chemotherapy was preferable for SD and that patients with MD should be carefully risk stratified for multiple treatment pathways, supporting the growing idea of personalised treatment regimens [27] with respect to not only timing of surgery but also to the type and combination of chemotherapy, biologic therapy or immunotherapy. This concept includes the evolving use of patient organoids to drug screen to determine optimal chemotherapy options [28].

Complications following CRS/HIPEC were stratified according to the Clavien-Dindo grade. Approximately 10% of patients had a grade III/IV complication. Only 4% of patients with lower PCI (<10) had a III/IV complication compared to 19% for those with PCI ≥ 10, again supporting the need for a more cautious approach when considering patients with PCI ≥ 10. There was no perioperative mortality.

This study showed no difference in survival outcomes between patients who were treated with Oxaliplatin HIPEC as compared to Mitomycin HIPEC. Our data is scanty with insufficient follow up on the Mitomycin group given the recency of the change in treatment protocol. In the near future trial data should emerge clarifying the question of HIPEC agent, and moreover the benefit of HIPEC itself given the questions raised by the PRODIGE trial [3].

5. Conclusions

CRS/HIPEC should continue to evolve into the standard of care for patients with low-volume CRC-PM, especially patients with a PCI of less than 10 in whom five-year OS and DFS of 60% and 29% are possible. The choice of HIPEC agent (Mitomycin or Oxaliplatin) remains unclear. It is clear however that many factors need to be taken into account for optimal patient selection. Further work is required to fully elucidate patient selection criteria, and no doubt personalised treatment regimens will play an increasing role in this expanding field.