Introduction

Malignant peritoneal mesothelioma is a rare but aggressive disease, with an incidence of 200–400 cases per year in the USA.1,2 Following pleural mesothelioma, the peritoneum is the second most common site for primary mesothelioma, accounting for an estimated 20–30% of mesotheliomas.1,2,3 The primary risk factor for disease is asbestos exposure, although prior radiation and exposure to other minerals may also increase the risk.2,4

Patients with malignant peritoneal mesothelioma often present late in the course of disease, as a result of the vague and indolent nature of its symptoms, including abdominal pain, abdominal distension or swelling, weight loss, and fever.1,2,3 Less commonly, patients may be diagnosed incidentally on laparoscopy, or acutely in the setting of bowel obstruction or perforation.4,5,6,7,8 Typically, the disease involves locoregional spread and remains confined to the peritoneal cavity, rarely metastasizing to involve lymph nodes or distant sites.1,2,9,10

Prior to the use of cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) in the late 1980s and early 1990s, median survival for malignant peritoneal mesothelioma was between 6 and 12 months when treated with systemic chemotherapy.1,2,9 Given the locoregional nature of disease, the current first-line treatment approach for malignant peritoneal mesothelioma is CRS to remove macroscopic disease in combination with HIPEC.2,3,5 This approach has been found to significantly improve survival, with a systematic review estimating survival from 34 months up to 92 months.11 Despite these advances in therapy, and although malignant peritoneal mesothelioma has been recognized as a distinct entity from pleural mesothelioma, there are no specific National Comprehensive Cancer Network (NCCN) guidelines that exist for malignant peritoneal mesothelioma.3,4,12 Furthermore, malignant peritoneal mesothelioma does not have an American Joint Committee on Cancer (AJCC) staging system.13,14

The Chicago Consensus Working Group, a panel of experts on peritoneal surface disease, published the Chicago Consensus Guidelines for the management of peritoneal mesothelioma.15 The guidelines recommend obtaining a tissue diagnosis with core-needle biopsy to determine management approach based on histologic subtype when peritoneal mesothelioma is suspected. Though there are some nuances to treatment approach based on malignant histology subtype, when complete cytoreduction is possible and the patient is deemed a surgical candidate, treatment generally involves CRS in combination with intraperitoneal chemotherapy (typically HIPEC), with or without early postoperative intraperitoneal chemotherapy (EPIC) or neoadjuvant intraperitoneal chemotherapy (NIPC).15 The Chicago guidelines suggest various intraperitoneal chemotherapy regimen options, the majority of which are a form of HIPEC, though they acknowledge that data is lacking to make a firm recommendation regarding a specific protocol. Furthermore, the management approach suggested by the Chicago Consensus Guidelines is occasionally dependent on stage of disease, though this is not based on an established AJCC system.15

With the lack of specific guidelines and expertise required to carry out CRS and HIPEC, we hypothesized that patients with malignant peritoneal mesothelioma treated at academic facilities would demonstrate superior survival outcomes compared to those treated at community facilities. In this study, we aim to compare survival outcomes of malignant peritoneal mesothelioma patients treated at academic versus community hospitals.

Materials and Methods

Data Source and Patient Selection

The National Cancer Database (NCDB) is jointly sponsored by the American College of Surgeons and the American Cancer Society. It contains demographic, disease, and treatment information for patients with a wide array of oncologic diseases treated at accredited cancer centers across the USA.16 Using the database’s participant user data files, we identified all patients diagnosed with primary malignant peritoneal mesothelioma between 2004 and 2016 (Fig. 1). Patients with benign mesothelioma histologies, notably fibrous mesothelioma and multicystic mesothelioma, were excluded.2,3 Histologies included were mesothelioma not otherwise specified (NOS), epithelioid mesothelioma, and biphasic mesothelioma. Though sarcomatoid mesothelioma is another histologic subtype of malignant peritoneal mesothelioma, it is very rare and there were no documented cases in our patient cohort.2,3

Figure 1
figure 1

Flowchart of patient selection from the National Cancer Database (NCDB). NOS, not otherwise specified

Full size image

Patients were divided according to treating facility type into two major groups: academic and community. The academic group included patients with a treatment facility that was categorized as academic/research. The community group included patients with a treatment facility that was categorized as one of the following: community cancer program, comprehensive community cancer program, or integrated network cancer program. Patients with a treatment facility type that was unknown were excluded. The study was deemed exempt by the Brigham and Women’s Hospital institutional review board.

Variables and Outcomes

Variables related to patient demographics and disease and treatment characteristics were collected and assessed. Demographic variables included age, sex, race, Charlson Comorbidity Index, insurance status, income, and education status. Disease and treatment variables included histologic subtype, surgical procedure, regional lymph nodes, receipt of chemotherapy (neoadjuvant, same-day as definitive surgery, adjuvant), and receipt of radiation therapy. Surgical procedure was divided into categories with peritoneum as the primary surgical site: local tumor destruction or excision, partial or total resection of the primary site (peritoneum), debulking surgery (surgery stated as “debulking”), radical surgery (partial or total resection of the primary site – peritoneum – and resection of other involved organs in continuity), surgery not otherwise specified (NOS), or unknown. Chemotherapy was categorized as neoadjuvant, same-day, or adjuvant. These categories were determined based on the timing of chemotherapy in reference to definitive surgical procedure. Since the NCDB does not specifically report the chemotherapy route of administration, as intravenous infusion for systemic chemotherapy or intraperitoneal for HIPEC, we added the category of “same-day” chemotherapy as day of definitive surgery to assume that this could be representative of receipt of HIPEC. The NCDB also does not collect information on specific chemotherapy agent administered.

The primary outcomes of interest were median, 1-year, and 5-year survival by treating facility type.

Statistical Analysis

Patients were divided according to treating facility type: academic or community. Demographic, disease, and treatment characteristics were assessed using descriptive statistics. Log-rank testing and Kaplan-Meier survival statistics were used to compare overall survival between groups. Unadjusted median, 1-, and 5-year survival were calculated. Adjusted overall survival for the entire patient cohort and for 3 subgroups was calculated and compared between academic and community facilities utilizing Kaplan-Meier plots. The 3 subgroups were patients treated with chemotherapy, patients treated with surgery, and patients treated with both chemotherapy and surgery. Cox proportional-hazard modeling was used to conduct a multivariable analysis and calculate hazard ratios (HRs). For adjusted analyses, we controlled for the following variables: age, sex, race, Charlson Comorbidity Index, insurance, income, education, histologic subtype, surgical procedure, regional lymph nodes, receipt of chemotherapy (neoadjuvant, same-day as definitive surgery, adjuvant), and receipt of radiation. Statistical significance was defined as p value < 0.05 for a two-sided test. Statistical analyses were performed using STATA statistical software, edition 16.1 (StataCorp, College Station, TX).

Results

Patient Characteristics

A total of 2682 patients with primary malignant peritoneal mesothelioma were identified. In total, 1272 (47.4%) of patients were treated at academic facilities and 1410 (52.6%) of patients were treated at community facilities. Demographic variables for each group are presented in Table 1. All demographic variables demonstrated an association with facility type (all p-values < 0.05) with the exception of sex (p = 0.935). In both groups, the majority of patients were male (56.4% in the academic group and 56.6% in the community group), and the majority of patients were non-Hispanic White (83.3% in the academic group and 87.6% in the community group). Fifty percent of patients treated at academic facilities were privately insured, while 51.2% of patients treated at community facilities were insured with Medicare. Patients treated at academic facilities tended to be younger, have higher incomes, and have higher levels of education compared to patients treated at community hospitals.

Table 1 Demographic characteristics of malignant peritoneal mesothelioma patients by treating facility type
Full size table

Disease and Treatment Characteristics

Table 2 presents disease and treatment characteristics by treating facility type. All variables were associated with facility type (all p values < 0.001), with the exception of radiation therapy (p = 1.00).

Table 2 Disease and treatment characteristics of malignant peritoneal mesothelioma patients by treating facility type
Full size table

For histologic subtype, epithelioid mesothelioma was more common in the academic group (52.1%) compared to the community group (40.3%). The majority of patients in the community group were categorized as mesothelioma NOS (55.8%). Few patients in both groups had biphasic mesothelioma, and no patients had sarcomatoid mesothelioma.

With regard to surgical treatment, patients in the academic group were more likely to have undergone surgery than those in the community group. In particular, patients treated at academic facilities were more likely to have undergone surgical debulking (32.2% for the academic group, 15.2% for the community group) or radical surgery (10.7% for the academic group, 5.0% for the community group) as part of their treatment, compared to those treated at community facilities.

In terms of chemotherapy, a slightly larger proportion of patients in the community group received neoadjuvant chemotherapy (34.0% in the community group, compared to 28.3% in the academic group). However, same-day chemotherapy was received by a much larger proportion of patients treated at academic facilities (28.8%) compared to community facilities (10.4%). Adjuvant chemotherapy was similar between the two groups, with 14.1% receiving adjuvant chemotherapy at academic facilities compared to 13.1% at community facilities.

Survival Outcomes

In unadjusted survival analysis, treatment at an academic center was associated with statistically significant improved survival (p < 0.001). Table 3 presents unadjusted survival data. Median survival for patients treated at academic centers was 24.8 months (95% CI 22.2–27.9), while median survival for patients treated at community centers was 11.6 months (95% CI 10.2–13.0). One-year survival was 66.2% (95% CI 63.3–68.9) and 48.9% (95% CI 46.1–51.7) for academic and community facilities, respectively. Five-year survival for patients in the academic group was 29.7% (95% CI 26.7–32.7), compared to 18.3% (95% CI 16.0–20.7) for patients in the community group.

Table 3 Unadjusted survival data for malignant peritoneal mesothelioma patients by treating facility type
Full size table

In adjusted survival analyses, academic facility demonstrated improved overall survival compared to community facility across the entire study period. This was true for the entire patient cohort, and for the 3 subgroups: patients treated with chemotherapy, patients treated with surgery, and patients treated with both chemotherapy and surgery (Fig. 2 A–D). The survival difference observed between groups widened in the subgroup analysis, with the largest difference seen among patients who were treated with both chemotherapy and surgery.

Figure 2
figure 2

Overall survival by facility type for malignant peritoneal mesothelioma. Panel A—Adjusted* survival for all patients. Panel B—Adjusted* survival by facility type for patients treated with chemotherapy. Panel C—Adjusted* survival by facility type for patients treated with surgery. Panel D—Adjusted* survival by facility type for patients treated with both chemotherapy and surgery. *Variables adjusted for: age, sex, race, Charlson Comorbidity Index, insurance, income, education, histology, surgery, regional lymph nodes, chemotherapy, and radiation

Full size image

Results from multivariable analysis are presented in Table 4. Treatment at a community facility was a statistically significant independent predictor of a higher risk of death (HR: 1.19, 95% CI 1.08–1.32, p = 0.001). Increasing age was associated with an incremental increase in the risk of death. Patients in the highest age group of 81–90 had an HR of 2.86 (95% CI 2.27–3.61, p < 0.001). Female sex was an independent predictor of improved survival (HR 0.65, 95% CI 0.58–0.72, p < 0.001). Compared to mesothelioma NOS, epithelioid mesothelioma was independently associated with a lower risk of death (HR 0.88, 95% CI 0.80–0.97, p = 0.013), while biphasic mesothelioma was independently associated with a higher risk of death (HR 1.94, 95% CI 1.55–2.43, p < 0.001). With regard to surgical procedures, local tumor destruction or tumor excision (HR 0.53, 95% CI 0.43–0.66, p < 0.001), partial or total peritoneal resection (HR 0.49, 95% CI 0.40–0.59), p < 0.001), debulking surgery (HR 0.62, 95% CI 0.52–0.73, p < 0.001), radical surgery (HR 0.69, 95% CI 0.55–0.86, p = 0.001), and unspecified surgery (HR 0.66, 95% CI 0.47–0.92, p = 0.014) were all associated with better survival compared to no surgery. Receipt of neoadjuvant, same-day, and adjuvant chemotherapy were all independently associated with lower risk of death compared to no chemotherapy. Specifically, receipt of same-day chemotherapy was associated with the lowest risk of death with an HR of 0.55 (95% CI 0.46–0.65, p < 0.001) compared to neoadjuvant chemotherapy with an HR 0.70 (95% CI 0.62–0.79, p < 0.001) and adjuvant therapy with an HR of 0.78 (95% CI 0.66–0.94, p = 0.007). Conversely, receipt of radiation was not a statistically significant predictor of survival (HR 0.76, 95% 0.44–1.30, p = 0.278).

Table 4 Cox proportional-hazards model for malignant peritoneal mesothelioma
Full size table

Discussion

In this retrospective national database study of malignant peritoneal mesothelioma, we found significantly improved median, 1-, and 5-year survival for patients treated at academic facilities compared to community facilities. Community facility was an independent predictor of higher risk of death compared to academic facility, and patients at academic facilities were more likely to undergo debulking or radical surgery and receive same-day chemotherapy compared to those at community facilities.

Our study identified important discrepancies in treatment of malignant peritoneal mesothelioma between the two groups that may help explain the differences in survival observed. Though CRS and HIPEC are the mainstay of therapy for this disease, 61.8% of patients at community centers did not undergo surgery as part of their disease treatment, compared to only 37.6% treated at academic centers. At academic facilities, 42.9% underwent debulking or radical surgery, while only 20.2% of patients at community facilities underwent these procedures. All categories of surgeries demonstrated an independent survival benefit compared to no surgery. There were also discrepancies in receipt of chemotherapy between the two groups. Though patients treated at community facilities received neoadjuvant therapy more frequently, those treated at academic facilities were more likely to receive same-day chemotherapy (28.8% at academic facilities vs. 10.4% at community facilities) which we estimate could represent receipt of intraperitoneal chemotherapy with HIPEC. Furthermore, same-day chemotherapy was independently associated with the lowest risk of death in reference to no chemotherapy, compared to either neoadjuvant or adjuvant chemotherapy.

A retrospective database study by Miura et al. of nearly 1600 patients from 1973 to 2010 assessing treatment trends of malignant peritoneal mesothelioma with CRS and HIPEC found that 61.6% of patients in the cohort did not receive surgery as part of their disease treatment.1 Though our study analyzed data from much more recent years, the no-surgery rate in this study was similar to our community cohort. This study also found that limited surgery (defined as partial or total resection of the primary site) and radical resection (defined as partial or total removal of the primary site with resection in continuity with other organs) were both independent predictors of improved survival.1 In our study, similarly defined surgical procedures were also independently associated with improved survival.

In the current study, we found a median survival of 24.8 months for patients with malignant peritoneal mesothelioma treated at academic facilities, compared to only 11.6 months for those treated at community facilities. The median survival of less than 1 year for those treated at community facilities is comparable to median survival prior to the introduction of CRS and HIPEC as a treatment approach for this disease.1,17 Furthermore, treatment at a community facility was associated with an independently increased risk of death. A recent study of patients who underwent radical surgery as part of their treatment demonstrated an overall survival of 38.4 months compared to 7.1 months for patients treated without surgery, while patients who were treated with both radical surgery and systemic chemotherapy demonstrated a survival of 41.8 months.18 In this same study, treatment at an academic facility was associated with a nearly three times higher likelihood of radical surgery compared to treatment at a community cancer program.18 A study of 62 patients with malignant peritoneal mesothelioma treated with CRS and HIPEC found an overall survival of 79 months and 1- and 5-year survival rates of 84% and 50% respectively,17 which were much higher than the survival estimates for either group in our study. A systematic review of studies conducted prior to 2006 of malignant peritoneal mesothelioma found median survival estimates between 34 and 92 months, 1-year survival rates between 60 and 88%, and 5-year survival rates between 29 and 59%.11 Most of these estimates were higher than the current studies’ survival results, though our 1- and 5-year results for patients treated at academic facilities were similar to the lower range estimates. A small prospective study of 65 patients undergoing CRS and HIPEC for malignant peritoneal mesothelioma between 2001 and 2010 found a median survival of 46.2 months and 1- and 5-year survival rates of 77% and 39%, respectively.10 Result differences may also reflect different proportions of histologic subtypes of mesothelioma, with the epithelioid histology having a better prognosis than other subtypes.2,3 It is important to note that in our study, epithelioid subtype was more common in the academic group and was associated with a lower risk of death in multivariable analysis. However, results are difficult to interpret given a large number of patients with mesothelioma NOS in our cohort, particularly in the community group. Furthermore, in our study, more patients in the academic group demonstrated biphasic histology, which was an independent predictor of a higher risk of death in multivariable analysis.

The survival differences observed between academic and community facilities may in part be explained by the lower rates of surgery and chemotherapy among patients treated at community hospitals. These treatment discrepancies may themselves result from a lack of standardized guidelines for malignant peritoneal mesothelioma, as well as a lack of expertise at community centers.4 Though not specific to peritoneal mesothelioma, studies have demonstrated improved outcomes over time with CRS and HIPEC with increasing center case volume and expertise.19 Currently, there are no dedicated NCCN guidelines for malignant peritoneal mesothelioma, even though the peritoneum accounts for nearly a third of malignant mesothelioma, and even though specific guidelines exist for pleural mesothelioma.12 In 2017, a set of management recommendations for peritoneal mesothelioma were published following the annual Peritoneal Surface Oncology Group International (PSOGI) meeting.4 Among these recommendations, the expert group suggested that all patients with peritoneal mesothelioma be evaluated at a peritoneal malignancy specialty center with experience in CRS techniques, that patients with malignant peritoneal mesothelioma histologies be offered or considered for CRS with HIPEC, and that burden of peritoneal disease and completeness of cytoreduction be clearly documented.4 The Chicago Consensus Guidelines for the management of peritoneal mesothelioma provide an algorithmic treatment approach based on histologic subtype and stage of disease.15 These recommendations are a step forward, though little guidance is given with regard to specific regimens for HIPEC or other intraperitoneal chemotherapy. Furthermore, as there is no formal staging system developed for malignant peritoneal mesothelioma, suggested stage-specific approaches are difficult to interpret.13 Yan et al. proposed a TNM staging system for malignant peritoneal mesothelioma, wherein the T-stage was based on the peritoneal cancer index (PCI), which quantifies peritoneal disease burden using lesion size within 12 abdomino-peritoneal regions.14 This study found important stage-specific differences in survival ranging from 29% for the highest stage of disease to 87% for the lowest stage of disease.14 Establishing a uniform staging approach could allow for more standardized treatment guidelines. Furthermore, though the Chicago Consensus Guidelines suggest core-needle biopsy to determine treatment approach, this may be impractical in some cases where presentation is acute or diagnosis is suggested on diagnostic laparoscopy.

Our study has a few important limitations. The study utilized the NCDB as its data source, therefore requiring a retrospective design. Though the NCDB contains a vast array of treatment information, it is not possible to differentiate between systemic intravenous and intraperitoneal chemotherapy at this time, and therefore, it was not possible to specify whether part or all of chemotherapy used to treat patients was intraperitoneal (HIPEC or otherwise) or not. To try to mitigate this limitation, we categorized chemotherapy as “same-day” if the day of chemotherapy was equal to the day of definitive surgery and assumed that those who received same-day chemotherapy as the day of definitive surgery received intraperitoneal chemotherapy in the form of HIPEC. This is a significant assumption to make and we cannot rely on this being an accurate assumption to compare utilization of HIPEC between academic and community facilities. Moreover, the NCDB does not specify specific chemotherapeutic agent used, so even assuming that same-day chemotherapy consisted of HIPEC, it would not be possible to compare specific regimens provided. Regardless, specific HIPEC agent and protocol remain important points of debate.

Furthermore, as we now know that CRS is a key component of treatment for malignant peritoneal mesothelioma, it would be ideal to obtain specific surgical information for outcomes assessment. In this case, using the NCDB, the types of surgery to the peritoneum as the primary site were defined according to the following categories: no surgery, local tumor destruction or excision, partial or total resection of the primary site, debulking surgery (stated as “debulking” surgery), radical resection (partial or total resection of the peritoneum with resection in continuity of other involved organs), and surgery NOS. Given the nature of malignant peritoneal mesothelioma, it can be difficult to interpret the extent of surgery involved in any particular category without knowing the extent of peritoneal disease. That said, the NCDB does not allow quantifying the extent of disease for malignant peritoneal mesothelioma, as it does not contain information on PCI, or on the completeness of cytoreduction, for example with the completeness of cytoreduction score (CC score), which assesses the amount of residual disease remaining following surgical resection.10,17 These variables would certainly allow a better assessment of therapeutic approach for patients with malignant peritoneal mesothelioma. In the absence of these variables, it is not possible to compare surgical outcomes by facility volume, as we would be unable to specifically define the volume of CRS and HIPEC cases at a given facility. As such, it is not possible to conduct a true volume-outcome analysis utilizing the data available in the NCDB. In light of this limitation, the current study sought to characterize and compare outcomes between academic and community facilities. However, it is important to note that academic facilities do not all necessarily have greater expertise in mesothelioma with a higher volume of CRS and HIPEC than all community facilities, with a couple single-institution studies reporting their success with CRS and HIPEC at a community center.20,21 While acknowledging the importance of multidisciplinary care, it has been estimated that between 140 and 220 cases are needed to reach technical proficiency in CRS and HIPEC.19

Furthermore, as previously mentioned, histology was classified as mesothelioma NOS for the majority of patients, which is a limitation because histologic subtype has important prognostic implications for this disease.3,4,10 Additionally, the NCDB does not provide information on pertinent risk factors for malignant peritoneal mesothelioma, particularly exposure to asbestos, which would have been interesting to assess and include in multivariable analysis. Finally, though the NCDB allows calculating overall survival, it does not provide information on disease-specific survival such that disease-free survival is unable to be estimated. Disease-specific recurrence is not reported, and therefore, we could not compare recurrence following surgery between facility types.

Conclusion

In conclusion, our study found that patients with primary malignant peritoneal mesothelioma demonstrated significantly improved survival outcomes when treated at academic facilities compared to community facilities. Patients with this disease may therefore be better served at experienced academic centers, given that CRS with HIPEC has become the mainstay of therapy and remains a fairly novel approach. There are significant gaps in treatment guidelines available for malignant peritoneal mesothelioma, and databases do not currently collect complete information on variables pertinent to the disease. As treatment with CRS and HIPEC becomes more widespread, experienced centers can collect and publish data on peritoneal mesothelioma outcomes and allow for the elaboration of treatment guidelines in a widely accessible source. Further guidelines could potentially facilitate standardization of the procedure and more widespread use, including outside of academic centers.