Abstract
BACKGROUND:
Cytoreductive surgery (CRS) is at the forefront of treatment for colorectal cancer with peritoneal metastasis or “carcinomatosis” (CRC-PC). We report outcomes of the operative management of CRC-PC at a single center.
STUDY DESIGN:
We retrospectively reviewed our database from 1992 through 2021. The Kaplan-Meier method was used to estimate survival. Proportional hazards regression and multivariable models were used for assessments.
RESULTS:
This study included 345 patients with mean age 53.5 years. Multivariate analysis revealed performance and resection status were associated with overall survival (OS; p < 0.001). Within the R0/R1 group, adverse impact on OS was found with increasing Peritoneal Cancer Index (PCI) score starting at 9 (hazard ratio [HR] = 1.98, CI 1.39–2.82, p = 0.0001) with the most significant hazard noted at PCI >14 (HR = 2.35, CI 1.52–3.63, p = 0.0001). Incomplete resection (R2) had significantly worse OS compared with complete CRS 33.4 (n = 206) vs R2: 12.7 months (n = 139; p < 0.0001. When stratified by PCI for the R0/R1 group, median OS for PCI less than 10, 10 to 15, and greater than 15 was 38.2, 19.7, and 22.2 m, respectively (p = 0.0007 comparing PCI less than 10 and greater than 15). Ten-year increments—1991 through 2000, 2001 through 2010, 2011 through 2020—revealed improvement in median OS (13.4 [n = 66], 19.3 [n = 139], and 29.1 months [n = 140]). However, by resection status, median OS remained stable for R0/R1 (32.3 [n = 23], 31.1 [n = 76], and 34.1 months [n = 107]) and improved for R2 (5.2 [n = 43], 14.4 [n = 63], and 14.6 months [n = 33]). Clavien-Dindo complication rate (greater than or equal to grade III) was 29.4%.
CONCLUSION:
CRS improves outcomes for CRC-PC compared with historic outcomes with nonoperative management. This benefit is greatest with complete resection and lower disease burden. Results of CRS (with or without heated intraperitoneal chemotherapy) are improving, and surgery for CRC-PC should be routinely considered.
Peritoneal metastases or “carcinomatosis” (PC) from colorectal cancer (CRC) is a justifiably feared type of advanced disease. In patients with colon and rectal malignancies, dissemination typically occurs via lymphatics and/or hematogenous spread.1 Between 8% and 13% of patients, however, will develop metastases presenting as PC.2 Development of PC is a poor prognostic indicator and the second leading cause of death in this patient population.3 In the past, PC was treated with therapeutic nihilism. Systemic therapy, with a considerable adverse effects and limited efficacy, was the only meaningful treatment. Over the last 30 years, there have been significant advances that can prolong survival in colorectal cancer in general and PC specifically. Although there have been significant improvements in systemic chemotherapy, CRS with heated intraperitoneal chemotherapy (HIPEC) remains at the forefront of treatment for this disease process for select patients.4,5
The routine use of Peritoneal Cancer Index (PCI) and Resection status (R0, R1, R2a–2c) to describe intraoperative findings of disease burden and completeness of cytoreduction, respectively, has helped to homogenize reporting and create uniformity in research related to this topic.6,7 In addition, evolving chemotherapeutic agents have affected the outcomes for patients undergoing CRS/HIPEC for CRC.
The consideration of CRS/HIPEC for CRC is primarily based on evidence from retrospective studies published from experienced centers.8 There have been 3 prospective, randomized trials: first comparing systemic chemotherapy vs CRS/HIPEC, then prophylactic HIPEC for patients with high-risk tumors, and finally, CRS alone vs CRS/HIPEC.9–12 In 2003, Verwaal and colleagues10 published data from a prospective, randomized trial that showed improved survival with CRS/HIPEC as compared with systemic chemotherapy and/or palliative surgery. Follow-up in 2008 reaffirmed the initial findings of improved survival.9 From 2015 through 2017, the COLOPEC trial examined perfusion at the time of resection of the primary tumor for patients with T4N0–2M0 or perforated tumors. They aimed to evaluate whether adjuvant HIPEC before development of PC for high-risk cancers would decrease the development of PC in the future. This was ultimately a negative result trial showing no difference in PC-free survival.
Important milestones for the management of PC were the consensus meetings in 2006 and 2010 by the American Society of Peritoneal Surface Malignancies, with the subsequent publication of guidelines in 2007 and 2014, respectively, and the University of Chicago guidelines in 2020.13,14 These meetings helped standardize treatment patterns and treatment parameters for CRS/HIPEC.
Most recently, the PRODIGE 7 trial evaluated CRS alone vs CRS with HIPEC.12 From 2008 through 2014, 265 patients were randomly assigned to CRS with HIPEC vs CRS alone. Median overall survival (OS) was 41.7 months and 41.2 months, respectively, with no statistically significant difference between the two groups. Patients who had HIPEC in addition to CRS had slightly higher complication rates at 60 days. However, subgroup analysis of patients with PCI 11–15 who underwent CRS with HIPEC showed a longer median OS and relapse-free survival, suggesting a subset of patients that do in fact benefit from perfusion.12 The authors suggested that cytoreduction alone should be the mainstay of treatment for patients with PC from CRC. However, the findings of the PRODIGE7 trial have been challenged as a result of 1) the relatively short (30 minutes) HIPEC perfusion time with oxaliplatin compared with longer perfusion times generally used in American centers and 2) a high postoperative complication rate.
Overall, the evidence to support HIPEC in patients with colon and rectal carcinoma is still widely debated. Which clinicopathologic and therapeutic factors influence response to this therapy remains controversial. We have previously reported our experience with patients treated with CRS and HIPEC from all types of peritoneal surface disease.15–17 The present study updates those reports with 30 years of our experience and outcomes of the operative management of patients with PC resulting from advanced colon and rectal carcinomas.
METHODS
The present study is a retrospective analysis of a prospectively maintained database at a single center between 1991 and 2021. The database and its analysis have been continuously approved by the Institutional Review Board at Wake Forest University. Clinical data on all patients were recorded in the database and updated by a dedicated data management unit. Pathologic and genetic results were obtained from available reports. We have previously described our techniques for patient selection, which involve detailed in-person evaluation in the Surgical Oncology Clinic and careful evaluation of objective data. Occasionally, diagnostic laparoscopy was used to determine PCI because imaging frequently underestimates disease burden. This was not routine practice but rather used selectively when there was suspicion that the patient had more severe peritoneal disease than imaging suggested. In addition, careful evaluation of the status of extraperitoneal metastases was be completed. If these extraperitoneal metastases were completely resectable or controlled on therapy, we frequently moved forward with cytoreductive surgery.
Our surgical technique for cytoreductive surgery is also standardized and described extensively in our previous patient reviews.15,17 During the 30-year experience, the goal of CRS has been removal of all gross disease in all cases. CRS consisted of the removal of gross tumor and involved organs, peritoneum, or tissue deemed technically feasible and safe for the patient, including routine omentectomy. On opening the abdomen, the quantity and distribution of disease and/or ascites present was noted and quantitated (since 2005) by the PCI.7 The resection status of patients was judged after completion of CRS using the following classification: R0, complete removal of all visible tumor and negative cytological findings or microscopic margins; R1, complete removal of all visible tumor and positive postperfusion cytological findings or microscopic margins; R2a, minimal residual tumor, nodule(s) measuring 0.5 cm or less; R2b, gross residual tumor, nodule greater than 0.5 cm but less than or equal to 2 cm; and R2c, extensive disease remaining, nodules greater than 2 cm.
The HIPEC procedure has been described previously.15,17 Briefly, near the completion of CRS, patients were cooled to a core temperature of approximately 34°C to 35°C by passive measures. Constant patient and perfusate temperature monitoring was performed in all cases. After CS was completed, peritoneal perfusion was facilitated via two 22 French inflow and two 32 French outflow catheters, placed percutaneously into the abdominal cavity. The abdominal skin incision was closed temporarily with a running cutaneous suture to prevent leakage of peritoneal perfusate. A perfusion circuit was established with approximately 3 L of crystalloid. Flow rates of approximately 1 L/min were maintained using a roller pump and a heat exchanger and then to the patient. Once a stable perfusion circuit was established and outflow temperature was >38.5°C, the chemotherapy was introduced into the perfusion circuit. A maximum inflow temperature of 43°C was tolerated during perfusion, with a target outflow temperature at the pelvis of 40°C. Total planned perfusion time after the initial addition of mitomycin c chemotherapy (MMC) was 120 minutes. The MMC was dosed based on volume of perfusate necessary to establish a stable circuit (typically 3 L). When MMC was used, 30 mg was added to the perfusate at the initiation of the HIPEC, and at 60 minutes, an additional 10 mg of MMC was added to keep MMC perfusate concentrations higher than 5 μg/mL.
Clinical follow-up occurred at 1 month after HIPEC, and then at least every 6 months thereafter for up to 5 years. After 5 years from the last HIPEC, follow-up was suggested on an annual basis. Blood counts, liver functions, and carcinoembryonic antigen (CEA; if initial level was elevated), as well as abdominal and pelvic CT or MRI scans with intravenous contrast, were obtained with each follow-up visit and when clinically indicated. Patients were followed jointly with medical oncologists. Some patients received systemic chemotherapy at the discretion of their medical oncologists.
Summary statistics for all study outcomes were evaluated, including means and standard deviations for continuous variables and frequencies, and proportions for categorical variables. These were then stratified by resection status (R0/R1 vs R2) and compared using Fisher’s Exact Test for categorical data and tests for continuous data. Patients with repeated or multiple cytoreductive surgeries and infusions were not included in this review. All clinicopathologic variables were not able to be collected for every patient owing to limitations in reporting and the electronic medical record. Genetic analyses were retrieved from available pathologic data and were not available for all patients.
All data were collected prospectively; descriptive statistics were generated for all measures, including means, ranges, and standard deviations for continuous variables and frequencies, and proportions for categorical data. OS was calculated from the date of CS and HIPEC to the last known date of follow-up or date of death. Estimates of survival were calculated using the Kaplan-Meier (product-limit) method; analysis using Cox proportional hazards was performed on all pertinent clinicopathologic variables to determine each one’s association with survival. Group comparisons of OS were performed using the approximate chi-square statistic for the log rank test. Additionally, the Cox proportional hazards regression model was used in a stepwise fashion to perform a multivariate analysis of clinic-pathologic factors to determine an overall model of independent predictors of OS. Proportional hazards regression models were created for single and multivariable models to assess the relationship between independent measures and survival. From these models, hazard ratios and their corresponding 95% confidence intervals were calculated. p values less than 0.05 were considered statistically significant. SAS (version 9.4, Cary, NC) was used for all analyses.
RESULTS
Patient and clinicopathologic characteristics
There were 345 patients included in this review treated between 1992 and 2021. Patient characteristics were stratified by resection status and are listed in Table 1. Mean age was 53.5 (± 12.4) years for all patients and similar between R0/R1 and R2 patients (53.8 [± 12.1] years and 53.2 [± 12.8] years, p = 0.65). There were 172 females and 173 males in total, with more females with an R1/R0 resection status (n = 118, 57%) and more males with R2 resection status (n = 84, 60%, p = 0.0014). The majority of patients were white (n = 301 [87%]), with no significant difference between the R0/R1 and R2 groups concerning race (p = 0.30). Patients that had R0/R1 resections tended to have better Eastern Cooperative Oncology Group (ECOG) scores with 112 patients (55%) with ECOG of 0 and only 44 patients (32%) of the R2 patients with an ECOG of 0 (p < 0.0001). More patients with incomplete resection fell into ECOG 1 and 2 (n = 67 [49%] and n = 20 [15%]) than those with complete resection (n = 80 [40%] and n = 8 [4%]). There were very few patients with ECOG 3 and 4 in both groups (n = 2 [< 1%] vs n = 6 [4%]).
Table 1.
Clinicopathologic Data
| Characteristic | Total | R0/R1 | R2 | p Value |
|---|---|---|---|---|
| Age, y, mean (± SD) | 53.5 (± 12.4) | 53.8 (±12.1) | 53.2 (±12.8) | 0.65 |
| Sex, n (%) | 0.0014* | |||
| Male | 173 (50) | 88 (43) | 84 (60) | |
| Female | 172 (50) | 118(57) | 55 (40) | |
| Race, n (%) | 0.30 | |||
| Black | 33 (10) | 16 (8) | 17 (12) | |
| White | 301 (87) | 181 (88) | 119 (86) | |
| Other | 11 (3) | 8 (4) | 3 (2) | |
| ECOG Performance Status, n (%) | <0.0001* | |||
| 0 | 156 (46) | 112 (55) | 44 (32) | |
| 1 | 147 (43) | 80 (40) | 67 (49) | |
| 2 | 28 (8) | 8 (4) | 20 (15) | |
| 3–4 | 8 (2) | 2 (1) | 6 (4) | |
| Preoperative chemotherapy, n (%) | 0.72 | |||
| Yes | 217 (67) | 134 (68) | 83 (66) | |
| No | 106 (33) | 63 (32) | 43 (34) | |
| Preoperative albumin, g/dL, mean (± SD) | 3.8 (± 0.6) | 3.9 (±0.5) | 3.7 (±0.6) | 0.031* |
| Preoperative CEA, ng/dL, mean (± SD) | 29.2 (± 88.0) | 17.3 (± 52.9) | 48.4 (± 125) | 0.017* |
| Laterality of primary tumor, n (%) | 0.54 | |||
| Left | 94 (48) | 67 (50) | 27 (44) | |
| Right | 101 (52) | 67 (50) | 34 (56) | |
| Smoker, n (%) | 0.52 | |||
| Yes | 39 (15) | 24 (17) | 15 (13) | |
| No | 168 (64) | 88 (61) | 80 (67) | |
| Former | 57 (22) | 33 (23) | 24 (20) | |
| PCI, mean (± SD) | 10.9 (± 7.1) | 8.3 (±4.9) | 18.5 (± 7.3) | <0.0001* |
| Incomplete resection, n (%) | - | |||
| R2a | NA | NA | 80 (58) | |
| R2b | NA | NA | 27 (19) | |
| R2c | NA | NA | 32 (23) | |
| Drug, n (%) | 0.81 | |||
| Mitomycin | 325 (94) | 193 (94) | 132 (95) | |
| Oxaliplatin | 20 (6) | 12 (6) | 7 (5) | |
| Time of perfusion, min, mean (± SD) | 116 (± 13.9) | 117 (± 12.4) | 115 (± 15.9) | 0.16 |
| Operation duration, h, mean (± SD) | 9.2 (± 2.9) | 8.7 (± 2.8) | 9.8 (± 2.9) | 0.0020* |
| Estimated blood loss, mL, mean (± SD) | 680 (± 622) | 598 (± 604) | 827 (± 629) | 0.0032* |
| Length of stay, d, mean (± SD) | 13.0 (± 13.0) | 10.8 (±11.3) | 15.7 (± 14.4) | 0.0019* |
| Clavien-Dindo, n (%) | 0.17 | |||
| 0–2 | 180 (71) | 101 (74) | 79 (66) | |
| 3–5 | 75 (29) | 35 (26) | 40 (34) |
Statistically significant.
CEA, carcinoembryonic antigen; ECOG, Eastern Cooperative Oncology group performance status; NA, not applicable; PCI, peritoneal cancer index.
Patients frequently underwent preoperative chemotherapy (n = 217 [67%]) with similar numbers in both R0/R1 and R2 patients (n = 134 [68%] vs n = 83 [63%], p = 0.72). Preoperative albumin was slightly higher in R0/R1 patients compared with R2 patients (3.9 [± 0.5] g/dL vs 3.7 [± 0.6] g/dL, p = 0.031). Preoperative CEA was lower in R0/R1 vs R2 patients at 17.3 (± 52.9) ng/mL and 48.4 (± 125) ng/mL respectively (p = 0.017).
Right- and left-sided tumors were equally common with no significant difference in the R0/R1 and R2 groups (p = 0.54). There were 94 left-sided tumors and 101 right-sided tumors in total. Patients smoking status was largely similar between the two groups, and the majority were nonsmokers with 88 (61%) R0/R1 and 80 (67%) R2 patients (p = 0.52). The mean PCI was 10.9 (± 7.1) and was expectedly higher in the R2 resection patients with mean of 18.5 (± 7.3) compared with 8.3 (±4.9, p < 0.0001). The average duration of surgery was 9.2 (± 2.9) hours with longer surgery times in R2 resections (9.8 [±2.9] hours) compared with R0/R1 (8.7 [± 2.8] hours, p = 0.002). This was similar to length of stay with mean of 13.0 (±13.0) days and longer stays for R2 patients 15.7 (± 14.4) days compared with R0/R1 patients (10.8 [±11.3] days, p = 0.0019). The average estimated blood loss was 680 (± 622) mL with higher volumes in R2 cases (827.0 [± 629] mL) compared with R0/R1 cases (598.0 [± 604] mL, p = 0.0032). Overall, the majority of complications fell into Clavien-Dindo (CD) classifications of 0 to 2 (n = 180 [71%]) with the rest 3 to 5. Complications were similar between the two groups. A total of 101 (74%) R1/R0 patients had CD 0 to 2 compared with 79 (66%) R2 patients, p = 0.17. Grade 3 or higher Clavien-Dindo complication rate was 29.4%.
Univariate analysis
Univariate analysis was carried out for clinicopathologic factors to determine influence on OS and is listed in Tables 2 and 3. There was no improvement in OS with preoperative chemotherapy. This was separated into preoperative treatment at any time in the past and within 3 months of surgery. For all patients, the median survival was 1.69 years vs 1.81 years for those with and without preoperative treatment, respectively (hazard ratio [HR] 1.11, 95% CI 0.76–1.63, p = 0.59). For patients with complete resection, median survival was 2.78 years (187) vs 3.11 years (19) for those with and without preoperative treatment respectively (HR = 1.05, 95% CI 0.63–1.76, p = 0.84). If chemotherapy was done within 3 months of surgery, OS in patients with complete resection was 2.42 years (134) vs 2.98 years (63) in those in whom it was done before the 3-month interval (HR = 1.09, 95% CI 0.76–1.57, p = 0.63). For all patients, OS was 1.77 vs 1.76 (HR = 1.10, 95% CI 0.85–1.42, p = 0.48). Postoperative chemotherapy did influence OS in all patients (2.48 vs 0.59, HR = 0.59, 95% CI 0.41–0.86, p = 0.0055). Within the R0/R1 group, this was not statistically significant (2.85 vs 1.42, HR = 1.26, 95% CI 0.70–2.27, p = 0.45). Within the R2 group, postoperative chemotherapy did have an effect with median survival of 1.69 vs 0.35 years (HR = 0.26, 95% CI 0.15–0.45, p < 0.0001). Laterality had no effect on survival in either group. Median survival of patients with incomplete resection and microsatellite instability was 0.08 years vs 1.16 years in microsatellite stable patients (HR = 2.53, 95% CI 2.24–2.86, p = 0.0090).
Table 2.
Univariate Analysis of Clinicopathologic Variables Affecting Overall Survival Stratified by R0/R1 and R2
| Variable | R0/R1 | R2 | ||||
|---|---|---|---|---|---|---|
| Median survival, y (n) | HR (95% CI) | p Value | Median survival, y (n) | HR (95% CI) | p Value | |
| Preoperative chemotherapy (yes vs no) | ||||||
| Within 3 mos | 2.42 (134) vs 2.98 (63) | 1.09 (0.76–1.57) | 0.63 | 1.07 (83) vs 1.28 (43) | 1.14 (0.77–1.69) | 0.51 |
| At any time | 2.78 (187) vs 3.11 (19) | 1.05 (0.63–1.76) | 0.84 | 1.06 (126) vs 0.84 (13) | 1.13 (0.63–2.03) | 0.68 |
| Postoperative chemotherapy (yes vs no) | 2.85 (94) vs 1.42 (24) | 1.26 (0.70–2.27) | 0.45 | 1.69 (39) vs 0.35 (27) | 0.26 (0.15–0.45) | <0.0001* |
| Laterality (left vs right) | 2.18 (67) vs 2.81 (67) | 1.05 (0.68–1.62) | 0.82 | 1.22 (27) vs 1.19 (34) | 0.80 (0.46–1.38) | 0.42 |
| Molecular marker (mutated vs WT) | ||||||
| KRAS | 1.77 (34) vs 3.16 (33) | 1.56 (0.81–2.99) | 0.18 | 1.36 (8) vs 1.22 (11) | 0.78 (0.25–2.40) | 0.29 |
| BRAF | 1.76 (4) vs 3.16 (19) | 3.53 (0.78–15.9) | 0.10 | NA | NA | NA |
| MSI vs MSS | 1.87 (10) vs 2.85 (54) | 1.44 (0.55–3.77) | 0.46 | 0.08(2) vs 1.16 (16) | 2.53 (2.24–2.86) | 0.0090* |
Statistically significant.
HR, hazard ratio; MSI, microsatellite instabilty; MSS, microsatellite stable; NA, not applicable; WT, wild type.
Table 3.
Univariate Analysis of Clinicopathologic Variables Affecting Overall Survival for All Patients
| Variable | All patients | ||
|---|---|---|---|
| Median survival, y (n) | HR (95% CI) | p Value | |
| Preoperative chemotherapy (yes vs no) | |||
| At any time | 1.69 (313) vs 1.81 (32) | 1.11 (0.76–1.63) | 0.59 |
| Within 3 mos | 1.77 (217) vs 1.76 (106) | 1.10 (0.85–1.42) | 0.48 |
| Postoperative chemotherapy (yes vs no) | 2.48 (133) vs 0.71 (51) | 0.59 (0.41–0.86) | 0.0055* |
| Laterality (left vs right) | 1.87 (94) vs 1.77 (101) | 0.95 (0.68–1.33) | 0.75 |
| Molecular marker (mutated vs WT) | |||
| KRAS | 1.76 (42) vs 3.12 (44) | 1.28 (0.73–2.23) | 0.60 |
| BRAF | 1.76 (4) vs 3.16 (23) | 2.51 (0.64–9.78) | 0.19 |
| MSI vs MSS | 1.76 (12) vs 2.84 (70) | 1.61 (0.72–3.64) | 0.25 |
Statistically significant.
HR, hazard ratio; MSI, microsatellite instabilty; MSS, microsatellite stable; WT, wild type.
Multivariate analysis
In multivariate analysis, performance status and resection status were factors independently associated with OS (p < 0.001). The other variables evaluated revealed no statistically significant impact on survival as demonstrated in Table 4.
Table 4.
Multivariate Analysis of Clinicopathologic Variables Associated with Overall Survival
| Characteristic | Hazard ratio (95% CI) | p Value |
|---|---|---|
| Preoperative chemotherapy, yes vs no | 1.10 (0.76, 1.58) | 0.62 |
| Preoperative albumin, per 1 unit | 1.12 (0.78, 1.61) | 0.54 |
| Age, per 5 y | 1.08 (0.99, 1.17) | 0.081 |
| Sex, m vs f | 0.83 (0.58, 1.18) | 0.29 |
| ECOG | 0.023* | |
| 1 vs 0 | 1.23 (0.87, 1.75) | |
| 2 vs 0 | 1.75 (0.83, 3.73) | |
| 3/4 vs 0 | 15.4 (2.39,100) | |
| 2 vs 1 | 1.42 (0.67, 3.02) | |
| 3/4 vs 1 | 12.5 (1.98, 76.9) | |
| 3/4 vs 2 | 8.77 (1.35, 58.8) | |
| Resection status | <0.0001* | |
| R2a vs R0/R1 | 2.13 (1.29, 3.53) | |
| R2bc vs R0/R1 | 8.20 (3.76, 17.9) | |
| R2bc vs R2a | 3.85 (1.83, 8.06) | |
| PCI | 0.62 | |
| 11–15 vs <10 | 1.22 (0.80, 1.86) | |
| 16+ vs <10 | 1.02 (0.59, 1.75) | |
| 16+ vs 11–15 | 0.84 (0.50, 1.41) | |
| Time of perfusion, per 5 min | 1.05 (0.96,1.14) | 0.29 |
ECOG, Eastern Cooperative Oncology group performance status; PCI, Peritoneal Cancer Index.
Survival
Median follow up was 88.5 months and median OS for all patients was 20.5 months. Median OS was significantly worse with incomplete cytoreduction (median OS 12.7 months, p < 0.0001; Fig. 1a). With complete resection, median OS was 33.4 months (n = 206). With incomplete resection status, OS was 16.4 months (n = 80) for R2a, 8.8 months (n = 27) for R2b, and 3.4 months (n = 32) for R2c patients (Fig. 1b). When stratified by PCI for the patients with complete resection, the median OS for less than 10 (n = 121), 10 to 15 (n = 47), and more than 15 (n = 19) was 38.2 months, 19.7 months, and 22.2 months, respectively (p = 0.0007; Fig. 2).
Figure 1.
(A) Overall survival for all patients, complete resection (R0/R1), and incomplete resection (R2), p < 0.0001. (B) Overall survival by resection status, p < 0.0001.
Figure 2.
Overall survival of patients with complete resection (R0/R1) stratified by Peritoneal Cancer Index (PCI), p = 0.0007.
For patients with complete resection, a significant adverse impact on OS was noticed with increasing PCI starting at PCI of 9 (HR = 1.98, 95% CI 1.39–2.82, p = 0.0001), with the largest hazard ratio noted at PCI greater than 14 (HR = 2.35, 95% CI 1.52–3.63, p = 0.0001). Other large hazard ratios were noted at values of 18 and 19, but the number of subjects with PCI values that large (11 and 8, respectively) provided significantly less stability in the estimate of the hazard of death compared to a cutoff value of 14.
Mortality
Overall, the 30-day mortality was 16 (4.6%) and 90-day mortality was 36 (10.4%).
Mortality within the first 30 days was 5 of 206 (2.4%), 3 of 80 (2.1%), and 8 of 59 (13.6%) for R0/R1, R2a, and R2b/c, respectively. Mortality within the first 90 days was 9 (4.4%), 9 (11.2%), and 18 (30.5%) for the aforementioned groups respectively.
Outcomes over time
A longitudinal analysis of the outcomes reveals improvement over the course of time. This is mainly seen in patients with incomplete resection while the outcomes of R0/1 subgroup has remained largely stable (Table 5).
Table 5.
Median Survival, Sample Size, and Proportions in 10-Year Intervals
| Resection type | 1991–2000 Median OS, mos, n (%) | 2001–2010 Median OS, mos, n (%) | 2010–2020 Median OS, mos, n (%) | p Value |
|---|---|---|---|---|
| Overall | 13.5 (66, 100) | 19.3 (139, 100) | 29.1 (140, 100) | 0.0082* |
| Complete resection | 32.3 (23, 35) | 31.1 (76, 55) | 34.1 (107, 76) | 0.52 |
| Incomplete resection | 5.2 (43, 65) | 14.4 (63, 45) | 14.6 (33, 24) | 0.041* |
Statistically significant.
OS, overall survival.
DISCUSSION
PC from CRC is a devastating diagnosis. Previously it was treated with a nihilistic approach limited to palliative surgery and chemotherapy. Then, however, CRS with HIPEC was developed to treat patients with PC from CRC without metastases outside of the abdomen. This aggressive approach offers a more promising alternative treatment that we have pursued since 1991 at our institution. CRS and HIPEC represent a substantial operative undertaking for both patient and surgeon. Average operative time is approximately 9 hours, with lengthy hospital stays that consume substantial resources. Outcomes have improved over the timeframe of this study, owing to several factors. Clearly, the efficacy of systemic therapy for CRC, consensus guidelines, clinical trials, and other developments in cancer treatment likely influenced outcomes. This study reflects evolution in the treatment of PC from CRC with advancement of our care over time. It reflects improvement in patient selection and continuous refining of technique to provide optimal patient care.
Several clinicopathologic variables predict outcomes. From previous studies, as well as this one, we have found that OS was influenced by tumor histology, resection status, complications, and performance status. The key outcome measure for CRC is completeness of resection, and the best index to predict completeness of resection is volume of peritoneal disease, measured via the PCI. It is not surprising that those with incomplete resections had worse preoperative factors, including poorer performance status, higher CEA, and lower albumin, likely owing to the extent of their disease. Further, early in the experience, we operated on patients with poor performance status, which we now know is associated with very poor outcomes; thus, we no longer offer the procedure to patients with an ECOG performance status of 3 or greater. Most patients who are candidates with an ECOG status of 2 are strongly considered for prehabilitation, which involves attempts to improve nutritional, medical, and functional status, potentially during ongoing preoperative chemotherapy. In addition, intraoperative variables including operative duration, estimated blood loss, and length of stay were also associated with poorer outcomes reflecting increased difficulty of resection. Despite recent studies, laterality of the tumor did not show an impact on OS (Table 2 and 3). There are studies that suggest that rectal tumors have a worse prognosis than colonic tumors.18 Age, race, preoperative chemotherapy, tumor laterality, smoking status, perfusate, time of perfusion, and CD complications were all similar between both groups.
In this study, postoperative chemotherapy was the only factor effecting OS in univariate analysis that was statistically significant. It is noteworthy that this benefit was limited to patients with incomplete (R2) resections. There were 313 patients who had preoperative chemotherapy and 133 patients who had postoperative treatment. The majority of patients underwent preoperative chemotherapy for their known stage 4 disease to facilitate and better select candidates for cytoreductive surgery (Table 2). Preoperative treatment did not portend a survival benefit and was not statistically different between those with complete and incomplete resections.
For those with postoperative chemotherapy, there was a statistically significant improvement in survival for all patients and patients with incomplete resection (Table 3). This may not be a causative relationship, because many of the patients who did not have preoperative chemotherapy were chosen for upfront CRS and HIPEC owing to low preoperative estimation of PCI scores, portending a likely complete cytoreduction. Further, this finding highlights a group of patients who were fit enough to make it through postoperative chemotherapy.19 Additionally, the benefit from postoperative chemotherapy seems to be limited to patients who were not able to be completely resected (R2). This cohort did not capture the intention to treat, so no conclusions can be drawn regarding the absolute impact of preoperative treatment; however, in those who are operative candidates, there was no difference in OS with preoperative chemotherapy. Despite lack of impact of preoperative treatments on OS, treatment response to preoperative chemotherapy may have implications regarding tumor biology and exclude those who may not benefit from extensive CRS/HIPEC. Specifically, if treatment failure to preoperative chemotherapy is indicative of aggressive tumor biology, then great care must be taken before moving forward with CRS and HIPEC. There is certainly a role for preoperative chemotherapy, but this must be balanced with the risks of delivering a treatment that has considerable side effects that may worsen functional status but may also show the true nature of their disease process. Thus, prospective studies with an intention-to-treat design are required to establish the role of preoperative treatment in treatment of this patient population, such as the ongoing CAIRO-6 trial.20
Molecular markers were also examined; however, these data were limited given lack of routine testing at our institution until recent years (Tables 2 and 3).21 Despite the modest sample size for molecular/genetic markers, it is noteworthy that the microsatellite instability-high lesions have a better prognosis when an R2 resection was performed. This confirms the benefits of treatments with checkpoint inhibitors in this subgroup. We anticipate that more complete analysis of the dataset is highly likely to yield genetic signatures, which could guide the decision to undertake CRS and HIPEC for CRC with peritoneal metastases (PM) in the future.
Median OS for all patients was 20.5 months. This is a significant improvement compared with OS in the review of our cohort in 2014, where median OS was 16.4 months.15 This is probably attributable to a combination of improved patient selection, better chemotherapeutic agents for systemic treatment, as well as ongoing improvements in postoperative management. When stratified by resection status, outcomes improved for those with incomplete resection but largely stayed the same for those with complete resection (Table 4). This is suggestive of better outcomes with postoperative chemotherapy in more recent decades with the addition of oxaliplatin, irinotecan, bevacizumab, and other agents.4,22,23 It is important to recognize that fewer patients have incomplete resections in the last 10 years as a result of improvements in patient selection, and the ones who do have incomplete resection have improved OS (Table 5).
Fundamental questions regarding HIPEC for peritoneal metastases still need to be addressed. Foremost among these is whether the addition of HIPEC after CRS is of value. This is precisely the question addressed in the French PRODIGE 7 trial. Compared with the PRODIGE 7 trial, our median OS was worse than both non-HIPEC and HIPEC groups which were 41.2 months (95% CI 35.1–49.7) and 41.7 months (95% CI 36.2–52.8), respectively. Our outcomes for R0/R1 patients with PCI less than 11 are, however, comparable with a median OS of 38 months. This is likely attributable to the stringent inclusion criteria and patient selection required for clinical trials, not reflected in our retrospective data. Interestingly, median PCI was 10 in this study as well as in PRODIGE 7. In evaluation of these data by PCI, the PRODIGE 7 trial showed that for PCI of 11 to 15, median OS was 32.7 months (95% CI 23.5–38.9) in the non-HIPEC arm and 41.6 months (95% CI 36.1 to not reached) in the HIPEC arm, (HR = 0.437, p = 0.0209).12 Our study attempted to replicate these findings by separating the R0/R1 patients into less than 10, 10 to 15, and greater than 15 as well (Fig. 2). Our numbers were not similar. However, we did find that above a PCI of 10, there was an adverse effect on OS (HR = 1.92, 95% CI 1.35–2.47, p = 0.0003). This is clearly explained at least in part by this study spanning the time period before oxaliplatin and irinotecan were available, whereas all patients in the PRODIGE 7 trial received a more modern systemic chemotherapy regimen.4,22,23 The PRODIGE 7 trial certainly establishes CRS as a standard of care for patients with PC from CRC with a PCI less than 11.
It is estimated that only a handful of patients who are potential candidates for CRS and HIPEC for CRC actually receive it, which is underscored by the relatively small number of patients in experience reports and accrued to trials. It is clear that expanding the number of centers should be done by surgical oncologists who have extensive knowledge of regional and systemic chemotherapy as well as being experienced in aggressive operative procedures of the abdomen. This suggestion is supported by several consensus statements.9,14
We acknowledge several potential weaknesses in this investigation. The nonrandomized retrospective design of our data analysis limits conclusions, but for rare entities such as CRC-PM these longitudinal data still provide valuable insight. The prospectively maintained nature of our database has allowed for a high but not complete follow-up of patients. Most of the cases were referred from outside of our institution, and thus we did not have control of the systemic therapy used. Further, we do not have complete data on genetics and mutation profiles available, because such analysis was not available during the first half of the study period. This study did not evaluate patients undergoing iterative CRS and HIPEC procedures, which may impact outcomes.24
Our data provide further evidence for the role of surgical management in significantly improving the outcomes of CRC-PM, compared with the historic outcomes with nonoperative management. Clearly, long-term survival is becoming possible for patients with CRC-PM, with survival rates approaching those seen for CRC with hepatic metastases, when complete resection is achieved along with performing HIPEC. Quality of life has been extensively studied at our institution and is not a barrier to this treatment pathway.25–27 Morbidity and mortality have improved over time but remain significant. Therefore, frank preoperative discussions with the patient and family are mandatory. However, properly selected patients have a meaningful chance at long-term survival that is rarely realized without such aggressive efforts. This benefit is greatest when complete macroscopic cytoreduction is feasible with lower disease burden.
Acknowledgment:
The authors acknowledge the substantial contributions of fellows, house officers, perfusionists, nurses, enterostomal therapists, psychosocial oncologists, pathologists, urologists, and radiologists who contributed to the care of these patients and Dr. Brian Loggie who initiated the program. We also thank Joan L. Feder for her editorial support.
Support:
This study was funded in part by the Smith Family Fund and the Comprehensive Cancer Center of Wake Forest University Biostatistics shared resource, supported by National Cancer Institute Cancer Center Support Grant P30CA012197.
Abbreviations and Acronyms
- CD
Clavien-Dindo
- CEA
carcinoembryonic antigen
- CRC
colorectal cancer
- CRS
cytoreductive surgery
- ECOG
Eastern Cooperative Oncology Group
- HIPEC
heated intraperitoneal chemotherapy
- HR
hazard ratio
- MMC
mitomycin c chemotherapy
- OS
overall survival
- PC
peritoneal carcinomatosis
- PCI
Peritoneal Cancer Index
- PM
peritoneal metastases
Footnotes
Disclosure Information: Nothing to disclose.
Presented at the Southern Surgical Association 133rd Annual Meeting, Hot Springs, VA, December 2021.
Contributor Information
Megan E Lundy, Department of General Surgery, Surgical Oncology, Wake Forest School of Medicine, Winston-Salem, NC.
Omeed Moaven, Department of General Surgery, Surgical Oncology, Wake Forest School of Medicine, Winston-Salem, NC.
Kathleen C Perry, Department of General Surgery, Surgical Oncology, Wake Forest School of Medicine, Winston-Salem, NC.
Christopher W Mangieri, Department of General Surgery, Surgical Oncology, Wake Forest School of Medicine, Winston-Salem, NC.
Cristian D Valenzuela, Department of General Surgery, Surgical Oncology, Wake Forest School of Medicine, Winston-Salem, NC.
Gregory B Russell, Department of Biostatistics and Data Science, Wake Forest School of Medicine, Winston-Salem, NC.
Rachel Bordelon, Department of General Surgery, Surgical Oncology, Wake Forest School of Medicine, Winston-Salem, NC.
Perry Shen, Department of General Surgery, Surgical Oncology, Wake Forest School of Medicine, Winston-Salem, NC.
Konstantinos I Votanopoulos, Department of General Surgery, Surgical Oncology, Wake Forest School of Medicine, Winston-Salem, NC.
Edward A Levine, Department of General Surgery, Surgical Oncology, Wake Forest School of Medicine, Winston-Salem, NC.
REFERENCES
- 1.Jayne D Molecular biology of peritoneal carcinomatosis. Cancer Treat Res 2007;134:21–33. [DOI] [PubMed] [Google Scholar]
- 2.Jayne DG, Fook S, Loi C, et al. Peritoneal carcinomatosis from colorectal cancer. Br J Surg 2002;89:1545–1550. [DOI] [PubMed] [Google Scholar]
- 3.Ihemelandu CU, Shen P, Stewart JH, et al. Management of peritoneal carcinomatosis from colorectal cancer. Paper presented at: Seminars in Oncology 2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Franko J, Ibrahim Z, Gusani NJ, et al. Cytoreductive surgery and hyperthermic intraperitoneal chemoperfusion versus systemic chemotherapy alone for colorectal peritoneal carcinomatosis. Cancer 2010;116:3756–3762. [DOI] [PubMed] [Google Scholar]
- 5.Glehen O, Kwiatkowski F, Sugarbaker PH, et al. Cytoreductive surgery combined with perioperative intraperitoneal chemotherapy for the management of peritoneal carcinomatosis from colorectal cancer: A multi-institutional study. J Clin Oncol 2004;22:3284–3292. [DOI] [PubMed] [Google Scholar]
- 6.Edge SB, Byrd DR, Carducci MA, et al. AJCC cancer staging manual. Vol 7: Springer; New York; 2010. [Google Scholar]
- 7.Jacquet P, Sugarbaker PH. Clinical research methodologies in diagnosis and staging of patients with peritoneal carcinomatosis. Cancer Treat Res 1996;82:359–374. [DOI] [PubMed] [Google Scholar]
- 8.Sadeghi B, Arvieux C, Glehen O, et al. Peritoneal carcinomatosis from non-gynecologic malignancies: Results of the EVOCAPE 1 multicentric prospective study. Cancer 2000;88:358–363. [DOI] [PubMed] [Google Scholar]
- 9.Verwaal VJ, Bruin S, Boot H, et al. 8-year follow-up of randomized trial: cytorEduction and hyperthermic intraperitoneal chemotherapy versus systemic chemotherapy in patients with peritoneal carcinomatosis of colorectal cancer. Ann Surg Oncol 2008;15:2426–2432. [DOI] [PubMed] [Google Scholar]
- 10.Verwaal VJ, van Ruth S, de Bree E, et al. Randomized trial of cytoreduction and hyperthermic intraperitoneal chemotherapy versus systemic chemotherapy and palliative surgery in patients with peritoneal carcinomatosis of colorectal cancer. J Clin Oncol 2003;21:3737–3743. [DOI] [PubMed] [Google Scholar]
- 11.Klaver CE, Musters GD, Bemelman WA, et al. Adjuvant hyperthermic intraperitoneal chemotherapy (HIPEC) in patients with colon cancer at high risk of peritoneal carcinomatosis: The COLOPEC randomized multicentre trial. BMC Cancer 2015;15:428. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Quenet F, Elias D, Roca L, et al. A UNICANCER phase III trial of Hyperthermic Intra-peritoneal Chemotherapy (HIPEC) for Colorectal Peritoneal Carcinomatosis. PRODIGE 7. European Journal of Surgical Oncology 2019;45(2):e17. [Google Scholar]
- 13.Turaga K, Levine E, Barone R, et al. Consensus guidelines from The American Society of Peritoneal Surface Malignancies on standardizing the delivery of hyperthermic intraperitoneal chemotherapy (HIPEC) in colorectal cancer patients in the United States. Ann Surg Oncol 2014;21:1501–1505. [DOI] [PubMed] [Google Scholar]
- 14.Group CCW, Izquierdo FJ, Schuitevoerder D, et al. The Chicago consensus on peritoneal surface malignancies: Management of colorectal metastases. Cancer 2020;126(11):2534–2540. [DOI] [PubMed] [Google Scholar]
- 15.Levine EA, Stewart JH 4th, Shen P, et al. Intraperitoneal chemotherapy for peritoneal surface malignancy: Experience with 1,000 patients. J Am Coll Surg 2014;218:573–585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Shen P, Hawksworth J, Lovato J, et al. Cytoreductive surgery and intraperitoneal hyperthermic chemotherapy with mitomycin C for peritoneal carcinomatosis from nonappendiceal colorectal carcinoma. Ann Surg Oncol 2004;11:178–186. [DOI] [PubMed] [Google Scholar]
- 17.Levine EA, Stewart JH 4th, Russell GB, et al. Cytoreductive surgery and intraperitoneal hyperthermic chemotherapy for peritoneal surface malignancy: Experience with 501 procedures. J Am Coll Surg 2007;204:943–953; discussion 953. [DOI] [PubMed] [Google Scholar]
- 18.Tonello M, Sommariva A, Pirozzolo G, et al. Colic and rectal tumors with peritoneal metastases treated with cytoreductive surgery and HIPEC: One homogeneous condition or two different diseases? A systematic review and meta-analysis. Eur J Surg Oncol 2019;45:2003–2008. [DOI] [PubMed] [Google Scholar]
- 19.Beal EW, Suarez-Kelly LP, Kimbrough CW, et al. Impact of neoadjuvant chemotherapy on the outcomes of cytoreductive surgery and hyperthermic intraperitoneal chemotherapy for colorectal peritoneal metastases: A multi-institutional retrospective review. J Clin Med 2020;9:E748. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Rovers KP, Bakkers C, Simkens GAAM, et al. ; Dutch Peritoneal Oncology Group (DPOG); Dutch Colorectal Cancer Group (DCCG). Perioperative systemic therapy and cytoreductive surgery with HIPEC versus upfront cytoreductive surgery with HIPEC alone for isolated resectable colorectal peritoneal metastases: Protocol of a multicentre, open-label, parallel-group, phase II-III, randomised, superiority study (CAIRO6). BMC Cancer 2019;19:390. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Levine EA, Blazer DG 3rd, Kim MK, et al. Gene expression profiling of peritoneal metastases from appendiceal and colon cancer demonstrates unique biologic signatures and predicts patient outcomes. J Am Coll Surg 2012;214:599–606; discussion 606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Rothenberg ML, Oza AM, Bigelow RH, et al. Superiority of oxaliplatin and fluorouracil-leucovorin compared with either therapy alone in patients with progressive colorectal cancer after irinotecan and fluorouracil-leucovorin: Interim results of a phase III trial. J Clin Oncol 2003;21:2059–2069. [DOI] [PubMed] [Google Scholar]
- 23.Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 2004;350:2335–2342. [DOI] [PubMed] [Google Scholar]
- 24.Votanopoulos KI, Ihemelandu C, Shen P, et al. Outcomes of repeat cytoreductive surgery with hyperthermic intraperitoneal chemotherapy for the treatment of peritoneal surface malignancy. J Am Coll Surg 2012;215:412–417. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Duckworth KE, McQuellon RP, Russell GB, et al. Caregiver quality of life before and after cytoreductive surgery and hyperthermic intraperitoneal chemotherapy. J Am Coll Surg 2020;230:679–687. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.McQuellon RP, Loggie BW, Lehman AB, et al. Long-term survivorship and quality of life after cytoreductive surgery plus intraperitoneal hyperthermic chemotherapy for peritoneal carcinomatosis. Ann Surg Oncol 2003;10:155–162. [DOI] [PubMed] [Google Scholar]
- 27.Dodson RM, McQuellon RP, Mogal HD, et al. Quality-of-life evaluation after cytoreductive surgery with hyperthermic intraperitoneal chemotherapy. Ann Surg Oncol 2016;23(Suppl 5):772–783. [DOI] [PMC free article] [PubMed] [Google Scholar]