- Email: [email protected]
Hydronephrosis as a prognostic factor in pelvic recurrence from rectal and colon carcinomas
The American Journal of Surgery 190 (2005) 55– 60
Clinical paper – International
Hydronephrosis as a prognostic factor in pelvic recurrence from rectal and colon carcinomas Stein G. Larsen, M.D.*, Johan N. Wiig, M.D., Ph.D., Karl-Erik Giercksky, M.D., Ph.D. Department of Surgical Oncology, The Norwegian Radium Hospital, N-0310 Oslo, Norway Manuscript received February 10, 2004; revised manuscript July 3, 2004
Abstract Background: After multimodal treatment estimated 5-year survival of locally recurrent rectal cancer is about 25%. Hydronephrosis secondary to pelvic recurrence of colorectal cancer is a condition claimed to represent a contraindication to surgery due to a dismal prognosis. Methods: Prospective registration of 193 consecutive patients operated for pelvic recurrence in rectal or colon cancer from January 1991 until March 2002 at a tertiary referral hospital, 121 men and 72 women, median age 67 years, all given irradiation preoperatively. Twenty-three of 193 had hydronephrosis prior to preoperative irradiation for recurrent disease. Results: R-0 stage resection was obtained in 22% of patients with hydronephrosis and in 41% without. The median survival times in patients without metastasis were 27 and 32 months, respectively, and 5-year survival rates were 11% and 25%. Conclusions: An aggressive surgical approach offers patients with pelvic recurrence from rectal and colon cancer the best potential for survival. The presence of hydronephrosis probably indicates a lower chance for complete surgical resection of the recurrence, but local control and improved survival may still be achieved, and about two thirds of patients may benefit from the operation. © 2005 Excerpta Medica Inc. All rights reserved. Keywords: Recurrent; Rectum; Carcinoma; Hydronephrosis; Surgery; Irradiation
The introduction of total mesorectal excision (TME) has lowered the local recurrence rate to around 5– 8% after surgery for rectal cancer [1,2]. Recurrence may cause a reduced quality of life, with pelvic pain, bleeding, stenosis, or fistulas to vagina, bowel, or bladder. An autopsy study from the time before the TME era demonstrated that as many as 25% of patients died of rectal cancer with local recurrence alone [3]. Thus it seems worthwhile to invest considerable effort in the treatment of local recurrence. The crude 5-year survival in patients with locally recurrent rectal cancer treated with preoperative irradiation and surgery is about 25% [4,5]. Without treatment patients with recurrent rectal cancer have a median survival of 3 to 6 months [3]. Some investigators have claimed that hydronephrosis secondary to recurrent rectal cancer represents a local manifestation of disseminated disease with almost no probability of long-term survival or cure and that adequate surgical result cannot be achieved [6,7]. Studies from Roswell Park * Corresponding author. Tel.: ⫹47 22 93 00; fax: ⫹47 22 93 59 44. E-mail address: [email protected]
Cancer Institute have suggested that hydronephrosis should be considered an absolute contraindication to surgery [8,9]. It has also been claimed that the 2 stomas after pelvic exenteration, which may be necessary in some patients with hydronephrosis, result in a poor quality of life [10]. In our institution we have had an aggressive approach with multimodal treatment for locally recurrent rectal cancer since 1990 and hydronephrosis per se has not been considered a contraindication.
Materials and Methods The Norwegian Radium Hospital is a tertiary referral university hospital for locally advanced and locally recurrent rectal cancer. From January 1991 until March 2002, 193 patients (median age 67 years) with pelvic recurrence from rectal or colon cancer were operated. The same team consisting of 2 general surgeons and 1 of 2 urologists or 1 reconstructive surgeon when necessary operated on all patients. All patients were included in a prospective investi-
0002-9610/05/$ – see front matter © 2005 Excerpta Medica Inc. All rights reserved. doi:10.1016/j.amjsurg.2004.07.043
56
S.G. Larsen et al. / The American Journal of Surgery 190 (2005) 55– 60
gation and recorded in a database. End points are local recurrence, distant metastases, or death. The preoperative evaluation was by rectal/vaginal exploration, sigmoideoscopy, computed tomography (CT) or magnetic resonance imaging (MRI) of the pelvis, transrectal ultrasonography when feasible, and cystoscopy when bladder involvement was suspected. Distant metastases were examined by chest x-ray and ultrasonography of the liver and/or CT of the abdomen or chest. Serum carcinoembryonal antigen (CEA) was measured preoperatively and as a part of follow-up. The patients have attended the outpatient clinic in our hospital postoperatively for 5 years with x-ray of the chest, ultrasonography of the liver, and CT of the pelvis and abdomen every 3 months for 2 years, then every 6 months for the next 3 years. Additional information was obtained from the referring hospitals or the patients’ general practitioners. Hydronephrosis Hydronephrosis was not routinely prospectively registered in our database and all reports from CT or MRI scans of the pelvis and abdomen and ultrasonography on these 193 patients were therefore evaluated retrospectively, both before radiation therapy and between radiation and operation. When necessary, signs of hydronephrosis and hydroureter were further investigated by ultrasonography, urography, and/or renography. We used the standard radiologic definition of hydronephrosis with dilation of renal pelvis and calyces, and dilation of ureter down to an obstruction. The authors performed the final review in agreement with the evaluation of 2 individual radiologists performing the examination and describing the hydronephrosis. These 23 patients with hydronephrosis were compared to the 170 patients without (106 males and 64 females, median age 67.5 years). Preoperative radiation All patients were given preoperative irradiation in a dose of 46 to 50 Gy with a daily fraction of 2 Gy [4]. None had been given irradiation for the primary cancer. Chemoradiation therapy was not used in our department in this period but is now part of our routine regimens.
Table 1 Patients operated for recurrent rectal cancer with (n ⫽ 23) or without hydronephrosis (n ⫽ 170)
No. of patients Median age Range Male Female
Hydronephrosis present
Hydronephrosis not present
23 65 y 49–81 y 15 (65%) 8
170 67.5 y 34–83 y 106 (62%) 64
The local surgical achievement was staged as follows: R-0 ⫽ microscopically free circumferential and distal margins, R-1 ⫽ microscopically involved margins, and R-2 ⫽ locally macroscopic residual cancer or no resection. Statistics Curves of actuarial survival and local recurrence were calculated with the Kaplan-Meier product-limit method and the difference between curves was calculated with the logrank test. Calculations were performed using the Statistical Package for the Social Sciences program (SPSS, Chicago, IL).
Results Patients Demographic data are given in Table 1. Fewer abdominoperineal or low anterior resections (61% vs. 84%) and more sigmoid resections (17% vs. 7%) were performed during primary surgery in the hydronephrosis group. Primary surgery consisted of 59 abdominoperineal resections, 98 low anterior resections, and 11 Hartmann’s procedures. The percentages of patients developing hydronephrosis were 8.5%, 9.2%, and 18%, respectively. Surgery The procedures performed are shown in Table 2. Resections must be relatively extensive if a curative goal or local control is the aim of the treatment. All patients underwent complete abdominal exploration to detect and evaluate met-
Intraoperative radiotherapy Intraoperative radiotherapy (IORT) was used during the first 8 years when technically feasible. The electron energy was selected on the basis of depth of tissue to be irradiated (median 12 MeV, range 9 –15 MeV). A dose of 15 Gy was given if no macroscopic tumor tissue remained and 17.5 to 20 Gy when macroscopic tumor tissue remained [4]. Nine patients with hydronephrosis (39%) and 61 without (36%) were given IORT. Due to the lack of a significant effect of IORT in these patients, IORT was abandoned in 1999 [4].
Table 2 Main operative procedures during surgery for recurrent rectal cancer in patients with (n ⫽ 23) or without hydronephrosis (n ⫽ 170) Operative procedures
Hydronephrosis
Not hydronephrosis
Abdominoperineal rectal resection Hartmann’s operation on rectum Low anterior resection Tumor resection only Explorative laparotomy
5 (22%) 8 (35%) 0 (0%) 7 (30%) 3 (13%)
46 (27%) 42 (25%) 16 (9%) 46 (27%) 20 (12%)
S.G. Larsen et al. / The American Journal of Surgery 190 (2005) 55– 60 Table 3 Additional resections during surgery for recurrent rectal cancer in patients with (n ⫽ 23) or without hydronephrosis (n ⫽ 170) Operative procedures
Hydronephrosis
Not hydronephrosis
Cystoprostatectomy (males) Cystectomy (females) Bladder resection Simple urinary diversion Reimplantation of ureter Prostatectomi (isolated) Resection of vesicles Hysterectomy Vaginal resection Small bowel resection Pelvic wall resection
7 (47%) 1 (13%) 0 1 (13%) 2 (25%) 0 8 (53%) 5 (63%) 4 (50%) 6 (40%) 16 (70%)
11 (10%) 3 (5%) 2 (1%) 0 0 3 (3%) 42 (40%) 22 (34%) 27 (42%) 38 (22%) 66 (39%)
astatic disease in peritoneum, liver, and lymphatic nodes. Abdominoperineal resections and Hartmann’s procedures were performed in more than half of the patients in both groups. Hartmann’s procedure was preferred when the risk of further recurrence was increased, the anal sphincter seemed weak, or the patient was satisfied with the previously prepared deviating stoma. Low anterior resection was only performed when the risk of local failure was considered relatively low or when the patient insisted on having an anastomosis. It was not performed in any of the patients with hydronephrosis and seldom in the group without. In order to obtain radical resections, a number of additional procedures were performed (Table 3). As expected, cystoprostatectomy was more frequently performed in patients with hydronephrosis than in the group without. Cystectomy as a single procedure was done in a few of the female patients. In both groups resection of 1 or both of the seminal vesicles was frequently necessary. Hysterectomy was performed twice as often in the hydronephrosis group, partly due to cancer infiltration and partly to facilitate the surgical procedure. Vaginal resection was done in almost half of the patients and resection of small bowel when necessary, partly due to tumor involvement and partly due to damage caused by irradiation or surgery. In total, 32 other organs within the pelvis were removed or resected in the 23 patients with hydronephrosis and 148 pelvic organs in the 170 patients without hydronephrosis. With extended surgery we believe it is possible to include some of the patients with pelvic sidewall involvement. We sometimes even perform these extended procedures in cases of metastatic disease to gain local control in the pelvis and thereby decrease the well-known risk of later quality of life reduction.
57
Twenty-three patients had unilateral or bilateral hydronephrosis: severe in 6 patients, moderate in 16 patients and light in 1 (15 males and 8 females, median age 65 years; Table 1). Two of the 23 patients had been treated for liver metastasis earlier. Twelve patients had hydronephrosis of the left kidney, 8 of the right kidney, and 3 had bilateral hydronephrosis. The hydronephrosis decreased only in 2 of 23 patients during irradiation. Two patients needed temporary nephrostomy drainage and 2 required internal stents in the ureter during irradiation. The peroperative treatment of the hydronephrosis was cystectomy in 8 patients and resection of ureter with reimplantation in 2. In 1 patient, a simple urinary diversion without resection was made, and in the last 12 patients the ureter was dissected free and decompressed during the tumor/bowel resection. Residual tumor stage The most common reason for refusing surgery in some reports [8,9] is pelvic sidewall involvement with or without carcinomatosis. In order to achieve R-0 or R-1 resections, we performed the resection outside the anatomical mesorectal fascia in 70% of the hydronephrosis patients and in 39% of the other patients. Despite this extensive surgical procedure in addition to preoperative radiation therapy, locally complete resections of the tumor (R-0) were achieved in 22% (5 patients) of the hydronephrosis group but 41% (69 patients) of the non-hydronephrosis group. R-1 resections were obtained in 43% (10 patients) of the hydronephrosis group and in 35% (61 patients) of the non-hydronephrosis group. We did not manage to remove all macroscopic tumor (R-2) in 35% (8 patients) of the hydronephrosis group and in 24% (40 patients) of the non-hydronephrosis group. Complications Complications are listed in Table 4. One patient (1%) without hydronephrosis died during the first postoperative month due to a subphrenic abscess complicated by pulmonary embolism; no patients in the hydronephrosis group died. Within 3 months postoperatively, 1 patient (4%) with hydronephrosis and 5 (3%) without hydronephrosis died. Five patients (22%) in the hydronephrosis group and 11 (6%) in the non-hydronephrosis group developed postoperative bowel leakage; similar numbers regarding urinary leakage were 1 (4%) and 17 (10%). Nearly 30% had signs of pelvic infection with increased C-reactive protein, fever and/or pelvic abscesses.
Hydronephrosis The obstructions were due to either ureteral tumor compression or tumor invasion. In 9 patients the obstruction was localized just below the common iliac vessels, in 7 along the pelvic sidewall, and in 7 at the base of the bladder/prostate.
Survival The median survival of all patients was 21 months in the hydronephrosis group and 32 months in the patients without
58
S.G. Larsen et al. / The American Journal of Surgery 190 (2005) 55– 60
Table 4 Number of patients with postoperative complications among 193 patients with pelvic recurrence of colorectal cancer
Mortality Leakage Bowel leakage Urinary leakage Infectious complications Deep abscess Septicaemia Wound infection Urinary tract infection Pneumonia Tromboembolic complications Deep venous thrombosis Pulmonary embolism Total
Hydronephrosis (n ⫽ 23)
Not hydronephrosis (n ⫽ 170)
0 5 patients (22%) 5 1 12 patients (52%) 7 0 3 5 3 2 patients (9%) 1 1 13 patients (56%)
1 (.6%) 28 patients (16%) 11 17 79 patients (45%) 42 7 22 45 16 6 patients (4%) 4 2 80 patients (47%)
hydronephrosis. The median survival times among the patients without metastasis (M-0), 16 with hydronephrosis and 133 without, were 27 and 32 months, respectively (Fig. 1). Estimated 5-year survival in patients without metastasis was 11% in the hydronephrosis group and 25% in the nonhydronephrosis group. This difference was statistically different (P ⫽ .05).
Secondary Local Recurrences Among the R-0 or R-1 resections, 9 (60%) patients with hydronephrosis are estimated free of local recurrence after 5 years, as are 69 (53%) among those without hydronephrosis (Fig. 2). This difference was not statistically significant (log-rank .54, P ⫽ .46).
Fig. 1. Kaplan-Meier survival plots for all M-0 patients (without metastasis) with pelvic recurrence from colorectal cancer. Patients with hydronephrosis (n ⫽ 16) and without (n ⫽ 133).
Fig. 2. Estimated Kaplan-Meier local rerecurrence plots for 145 patients with R-0 and R-1 resections after pelvic recurrence from colorectal cancer. Patients with hydronephrosis (n ⫽ 15) and without (n ⫽ 130).
Comments Hydronephrosis developing in patients with a history of colorectal cancer is usually of malignant origin. LevChelouche et al [7] found a benign cause in less than 8% of cases. Our material considers only malignant disease and does not include patients who were rejected for surgery. All patients had hydronephrosis and dilation of the ureter down to an obstruction near the recurrent tumor. Still, the 12% frequency of patients with hydronephrosis is similar to reported frequencies of 7% [11] and 19% [8] in publications regarding recurrent rectal cancers, as well as the 13% found in a mixture of primary and recurrent colorectal cancers [12]. Some reports on recurrent rectal cancer and hydronephrosis were heterogeneous in procedures performed and irradiation given [7–9]. The studies were small and contained 38, 27, and 16 patients with hydronephrosis and recurrent rectal cancer. All patients in 1 study [9] were surgically explored, but only 6 (22%) [8] and 10 (26%) [7] underwent surgical exploration in the 2 other studies. In the latter report, 3 patients were operated on with radical intent and 7 palliatively. In the 2 reports from Roswell Park Cancer Institute [8,9], none of the patients had tumor resections, particularly due to pelvic sidewall involvement. We have resected the tumor in 20 of our 23 patients (87%) with recurrent cancer and hydronephrosis. None of our patients had irradiation as part of the treatment for their primary cancer. Accordingly, they could be given preoperative irradiation for their recurrent cancer, which we consider to be of significant importance. Irradiation did not seem to be an option in the preoperative treatment in the other studies concerning recurrent rectal cancer and hydronephrosis [7–9]. In the present study, locally complete resection was achieved in 22% of the patients with preoperative hydronephrosis. This is half the frequency of R-0 resections in the
S.G. Larsen et al. / The American Journal of Surgery 190 (2005) 55– 60
non-hydronephrosis group, although the difference is not statistically significant. Totally, about two thirds of patients had R-0/R-1 resections in the hydronephrosis group, slightly lower than in the other group. In a previous report we found that patients with R-1 resection appeared to fare better than those with R-2 resections [13]. Thus we suggest that both patients with R-0 and R-1 resections (65%) may benefit from the operation. This correlation between surgical resection stage and prognosis seems to be in accordance with the suggestion of other authors [14,15]. Hydronephrosis is caused by either infiltration of tumor into the ureter or by tumor compressing the ureter. Thus, an R-0 resection may in some cases be achieved without resection of the ureter. Resections of the ureter and/or bladder were performed in 10 of 23 patients (43%) in the hydronephrosis group and decompression in the rest. Investigation of the specimens in the resection group showed direct invasion into or through the affected ureter in 5 patients, tumor in very close distance in 2 (ostium and vesicle), fibrosis after irradiation in 1, and in 2 patients invasion could not be further evaluated. Among the unresected cases, we believe that only the patient with simple urinary diversion had direct tumor invasion, giving an invasion rate after irradiation of at least 23% in the hydronephrosis group, a little higher than the 17% described earlier [7]. Lopez et al [6] described complete ureteral obstruction as “a local sign of invasive carcinoma beyond the limits of operability to cure.” When obstruction was due to compression rather than invasion, they suggested that resection might not be contraindicated. We no longer perform IORT because we did not find any difference between the IORT and non-IORT groups in any R-stages regarding overall survival or local recurrence [4]. We therefore do not think the IORT treatment influenced the results in this study. Cheng and Rodriguez-Bigas et al [8,9] from Roswell Park Cancer Institute reported 22 patients who underwent exploration without any resection of tumor. Their results show survival times of mean 14 and median 8 months, which they claim are similar to those in patients with distant metastasis. Their 5-year survival rate in patients undergoing exploratory laparotomy alone is 0. In our patients with hydronephrosis, the 5-year survival rate for those without metastasis (M-0) is 11%. Our median survival time is 27 months in the hydronephrosis group. It is difficult to determine if it is the operation, adjuvant treatment, or selection of patients that causes such a wide variation in results. In our patients without hydronephrosis and without distant metastasis, the median survival (32 months) is slightly longer and the 5-year survival (25%) higher, which suggests hydronephrosis to be a factor complicating recurrent rectal cancer. An aggressive treatment approach to recurrent rectal cancer may lead to 5-year survival rates up to 54% by selecting patients who could be surgically resected with negative margins [16] and who tolerated extensive surgery. Five-year survival rates of 20% in all patients independent of hydronephrosis status seems acceptable compared to
59
other reports [17]. Some investigators have found the extent of resection to be a significant predictor of improved survival [14,15,18]. It was possible to obtain 38% R-0 resections after extensive surgery, even in the treatment of recurrent cancer. It is generally accepted that the surgical stage is of importance also for the frequency of local re-recurrences. Although a large number of patients with pelvic recurrence following rectal surgery may be re-explored, the number of patients who can be resected for cure (R-0) is often less than 50% [19,20] and varies from 15% [5] to 80% [21]. These variations may partly reflect differences in surgical aggressiveness in operations performed, and they underline the importance of obtaining, at least microscopically, free resection margins even in recurrent disease. Despite extensive surgery, the postoperative mortality rate was low in the total group of patients (.5%) with recurrent colorectal cancer both with regard to 30-day and in-hospital mortality. No patients died within 1 month postoperatively in the hydronephrosis group. In order to operate radically, cystoprostatectomy was performed 5 times more often in the hydronephrosis group and hysterectomy twice as often. Even so, the complication rate was about the same in the groups. We had 193 complications among the 193 patients with apparently more bowel leakage in the hydronephrosis group: 5 patients (22%) in contrast to 11 patients (6%) in the group without hydronephrosis. This could be due to an increased number of resections on small bowel, use of small bowel in deviation after cystoprostatectomy, or simple bias because of small numbers. The lower urinary leakage rate in the hydronephrosis group is again probably due to the smaller number of subjects in this group. The infectious complications from the pelvis may be due to the use of preoperative irradiation [22,23]. Irradiation of the bowel may further enhance bacterial translocation with subsequent infections. Irradiation also depresses the immune system through an effect on resident lyphocytes that may influence the risk of postoperative infections [24]. Patients who undergo major cancer surgery after heavy pelvic irradiation are known to have a high risk of deep venous thrombosis or pulmonary embolism. Accordingly, the result of 3% clinical deep venous thrombosis and 2% clinical pulmonary embolism, including 1 fatal, seems low and acceptable [25]. The morbidity rate varies between 14% and 49%, mostly around 25%, and mortality rates are low, varying from 0% to 3% [18,26]. Higher complication rates were observed in patients who underwent extended resections (32%) and in patients with fixed tumors [26]. In our experience, the relatively high complication rates but low mortality rates seem acceptable when the prolonged survival is taken into account. The 8 patients in need of cystectomy were required double stomas. In a recent report from our department, we found somewhat to our surprise that pelvic exenteration
60
S.G. Larsen et al. / The American Journal of Surgery 190 (2005) 55– 60
with 2 stomas was still consistent with a good quality of life [27]. Multidisciplinary teams Multidisciplinary teams in highly specialized institutions with profound knowledge in preoperative staging and extensive surgical procedures should evaluate patients with advanced tumors.
Conclusion All previous studies on recurrent cancer with hydronephrosis, as well as the present study, included few patients. Thus, no firm conclusions can be drawn. Nevertheless, our study suggests that the presence of hydronephrosis reduces the chance of locally complete resection of recurrence after colorectal cancer. Still, probably about two thirds of patients may benefit from extensive resections, and surgery may be performed with acceptable postoperative mortality and morbidity rates. Surgery also may be an option in patients with local recurrence presenting with hydronephrosis.
References [1] MacFarlane JK, Ryall RDH, Heald RJ. Mesorectal excision for rectal cancer. Lancet 1993;341:457– 60. [2] Enker WE, Thaler H, Cranor ML, et al. Total mesorectal excision in the operative treatment of carcinoma of the rectum. J Am Coll Surg 1995;181:335– 46. [3] Welch JP, Donaldson GA. Detection of treatment of recurrent cancer of the colon and rectum. Am J Surg 1978;135:505– 8. [4] Wiig JN, Tveit KM, Poulsen JP, et al. Preoperative irradiation and surgery for recurrent rectal cancer. Will intraoperative radiotherapy (IORT) be of addidional benefit? A prospective study. Radiother Oncol 2002;62:207–13. [5] Gunderson LL, Haddock MG, Nelson H, et al. Locally recurrent colorectal cancer. IOERT and EBRT ⫹/⫺ 5-FU and maximal resection. Front Radiot Ther Oncol 1997;31:224 – 8. [6] Lopez MJ, Kraybill WG, Downey RS, et al. Exenterative surgery for locally advanced rectosigmoid cancers. Is it worthwhile? Surgery 1987;102:644 –51. [7] Lev-Chelouche D, Keidar A, Rub R, et al. Hydronephrosis associated with colorectal carcinoma: treatment and outcome. Eur J Surg Oncol 2001;27:482– 6. [8] Cheng C, Rodriguez-Bigas M, Petrelli N. Is there a role for curative surgery for pelvic recurrence from rectal carcinoma in the presence of hydronephrosis? Am J Surg 2001;182:274 –7.
[9] Rodriguez-Bigas M, Herrara L, Petrelli N. Surgery for recurrent rectal adenocarcinoma in the presence of hydronephrosis? Am J Surg 1992; 164:18 –21. [10] Baigrie RJ, Berry AR. Management of advanced rectal cancer. Br J Surg 1994;8:343–52. [11] Ulm AK, Klein E. Management of ureteral obstruction produced by recurrent cancer of the rectosigmoid colon. Surg Gynecol Obstet 1960;110:413– 8. [12] Lee PH, Khauli RB, Baker S, et al. Prognostic and therapeutic observations of manifestations in the genitourinary tract of adenocarcinoma of the colon and rectum. Surg Gynecol Obstet 1989;169: 511– 8. [13] Wiig JN, Poulsen JP, Larsen S, et al. Total pelvic exenteration with preoperative irradiation for advanced primary and recurrent rectal cancer. Eur J Surg 2002;168:42– 8. [14] Shoup M, Guillem JG, Alektiar KM, et al. Predictors of survival in recurrent rectal cancer after resection and intraoperative radiotherapy. Dis Colon Rectum 2002;45:585–91. [15] Wallace HJ, Willett CG, Shellito PC, et al. Intraoperative radiation therapy for locally advanced recurrent rectal or rectosigmoid cancer. J Clin Oncol 1995;60:122–7. [16] Curley SA, Carlson GW, Shumate CR, et al. Extended resection for locally advanced colorectal carcinoma. Am J Surg 1992;163:553–9. [17] Abulafi AM, Williams NS. Local recurrence of colorectal cancer: the problem, mechanisms, management and adjuvant therapy. Review. Br J Surg 1994;81:7–19. [18] Wanebo HJ, Koness RJ, Turk PS, et al. Composite resection of posterior pelvic malignancy. Ann Surg 1992;215:685–95. [19] Cunningham JD, Enker W, Cohen A. Salvage therapy for pelvic recurrence following curative rectal cancer resection. Dis Colon Rectum 1997;40:393– 400. [20] Holm T, Cedermark B, Rutqvist LE. Local recurrence of rectal adenocarcinoma after “curative” surgery with and without preoperative radiotherapy. Br J Surg 1994;81:452–5. [21] Lowy AM, Rich TA, Skibber JM, et al. Preoperative infusional chemoradiation, selective intraoperative radiation, and resection for locally advanced pelvic recurrence of colorectal adenocarcinoma. Ann Surg 1996;223:177– 85. [22] Stockholm Rectal Cancer Study Group. Preoperative short-term radiation therapy in operable rectal carcinoma. A prospective randomized trial. Cancer 1990;66:49 –55. [23] Swedish Rectal Cancer Trial. Initial report from a Swedish multicentre study examining the role of preoperative irradiation in the treatment of patients with respectable rectal carcinoma. Br J Surg 1993; 80:1133– 6. [24] Order SE. The effects of therapeutic irradiation on lymphocytes and immunity. Cancer 1977;39:737– 43. [25] Holm T, Rutqvist LE, Johansson H, et al. Postoperative mortality in rectal cancer treated with or without preoperative radiotherapy: causes and risk factors. Br J Surg 1996;83:964 – 8. [26] Hahnloser D, Nelson H, Gunderson LL, et al. Curative potential of multimodality therapy for locally recurrent rectal cancer. Ann Surg 2003;237:502– 8. [27] Guren MG, Wiig JN, Dueland S, et al. Quality of life in patients with urinary diversion after operation for locally advanced rectal cancer. Eur J Surg Oncol 2001;27:502– 8.
Clinical paper – International
Hydronephrosis as a prognostic factor in pelvic recurrence from rectal and colon carcinomas Stein G. Larsen, M.D.*, Johan N. Wiig, M.D., Ph.D., Karl-Erik Giercksky, M.D., Ph.D. Department of Surgical Oncology, The Norwegian Radium Hospital, N-0310 Oslo, Norway Manuscript received February 10, 2004; revised manuscript July 3, 2004
Abstract Background: After multimodal treatment estimated 5-year survival of locally recurrent rectal cancer is about 25%. Hydronephrosis secondary to pelvic recurrence of colorectal cancer is a condition claimed to represent a contraindication to surgery due to a dismal prognosis. Methods: Prospective registration of 193 consecutive patients operated for pelvic recurrence in rectal or colon cancer from January 1991 until March 2002 at a tertiary referral hospital, 121 men and 72 women, median age 67 years, all given irradiation preoperatively. Twenty-three of 193 had hydronephrosis prior to preoperative irradiation for recurrent disease. Results: R-0 stage resection was obtained in 22% of patients with hydronephrosis and in 41% without. The median survival times in patients without metastasis were 27 and 32 months, respectively, and 5-year survival rates were 11% and 25%. Conclusions: An aggressive surgical approach offers patients with pelvic recurrence from rectal and colon cancer the best potential for survival. The presence of hydronephrosis probably indicates a lower chance for complete surgical resection of the recurrence, but local control and improved survival may still be achieved, and about two thirds of patients may benefit from the operation. © 2005 Excerpta Medica Inc. All rights reserved. Keywords: Recurrent; Rectum; Carcinoma; Hydronephrosis; Surgery; Irradiation
The introduction of total mesorectal excision (TME) has lowered the local recurrence rate to around 5– 8% after surgery for rectal cancer [1,2]. Recurrence may cause a reduced quality of life, with pelvic pain, bleeding, stenosis, or fistulas to vagina, bowel, or bladder. An autopsy study from the time before the TME era demonstrated that as many as 25% of patients died of rectal cancer with local recurrence alone [3]. Thus it seems worthwhile to invest considerable effort in the treatment of local recurrence. The crude 5-year survival in patients with locally recurrent rectal cancer treated with preoperative irradiation and surgery is about 25% [4,5]. Without treatment patients with recurrent rectal cancer have a median survival of 3 to 6 months [3]. Some investigators have claimed that hydronephrosis secondary to recurrent rectal cancer represents a local manifestation of disseminated disease with almost no probability of long-term survival or cure and that adequate surgical result cannot be achieved [6,7]. Studies from Roswell Park * Corresponding author. Tel.: ⫹47 22 93 00; fax: ⫹47 22 93 59 44. E-mail address: [email protected]
Cancer Institute have suggested that hydronephrosis should be considered an absolute contraindication to surgery [8,9]. It has also been claimed that the 2 stomas after pelvic exenteration, which may be necessary in some patients with hydronephrosis, result in a poor quality of life [10]. In our institution we have had an aggressive approach with multimodal treatment for locally recurrent rectal cancer since 1990 and hydronephrosis per se has not been considered a contraindication.
Materials and Methods The Norwegian Radium Hospital is a tertiary referral university hospital for locally advanced and locally recurrent rectal cancer. From January 1991 until March 2002, 193 patients (median age 67 years) with pelvic recurrence from rectal or colon cancer were operated. The same team consisting of 2 general surgeons and 1 of 2 urologists or 1 reconstructive surgeon when necessary operated on all patients. All patients were included in a prospective investi-
0002-9610/05/$ – see front matter © 2005 Excerpta Medica Inc. All rights reserved. doi:10.1016/j.amjsurg.2004.07.043
56
S.G. Larsen et al. / The American Journal of Surgery 190 (2005) 55– 60
gation and recorded in a database. End points are local recurrence, distant metastases, or death. The preoperative evaluation was by rectal/vaginal exploration, sigmoideoscopy, computed tomography (CT) or magnetic resonance imaging (MRI) of the pelvis, transrectal ultrasonography when feasible, and cystoscopy when bladder involvement was suspected. Distant metastases were examined by chest x-ray and ultrasonography of the liver and/or CT of the abdomen or chest. Serum carcinoembryonal antigen (CEA) was measured preoperatively and as a part of follow-up. The patients have attended the outpatient clinic in our hospital postoperatively for 5 years with x-ray of the chest, ultrasonography of the liver, and CT of the pelvis and abdomen every 3 months for 2 years, then every 6 months for the next 3 years. Additional information was obtained from the referring hospitals or the patients’ general practitioners. Hydronephrosis Hydronephrosis was not routinely prospectively registered in our database and all reports from CT or MRI scans of the pelvis and abdomen and ultrasonography on these 193 patients were therefore evaluated retrospectively, both before radiation therapy and between radiation and operation. When necessary, signs of hydronephrosis and hydroureter were further investigated by ultrasonography, urography, and/or renography. We used the standard radiologic definition of hydronephrosis with dilation of renal pelvis and calyces, and dilation of ureter down to an obstruction. The authors performed the final review in agreement with the evaluation of 2 individual radiologists performing the examination and describing the hydronephrosis. These 23 patients with hydronephrosis were compared to the 170 patients without (106 males and 64 females, median age 67.5 years). Preoperative radiation All patients were given preoperative irradiation in a dose of 46 to 50 Gy with a daily fraction of 2 Gy [4]. None had been given irradiation for the primary cancer. Chemoradiation therapy was not used in our department in this period but is now part of our routine regimens.
Table 1 Patients operated for recurrent rectal cancer with (n ⫽ 23) or without hydronephrosis (n ⫽ 170)
No. of patients Median age Range Male Female
Hydronephrosis present
Hydronephrosis not present
23 65 y 49–81 y 15 (65%) 8
170 67.5 y 34–83 y 106 (62%) 64
The local surgical achievement was staged as follows: R-0 ⫽ microscopically free circumferential and distal margins, R-1 ⫽ microscopically involved margins, and R-2 ⫽ locally macroscopic residual cancer or no resection. Statistics Curves of actuarial survival and local recurrence were calculated with the Kaplan-Meier product-limit method and the difference between curves was calculated with the logrank test. Calculations were performed using the Statistical Package for the Social Sciences program (SPSS, Chicago, IL).
Results Patients Demographic data are given in Table 1. Fewer abdominoperineal or low anterior resections (61% vs. 84%) and more sigmoid resections (17% vs. 7%) were performed during primary surgery in the hydronephrosis group. Primary surgery consisted of 59 abdominoperineal resections, 98 low anterior resections, and 11 Hartmann’s procedures. The percentages of patients developing hydronephrosis were 8.5%, 9.2%, and 18%, respectively. Surgery The procedures performed are shown in Table 2. Resections must be relatively extensive if a curative goal or local control is the aim of the treatment. All patients underwent complete abdominal exploration to detect and evaluate met-
Intraoperative radiotherapy Intraoperative radiotherapy (IORT) was used during the first 8 years when technically feasible. The electron energy was selected on the basis of depth of tissue to be irradiated (median 12 MeV, range 9 –15 MeV). A dose of 15 Gy was given if no macroscopic tumor tissue remained and 17.5 to 20 Gy when macroscopic tumor tissue remained [4]. Nine patients with hydronephrosis (39%) and 61 without (36%) were given IORT. Due to the lack of a significant effect of IORT in these patients, IORT was abandoned in 1999 [4].
Table 2 Main operative procedures during surgery for recurrent rectal cancer in patients with (n ⫽ 23) or without hydronephrosis (n ⫽ 170) Operative procedures
Hydronephrosis
Not hydronephrosis
Abdominoperineal rectal resection Hartmann’s operation on rectum Low anterior resection Tumor resection only Explorative laparotomy
5 (22%) 8 (35%) 0 (0%) 7 (30%) 3 (13%)
46 (27%) 42 (25%) 16 (9%) 46 (27%) 20 (12%)
S.G. Larsen et al. / The American Journal of Surgery 190 (2005) 55– 60 Table 3 Additional resections during surgery for recurrent rectal cancer in patients with (n ⫽ 23) or without hydronephrosis (n ⫽ 170) Operative procedures
Hydronephrosis
Not hydronephrosis
Cystoprostatectomy (males) Cystectomy (females) Bladder resection Simple urinary diversion Reimplantation of ureter Prostatectomi (isolated) Resection of vesicles Hysterectomy Vaginal resection Small bowel resection Pelvic wall resection
7 (47%) 1 (13%) 0 1 (13%) 2 (25%) 0 8 (53%) 5 (63%) 4 (50%) 6 (40%) 16 (70%)
11 (10%) 3 (5%) 2 (1%) 0 0 3 (3%) 42 (40%) 22 (34%) 27 (42%) 38 (22%) 66 (39%)
astatic disease in peritoneum, liver, and lymphatic nodes. Abdominoperineal resections and Hartmann’s procedures were performed in more than half of the patients in both groups. Hartmann’s procedure was preferred when the risk of further recurrence was increased, the anal sphincter seemed weak, or the patient was satisfied with the previously prepared deviating stoma. Low anterior resection was only performed when the risk of local failure was considered relatively low or when the patient insisted on having an anastomosis. It was not performed in any of the patients with hydronephrosis and seldom in the group without. In order to obtain radical resections, a number of additional procedures were performed (Table 3). As expected, cystoprostatectomy was more frequently performed in patients with hydronephrosis than in the group without. Cystectomy as a single procedure was done in a few of the female patients. In both groups resection of 1 or both of the seminal vesicles was frequently necessary. Hysterectomy was performed twice as often in the hydronephrosis group, partly due to cancer infiltration and partly to facilitate the surgical procedure. Vaginal resection was done in almost half of the patients and resection of small bowel when necessary, partly due to tumor involvement and partly due to damage caused by irradiation or surgery. In total, 32 other organs within the pelvis were removed or resected in the 23 patients with hydronephrosis and 148 pelvic organs in the 170 patients without hydronephrosis. With extended surgery we believe it is possible to include some of the patients with pelvic sidewall involvement. We sometimes even perform these extended procedures in cases of metastatic disease to gain local control in the pelvis and thereby decrease the well-known risk of later quality of life reduction.
57
Twenty-three patients had unilateral or bilateral hydronephrosis: severe in 6 patients, moderate in 16 patients and light in 1 (15 males and 8 females, median age 65 years; Table 1). Two of the 23 patients had been treated for liver metastasis earlier. Twelve patients had hydronephrosis of the left kidney, 8 of the right kidney, and 3 had bilateral hydronephrosis. The hydronephrosis decreased only in 2 of 23 patients during irradiation. Two patients needed temporary nephrostomy drainage and 2 required internal stents in the ureter during irradiation. The peroperative treatment of the hydronephrosis was cystectomy in 8 patients and resection of ureter with reimplantation in 2. In 1 patient, a simple urinary diversion without resection was made, and in the last 12 patients the ureter was dissected free and decompressed during the tumor/bowel resection. Residual tumor stage The most common reason for refusing surgery in some reports [8,9] is pelvic sidewall involvement with or without carcinomatosis. In order to achieve R-0 or R-1 resections, we performed the resection outside the anatomical mesorectal fascia in 70% of the hydronephrosis patients and in 39% of the other patients. Despite this extensive surgical procedure in addition to preoperative radiation therapy, locally complete resections of the tumor (R-0) were achieved in 22% (5 patients) of the hydronephrosis group but 41% (69 patients) of the non-hydronephrosis group. R-1 resections were obtained in 43% (10 patients) of the hydronephrosis group and in 35% (61 patients) of the non-hydronephrosis group. We did not manage to remove all macroscopic tumor (R-2) in 35% (8 patients) of the hydronephrosis group and in 24% (40 patients) of the non-hydronephrosis group. Complications Complications are listed in Table 4. One patient (1%) without hydronephrosis died during the first postoperative month due to a subphrenic abscess complicated by pulmonary embolism; no patients in the hydronephrosis group died. Within 3 months postoperatively, 1 patient (4%) with hydronephrosis and 5 (3%) without hydronephrosis died. Five patients (22%) in the hydronephrosis group and 11 (6%) in the non-hydronephrosis group developed postoperative bowel leakage; similar numbers regarding urinary leakage were 1 (4%) and 17 (10%). Nearly 30% had signs of pelvic infection with increased C-reactive protein, fever and/or pelvic abscesses.
Hydronephrosis The obstructions were due to either ureteral tumor compression or tumor invasion. In 9 patients the obstruction was localized just below the common iliac vessels, in 7 along the pelvic sidewall, and in 7 at the base of the bladder/prostate.
Survival The median survival of all patients was 21 months in the hydronephrosis group and 32 months in the patients without
58
S.G. Larsen et al. / The American Journal of Surgery 190 (2005) 55– 60
Table 4 Number of patients with postoperative complications among 193 patients with pelvic recurrence of colorectal cancer
Mortality Leakage Bowel leakage Urinary leakage Infectious complications Deep abscess Septicaemia Wound infection Urinary tract infection Pneumonia Tromboembolic complications Deep venous thrombosis Pulmonary embolism Total
Hydronephrosis (n ⫽ 23)
Not hydronephrosis (n ⫽ 170)
0 5 patients (22%) 5 1 12 patients (52%) 7 0 3 5 3 2 patients (9%) 1 1 13 patients (56%)
1 (.6%) 28 patients (16%) 11 17 79 patients (45%) 42 7 22 45 16 6 patients (4%) 4 2 80 patients (47%)
hydronephrosis. The median survival times among the patients without metastasis (M-0), 16 with hydronephrosis and 133 without, were 27 and 32 months, respectively (Fig. 1). Estimated 5-year survival in patients without metastasis was 11% in the hydronephrosis group and 25% in the nonhydronephrosis group. This difference was statistically different (P ⫽ .05).
Secondary Local Recurrences Among the R-0 or R-1 resections, 9 (60%) patients with hydronephrosis are estimated free of local recurrence after 5 years, as are 69 (53%) among those without hydronephrosis (Fig. 2). This difference was not statistically significant (log-rank .54, P ⫽ .46).
Fig. 1. Kaplan-Meier survival plots for all M-0 patients (without metastasis) with pelvic recurrence from colorectal cancer. Patients with hydronephrosis (n ⫽ 16) and without (n ⫽ 133).
Fig. 2. Estimated Kaplan-Meier local rerecurrence plots for 145 patients with R-0 and R-1 resections after pelvic recurrence from colorectal cancer. Patients with hydronephrosis (n ⫽ 15) and without (n ⫽ 130).
Comments Hydronephrosis developing in patients with a history of colorectal cancer is usually of malignant origin. LevChelouche et al [7] found a benign cause in less than 8% of cases. Our material considers only malignant disease and does not include patients who were rejected for surgery. All patients had hydronephrosis and dilation of the ureter down to an obstruction near the recurrent tumor. Still, the 12% frequency of patients with hydronephrosis is similar to reported frequencies of 7% [11] and 19% [8] in publications regarding recurrent rectal cancers, as well as the 13% found in a mixture of primary and recurrent colorectal cancers [12]. Some reports on recurrent rectal cancer and hydronephrosis were heterogeneous in procedures performed and irradiation given [7–9]. The studies were small and contained 38, 27, and 16 patients with hydronephrosis and recurrent rectal cancer. All patients in 1 study [9] were surgically explored, but only 6 (22%) [8] and 10 (26%) [7] underwent surgical exploration in the 2 other studies. In the latter report, 3 patients were operated on with radical intent and 7 palliatively. In the 2 reports from Roswell Park Cancer Institute [8,9], none of the patients had tumor resections, particularly due to pelvic sidewall involvement. We have resected the tumor in 20 of our 23 patients (87%) with recurrent cancer and hydronephrosis. None of our patients had irradiation as part of the treatment for their primary cancer. Accordingly, they could be given preoperative irradiation for their recurrent cancer, which we consider to be of significant importance. Irradiation did not seem to be an option in the preoperative treatment in the other studies concerning recurrent rectal cancer and hydronephrosis [7–9]. In the present study, locally complete resection was achieved in 22% of the patients with preoperative hydronephrosis. This is half the frequency of R-0 resections in the
S.G. Larsen et al. / The American Journal of Surgery 190 (2005) 55– 60
non-hydronephrosis group, although the difference is not statistically significant. Totally, about two thirds of patients had R-0/R-1 resections in the hydronephrosis group, slightly lower than in the other group. In a previous report we found that patients with R-1 resection appeared to fare better than those with R-2 resections [13]. Thus we suggest that both patients with R-0 and R-1 resections (65%) may benefit from the operation. This correlation between surgical resection stage and prognosis seems to be in accordance with the suggestion of other authors [14,15]. Hydronephrosis is caused by either infiltration of tumor into the ureter or by tumor compressing the ureter. Thus, an R-0 resection may in some cases be achieved without resection of the ureter. Resections of the ureter and/or bladder were performed in 10 of 23 patients (43%) in the hydronephrosis group and decompression in the rest. Investigation of the specimens in the resection group showed direct invasion into or through the affected ureter in 5 patients, tumor in very close distance in 2 (ostium and vesicle), fibrosis after irradiation in 1, and in 2 patients invasion could not be further evaluated. Among the unresected cases, we believe that only the patient with simple urinary diversion had direct tumor invasion, giving an invasion rate after irradiation of at least 23% in the hydronephrosis group, a little higher than the 17% described earlier [7]. Lopez et al [6] described complete ureteral obstruction as “a local sign of invasive carcinoma beyond the limits of operability to cure.” When obstruction was due to compression rather than invasion, they suggested that resection might not be contraindicated. We no longer perform IORT because we did not find any difference between the IORT and non-IORT groups in any R-stages regarding overall survival or local recurrence [4]. We therefore do not think the IORT treatment influenced the results in this study. Cheng and Rodriguez-Bigas et al [8,9] from Roswell Park Cancer Institute reported 22 patients who underwent exploration without any resection of tumor. Their results show survival times of mean 14 and median 8 months, which they claim are similar to those in patients with distant metastasis. Their 5-year survival rate in patients undergoing exploratory laparotomy alone is 0. In our patients with hydronephrosis, the 5-year survival rate for those without metastasis (M-0) is 11%. Our median survival time is 27 months in the hydronephrosis group. It is difficult to determine if it is the operation, adjuvant treatment, or selection of patients that causes such a wide variation in results. In our patients without hydronephrosis and without distant metastasis, the median survival (32 months) is slightly longer and the 5-year survival (25%) higher, which suggests hydronephrosis to be a factor complicating recurrent rectal cancer. An aggressive treatment approach to recurrent rectal cancer may lead to 5-year survival rates up to 54% by selecting patients who could be surgically resected with negative margins [16] and who tolerated extensive surgery. Five-year survival rates of 20% in all patients independent of hydronephrosis status seems acceptable compared to
59
other reports [17]. Some investigators have found the extent of resection to be a significant predictor of improved survival [14,15,18]. It was possible to obtain 38% R-0 resections after extensive surgery, even in the treatment of recurrent cancer. It is generally accepted that the surgical stage is of importance also for the frequency of local re-recurrences. Although a large number of patients with pelvic recurrence following rectal surgery may be re-explored, the number of patients who can be resected for cure (R-0) is often less than 50% [19,20] and varies from 15% [5] to 80% [21]. These variations may partly reflect differences in surgical aggressiveness in operations performed, and they underline the importance of obtaining, at least microscopically, free resection margins even in recurrent disease. Despite extensive surgery, the postoperative mortality rate was low in the total group of patients (.5%) with recurrent colorectal cancer both with regard to 30-day and in-hospital mortality. No patients died within 1 month postoperatively in the hydronephrosis group. In order to operate radically, cystoprostatectomy was performed 5 times more often in the hydronephrosis group and hysterectomy twice as often. Even so, the complication rate was about the same in the groups. We had 193 complications among the 193 patients with apparently more bowel leakage in the hydronephrosis group: 5 patients (22%) in contrast to 11 patients (6%) in the group without hydronephrosis. This could be due to an increased number of resections on small bowel, use of small bowel in deviation after cystoprostatectomy, or simple bias because of small numbers. The lower urinary leakage rate in the hydronephrosis group is again probably due to the smaller number of subjects in this group. The infectious complications from the pelvis may be due to the use of preoperative irradiation [22,23]. Irradiation of the bowel may further enhance bacterial translocation with subsequent infections. Irradiation also depresses the immune system through an effect on resident lyphocytes that may influence the risk of postoperative infections [24]. Patients who undergo major cancer surgery after heavy pelvic irradiation are known to have a high risk of deep venous thrombosis or pulmonary embolism. Accordingly, the result of 3% clinical deep venous thrombosis and 2% clinical pulmonary embolism, including 1 fatal, seems low and acceptable [25]. The morbidity rate varies between 14% and 49%, mostly around 25%, and mortality rates are low, varying from 0% to 3% [18,26]. Higher complication rates were observed in patients who underwent extended resections (32%) and in patients with fixed tumors [26]. In our experience, the relatively high complication rates but low mortality rates seem acceptable when the prolonged survival is taken into account. The 8 patients in need of cystectomy were required double stomas. In a recent report from our department, we found somewhat to our surprise that pelvic exenteration
60
S.G. Larsen et al. / The American Journal of Surgery 190 (2005) 55– 60
with 2 stomas was still consistent with a good quality of life [27]. Multidisciplinary teams Multidisciplinary teams in highly specialized institutions with profound knowledge in preoperative staging and extensive surgical procedures should evaluate patients with advanced tumors.
Conclusion All previous studies on recurrent cancer with hydronephrosis, as well as the present study, included few patients. Thus, no firm conclusions can be drawn. Nevertheless, our study suggests that the presence of hydronephrosis reduces the chance of locally complete resection of recurrence after colorectal cancer. Still, probably about two thirds of patients may benefit from extensive resections, and surgery may be performed with acceptable postoperative mortality and morbidity rates. Surgery also may be an option in patients with local recurrence presenting with hydronephrosis.
References [1] MacFarlane JK, Ryall RDH, Heald RJ. Mesorectal excision for rectal cancer. Lancet 1993;341:457– 60. [2] Enker WE, Thaler H, Cranor ML, et al. Total mesorectal excision in the operative treatment of carcinoma of the rectum. J Am Coll Surg 1995;181:335– 46. [3] Welch JP, Donaldson GA. Detection of treatment of recurrent cancer of the colon and rectum. Am J Surg 1978;135:505– 8. [4] Wiig JN, Tveit KM, Poulsen JP, et al. Preoperative irradiation and surgery for recurrent rectal cancer. Will intraoperative radiotherapy (IORT) be of addidional benefit? A prospective study. Radiother Oncol 2002;62:207–13. [5] Gunderson LL, Haddock MG, Nelson H, et al. Locally recurrent colorectal cancer. IOERT and EBRT ⫹/⫺ 5-FU and maximal resection. Front Radiot Ther Oncol 1997;31:224 – 8. [6] Lopez MJ, Kraybill WG, Downey RS, et al. Exenterative surgery for locally advanced rectosigmoid cancers. Is it worthwhile? Surgery 1987;102:644 –51. [7] Lev-Chelouche D, Keidar A, Rub R, et al. Hydronephrosis associated with colorectal carcinoma: treatment and outcome. Eur J Surg Oncol 2001;27:482– 6. [8] Cheng C, Rodriguez-Bigas M, Petrelli N. Is there a role for curative surgery for pelvic recurrence from rectal carcinoma in the presence of hydronephrosis? Am J Surg 2001;182:274 –7.
[9] Rodriguez-Bigas M, Herrara L, Petrelli N. Surgery for recurrent rectal adenocarcinoma in the presence of hydronephrosis? Am J Surg 1992; 164:18 –21. [10] Baigrie RJ, Berry AR. Management of advanced rectal cancer. Br J Surg 1994;8:343–52. [11] Ulm AK, Klein E. Management of ureteral obstruction produced by recurrent cancer of the rectosigmoid colon. Surg Gynecol Obstet 1960;110:413– 8. [12] Lee PH, Khauli RB, Baker S, et al. Prognostic and therapeutic observations of manifestations in the genitourinary tract of adenocarcinoma of the colon and rectum. Surg Gynecol Obstet 1989;169: 511– 8. [13] Wiig JN, Poulsen JP, Larsen S, et al. Total pelvic exenteration with preoperative irradiation for advanced primary and recurrent rectal cancer. Eur J Surg 2002;168:42– 8. [14] Shoup M, Guillem JG, Alektiar KM, et al. Predictors of survival in recurrent rectal cancer after resection and intraoperative radiotherapy. Dis Colon Rectum 2002;45:585–91. [15] Wallace HJ, Willett CG, Shellito PC, et al. Intraoperative radiation therapy for locally advanced recurrent rectal or rectosigmoid cancer. J Clin Oncol 1995;60:122–7. [16] Curley SA, Carlson GW, Shumate CR, et al. Extended resection for locally advanced colorectal carcinoma. Am J Surg 1992;163:553–9. [17] Abulafi AM, Williams NS. Local recurrence of colorectal cancer: the problem, mechanisms, management and adjuvant therapy. Review. Br J Surg 1994;81:7–19. [18] Wanebo HJ, Koness RJ, Turk PS, et al. Composite resection of posterior pelvic malignancy. Ann Surg 1992;215:685–95. [19] Cunningham JD, Enker W, Cohen A. Salvage therapy for pelvic recurrence following curative rectal cancer resection. Dis Colon Rectum 1997;40:393– 400. [20] Holm T, Cedermark B, Rutqvist LE. Local recurrence of rectal adenocarcinoma after “curative” surgery with and without preoperative radiotherapy. Br J Surg 1994;81:452–5. [21] Lowy AM, Rich TA, Skibber JM, et al. Preoperative infusional chemoradiation, selective intraoperative radiation, and resection for locally advanced pelvic recurrence of colorectal adenocarcinoma. Ann Surg 1996;223:177– 85. [22] Stockholm Rectal Cancer Study Group. Preoperative short-term radiation therapy in operable rectal carcinoma. A prospective randomized trial. Cancer 1990;66:49 –55. [23] Swedish Rectal Cancer Trial. Initial report from a Swedish multicentre study examining the role of preoperative irradiation in the treatment of patients with respectable rectal carcinoma. Br J Surg 1993; 80:1133– 6. [24] Order SE. The effects of therapeutic irradiation on lymphocytes and immunity. Cancer 1977;39:737– 43. [25] Holm T, Rutqvist LE, Johansson H, et al. Postoperative mortality in rectal cancer treated with or without preoperative radiotherapy: causes and risk factors. Br J Surg 1996;83:964 – 8. [26] Hahnloser D, Nelson H, Gunderson LL, et al. Curative potential of multimodality therapy for locally recurrent rectal cancer. Ann Surg 2003;237:502– 8. [27] Guren MG, Wiig JN, Dueland S, et al. Quality of life in patients with urinary diversion after operation for locally advanced rectal cancer. Eur J Surg Oncol 2001;27:502– 8.