Cytoreductive surgery (peritonectomy procedures) combined with hyperthermic intraperitoneal chemotherapy (HIPEC) in the treatment of diffuse peritoneal carcinomatosis from ovarian cancer

Authors


Abstract

BACKGROUND.

Because of scarce data from larger series and nonhomogeneous selection criteria, further information is needed on peritonectomy with hyperthermic intraperitoneal chemotherapy (HIPEC) in managing patients with ovarian peritoneal carcinomatosis.

METHODS.

In an open, prospective, single-center nonrandomized phase 2 study conducted from November 2000 to April 2007, 47 patients with primary advanced or recurrent ovarian cancer and diffuse peritoneal carcinomatosis were enrolled; 22 underwent primary and 25 secondary cytoreduction plus immediate HIPEC followed by systemic chemotherapy.

RESULTS.

The overall mean Sugarbaker peritoneal cancer index was 14.9 (range, 6–28). A mean of 6 surgical procedures were required per patient (range, 4–10). In 87.3% of the patients debulking achieved optimal cytoreduction (Sugarbaker completeness of cytoreduction [CC] score 0–1), whereas in 12.7% it left macroscopic residual disease (CC-2 or CC-3). Major complications developed in 21.3% of the patients and the in-hospital mortality rate was 4.2%. The mean overall survival was 30.4 months, median survival was 24 months, and mean disease-free survival was 27.4 months. Five-year survival was 16.7%. Univariate (log-rank test and analysis of variance) and multivariate analyses (Cox proportional-hazard model) identified the CC score as the main factor capable of independently influencing survival.

CONCLUSIONS.

Peritonectomy procedures combined with HIPEC offer promising long-term survival in patients with diffuse peritoneal ovarian carcinomatosis. They achieve high adequate primary and secondary surgical cytoreduction rates with acceptable morbidity and mortality. Cancer 2008. © 2008 American Cancer Society.

Peritoneal carcinomatosis is a typical feature of cancer spread in patients with primary advanced or recurrent epithelial ovarian cancers. Ascites and locoregional lymph node involvement are frequently associated with peritoneal carcinomatosis, whereas distant metastases appear later.

Although ovarian cancer is 1 of the most chemotherapy-sensitive solid tumors and 1 of the few in which the 5-year survival rate has improved in recent years,1 in most women with locally advanced disease long-term survival remains poor.2 Standard therapy for advanced ovarian cancer consists of 2 steps: cytoreductive surgery followed by adjuvant platinum and taxane chemotherapy. Although this combined therapy initially seems effective, insofar as high rates of patients achieve a good response,3 about half of the patients relapse within 5 years and disease-free survival rarely exceeds 18 months.4, 5

Since the first report of Griffith in 1975,6 multiple retrospective series have shown that survival time is inversely proportional to residual tumor size,7, 8 and a recent meta-analysis confirms that 1 of the most powerful determinants of cohort survival among patients in stage III or IV ovarian cancer is maximal cytoreduction.9 Although maximal tumor debulking has now achieved consensus for women with advanced ovarian cancer,10–12 important questions remain about the complex strategy for treating diffuse ovarian peritoneal carcinomatosis, a condition characterized by severe macroscopic disease involving the upper and lower abdominal quadrants.12, 13

In other nonovarian carcinomatoses, including cancer of the appendix and colon, maximum cytoreduction, namely, peritonectomy according to the Sugarbaker criteria,14 combined with hyperthermic intraperitoneal chemotherapy (HIPEC), is a promising locoregional treatment that offers acceptable morbidity and mortality rates and the possibility of long survival.15, 16 Although peritonectomy procedures with HIPEC found favor later in ovarian carcinomatosis than in other carcinomatoses, in recent years this therapeutic strategy has attracted growing interest.17–24 The discrepant results from the various case series nevertheless leave several questions unclear, most important, the selection of patients according to the disease setting (primary or secondary cytoreduction) and extent of carcinomatosis, criteria for classifying intraperitoneal disease spread, and surgical techniques.

In this study, we reviewed our prospective experience in a consecutive series of patients with primary advanced or recurrent ovarian cancer and diffuse peritoneal carcinomatosis who underwent extensive tumor debulking surgery aimed at maximal surgical cytoreduction with peritonectomy procedures and HIPEC. We focused our attention on patient selection criteria, surgical techniques, and immediate and long-term results.

MATERIALS AND METHODS

The study was an open, prospective, single-center nonrandomized phase 2 study conducted at the Department of Surgery, Pietro Valdoni, University La Sapienza of Rome, Italy, between November 2000 and April 2007. Patients with primary advanced or recurrent ovarian cancer (TNM-FIGO stages IIIc-IV) and evidence of peritoneal diffuse carcinomatosis were considered eligible. The inclusion criteria were age younger than 75 years; a histologically or cytologically confirmed diagnosis; performance status 0–2 (WHO)25; adequate cardiac, renal, hepatic, and bone marrow function; resectable disease; and informed written consent. The exclusion criteria were extraabdominal metastases; other malignancies except breast cancer; unresectable disease; and active infections or severe associated medical conditions.

The treatment plan envisaged extensive surgical cytoreduction aimed at removing all visible disease plus immediate HIPEC and adjuvant systemic chemotherapy according to the patient's general status. The study protocol was approved by the Hospital Medical Ethical Committee.

Surgical Procedure (Peritonectomy)

At laparotomy the extent of peritoneal carcinomatosis was recorded according to the peritoneal cancer index (PCI).26 Aggressive surgical cytoreduction was then undertaken to leave the patient with no visible disease. One or more peritonectomy procedures14 were required, depending on the distribution and volume of peritoneal carcinomatosis. Small numbers of scattered peritoneal implants were resected or ablated with high-voltage electrocautery, radiofrequency technology (TissueLink BPS 6.0, Dover, NH), or an argon beam coagulator. Bulky disease such as confluent implants involving an extensive area were usually removed with 1 of the major peritonectomy procedures.

The completeness of cytoreduction (CC) was scored as proposed by Sugarbaker: CC0: no residual disease; CC1: residual nodules measuring less than 2.5 mm; CC2: residual nodules measuring between 2.5 mm and 2.5 cm; and CC3: residual nodules greater than 2.5 cm.27

HIPEC

At the end of each surgical procedure, HIPEC was given with the closed technique. Four drains were positioned for HIPEC inflow/outflow and temperature monitoring. The inflow/outflow drains were connected to a closed extracorporeal sterile circuit in which a 4 to 6 L perfusate was circulated by means of a peristaltic pump at a flow rate of 500 mL/min. The closed sterile circuit was heated by means of a thermal exchanger connected to the heating circuit (EXIPER, Euromedical, Italy).

HIPEC at temperatures ranging from 42–43°C was given for 60 minutes with cisplatin at a dose of 75 mg/m2. Trendelenburg/anti-Trendelenburg and laterolateral inclinations of the recumbent patient were changed every 5 minutes to guarantee that the entire endoabdominal surface was perfused. After the procedure the abdomen was washed with 3–4 L of sterile saline solution at 37°C.

Morbidity and Toxicity

During the immediate postoperative period patients were assisted in an intensive care unit (ICU) for at least 24 hours. Cisplatin toxicity was scored using the WHO toxicity grading scale for chemotherapy.25 Treatment-related morbidity and mortality were classified as grade 1: no complication; grade 2: minor complications; grade 3: major complications requiring reoperation, ICU admission, or interventional radiology; and grade 4: in-hospital mortality.28

Follow-up

After hospital discharge, patients were referred to the medical oncologic staff to plan systemic chemotherapy. Patients were followed up every 6 months with abdominal computed tomography (CT) scan, positron emission tomography (PET) scan, measurement of serum markers (Ca 125 and Ca 199), and any further evaluation indicated by the patients' clinical presentation.

Survival and Statistical Analysis

Differences between groups of observations were analyzed by a chi-square test. The Kaplan-Meier method was used to construct survival curves and a log-rank test was used to assess the significance of the differences between curves. Survival was also expressed as mean and median. Analysis of variance (ANOVA) was used to assess the statistical significance of differences in mean survival between groups. The Cox regression model was used to determine the prognostic value of independent variables. P-values <0.05 were considered statistically significant. The NCSS (Kaysville, Utah) package was used to analyze the database and perform statistical tests.

RESULTS

Patients

Forty-seven patients with a median age of 60.3 years (range, 32–75) with primary advanced or recurrent ovarian cancer and evidence of diffuse peritoneal carcinomatosis were prospectively enrolled in the study from November 2000 to April 2007. Four more patients initially enrolled underwent a palliative procedure alone and are not considered in this report.

Of the 47 patients enrolled, 22 (46.8%) underwent primary cytoreduction; 4 of them received neoadjuvant chemotherapy with cisplatin and paclitaxel and 2 had previously undergone ileostomy for intestinal occlusion elsewhere. The other 25 patients underwent secondary cytoreduction at 4 to 312 months after the last operation (mean, 52.4 months; median, 26 months): 18 patients for recurrent disease, 4 patients as interval debulking, and 3 as second-look surgery.

All 25 patients undergoing secondary cytoreduction had already undergone 1 or more operations for ovarian cancer (range, 1–4 laparotomies) plus chemotherapy (range, 1–3 lines). Two patients had undergone 6 laparotomies either for ovarian cancer or for other abdominal nonneoplastic diseases. Patient characteristics are summarized in Table 1.

Table 1. Patients' Demographic and Clinical Characteristics (47 Patients)
VariablesNo. (%)
Age, y, mean [range]60.3 [32–75]
Disease setting
 Primary cytoreduction:22 (46.8)
 Secondary cytoreduction:25 (53.2)
Previous chemotherapy:
 No18 (38.3)
 Yes29 (61.7)
  Neoadjuvant4
  Adjuvant25
Performance status (WHO)
 018 (38.3)
 113 (27.6)
 216 (34.1)
Intestinal obstruction
 Absent32 (68.1)
 Present15 (31.9)
AscitesNo. (%)
 Absent18 (38.3)
 Present29 (61.7)
Comorbidity
 Absent37 (78.7)
 Present10 (21.3)
Ca125 level, U/mL, mean [range]528.4 (20–6800)

Surgical Procedure

The mean PCI calculated according to Sugarbaker criteria was 14.9 (range, 6–28) for the study sample overall, without differences between primary or secondary cytoreduction. The 47 patients treated required a total 305 surgical procedures, mean 6.4 per patient (range, 4–10) (Table 2).

Table 2. Surgical Treatment
Type of resectionNo. primary cytoreductionNo. secondary cytoreductionTotal no.
Hysterectomy ± adnexectomy22325
Pelvic mass resection 77
Omental resection221840
Intestinal resection
 Colorectal resection17421
 Total colectomy with rectal resection51015
 Other colic resection246
 Appendicectomy10212
 Small bowel resection6713
 Gastric resection11
Splenectomy61218
Cholecystectomy5712
Liver resection123
Total cystectomy1 1
Bladder resection 22
Total peritonectomy369
Partial peritonectomy191534
Abdominal wall resection336
Resection or reduction cancer implants132235
Paraortic and pelvic lymphadenectomy22830
Lymphadenectomy in other sites9615
Total166139305
Mean no. of procedures (range)7.4 (4–10)5.6 (4–9)6.4 (4–10)

In patients managed by en bloc rectal resection with hysterectomy-adnexectomy or removal of the recurrent pelvic mass, intestinal continuity was restored later, after the end of adjuvant chemotherapy and an adequate follow-up.

All patients who underwent primary cytoreduction also had systematic pelvic and paraaortic lymphadenectomy. Of the 25 patients who underwent secondary cytoreduction only 8 underwent these procedures. In 15 patients from the 2 groups lymphadenectomy included other sites, especially for enlarged mesenteric, splenic hilus, and splenic and hepatic artery lymph nodes.

In 41 of the 47 patients debulking achieved optimal cytoreduction (scored CC-0 or CC-1), whereas in 6 patients (12.8%) it left macroscopic residual disease (scored CC-2 or CC-3). In all patients who had a low PCI (less than or equal to 15) debulking achieved complete cytoreduction, whereas in patients with PCI >15 debulking achieved optimal cytoreduction in 76% (P = .0006, chi-square test). Primary and secondary cytoreduction procedures yielded similar CC scores (Table 3).

Table 3. Impact of Peritoneal Cancer Index and Disease Setting on Completeness of Cytoreduction
CC scorePCI≤15*PCI>15*Primary cytoreductionSecondary cytoreduction
No. of cases (%)No. of cases (%)No. of cases (%)No. of cases (%)
  • PCI indicates peritoneal cancer index; CC score, completeness of cytoreduction.

  • *

    Chi-square test P = .0006.

  • Chi-square test P was not significant.

020 (90.9)8 (32)15 (68.2)15 (60.0)
12 (9.1)11 (44)5 (22.7)7 (28.0)
204 (16)1 (4.5)2 (8.0)
302 (8)1 (4.5)1 (4.0)
Total22252225

The mean overall duration of the operative procedure was 528 minutes (range, 180-780) including HIPEC. Primary cytoreduction took longer than secondary cytoreduction (600 minutes vs 465). The mean blood loss was 1494 mL (range, 100–4900); a mean 3.4 blood units and 5.2 plasma units were transfused. The mean ICU stay was 40.1 hours (range, 12–96) and the mean hospital stay 22.1 days, and differed slightly for primary and secondary cytoreduction (Table 4).

Table 4. Surgical Outcome
ProceduresNo. [range] (%)No. of complicationsTreatment
Duration of procedure, min528 [180–780)
 Primary cytoreduction600 [300–780]
 Secondary cytoreduction465 [180–720]
Blood loss, mL1494 [100–4900]
 Primary cytoreduction1638 [600–4900]
 Secondary cytoreduction1368 [100–3100]
Blood transfusion, units3.4 [0–8]
Plasma transfusion, units5.2 [2–10]
ICU stay, h40.1 [12–96]
Morbidity 
 Grade 124 (51.1)
 Grade 210 (21.3)4 pleural effusion
4 wound infection
2 transient ischemic attacks TIA
 Grade 310 (21.3)3 Small or large bowel fistulaSurgery
2 Endoperitoneal bleedingSurgery
1 EventrationSurgery
1 Gastric bleedingInterventional radiology
1 Urinary fistulaInterventional radiology
1 Deep venous thrombosisInterventional radiology
1 Myocardial infarctionIntensive care unit
 Grade 42 (4.2)2 Pulmonary embolism
Postoperative stay, d22.1 [8–93]
 Primary cytoreduction23.6 [9–90]
 Secondary cytoreduction20.8 [8–93]

HIPEC

Of the total 47 patients in the series, 45 underwent HIPEC at the end of surgery. Two patients did not undergo HIPEC owing to their unstable postoperative hemodynamic conditions and underwent a 5-day course of postoperative intraperitoneal unheated chemotherapy after ICU discharge. The mean time to reach a temperature higher than 40°C in the abdominal cavity was 10 minutes and the intraperitoneal temperature invariably remained over 41°C and mainly over 42°C (mean, 42.7°C, range, 41.0–43.3°C). None of the HIPEC procedures led to the development of thermal intolerance. The only adverse event related to HIPEC was cisplatin toxicity (2 cases). The major cisplatin-related renal toxicity was grade 1 (WHO) in 1 patient and grade 2 in the other; medical treatment reversed both drug-induced reactions.

Morbidity and Mortality

Of the 47 patients treated, 24 (51.1%) had no complications, 10 (21.3%) had minor complications (grade 2) mainly accounted for by pleural effusion and wound infection, and 10 patients (21.3%) had grade 3 complications: in 6 patients requiring reoperation, in 3 an interventional radiologic procedure, and in 1 patient ICU readmission. Two patients died of pulmonary embolisms during the postoperative course, accounting for an in-hospital mortality of 4.2% (Table 4).

Histologic Features

In most of the 47 patients treated (66%) histologic analysis of the resected specimen disclosed serous adenocarcinomas; 19.1% had mucinous carcinomas, and the remaining 14.9% had endometrioid carcinomas. In many patients (61.7%) lymph nodes removed during lymphadenectomy and cytoreduction were free of metastases. Of the 47 patients, 18 (38.3%) had lymph-node metastases; 8.6% had metastases in locoregional nodes, 17% in nonlocoregional stations alone, and 12.7% in both locoregional and extraregional stations according to TNM and FIGO staging29 (Table 5).

Table 5. Histology
Histological subtypesNo. of cases (%)
Serous carcinoma31 (66)
 Mucinous carcinoma9 (19.1)
 Endometrioid carcinoma7 (14.9)
Lymph node status
 Negative29 (61.7)
 Positive18 (38.3)
  Locoregional4 (8.6)
  Extraregional8 (17)
  Loco + extraregional6 (12.7)

Follow-up and Survival

Assessment at discharge showed that most patients (95.6%) had a satisfactory performance status (equal to or less than 2). In only 2 cases did the patient's poor general condition contraindicate the postoperative systemic chemotherapy foreseen in the study protocol.

No patient was lost to follow-up. Of the 45 patients who survived after the surgical procedure, 23 (51.1%) are alive and disease-free, 3 (6.7%) are alive with disease, 16 (35.5%) died of disease-related causes, and 3 patients (6.7%) died of causes unrelated to cancer.

The mean overall survival was 30.4 months (standard deviation [SD] 22.1 months; range, 6–83); median survival was 24 months. Mean disease-free survival was 27.4 months (SD 23.2, range, 0–8); median disease-free survival was 20 months. The overall 5-year survival was 16.7% (Fig. 1).

Figure 1.

Overall 5-year survival.

In the univariate analysis, the only 2 factors that were associated with prognosis were CC score and age. These 2 prognostic variables yielded significantly different distribution curves (CC score, P = .02; and age P = .01 by log-rank test) (Figs. 2, 3) and correlated significantly with mean survival (ANOVA) (Table 6). The analysis of the sample stratified into 3 groups according to CC scores and into 2 groups according to age identified as the best predictors of survival a CC score of 0 and an age younger than 60 years. No significant differences were found in survival according to primary or secondary surgical cytoreduction, the presence or absence of neoplastic ascites, presence or absence of intestinal obstruction, Ca 125 level, PCI, number of surgical procedures, histology, or lymph node status (Table 6).

Figure 2.

Five-year survival and completeness of cytoreduction (CC) score (P = .0247).

Figure 3.

Five-year survival and age.

Table 6. Univariate Analyses of Prognostic Factors and Survival
CharacteristcsMonths of overall survivalMonths of disease-free survivalP
Mean(Median)PANOVAMean(Median)PANOVALog-rank test
  1. PCI indicates peritoneal cancer index; CC score, completeness of cytoreduction; ANOVA, analysis of variance; ns, not significant.

Age,y  .008  .009.01
 ≤6040.6(32) 37.9(27)  
 >6022.8(16) 19.6(12)  
Primary cytoreduction32.2(27)ns29.7(25.5)nsNs
Secondary cytoreduction28.5(22.5) 24.9(15.5)  
Obstruction  ns  nsns
 Yes30.2(26.5) 25(21)  
 No30.5(23.5) 28.6(17)  
Ascites  ns  nsns
 Yes28(21.3) 24.6(16)  
 No33.9(23.6) 31.6(24)  
CA 125  ns  nsns
 ≤50032.7(25) 30.5(24)  
 >50020.9(20.5) 14.4(14)  
PCI  ns  nsns
 ≤1530(24) 27.6(21)  
 >1530.9(26) 27.2(20)  
No. of procedures  ns  nsns
 ≤630.7(24) 28.2(22)  
 >630.1(24) 26.5(18)  
CC 036.8(26).00334.6(24).002.02
CC 119.7(13) 16.2(13)  
CC 2–316.7(12) 9.7(6)  
Histology  ns  nsns
 Serous carcinoma28.4(24) 24.8(21)  
 Mucinous carcinoma41.1(37) 40.1(37)  
 Endometrioid carcinoma24.3(19) 20.7(12)  
Lymph-node status  ns  nsns
 N030.9(24) 28.2(22)  
 N+29.7(22.5) 26.2(20)  

In the multivariate analysis the Cox proportional-hazard model used to test the simultaneous effect on survival of primary or secondary surgical cytoreduction, CC score, lymph node status, PCI score, and age, the only prognostic factor capable of independently influencing survival was the CC score (Table 7).

Table 7. Cox Regression Model of Prognostic Factors and Survival
VariableDeviance (Chi-square)P
  1. CC score indicates completeness of cytoreduction; ns, not significant.

Primary cytoreduction/ secondary cytoreduction2.7150ns
CC score7.2289.0072
Lymph node status2.9063ns
Peritoneal Cancer Index2.2495ns
Age3.4791ns

DISCUSSION

Our findings from this review of a consecutive series of 47 patients with primary advanced or recurrent ovarian cancer with diffuse peritoneal carcinomatosis who underwent extensive tumor debulking surgery aimed at maximal surgical cytoreduction (peritonectomy procedures) plus HIPEC suggest that this procedure is a reasonable evidence-based therapeutic approach. Our study underlines that according to the disease setting, peritonectomy plus HIPEC seems equally warranted for primary and secondary cytoreduction. Until now this combined procedure was reserved mainly for patients undergoing secondary cytoreduction because it was merely considered as salvage treatment in patients with recurrent or chemotherapy-resistant ovarian cancer. Nevertheless, in agreement with others,24, 30 considering the high recurrence rates after standard treatment for advanced ovarian cancer, and the good results the combined procedures achieved in our series and in others, we suggest that maximal cytoreduction (peritonectomy procedures) plus HIPEC should now be the up-front treatment for selected patients with ovarian carcinomatosis who have diffuse peritoneal spread. When we analyzed the main patient selection criteria—age, performance status, comorbidity, and extent of disease—we found that, as others have underlined, patients aged older than 75 years with severely compromised performance status or severe comorbidity are generally unsuitable to undergo peritonectomy plus HIPEC.18, 21, 23 Diagnostic techniques are still lacking that will reliably select patients with extensive peritoneal carcinomatosis suitable for surgery. The previously reported criteria for patient selection include clinical and radiographic characteristics ranging from a general description of unresectable disease to specific disease locations or clinical conditions and genetic profile.31, 32 In this series, unlike others, to assess the extent of disease and predict the surgical outcome we never used diagnostic laparoscopy33 because we consider it of no help in selecting patients to undergo peritonectomy plus HIPEC and may even bias the decision in favor of inappropriate nonexploration, thus reserving operation for the less problematic cases. Although most of our patients had extensive ovarian carcinomatosis (mean PCI 14.9), in 87.2% of the patients aggressive debulking surgery achieved satisfactory cytoreduction (CC-0/CC-1) with a rate of complete cytoreduction (CC0) of 59.6%. The degree of cytoreduction we obtained ranks high among that from published series in patients with advanced or recurrent ovarian cancer treated with standard techniques or with peritonectomy plus HIPEC.12, 13, 17–24, 34–36 Our experience in ovarian carcinomatosis suggests that although a PCI lower than 15 is a determinant factor in obtaining optimal cytoreduction, a PCI higher than 15 does not invariably preclude satisfactory surgical results (Table 3). Our opinion therefore goes against those who recommend a strict threshold value, considering that, as Eisenkop et al.34 observed in ovarian cancer, prognosis seems to be influenced less by initial tumor volume and more by the extent of cytoreduction achieved. Our experience in this series adds useful personal technical considerations specific to ovarian carcinomatosis. During pelvic peritonectomy according to Sugarbaker,14 which combines en bloc removal of the female reproductive organs, pouch of Douglas, and rectum, we always extended parietal peritonectomy to the transverse umbilical line. For several reasons we rarely restored intestinal continuity with the rectal stump immediately, during peritonectomy procedures and HIPEC. First, because most of our patients presented with intestinal subocclusion that precluded proper preoperative large bowel preparation, we had to limit the development of complications in patients who had undergone especially aggressive interventions often entailing numerous anastomoses and in whom the protocol envisaged HIPEC. Besides, because the pelvis is the main site of malignant recurrence we preferred to opt for a terminal colostomy in the iliac fossa and postpone restoring intestinal continuity for a second look. In most patients we waited for at least 6 months after postperitonectomy systemic chemotherapy before restoring intestinal continuity. In 15 patients in our series extensive large bowel carcinomatosis necessitated a total colectomy.

An especially important technical point concerns the best strategy for treating the small bowel: the number of resections and the extension of resection need careful thought to limit functional damage and allow adequate debulking. Ileal tracts less severely involved by the malignancy are more amenable to in situ treatment by means of argon beam laser, electrovaporization, or radiofrequency technology (TissueLink, BPS 6.0; Salient Surgical Technologies, Dover, NH). In most patients in this series, extensive tumor spread necessitated upper abdominal procedures to obtain optimal debulking and in all cases envisaged complete greater omentectomy, and never infracolic omentectomy alone. In all the patients whom we operated on for recurrence and who had undergone surgery elsewhere, laparotomy disclosed carcinomatosis in the residual omentum. Precisely when debulking surgery should include removal of the spleen remains unclear.37 In our experience, apart from patients with evident direct tumor involvement of the spleen, we undertook splenectomy whenever macroscopic malignant implants involved the gastrosplenic ligament. In these cases possible lymph-node metastases should be sought also in the splenic hilum (11.1% in our case series).

According to others, the presence of liver metastases seems not per se sufficient to contraindicate cytoreduction.38 In our experience in a patient who underwent secondary cytoreduction a liver resection (left hepatic lobectomy) provided an excellent outcome. The patient is alive and has been disease-free for 83 months.

In this series we routinely removed pelvic and paraaortic lymph nodes in all patients who underwent primary cytoreduction and in many patients who had secondary cytoreduction. An analysis of the site of nodal metastases nonetheless shows how diffuse peritoneal spread of ovarian malignancies favors extensive lymph node diffusion spreading well beyond the regional drainage to involve extraregional nodes. Most patients who had lymph node metastases (77.8%) had them in extraregional sites with or without associated regional nodes.

Considering the numerous surgical procedures for each patient (6.4 per patient) and the lengthy duration of peritonectomy procedures plus HIPEC (mean, 528 minutes) the number of major complications and operative mortality rate is reasonably low and comes within the range reported by other similar case series17–24 (Table 8). The most frequent major complications (grade 3) were intestinal fistulas owing to the large number of multiple intestinal resections required, bowel-wall weakening during maneuvers to destroy malignant implants in situ, and in our series the increased risk of applying closed HIPEC after completing surgery and therefore after anastomoses had been constructed.

Table 8. International Experience With Cytoreduction Plus HIPEC in Advanced Ovarian Cancer
AuthorsYearNo. of casesDisease settingFollow-up, mo% PCI% CC score% Major complications% Mortality% 5-Year SurvivalMean overall and DF survival, mo
  • PCI indicates peritoneal cancer index; NR, not reported; CC score, completeness of cytoreduction; DF, disease-free.

  • *

    Median.

  • Disease-free.

  • At 3 years.

  • §

    Primary.

  • Recurrent.

Look17200428Primary/Secondary0.5–79.19.8 (0–26)CC 0–1, 57110NR45.8
26.9*CC 2–3, 4317.3
Zanon18200430Primary/Secondary2–68.2<12, 44.3CC 0–1, 7716.73.30NR28.1
18.9*≥12, 56.717.1
Piso19200419Primary/Secondary24*<15, 47.3CC 0/1, 47.42851533
≥15, 52.7CC 2/3, 52.6
Reichman20200513Primary/Secondary6–366 (2–16)CC 0, 38.5NR011NR
13.7*CC 2/3, 61.5
Rufian21200633Primary/Secondary7–91NRR0, 5236037§NR
R1≤1 cm, 3351
R2>1 cm, 15
Raspaglieli22200640Secondary0.3–117.6NRCC 0/1, 83501541.4
26.1*23.9
Cotte23200781Secondary9–23711.5 (1–30)CC 0, 55.613.62.50NR28.4
47.1*CC 1, 24.719.2
CC 3, 19.7
Helm24200718Secondary30NRCC 0, 61.1≥22.25.5NR31
CC 1, 22.210
CC 2/3, 16.7
Current series200747Primary/Secondary0.5–8314.9 (6–28)CC 0, 59.621.34.2016.7030.4
<15, 46.8CC 1, 27.727.4
<15, 53.2CC 2/3, 12.7

Two patients died during the postoperative course, 1 on Day 2 and the other on Day 5, both of massive pulmonary embolisms. The risk of this complication developing is high even despite routine preventive treatment. Hence, to reduce mortality from pulmonary embolism and prevent these complications in future we intend to use temporary inferior vena cava filters especially for patients undergoing more aggressive surgical debulking or longer procedures. Most of our patients had a long postoperative course (about 20 days) and the duration correlated with the extent of surgical resection. Conversely, the relatively short mean ICU stay provided evidence that strict patient selection and proper surgical and anesthesiology techniques can keep the procedural-related risks within acceptable limits. Here we underline the importance of having an experienced operating team. All the interventions in our series were done by the same surgeon (A.D.) with the same surgical and anesthesiology team.

When we analyzed our long-term results, the survival analysis showed that the overall 5-year survival was 16.7%, similar to that of most investigators who reported treating patients with widespread peritoneal carcinomatosis (IIIc) and used cytoreduction surgery plus HIPEC. The mean survival in our series also matches the rate in published series reporting a mean survival ranging from 28.1 to 45.8 months (Table 8). The major discrepancies between our prognostic results and those of others arise from differences in the length of follow-up and in initial selection of patients, particularly the extent of the endoperitoneal disease (PCI). Our patients' PCI values were among the highest reported in the literature, with a mean of 15, ranging from 6 to 28. Those who reported considerably higher survival values than we did included in their series patients staged lower than IIIc FIGO21 or made no mention of how carcinomatosis was classified.22

The prognostic indicators tested in our univariate analysis that correlated significantly with survival were the degree of cytoreduction and the patient's age. Cox regression analysis, including along with the CC score and patient's age other prognostic factors such as lymph node status, PCI, and primary or secondary surgical cytoreduction, identified as the most significant factor in line with other series the completeness of cytoreduction.9, 39 Unlike others a prognostic factor that weakens significance in our multivariate analysis is age.39 Another factor that seems to weigh less heavily on prognosis is the disease setting. Patients who underwent secondary cytoreduction for recurrent disease fared as well as those who underwent primary cytoreduction because we achieved a similar degree of cytoreduction in the 2 groups. Accordingly, as the univariate and multivariate analyses showed, the CC score, the most significant outcome variable, was equally distributed in the 2 groups (Table 3).

Conclusions

In conclusion, peritonectomy procedures combined with HIPEC provide a promising approach for long-term survival in patients with diffuse peritoneal ovarian carcinomatosis. The combined procedure appears effective both for primary and secondary surgical cytoreduction. The rate of adequate cytoreduction is high. Morbidity and mortality rates are acceptable. The role of neoadjuvant systemic chemotherapy in primary cytoreduction remains to be investigated in prospective trials.

Acknowledgements

The authors thank the “Azienda Policlinico Umberto I” and “Eleonora Lorillard Spencer Cenci Foundation” for economic support and the “B.T. Medical srl” for technical assistance.

Ancillary