Outcomes of Cytoreduction and Oxaliplatin-Based Hyperthermic Intraperitoneal Chemotherapy in Patients With Peritoneal Carcinomatosis From Colorectal Cancer

Among patients with metastatic colorectal cancer, 25% have isolated peritoneal carcinomatosis. We performed a systematic review and meta-analysis to assess the disease-free survival (DFS) and overall survival (OS) of patients undergoing hyperthermic intraperitoneal chemotherapy with oxaliplatin. Eleven studies were included in the final assessment. Pooled three- and five-year OS rates were 58.60% and 42.19%, respectively. The estimated pooled three- and five-year DFS rates were 23.47% and 14.26%, respectively.


Introduction And Background
Worldwide, colorectal cancer (CRC) is the second most common cancer in women and the third most common in men. The global incidence of CRC in 2018 accounted for 10.2% of all cancers [1]. Peritoneal carcinomatosis (PC) affects around 30% to 40% of patients with CRC, 5% to 10% of whom will have the synchronous disease [2,3]. It is the sole metastasis site in approximately 25% of patients [4]. Although patients presenting with PC arising from CRC previously received palliative treatment, in carefully selected patients, it is no longer considered a terminal condition [5]. In the past few decades, such patients have instead been managed with cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) [6][7][8][9][10][11][12][13].
CRS plus HIPEC has become a promising management modality that improves disease-free survival (DFS) and overall survival (OS) rates in PC secondary to CRC. This modality can yield five-year OS rates of 19% to 51%, which in the case of palliative systemic treatment is only 13% [5,7]. Approaching a locoregional disease is best achieved by locoregional modalities [14]. The rationale behind this treatment modality is to remove all macroscopic diseases with CRS and microscopic cells by HIPEC [15]. A major advantage of HIPEC is the exposure of the malignant cells to a hyperthermic chemotherapeutic agent, thus enhancing its activity and increasing the toxicity to neoplastic cells [16][17][18].
Multiple cytotoxic agents can be used for HIPEC as intraperitoneal (IP) chemotherapy. As there is no solid evidence for the superiority of one chemotherapeutic agent over another in PC arising from CRC, the decision about which chemotherapeutic agent to use is based on multiple variables, including, but not limited to, the tumor histology and individual experience at the PC treatment center [5]. Mitomycin C (MMC) and oxaliplatin are currently the most commonly used IP chemotherapeutic agents for HIPEC. Compared with systemic infusion, in HIPEC, the high molecular weights of oxaliplatin and MMC permit their exposure to malignant cells in high concentrations and to prompt their absorption [19,20]. Studies that compared these two agents reported varying results regarding survival benefits [5,[21][22][23][24]. More recently, surgeons have used IP irinotecan in combination with oxaliplatin, but with no improvement in OS or DFS [8].
Despite oxaliplatin being the drug of choice in systemic adjuvant and palliative settings [25], the survival benefit in the setting of HIPEC has not been compared with other treatment modalities in randomized settings. Whether the use of oxaliplatin in HIPEC provides additional survival benefit to patients is controversial, given the preliminary results of the PRODIGE 7 and the PROPHYLOCHIP trials [26,27]. Nonrandomized data suggest that oxaliplatin may be superior to other IP agents.
In this systematic review and meta-analysis, we assessed studies for the survival benefit of patients with CRC and PC who underwent treatment with CRS plus HIPEC in which oxaliplatin was used.

Study selection
Studies were screened by title and/or abstract. Reviewers assessed the retrieved articles for inclusion and exclusion criteria, and independently assessed all of the articles. The senior author resolved any disagreements between reviewers.

Inclusion and exclusion criteria
For a study to be considered, it had to include primary CRC patients with PC who underwent CRS plus HIPEC, with oxaliplatin being used in at least one of the HIPEC groups. It also had to be published in English and to report OS and/or DFS for the oxaliplatin group. We excluded case reports, review studies, phase I clinical trials, non-human studies, studies that included pediatric patients (<18 years of age), pharmacokinetics and dynamics studies, studies in which the primary tumor origin was other than colorectal, studies of pressurized IP aerosol chemotherapy, studies of non-oxaliplatin-based HIPEC, and overlap in patients/data between studies. If the same trial was reported several times, we used the most recent publication to avoid replication of study participants.

Quality assessment
The methodological index for non-randomized studies (MINORS) criteria were used to assess the quality of the studies in this systematic review and meta-analysis [28]. Three authors (MA, SS, and EF) assessed the quality individually, which was then reviewed by NT for any disagreement. The cutoff for an acceptable quality assessment score for non-comparative studies was set at 75% (equivalent to 12 points) and for comparative studies at 75% (18 points).

Data extraction
Three authors (MA, SS, and EF) independently extracted data from the included studies by using a predesigned extraction form created by the most senior author, containing the following: author, publication year, setting, the total number of patients in the study, number of patients receiving oxaliplatin, the overall male proportion of patients and male proportion of patients receiving oxaliplatin, median and mean age of patients overall and those receiving oxaliplatin chemotherapeutic agents used for HIPEC, comparison chemotherapeutic agent (for either HIPEC or systemic use), median DFS and OS for the oxaliplatin group, DFS and OS overtime for the oxaliplatin group, and median follow-up period. Three authors (MA, SS, and EF) reviewed the extracted data for further validation of values and minimization of error during the extraction process. After data extraction, the authors of the included papers were contacted to provide additional data that might help to expand the systemic review. However, most of the authors did not respond.

Statistical analyses
The summary estimates of the one-, two-, three-, and five-year DFS and OS rates were calculated after applying weights that were based on the sample size of the individual studies. Random-effects models were used in the presence of heterogeneity, while fixed-effects models were used in their absence [29]. Forest plots were created to graphically represent the summary estimates and the contributions of the individual studies.
Heterogeneity among the studies was assessed by the index of heterogeneity (I 2 ) [30,31]. This test statistic is presented in percentage form, expressing the proportion of variation across the different publications attributed to heterogeneity [30]. The lower the number, the less the heterogeneity present. Publication bias was assessed by visually inspecting Begg's funnel plots for five-year DFS and five-year OS rates and by Egger's regression asymmetry tests [29].
P-values of ≤ 0.05 indicated statistically significant results. All statistical significance tests were two sided, and the analyses were performed by using Stata 12.1 (StataCorp LP, College Station, TX, USA).

Study Selection and Literature Search
In total, 2,939 articles were identified through the databases ( Figure 1). First these studies were assessed by title and/or abstract. Of these studies, 2,818 were excluded because they were irrelevant, not English, nonhuman, or duplicates. The initially included studies narrowed down to 121. After a full-text review of these studies, 11 were included in the final quantitative synthesis (

Study Characteristics
An outline of the studies included in this analysis, with their references, is displayed in Table 1. A total of 11 studies were included in the meta-analyses. The publication years ranged between 2007 and 2017. Of the 11 studies, seven were prospective [7,8,22,32,35,37,38] and four were retrospective [5,33,34,36]. None of the included studies were randomized controlled trials (RCTs). Across all studies, 797 of 1,630 patients underwent HIPEC with oxaliplatin. Among these studies, three included the addition of irinotecan to oxaliplatin in HIPEC (294 patients) [7,8,32]. The remaining 833 patients were excluded because they received HIPEC chemotherapeutic agents other than oxaliplatin (MMC, cisplatinum, or irinotecan).
The mean or median age of the patients is presented in Table 1. The proportions of male patients ranged between 33.3% and 65.6%. The median follow-up time ranged from 13 to 63 months. Patients in seven of the included studies received neoadjuvant treatment [7,8,[32][33][34]37,38]. All included studies used a regimen based on oxaliplatin only, except for the study by Faron et al, which used oxaliplatin plus irinotecan [32]. When present, for comparison, the second arm of the studies varied between using HIPEC based on nonoxaliplatin only [5,8,22,34,36] or systemic chemotherapy [7]. Other HIPEC agents based on non-oxaliplatin only included MMC [5,22,36], cisplatinum [22], and irinotecan alone [34] or combined with oxaliplatin [8].
Not all studies reported one-, two-, three-, and five-year survival estimates. We present here the summary estimate of the OS and DFS for the studies that reported them. The pooled estimates of two-, three-, and five-year OS are presented in Figure 2.     Three studies included in our systematic review assessed the addition of irinotecan to oxaliplatin [7,8,32]. The total number of patients across these studies was 294. The five-year OS rates reported by Elias et al. [7], Faron et al. [32], and Quenet et al. [8] were 51%, 42%, and 42.4%, respectively. The five-year DFS rates reported by Faron et al. [32] and Quenet et al. [8] were 14% and 14.2%, respectively, whereas this rate was not reported by Elias et al. [7].

Study Quality
Using MINORS criteria for the quality assessment of the included studies, we found that the median for the comparative studies [5,8,34,36,38] was 15/24 with a range of 13-20, with only 40% achieving a score of 75% or more. On the other hand, the median score in the non-comparative studies [7,22,32,33,35,37] was 12/16 with a score ranging from 10 to 12, with 66.7% achieving a score of 75% or higher.

Publication Bias
Begg's funnel plots for the five-year DFS and five-year OS rates are shown in Figure 4. No publication bias was observed.

5-year disease-free survival (DFS) and 5-year overall survival (OS) rates.
The horizontal line illustrates the fixed-effects summary estimate, while the sloping lines indicate the expected 95% confidence intervals for a given standard error (SE).

Discussion
With medical and surgical advancements, PC in selected patients with primary CRC is no longer incurable, and surgical treatment is gaining acceptability. Before the HIPEC era, survival rates with traditional systemic chemotherapy for patients with PC were between 5.2 and 12.6 months [16].
During our literature search, no randomized clinical trials matched our inclusion criteria. This probably relates to the fact that PC from CRC is a complex disease that is just beginning to be understood. Our systematic review calculated the pooled estimate of survival rates for patients who underwent HIPEC with oxaliplatin. In our study, the two-, three-, and five-year OS rates were 79. No standardized HIPEC regimen has been proven to have a superior survival outcome compared with other regimens. Some centers use MMC as an agent of choice and others prefer oxaliplatin. As oxaliplatin is one of the agents of choice for systemic treatment [39], logic suggests that it is superior to MMC in HIPEC for CRC.
No previous meta-analysis has been conducted, however, that compares oxaliplatin head to head with MMC. We aimed in this systematic review to contribute to this debate by assessing oxaliplatin survival outcomes.
Oxaliplatin is suggested as initial therapy for metastatic CRC in patients who are not candidates for CRS plus HIPEC [39]. In patients who are found to be sensitive to platinum-based chemotherapy, it has been reported that repeated administration of oxaliplatin continued to show a systemic response [40,41]. In animal models, local infusion of oxaliplatin in the peritoneal cavity results in a high local concentration with minimal systemic absorption and toxicity [42].
PRODIGE 7 is a phase III RCT that compares CRS plus HIPEC with oxaliplatin against seldom CRS. The preliminary results showed the benefit of oxaliplatin-based HIPEC in patients with a mid-range (11-15) PC index (PCI), as opposed to no benefit for patients with a low-range PCI (<11) [26]. Our meta-analysis did not assess the effect of PCI on the DFS and OS of patients receiving HIPEC with oxaliplatin. Another RCT with recently released preliminary results is the PROPHYLOCHIP trial. It evaluated the benefit of prophylactic oxaliplatin-based HIPEC in patients with CRC at high risk of PC. Early results showed no difference in outcome for patients at high risk of developing PC undergoing prophylactic HIPEC compared with that of controls [27]. Our opinion is that select patients with a peritoneal recurrence may still benefit from oxaliplatin-based HIPEC.
Huang et al. [43] conducted a systematic review of PC secondary to CRC. Their primary endpoint was an evaluation of the efficacy and effectiveness of HIPEC in CRC. In seven studies, 614 patients received MMCbased HIPEC. Four studies with a total of 283 patients received oxaliplatin-based HIPEC. For MMC-based HIPEC, the one-, three-, and five-year OS was 79.5%, 38.8%, and 34%, respectively. For oxaliplatin-based HIPEC, the one-, three-, and five-year OS was 93%, 59%, and 43%, respectively. These findings are consistent with the OS of our systemic review and meta-analysis of our 11 included studies and 797 patients.
Postoperative complications in HIPEC include hematological complications related to the therapeutic agents, which have been reported to occur in 13.6% of patients [44]. One of the arguments against the use of oxaliplatin-based HIPEC is that, compared with MMC, it is associated with more such complications [45]. The literature on this is inconclusive. In a retrospective study, Votanopoulos et al. [46] compared toxicities between oxaliplatin and MMC in the setting of HIPEC. Oxaliplatin had statistically significant grade 3 or 4 platelet toxicity. Their study also demonstrated a trend toward a significantly higher grade 2-4 neutrophil toxicity [46]. Other studies demonstrated that MMC is associated with a higher number of neutropenic toxicities [21]. Another aspect to consider when using oxaliplatin-based HIPEC is that it has been shown to be an independent risk factor for hemorrhagic complications in the postoperative period [45,46].
Our systematic review has several limitations. None of the included studies were RCTs. Regarding the quality of the included studies, five studies scored less than 75% based on the MINORS criteria. Five studies did not mention whether there was any loss of follow-up, which could contribute to attrition bias. We believe that the other items in the MINORS score do not directly influence the validity of our results. We did not assess the role of PCI on survival in this systematic review, as it was not available in all included studies, which is increasingly thought to be a major factor in this population. Another limitation is that the included studies did not report the same survival outcomes. Results of publication bias should be interpreted with caution, as there were a few studies assessed in the funnel plot. There was a low response rate from the study authors who we contacted in order to obtain survival data. Our study was limited to the English language, which may contribute to language bias.
Our study's strengths are that, to our knowledge, it is the largest systematic review and meta-analysis to assess survival for oxaliplatin-based HIPEC in PC secondary to CRC. Our pooled estimates also had low heterogeneity among the included studies, making the summary estimates more reliable. Moreover, our results are consistent with the previously published literature on this topic.

Conclusions
This study supplies healthcare providers with a summary of survival results associated with the use of oxaliplatin-based HIPEC in the literature. Having these estimates aids clinicians and their patients in deciding the appropriateness and benefit of oxaliplatin-based HIPEC. The limited number of good-quality prospective RCTs may affect the results. Moreover, assessment of the effect of PCI on DFS and OS of patients receiving oxaliplatin-based HIPEC is likely to help determine the population that would benefit from this regimen. Finally, definitive conclusions await an RCT that compares oxaliplatin-based HIPEC versus MMC-based HIPC versus CRS alone.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.