Abstract
Aim: To assess the incidence of morbidity and mortality of Cytoreductive Surgery plus Hyperthermic Intraperitoneal Chemotherapy. Patients and Methods: A retrospective multicentric study was performed. Six hundred and eighty-three patients were recorded. Predictors of morbidity and mortality were evaluated with univariate and multivariate analysis. Results: In univariate analysis, older age, Eastern Cooperative Oncology Group score, a greater value of Peritoneal Cancer Index (PCI) and sub-optimal cytoreduction were correlated with higher mortality, while older age, presence of ascites, ovarian origin of carcinomatosis, closed technique, a greater value of PCI, longer operative time and sub-optimal cytoreduction were predictors of higher morbidity. In multivariate analysis, older age and a greater value of PCI were correlated with higher mortality; older age, ovarian origin of tumor, presence of ascites, closed technique and longer operative time were predictors of higher morbidity. Conclusion: Careful patient selection has to be performed to improve clinical outcomes.
Cytoreductive surgery (CRS) plus Hyperthermic Intraperitoneal Chemotherapy (HIPEC) represent a promising approach to treat peritoneal surface malignancies (PSM) (1-3). Actually, mortality and morbidity rates remain the most comprehensive measures used to assess short-term outcomes of a specific procedure. Surgical complications are frequently the main reason to modify patient treatment. Nevertheless, nowadays these two outcome measures remain hard to assess and identify by the surgical literature for the lack of standardization and under-reporting. For this reason, presently, in the surgical field, there is an absence of a clear system to classify complications. These considerations are valid also for CRS plus HIPEC. Furthermore, the issue is relatively complex because the appearance of postoperative complications related to surgery can be confounded with toxic side effects of the intraperitoneal chemotherapy.
To provide a substantial contribution to this field, we carried out a multicentre retrospective study. We report on the preliminary analysis of this experience.
Patients and Methods
A retrospective multicentre study from eight Italian Centres was performed. Six hundred and eighty-three patients submitted in the November 2000 – January 2014 period to CRS-plus-HIPEC, were recorded.
Two centres recorded more than 100 cases, four centres between 50 to 100 and two centres less than 50. To satisfy the statistical criteria, only 507 patients were enrolled. Morbidity was evaluated in accordance with the National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events (CTCAE) v4.0.
Predictors age, gender, body-mass index (BMI), primary tumour site, Eastern Cooperative Oncology Group (ECOG) score, ascites, neoadjuvant chemotherapy (NACT), HIPEC technique, Peritoneal Cancer Index (PCI), operative time, Completeness of Cytoreduction score (CC-score) of morbidity and mortality were evaluated with univariate and multivariate analysis. p-Values <0.05 were considered statistically significant.
Results
The mean age of the patients was 56.6 years standard deviation (SD)±11.5; 83% were female. Four hundred and sixty-one (90.9%) patients had peritoneal carcinomatosis, 293 were of ovarian, 106 of colon and 62 of gastric origin; 46 (9.1%) patients had other peritoneal malignancies. The ECOG score was 0 for 193 patients (38.1 %), 1 for 150 (29.6%), 2 for 136 (26.8%) and 3 for 28 (5.5%). One hundred and seventy-four patients (50.7%) had preoperative ascites. Average BMI was 24.9 (±4.6). Ninety-three patients (18.3%) were submitted to NACT. The average PCI was 12 (±9.4). The closed abdomen technique was used in 396 cases (78.1%) and the Coliseum technique in 111 (21.9%). Three hundred and eighty-seven (76.3%) patients did not have visible residual disease (CC-0 score), 73 (14.4%) had residual disease ≤2.5 mm (CC-1 score), 29 (5.7%) had residual disease between 2.5 mm and 2.5 cm (CC-2 score) and 18 (3.6%) had residual disease >2.5 cm (CC-3 score). The average operative time was 504 min (±156,2) (Table I). The mean stay in the intensive care unit was 2.1 days (range=0-18 days). Mean hospital stay was 16.5 days (range=1-95 days). Grade 2-4 morbidity occurred in 280 patients (55.2%). Between the recorded complications (Table II), the most frequent were: pleural effusion (10.2%), anemia (9.8%), pancreatitis (9.4%), nausea (8%), postoperative ileus (7.6%), renal deficiency (6.9%), sepsis (5.3%), wound infections (4.9%) and intestinal perforations (4.3%). Overall re-operation percentage was 7.2% (intestinal leakage: 18 cases, haemoperitoneum: 14 cases, abdominal abscess: 3 cases and colostomy dehiscence: 2 cases). Mortality occurred in 18 cases (3.6%). The causes of death were: 1 Wernicke syndrome, 1 pulmonary embolism, 3 haemoperitoneum, 5 cardio-respiratory failure, and 8 multi-organ failure (5 correlated with an intestinal fistula, 2 determined by sepsis and 1 post-haemoperitoneum). In univariate analysis, the older age (p=0.027), the ECOG score 3 vs. 0 (p=0.016), a greater value of PCI (p=0.014) and CC-score 2 vs. 0 (p=0.044) were correlated with a higher mortality, while the older age (p<0.01), the presence of ascites (p<0.01), the ovarian origin of carcinomatosis (p<0.01), the closed abdomen technique (p=0.015), a greater value of PCI (p=0.014), longer operative time (p<0.01) and CC-score 1 vs. 0 (p=0.022) and 2 vs. 0 (p=0.043) were predictors of higher morbidity. In multivariate analysis the older age (p=0.047) and a greater value of PCI (p=0.026) were correlated with a higher mortality; the older age (p<0.01), the ovarian origin of tumour (p<0.01), the presence of ascites (p=0.011), the closed abdomen technique (p<0.01) and longer operative time (p<0.01) were predictors of higher morbidity (Tables III and IV).
Discussion
PSM has been regarded as inoperable condition and treated by systemic chemotherapy or palliative therapies. Based on the theory that PSM is a locoregional disease, CRS-plus-HIPEC have been applied in selected patients with these pathologies. Chua et al. (4) have showed that the morbidity and mortality of CRS and HIPEC were similar to other major gastrointestinal interventions. In literature, major morbidity rates have been reported ranging from 12% to 57% in high-volume centres (4-5), even if, in recent publications, the overall grade III-IV morbidity rates are shown to be between 7 and 41% (6-11). These results are due to the learning curve that has been demonstrated to be necessary to minimize mortality and morbidity after the procedure (12, 13). Instead, regarding the mortality, it is reported that it ranges from 0.9% to 11% (4, 14, 15).
Among the numerous classifications used to report morbidity, some investigators have preferred to use the Clavien system (16), others the Feldman or modified Clavien system (17), still others the Elias classification (18), the Bozzetti classification (19) or the CTCAE (20). This heterogeneity in the adopted systems impedes the comparison between treatment-related complication rates among the various reports. All these classification systems, beyond to be different from each other, are no intended, except CTCAE, to account for toxicities related to the use of chemotherapy during HIPEC. For this reason, Kusamura proposed to use the World Health Organization toxicity scale (6). Others authors, Elias (18), Smeenek (21) and Glehen (7), prefer to use the CTCAE version 3.0 (20) that, in 2006, during the 5th International Workshop on Peritoneal Surface Malignancy (22), was adopted by an international panel of experts as the definitive classification system for complications resulting from CRS plus HIPEC. In 2009, an updated version of CTCAE called version 4.0 was published (20).
There are many risk factors, as described in the literature, associated with the appearance of complications related to CRS plus HIPEC. However, it is not possible to make a comparison between the various series because of the numerous variables used by the different authors. Thus, we report the most representative series in this field: Kusamura et al. (6), in their report, by a univariate analysis, found that the following variables were proven to have a statistically significant correlation with major morbidity: male gender, ECOG score, no previous systemic chemotherapy, carcinomatosis extension, number of bowel anastomoses >2, duration of procedure, extent of cytoreduction and dose of CDDP >240 mg. In multivariate analysis, no previous systemic chemotherapy, extent of cytoreduction and fractionated dose of cisplatin (CDDP) >240 mg were independent risk factors for major morbidity. Glehen et al. (7) have carried out only a univariate analysis reporting that major morbidity was statistically linked with the carcinomatosis stage, duration of surgery and number of resections and of peritonectomy procedures. Hansson et al. (8), instead, in their multivariate analysis, have observed that morbidity was associated with stoma formation, duration of surgery, perioperative blood loss and PCI. Casado-Adam et al. (9), in their univariate analysis, have found a statistically significant correlation between morbidity, histological grade, PCI, small bowel resections, colorectal anastomosis and the number of anastomoses performed; their multivariate analysis shown that PCI was the only independent risk factor for gastrointestinal complications. Finally, Mizumoto et al. (10), in their univariate analysis, have showed that PCI greater than 20, operation time longer than 5 h and blood loss greater than 2.5 l were significant risk factors for the occurrence of postoperative complications. Multivariate analysis showed that PCI higher than 20 was the only significant factor which increased the occurrence of postoperative complications. A PCI greater than 20 was associated with 2.8-times increased risk of postoperative complications.
The analysis of our experience documented that mortality and morbidity rates of CRS-plus-HIPEC are acceptable and within the literature range. Morbidity was correlated with some predictive factors already documented in the literature (6, 9, 10), like PCI, operation time and extension of cytoreduction in univariate analysis, with the operative time being the predictive factor for the multivariate one. Instead, original results are represented by the statistical correlation, both on uni- and multivariate evaluation, between morbidity and the older age of patients, the presence of preoperative ascites, the ovarian origin of carcinomatosis and the closed abdomen technique used to perform the HIPEC.
While the presence of ascites, indirect index of diffuse disease, should be assimilable to the role of the PCI, statistically significant also in the multivariate analyses of Hasson (8), Casado-Adam (9) and Mizumoto (10), the older age (23, 24), the ovarian origin of tumor and HIPEC technique are not supported by the literature. We retain that the higher percentage of morbidity correlated with closed abdomen technique is particularly interesting, because this finding has never been reported in the literature. We think that further analyses are useful to understand if the lower complication rates correlated to the Coliseum technique are dependent on the technique or the two samples of patients have different characteristics.
Concerning mortality, we did not find any items against which to compare our results, therefore we believe that they can be considered a landmark in this topic. In particular, mortality was statistically correlated, in univariate analysis, with older age of patients, higher ECOG score, greater value of PCI and CC score 3 vs. 0 and, in multivariate one, only with older age and higher PCI.
In conclusion, in light of this preliminary analysis, we retain that, to improve short-term outcome of patients submitted to CRS-plus-HIPEC, an optimal selection is necessary, referring especially to PCI and age, with particular care to the performance status.
With respect to the higher morbidity and mortality correlated to an incomplete cytoreduction, we retain that it is mandatory to carry-out the preoperative staging as accurately as possible, considering the possibility to include, in the diagnostic algorithm of doubtful cases, the exploratory laparoscopy approach.
Our intention is to subsequently provide more detailed evaluation of the role of the HIPEC technique after performing a more accurate assessment of the two groups of patients subjected to open or closed approach.
- Received June 26, 2014.
- Revision received July 29, 2014.
- Accepted July 30, 2014.
- Copyright© 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved