The American Journal of Surgery (2014) 207, 760-765
North Pacific Surgical Association
Preoperative computed tomography does not predict resectability in peritoneal carcinomatosis Justin D. Rivard, M.D., F.R.C.S.C.*, Walley J. Temple, M.D., F.R.C.S.C., F.A.C.S., Yarrow J. McConnell, M.D., F.R.C.S.C., Hisham Sultan, M.D., Lloyd A. Mack, M.D., F.R.C.S.C., F.A.C.S. Department of Surgery and Oncology, University of Calgary, Calgary, Alberta, Canada
KEYWORDS: Peritoneal carcinomatosis; Cytoreduction; Computed tomography; Resectability
Abstract BACKGROUND: Obtaining a complete cytoreduction in patients with peritoneal carcinomatosis (PC) is one of the most significant prognostic variables for long-term survival. This study explored features on preoperative computed tomography (CT) to predict unresectability. METHODS: A retrospective case–control study was conducted of 15 patients with unresectable PC and 15 patients with completely resected PC matched by intraoperative peritoneal cancer index (PCI) and pathology type. Two surgical oncologists blindly analyzed all abdominopelvic CT scans. RESULTS: PCI estimated on imaging was not higher in unresectable patients (P 5 .851) and significantly underestimated intraoperative PCI measurement (P 5 .003). No single concerning feature was associated with unresectability. However, patients with 2 or more concerning features were more likely to be unresectable (87.5% vs 36.4%, P 5 .035). CONCLUSIONS: Two or more concerning CT imaging features appear to be associated with a higher risk of unresectability in patients with PC. However, no specific imaging feature should exclude a patient from an attempted cytoreduction. Ó 2014 Elsevier Inc. All rights reserved.
Peritoneal carcinomatosis (PC) occurs in 10% to 40% of gastrointestinal malignancies either at initial presentation or upon disease recurrence.1–3 Encouraging progress has been made in recent years toward the treatment of patients with peritoneal metastasis with the use of cytoreduction surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC). One randomized trial and several prospective cohort
There were no relevant financial relationships or any sources of support in the form of grants, equipment, or drugs. * Corresponding author. Tel.: 11-403-521-3070; fax: 11-403-5213744. E-mail address: [email protected] Manuscript received October 28, 2013; revised manuscript December 21, 2013 0002-9610/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjsurg.2013.12.024
trials have shown improvements in disease-free and overall survival for those patients undergoing CRS 1 HIPEC over palliative surgery and/or systemic chemotherapy alone.4–7 However, CRS 1 HIPEC has been associated with considerable morbidity. Additionally, despite attempting to select which patients to perform CRS 1 HIPEC on, there is an unresectability rate ranging from 7% to 25%.4,5,8 Therefore, identifying which patients would most likely benefit from this surgical approach in the preoperative setting is desirable. The Peritoneal Cancer Index (PCI) is a scoring system that quantifies the severity of PC. The extent of disease is scored on a scale with a maximum value of 39 by dividing the abdomen into 13 distinct regions, as described by Harmon and Sugarbaker.9 The PCI has been shown to
J.D. Rivard et al.
Preoperative CT does not predict resectability
predict the surgical team’s ability to perform a complete cytoreduction, which is known to be one of the principal prognostic variables in survival. Patients who undergo a complete cytoreduction with appendiceal mucinous neoplasms and colorectal cancer have been recently shown to have 5-year survival rates of 74% to 85% and 17% to 45%, respectively.4,6,10 This is compared with 5-year survival rates of 24% and 0% for an incomplete cytoreduction in appendiceal mucinous neoplasms10 and colorectal patients.4 Furthermore, PCI has been shown to be an independent prognostic indicator for long-term survival,11–13 even in the presence of a complete cytoreduction.8,13 Preoperative staging is most commonly performed using computed tomography (CT) scanning. A recent international workshop on peritoneal surface malignancy devised an expert consensus statement concluding that contrastenhanced CT scans are the preferred modality to assess a patient’s suitability for CRS.14 However, CT has been identified as a poor predictor of intraoperative PCI (oPCI),15–17 particularly for lesions smaller than .5 cm where the sensitivity has been reported between 11% and 28% compared with 90% and 94% for lesions larger than 5 cm.15,17 The purpose of this study is to identify whether there are any features present on preoperative abdominopelvic CT imaging that can accurately predict unresectability in patients with PC.
761
0 to 39 as reported by Harmon and Sugarbaker.9 PCI was not used as a selection criterion for which patients would undergo an attempted CRS. However, those patients who were found at laparotomy to have disease involving the retroperitoneal or periaortic lymph nodes, invasive disease involving portal triad, extensive small bowel involvement, unexpected liver metastasis, or other unresectable structures did not undergo CRS 1 HIPEC. All cytoreductive surgery included resection of all visible peritoneal tumor deposits, except where such resections would compromise vital functions. At the completion of surgery, the CCR score was recorded as described by Harmon and Sugarbaker9: CCR0 (no residual macroscopic tumor); CCR-1 (residual tumor deposits ,2.5 mm in diameter); and CCR-2 (residual tumor deposits .2.5 mm in diameter). Patients with a CCR-2 score did not undergo intraperitoneal chemotherapy.
Radiologic evaluation
Between 2007 and 2011, 15 consecutive patients with PC who were determined to be unresectable (completeness of cytoreduction score 2 (CCR-2)) after intraoperative evaluation were case matched to 15 patients with PC who underwent a complete cytoreduction (CCR-0). Cases were matched based on oPCI and pathologic type. All patients were operated on at the University of Calgary by 1 of 2 surgical oncologists (W.T., L.M.). Preoperatively, those patients with PC and a good functional status with no evidence of distant disease were considered for treatment with CRS 1 HIPEC. Preoperative investigations included clinical examination; complete colonoscopy; CT scans of the chest, abdomen, and pelvis; and whole body positron emission tomography/CT scans. Most patients were discussed at a multidisciplinary tumor board before surgery. All patients before surgery provided informed consent. Data regarding patient, tumor, and perioperative treatment factors were collected and maintained in a prospective database. The Alberta Cancer Research Ethics Committee approved this study.
The included patients’ preoperative abdominal and pelvic CT scans enhanced with intravenous and oral contrast were interpreted blindly by 2 experienced surgical oncologists who each perform more than 30 cytoreductive surgeries per year. CT scans had been performed at different sites as many patients had imaging investigations before being referred to our hospital. All images had been reconstructed at 5-mm intervals and were read at a digital workstation. Median time between CT scan and cytoreductive surgery was 2 months and 9 days. If the surgeon felt the patient was a candidate for surgery at the initial surgical consult based on the above criteria, a repeat CT scan closer to the surgery date was not required. After blindly reviewing each patient’s CT scans in random order, the surgeons calculated the estimated PCI, which was calculated using the same scale to calculate the oPCI. They also documented whether there were any other concerning factors identified (bowel obstruction, ascites, retroperitoneal lymphadenopathy, pelvic side wall involvement, portal involvement, bilateral hydroureter, liver parenchymal metastases, lung metastases) and whether they felt the patient was technically resectable (technical consensus). The likelihood of resectability was scored on a Likert scale with values ranging from 0 to 5, with 0 being not resectable and 5 being definitely resectable. Following this, the surgeons were given patient and tumor characteristics including age, primary pathology, disease-free interval, administration of neoadjuvant systemic chemotherapy, and response to chemotherapy. With the additional information, the surgeons used the same 5-point scale to record whether they felt the patient was resectable and if they would attempt a cytoreductive surgery (clinical consensus).
Cytoreductive surgery
Statistical analysis
For all patients undergoing attempted CRS, the extent of PC was scored intraoperatively using the PCI on a scale from
Descriptive analyses are presented as medians and ranges for continuous data and frequencies and percentages
Methods Patient selection
762
The American Journal of Surgery, Vol 207, No 5, May 2014
for categorical variables. Comparisons were made using the Mann–Whitney U test and Wilcoxon signed-rank test for quantitative data and Fisher’s exact test for categorical data. All tests were 2-sided with a significance level set at P , .05. All analysis was performed using SPSS version 21.0 (IBM Corp, Armonk, NY).
Results Patients A total of 30 patients were selected for this study. The unresectable and resectable cohorts were each made up of 15 patients and were case-matched by PCI and pathology. Each group consisted of 8 colorectal cancer, 1 high-grade appendix cancer, 3 low-grade appendix cancer, 1 mesothelioma, and 2 appendix adenocarcinoid patients. There were no differences seen between the groups (Table 1).
Accuracy of CT-PCI In comparison with the oPCI, the CT-PCI significantly underestimated the true burden of the disease (median oPCI 5 39, interquartile range (IQR) 24.75 to 39; median CTPCI 5 25.5, IQR 11.25 to 36.75; P 5 .003) (Fig. 1). However, despite the underestimation, there appears to be good correlation (r2 5 .67). The CT-PCI was not higher in unresectable patients (CCR-2 median CT-PCI 5 23, IQR 12 to 30; CCR-0 median CT-PCI 5 27, IQR 7 to 39; P 5 .851).
Predictability of CT-PCI At least one concerning imaging feature was seen in 87% of resectable patients and 93% of unresectable patients (P 5
Table 1 Comparison of clinical and demographic data for the complete cytoreduction (CCR-0) and the unresectable (CCR-2) groups CCR-0 Sex Male Female Age (years) Disease-free interval Synchronous 1–12 months .12 months Neoadjuvant chemotherapy Yes No Response to chemotherapy Response/stable Progression Age expressed as mean.
CCR-2 6 (40%) 9 (60%) 55.8
9 (60%) 4 (27%) 2 (13%)
9 (60%) 1 (6%) 5 (33%)
4 (27%) 11 (73%)
8 (53%) 7 (47%)
3 (75%) 1 (25%)
5 (63%) 3 (37%)
1.00). No single concerning feature was associated with unresectability (Table 2). There were a total of 8 patients with 2 or more concerning features identified on CT scan, of which 7 were from the unresectable group (CCR-2 5 87.5%; CCR0 5 12.5%; P 5 .04). Of the 22 patients who had less than 2 concerning features, only 8 (36.4%) were from the unresectable group. When imaging features alone were used (technical consensus), the median likelihood of resectability scores were similar between the completely cytoreduced (CCR-0 5 2, IQR 2 to 4) and unresectable patients (CCR2 5 2, IQR 2 to 3; P 5 .76). After the addition of the clinical features (clinical consensus), the median likelihood of resectability scores remained similar between the completely cytoreduced (CCR-0 5 2, IQR 1 to 3) and unresectable patients (CCR-2 5 2, IQR 1 to 3; P 5 .93).
Comments Our case-matched study compared the preoperative CT scans between patients with PC found intraoperatively to be unresectable with those who underwent a complete cytoreduction. We used 2 surgical oncologists to interpret the
P value .47
9 (60%) 6 (40%) 51.7
Figure 1 Comparison between CT-PCI and intraoperative PCI for each patient (some lines represent multiple patients).
.34 .21
.27
1.00
Table 2 Comparison of concerning features seen on CT imaging for the complete cytoreduction (CCR-0) and the unresectable (CCR-2) groups CCR-0
CCR-2
P value
Bowel obstruction 2 (13) 2 (13) 1.00 Ascites 1 (7) 3 (20) .60 Retroperitoneal lymphadenopathy 0 (0) 2 (13) .48 Hydroureter 2 (13) 0 (0) .48 Porta hepatis involvement 8 (53) 12 (80) .25 Parenchymal liver involvement 1 (7) 4 (27) .33 Pelvic sidewall involvement 0 (0) 1 (7) 1.00 R1 Concerning features 13 (87) 14 (93) 1.00 R2 Concerning features 1 (7) 7 (47) .04 Data are expressed as number (percent).
J.D. Rivard et al.
Preoperative CT does not predict resectability
preoperative CT scans rather than radiologists because commonly the ultimate decision to operate lies with the surgeon. Although radiologists are an invaluable part of the multidisciplinary team necessary for this difficult patient population, they are not always present when the decision of whether to operate is made. Several centers currently use oPCI as an exclusion criterion with patients having a PCI greater than 20 not undergoing an attempted CRS 1 HIPEC. Additionally, several prognostic stratification systems have been developed to help determine which patients are candidates for CRS 1 HIPEC. Unfortunately, many are based on data that can only be obtained at the time of surgery.18,19 The ability to accurately determine who would most benefit from surgery in the preoperative setting could potentially spare certain patients from an unnecessary operation. Few studies have focused on outcomes based on preoperative PCI as determined by CT imaging. Pelz et al conducted one such study where they reported a peritoneal disease severity score (PDSS) that had the ability to separate patients with the PC of colonic origin into prognostic groups. They used clinical symptoms, preoperative PCI, and histologic grade to separate patients into 1 of 4 groups.20 Those patients with a higher score were found to have worse overall survival independent of whether CRS 1 HIPEC was performed. Unfortunately, specific patient symptoms were not prospectively collected in the current cohort of patients, making the analysis of PDSS beyond the scope of this study. Preoperative prognostic scoring systems have tremendous potential in patients with PC. However, currently the strongest predictor of long-term survival is the ability to perform a complete cytoreduction. Our results do not indicate that CT-PCI is associated with the ability to perform a complete cytoreduction. Several anatomical regions, if involved with peritoneal metastases, may preclude the ability to achieve a complete cytoreduction. These include the porta hepatis, retroperitoneum, periaortic lymph nodes, pelvic sidewall, and extensive small bowel involvement. PC smaller than .5 cm are commonly not identified on CT scans where the sensitivity can be as low as 11%.15 The small bowel can be underrepresented on the preoperative CT scan where the false-negative rate has been reported to be as high as 35%.16 Evidence of a small bowel obstruction on preoperative imaging is indicative of a more infiltrative disease process that may limit a complete cytoreduction.9 However, if the area of concern is limited to a small section that can be safely included in a bowel resection, a complete cytoreduction is still achievable. Our results show that the presence of a small bowel obstruction is not associated with unresectability. We found that no individual concerning feature seen on preoperative imaging is associated with the inability to perform a complete cytoreduction. However, 2 or more concerning features seen on preoperative imaging appears to be associated with unresectability and 3 concerning features were only seen in the unresectable group. When an overall likelihood of resectability score
763
was given based solely on the CT scans, there was no difference between the resectable and unresectable patients. Even with the additional knowledge of clinical factors, the results remained the same. This underlines the difficulties encountered by surgeons while attempting to determine if a patient is potentially resectable based on a combination of imaging and clinical features. The limitations of this study include the small sample size and the fact that the 2 surgeons reading the CT scans were the same surgeons performing the operations. Despite only having 30 patients in our study, this is comparable to previous studies in the literature assessing preoperative CT scans. Additionally, our numbers were limited because we matched resectable and unresectable patients based on oPCI and pathologic type. The knowledge of the surgeons’ previous decision to operate was minimized by having between 1 and 6 years pass between the actual surgery and the blind review of the CT scans. Furthermore, each of the 2 surgeons perform more than 30 cytoreductive cases per year, making it less likely that an individual patient’s CT scan would be recognized. Finally, an argument could be made that a median of 2 months between preoperative imaging and surgery is a long time. We did not repeat CT scans in the weeks leading up to a surgery as long as the patient was deemed a surgical candidate at the initial consultation. In conclusion, we found that CT-PCI underestimates the oPCI. On blinded review by 2 experienced peritoneal surgeons, 2 or more concerning CT imaging features appear to be associated with a higher risk of unresectability in patients with PC. However, no specific imaging feature should exclude a patient from an attempted cytoreduction.
References 1. Sugarbaker PH, Cunliffe WJ, Belliveau J, et al. Rationale for integrating early postoperative intraperitoneal chemotherapy into the surgical treatment of gastrointestinal cancer. Semin Oncol 1989;16(4 Suppl 6): 83–97. 2. Koppe MJ, Boerman OC, Oyen WJ, et al. Peritoneal carcinomatosis of colorectal origin: incidence and current treatment strategies. Ann Surg 2006;243:212–22. 3. Smeenk RM, van Velthuysen ML, Verwaal VJ, et al. Appendiceal neoplasms and pseudomyxoma peritonei: a population based study. Eur J Surg Oncol 2008;34:196–201. 4. Verwaal VJ, Bruin S, Boot H, et al. 8-year follow-up of randomized trial: cytoreduction and hyperthermic intraperitoneal chemotherapy versus systemic chemotherapy in patients with peritoneal carcinomatosis of colorectal cancer. Ann Surg Oncol 2008;15:2426–32. 5. Glehen O, Kwiatkowski F, Sugarbaker PH, et al. Cytoreductive surgery combined with perioperative intraperitoneal chemotherapy for the management of peritoneal carcinomatosis from colorectal cancer: a multi-institutional study. J Clin Oncol 2004;22:3284–92. 6. Bijelic L, Yan TD, Sugarbaker PH. Treatment failure following complete cytoreductive surgery and perioperative intraperitoneal chemotherapy for peritoneal dissemination from colorectal or appendiceal mucinous neoplasms. J Surg Oncol 2008;98:295–9. 7. Elias D, Lefevre JH, Chevalier J, et al. Complete cytoreductive surgery plus intraperitoneal chemohyperthermia with oxaliplatin for peritoneal carcinomatosis of colorectal origin. J Clin Oncol 2009; 27:681–5.
764
The American Journal of Surgery, Vol 207, No 5, May 2014
8. Elias D, Glehen O, Pocard M, et al. A comparative study of complete cytoreductive surgery plus intraperitoneal chemotherapy to treat peritoneal dissemination from colon, rectum, small bowel, and nonpseudomyxoma appendix. Ann Surg 2010;251:896–901. 9. Harmon RL, Sugarbaker PH. Prognostic indicators in peritoneal carcinomatosis from gastrointestinal cancer. Int Semin Surg Oncol 2005;2:3. 10. Chua TC, Moran BJ, Sugarbaker PH, et al. Early- and long-term outcome data of patients with pseudomyxoma peritonei from appendiceal origin treated by a strategy of cytoreductive surgery and hyperthermic intraperitoneal chemotherapy. J Clin Oncol 2012;30:2449–56. 11. Verwaal VJ, van Ruth S, de Bree E, et al. Randomized trial of cytoreduction and hyperthermic intraperitoneal chemotherapy versus systemic chemotherapy and palliative surgery in patients with peritoneal carcinomatosis of colorectal cancer. J Clin Oncol 2003;21:3737–43. 12. Elias D, Blot F, El Otmany A, et al. Curative treatment of peritoneal carcinomatosis arising from colorectal cancer by complete resection and intraperitoneal chemotherapy. Cancer 2001;92:71–6. 13. da Silva RG, Sugarbaker PH. Analysis of prognostic factors in seventy patients having a complete cytoreduction plus perioperative intraperitoneal chemotherapy for carcinomatosis from colorectal cancer. J Am Coll Surgeons 2006;203:878–86. 14. Yan TD, Morris DL, Shigeki K, et al. Preoperative investigations in the management of peritoneal surface malignancy with cytoreductive surgery and perioperative intraperitoneal chemotherapy: expert consensus statement. J Surg Oncol 2008;98:224–7. 15. Koh JL, Yan TD, Glenn D, et al. Evaluation of preoperative computed tomography in estimating peritoneal cancer index in colorectal peritoneal carcinomatosis. Ann Surg Oncol 2009;16:327–33. 16. Esquivel J, Chua TC, Stojadinovic A, et al. Accuracy and clinical relevance of computed tomography scan interpretation of peritoneal cancer index in colorectal cancer peritoneal carcinomatosis: a multiinstitutional study. J Surg Oncol 2010;102:565–70. 17. Jacquet P, Jelinek JS, Steves MA, et al. Evaluation of computed tomography in patients with peritoneal carcinomatosis. Cancer 1993; 72:1631–6. 18. Shen P, Hawksworth J, Lovato J, et al. Cytoreductive surgery and intraperitoneal hyperthermic chemotherapy with mitomycin C for peritoneal carcinomatosis from nonappendiceal colorectal carcinoma. Ann Surg Oncol 2004;11:178–86. 19. Goere D, Malka D, Tzanis D, et al. Is there a possibility of a cure in patients with colorectal peritoneal carcinomatosis amenable to complete cytoreductive surgery and intraperitoneal chemotherapy? Ann Surg 2013;257:1065–71. 20. Pelz JO, Stojadinovic A, Nissan A, et al. Evaluation of a peritoneal surface disease severity score in patients with colon cancer with peritoneal carcinomatosis. J Surg Oncol 2009;99:9–15.
appendiceal cancer of 12 to 36 months, colorectal cancer of 18 to 48 months. Experience over the last several decades has shown considerable improvement in Cytoreductive surgery (CRS) and Hyperthermic Intraperitoneal chemotherapy (HIPEC) in some centers even approaching a clinical cure in some selected cases. The largest study included over 500 patients treated at 28 centers that showed an improvement in mean survival from 19 to 62 months. (JGO Mar 13.) This harkens back to Oliver Bearht’s original study of routine re-exploration of colon cancer patients and resecting any residual disease with a measurable improvement in survival, published in and the old surgical dictum of signs of unresectability on physical exam in pancreatic and gastric cancers being either a Blumer’s shelf or malignant ascites. The authors present their experience in a single institution in which 15 patients’ preoperative CT scans were reviewed in a blinded fashion by their surgical oncologists to see if any one particular CT finding correlated with unresectability in the hopes of sparing patients the routine operative exploration via either minimally invasive or open approach. Most of the patients (8) suffered from colorectal cancers with the other 7 various appendiceal malignancies. Unfortunately, the authors could not identify any particular sign that unequivocally deemed the patient as ‘‘unresectable.’’ However, patients with 2 or more worrisome signs were more likely to have been found to be unresectable at exploration. This is a small series that attempted to spare patients the pain and expense of exploration, but unfortunately was a ‘‘negative’’ study in that none could be found. I offer to the authors the following questions and observations:
Discussion Hugh Foy, M.D.: The authors present their experience with using CT scan to help screen those patients who are most likely to benefit from operative excision and intraperitoneal hyperthermic chemotherapy for intraperitoneal recurrence of Gastrointestinal malignancy. As we nononcologic general surgeons know, this is a condition that was once considered a death knell carrying a dismal prognosis and a mean survival measured in months. The most notorious of which is pancreatic cancer with a mean survival of 6 months. At the other extreme, a low grade appendiceal cancer patient may enjoy a mean survival ten times longer, 60 months or better. In the middle, mean survival in other GI malignancies with peritoneal dissemination vary; Gastric cancer of 6 to 12 months, high-grade
1. As you the 2 senior surgeons in your group each do more that 30 CRS/HIPEC cases a year, why is your series only 15. 2. The mean time between the CT scan and operation was more than 2 months, why were they not scanned immediately before the operation in this disease process where a month can represent a significant interval? 3. The scans were viewed by the same surgeons who operated on the patients within 1 to 6 years in the past. Do you think there was any influence on their assessment due to their recollection of some of the details of the case? 4. The CT model and technique is not stated in the manuscript. Our radiologist state that their protocol would be to use a 16 Array, Spiral scanner with intervals set at 2.5 mm with reconstruction at 5 mm. Any suspicious areas would be reconstructed at 2.5 mm. 5. If CT is so poor, would positron emission tomography be better? 6. The data from Sugarbaker’s initial study were not given in the discussion. This and other more recent
J.D. Rivard et al.
Preoperative CT does not predict resectability
data would help better educate the reader to put the problem in the proper perspective. 7. The disease free interval in the unresectable patients was higher (13 vs 33 months) Do you have a sense that this was due to the inherent nature of their disease or less rigorous follow up?
765 8. Twice as many CCR-2 patients had neoadjuvant chemotherapy than the CCR-0 patients (27 vs 53%). No data is given of the patient’s initial clinical or surgical staging. Again, this would put the paper in a better overall perspective for the reader.
North Pacific Surgical Association
Preoperative computed tomography does not predict resectability in peritoneal carcinomatosis Justin D. Rivard, M.D., F.R.C.S.C.*, Walley J. Temple, M.D., F.R.C.S.C., F.A.C.S., Yarrow J. McConnell, M.D., F.R.C.S.C., Hisham Sultan, M.D., Lloyd A. Mack, M.D., F.R.C.S.C., F.A.C.S. Department of Surgery and Oncology, University of Calgary, Calgary, Alberta, Canada
KEYWORDS: Peritoneal carcinomatosis; Cytoreduction; Computed tomography; Resectability
Abstract BACKGROUND: Obtaining a complete cytoreduction in patients with peritoneal carcinomatosis (PC) is one of the most significant prognostic variables for long-term survival. This study explored features on preoperative computed tomography (CT) to predict unresectability. METHODS: A retrospective case–control study was conducted of 15 patients with unresectable PC and 15 patients with completely resected PC matched by intraoperative peritoneal cancer index (PCI) and pathology type. Two surgical oncologists blindly analyzed all abdominopelvic CT scans. RESULTS: PCI estimated on imaging was not higher in unresectable patients (P 5 .851) and significantly underestimated intraoperative PCI measurement (P 5 .003). No single concerning feature was associated with unresectability. However, patients with 2 or more concerning features were more likely to be unresectable (87.5% vs 36.4%, P 5 .035). CONCLUSIONS: Two or more concerning CT imaging features appear to be associated with a higher risk of unresectability in patients with PC. However, no specific imaging feature should exclude a patient from an attempted cytoreduction. Ó 2014 Elsevier Inc. All rights reserved.
Peritoneal carcinomatosis (PC) occurs in 10% to 40% of gastrointestinal malignancies either at initial presentation or upon disease recurrence.1–3 Encouraging progress has been made in recent years toward the treatment of patients with peritoneal metastasis with the use of cytoreduction surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC). One randomized trial and several prospective cohort
There were no relevant financial relationships or any sources of support in the form of grants, equipment, or drugs. * Corresponding author. Tel.: 11-403-521-3070; fax: 11-403-5213744. E-mail address: [email protected] Manuscript received October 28, 2013; revised manuscript December 21, 2013 0002-9610/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjsurg.2013.12.024
trials have shown improvements in disease-free and overall survival for those patients undergoing CRS 1 HIPEC over palliative surgery and/or systemic chemotherapy alone.4–7 However, CRS 1 HIPEC has been associated with considerable morbidity. Additionally, despite attempting to select which patients to perform CRS 1 HIPEC on, there is an unresectability rate ranging from 7% to 25%.4,5,8 Therefore, identifying which patients would most likely benefit from this surgical approach in the preoperative setting is desirable. The Peritoneal Cancer Index (PCI) is a scoring system that quantifies the severity of PC. The extent of disease is scored on a scale with a maximum value of 39 by dividing the abdomen into 13 distinct regions, as described by Harmon and Sugarbaker.9 The PCI has been shown to
J.D. Rivard et al.
Preoperative CT does not predict resectability
predict the surgical team’s ability to perform a complete cytoreduction, which is known to be one of the principal prognostic variables in survival. Patients who undergo a complete cytoreduction with appendiceal mucinous neoplasms and colorectal cancer have been recently shown to have 5-year survival rates of 74% to 85% and 17% to 45%, respectively.4,6,10 This is compared with 5-year survival rates of 24% and 0% for an incomplete cytoreduction in appendiceal mucinous neoplasms10 and colorectal patients.4 Furthermore, PCI has been shown to be an independent prognostic indicator for long-term survival,11–13 even in the presence of a complete cytoreduction.8,13 Preoperative staging is most commonly performed using computed tomography (CT) scanning. A recent international workshop on peritoneal surface malignancy devised an expert consensus statement concluding that contrastenhanced CT scans are the preferred modality to assess a patient’s suitability for CRS.14 However, CT has been identified as a poor predictor of intraoperative PCI (oPCI),15–17 particularly for lesions smaller than .5 cm where the sensitivity has been reported between 11% and 28% compared with 90% and 94% for lesions larger than 5 cm.15,17 The purpose of this study is to identify whether there are any features present on preoperative abdominopelvic CT imaging that can accurately predict unresectability in patients with PC.
761
0 to 39 as reported by Harmon and Sugarbaker.9 PCI was not used as a selection criterion for which patients would undergo an attempted CRS. However, those patients who were found at laparotomy to have disease involving the retroperitoneal or periaortic lymph nodes, invasive disease involving portal triad, extensive small bowel involvement, unexpected liver metastasis, or other unresectable structures did not undergo CRS 1 HIPEC. All cytoreductive surgery included resection of all visible peritoneal tumor deposits, except where such resections would compromise vital functions. At the completion of surgery, the CCR score was recorded as described by Harmon and Sugarbaker9: CCR0 (no residual macroscopic tumor); CCR-1 (residual tumor deposits ,2.5 mm in diameter); and CCR-2 (residual tumor deposits .2.5 mm in diameter). Patients with a CCR-2 score did not undergo intraperitoneal chemotherapy.
Radiologic evaluation
Between 2007 and 2011, 15 consecutive patients with PC who were determined to be unresectable (completeness of cytoreduction score 2 (CCR-2)) after intraoperative evaluation were case matched to 15 patients with PC who underwent a complete cytoreduction (CCR-0). Cases were matched based on oPCI and pathologic type. All patients were operated on at the University of Calgary by 1 of 2 surgical oncologists (W.T., L.M.). Preoperatively, those patients with PC and a good functional status with no evidence of distant disease were considered for treatment with CRS 1 HIPEC. Preoperative investigations included clinical examination; complete colonoscopy; CT scans of the chest, abdomen, and pelvis; and whole body positron emission tomography/CT scans. Most patients were discussed at a multidisciplinary tumor board before surgery. All patients before surgery provided informed consent. Data regarding patient, tumor, and perioperative treatment factors were collected and maintained in a prospective database. The Alberta Cancer Research Ethics Committee approved this study.
The included patients’ preoperative abdominal and pelvic CT scans enhanced with intravenous and oral contrast were interpreted blindly by 2 experienced surgical oncologists who each perform more than 30 cytoreductive surgeries per year. CT scans had been performed at different sites as many patients had imaging investigations before being referred to our hospital. All images had been reconstructed at 5-mm intervals and were read at a digital workstation. Median time between CT scan and cytoreductive surgery was 2 months and 9 days. If the surgeon felt the patient was a candidate for surgery at the initial surgical consult based on the above criteria, a repeat CT scan closer to the surgery date was not required. After blindly reviewing each patient’s CT scans in random order, the surgeons calculated the estimated PCI, which was calculated using the same scale to calculate the oPCI. They also documented whether there were any other concerning factors identified (bowel obstruction, ascites, retroperitoneal lymphadenopathy, pelvic side wall involvement, portal involvement, bilateral hydroureter, liver parenchymal metastases, lung metastases) and whether they felt the patient was technically resectable (technical consensus). The likelihood of resectability was scored on a Likert scale with values ranging from 0 to 5, with 0 being not resectable and 5 being definitely resectable. Following this, the surgeons were given patient and tumor characteristics including age, primary pathology, disease-free interval, administration of neoadjuvant systemic chemotherapy, and response to chemotherapy. With the additional information, the surgeons used the same 5-point scale to record whether they felt the patient was resectable and if they would attempt a cytoreductive surgery (clinical consensus).
Cytoreductive surgery
Statistical analysis
For all patients undergoing attempted CRS, the extent of PC was scored intraoperatively using the PCI on a scale from
Descriptive analyses are presented as medians and ranges for continuous data and frequencies and percentages
Methods Patient selection
762
The American Journal of Surgery, Vol 207, No 5, May 2014
for categorical variables. Comparisons were made using the Mann–Whitney U test and Wilcoxon signed-rank test for quantitative data and Fisher’s exact test for categorical data. All tests were 2-sided with a significance level set at P , .05. All analysis was performed using SPSS version 21.0 (IBM Corp, Armonk, NY).
Results Patients A total of 30 patients were selected for this study. The unresectable and resectable cohorts were each made up of 15 patients and were case-matched by PCI and pathology. Each group consisted of 8 colorectal cancer, 1 high-grade appendix cancer, 3 low-grade appendix cancer, 1 mesothelioma, and 2 appendix adenocarcinoid patients. There were no differences seen between the groups (Table 1).
Accuracy of CT-PCI In comparison with the oPCI, the CT-PCI significantly underestimated the true burden of the disease (median oPCI 5 39, interquartile range (IQR) 24.75 to 39; median CTPCI 5 25.5, IQR 11.25 to 36.75; P 5 .003) (Fig. 1). However, despite the underestimation, there appears to be good correlation (r2 5 .67). The CT-PCI was not higher in unresectable patients (CCR-2 median CT-PCI 5 23, IQR 12 to 30; CCR-0 median CT-PCI 5 27, IQR 7 to 39; P 5 .851).
Predictability of CT-PCI At least one concerning imaging feature was seen in 87% of resectable patients and 93% of unresectable patients (P 5
Table 1 Comparison of clinical and demographic data for the complete cytoreduction (CCR-0) and the unresectable (CCR-2) groups CCR-0 Sex Male Female Age (years) Disease-free interval Synchronous 1–12 months .12 months Neoadjuvant chemotherapy Yes No Response to chemotherapy Response/stable Progression Age expressed as mean.
CCR-2 6 (40%) 9 (60%) 55.8
9 (60%) 4 (27%) 2 (13%)
9 (60%) 1 (6%) 5 (33%)
4 (27%) 11 (73%)
8 (53%) 7 (47%)
3 (75%) 1 (25%)
5 (63%) 3 (37%)
1.00). No single concerning feature was associated with unresectability (Table 2). There were a total of 8 patients with 2 or more concerning features identified on CT scan, of which 7 were from the unresectable group (CCR-2 5 87.5%; CCR0 5 12.5%; P 5 .04). Of the 22 patients who had less than 2 concerning features, only 8 (36.4%) were from the unresectable group. When imaging features alone were used (technical consensus), the median likelihood of resectability scores were similar between the completely cytoreduced (CCR-0 5 2, IQR 2 to 4) and unresectable patients (CCR2 5 2, IQR 2 to 3; P 5 .76). After the addition of the clinical features (clinical consensus), the median likelihood of resectability scores remained similar between the completely cytoreduced (CCR-0 5 2, IQR 1 to 3) and unresectable patients (CCR-2 5 2, IQR 1 to 3; P 5 .93).
Comments Our case-matched study compared the preoperative CT scans between patients with PC found intraoperatively to be unresectable with those who underwent a complete cytoreduction. We used 2 surgical oncologists to interpret the
P value .47
9 (60%) 6 (40%) 51.7
Figure 1 Comparison between CT-PCI and intraoperative PCI for each patient (some lines represent multiple patients).
.34 .21
.27
1.00
Table 2 Comparison of concerning features seen on CT imaging for the complete cytoreduction (CCR-0) and the unresectable (CCR-2) groups CCR-0
CCR-2
P value
Bowel obstruction 2 (13) 2 (13) 1.00 Ascites 1 (7) 3 (20) .60 Retroperitoneal lymphadenopathy 0 (0) 2 (13) .48 Hydroureter 2 (13) 0 (0) .48 Porta hepatis involvement 8 (53) 12 (80) .25 Parenchymal liver involvement 1 (7) 4 (27) .33 Pelvic sidewall involvement 0 (0) 1 (7) 1.00 R1 Concerning features 13 (87) 14 (93) 1.00 R2 Concerning features 1 (7) 7 (47) .04 Data are expressed as number (percent).
J.D. Rivard et al.
Preoperative CT does not predict resectability
preoperative CT scans rather than radiologists because commonly the ultimate decision to operate lies with the surgeon. Although radiologists are an invaluable part of the multidisciplinary team necessary for this difficult patient population, they are not always present when the decision of whether to operate is made. Several centers currently use oPCI as an exclusion criterion with patients having a PCI greater than 20 not undergoing an attempted CRS 1 HIPEC. Additionally, several prognostic stratification systems have been developed to help determine which patients are candidates for CRS 1 HIPEC. Unfortunately, many are based on data that can only be obtained at the time of surgery.18,19 The ability to accurately determine who would most benefit from surgery in the preoperative setting could potentially spare certain patients from an unnecessary operation. Few studies have focused on outcomes based on preoperative PCI as determined by CT imaging. Pelz et al conducted one such study where they reported a peritoneal disease severity score (PDSS) that had the ability to separate patients with the PC of colonic origin into prognostic groups. They used clinical symptoms, preoperative PCI, and histologic grade to separate patients into 1 of 4 groups.20 Those patients with a higher score were found to have worse overall survival independent of whether CRS 1 HIPEC was performed. Unfortunately, specific patient symptoms were not prospectively collected in the current cohort of patients, making the analysis of PDSS beyond the scope of this study. Preoperative prognostic scoring systems have tremendous potential in patients with PC. However, currently the strongest predictor of long-term survival is the ability to perform a complete cytoreduction. Our results do not indicate that CT-PCI is associated with the ability to perform a complete cytoreduction. Several anatomical regions, if involved with peritoneal metastases, may preclude the ability to achieve a complete cytoreduction. These include the porta hepatis, retroperitoneum, periaortic lymph nodes, pelvic sidewall, and extensive small bowel involvement. PC smaller than .5 cm are commonly not identified on CT scans where the sensitivity can be as low as 11%.15 The small bowel can be underrepresented on the preoperative CT scan where the false-negative rate has been reported to be as high as 35%.16 Evidence of a small bowel obstruction on preoperative imaging is indicative of a more infiltrative disease process that may limit a complete cytoreduction.9 However, if the area of concern is limited to a small section that can be safely included in a bowel resection, a complete cytoreduction is still achievable. Our results show that the presence of a small bowel obstruction is not associated with unresectability. We found that no individual concerning feature seen on preoperative imaging is associated with the inability to perform a complete cytoreduction. However, 2 or more concerning features seen on preoperative imaging appears to be associated with unresectability and 3 concerning features were only seen in the unresectable group. When an overall likelihood of resectability score
763
was given based solely on the CT scans, there was no difference between the resectable and unresectable patients. Even with the additional knowledge of clinical factors, the results remained the same. This underlines the difficulties encountered by surgeons while attempting to determine if a patient is potentially resectable based on a combination of imaging and clinical features. The limitations of this study include the small sample size and the fact that the 2 surgeons reading the CT scans were the same surgeons performing the operations. Despite only having 30 patients in our study, this is comparable to previous studies in the literature assessing preoperative CT scans. Additionally, our numbers were limited because we matched resectable and unresectable patients based on oPCI and pathologic type. The knowledge of the surgeons’ previous decision to operate was minimized by having between 1 and 6 years pass between the actual surgery and the blind review of the CT scans. Furthermore, each of the 2 surgeons perform more than 30 cytoreductive cases per year, making it less likely that an individual patient’s CT scan would be recognized. Finally, an argument could be made that a median of 2 months between preoperative imaging and surgery is a long time. We did not repeat CT scans in the weeks leading up to a surgery as long as the patient was deemed a surgical candidate at the initial consultation. In conclusion, we found that CT-PCI underestimates the oPCI. On blinded review by 2 experienced peritoneal surgeons, 2 or more concerning CT imaging features appear to be associated with a higher risk of unresectability in patients with PC. However, no specific imaging feature should exclude a patient from an attempted cytoreduction.
References 1. Sugarbaker PH, Cunliffe WJ, Belliveau J, et al. Rationale for integrating early postoperative intraperitoneal chemotherapy into the surgical treatment of gastrointestinal cancer. Semin Oncol 1989;16(4 Suppl 6): 83–97. 2. Koppe MJ, Boerman OC, Oyen WJ, et al. Peritoneal carcinomatosis of colorectal origin: incidence and current treatment strategies. Ann Surg 2006;243:212–22. 3. Smeenk RM, van Velthuysen ML, Verwaal VJ, et al. Appendiceal neoplasms and pseudomyxoma peritonei: a population based study. Eur J Surg Oncol 2008;34:196–201. 4. Verwaal VJ, Bruin S, Boot H, et al. 8-year follow-up of randomized trial: cytoreduction and hyperthermic intraperitoneal chemotherapy versus systemic chemotherapy in patients with peritoneal carcinomatosis of colorectal cancer. Ann Surg Oncol 2008;15:2426–32. 5. Glehen O, Kwiatkowski F, Sugarbaker PH, et al. Cytoreductive surgery combined with perioperative intraperitoneal chemotherapy for the management of peritoneal carcinomatosis from colorectal cancer: a multi-institutional study. J Clin Oncol 2004;22:3284–92. 6. Bijelic L, Yan TD, Sugarbaker PH. Treatment failure following complete cytoreductive surgery and perioperative intraperitoneal chemotherapy for peritoneal dissemination from colorectal or appendiceal mucinous neoplasms. J Surg Oncol 2008;98:295–9. 7. Elias D, Lefevre JH, Chevalier J, et al. Complete cytoreductive surgery plus intraperitoneal chemohyperthermia with oxaliplatin for peritoneal carcinomatosis of colorectal origin. J Clin Oncol 2009; 27:681–5.
764
The American Journal of Surgery, Vol 207, No 5, May 2014
8. Elias D, Glehen O, Pocard M, et al. A comparative study of complete cytoreductive surgery plus intraperitoneal chemotherapy to treat peritoneal dissemination from colon, rectum, small bowel, and nonpseudomyxoma appendix. Ann Surg 2010;251:896–901. 9. Harmon RL, Sugarbaker PH. Prognostic indicators in peritoneal carcinomatosis from gastrointestinal cancer. Int Semin Surg Oncol 2005;2:3. 10. Chua TC, Moran BJ, Sugarbaker PH, et al. Early- and long-term outcome data of patients with pseudomyxoma peritonei from appendiceal origin treated by a strategy of cytoreductive surgery and hyperthermic intraperitoneal chemotherapy. J Clin Oncol 2012;30:2449–56. 11. Verwaal VJ, van Ruth S, de Bree E, et al. Randomized trial of cytoreduction and hyperthermic intraperitoneal chemotherapy versus systemic chemotherapy and palliative surgery in patients with peritoneal carcinomatosis of colorectal cancer. J Clin Oncol 2003;21:3737–43. 12. Elias D, Blot F, El Otmany A, et al. Curative treatment of peritoneal carcinomatosis arising from colorectal cancer by complete resection and intraperitoneal chemotherapy. Cancer 2001;92:71–6. 13. da Silva RG, Sugarbaker PH. Analysis of prognostic factors in seventy patients having a complete cytoreduction plus perioperative intraperitoneal chemotherapy for carcinomatosis from colorectal cancer. J Am Coll Surgeons 2006;203:878–86. 14. Yan TD, Morris DL, Shigeki K, et al. Preoperative investigations in the management of peritoneal surface malignancy with cytoreductive surgery and perioperative intraperitoneal chemotherapy: expert consensus statement. J Surg Oncol 2008;98:224–7. 15. Koh JL, Yan TD, Glenn D, et al. Evaluation of preoperative computed tomography in estimating peritoneal cancer index in colorectal peritoneal carcinomatosis. Ann Surg Oncol 2009;16:327–33. 16. Esquivel J, Chua TC, Stojadinovic A, et al. Accuracy and clinical relevance of computed tomography scan interpretation of peritoneal cancer index in colorectal cancer peritoneal carcinomatosis: a multiinstitutional study. J Surg Oncol 2010;102:565–70. 17. Jacquet P, Jelinek JS, Steves MA, et al. Evaluation of computed tomography in patients with peritoneal carcinomatosis. Cancer 1993; 72:1631–6. 18. Shen P, Hawksworth J, Lovato J, et al. Cytoreductive surgery and intraperitoneal hyperthermic chemotherapy with mitomycin C for peritoneal carcinomatosis from nonappendiceal colorectal carcinoma. Ann Surg Oncol 2004;11:178–86. 19. Goere D, Malka D, Tzanis D, et al. Is there a possibility of a cure in patients with colorectal peritoneal carcinomatosis amenable to complete cytoreductive surgery and intraperitoneal chemotherapy? Ann Surg 2013;257:1065–71. 20. Pelz JO, Stojadinovic A, Nissan A, et al. Evaluation of a peritoneal surface disease severity score in patients with colon cancer with peritoneal carcinomatosis. J Surg Oncol 2009;99:9–15.
appendiceal cancer of 12 to 36 months, colorectal cancer of 18 to 48 months. Experience over the last several decades has shown considerable improvement in Cytoreductive surgery (CRS) and Hyperthermic Intraperitoneal chemotherapy (HIPEC) in some centers even approaching a clinical cure in some selected cases. The largest study included over 500 patients treated at 28 centers that showed an improvement in mean survival from 19 to 62 months. (JGO Mar 13.) This harkens back to Oliver Bearht’s original study of routine re-exploration of colon cancer patients and resecting any residual disease with a measurable improvement in survival, published in and the old surgical dictum of signs of unresectability on physical exam in pancreatic and gastric cancers being either a Blumer’s shelf or malignant ascites. The authors present their experience in a single institution in which 15 patients’ preoperative CT scans were reviewed in a blinded fashion by their surgical oncologists to see if any one particular CT finding correlated with unresectability in the hopes of sparing patients the routine operative exploration via either minimally invasive or open approach. Most of the patients (8) suffered from colorectal cancers with the other 7 various appendiceal malignancies. Unfortunately, the authors could not identify any particular sign that unequivocally deemed the patient as ‘‘unresectable.’’ However, patients with 2 or more worrisome signs were more likely to have been found to be unresectable at exploration. This is a small series that attempted to spare patients the pain and expense of exploration, but unfortunately was a ‘‘negative’’ study in that none could be found. I offer to the authors the following questions and observations:
Discussion Hugh Foy, M.D.: The authors present their experience with using CT scan to help screen those patients who are most likely to benefit from operative excision and intraperitoneal hyperthermic chemotherapy for intraperitoneal recurrence of Gastrointestinal malignancy. As we nononcologic general surgeons know, this is a condition that was once considered a death knell carrying a dismal prognosis and a mean survival measured in months. The most notorious of which is pancreatic cancer with a mean survival of 6 months. At the other extreme, a low grade appendiceal cancer patient may enjoy a mean survival ten times longer, 60 months or better. In the middle, mean survival in other GI malignancies with peritoneal dissemination vary; Gastric cancer of 6 to 12 months, high-grade
1. As you the 2 senior surgeons in your group each do more that 30 CRS/HIPEC cases a year, why is your series only 15. 2. The mean time between the CT scan and operation was more than 2 months, why were they not scanned immediately before the operation in this disease process where a month can represent a significant interval? 3. The scans were viewed by the same surgeons who operated on the patients within 1 to 6 years in the past. Do you think there was any influence on their assessment due to their recollection of some of the details of the case? 4. The CT model and technique is not stated in the manuscript. Our radiologist state that their protocol would be to use a 16 Array, Spiral scanner with intervals set at 2.5 mm with reconstruction at 5 mm. Any suspicious areas would be reconstructed at 2.5 mm. 5. If CT is so poor, would positron emission tomography be better? 6. The data from Sugarbaker’s initial study were not given in the discussion. This and other more recent
J.D. Rivard et al.
Preoperative CT does not predict resectability
data would help better educate the reader to put the problem in the proper perspective. 7. The disease free interval in the unresectable patients was higher (13 vs 33 months) Do you have a sense that this was due to the inherent nature of their disease or less rigorous follow up?
765 8. Twice as many CCR-2 patients had neoadjuvant chemotherapy than the CCR-0 patients (27 vs 53%). No data is given of the patient’s initial clinical or surgical staging. Again, this would put the paper in a better overall perspective for the reader.
