Abstract
Background: Cytoreductive surgery and hyperthermic intraperitoneal chemotherapy are the treatment of choice for selected patients with peritoneal metastasis. Despite a stringent selection process, some patients were found to be unresectable only at surgery, which leads to disappointment and poor utilisation of limited infrastructural resources. This study aims to determine the pre-operative factors associated with unresectability in planned CRS and HIPEC.
Methods: Retrospective analysis of 172 consecutive patients eligible for CRS and HIPEC at the National Cancer Centre Singapore from April 2004 to May 2014 was performed. Pre-operative factors (clinical presentation, disease factors, and investigation findings) between the unresectable (13%) and the successful groups (87%) were compared.
Results: Patient demographics between the two cohorts were comparable. In terms of clinical presentation, the unresectable group was more likely to present with bloating (p = .00), altered bowel habits (p = .04), abdominal distension (p = .00), palpable abdominal masses (p = .00) and palpable pouch of Douglas nodules (p = .00). Differences were also noted in disease factors with the unresectable group having more high-grade tumours (p = .01), inadequate initial resections (p = .01), progression through chemotherapy (p = .00) and shorter median disease-free intervals (p = .03). In addition, investigations in the unresectable group revealed more patients with elevated tumour markers (p = .01), thrombocytosis (p = .00) and computed tomography findings of ascites (p = .00), omental thickening (p = .00), lymphadenopathy (p = .02) and small bowel disease (p = .00).
Conclusions: Significant factors associated with unresectability that were identified in our study could potentially create a new treatment algorithm and refine current selection process to exclude patients at risk of unresectability in planned CRS and HIPEC.
Introduction
Peritoneal metastasis was once regarded a terminal condition with a median survival of 6 months [Citation1]. Cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) have produced promising results demonstrating an increased in median survival in many different cancers including colorectal and ovarian cancers [Citation2]. CRS and HIPEC are only beneficial in patients whom a complete cytoreduction is achieved [Citation3]. Hence, during surgical exploration, if complete cytoreduction is deemed to be unachievable, CRS and HIPEC will be abandoned resulting in an unnecessary laparotomy. This has resulted in the development of selection criteria (e.g. Peritoneal Carcinomatosis Index (PCI) [Citation4], Pertioneal Surface Disease Severity Score (PSDSS) [Citation5], COREP (ColoREctal-Pc) score [Citation6] and Verwaal prognostic score [Citation7]) to identify patients likely to achieve complete cytoreduction. However, a significant proportion of patients still undergo unnecessary laparotomies [Citation8,Citation9]. It was noted that 20–30% of surgical exploration found lesions unamenable to complete cytoreduction when no contraindications were found on pre-operative exam [Citation10]. Our study aims to determine the pre-operative factors that are significantly associated with unresectability in those planned for CRS and HIPEC.
Material and methods
This study was carried out under the approval of the Centralised Institutional Review Board of Singapore Health Services. Data were retrospectively collected from a prospective database of patients with peritoneal metastasis treated with CRS and HIPEC at the National Cancer Centre Singapore from April 2004 to May 2014. Pre-operative parameters such as patients’ clinical presentations, disease factors and investigations findings were analysed.
Patient selection
Patients referred to our centre had documented peritoneal metastases either on radiological imaging or from previous surgery. All patients underwent pre-operative imaging with a Computed Tomography (CT) scan of the chest, abdomen and pelvis, or a Positron Emission Tomography (PET)-CT scan. The patients’ clinical histories, including details of primary tumour histology and stage, the disease free-interval (DFI), response to chemotherapy and previous surgeries and their imaging findings were reviewed at a multidisciplinary tumour board meeting. Patients recommended to undergo CRS and HIPEC had favourable disease biology, with no distant metastases. In addition, they had to be of Eastern Cooperative Oncology Group (ECOG) status 0 or 1.
Pre-operative factors
Clinical factors included patients’ presentation of bloating, changed in bowel habits, abdominal pain, nausea, vomiting and weight loss of ≥10% of body weight. Disease factors included patients’ response to adjuvant and neoadjuvant chemotherapy (complete, partial, no response or progressive disease on chemotherapy), the need for multiple lines of chemotherapy, the adequacy of prior resection and their disease-free interval. For pre-operative investigations, routine blood investigations and radiological evidence of ascites, omental thickening, lymphadenopathy and small bowel disease on CT imaging were reviewed.
Intra-operative factors
Before commencing with CRS, intra-operative assessment was achieved with an exploratory laparotomy. The extent of small bowel involvement, adhesions, involvement of neurovascular or bony structures especially in the pelvis, presence of liver metastases and the PCI score were evaluated for each patient. A frozen pelvis was defined as pelvis obliterated by tumour deposits and adhesions. The unresectable group was defined as inoperability which was determined intra-operatively that had resulted in the abandonment of planned CRS and HIPEC. The successfully resected group was defined as completion of CRS and HIPEC regardless if complete cytoreduction has been achieved.
Statistical analysis
Odds ratios (OR) with 95% confidence intervals (CI) and p values were estimated using logistic regression. Haldane’s approximation was utilised by adding a constant of 0.5 to each cell if any observed frequencies was 0 [Citation11]. p < .05 was considered statistically significant. Statistical analysis was performed by SPSS version 19.0 (SPSS Inc., Chicago, IL).
Results
From April 2004 to May 2014, there were 172 patients eligible for CRS and HIPEC. Twenty-three patients (13%) were found to be unresectable intra-operatively. The reasons for unresectability () were extensive small bowel disease (15/23, 65%), frozen pelvis (7/23, 30%) and unexpected liver metastases (1/23, 4%). The average PCI score for the unresectable group was 27. There were no significant differences between the unresectable group and the successful group in terms of age (47 vs. 52 years, OR 0.99, CI 0.95–1.02, p = .41), gender (61% vs. 77%, OR 0.46, CI 0.18–1.16, p = .10) and origin of primary tumour (gastrointestinal tumours 61% vs. 49%; gynaecological tumours 26% vs. 40%, OR 0.53, CI 0.19–1.47, p = .22 and primary peritoneal tumours 13% vs. 11%, OR 0.92, CI 0.24–3.56, p = .90) ().
Table 1. Reasons for unresectability.
Table 2. Demographics of the unresectable and the successful group.
With regard to the clinical presentation (), the unresectable group was more likely to complain of bloatedness (70% vs. 23%, OR 7.74, CI 2.92–20.51, p = .00) and altered bowel habits (26% vs. 10%, OR 3.05, CI 1.03–9.01, p = .04). Differences were also noted in physical examination, with abdominal distension (30% vs. 8%, OR 4.92, CI 1.67–14.54, p = .00), palpable abdominal masses (26% vs. 2%, OR 23.47, CI 4.38–125.67, p = .00) and pouch of Douglas nodules (9% vs. 0%, OR 31.51, CI 1.46–679.04, p = .00) being significantly more common in the unresectable group.
Table 3. Clinical presentation in the unresectable and the successful group.
In terms of disease factors (), there was a significant proportion of high-grade tumours in the unresectable group (35% vs. 11%, OR 4.47, CI 1.45–13.85, p = .01). There was no difference in both groups in terms of percentage of patients who underwent neo-adjuvant chemotherapy prior to the planned CRS and HIPEC (70% vs. 70%, OR 1.01, CI 0.39–2.63, p = .98). In patients who received neoadjuvant chemotherapy, the difference in the median number of chemotherapy cycles did not reach statistical significance, although the unresectable group underwent more chemotherapy cycles (11 vs. 6, OR 1.11, CI 0.99–1.24, p = .07). The unresectable group was also found to have a higher proportion of patients who underwent multiple lines of chemotherapy (79% vs. 30%, OR 8.43, CI 2.19–32.42, p = .00) and displayed progression of disease during chemotherapy (50% vs. 4%, OR 42, CI 7.92–222.65, p = .00). The unresectable group had fewer patients who underwent prior surgery (61% vs. 81%, OR 0.38, CI 0.15–0.95, p = .04) and for those who did undergo surgery, they were mostly sub-optimally resected as compared with the successful group (45% vs. 19%, OR 3.54, CI 1.34–9.35, p = .01). Overall, there was also a shorter disease-free interval in the unresectable group, with a median DFI of 5 months in this group, compared with 18 months in the patients who successfully underwent CRS and HIPEC (OR 0.86, CI 0.76–0.98, p = .03).
Table 4. Disease factor in the unresectable and the successful group.
For pre-operative investigations (), elevated tumour markers (90% vs. 50%, OR 8.37, CI 1.85–37.74, p = .01) and thrombocytosis (36% vs. 11%, OR 4.68, CI 1.70–12.87, p = .00), as well as CT scan findings of ascites (100% vs. 31%, OR 93.45, CI 5.51–1584.33, p = .00), omental thickening (95% vs. 18%, OR 92.38, CI 11.74–727.01, p = .00), lymphadenopathy (48% vs. 23%, OR 3.06, CI 1.18–7.99, p = .02) and small bowel disease (29% vs. 4%, OR 9.04, CI 2.46–33.28, p = .00) were also significantly more common in the unresectable group.
Table 5. Pre-operative investigation in the unresectable and the successful group.
Discussion
The Peritoneal Cancer Index (PCI) has been recommended by a panel of experts to be the best quantitative prognostic tool [Citation12]. However, it is scored during laparotomy and is, therefore, not useful to predict unresectability pre-operatively. Less invasive alternatives of PCI include computed tomography-derived PCI (CT-PCI) and laparoscopic-derived PCI. However, CT-PCI underestimated PCI in 33% of cases and failed to avoid unnecessary laparotomy in 12% of cases [Citation13]. On the contrary, laparoscopic evaluation is capable to achieve full laparoscopic PCI assessment with minimal underestimation [Citation14]. However, laparoscopy has risk of port track seeding, cause inadvertent damage due to adhesions from previous surgeries and may fail to assess the retroperitoneal space [Citation15]. Jayakrishnan et al. [Citation16] demonstrated that laparoscopic exploration can exclude 15.1% patients from unnecessary laparotomy due to extensive disease. However, they also experienced 13.7% laparotomic conversions due to poor visualisation from adhesions. This suggests that approximately 50% of laparoscopic assessment may result in laparotomic conversion. Many of the patients referred to our tertiary centre have had multiple laparotomies and attempted resections prior to their planned CRS and HIPEC. Hence, the role of laparoscopy cannot be yet established as a critical component for patient selection in CRS and HIPEC in our centre. In addition, the international consensus in Milan in 2006 did not recommend laparoscopic exploration as an essential investigational modality [Citation17].
Pre-operative selection criteria such as PSDSS [Citation5] and Verwaal prognostic score [Citation7] were established from studies which only included patients who had successful CRS and HIPEC with complete cytoreduction. Prognostic indicators that favoured complete resection or improved survival were identified and subsequently adopted to become the selection criteria. Therefore, these selection criteria may not be sensitive in excluding unresectable cases pre-operatively [Citation8,Citation9]. COREP has found to be superior to PSDSS and Verwaal prognostic score for unresectable case prediction with accuracy of 87% [Citation18]. While COREP score included histopathological variable, haematological counts, serum tumours markers and their changes over time into their calculation, our data have showed that radiological factors may have a role in predicting unresectability as well. Therefore, our findings warrant future research to evaluate if radiological factors can be a complementary factor, or even improve the prediction accuracy for unresectability. Our study recorded outcomes on an intention-to-treat basis, in which all patients who underwent CRS and HIPEC were included regardless of the surgical outcome. In our centre, we experienced 13% of unresectable cases, which was less than the reported figures of 20–30% [Citation10]. One reason may be our stringent criteria for patient selection, with a significant emphasis placed on determining the likelihood of a complete cytoreduction based on radiological imaging. As our experience grew, this method for patient selection resulted in an initial decrease in the percentage of unresectable cases. With the greater acceptance of CRS and HIPEC as a treatment option for peritoneal metastasis, referral rates to our tertiary centre is on the rise, with increasingly complex cases, some of which may have poor tumour biology and we are noticing an upward trend in unresectable cases once again, in the past 4 years from 7% to 20%. We believe that the factors identified from our study have the potential to strengthen current pre-operative selection process to identify unresectable cases.
Clinical presentation
The main reasons for unresectability were extensive small bowel disease and frozen pelvis. Cashin et al. [Citation6] have demonstrated that clinical examination can help exclude patients from CRS and HIPEC and does so to a greater extent than radiological evaluation. As such, we have identified symptoms and signs that may be useful to identify these cases. Altered bowel habits and thrombocytosis have been shown to be present in patients with extensive small bowel disease and frozen pelvises [Citation19]. In our study, clinical findings of abdominal bloatedness, abdominal distension and abdominal mass may be suggestive of underlying extensive small bowel disease [Citation20]. Similarly, it has been suggested that a detailed physical examination may be more superior than imaging studies in frozen pelvis [Citation21]. Besides the presence of nodules in the pouch of Douglas, it may be worthwhile to perform a bimanual vaginal examination to check for fixity pelvic organs as well [Citation22].
Disease factors
We found out that unresectable cases tend to have higher grade tumours, received multiple lines of chemotherapy and progressive disease whilst undergoing chemotherapy before surgery. On top of that, chemotherapy did not add to successful CRS and HIPEC unless there was a complete response. In addition, the unresectable group had more cases of suboptimal initial resection, which may be due to greater intra-peritoneal spillage of tumour emboli resulting in a greater volume of peritoneal disease at the planned CRS and HIPEC surgery [Citation23]. Progressive disease while undergoing systemic chemotherapy indicates biological aggressiveness of the primary tumour [Citation10]. Even in the absence of metastatic disease, high-grade tumours have a higher propensity to be locally invasive, with increased risks of peritoneal seeding of the small bowel serosa resulting in hostile abdomens and frozen pelvises [Citation23]. A study conducted by Passot et al. [Citation24] has found out that progression following neoadjuvant therapy did not have a significant impact on survival and should, therefore, not constitute a contraindication to CRS and HIPEC. However, only successful resection cases were included and thus, no conclusion can be made regarding the impact of progressive disease on unresectability. Our data also showed that the use of neoadjuvant therapy was not a protective factor for unresectability. In fact, the protracted use of neoadjuvant therapy may to put that patient at risk of unresectability. Again, there are limited literature documenting this phenomenon as most studies evaluating the role of neoadjuvant therapy in peritoneal metastasis prior to CRS and HIPEC only included patients who had underwent successful CRS and HIPEC [Citation25–27]. Although the above studies have shown survival benefit and a prognostic role of neoadjuvant therapy, more studies are needed to show that it is not achieved in the expense of increased unresectable cases.
Pre-operative investigation
It has been shown that in advanced peritoneal metastasis, the small bowels may be affected by serosal implants, frank wall invasion and substantial adhesion formations. In addition, ascites formation is common due to tumour cells occluding diaphragmatic lymphatic absorption of peritoneal fluid [Citation28]. This is in keeping with our finding of CT-detected omental thickening, small bowel disease and ascites in the unresectable group. CT-detected lymphadenopathy has also been found to be a significant finding in the unresectable group. However, a study by O’Neill et al. [Citation29] showed CT lymphadenopathy to be a non-specific sign in peritoneal metastasis. The discrepancy may be due to a younger patient cohort in our study (47 years, SD ±12.0) compared to O’Neill’s patient cohort (59.1 years, SD ±14.0). Current evidences have shown that younger cancer patients tended to have more advanced stage disease than older cancer patients which may imply that lymphadenopathy is a sign of advanced peritoneal metastasis [Citation30]. On top of that, we found out that 70% of the lymphadenopathy in the unresectable group was found either at midline (pre-sacral or para-aortic region) or contralateral side. This finding further suggests that lymphadenopathy may be a marker for advanced peritoneal metastasis, as with omental thickening, small bowel disease and ascites, and are reflective of high volume of disease in the peritoneum.
Limitation
This is a retrospective study with a paucity of data in the unresectable group. Due to the small numbers, we were unable to stratify our data according to primary site as the numbers became too small for the results to be meaningful. In addition, the small number of patients also limited the statistical power of the analyses. Further evaluation into each specific primary site will be needed to find out if the findings are applicable to all types of primary tumour (i.e. peritoneal, gastrointestinal and gynaecological).
Conclusion
Our study has identified pre-operative factors in clinical presentation, disease factors and investigation findings that were associated with unresectability. Patients who present with abdominal bloatedness, distension, abdominal mass and pouch of Douglas nodules, and who have high-grade primary tumours precluding optimal resection and require more lines of chemotherapy are at higher risk of having unresectable disease at planned CRS and HIPEC, as are those with obvious radiological involvement of small bowel, regional lymphadenopathy, omental thickening and ascites.
Disclosure statement
The authors report no declaration of interest.
References
- Terzi C, Arslan NC, Canda AE. (2014). Peritoneal carcinomatosis of gastrointestinal tumours: where are we now? World J Gastroenterol 20:14371–80.
- Klaver YL, Simkens LH, Lemmens VE, et al. (2012). Outcomes of colorectal cancer patients with peritoneal carcinomatosis treated with chemotherapy with and without targeted therapy. Eur J Surg Oncol 38:617–23.
- Esquivel J, Sticca R, Sugarbaker P, et al. (2007). Cytoreductive surgery and hyperthermic intraperitoneal chemotherapy in the management of peritoneal surface malignancies of colonic origin: a consensus statement. Soc Surg Oncol. Ann Surg Oncol 14:128–33.
- Jacquet P, Sugarbaker PH. (1996). Clinical research methodologies in diagnosis and staging of patients with peritoneal carcinomatosis. Cancer Treat Res 82:359–74.
- Pelz JO, Chua TC, Esquivel J, et al. (2010). Evaluation of best supportive care and systemic chemotherapy as treatment stratified according to the retrospective peritoneal surface disease severity score (PSDSS) for peritoneal carcinomatosis of colorectal origin. BMC Cancer 10:689.
- Cashin PH, Graf W, Nygren P, Mahteme H. (2012). Patient selection for cytoreductive surgery in colorectal peritoneal carcinomatosis using serum tumour markers: an observational cohort study. Ann Surg 256:1078–83.
- Verwaal VJ, van Tinteren H, van Ruth S, Zoetmulder FA. (2004). Predicting the survival of patients with peritoneal carcinomatosis of colorectal origin treated by aggressive cytoreduction and hyperthermic intraperitoneal chemotherapy. Br J Surg 91:739–46.
- Aguiar S Jr. (2010). Peritoneal carcinomatosis from colorectal cancer: the need for improving selection criteria for cytoreductive surgery and intraperitoneal chemotherapy. J Surg Oncol 102:563–4.
- Yan TD, Sim J, Morris DL. (2007). Selection of patients with colorectal peritoneal carcinomatosis for cytoreductive surgery and perioperative intraperitoneal chemotherapy. Ann Surg Oncol 14:1807–17.
- Cotte E, Passot G, Gilly F-N, Glehen O. (2010). Selection of patients and staging of peritoneal surface malignancies. World J Gastrointest Oncol 2:31–5.
- Sheehe P. (1966). Combination of log relative risk in retrospective studies of disease. Am J Public Health Nations Heath 56:1745–50.
- Portilla AG, Shigeki K, Dario B, Marcello D. (2008). The intraoperative staging systems in the management of peritoneal surface malignancy. J Surg Oncol 98:228–31.
- Esquivel J, Chua TC, Stojadinovic A, et al. (2010). Accuracy and clinical relevance of computed tomography scan interpretation of peritoneal cancer index in colorectal cancer peritoneal carcinomatosis: a multi-institutional study. J Surg Oncol 102:565–70.
- Pomel C, Appleyard TL, Gouy S, et al. (2005). The role of laparoscopy to evaluate candidates for complete cytoreduction of peritoneal carcinomatosis and hyperthermic intraperitoneal chemotherapy. Eur J Surg Oncol 31:540–3.
- Riss S, Mohamed F, Dayal S, et al. (2013). Peritoneal metastases from colorectal cancer: patient selection for cytoreductive surgery and hyperthermic intraperitoneal chemotherapy. Eur J Surg Oncol 39:931–7.
- Jayakrishnan TT, Zacharias AJ, Sharma A, et al. (2014). Role of laparoscopy in patients with peritoneal metastases considered for cytoreductive surgery and hyperthermic intraperitoneal chemotherapy (HIPEC). World J Surg Oncol 21:270.
- Yan TD, Morris DL, Shigeki K, et al. (2008). Preoperative investigations in the management of peritoneal surface malignancy with cytoreductive surgery and perioperative intraperitoneal chemotherapy: expert consensus statement. J Surg Oncol 98:224–7.
- Cashin PH, Graf W, Nygren P, Mahteme H. (2013). Comparison of prognostic scores for patients with colorectal cancer peritoneal metastases treated with cytoreductive surgery and hyperthermic intraperitoneal chemotherapy. Ann Surg Oncol 20:4183–9.
- Nakamoto H. (2005). Encapsulating peritoneal sclerosis – a clinician’s approach to diagnosis and medical treatment. Perit Dial Int 25:S30–8.
- Caldwell J, Dyer A. (2013). Sclerosing encapsulating peritonitis (cocoon bowel) presenting after laparotomy for splenic abscess. J Radiol Case Rep 7:17–23.
- Goldstein DP. (2007). Surgical strategies to untangle a frozen pelvis. OBG Manag 19:65–70.
- Hsu AL, Khachikyan I, Stratton P. (2010). Invasive and non-invasive methods for the diagnosis of endometriosis. Clin Obstet Gynecol 53:413–19.
- Sugarbaker P. (1995). Patient selection and treatment of peritoneal carcinomatosis from colorectal and appendiceal cancer. World J Surg 19:235–40.
- Passot G, Vaudoyer D, Cotte E, et al. (2012). Progression following neoadjuvant systemic chemotherapy may not be a contraindication to a curative approach for colorectal carcinomatosis. Ann Surg 256:125–9.
- Turner KM, Hanna NN, Zhu Y, et al. (2013). Assessment of neoadjuvant chemotherapy on operative parameters and outcome in patients with peritoneal dissemination from high-grade appendiceal cancer. Ann Surg Oncol 20:1068–73.
- Ceelen W, Van Nieuwenhove Y, Putte DV, Pattyn P. (2014). Neoadjuvant chemotherapy with bevacizumab may improve outcome after cytoreduction and hyperthermic intraperitoneal chemoperfusion (HIPEC) for colorectal carcinomatosis. Ann Surg Oncol 21:3023–8.
- Passot G, You B, Boschetti G, et al. (2014). Pathological response to neoadjuvant chemotherapy: a new prognosis tool for the curative management of peritoneal colorectal carcinomatosis. Ann Surg Oncol 21:2608–14.
- Woodward P, Hosseinzadeh K, Saenger J. (2004). From the archives of the AFIP: radiologic staging of ovarian carcinoma with pathologic correlation. Radiographics 24:225–46.
- O'Neill AC, Shinagare AB, Rosenthal MH, et al. (2014). Differences in CT features of peritoneal carcinomatosis, sarcomatosis, and lymphomatosis: retrospective analysis of 122 cases at a tertiary cancer institution. Clin Radiol 69:1219–27.
- O’Connell JB, Maggard MA, Liu JH, et al. (2004). Do young colon cancer patients have worse outcomes? World J Surg 28:558–62.