Introduction

Recurrent epithelial ovarian cancer (REOC) occurs in about 70 % after primary treatment consisting of radical surgical resection and platin-based systemic chemotherapy. Treatment protocols at this point of the disease are mainly based on palliative chemotherapy [1–3].

An alternative approach is repetitive surgery which is up to now restricted to clinically symptomatic patients. In patients with REOC and peritoneal metastases (PM), secondary radical surgery, however, provided a similar outcome compared with patients suffering from recurrent disease without PM as long as complete cytoreduction could be achieved. Moreover, in this particular patient population, survival was mainly dependent on the completeness of cytoreduction [4, 5] supporting again the pivotal role of radical surgery in the treatment of REOC.

It remains unclear, however, whether simultaneous hyperthermic intraperitoneal chemotherapy as adjunct to radical surgery could further improve prognosis in REOC. In prospective trials, normothermic intraperitoneal chemotherapy has already shown promising results prolonging overall survival in primary advanced ovarian cancer [6–8].

Radical cytoreductive surgery (CRS) along with hyperthermic intraperitoneal chemotherapy (HIPEC) revolutionized treatment in various stage IV cancers [9, 10]. There is a growing body of retrospective data favourizing this multimodal strategy [11–17] also in epithelial ovarian cancer even though it is still discussed controversially. At least in Germany, many gynaecological oncologists discourage from using CRS and HIPEC outside of well-designed, prospective and controlled clinical trials claiming this multimodal procedure to be associated with increased morbidity in particular due to the HIPEC itself [18]. Recently, several prospective trials in primary and recurrent ovarian cancer have been set up to evaluate the beneficial effect of HIPEC (www.clinicaltrials.gov).

The objective of our retrospective analysis was to evaluate morbidity, mortality and postoperative outcomes of CRS and HIPEC in REOC with PM.

Patients and methods

Patient selection

Between February 2007 and June 2012, 90 consecutive patients with REOC and PM were enrolled in the Peritoneal Surface Malignancy Program at the University of Tübingen, Germany. Preoperative diagnostics consisted of a clinical examination, blood test and CT scan and laparoscopy in some cases to rule out distant metastases and local irresectability, i.e. infiltration of the mesenteric axis, retroperitoneum or diffuse extensive tumour growth on the entire small bowel surface. Eligibility for CRS and HIPEC was assessed by a surgical oncologist, a gynaecological oncologist, a radiologist and a clinical pathologist attending a weekly interdisciplinary oncologic team meeting presenting patient demographics and imaging results.

Patients were retrospectively categorized into two groups with respect to the completeness of cytoreduction (CC-0/1 vs CC-2/3).

Surgical procedure

After laparotomy through a midline incision, complete adhesiolysis and exploration of the entire abdominal cavity, the peritoneal carcinomatosis index (PCI) was determined following the criteria described by Sugarbaker et al. [19, 20].

If PM were interpreted to be potentially resectable, cytoreductive surgery was started aiming for complete cytoreduction (CC-0, CC-1; CC-0 meaning no visible disease; CC-1 meaning nodules smaller than 0.25 cm).

After complete cytoreduction and fashioning of intestinal anastomoses, if necessary, HIPEC with cisplatin 50 mg/m2 for 90 min at 42 °C using the open coliseum technique was administered.

Postoperative morbidity and mortality

Adverse events were categorized according to the Clavien-Dindo (CD) complication score [21]. Grade 1 was defined as any deviation from the normal postoperative course, and grade 2 required pharmacological treatment. In grade 3 complications, there was a need for radiological, endoscopic or surgical intervention. Life-threatening complications were classified as grade 4 and death as grade 5. Surgery-related mortality was defined as death within 30 days of surgery. Overall complication rate was defined as the percentage of patients developing as least one CD I-V adverse event within the postoperative period.

Statistics

Data are presented as median (min-max) or n (%) unless otherwise stated. Qualitative differences were compared using Χ 2 test and quantitative differences using Mann-Whitney U test. Survival analysis was performed by the Kaplan-Meier method. For overall survival (OS), time to event was calculated as time from cytoreductive surgery until death or time to last contact, if the patient was alive. Recurrence was calculated from the date of surgery to disease relapse or to the last known date of follow-up evaluation or date of death using the Kaplan-Meier method. A p value less than 0.05 was considered significant. SPSS version 13.0 software (SPSS, Chicago, Illinois, USA) was used for all statistical analysis.

Results

Operative outcomes

A total of 90 patients were scheduled to cytoreductive surgery and HIPEC when there was evidence of peritoneal disease without signs of irresectability. Complete cytoreduction and HIPEC could be performed in 62 patients (69 %) (CC-0 47 patients (52 %), CC-1 15 patients (17 %)), whereas in 5 patients (6 %), residual tumour volume was greater than 0.25 cm (CC-2) located either in the hepatoduodenal ligament or the mesenteric axis. Twenty-three patients (25 %) only underwent explorative laparotomy or tumour debulking either due to retraction of the mesenteric axis or diffuse extensive tumour growth on the entire small bowel surface. Eight of these 23 patients (35 %) have been operated on for symptomatic bowel obstruction or to prevent mechanical ileus.

Patients were retrospectively categorized with respect to the completeness of cytoreduction (CC-0/1 vs CC-2/3). Both groups were comparable for baseline demographic characteristics except intra-abdominal tumour load (PCI), time in operating room and postoperative hospital stay (Table 1).

Table 1 Demographic characteristics
Full size table

Types of visceral resections are shown in Table 2. Patients with complete cytoreduction received more visceral resections as opposed to patients with incomplete cytoreductive surgery. In particular, there was a higher incidence of rectosigmoid resection, greater omentectomy, diaphragma resection and splenectomy. In total, an ostomy had to be created in seven patients.

Table 2 Visceral resections
Full size table

Morbidity and mortality

Thirty-eight patients suffered from postoperative adverse events resulting in an overall complication rate (Clavien-Dindo I-V) of 42 %. There was no significant difference in morbidity rates between CC-0/1 and CC-2/3 patients (28 patients (45 %) vs 10 patients (36 %); p = 0,254) even though there was a trend to a higher frequency of CD ≥3 adverse events in the CC-0/1 group (p = 0.086). Re-laparotomy had to be performed in nine patients (15 %) in the CC-0/1 as opposed to one patient (4 %) in the CC-2/3 group. Following complete cytoreduction and HIPEC, four patients (6 %) suffered from postoperative pneumonia and seven patients (11 %) from pulmonary embolism, whereas these adverse events were not observed after exploratory laparotomy (Table 3).

Table 3 Postoperative adverse events classified according to Clavien-Dindo (CD)
Full size table

There was no surgery-related and 90-day postoperative mortality.

Survival

Median follow-up was 19 months [2–62]. In patients with complete cytoreduction (CC-0/1), median overall survival was calculated to 35 months (95 % CI 23–46) as opposed to 14 months (95 % CI 4–25) in patients with incomplete cytoreduction (CC-2/3) (Fig. 1a). In addition, there was no difference in survival with respect to PCI as long as complete cytoreduction could be achieved (Fig. 1b).

Fig. 1
figure 1

a Median overall survival was calculated to 35 months (95 % CI 23–46) as opposed to 14 months (95 % CI 4–25) in patients with incomplete cytoreduction (CC-2/3). b In addition, there was no difference in survival with respect to PCI as long as complete cytoreduction could be achieved

Full size image

Discussion

CRS and HIPEC are a multimodal treatment strategy for intra-abdominal tumour spread without organ metastases or extra-abdominal seedings currently utilized for various gastrointestinal cancers especially colorectal cancer [9, 10]. This therapy protocol is based on the rationale that the peritoneal surface compartment is considered a natural barrier against systemic tumour spread limiting tumour growth to the peritoneal cavity. While radical surgery is intended for eliminating macroscopic disease, HIPEC is supposed to erase non-visible remaining tumour or free tumour cells.

Standard of care for recurrent ovarian cancer with PM, so far, is palliative systemic chemotherapy, whereas surgery is only considered in case of mechanical bowel obstruction. The multicentre DESKTOP I trial for recurrent ovarian cancer, however, suggested radical debulking surgery as preferable treatment strategy as long as complete cytoreduction can be achieved [4, 5].

Since phase III trials are not available, HIPEC is still considered as “experimental” in ovarian cancer and is restricted to selected individual cases or interventional institutional studies [11–17]. Some gynaecological oncologists argue that HIPEC itself causes major morbidity postponing systemic chemotherapy in case of severe postoperative adverse events. In our series, overall morbidity was 42 % which is accordance to the current literature where morbidity ranges between 20 and 66 %, mainly due to the extent of surgery [11, 12, 14, 22]. Even though our re-operation rate was 15 %, there was neither surgery-related death nor any patient died within 90 days from surgery. Interestingly, there was no significant difference in the overall cumulative complication rate between patients who underwent radical cytoreductive surgery plus HIPEC compared with those who did not even though visceral resections were much more frequent in the CC-0/1 group. The trend to a higher frequency of Clavien-Dindo ≥3 adverse events is more likely related to the extent of surgery rather than the HIPEC procedure itself. Despite of numerous visceral resections, ostomy had to be created in only six patients (10 %).

Four patients (6 %) developed transient leucopenia (<2,000/μl) following HIPEC which resolved spontaneously with no need for granulocyte colony-stimulating factors. In addition, we did not find more infectious complications such as surgical site infections following HIPEC. HIPEC itself—at least according to our data—does not seem to necessarily increase morbidity not allowing early postoperative chemotherapy.

We found a median overall survival of 35 months (95 % CI 23–46) in patients with complete cytoreduction plus HIPEC as opposed to 14 months (95 % CI 4–25) in patients with incomplete cytoreduction (CC-2/3). Recently, Bakrin et al. demonstrated a 45.7-month overall survival in a French retrospective multicentre cohort study [11]. Median overall survival for chemotherapy alone in EORC was reported to be as high as 29 months [1] or 18 months [2]. It remains unclear, however, how many patients suffered from PM in these particular investigations. In the DESKTOP I trial, Harter et al. reported a median overall survival of 19.9 months following cytoreductive surgery without additional HIPEC in patients suffering from REOC with PM. In this study, however, complete cytoreduction could only be achieved in 26 % of patients [4, 5]. Those particular patients showed a significant increase in overall survival with a 2-year survival rate of 77 % even without additional HIPEC. Compared with our study, however, PM were defined as present according to the surgical report [5] without any quantification by PCI, and the extent of surgery for achieving complete cytoreduction was not reported. In addition, approximately 90 % of patients received postoperative chemotherapy which has been proven by the authors as an independent prognostic factor by multivariate analysis [4].

In summary, complete cytoreduction seems to be the major prognostic factor in REOC for long-term outcome [4, 5]. Even in primary ovarian cancer, progression-free and overall survivals were dependent on surgical radicalness and could be prolonged by more radical surgery exploring the entire abdominal cavity [23].

In our series, we were able to achieve complete cytoreduction in 69 % of the patients. However, we did not find appropriate clinical selection criteria to anticipate the probability of complete cytoreduction. Selecting these particular patients in whom complete cytoreduction is likely to be achieved remains a major challenge for the future.

Interestingly, PCI did not impact survival in our series as long as complete cytoreduction could be achieved. This is in discordance to a recently published multicentre paper where Bakrin et al. observed that patients with a PCI ≤8 had a significant better survival than patients with a PCI >8. However, the percentage of “incomplete cytoreduction” in the PCI >8 group is not documented and may lead to bias in this study. Therefore, at least in our opinion, we should not a priori exclude patients with high intra-abdominal tumour load from the opportunity of radical cytoreductive surgery.

Most importantly, patients should be treated by surgeons and gynaecological oncologists who are familiar with cytoreductive/visceral surgery in specialized centres. In order to reduce morbidity, we prefer an organ-preserving approach rather than radical organ resections. Recently, our group was able to demonstrate that colon-preserving surgery along with meticulous “cleaning” of the bowel instead of colectomy in REOC can keep morbidity down at an acceptable level [24].

Due to the retrospective evaluation of our experience, however, our results have to be interpreted with caution, and as previously mentioned, the beneficial effect of HIPEC remains to be proven. However, our series is one of the largest single centre reports on REOC showing again that this multimodal option can be offered in patients with acceptable morbidity.

Pushing surgical efforts seem to result in favourable survival data, and organ-preserving surgery may diminish morbidity. The impact of HIPEC on non-visible disease remains to be evaluated in prospective trials. We are urgently awaiting the results of the CHIPOR study [25]. This randomized controlled phase III compares cytoreductive surgery and HIPEC to cytoreductive surgery alone with respect to overall survival. In addition, preoperative neoadjuvant chemotherapy using either carboplatin-paclitaxel or carboplatin-caelyx is performed in both arms testing the hypothesis that adjunction of HIPEC in REOC is able to improve the median overall survival by 12 months.