Long-Term Survival in Patients Treated with Cytoreduction and Heated Intraperitoneal Chemotherapy for Peritoneal Mesothelioma at a Single High-Volume Center

Abstract

Background.

Malignant peritoneal mesothelioma (MPM) is a rare diagnosis with a dismal prognosis if untreated. Cytoreductive surgery with hyperthermic intraperitoneal chemotherapy (CRS-HIPEC) is shown to significantly improve survival. Our institution is uniquely positioned to report long-term outcomes in MPM with CRS-HIPEC, due to our robust peritoneal surface disease program existing over the past three decades.

Methods.

Our prospectively maintained, single-institution database of CRS-HIPEC cases was reviewed, identifying 111 consecutive patients with MPM over 28 years (1993–2021). Prognostic, operative, and pathologic factors were reviewed. Overall survival (OS) and conditional survival (CS) analyses were performed.

Results.

The average age was 55.1 years; 58.6% of patients were male; 17 of 111 patients (15.3%) had a second CRS-HIPEC. At first CRS-HIPEC, the average PCI score was 18.7, and the perfusate drugs were platinum-based (72.1%) and mitomycin C (27.9%). The resection status at first CRS-HIPEC was R2a (46.4%), followed by R0–1 (29.1%), and R2b-c (24.5%). Median OS was 3.3 years for the entire cohort, with 75th and 25th percentiles at 10.7 months and 10.6 years. Median CS was improved if patients survived to the 1-year postoperative mark (4.9 years, p < 0.01) and trended toward further improvement with each passing year. If 3-year postoperative survival was achieved, the median CS improved to 6.1 years.

Conclusions.

This represents one of the largest and lengthiest, single-center, longitudinal, case series of peritoneal mesothelioma treated with CRS-HIPEC. The OS suggests efficacy for CRS-HIPEC for MPM. Long-term survival improves significantly after patients achieve the 1-year, postoperative mark.

Peritoneal surface malignancy may originate from a wide variety of cancer lineages, including appendiceal, colorectal, ovarian, gastric, and mesothelial.^1–3 Of these, malignant peritoneal mesothelioma (MPM) has remained among the most challenging to treat effectively. MPM is extremely rare, occurring at a rate of approximately 1.0 per million person-years, and incidence has remained stable between 1973–2005, despite a concurrent decrease in pleural mesothelioma rates associated with decreased asbestos usage.⁴ Thus, MPM has remained exceptionally difficult to study. Before the routine use of cytotoxic chemotherapy for cancers, the median overall survival (OS) time of MPM from symptom onset to death was 11 (range 1–60) months.⁵ Attempts to treat with abdominal radiation therapy also failed to improve outcomes, with a cost of significant toxicity.⁵ The routine use of systemic chemotherapy using cyclophosphamide/Adriamycin-based regimens in the 1960s-1970s improved the dismal natural history of MPM somewhat, raising the median survival to 13.6 months.⁶ Even a more modern systemic chemotherapy combination within a Phase III trial (pemetrexed/cisplatin) failed to appreciably improve survival of MPM, with median survival time remaining at 12.1 months.⁷

Lacking satisfactory alternatives, cytoreductive surgery with heated intraperitoneal chemotherapy (CRS-HIPEC) began to be utilized to treat MPM. Initial reports suggested improved outcomes, extending life expectancy from months to years.^8–10 These encouraging results prompted some to explore additional intraperitoneal regional treatments, such as early postoperative intraperitoneal chemotherapy (EPIC), normothermic intraperitoneal chemotherapy (NIPEC),¹¹ and pressurized intraperitoneal aerosol chemotherapy (PIPAC).¹² Additionally, generally good responses and low toxicity opened the door to the performance of additional repeat CRS-HIPEC procedures for recurrence in some patients, resulting in improved survival outcomes, including at our institution.^13–16 In addition to the better survival outcomes, quality of life after treatment of MPM with CRS-HIPEC has been objectively demonstrated to improve to baseline or better within a year postoperatively.¹⁷

However, despite the success of HIPEC in improving outcomes for MPM patients at an increasing number of centers worldwide, there remains paucity of prognostic and long-term data for MPM treated with CRS-HIPEC. Attempts to fill this knowledge gap have been severely limited by the rarity of this entity, and the worldwide paucity of highly specialized centers. Thus, published reports focusing on MPM have consisted of single-center, case series with relatively small numbers, or pooled case series with larger numbers across multiple collaborative institutions.^18–28 Conditional long-term survival of MPM after CRS-HIPEC has never been reported in the literature.

With 28 years of ongoing experience performing HIPEC operations for MPM (1993 to present time), our institution is uniquely positioned to report longitudinal outcomes for our set of 111 patients who presented to us with MPM and were operated on with CRS-HIPEC. This represents one of the largest series of peritoneal mesotheliomas treated with CRS-HIPEC and possibly the largest, single-center series for this treatment population.

Herein, we first examine some of the basic demographics of our MPM population and compare survival of MPM with other cancers after CRS-HIPEC. Next, we examine long-term survival in relation to several key factors: having received one or two CRS-HIPECs, perfusate type, and resection status. Lastly, we perform two analyses to examine how time and interconnected relevant variables affect long-term survival in MPM cases undergoing CRS-HIPEC.

METHODS

All patients who underwent CRS-HIPEC operation at Wake Forest Baptist Medical Center were registered into a prospectively maintained database, since the inception our peritoneal surface malignancy program in December 1991. The institutional review board approved maintenance of this database and approved this study. The first 1,580 consecutive patients in this database were reviewed. Patients of all tumor etiologies (mesothelioma, appendix, colorectal, as well as miscellaneous tumor types) were included in our preliminary analysis comparing overall survival between tumor types. For the MPM specific portion of this study, all data was reviewed pertaining to 111 peritoneal mesothelioma patients who received CRS-HIPEC operations between August 1993 and December 2021. The vast majority of patients in our cohort were of the epithelioid variant,²⁸ and patients presenting with pure sarcomatoid variant MPM were generally not offered CRS-HIPEC.

Recorded variables included patient demographics, Eastern Cooperative Oncology Group (ECOG) score,²⁹ intraoperative and perioperative characteristics, pathologic classifications, follow-up time, recurrences, and survival data. Using previously defined criteria,³⁰ resection status was classified as: R0–1, complete cytoreduction without residual macroscopic disease; R2a, residual disease <5 mm; R2b, residual disease between 5–20 mm; R2c, residual disease >20 mm.

Operative techniques have been well described³¹; briefly, cytoreductive surgery (CRS) preceded the HIPEC portion of the operation, in which all visible tumor deposits were removed if possible, by sharp dissection, electrocautery, blunt dissection, and stripping of visceral and parietal peritoneal surfaces as appropriate. During initial assessment, the peritoneal cancer index (PCI) score was calculated by using an established method.³² A supracolic omentectomy was routinely performed, and any other nonvital organ with gross disease upon it was resected (e.g., spleen, portions of stomach, colon, small bowel, abdominal wall, minor hepatic resections). The HIPEC procedures were performed using a closed-abdomen technique, with mitomycin C or platinum-based perfusate: most commonly cisplatin at 200–250 mg/M² for 90 min, except for two occasions when carboplatin at 500–1000 mg/M² for 90 min was used. Mitomycin C was used routinely during the early years of our program, but the first agent of choice was changed to cisplatin in 2010, after our group reported improved outcomes with the latter agent for MPM.¹⁰ Per-fusate temperatures were manipulated to maintain a target outflow temperature of 40 °C. Protective ileostomy or colostomy was occasionally performed at the conclusion of the operation, at the surgeon’s discretion, for example when rectal resection was needed. Patients then routinely underwent postoperative follow-up surveillance in clinic every 3–6 months with physical exam and CT scan until 3 years postoperative, then yearly until postoperative year 6, and then every 2 years until 10 years after surgery, after which time they were seen on an as-needed basis. Patients who were lost to follow-up were censored at the last point of contact.

Descriptive statistics performed on applicable patient variables were the chi-square test and the Student’s t-test. Kaplan–Meier survival analysis with the log-rank (Mantel-Cox) test was used to estimate OS. Conditional survival (CS) analysis was calculated using Bayes’ theorem, with established clinical methods that combine this statistical test with Kaplan–Meier analysis.^33–36 Briefly, CS is the probability of long-term patient survival, under the condition that the patient has already survived to earlier defined benchmark postoperative timepoints (we used postoperative years 1, 2, and 3). For multivariate analysis, Cox proportional hazards regression was performed using relevant covariates to determine hazard ratios for OS. Significance was set at p < 0.05 for all statistical tests.

The internal clinical database was maintained prospectively with Microsoft Excel (Microsoft, Redmond, WA). The software used for statistical analysis was SPSS, ver. 26 (IBM, Armonk, NY), as well as with Microsoft Excel ver. 2016.

RESULTS

Our database query found 111 patients with MPM who underwent CRS-HIPEC between December 1993 and December 2021 (Table 1). The average age at first CRS-HIPEC was 55.1 years, and the majority of patients were male (58.6%). Postoperative 30-day mortality was 2.7% (n = 3). A subset of 17 patients received a second CRS-HIPEC operation for recurrence during their clinical course (15.3%). In these patients, average age was significantly younger (47.1 years, p<0.01), and the sex distribution was closer to even (52.9% male and 47.1% female). The majority of patients were white/Caucasian 84.7%, 11.7% were African American, and 3.6% were of other races. The racial composition did not change significantly within the subpopulation that went on to receive a second CRS-HIPEC (p = 0.28). A plurality of patients were ECOG 0 or ECOG 1 at first operation (30.3% and 39.4% respectively), and there were smaller subsets of ECOG 2 and ECOG 3 patients in this population (20.2% and 10.1%, respectively). The vast majority of mesotheliomas in our cohort were epithelioid (n = 99), and the remaining tumor variants were papillary (n = 7), sarcomatoid (n = 3), and cystic (n = 2). There were no biphasic variants in our cohort. Kaplan–Meier survival estimates were calculated for epithelioid, papillary, and sarcomatoid variants. Median OS for those cases were 38.1, 70.3, and 2.0 months, respectively (Supplemental Fig. 1).

TABLE 1.

Baseline patient characteristics

Characteristic	Patient Received One CRS-HIPEC	Patient Received Two CRS-HIPECs	Statistical Test
Patients who Received One or Two CRS-HIPECs for Mesothelioma, n	111	17
Age at First or Second CRS-HIPEC, mean years ± SD	55.1 ± 14.0	47.1 ± 10.1	t-test p = 0.01
Sex, n (%)
Male	65 (58.6%)	9 (52.9%)	χ2 0.19, p = 0.66
Female	46 (41.4%)	8 (47.1%)
Race, n (%)
White	94 (84.7%)	14 (82.4%)	χ2 2.56, p = 0.28
African-American	13 (11.7%)	1 (5.9%)
Other/Unknown	4 (3.6%)	2 (11.8%)
ECOG Performance Status at First or Second CRS-HIPEC
ECOG 0	33 (30.3%)	3 (21.4%)	χ2 8.89, p = 0.03
ECOG 1	43 (39.4%)	11 (78.6%)
ECOG 2	22 (20.2%)	0 (0.0%)
ECOG 3	11 (10.1%)	0 (0.0%)
Resection Status at First or Second CRS-HIPEC
R0/1	32 (29.1%)	8 (47.1%)	χ2 7.26, p = 0.06
R2a	51 (46.4%)	2 (11.8%)
R2b	15 (13.6%)	4 (23.5%)
R2c	12 (10.9%)	3 (17.6%)
Perfusate Type Used at First or Second CRS-HIPEC
Mitomycn C	31 (27.9%)	11 (64.7%)	χ2 0.19, p = 0.003
Platinum-based	80 (72.1%)	6 (35.3%)
Time Interval Between Operations, median ± SD
Months between First and Second CRS-HIPEC	N/A	23.6 ± 15.2
Mean PCI Score At First or Second CRS-HIPEC ± SD	18.7 ± 10.3	15.0 ± 2.9	t-test p = 0.09
Time to Recurrence, median months ± SD	11.0 ± 12.4	7.6 ± 15.0	t-test p = 0.83

Bold values reflect a p-value less than 0.05 (p < 0.05)

CRS-HIPEC, Cytoreductive Surgery-Hyperthermic Intraperitoneal Chemotherapy; ECOG, Eastern Cooperative Oncology Group; SD, Standard Deviation; PCI, Peritoneal Cancer Index.

During this time interval, 579 appendix cancers, 288 colorectal cancers, and 224 miscellaneous cancers who received CRS-HIPEC at our institution also were queried for comparison of OS to mesothelioma cases (Fig. 1). Median OS for MPM was 39.0 months, which was not statistically different from appendiceal cancers by log-rank test, which had a median OS 47.7 months (p = 0.097). The OS 75th and 25th percentiles for MPM were 10.7 months and 10.6 years, respectively. However, median OS for MPM was significantly better than that of colorectal cancer (17.3 months, p < 0.001), and miscellaneous cancers (16.7 months, p < 0.001).

FIG. 1.

Overall survival of mesothelioma versus other cancers after CRS-HIPEC

Seventeen patients (15.3%) in our set received an iterative (i.e., second) CRS-HIPEC operation, when recurrence was detected on postoperative surveillance. This is notably less than our reported CRS-HIPEC repeat case rate for low-grade appendiceal mucinous neoplasms (40.5%).³⁷ From the time of first operation, the median OS for patients with MPM who received two CRS-HIPECs was 67.7 months versus 35.9 months for those who received one CRS-HIPEC operation (p = 0.037; Fig. 2).

FIG. 2.

Overall survival of mesothelioma after single versus repeat CRS-HIPEC

We found a significant difference in survival for patients with MPM who received HIPEC perfusion with mitomycin C versus platinum-based (cisplatin or carboplatin) perfusates (Fig. 3). Postoperative OS was 42.4 months for platinum-based perfusates versus 11.6 months for mitomycin C (p = 0.007). Of note, our institution switched from the routine use of mitomycin C before 2010 to the routine use of platinum-based agents (most commonly cisplatin) for MPM, after our group published significantly better OS, and a trend toward longer disease-free and progression-free survival with cisplatin in this context.¹⁰ The results of this study confirm our earlier findings.

FIG. 3.

Overall survival of mesothelioma after CRS-HIPEC by perfusate type

Data were available for 65% of our cohort (n = 72) regarding the use of systemic chemotherapy preoperatively and/or postoperatively. Data unavailability was typically due to medical oncologic management performed outside our institution. Of cases with available data, 56% (n = 40) received no systemic chemotherapy in addition to CRS-HIPEC, 22% (n = 16) received preoperative systemic chemotherapy, 14% (n = 10) received postoperative systemic chemotherapy, and 8% (n = 6) received preoperative and postoperative systemic chemotherapy. Stratified by these groups, there were no significant differences in postoperative OS, which ranged from 29.7 to 41.1 months (Supplemental Fig. 2).

When Kaplan–Meier analysis of OS is stratified by resection status, cases that achieved R0-R1 (i.e. complete cytoreduction) fared best, with median OS of 127 months (Fig. 4). R0-R1 cases survived significantly longer than any other resection status: R2a, R2b, or R2c (p<0.001 for each pairwise comparison to R0-R1). The second-best survival stratum was achieved by R2a resections, with a median OS of 35.9 months, followed by R2b at 11.4 months OS, and R2c at 3.0 months OS. Pairwise comparison between R2a and R2c was statistically significant at p = 0.0046, but comparisons of R2a to R2b, and R2b to R2c, were not significantly different.

FIG. 4.

Overall survival after CRS-HIPEC for mesothelioma, by resection status

Because CRS-HIPEC treatment may potentially convert MPM into an indolent disease with extended life expectancy, we performed conditional survival (CS) analysis on our set of 111 MPM patients who received CRS-HIPEC operations, using survival from first operation to postoperative years 1, 2, and 3 as the condition for stratification (Fig. 5). For example, if a patient remains alive at his or her 1-year postoperative anniversary, then they have met the prerequisite condition, and thus is included in the 1-year CS curve. OS for the entire cohort without condition was 39.0 months. For patients who survived to the 1-year mark following CRS-HIPEC, CS from the time of first CRS-HIPEC was 58.4 months, which was significantly longer than the overall cohort OS (p = 0.007). For patients who survived to the 2- and 3-year marks following CRS-HIPEC, CS was 67.7 and 73.3 months, respectively. However, there was no statistical difference between 1-year and 2-year CS or between 2-year and 3-year CS. Thus, although subsequent years fared better sequentially, the only statistically significant improvement to baseline occurred when the patient reached the 1-year mark alive.

FIG. 5.

Conditional survival after CRS-HIPEC for mesothelioma, by postoperative year

We performed a multivariate Cox regression analysis to examine whether relevant preoperative and operative factors are associated with increased hazard of postoperative death (Table 2). We found increased age was independently associated with a higher postoperative mortality hazard (p = 0.008). Resection status of R2a or worse (compared with R0-R1) also was independently associated with higher postoperative mortality hazard (p < 0.005). Additionally, a functional status of ECOG 3 was independently associated with higher postoperative mortality hazard compared with ECOG 0 (p <0.001). We found that body mass index, sex, race, perfusate, and intraoperative red blood transfusions did not have a significant independent correlations with postoperative mortality hazard; however, there was a trend toward worse outcomes with mitomycin C compared with cisplatin.

TABLE 2.

Multivariate Cox regression analysis of mesotheliomas undergoing CRS-HIPEC

	Hazard Ratio [Exp(B)]	95% CI for Exp(B)		Significance (p value)
		Lower	Upper
Age	1.030	1.008	1.052	0.008
R-Status (R0–1, Ref.)				< 0.001
R-Status R2a	3.109	1.430	6.762	0.004
R-Status R2b-c	6.394	2.657	15.388	< 0.001
BMI	0.959	0.905	1.016	0.154
Sex (F, Ref.)	1.137	0.625	2.070	0.674
ECOG (0, Ref.)				< 0.001
ECOG (1)	1.524	0.727	3.193	0.264
ECOG (2)	2.312	0.994	5.379	0.052
ECOG (3)	13.269	4.790	36.758	< 0.001
Race (White, Ref.)				0.904
African-American	0.816	0.326	2.043	0.663
Other/Unknown	1.030	0.221	4.795	0.97
Perfusate (Cisplatin, Ref.)				0.523
Perfusate (Mitomycin C)	1.253	0.727	2.158	0.417
Perfusate (Others)	1.780	0.511	6.199	0.365
PRBC (Units) given in OR	0.984	0.882	1.098	0.776

Bold values reflect a p-value less than 0.05 (p < 0.05)

DISCUSSION

Because MPM is such a rare entity, there have been no prospective, randomized trials comparing CRS-HIPEC versus other treatment modalities. One can reasonably assume that equipoise has been lost between consideration of CRS-HIPEC against systemic chemotherapies alone, after such marked improvement in survival was demonstrated with CRS-HIPEC operations.⁸ Aside from one recently published, prospective, early-phase, clinical trial for palliative intraperitoneal chemotherapy, and similar trial designs in the future for second-line approaches, prospective trials on CRS-HIPEC in MPM will be very difficult to perform and remain unlikely to be completed.³⁸ Thus, long-term, survival data have remained reliant on retrospective, small, single-institutional or larger, multi-institutional, collaborative datasets. Our findings are important, because our study provides long-term outcome data from a very large, prospectively maintained dataset of MPM patients treated with CRS-HIPEC at a single institution. Additionally, our very low, 30-day, postoperative mortality (2.7%) indicates CRS-HIPEC for MPM is a safe treatment modality for carefully chosen operative candidates. Our data avoid the heterogeneity and imprecisions of mixing databases from multiple institutions and is strengthened by consistent operative methods and data-keeping throughout our institution’s nearly 30-year experience. The predominant histologic variant was epithelioid throughout the study, but examples of other variants were occasionally encountered. We typically did not offer CRS-HIPEC to patients with the sarcomatoid variant, but for our three cases in which it was indeed performed, prognosis remained abysmal. The only significant change to our methods throughout the study timeframe was changing the first-choice perfusate from mitomycin C to cisplatin, which was prompted by earlier findings of comparably better survival with cisplatin.¹⁰ Our reported median OS for MPM treated with CRS-HIPEC was 39.0 months, which is in keeping with previous reports.^8,9,19,39 Given the flattening of our survival curves after approximately 6 years postoperatively, suggesting long-term cure in these patients, we support the emerging sentiment that it is no longer unreasonable to consider MPM an incurable disease.²² Additionally, our data contribute to existing reports by providing additional granularity on survival related to resection status, perfusion agent, repeat-HIPEC status, and additional perioperative factors.

A related finding was that the performance of a second CRS-HIPEC in a subset of our patients was associated with significantly improved median overall survival: approximately 5.6 years compared with 3.0 years. It should be noted that patients who were offered a second CRS-HIPEC for disease recurrence were selected at the surgeon’s discretion for their relative overall good health and performance status predicted ability to achieve a complete cytoreduction based on CT-scan findings and perceived ability and willingness to undergo a second major operation. Our findings demonstrate a second CRS-HIPEC in the setting of disease recurrence may be significantly beneficial for long-term outcomes. Our group has recently published outcomes data on repeat hepatectomy in colorectal liver metastasis and on repeat CRS-HIPEC operations in low-grade appendiceal neoplasms.^37,40 Our analysis of repeat CRS-HIPEC for MPM will be expanded upon in later reports, as our cohort progresses and additional repeat cases are performed.

Although data were incomplete when our patients received systemic chemotherapy at outside centers, our data suggest that systemic chemotherapy given preoperatively, postoperatively, or both had no appreciable effect on survival if delivered in addition to CRS-HIPEC. As such, regional therapy will likely remain the cornerstone of curative-intent treatment for peritoneal mesothelioma.

This is the first report in the literature of conditional survival data in MPM treated with CRS-HIPEC. Conditional survival is a valuable metric for measuring how the likelihood of survival changes over time, as certain conditions are met: in other words, the dynamic probability of survival. When the surgeon sees the patient at given postoperative timepoints (such as at the 1-year, 2-year, or 3-year mark) in clinic, conditional survival data allow the surgeon to provide updated prognostic information from the time of that specific timepoint—not only from the time of operation. This is particularly important in the structured long-term, follow-up context of HIPEC for rare entities, such as MPM. After postoperative year 6, the OS and CS curves appear to flatten, implying long-term follow-up and surveillance imaging in this population of survivors should be continued but may be continued less aggressively.

The primary limitation in our study is its retrospective design, but this is largely unavoidable given the extreme rarity of MPM presenting with potentially curable regional extent of disease. A prospective clinical trial for primary treatment of MPM with CRS-HIPEC for curative intent versus other treatments has not been performed and likely will never be attempted, due to the marked improvement in survival that CRS-HIPEC operations offer for this entity. However, our dataset of 111 patients is likely the largest single-institutional outcomes report for MPM treated with CRS-HIPEC published to date. We achieved a comparable cohort size to published multi-institutional reports, without the statistical drawbacks produced by combining different databases and divergent treatment methods of multiple clinical centers.

We speculate that in the future, treatment for MPM will be guided by molecular data and genetic fingerprinting obtained at biopsy.^41–43 Additionally, progress is being made in patient-derived organoids from MPM, for in vitro mutation-specific drug testing and selection of individualized treatments.⁴⁴ In fact, MPM is a perfect candidate for development of personalized treatment guided by genomic analysis, due to its extreme rarity, inability to be detectable by screening, and devastating natural history.

We conclude that our large, institutional, cohort confirms the findings of collaborative groups worldwide, as well as smaller institutional reports, that MPM treated with CRS-HIPEC remains the optimal treatment option at this time to achieve the best long-term survival. We reported additional information on several key patient and operative variables, and their relationship to long-term survival in this clinical scenario: repeat-HIPEC status, patient functional status, resection status, and perfusate selection. Lastly, we reported conditional survival outcomes for MPM treated with CRS-HIPEC, showing that the 1-year postoperative anniversary is an important milestone for a patient to reach, after which his or her chances of long-term survival increase significantly. Large, international collaborations and individualized, tumor genetic analyses will be key to expanding treatment options and further improving outcomes for MPM in the future.