Biphasic learning curve of cytoreductive surgery and hyperthermic intraperitoneal chemotherapy:technical competence and refinement of patient selection

Introduction

Cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) have achieved good long-term results in treating peritoneal metastasis from colorectal [1], ovarian [2], appendiceal [3], mesothelioma [4] and primary peritoneal neoplasms. An increase in survival durations has been reported at the expense of considerable perioperative morbidity and mortality although patients are able to return to baseline or improved quality of life by 6–12 months [5]. This high complication rate has been attributed to the steep learning curve associated with the procedure, although even in several high volume centers, the complication rate remains high and overall is comparable to major gastrointestinal surgery [6].

Major morbidity in CRS/HIPEC varies widely 12 [7]–67.6 % [8], with a median of 31 % reported from an analysis of several studies in the literature [9]. Rates of perioperative mortality vary from 0 % [10] to 9 % [11], with a median of 4 % [9].

Traditionally the two key parts of the learning curve assessed are an increase in rates of complete cytoreduction and decreases in procedure-related serious complications or mortality. Estimates from previous studies put the number of cases required to reach surgical proficiency at 130–140 [12, 13, 14]. Kusamura et al. reported in 2012 that 140 cases are necessary to ensure surgical proficiency in CRS/HIPEC on the basis of completeness of cytoreduction and serious morbidity [13], and in 2013 reported a similar number required in a second center [12]. Similarly, Smeenk et al. reported a peak in the learning curve after around 130 procedures had been performed out of a total cohort of 323 on the basis of completeness of cytoreduction [14].

The demonstration from initial studies that the maximal benefit of CRS/HIPEC can only be achieved with complete resection (completeness of cytoreduction score 0 or 1, representing no nodules>0.5 cm remaining) [15, 16] has established this as an important factor in patient selection, placing the emphasis of learning curve on procedure-related serious complications or mortality.

Materials and methods

Results

Discussion

CRS/HIPEC is a relatively recent development in the treatment of peritoneal metastasis, and has demonstrated an increase in survival times at the expense of considerable perioperative morbidity and mortality. Developments in this technique have largely focused on the selection criteria utilized for patients. The demonstration of initial studies that the maximal benefit of CRS/HIPEC can only be achieved with complete resection [15, 16], has been an important factor driving evolution of selection criteria. Improvements have also been made in the perioperative management of patients suggesting that it is not just the learning curve of surgical technique which is important in reducing the high morbidity associated with CRS/HIPEC. In contrast to previous studies which have focused solely on rates of incomplete cytoreduction, mortality and serious morbidity to assess learning curve, we have identified a secondary component in the learning curve related to trends of patient selection and perioperative management.

In our institution, the extent of intra-abdominal disease and its amenability to complete CRS is determined at a multidisciplinary tumor board and with expert radiologists present. Only patients who have absence of extra-abdominal metastases and a high likelihood of complete CRS are subjected to the procedure, and in line with the literature, our practice is not to administer HIPEC if optimal CRS cannot be performed.

In order to assess the learning curve of CRS/HIPEC, we have examined trends in patient selection, technical competence and perioperative management in our institution, representing the largest Asian cohort to date.

Trends in primary tumor selection included an increase, followed by subsequent decrease in the last cohort in tumors presenting with positive nodal status (20 %, 24 %, 48 %, 18 %, p<0.01), whilst there was an increase in the final cohort of patients found to have metastatic disease at initial presentation (i.e. patients presenting for CRS and HIPEC with synchronous peritoneal metastasis) (29, 34, 35, 59 %, p=0.012). There was no significant difference in T-stage across the cohorts (p=0.47). These trends reflect development of a hesitancy about aggressive locoregional surgical treatment in those with node positive tumors, as such tumors carry a generally higher risk of failure at distant sites. However, there is a recognition of the role of CRS and HIPEC in patients who present with limited synchronous peritoneal metastasis at the time of their diagnosis, although it has traditionally been thought that patients with synchronous metastases tended to have poorer tumor biology compared to those who developed their metastases in a metachronous fashion. We are increasingly seeing referrals for such patients resulting in the increased proportion of patients presenting with synchronous metastases across the cohorts.

As peritoneal disease often responds poorly to systemic chemotherapy, there is a risk that staged resection, with CRS and HIPEC planned for after pseudo-neoadjuvant chemotherapy after resection of the primary tumor, may result in the loss of the window of opportunity for patients who present with synchronous peritoneal metastasis. We have adopted the strategy to administer a short course of pseudo-neoadjuvant chemotherapy, to assess for favorable tumor biology (response or stable disease) and absence of new abdominal or distant metastases prior to CRS/HIPEC, particularly in patients presenting with synchronous metastases.

Although frequency of subdiaphragmatic stripping decreased across the study (86 %, 72 %, 38 %, 46 %, p<0.01), there was a nonsignificant increase in overall PCI score (average 8, 14, 12, 13). One explanation is that the distribution of disease may play a role rather than overall extent. The primary tumor type did not change significantly in the first three cohorts, although in the final cohort there was an increased proportion of primary peritoneal (18 %) and other primary tumor types (20 %) (p<0.01). Distribution of disease may reflect the cases that are being referred from gynecology and colorectal surgeons. In the early cohorts, frequent requirement for subdiaphragmatic stripping may reflect a tendency for referrals for disease in the upper abdomen, outside of the normal operative zone of the respective surgeons, whilst they resect pelvic peritoneal lesions themselves. With increasing awareness of the role of CRS/HIPEC, referrals are likely to reflect patients with known peritoneal disease, regardless of the location of the disease.

In this study we observed a decrease in two outcomes over the cohorts. A significant decrease in the primary outcome of serious complications was observed only after 100 consecutive operations (34, 30, 12, 14 % in consecutive cohorts). However, a significant decrease in the secondary outcome of duration of CRS/HIPEC was observed after just the first 50 operations (average of 600, 470, 470, 430 min in consecutive cohorts). Another way of putting this is that there is a decrease in time taken per PCI after the first cohort (1.25, 0.56, 0.65, 0.55 h taken per PCI, or 1.125, 0.49, 0.57, 0.47 h/PCI excluding HIPEC time). This measure remains similar after the first cohort, suggesting that this measure reflects improvement in technical expertise rather than patient selection.

Previous studies have reported that around 130–140 procedures are required to reach competence with CRS/HIPEC [12, 13, 14], hence we expected to see an improvement in outcomes by the third cohort of 50 patients. However previous studies have largely focused on rates of incomplete cytoreduction and serious morbidity as surrogates of surgical competency, although these are closely related to patient selection. The practice at our institution to only proceed with CRS/HIPEC if a CC0 or 1 resection is possible limits the comparison to rates of serious complications, with a significant drop most apparent after the first 100 cases (34, 30, 12, 14 %) similar to previously reported learning curves.

A proportion of this decrease is due to changes in management made as a result of increased experience. After the first two cohorts, we ceased early postoperative intraperitoneal chemotherapy (EPIC; 68, 72, 0, 0 %) due to the high rate of intra-abdominal collections and following this, were able to demonstrate a significant decrease in the rate of intra-abdominal collections (10, 22, 2, 2 %). We also began to routinely insert chest tubes intraoperatively for those patients undergoing subdiaphragmatic stripping to avoid the development of postoperative pleural effusion requiring insertion of a chest tube at the bedside, that may explain the decrease in respiratory complications seen after the initial cohort (30, 14, 6, 20 %), which was limited to patients with extensive disease in the last cohort (median PCI 18 vs. 5 in the first cohort). A nonsignificant decrease in rate of acute renal impairment (29, 31, 16, 13 %) was seen after the second cohort and may reflect increased awareness of this complication and optimization of postoperative monitoring and fluid management in response to this.

Finally, a decreased rate of ICU admission (82, 76, 66, 46 %) may reflect increased surgeon experience resulting in decreased operation time and bleeding as well as anesthetist experience with intraoperative fluid resuscitation in this complex combined modality treatment of CRS and HIPEC.

Conclusions

In our experience the learning curve for CRS/HIPEC consisted of two phases; the first 50 cases were adequate to achieve surgical technical familiarity with the procedure as demonstrated by decreased average operation time and a further 50 case for a significant reduction in the rate of serious morbidity. We propose that this represents an initial period of training in which technical competence, reflected by a decreased duration of the procedure is achieved followed by a subsequent period of training resulting in a significant gain in experience, leading to better selection and perioperative management of the patients, and less serious morbidity despite the increasingly complex cases being subjected to CRS/HIPEC.