Outcome of acutely perforated colorectal cancers: Experience of a single district general hospital

Outcome of acutely perforated colorectal cancers: Experience of a single district general hospital

ARTICLE IN PRESS Surgical Oncology (2006) 15, 91–96 Available at www.sciencedirect.com journal homepage: www.elsevier.com/locate/suronc Outcome of ...

202KB Sizes 0 Downloads 6 Views

ARTICLE IN PRESS Surgical Oncology (2006) 15, 91–96

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/suronc

Outcome of acutely perforated colorectal cancers: Experience of a single district general hospital Muzaffar Ali Anwar, Francis D’Souza, Rahael Coulter, Breda Memon, Iftikhar M Khan, Muhammed Ashraf Memon Department of Surgery, Whiston Hospital, Warrington Road, Prescot, Merseyside L35 5DR, UK

KEYWORDS Colorectal cancer; Perforation; Mortality

Abstract Introduction: Perforation of colorectal cancer (CRC) is rare and is associated with a significantly high mortality and morbidity. The aim of the current study was to evaluate various factors influencing the outcome in these patients. Material and Methods: A retrospective analysis of 42 patients with perforated CRC between 1999 and 2003 was performed. A number of variables including age, sex, site of perforation, presence of faecal peritonitis, grade of surgeon, presence of metastasis, stage of tumour, type of surgery, ASA grade and CR POSSUM score were analysed for their influence on the outcome in these patients using MS Excel, MS Access and Stata. Results: Of the 42 patients 19 were female and 23 were male. The mean age of the patients was 70.5 (range 44–96 yr). Thirty patients had perforation at the tumour, 10 proximal to the tumour, and one distal to the primary tumour. The perforation was localised in 25 patients. However, 17 patients had free perforation with frank faecal peritonitis. Twenty-one patients had resection and anastomosis, 18 patients had resection without restoration of bowel continuity and 3 had palliative colostomy. The in-hospital mortality (within 30 d) was 40.5% (n ¼ 17) with only 15 patients being alive at the end of 2 yr with an overall mortality of 64.3% (n ¼ 27). The outcome was not altered by variables such as sex, surgeon’s grade, surgical procedure, Dukes’ staging or the site of perforation (p40.5). Univariate analysis showed that advanced age (po0:01), higher ASA grade (po0:001), higher CR POSSUM score (po0:001) and degree of peritonitis (po0:01) were strongly associated with adverse outcomes. However, in stepwise multivariate logistic regression analysis ASA grade (p ¼ 0:01) and CR POSSUM score (p ¼ 0:01) were the only significant predictors of in-hospital mortality. Conclusion: The outcome of perforated colonic cancer continues to be poor. ASA score and CR POSSUM score are good predictors of the short-term outcome. & 2006 Elsevier Ltd. All rights reserved.

Introduction Corresponding author. Astley House, Whitehall Road, Darwen,

Lancashire BB3 2LH, UK. Tel./fax: +44 1254 760717. E-mail address: [email protected] (M.A. Memon). 0960-7404/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.suronc.2006.09.001

Colorectal cancer (CRC) is the second most common cause of cancer related deaths in the UK. Some 34,000 cases are

ARTICLE IN PRESS 92 diagnosed in the UK each year with 60% being colonic and 40% being rectal cancers [1]. Perforation of CRC is an uncommon complication (2.6–9%) [2–5], but is associated with high mortality and morbidity. In-hospital mortality in cases of perforated CRC has been reported to be between 5% and 40% [4,6] despite advances in modern management of sepsis and peri- and post-operative intensive care medicine. Studies [2–12] have shown various risk factors responsible for the poor outcome of acutely perforated CRC patients and include degree of peritonitis, site of perforation, septic state, stage of the tumour, age and ASA grade of the patient. The aim of this study was to evaluate a number of risk factors that may have a direct bearing on the short and long term outcome of perforated CRC in our hospital.

Method and materials Data was analysed from a prospectively maintained CRC Database on 762 consecutive patients presented between January 1999 and December 2003. Of these 42 (5.5%) patients presented with acute colorectal (CR) perforation. Diagnosis of perforation was made on the basis of acute presentation and surgical intervention and confirmed by histology. The case notes, operating notes and histopathology reports of these 42 patients were then reviewed in detail and a number of important variables were analysed using MS Excel, MS Access and Stata. The variables analysed against in-hospital mortality (within 30 d) include age, sex, site of perforation, presence or absence of faecal peritonitis, stage of tumour, type of surgery, grade of operating surgeon, ASA grade and colorectal (CR) POSSUM score.

Statistical analysis All data were analysed using Stata version 7 (intercooled for windows), Microsoft Excel and Microsoft Access. Continuous variables are summarized with the mean7standard deviation or median (range). Categoric variables are summarized with frequencies and percentages. Univariate analysis of all categorical variables was performed either by chi-square test or the Fischer’s exact test. One-way analysis of variance was used for comparison of all continuous variables. A p value of 0.05 or less was considered statistically significant in univariate analysis. For assessment of predictors of inhospital mortality, multivariate logistic regression analysis was performed. Candidate variables with a p value less than 0.05 were entered using a backwards, stepwise approach. Results were presented as odds ratios (ORs) with 95% confidence intervals (95% CIs). A p value of less than 0.05 was taken as statistically significant. Lastly, the surgical outcome was monitored using cumulative sum analysis (cusum) which is a well accepted visual representation used to detect change in the level of surgical practice over time [13]. It is designed to quickly detect change in performance assoicated with an unacceptable rate of adverse outcome. At an acceptable level of performance, the cusum curve runs randomly at or above a horizontal line (no slope). However, when performance is at an unacceptable level, the cusum slope changes [14]. Cusum is calculated using the formula Sn ¼ SðX 0 2X i Þ, where X i ¼ 1 for a success, X i ¼ 0

M.A. Anwar et al. for a failure, and X 0 represents for a predetermined acceptable level of performance. In our study, the success rate of 90% and 60% over a 30 d period was calculated.

Results Demographic, pre-operative, operative data and Duke’s grading are detailed in Table 1. Of the 42 patients with CRC perforation, 23 (55%) were male and 19 (45%) were female with the mean age 70 (range 44–96 yr). In only 5 patients the diagnosis of CRC cancer was known prior to the acute presentation. Thirty-four (81%) patients were admitted directly via the accident and emergency department whereas 5 patients were in-patients in non-surgical wards and were referred to the on-call surgical team with suspicion of either a large bowel obstruction or perforation. Three patients were admitted through outpatient clinics for suspicion of obstructed large bowel cancer and were found to have perforation when operated upon urgently. There were three patients who had surgery for perforation on

Table 1 Patient’s demographic, preoperative and operative data. Parameter

Number or mean7SD

Percentange or range

Number of patients Age (years)

42 70.5711.5

100% 44–96

Gender  Male  Female

19 23

45.2% 54.8%

Site of perforation Proximal to tumour At the site of tumour Distal to tumour

10 31 1

23.8% 73.8% 2.4%

21

50.0%

18

42.9%

3

7.1%

 

Surgeon’s grade Consultant Trainee

22 20

52.4% 47.6%

 

Peritonitis Faecal (generalised) Purulent (localised)

17 25

40.5% 59.5%

   

Dukes stage A+B C D Not reported

17 23 1 1

40.5% 54.8% 2.4% 2.4%

CR Possum score

10.472.51

6–18

  

Type of Surgery Resection with anastomosis  Resection without anastomosis  By-pass/ defunctioning stoma



ARTICLE IN PRESS Acutely perforated colorectal cancers

93 significant (p40.5). Twenty-five patients had localized peritonitis, of which 6 (24%) died in-hospital while out of 17 patients who had free faecal peritonitis, 11 (65%) died inhospital (po0:05). There was no statistical difference between grades of operating surgeons and in-hospital mortality (Table 3). There were total of 17 deaths (M ¼ 10, F ¼ 7), the causes of which include sepsis (n ¼ 6), multiorgan failure (n ¼ 5), pneumonia (n ¼ 1), myocardial infarction (n ¼ 1) and disseminated disease (n ¼ 4). Interestingly enough, 21 patients had unsuspected metastatic disease discovered at the time of operation and 8 out of them died in-hospital. However in only 4 of them metastasis was considered to be the direct cause of death. CR POSSUM [15], a modification of POSSUM, is a severity score, which is used to measure the operative mortality of CR surgical patients based on their physiological condition

27th, 28th and 29th days of admission. This was because their immediate presentation was that of subacute small bowel obstruction which was managed conservatively. However, following a period of improvement, they deteriorated subsequently and therefore underwent emergency surgery. In analysing the in-hospital mortality with respect to time between admission and surgery, there was no significant difference between the groups of patients who had operation within 24 h or after that (48% and 50%, respectively). Table 2 provides detailed account of univariate analysis of risk factors associated with in-hospital mortality. Regarding the site of perforations, 31 patients had perforation at tumour site, 10 proximal to the tumour (diastatic), and one distal to the tumour. Eleven (35.4%) patients with tumour site perforation and 6 (60%) with proximal perforation died. This result was not statistically

Table 2

Univariate analysis of risk factors associated with in-hospital mortality.

Parameters

Deaths (n ¼ 17)

Survivors ðn ¼ 25Þ

P value

n

%

n

%

o70 X70

3 14

15.8 60.9

16 9

84.2 39.1

0.004a

Gender  Male  Female

10 7

43.5 36.8

13 12

56.5 63.2

0.757

  

Site of perforation Colon Rectum Not specified

17 0 0

43.6

22 2 1

56.4 100 100

0.18

 

Peritonitis Faecal Localised

11 6

64.7 24.0

6 19

35.3 76.0

0.012a

   

Dukes Stage A+B C D Not specified

6 9 1 1

35.3 39.1 100 100

11 14 0 0

64.7 60.9

0.593

  

Surgery Resection with anastomosis Resection without anastomosis By-pass/stoma

9 7 1

42.9 38.9 33.3

12 11 2

57.1 61.1 66.7

1.00

 

Surgeon’s grade Consultant Trainee

8 9

36.4 45.0

14 11

63.6 55.0

0.754

 

ASA score I–II III–IV

3 14

13.6 70.0

19 6

86.4 30.0

o0:001a

CR Possum o12 X12

7 10

22.6 90.9

24 1

77.4 9.1

o0:001a

Age

 

 

a

Statistically significant with p value o0:05:

ARTICLE IN PRESS 94

M.A. Anwar et al.

Table 3

Type of surgery and in-hospital mortality.

Type of surgery

Total

Alive

Dead

Anterior resection (n ¼ 1) Left hemicolectomy (n ¼ 3) Right hemicolectomy (n ¼ 11) Extended right hemicolectomy Simoid colectomy Subtotal colectomy Resection with primary anastomosis

1 3 11 2 3 1 21

1 2 6 1 1 1 12 (57%)

0 1 5 1 2 0 9 (43%)

Hartmann’s procedure Right hemicolectomy+Hartmann’s Left hemicolectomy+defunctioning stoma Sigmoid colectomy+defunctioning stoma Resection without anastomosis By-pass or defunctioning Total

12 2 1 3 18 3 42

7 1 1 2 11 (61%) 2 (67%) 25 (59.5%)

5 1 0 1 7 (39%) 1 (33%) 17 (40.5%)

Table 4

Multivariate analysis showing factors independently affecting in-hospital mortality.

Variable

Comparison

Odds ratio

95% confidence interval

P value

ASA group CR-POSSUM score

ASA I–II vs. ASA III–V Score o12 vs. score X12

10.52958 24.48149

1.75–66.10 2.15–278.86

0.010 0.010

Table 5

Showing the difference between the observed and predicted mortality by CR-Posum.

Patient group

CR possum score

No of patients

Predicted mortality (%)

Expected deaths

Observed deaths

Oberved to expected ratio

1 2 3 4

0–9 10–19 20–29 429

23 9 2 8

5.33 14.22 21.85 41.44

1.23 1.3 0.45 3.32

5 4 2 6

4.1 3.1 4.4 1.8

Fisher’s exact ¼ 0.012.

and operative severity score. The analysis of this severity score in our perforated CRC patients showed that there was a direct correlation between a higher physiological score and the mortality and it was 100% for patients who had CR POSSUM score of X13 (Table 2). When analysing various risk factors (as shown in Table 1) against in-hospital mortality, univariate analysis showed that advanced age (po0:01), higher ASA grade (po0:001), higher CR POSSUM score (po0:001) and degree of peritonitis (po0:01) were strongly associated with adverse outcomes. However, multivariate logistic regression analysis revealed that ASA grade (p ¼ 0:01) and CR POSSUM score (p ¼ 0:01) were the only significant predictors of in-hospital mortality (Table 4). Table 5 shows the comparison between the predicted mortality by CR POSSUM and the observed mortality. Clearly the observed mortality was two to four times more than the predicted mortality. CR POSSUM therefore under estimated

the mortality in this patient sub-group, which was statistically significant. Lastly, the cusum plots were calculated for a success rate of 90% and 60% (mortality rate 10% and 40%) over a 30 d period. Fig. 1 shows that at 60% success rate (cusum 40), the curve remains flat from the beginning of the series, indicating that there is no learning curve. On the other hand the curve at 90% success (cusum 10) continues to slope upwards with the passage of time reflecting the high incidence of mortality in this group.The curve does not achieve plateau at any stage, indicating that despite adequate experience the mortality still remains high, which is likely to be due to factors other than operator competency affecting the outcome. The traditional cusum used here makes no adjustment for different risk profiles, clinical presentation and CR POSSUM score. A risk adjusted cusum could be a more accurate representation of the level of performance.

ARTICLE IN PRESS Acutely perforated colorectal cancers

95

CUSUM ANALYSIS OF THE INSTITUTIONAL RESULTS OF THE 30 DAY MORTALITY 14 CUSUM 10 CUSUM 40

12 10

Cusum

8 6 4 2 0 -2

1

6

11

16

21

26

31

36

41

-4 Number of Patients

Fig. 1 Cusum analysis of the institutional results of the 30 d morality.

Discussion As perforated CRCs are quite uncommon (2.6–9%) [2–5], the majority of the published series span over a long period of time to achieve adequate numbers for any meaningful analysis [2–12]. Our series analysed the 5 yr data on 762 consecutive CRCs which revealed a perforation rate of 5.5%, very much in keeping with the other large-scale series. A number of previously [2–12] published series have revealed that perforated CRC patients have a high perioperative mortality and low short- and long-term survival. This poor outcome is based on a number of risk factors which include older age of the patients, degree of peritonitis, septic stage, site of perforation, stage of tumour (locally advanced malignancy), cardiopulmonary co-morbidies, ASA grade and the presence or absence of distant metastasis at presentation. In our series we analysed not only all these risk factors but additionally looked at the timing of surgery, type of the surgical procedures, the grade of operating surgeon(s) and the role CR POSSUM score. There are two mechanisms why CRC perforation occurs; (a) due to tumour necrosis (site of tumour) and (b) due to proximal colon blow out from an obstructed tumour and a competent ileocecal valve (diastatic perforation). A number of studies have shown that diastatic perforation carries poorer prognosis than tumour site perforation. In the former group, the contamination is diffuse and faecal, leading to severe septic shock, resulting in higher perio-operative deaths [3,4,8]. Compare this with a tumour site perforation, where perforation is usually localised and leads to the development of phlegmon, thereby reducing greatly the risk of generalised peritonitis. Furthermore, this type of perforation usually leads to purulent collection. In our series, patients with diastatic perforation had a peri-operative mortality of 60% compared to tumour site perforation whose mortality rate was 37%. This difference, although not statistically significant, showed a trend towards poorer outcome in patients with diastatic perforation. Previous studies have found a direct correlation between advancing age and perforation [4,16,17]. Our study similarly

confirmed this observation. It is therefore imperative that a higher degree of suspicion for CR carcinoma must be maintained when older patients present with any sort of bowel symptoms such as altered bowel habits, PR bleeding etc. In our study, univariate analysis showed that advanced age was associated with adverse postoperative outcome. However, age was not a significant predictor of poor outcome in the stepwise multivariate analysis. There seems to be no consensus between tumour (Dukes’) stages and perforated CRCs. A number of authors [4,11,16,17] have reported higher stages in perforated CRCs compared to uncomplicated CRCs. However, others [2] like our study have failed to confirm these results. Except for two patients, all the patients were either in Dukes’ B or C and there was no statistically significant difference in the mortality rate between the two groups (Table 2). However, it is important to bear in mind that the number of patients in each group was quite small. Since the 1960s, when Glenn and McSherry [18] showed primary resection of the perforated CR lesion carried a better prognosis, most of the surgeons endeavour to perform resection with or without primary anastomosis as the minimum. In our series, except for three patients, 39 patients underwent resection with or without anastomosis (Table 3). There was no difference in the mortality rate between those who underwent resection and primary anastomosis versus those who underwent resection and formation of stoma. Our study, like that of Runkel et al. [6], support the notion that the one-stage procedure has no adverse outcome in patients with perforated CRCs and whenever feasible should be performed. None of the previous series have looked at the role of the grade of operating surgeon and in-hospital mortality in perforated CRCs. Our analysis failed to reveal a distinct correlation between the grade of the operating surgeons, i.e. trainees versus consultant and in-hospital mortality (Table 2). There may be a number of reasons behind this. Firstly the number of perforated CRC in this series is small. Secondly one can assume that the trainees carrying out emergency colonic procedure on their own must be experienced and confident enough, otherwise they would have sought the help of a senior surgeon. Thirdly, when looking at the consultant data, not all the surgeons are dedicated CR specialists (although a vast majority are trained in CR surgery) and they may have been responsible for diluting the better results produced by the specialist CR surgeon in such a situation. We have no way of proving or disproving these assumptions. However, the only way forward would be to perform a prospective audit into the results of specialist versus non-specialist CR surgeons, especially in emergency CR situations such as perforation and obstruction. However, such a dedicated prospective audit will require a fairly large number of units in order to accrue reasonable number of patients over a long period of time to get some meaningful comparison. Biondo et al. [19] assessed the prognostic value of the ASA score in patients with left colonic peritonitis and concluded that only the ASA score and preoperative organ failure were significantly associated with postoperative mortality in multivariate logistic regression analysis. Similarly in our study a higher ASA score was strongly associated with adverse outcome both in the univariate and multivariate analyses.

ARTICLE IN PRESS 96 A number of studies have evaluated the importance of POSSUM and its modifications in predicting postoperative mortality in patients undergoing both elective and emergency CR surgery [20–22]. These scoring systems although not perfect provide valuable information of the approximate risk of morbidity and mortality before surgery both to patients, their next of kin, surgeons and the anaesthetists. Furthermore, such tools can be used for a comparative audit. We similarly evaluated our patient’s series using one of the recent modifications of POSSUM, the CR POSSUM [15,21] applicable to patients undergoing CR surgery only. This modification uses 10 of the 18 POSSUM risk factors and requires linear analysis. The overall accuracy of predicting mortality using the CR POSSUM was good. However, our results are in agreement with Metcalfe et al. [23] who found that the CR POSSUM model significantly under-estimated the predicted mortality in patients undergoing emergency CR surgery. They attributed the higher mortality to possible failure to incorporate adequate weighting for faecal peritonitis and the associated systemic insult into this model. Lastly we feel that the cusum technique does provide a simple form of objective monitoring even of an established practice in an emergency situation (e.g., perforated CRC) and any sub-optimal performance in relation to an agreed standard can be recognised and remedied.

Conclusion Perforated CRCs, despite advances in modern management of sepsis and peri- and post-operative intensive care medicine, carry a substaintial mortality. The two major predictors of poor outcome are ASA grade and CR POSSUM. It is recommended that these patients should be managed expeditiously and aggresively with a multidisciplinary approach involving surgeons, anaesthetist and intensivists to optimise their perioperative and postoperative care.

References [1] Office for National Statistics. Cancer statistics: registrations. Registrations of cancer diagnosed in 2001. England. Series MBI no 32, /http://www.statistics.gov.ukS. London:National Statistics; 2004. [2] Chen HS, Sheen-Chen SM. Obstruction and perforation in colorectal adenocarcinoma: an analysis of prognosis and current trends. Surgery 2000;127:370–6. [3] Mandava N, Kumar S, Pizzi WF, Aprile IJ. Perforated colorectal carcinomas. American Journal of Surgery 1996;172:236–8. [4] Kriwanek S, Armbruster C, Dittrich K, Beckerhinn P. Perforated colorectal cancer. Diseases of the Colon and Rectum 1996; 39:1409–14. [5] Kelley Jr. WE, Brown PW, Lawrence Jr. W, Terz JJ. Penetrating, obstructing, and perforating carcinomas of the colon and rectum. Archives of Surgery 1981;116:381–4.

M.A. Anwar et al. [6] Runkel NS, Hinz U, Lehnert T, Buhr HJ, Herfarth CH. Improved outcome after emergency surgery for cancer of the large intestine. British Journal of Surgery 1998;85:1260–5. [7] Shinkawa H, Yasuhara H, Naka S, Yanagie H, Nojiri T, Furuya Y, et al. Factors affecting the early mortality of patients with nontraumatic colorectal perforation. Surgery Today 2003;33: 13–7. [8] Carraro PG, Segala M, Orlotti C, Tiberio G. Outcome of largebowel perforation in patients with colorectal cancer. Diseases of the Colon and Rectum 1998;41:1421–6. [9] Khan S, Pawlak SE, Eggenberger JC, Lee CS, Szilagy EJ, Margolin DA. Acute colonic perforation associated with colorectal cancer. American Surgeon 2001;67:261–4. [10] Michowitz M, Avnieli D, Lazarovici I, Solowiejczyk M. Perforation complicating carcinoma of colon. Journal of Surgical Oncology 1982;19:18–21. [11] Peloquin AB. Factors influencing survival with complete obstruction and free perforation of colorectal cancers. Diseases of the Colon and Rectum 1975;18:11–21. [12] Willett C, Tepper JE, Cohen A, Orlow E, Welch C. Obstructive and perforative colonic carcinoma: patterns of failure. Journal of Clinical Oncology 1985;3:379–84. [13] De Leval MR. Analysis of a cluster of surgical failures: application to a series of neonatal arterial switch operations. Journal of Thoracic and Cardiovascular Surgery 1994;107: 914–24. [14] Change WR, McLean IP. CUSUM: a tool for early feedback about performance? BMC Medical Research Methodology 1996;6:8. [15] Tekkis PP, Prytherch DR, Kocher HM, Senapati A, Poloniecki JD, Stamatakis JD, et al. Development of a dedicated riskadjustment scoring system for colorectal surgery (colorectal POSSUM). British Journal of Surgery 2004;91:1174–82. [16] Kyllonen LE. Obstruction and perforation complicating colorectal carcinoma. An epidemiologic and clinical study with special reference to incidence and survival. Acta Chirurgica Scandinavica 1987;153:607–14. [17] Runkel NS, Schlag P, Schwarz V, Herfarth C. Outcome after emergency surgery for cancer of the large intestine. British Journal of Surgery 1991;78:183–8. [18] Glenn F, McSherry CK. Obstruction and perforation in colorectal cancer. Annals of Surgery 1971;173:983–92. [19] Biondo S, Ramos E, Deiros M, Rague JM, De Oca J, Moreno P, et al. Prognostic factors for mortality in left colonic peritonitis: a new scoring system. Journal of the American College of Surgeons 2000;191:635–42. [20] Slim K, Panis Y, Alves A, Kwiatkowski F, Mathieu P, Mantion G. Association Francaise de Chirurgie. Predicting postoperative mortality in patients undergoing colorectal surgery. World Journal of Surgery 2006;30:100–6. [21] Senagore AJ, Warmuth AJ, Delaney CP, Tekkis PP, Fazio VW. POSSUM, p-POSSUM, and Cr-POSSUM: implementation issues in a United States health care system for prediction of outcome for colon cancer resection. Diseases of the Colon and Rectum 2004;47:1435–41. [22] Poon JT, Chan B, Law WL. Evaluation of P-POSSUM in surgery for obstructing colorectal cancer and correlation of the predicted mortality with different surgical options. Diseases of the Colon and Rectum 2005;48:493–8. [23] Metcalfe MS, Norwood MG, Miller AS, Hemingway D. Unreasonable expectations in emergency colorectal cancer surgery. Colorectal Diseases 2005;7:275–8.