Available online at www.sciencedirect.com
ScienceDirect EJSO xx (2013) 1e6
www.ejso.com
Bowel obstruction in recurrent gynecologic malignancies: Defining who will benefit from surgical intervention T. Perri a,c,*,d, J. Korach a,c,d, G. Ben-Baruch a,c, A. Jakobson-Setton a,c, L. Ben-David Hogen a,c, S. Kalfon a,c, M. Beiner a,c, L. Helpman a,c, D. Rosin b,c a
Department of Gynecologic Oncology, Sheba Medical Center, Tel Hashomer, Israel b Department of Surgery, Sheba Medical Center, Tel Hashomer, Israel c Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel Accepted 30 October 2013 Available online - - -
Abstract Aim: To define factors that could help select, in a cohort of gynecologic cancer patients with malignant gastro-intestinal obstruction, those most likely to benefit from palliative surgery. Methods: In this retrospective study of patients with malignant gastro-intestinal obstruction who underwent palliative surgery in our institute over 7 years, outcome measures were oral intake, chemotherapy, and 30-day, 60-day and overall survival. Based on Cox proportionalhazards regression models and KaplaneMeier curves with log-rank tests, a prognostic score was developed to identify those most likely to benefit from surgery. Results: Sixty-eight palliative surgeries were performed in 62 patients with ovarian (69.1%), primary-peritoneal (8.8%), cervical (11.8%) or uterine (10.3%) malignancies. Procedures were colostomy (26.5%), ileostomy (39.7%), colonic stent (1.5%), gastrostomy (7.3%), gastroenterostomy (5.9%) and bypass/resection and anastomosis (19.1%). Eighteen patients died prior to discharge, within 3e81 days (median 25 days). The 30-day and 60-day mortality rates were 14.7% and 29.4%, respectively. Postoperative oral-intake and chemotherapy rates were 65% and 53%, respectively, with albumin level identified on multivariate analysis as the only significant predictor of both. Median postoperative survival was 106 days (3e1342). Bypass/resection and anastomosis was associated with improved survival. Ascites below 2 L, younger age, ovarian primary tumor, and higher blood albumin correlated with longer postoperative survival. A prognostic index based on these factors was found to identify patients with increased 30-day and 60-day mortality. Conclusions: Our proposed prognostic index, based on age, primary tumor, albumin and ascites, might help select those gynecological cancer patients most likely to benefit from palliative surgery. Ó 2013 Published by Elsevier Ltd. Keywords: Gynecological malignancies; Palliative care; Bowel surgery; Scoring system
Introduction Patients with recurrent gynecologic malignancies may suffer from bowel obstruction once they become incapable of responding to therapy, usually near end-stage disease. The reported frequency of malignant bowel obstruction (MBO) in ovarian cancer patients might be as high as 50%.1 The usual cause is either extrinsic, due to * Corresponding author. Department of Gynecologic Oncology, Sheba Medical Center, 52621 Tel Hashomer, Israel. Tel.: þ972 3 530286; fax: þ972 3 5304794. E-mail address: [email protected] (T. Perri). d Tamar Perri and Jacob Korach contributed equally to this work.
compression of the bowel by a tumor mass, or intrinsic with infiltration of tumor into the bowel wall or mesentery.1 Once the disease recurs it is usually incurable, and the goal of treatment is to improve quality of life and, if possible, to prolong survival.2e4 Since the quality of life experienced by patients with bowel obstruction is severely impaired, physicians usually initiate conservative treatment.5,6 In many cases, unfortunately, these measures fail7 and the option of palliative surgery is considered. Possible procedures are colostomy, ileostomy, gastrostomy, and Bypass/resection and anastomosis or stent placement. However, the role of palliative procedures in this situation, as well as their true palliative potential, is controversial. There is no
0748-7983/$ - see front matter Ó 2013 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.ejso.2013.10.025 Please cite this article in press as: Perri T, et al., Bowel obstruction in recurrent gynecologic malignancies: Defining who will benefit from surgical intervention, Eur J Surg Oncol (2013), http://dx.doi.org/10.1016/j.ejso.2013.10.025
2
T. Perri et al. / EJSO xx (2013) 1e6
consensus with regard to their safety in frail patients, the operative mortality and morbidity rates are high,8,9 and the reported overall survival following surgery is about 190 days.10,11 There are no generally accepted criteria for selecting patients with MBO who could benefit from palliative procedures, although some prognostic factors have been proposed, including the patient’s age, platinum sensitivity, extent of disease, presence of ascites, and nutritional status.10e14 Our aim was to detect, in gynecologic cancer patients with MBO, prognostic factors for durable palliative surgical procedures that would benefit such patients in terms of symptom control, oral intake, ability to tolerate further chemotherapy, and survival. In particular, we aimed to develop a scoring system that would help identify those patients likely to benefit most. Patients and methods Patients The study was approved by the Ethics Committee of our medical center. All consecutive patients with gynecologic malignancies who underwent bowel surgery performed by a single surgeon (D.R.) in our institute between October 2004 and January 2013 were identified from our prospectively created operating-room database. Of the 88 patients whose medical records were reviewed, those with recurrent gynecologic malignancy and MBO were enrolled in this study. Excluded were patients who were operated on for a recto-vaginal fistula (6 patients), bowel resection during primary cytoreductive surgery due to malignant involvement of the gastrointestinal tract (16 patients), and bowel resection at secondary debulking without obstruction (4 patients). The remaining 62 patients, who had a total of 68 procedures, constituted the study group. In all cases, diagnosis of obstruction was suspected by clinical symptoms and signs and confirmed by a computed tomography scan. The collected data consisted of demographic, clinical, and histological characteristics at diagnosis, course of disease, and clinical details of obstruction and surgery. Ascites fluid amount was estimated pre operatively by imaging and measured at surgery. Postoperative outcome measures for successful palliation were oral intake, chemotherapy, and 30-day, 60-day, and overall survival. Statistical analysis IBM SPSS Statistics for Windows, Version 21.0 (Armonk, NY: IBM Corp.) was used for statistical analyses, which included KaplaneMeier survival analysis and logrank (ManteleCox) tests. Cox proportional-hazards regression models were used to assess the association between different characteristics and the above outcome measures. Logelog plots for each variable were inspected to verify the assumption of proportionality of the hazards, which
was confirmed for all variables studied. A value of P < 0.05 was considered statistically significant. The variables identified as significant for 30- and 60-day survival were then incorporated in the development of a scoring system to predict short-term survival after surgery for MBO. Groupings by score were analyzed for significance.
Results Baseline and patient characteristics Demographic and clinical characteristics of the 68 cases are presented in Table 1. The primary malignancy was ovarian or primary peritoneal in 53 cases (78%). Of these, neoadjuvant chemotherapy was administered in 20 cases (37.7%). Optimal cytoreduction at primary surgery was achieved in 40 cases (75.4%), and 25 cases (47%) were platinum resistant. Median overall survival after surgery for MBO was 106.5 days (range, 3e1342 days). Outcomes of the 68 various procedures are presented in Table 2. Severe complications were sepsis (5 cases), leak from Table 1 Characteristics of 68 palliative surgeries for malignant bowel obstruction. Variable
n (%)
Age in years, median (range) Primary malignancy Epithelial ovarian Low malignant potential ovarain Primary peritoneal Cervix Uterine FIGO stage I‒II III‒IV Unknown Number of chemotherapy lines prior to MBO 1 2 3 4 5 Time in months from diagnosis to MBO, median (range) Number of previous abdominal operations at time of MBO 0 1 2 3 4 Peritoneal carcinomatosis at MBO Ascites>2 L at MBO Serum albumin (g/dl) at MBO, median (range) Serum lymphocytes (absolute count) at MBO, median (range) Serum hematocrit (%) at MBO, median (range)
54 (28e75) 43 (63.2) 4 (5.9) 6 (8.8) 8 (11.8) 7 (10.3) 8 (11.8) 56 (82.3) 4 (5.9)
9 (13.2) 18 (26.5) 17 (25.0) 7 (10.3) 17 (25.0) 32 (3.5e133.8)
3 (4.4) 30 (44.1) 15 (22.1) 9 (13.2) 11 (16.2) 58 (85.3) 26 (38.2) 2.8 (1.6e4.7) 0.8 103 (0.3‒3.5 103) 31.5 (21e45)
Please cite this article in press as: Perri T, et al., Bowel obstruction in recurrent gynecologic malignancies: Defining who will benefit from surgical intervention, Eur J Surg Oncol (2013), http://dx.doi.org/10.1016/j.ejso.2013.10.025
T. Perri et al. / EJSO xx (2013) 1e6
3
Table 2 Operative procedures and outcome of surgery for malignant bowel obstruction. Procedure
Oral feeding after procedure n (%)
Chemotherapy after procedure n (%)
Survival at 30 days n (%)
Survival at 60 days n (%)
Colostomy (n ¼ 18) Ileostomy (n ¼ 27) Resection/bypass & anastomosis (n ¼ 13) Colonic stent (n ¼ 1) Gastrostomy (n ¼ 5) Gastroenterostomy (n ¼ 4) Total (n ¼ 68)
11 (61) 18 (66.6) 11 (84.6) 0 1 (20) 3 (75) 44 (64.7)
10 (55.6) 12 (44.4) 9 (69.2) 0 2 (40) 3 (75) 36 (52.9)
15 25 11 1 2 4 58
13 (72.2) 18 (66.7) 11 (84.6) 0 2 (40) 4 (100) 48 (70.6)
(83.3) (92.6) (84.6) (100) (40) (100) (85.3)
n number of cases.
anastomosis (6 cases), and necrotizing fasciitis (2 cases). After 44 (64.7%) procedure, the patients were able to tolerate oral intake, and in 36 (53%) additional chemotherapy was given. Five patients (4 with ovarian cancer and one with cervical cancer) had percutaneous endoscopic gastrostomy, 4/5 as a primary (and only) procedure and 1/5 as a second procedure at recurrent bowel obstruction (Table 3, patient no.4). Four patients (3 with ovarian cancer and one with cervical cancer) had gastroenterostomy, and one ovarian cancer patient had a colonic stent. In all of those patients, it was estimated as the only possible procedure. All procedures were considered “semi-elective”. Re-obstruction occurred in 10 patients within a median period of 4.3 months (range, 1e36 months), of whom 6 underwent bowel surgery for a second time during the course of their disease (Table 3). Eighteen patients died within 3e81 days (median 25 days) prior to discharge. Overall mortality rates within 30 days and 60 days were 14.7% and 29.4%, respectively. Predictors of survival The following variables were tested by Cox regression analysis as potential predictors of surgical outcome: age at diagnosis and at obstruction, site and stage of primary
disease site, neoadjuvant treatment at diagnosis, cytoreduction to 2 L Age>60 Non-ovarian primary tumor
2 1 1 1
are faced with this situation the procedure most often adopted is to initiate conservative treatment, usually including no oral intake, intravenous fluids or total parenteral nutrition, use of a nasogastric decompression tube, and medical measures to relieve symptoms (narcotics, antiemetics, corticosteroids) and reduce gastrointestinal secretions (anticholinergics and somatostatin receptor agonists).5,6 With complete bowel obstruction, however, the conservative treatment frequently fails. Tang et al.7 reported that out of 26 episodes of complete obstruction, only one (3.8%) resolved with conservative management alone. In a later study,16 43% of 329 patients were successfully treated conservatively but with a re-obstruction rate of 40.5%. Once conservative measures fail, and subject to the patients’ choice, surgical intervention may be considered. However, the associated high incidence of morbidity and mortality should be taken into account. Severe complications have been reported in as many as 45%e78% of patients.17e19 In our patient cohort the rate of severe morbidity was 20%, close to that reported by others.8,9,14,17,20,21 Our finding of 15% and 29% postoperation mortality rates within 30 days and 60 days, respectively, as well as a median survival of 107 days, is also in agreement with other reports.8,9,17,21,22 The longer median survival of 6e8 months after palliative surgery reported by some authors10,20,23 might be attributable to the nature of the study population (exclusively ovarian cancer patients as opposed to all gynecologic cancer patients) or the fact that it was possible to administer chemotherapy postoperatively. In most studies, as in ours, palliation was achieved in the majority of treated patients, yet morbidity and mortality were high. It thus seems that an acceptable strategy might be to identify those patients for whom surgical intervention has a reasonable chance of success and then restrict surgical intervention to those patients. Established criteria and recommendations that might help patients and clinicians in this decision-making process are still lacking. Several criteria have been suggested as prognostic. In most studies, including ours, one of the factors found to be predictive of survival is the patient’s age.12,14,21,24 Several additional criteria that predate the MBO have been suggested, including time from diagnosis to obstruction,13,24 stage of disease,14 platinum sensitivity,13,24,25 prior treatment lines,14 treatment-free interval,26e28 and disease dissemination.12,14,25 None of these additional criteria was found in our study to be significant.
p < 0.02). On univariate analysis, significant factors for post-procedure survival were older age (HR, 1.033; 95% CI, 1.006e1.061; P < 0.02), primary disease site other than ovarian (HR, 5.98; 95% CI, 2.3e15.62; P < 0.05), ascites of more than 2 L (HR, 2.54; 95% CI, 1.42e4.56; P < 0.02), albumin level (HR, 0.53; 95% CI, 0.32e0.88; P < 0.02), absolute lymphocyte count (HR, 0.59; 95% CI, 0.35e0.97; P < 0.04), and bypass/resection and anastomosis surgery (HR, 3.025; 95% CI, 1.35e6.78; P < 0.05). On multivariate analysis, significant factors for shorter post-procedure survival were older age (HR, 1.052; 95% CI, 1.02e1.09: P < 0.03), non-ovarian tumor origin, low albumin level, and type of procedure (bypass/resection and anastomosis, 270 days (213e696), significantly better than colostomy and ileostomy, 154 days (134e474), and 82 days (86e187) respectively, P < 0.03). Scoring system Factors that correlated with increased 30- and 60-day operative mortality were entered into a scoring system. These included age above 60, ascites of more than 2 L, non-ovarian primary tumor, and albumin
ScienceDirect EJSO xx (2013) 1e6
www.ejso.com
Bowel obstruction in recurrent gynecologic malignancies: Defining who will benefit from surgical intervention T. Perri a,c,*,d, J. Korach a,c,d, G. Ben-Baruch a,c, A. Jakobson-Setton a,c, L. Ben-David Hogen a,c, S. Kalfon a,c, M. Beiner a,c, L. Helpman a,c, D. Rosin b,c a
Department of Gynecologic Oncology, Sheba Medical Center, Tel Hashomer, Israel b Department of Surgery, Sheba Medical Center, Tel Hashomer, Israel c Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel Accepted 30 October 2013 Available online - - -
Abstract Aim: To define factors that could help select, in a cohort of gynecologic cancer patients with malignant gastro-intestinal obstruction, those most likely to benefit from palliative surgery. Methods: In this retrospective study of patients with malignant gastro-intestinal obstruction who underwent palliative surgery in our institute over 7 years, outcome measures were oral intake, chemotherapy, and 30-day, 60-day and overall survival. Based on Cox proportionalhazards regression models and KaplaneMeier curves with log-rank tests, a prognostic score was developed to identify those most likely to benefit from surgery. Results: Sixty-eight palliative surgeries were performed in 62 patients with ovarian (69.1%), primary-peritoneal (8.8%), cervical (11.8%) or uterine (10.3%) malignancies. Procedures were colostomy (26.5%), ileostomy (39.7%), colonic stent (1.5%), gastrostomy (7.3%), gastroenterostomy (5.9%) and bypass/resection and anastomosis (19.1%). Eighteen patients died prior to discharge, within 3e81 days (median 25 days). The 30-day and 60-day mortality rates were 14.7% and 29.4%, respectively. Postoperative oral-intake and chemotherapy rates were 65% and 53%, respectively, with albumin level identified on multivariate analysis as the only significant predictor of both. Median postoperative survival was 106 days (3e1342). Bypass/resection and anastomosis was associated with improved survival. Ascites below 2 L, younger age, ovarian primary tumor, and higher blood albumin correlated with longer postoperative survival. A prognostic index based on these factors was found to identify patients with increased 30-day and 60-day mortality. Conclusions: Our proposed prognostic index, based on age, primary tumor, albumin and ascites, might help select those gynecological cancer patients most likely to benefit from palliative surgery. Ó 2013 Published by Elsevier Ltd. Keywords: Gynecological malignancies; Palliative care; Bowel surgery; Scoring system
Introduction Patients with recurrent gynecologic malignancies may suffer from bowel obstruction once they become incapable of responding to therapy, usually near end-stage disease. The reported frequency of malignant bowel obstruction (MBO) in ovarian cancer patients might be as high as 50%.1 The usual cause is either extrinsic, due to * Corresponding author. Department of Gynecologic Oncology, Sheba Medical Center, 52621 Tel Hashomer, Israel. Tel.: þ972 3 530286; fax: þ972 3 5304794. E-mail address: [email protected] (T. Perri). d Tamar Perri and Jacob Korach contributed equally to this work.
compression of the bowel by a tumor mass, or intrinsic with infiltration of tumor into the bowel wall or mesentery.1 Once the disease recurs it is usually incurable, and the goal of treatment is to improve quality of life and, if possible, to prolong survival.2e4 Since the quality of life experienced by patients with bowel obstruction is severely impaired, physicians usually initiate conservative treatment.5,6 In many cases, unfortunately, these measures fail7 and the option of palliative surgery is considered. Possible procedures are colostomy, ileostomy, gastrostomy, and Bypass/resection and anastomosis or stent placement. However, the role of palliative procedures in this situation, as well as their true palliative potential, is controversial. There is no
0748-7983/$ - see front matter Ó 2013 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.ejso.2013.10.025 Please cite this article in press as: Perri T, et al., Bowel obstruction in recurrent gynecologic malignancies: Defining who will benefit from surgical intervention, Eur J Surg Oncol (2013), http://dx.doi.org/10.1016/j.ejso.2013.10.025
2
T. Perri et al. / EJSO xx (2013) 1e6
consensus with regard to their safety in frail patients, the operative mortality and morbidity rates are high,8,9 and the reported overall survival following surgery is about 190 days.10,11 There are no generally accepted criteria for selecting patients with MBO who could benefit from palliative procedures, although some prognostic factors have been proposed, including the patient’s age, platinum sensitivity, extent of disease, presence of ascites, and nutritional status.10e14 Our aim was to detect, in gynecologic cancer patients with MBO, prognostic factors for durable palliative surgical procedures that would benefit such patients in terms of symptom control, oral intake, ability to tolerate further chemotherapy, and survival. In particular, we aimed to develop a scoring system that would help identify those patients likely to benefit most. Patients and methods Patients The study was approved by the Ethics Committee of our medical center. All consecutive patients with gynecologic malignancies who underwent bowel surgery performed by a single surgeon (D.R.) in our institute between October 2004 and January 2013 were identified from our prospectively created operating-room database. Of the 88 patients whose medical records were reviewed, those with recurrent gynecologic malignancy and MBO were enrolled in this study. Excluded were patients who were operated on for a recto-vaginal fistula (6 patients), bowel resection during primary cytoreductive surgery due to malignant involvement of the gastrointestinal tract (16 patients), and bowel resection at secondary debulking without obstruction (4 patients). The remaining 62 patients, who had a total of 68 procedures, constituted the study group. In all cases, diagnosis of obstruction was suspected by clinical symptoms and signs and confirmed by a computed tomography scan. The collected data consisted of demographic, clinical, and histological characteristics at diagnosis, course of disease, and clinical details of obstruction and surgery. Ascites fluid amount was estimated pre operatively by imaging and measured at surgery. Postoperative outcome measures for successful palliation were oral intake, chemotherapy, and 30-day, 60-day, and overall survival. Statistical analysis IBM SPSS Statistics for Windows, Version 21.0 (Armonk, NY: IBM Corp.) was used for statistical analyses, which included KaplaneMeier survival analysis and logrank (ManteleCox) tests. Cox proportional-hazards regression models were used to assess the association between different characteristics and the above outcome measures. Logelog plots for each variable were inspected to verify the assumption of proportionality of the hazards, which
was confirmed for all variables studied. A value of P < 0.05 was considered statistically significant. The variables identified as significant for 30- and 60-day survival were then incorporated in the development of a scoring system to predict short-term survival after surgery for MBO. Groupings by score were analyzed for significance.
Results Baseline and patient characteristics Demographic and clinical characteristics of the 68 cases are presented in Table 1. The primary malignancy was ovarian or primary peritoneal in 53 cases (78%). Of these, neoadjuvant chemotherapy was administered in 20 cases (37.7%). Optimal cytoreduction at primary surgery was achieved in 40 cases (75.4%), and 25 cases (47%) were platinum resistant. Median overall survival after surgery for MBO was 106.5 days (range, 3e1342 days). Outcomes of the 68 various procedures are presented in Table 2. Severe complications were sepsis (5 cases), leak from Table 1 Characteristics of 68 palliative surgeries for malignant bowel obstruction. Variable
n (%)
Age in years, median (range) Primary malignancy Epithelial ovarian Low malignant potential ovarain Primary peritoneal Cervix Uterine FIGO stage I‒II III‒IV Unknown Number of chemotherapy lines prior to MBO 1 2 3 4 5 Time in months from diagnosis to MBO, median (range) Number of previous abdominal operations at time of MBO 0 1 2 3 4 Peritoneal carcinomatosis at MBO Ascites>2 L at MBO Serum albumin (g/dl) at MBO, median (range) Serum lymphocytes (absolute count) at MBO, median (range) Serum hematocrit (%) at MBO, median (range)
54 (28e75) 43 (63.2) 4 (5.9) 6 (8.8) 8 (11.8) 7 (10.3) 8 (11.8) 56 (82.3) 4 (5.9)
9 (13.2) 18 (26.5) 17 (25.0) 7 (10.3) 17 (25.0) 32 (3.5e133.8)
3 (4.4) 30 (44.1) 15 (22.1) 9 (13.2) 11 (16.2) 58 (85.3) 26 (38.2) 2.8 (1.6e4.7) 0.8 103 (0.3‒3.5 103) 31.5 (21e45)
Please cite this article in press as: Perri T, et al., Bowel obstruction in recurrent gynecologic malignancies: Defining who will benefit from surgical intervention, Eur J Surg Oncol (2013), http://dx.doi.org/10.1016/j.ejso.2013.10.025
T. Perri et al. / EJSO xx (2013) 1e6
3
Table 2 Operative procedures and outcome of surgery for malignant bowel obstruction. Procedure
Oral feeding after procedure n (%)
Chemotherapy after procedure n (%)
Survival at 30 days n (%)
Survival at 60 days n (%)
Colostomy (n ¼ 18) Ileostomy (n ¼ 27) Resection/bypass & anastomosis (n ¼ 13) Colonic stent (n ¼ 1) Gastrostomy (n ¼ 5) Gastroenterostomy (n ¼ 4) Total (n ¼ 68)
11 (61) 18 (66.6) 11 (84.6) 0 1 (20) 3 (75) 44 (64.7)
10 (55.6) 12 (44.4) 9 (69.2) 0 2 (40) 3 (75) 36 (52.9)
15 25 11 1 2 4 58
13 (72.2) 18 (66.7) 11 (84.6) 0 2 (40) 4 (100) 48 (70.6)
(83.3) (92.6) (84.6) (100) (40) (100) (85.3)
n number of cases.
anastomosis (6 cases), and necrotizing fasciitis (2 cases). After 44 (64.7%) procedure, the patients were able to tolerate oral intake, and in 36 (53%) additional chemotherapy was given. Five patients (4 with ovarian cancer and one with cervical cancer) had percutaneous endoscopic gastrostomy, 4/5 as a primary (and only) procedure and 1/5 as a second procedure at recurrent bowel obstruction (Table 3, patient no.4). Four patients (3 with ovarian cancer and one with cervical cancer) had gastroenterostomy, and one ovarian cancer patient had a colonic stent. In all of those patients, it was estimated as the only possible procedure. All procedures were considered “semi-elective”. Re-obstruction occurred in 10 patients within a median period of 4.3 months (range, 1e36 months), of whom 6 underwent bowel surgery for a second time during the course of their disease (Table 3). Eighteen patients died within 3e81 days (median 25 days) prior to discharge. Overall mortality rates within 30 days and 60 days were 14.7% and 29.4%, respectively. Predictors of survival The following variables were tested by Cox regression analysis as potential predictors of surgical outcome: age at diagnosis and at obstruction, site and stage of primary
disease site, neoadjuvant treatment at diagnosis, cytoreduction to 2 L Age>60 Non-ovarian primary tumor
2 1 1 1
are faced with this situation the procedure most often adopted is to initiate conservative treatment, usually including no oral intake, intravenous fluids or total parenteral nutrition, use of a nasogastric decompression tube, and medical measures to relieve symptoms (narcotics, antiemetics, corticosteroids) and reduce gastrointestinal secretions (anticholinergics and somatostatin receptor agonists).5,6 With complete bowel obstruction, however, the conservative treatment frequently fails. Tang et al.7 reported that out of 26 episodes of complete obstruction, only one (3.8%) resolved with conservative management alone. In a later study,16 43% of 329 patients were successfully treated conservatively but with a re-obstruction rate of 40.5%. Once conservative measures fail, and subject to the patients’ choice, surgical intervention may be considered. However, the associated high incidence of morbidity and mortality should be taken into account. Severe complications have been reported in as many as 45%e78% of patients.17e19 In our patient cohort the rate of severe morbidity was 20%, close to that reported by others.8,9,14,17,20,21 Our finding of 15% and 29% postoperation mortality rates within 30 days and 60 days, respectively, as well as a median survival of 107 days, is also in agreement with other reports.8,9,17,21,22 The longer median survival of 6e8 months after palliative surgery reported by some authors10,20,23 might be attributable to the nature of the study population (exclusively ovarian cancer patients as opposed to all gynecologic cancer patients) or the fact that it was possible to administer chemotherapy postoperatively. In most studies, as in ours, palliation was achieved in the majority of treated patients, yet morbidity and mortality were high. It thus seems that an acceptable strategy might be to identify those patients for whom surgical intervention has a reasonable chance of success and then restrict surgical intervention to those patients. Established criteria and recommendations that might help patients and clinicians in this decision-making process are still lacking. Several criteria have been suggested as prognostic. In most studies, including ours, one of the factors found to be predictive of survival is the patient’s age.12,14,21,24 Several additional criteria that predate the MBO have been suggested, including time from diagnosis to obstruction,13,24 stage of disease,14 platinum sensitivity,13,24,25 prior treatment lines,14 treatment-free interval,26e28 and disease dissemination.12,14,25 None of these additional criteria was found in our study to be significant.
p < 0.02). On univariate analysis, significant factors for post-procedure survival were older age (HR, 1.033; 95% CI, 1.006e1.061; P < 0.02), primary disease site other than ovarian (HR, 5.98; 95% CI, 2.3e15.62; P < 0.05), ascites of more than 2 L (HR, 2.54; 95% CI, 1.42e4.56; P < 0.02), albumin level (HR, 0.53; 95% CI, 0.32e0.88; P < 0.02), absolute lymphocyte count (HR, 0.59; 95% CI, 0.35e0.97; P < 0.04), and bypass/resection and anastomosis surgery (HR, 3.025; 95% CI, 1.35e6.78; P < 0.05). On multivariate analysis, significant factors for shorter post-procedure survival were older age (HR, 1.052; 95% CI, 1.02e1.09: P < 0.03), non-ovarian tumor origin, low albumin level, and type of procedure (bypass/resection and anastomosis, 270 days (213e696), significantly better than colostomy and ileostomy, 154 days (134e474), and 82 days (86e187) respectively, P < 0.03). Scoring system Factors that correlated with increased 30- and 60-day operative mortality were entered into a scoring system. These included age above 60, ascites of more than 2 L, non-ovarian primary tumor, and albumin
