Annals of Surgical Oncology Sign the Guestbook
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
10.1245/s10434-006-9082-0
Annals of Surgical Oncology 14:74-83 (2007)
© 2007 Society of Surgical Oncology
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Akasu, T.
Right arrow Articles by Moriya, Y.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Akasu, T.
Right arrow Articles by Moriya, Y.

Original Article

Abdominal Sacral Resection for Posterior Pelvic Recurrence of Rectal Carcinoma: Analyses of Prognostic Factors and Recurrence Patterns

Takayuki Akasu, MD, Takashi Yamaguchi, MD, Yoshiya Fujimoto, MD, Seiji Ishiguro, MD, Seiichiro Yamamoto, MD, Shin Fujita, MD and Yoshihiro Moriya, MD

Colorectal Surgery Division, National Cancer Center Hospital, 5-1-1, Tsukiji, Chuo-ku, Tokyo, 104-0045, Japan

Correspondence: Address correspondence and reprint requests to: Takayuki Akasu, MD; E-mail: takasu{at}ncc.go.jp


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Background: Local recurrence of rectal cancer presents challenging problems. Although abdominal sacral resection (ASR) provides pain control, survival prolongation, and possibly cure, reported morbidity and mortality are still high, and survival is still low. Thus, appropriate patient selection and adjuvant therapy based on prognostic factors and recurrence patterns are necessary. The purpose of this study was to evaluate the results of ASR for posterior pelvic recurrence of rectal carcinoma and to analyze prognostic factors and recurrence patterns.

Methods: Forty-four patients underwent ASR for curative intent in 40 and palliative intent in 4 cases. All but one could be followed up completely. Multivariate analyses of factors influencing survival and positive surgical margins were conducted.

Results: Morbidity and mortality were 61% and 2%, respectively. Overall 5-year survival was 34%. The Cox regression model revealed a positive resection margin (hazard ratio, 10 [95% confidence interval, 3.8–28]), a local disease–free interval of <12 months (4.2 [1.8–9.8]), and pain radiating to the buttock or further (4.2 [1.6–11]) to be independently associated with poor survival. The logistic regression model showed that macroscopic multiple expanding or diffuse infiltrating growths were independently associated with a positive margin (7.5 [1.4–40]). Of the patients with recurrence, 56% had failures confined locally or to the lung.

Conclusions: ASR is beneficial to selected patients in terms of survival. To select patients, evaluation of the resection margin, the local disease–free interval, pain extent, and macroscopic growth pattern is important. To improve survival, adjuvant treatment should be aimed at local and lung recurrences.

Key Words: Therapy • Surgery • Rectal cancer • Local recurrence • Recurrence • Prognostic factor


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Posterior pelvic recurrence1–3 (PPR) of rectal carcinoma, which involves the sacrum and/or sacral nerves, presents challenging clinical problems. It may cause sacral nerve pain, perineal ulcers, fistula formation, bleeding, bowel and/or urinary tract obstruction, sepsis, and, finally, death.4 These conditions are difficult to treat, and chemotherapy provides only minimal benefits at present.4–6 Radiotherapy may give pain relief, but its effectiveness is limited and temporary.4,7–9 Conventional abdominoperineal resection or local excision is only palliative.10,11

In 1981, Wanebo and Marcove11 reported the advantage of the abdominal sacral resection (ASR), which was first described by Brunschwig and Barber12 in 1969, for PPR of rectal carcinoma. Although published data on this operation are still limited and there have been few long-term follow-up studies, this aggressive operation provides pain control, prolongation of survival, and possibly cure.13–22 However, reported morbidity and mortality are significantly high,13–22 and survival is still low.13–22 Therefore, appropriate selection of patients, especially with reference to the probable prognosis, is necessary. In addition, adjuvant therapy based on recurrence patterns may be required. The purpose of this study was to evaluate the results of ASR for PPR of rectal carcinoma and to analyze prognostic factors and recurrence patterns.


   PATIENTS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Between March 1983 and May 2000, 44 patients with PPR of rectal carcinoma that involved the sacrum on computed tomography (CT) were considered candidates for ASR and admitted to the National Cancer Center Hospital, Tokyo. There were 35 men and 9 women, with a median age of 55 years (range, 32–73 years). Of these, 40 patients underwent initial operation at other hospitals. Selection criteria for curative-intent ASR were as follows: (1) medical fitness for ASR; (2) no signs of disseminated disease on preoperative imaging; (3) tumors involving the sacrum but not the first sacral bone and the bony lateral walls; and (4) tumors anatomically confined within the pelvis, with or without resectable solitary liver metastasis. The imaging studies routinely performed before resection were abdominal and pelvic CT, abdominal ultrasonography, and chest roentgenogram until 1989; pelvic magnetic resonance imaging and chest CT were added thereafter.

Of the 44 patients for whom ASR was attempted, 40 received curative-intent ASR, and 4 received palliative-intent ASR because of 1 or 2 lung metastases in 3 and 3 liver metastases in 1. Of the 40 who received curative-intent ASR, 33 patients underwent macroscopic curative ASR, 2 with solitary liver metastasis underwent macroscopic curative ASR with complete resection of liver metastasis, 1 with 4 peritoneal metastases adjacent to the main tumor underwent macroscopic curative ASR with complete resection of peritoneal metastases, and the remaining 4 underwent palliative ASR because of macroscopic residual local tumor in 3 and residual lymph node metastases in 1. Of the four who received palliative-intent ASR, three with lung metastases underwent palliative ASR leaving only residual lung metastases in two and both residual lung and local tumors in one, and one with three liver metastases underwent macroscopic curative ASR with complete resection of liver metastases. Conseuently, 37 underwent macroscopic curative resection, and 7 underwent macroscopic palliative resection. Of them, 27 patients received no radiation, 13 received preoperative adjuvant radiation of 30 to 73 Gy (median, 44 Gy), and 4 received 44 to 50 Gy (median, 50 Gy) as previous treatment.

Data for these patients were collected and entered prospectively into the database of the Colorectal Surgery Division. They included the following: (1) patient demographics; (2) treatment and pathology of the primary rectal cancer; (3) presentation of PPR; (4) treatment and pathology of recurrent tumor; (5) operative details; (6) hospital course, including complications; and (7) outcome. Of these, 15 variables were selected for prognostic factor analysis (Table 1Go) by consideration of their potential relationship to survival after ASR, as indicated by previous studies.13–15,17–19,22 The local disease–free interval (LDFI) was defined as the interval between the initial curative operation and the occurrence of symptoms or detection of asymptomatic PPR by CT.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Univariate Predictors of Adverse Outcome
 
Surgical Procedure
Our surgical procedure was basically similar to that originally described by Wanebo and Marcove11 and Wanebo et al.;13 however, it was slightly modified.23 Our sacral resection was performed immediately after the abdominal phase as a one-stage procedure instead of a two-stage procedure.13 The presence of liver metastasis did not preclude continuation of the procedure if it was solitary and if the disease-free interval was sufficiently long. Solitary liver metastasis was resected simultaneously. We did not make full-thickness fascial myocutaneous flaps for sacroperineal wound closure but sutured the wound simply because there were no patients with large exposed tumors at the perineum.

After the patient was placed in a supine position with flexed and abducted thighs, dissection was started at the aortic bifurcation, and the common and external iliac vessels were dissected. The internal iliac vessels were divided at their root or beyond the superior gluteal artery. Adipose tissue, lymphatics, and the nodes surrounding these vessels, including obturator nodes, were removed completely, and the muscular pelvic side walls and the sacral nerve roots were exposed. The upper limit of the tumor was identified, and the anterior surface of the sacrum was dissected down to the planned level of sacral transection. When the tumor adhered or invaded into urogenital organs, the remaining rectum, pelvic nerves or muscles, and involved organs were all resected en bloc to avoid incomplete resection and cancer cell spillage. To facilitate resection and hemostasis and to shorten operating time, a combined abdominal and perineal approach was used.

After dissection of the lateral, cephalad, anterior, and caudal aspects of the tumor with surrounding organs to be resected was accomplished, the patient was placed in a prone position with flexed and abducted thighs. A posterior sacral incision including the perineal lesion was made, and the sacrum and gluteal muscles were exposed. The gluteal muscles, sacrotuberous ligament, sacrospinous ligaments, and pyriformis muscles were divided as far from the tumor as possible. After the level of abdominal dissection and the extent of the tumor were confirmed by hand in the pelvic cavity, a laminectomy proximal to the planned level of sacral transection was performed to preserve the noninvolved sacral nerve roots and ligate the dura. The sacrum was transected by an osteotome, and en-bloc resection of the tumor with the sacrum and the surrounding organs was accomplished. The gluteal muscles and skin were closed primarily. Again, the patient was placed in a supine position with flexed and abducted thighs. A colostomy and an ileal conduit were made.

Extent of Resection
Levels of sacral transection included S2 in 6 patients, S2–3 in 19, S3 in 5, S3–4 in 11, S4 in 1, and S4–5 in 2. Thirty-nine patients underwent total pelvic exenteration, one underwent posterior pelvic exenteration, and four underwent abdominoperineal resection. En-bloc resection of entire pelvic lymph nodes with the bilateral internal iliac arteries and veins was performed for all patients. Resected organs included the rectum in 20 cases, the urinary bladder in 39, the uterus and vagina in 8, the external genitalia in 2, the obturator internis muscle in 12, the gluteus maximus muscle in 5, and the small intestine in 7. Urinary diversions were an ileal conduit in 37 patients and a ureterocutaneostomy in 2. Three patients underwent complete resection of one, one, and three synchronous liver metastases. In addition, one patient underwent complete resection of four peritoneal metastases.

Follow-Up
One patient returned to Indonesia and was lost to follow-up. The other 43 were followed up completely, with a median follow-up time for live patients of 4.7 years (range, 1.2–15.8 years). They were examined with abdominal and pelvic CT, chest roentgenogram or CT, and carcinoembryonic antigen (CEA) measurement every 4 months for 0 to 1 years, every 6 months for 2 to 4 years, and annually for 5 to 10 years.

Statistical Analysis
Survival, disease-free survival, and local disease–free survival distributions were estimated by using the Kaplan-Meier product-limit method. Univariate comparisons of survival were made by using the log-rank test, and multivariate analysis was performed by using the Cox regression model with the forward stepwise method (likelihood ratio). All variables were dichotomized for analysis. Differences in proportions were analyzed by Fisher’s exact test and by multivariate analysis with the logistic regression model and the forward stepwise method (likelihood ratio). All statistical analyses were performed with SPSS for Windows, version 10.0J (SPSS-Japan Inc., Tokyo, Japan). All P values were two sided, and a P value of <.05 was considered to be statistically significant.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Pathologic Findings
Histological diagnoses of the PPR cases are listed in Table 1Go. The bone marrow or periosteum of the sacrum was histologically involved in 17 patients. The remaining 27 had no sacral invasion, but dense fibrotic tissues adhered extensively to the sacrum, and cancer cells were found within them. Of 13 patients with pelvic lymph node involvement, 12 had intra-pelvic metastases alone, and 1 had both intrapelvic and extrapelvic metastases. Eight patients had distant metastasis, including liver metastasis in three, lung metastasis in three, peritoneal metastasis in one, and distant lymph node metastasis in one.

Resection margins were microscopically negative in 24 patients, microscopically positive in 13, macroscopically positive in 3, and grossly residual in 4 (lung, n = 2; lung and local, n = 1; lymph node, n = 1; Table 1Go). The sites of macroscopic positive margins included cut ends of the sacrum and/or presacral connective tissue in two, cut ends of the sacral nerves and the external iliac artery in one, and the lateral pelvic sidewall in one. The major artery was involved only in one patient with prior extended lateral pelvic lymph node dissection. The sites of microscopic positive margins included the cut end of the sacrum in two, the cut end of the presacral connective tissue in three, the cut ends of the sacrospinous ligament and sacrotuberous ligament in one, the cut ends of the sacrospinous ligament and obturator internis muscle in one, the cut end of the obturator lymph node in one, and the cut ends of the sacral nerves in one.

Macroscopic growth patterns were based on macroscopic views of sections of resected specimens and were classified as solitary expanding growth, multiple expanding growth, and diffuse infiltrating growth (Fig. 1Go; Table 1Go). Expanding growth featured smooth and clear margins. Any tumors showing irregular or obscure margins were therefore classified into the diffuse infiltrating category.


Figure 1
View larger version (57K):
[in this window]
[in a new window]

  		FIG. 1. (A) A section after abdominal sacral resection for posterior pelvic recurrence of rectal carcinoma. This tumor was macroscopically classified as solitary expanding growth. (B) Corresponding magnetic resonance image of (A). (C) A section of tumor macroscopically classified as multiple expanding growth. (D) Corresponding magnetic resonance image of (C). (E) A section of tumor macroscopically classified as diffuse infiltrating growth. (F) Corresponding computed tomography of (E). Arrowheads, main tumor; arrow, satellite tumor. *Sacrum.

 
Morbidity and Mortality
The median operating time was 751 minutes (range, 263–1377 minutes). The median blood loss was 3208 mL (range, 856–26160 mL), and all of the patients underwent transfusion. Of the 27 patients with postoperative complications (morbidity, 61%), 10 (23%) had major complications that necessitated surgical interventions or resulted in hospital death, and 17 (38%) had minor complications that could be managed conservatively (Table 2Go). The number of complications per patient was as follows: 4 in 1 patient, 3 in 5 patients, 2 in 10 patients, and 1 in 11 patients. One patient who had pelvic sepsis, residual tumor regrowth, bowel obstruction, and renal failure died on the 66th postoperative day (mortality, 2%).


View this table:
[in this window]
[in a new window]

  		TABLE 2. Complications
 
Eleven (65%) of 17 patients who had received adjuvant or previous radiation had postoperative complications, compared with 16 (59%) of 27 who had not received radiation (P = .76). In contrast, 7 (41%) of 17 with adjuvant or previous radiation experienced major complications, compared with 3 (11%) of 27 without irradiation (P = .03). The median hospital stay was 38 days (range, 22–316 days).

Survival
The median survival for all the patients undergoing ASR was 2.3 years (range, .1–15.8 years). The estimated overall 1-, 3-, and 5-year survival rates were 90%, 47%, and 34%, respectively, including one hospital death (Fig. 2Go). Of the 15 patients who survived >4 years, 9 were disease free, and 5 survived >8 years. The disease-free 1-, 3-, and 5-year survival rates were 44%, 26%, and 24%, respectively. The local disease–free 1-, 3-, and 5-year survival rates were 63%, 47%, and 47%, respectively (Fig. 2Go).


Figure 2
View larger version (11K):
[in this window]
[in a new window]

  		FIG. 2. Overall, disease-free, and local disease–free survival distributions for the 44 patients undergoing abdominal sacral resection for posterior pelvic recurrence of rectal carcinoma. The numbers in parentheses for the overall survival curve indicate the patients alive at 3 and 5 years.

 
Prognostic Factors
Results of univariate analysis of prognostic factors are summarized in Table 1Go. The overall survival of the patients with microscopic positive resection margins was significantly worse than that of those with microscopic negative margins (P < .0001) but was not significantly better than that of those with macroscopic positive margins or macroscopic residual tumor (P = .11). Patients with macroscopic positive margins or macroscopic residual tumor did not survive >2.3 years.

The survival of patients with buttock pain was significantly worse than that of those without pain or with perineal pain (P = .043) and was significantly better than that of those with thigh or leg pain (P = .0046). The latter died within 1.2 years.

Of the eight patients with distant metastasis, two undergoing resection of solitary liver metastasis were alive and disease free for 7.6 and 2.7 years, one undergoing resection of three liver metastases died of disease at 1.3 years, one undergoing resection of four peritoneal metastases was alive with disease at 1.1 years, three with one or two lung metastases died of disease at 2.3, 2.0, and 1.6 years, and one with para-aortic lymph node metastasis died at 1.7 years.

The univariate analysis of the 15 variables (Table 1Go), when dichotomized, showed a positive resection margin, pain extending to the buttock or further, multiple growths or diffuse infiltrating growth, LDFI of <12 months, a preoperative CEA level >10 ng/mL, and primary cancer stage IV to be associated with significantly worse survival. The other nine factors did not show any significant association with outcome.

The multivariate analysis of the 15 dichotomized variables revealed that only a positive resection margin (hazard ratio, 10 [95% confidence interval, 3.8–28]; P < .001), an LDFI of <12 months (4.2 [1.8–9.8]; P = .001), and pain radiating to the buttock or further (4.2 [1.6–11]; P = .004) were independently associated with worse survival.

When the most significant independent factors were considered together, the 5-year overall survival rates of the 17 patients with microscopic negative margins and an LDFI >12 months (group I), the 7 with microscopic negative margins and an LDFI <12 months (group II), the 10 with positive margins and an LDFI >12 months (group III), and the 10 with positive margins and an LDFI <12 months (group IV) were 67%, 51%, 10%, and 0%, respectively (Fig. 3Go). There were significant survival differences between group I and group III (P < .0001), group III and group IV (P = .0014), and group II and group IV (P = .01). Group IV patients did not survive >2.3 years.


Figure 3
View larger version (11K):
[in this window]
[in a new window]

  		FIG. 3. Overall survival curves for group I (microscopic negative margin and local disease–free interval [LDFI] of >12 months), group II (microscopic negative margin and LDFI <12 months), group III (positive margin and LDFI >12 months), and group IV (positive margin and LDFI <12 months). The numbers in parentheses for each curve indicate the patients alive at 3 and 5 years.

 
Risk Factors for a Positive Resection Margin
To clarify the risk factors for a positive resection margin, the most significant prognostic factor on multivariate analysis, univariate and multivariate analyses were conducted. Three patients who under- went palliative-intent resection as a result of gross residual lung metastases were excluded from this study. Univariate analysis revealed that the incidences of microscopic positive margins were significantly higher in patients with multiple expanding or diffuse infiltrating growth (56% vs. 14%; P = .018) and in patients with pain extending to the buttock or further (72% vs. 30%; P = .029; Table 3Go). On multivariate analysis of the 14 dichotomized variables, excluding resection margin, multiple expanding or diffuse infiltrating growth was independently associated with positive margin (hazard ratio, 7.5 [95% confidence interval, 1.4–40]; P = .019).


View this table:
[in this window]
[in a new window]

  		TABLE 3. Univariate predictors of positive resection margin
 
Recurrence Patterns
Of the 37 patients who underwent macroscopic curative resection, 25 (68%) experienced further recurrence. Sites of their first recurrence after ASR are listed in Table 4Go. Of them, 13 patients (52%) had local failure, 7 (28%) had lung metastasis, and 14 (56%) had failures confined locally or to the lung. Sites of local failure were the cut end of the sacrum in five, the sacral cut end and buttock in one, and the pelvic side wall or ischium in 3. None of the 25 patients with recurrence was treatable by surgery, so these patients were given chemotherapy, radiotherapy, and/or best supportive care.


View this table:
[in this window]
[in a new window]

  		TABLE 4. Sites of first recurrence after abdominal sacral resection in 37 patients undergoing macroscopic curative resection
 
Of the 13 patients who developed local failure, 9 had positive margins, and 4 had negative margins on histological analysis. Of the 24 patients without local failure, 20 had microscopic negative margins, and 4 had microscopic positive margins. The rate for local failure was significantly higher in patients with microscopic positive margins than in those with microscopic negative margins (69% [9 of 13] vs. 17% [4 of 20]; P = .003). When the accuracy of the microscopic status of surgical margins in prediction of local failure was evaluated, the sensitivity was 69% (9 of 13), the specificity was 83% (20 of 24), the positive predictive value was 69% (9 of 13), the negative predictive value was 83% (20 of 24), and the overall accuracy rate was 78% (29 of 37). Of the 13 patients with microscopic positive margins, 9 developed local recurrence that corresponded well to histological findings, 1 experienced local failure at a different site with a positive margin, and 3 had no obvious local failure at the last follow-up.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The most effective treatment for PPR of rectal carcinoma is a curative resection, that is, complete resection with microscopic negative margins.13,15,17–19,22 Because the tumor involves contiguous organs, including the sacrum, retained rectum, internal iliac vessels, and genitourinary organs, by either invasion or dense adhesion, combined resection of these organs—that is, ASR—is mandatory for clear surgical margins and possible cure. The overall 5-year survival rate after ASR is reported to be 25% to 31% in the largest series13,14 and was 34% in this study. Such results have never been achieved with other therapeutic modalities, including chemotherapy and radiotherapy.4–9

However, morbidity and mortality after ASR are reported to be 26% to 82%13,15–18,21,22 and 0% to 9%,13–22 respectively. In our series, they were 61% and 2%, and 23% of our patients experienced major complications resulting in reoperation or death, and their mean hospital stay was 135 days. In addition, most patients lose genitourinary functions and must endure permanent stomas. These costs are very high and sometimes even catastrophic for those who nevertheless do not obtain long-term survival. Therefore, appropriate patient selection based on survival benefit determined on the basis of prognostic factors is necessary. Also, efforts toward seeking effective adjuvant therapy aiming at the most common sites of recurrence are mandatory. Thus, we analyzed prognostic factors and recurrence patterns after ASR in this study.

Several factors that can be estimated before surgery have been reported to be significantly associated with prognosis on either univariate or multivariate analysis. These include residual tumor extent,13,15,17–19,22 distant metastasis,14 initial operation,13 disease-free interval,14 preoperative CEA level,13,14 preoperative CEA doubling time,14 and proliferating cell nuclear antigen labeling index.24 In addition, whether significant or not, there are factors definitely indicative of a poor prognosis. Wanebo et al.13,25 reported that patients with positive margins, bone marrow involvement, or pelvic lymph node involvement had a median survival of only 10 months. Strong suspicion of such factors thus contraindicates ASR. However, the number of patients so far studied is still not sufficiently large to allow definitive patient selection criteria to be established.

We tested 15 factors in multivariate analysis because previous studies indicated their potential relationship to survival after ASR.13–15,17–19,22,24,25 Of these, microscopic positive margins, LDFI <1 year, and preoperative pain exceeding the buttock showed a significant independent association with a poor prognosis. Microscopic margin status is the most significant, as reported so far.13,15,17–19,22 Of our patients with microscopic positive margins, 69% developed local recurrence, and this caused persistent pain and a poor prognosis. Although some previous studies claimed a benefit of palliative resection for both survival and pain,26 it usually leads to a very poor prognosis and fails to relieve pain, as previously reported.25,27 Therefore, palliative resection leaving a gross residual tumor should not be attempted. In addition to conventional imaging,28,29 recent advances in radiological imaging, including thin-section magnetic resonance imaging30 and multidetector row CT,31 allow us to accurately evaluate tumor extent so that cautious interpretation can preclude such unnecessary surgery.

The extent of preoperative pain corresponds well with tumor extent and invasiveness and therefore predicts survival.17 In this study, the survival of the patients with buttock pain was significantly worse than that of patients without pain or with perineal pain and was significantly better than that of patients with thigh or leg pain. Thigh or leg pain, caused by involvement of the first or second sacral nerves, indicates lateral and/or cephalad extension of the tumor, which usually renders curative resection impossible. Indeed, in our series, the affected patients died within 1.2 years. In contrast, if the pain remains within the buttock, there is the possibility of curative resection.

The factors relating to tumor growth rate can predict prognosis only if patients have residual tumors after ASR. Maetani et al.14 and Onodera et al.24 reported a significant association of disease-free interval14 and preoperative CEA doubling time14 with survival. These parameters reflect not only the growth rate of locally recurrent tumors, but also that of distant metastases. The proliferating cell nuclear antigen labeling index24 can reflect a growth rate specific to local recurrence, so it may predict prognosis more accurately. Although LDFI has not been studied so far, it is easier to measure than the labeling index, and it is also specific to local recurrence. As this study showed, patients with an LDFI of >12 months and clear surgical margins are the best candidates for ASR, and a 5-year survival of 67% can be expected. Conversely, if the LDFI is <12 months, thus indicating rapid tumor growth, and resection is palliative, a 2-year survival of only 11% is expected. In such cases, ASR should not be attempted. Palliative resection is indicated only for patients with an LDFI of >12 months and preferably >18 months.11

Primary cancer stage, preoperative CEA level, and macroscopic growth pattern were prognostically significant only in univariate analysis in this study. Thus, they are related to any of the previously described independent factors, but they are worth considering to a certain degree when decisions are made. Macroscopic growth pattern, which has not been investigated so far, especially influences the surgical margin status and is important when deciding the extent of resection.

As our logistic regression model showed, multiple expanding or diffuse infiltrating growth is independently associated with positive resection margins. The curative resection of the tumors with multiple expanding or infiltrating growth (44%) is clearly more difficult than with solitary expanding growth (86%). Therefore, cautious evaluation of both growth pattern and tumor extent by magnetic resonance imaging or CT is needed to determine a correct line of resection.

Although tumor extent (distant and pelvic metastases)14,25 and initial operation type13,25 have been reported to be significant prognostic factors, this was not confirmed here, presumably at least partly because of differences in patient backgrounds and selection criteria. As described previously,11 the presence of pulmonary, multiple liver, peritoneal, and extrapelvic lymph node metastases leads to a very poor prognosis, with a median survival of only 1.6 years in our cases, so these patients should not undergo ASR. However, solitary liver metastasis may be an exception. Indeed, in our series, two patients with solitary liver metastases survived disease free for 7.6 and 2.7 years after ASR and liver resection. In such cases, aggressive surgery seems justified.

Because adjuvant external beam radiotherapy has been reported to be beneficial for local control and prolongation of survival in primary rectal carcinoma,32,33 many surgeons have recommended its application for ASR.13,15–18,20 In this multivariate study, however, a prognostic benefit of preoperative radiotherapy could not be detected. This may be at least partly caused by the small number of patients, so further investigation is necessary. Marijnen et al.34 reported that preoperative radiotherapy for primary rectal cancer has a beneficial effect in patients with more than 1-mm resection margins but that it cannot compensate for microscopically nonradical resection resulting in positive margins. Therefore, preoperative radiation should be given only to patients for whom surgical margins are expected to be attained but insufficient.

The situation with intraoperative radiotherapy may be different.13,15–17 Hahnloser et al.17 reported that the overall 5-year survival rate of patients undergoing palliative resection and intraoperative radiotherapy with or without external beam radiotherapy was 21%. Survival rates for their patients with no fixation, one fixation, two fixations, and three or more fixations were 43%, 24%, 20%, and 0%, respectively. Although candidates for ASR usually have two or more fixations and the expected survival of those with positive margins is not good, intraoperative radiotherapy may benefit those undergoing ASR despite a positive margin.

As to recurrence patterns after ASR, this study showed that, in 56% of our patients, recurrence was confined locally or to the lung. Wanebo et al.13 reported this to be the case for 68% of their series, in line with other previous studies.35,36 Thus, in addition to precise resection based on precise evaluation of tumor extent with thin-section magnetic resonance imaging or multidetector row CT, adjuvant therapies aiming at local and lung recurrences may be necessary. For local control, preoperative and intraoperative radiotherapy may be helpful. For lung metastases, systemic adjuvant chemotherapy using 5-fluorouracil–based chemotherapy or newly developed drugs (or their combination) may be effective.5,6

Although this retrospective exploratory study featured only a relatively small number of patients, we conclude that ASR is beneficial for a selected subset of patients in terms of survival prolongation and even cure. To select appropriate patients, evaluation of resection margin, LDFI, pain extent, and growth pattern is important. To improve survival, adjuvant treatment should be aimed at local and lung recurrences.


   ACKNOWLEDGMENTS
 
Supported by a Grant-in-Aid for Clinical Research for Evidence Based Medicine and a Grant-in-Aid for Cancer Research from the Ministry of Health Labour and Welfare and a grant from the Foundation for Promotion of Cancer Research in Japan.


   FOOTNOTES
 
Presented at the Annual Meeting of the Society of Surgical Oncology, San Diego, California, March 23–26, 2006.

Received for publication April 8, 2006. Accepted for publication May 18, 2006.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Stearns MWJ. Diagnosis and management of recurrent pelvic malignancy following combined abdominoperineal resection. Dis Colon Rectum 1980; 23:359–61.[Medline]
  2. Wanebo HJ, Whitehill R, Gaker D, Wang GJ, Morgan R, Constable W. Composite pelvic resection, an approach to advanced pelvic cancer. Arch Surg 1987; 122:1401–6.[Abstract/Free Full Text]
  3. Wanebo HJ, Koness RJ, Turk PS, Cohen SI. Composite resection of posterior pelvic malignancy. Ann Surg 1992; 215:685–95.[Medline]
  4. Pearlman NW, Stiegmann GV, Donohue RE. Extended resection of fixed rectal cancer. Cancer 1989; 63:2438–41.[CrossRef][Medline]
  5. Simmonds PC. Palliative chemotherapy for advanced colorectal cancer: systematic review and meta-analysis. Colorectal Cancer Collaborative Group. BMJ 2000; 321:531–5.[Abstract/Free Full Text]
  6. Meyerhardt JA, Mayer RJ. Systemic therapy for colorectal cancer. N Engl J Med 2005; 352:476–87.[Free Full Text]
  7. Hoffman JP, Riley L, Carp NZ, Litwin S. Isolated locally recurrent rectal cancer: a review of incidence, presentation, and management. Semin Oncol 1993; 20:506–19.[Medline]
  8. Lingareddy V, Ahmad NR, Mohiuddin M. Palliative reirradiation for recurrent rectal cancer. Int J Radiat Oncol Biol Phys 1997; 38:785–90.[CrossRef][Medline]
  9. Wong CS, Cummings BJ, Brierley JD, et al. Treatment of locally recurrent rectal carcinoma: results and prognostic factors. Int J Radiat Oncol Biol Phys 1998; 40:427–35.[CrossRef][Medline]
  10. Cohen AM, Minsky BD. Aggressive surgical management of locally advanced primary and recurrent rectal cancer. Dis Colon Rectum 1990; 33:432–8.[CrossRef][Medline]
  11. Wanebo HJ, Marcove RC. Abdominal sacral resection of locally recurrent rectal cancer. Ann Surg 1981; 194:458–71.[Medline]
  12. Brunschwig A, Barber HR. Pelvic exenteration combined with resection of segments of bony pelvis. Surgery 1969; 65:417–20.[Medline]
  13. Wanebo HJ, Antoniuk P, Koness RJ, et al. Pelvic resection of recurrent rectal cancer: technical considerations and outcomes. Dis Colon Rectum 1999; 42:1438–48.[CrossRef][Medline]
  14. Maetani S, Onodera H, Nishikawa T, et al. Significance of local recurrence of rectal cancer as a local or disseminated disease. Br J Surg 1998; 85:521–13.[CrossRef][Medline]
  15. Mannaerts GH, Rutten HJ, Martijn H, Groen GJ, Hanssens PE, Wiggers T. Abdominosacral resection for primary irresectable and locally recurrent rectal cancer. Dis Colon Rectum 2001; 44:806–14.[CrossRef][Medline]
  16. Magrini S, Nelson H, Gunderson LL, Sim FH. Sacropelvic resection and intraoperative electron irradiation in the management of recurrent anorectal cancer. Dis Colon Rectum 1996; 39:1–9.[CrossRef][Medline]
  17. Hahnloser D, Nelson H, Gunderson LL, et al. Curative potential of multimodality therapy for locally recurrent rectal cancer. Ann Surg 2003; 237:502–8.[CrossRef][Medline]
  18. Pearlman NW, Stiegmann GV, Donohue RE. Extended resection of fixed rectal cancer. Cancer 1989; 63:2438–41.[CrossRef][Medline]
  19. Schiessel R, Wunderlich M, Herbst F. Local recurrence of colorectal cancer: effect of early detection and aggressive surgery. Br J Surg 1986; 73:342–4.[Medline]
  20. Temple WJ, Ketcham AS. Sacral resection for control of pelvic tumors. Am J Surg 1992; 163:370–4.[CrossRef][Medline]
  21. Takagi H, Morimoto T, Hara S, Suzuki R, Horio S. Seven cases of pelvic exenteration combined with sacral resection for locally recurrent rectal cancer. J Surg Oncol 1986; 32:184–8.[Medline]
  22. Yamada K, Ishizawa T, Niwa K, Chuman Y, Aikou T. Pelvic exenteration and sacral resection for locally advanced primary and recurrent rectal cancer. Dis Colon Rectum 2002; 45:1078–84.[CrossRef][Medline]
  23. Moriya Y, Akasu T, Fujita S, Yamamoto S. Total pelvic exenteration with distal sacrectomy for fixed recurrent rectal cancer in the pelvis. Dis Colon Rectum 2004; 47:2047–53.[CrossRef][Medline]
  24. Onodera H, Maetani S, Kawamoto K, Kan S, Kondo S, Imamura M. Pathologic significance of tumor progression in locally recurrent rectal cancer: different nature from primary cancer. Dis Colon Rectum 2000; 43:775–81.[CrossRef][Medline]
  25. Wanebo HJ, Koness RJ, Vezeridis MP, Cohen SI, Wrobleski DE. Pelvic resection of recurrent rectal cancer. Ann Surg 1994; 220:586–95.[CrossRef][Medline]
  26. Deckers PJ, Olsson C, Williams LA, Mozden PJ. Pelvic exenteration as palliation of malignant disease. Am J Surg 1976; 131:509–15.[CrossRef][Medline]
  27. Cunningham JD, Enker W, Cohen A. Salvage therapy for pelvic recurrence following curative rectal cancer resection. Dis Colon Rectum 1997; 40:393–400.[CrossRef][Medline]
  28. de Lange EE, Fechner RE, Wanebo HJ. Suspected recurrent rectosigmoid carcinoma after abdominoperineal resection: MR imaging and histopathologic findings. Radiology 1989; 170:323–8.[Abstract/Free Full Text]
  29. Farouk R, Nelson H, Radice E, Mercill S, Gunderson L. Accuracy of computed tomography in determining resectability for locally advanced primary or recurrent colorectal cancers. Am J Surg 1998; 175:283–7.[CrossRef][Medline]
  30. Brown G, Richards CJ, Newcombe RG, et al. Rectal carcinoma: thin-section MR imaging for staging in 28 patients. Radiology 1999; 211:215–22.[Abstract/Free Full Text]
  31. Fletcher JG, Johnson CD, Krueger WR, et al. Contrast-enhanced CT colonography in recurrent colorectal carcinoma: feasibility of simultaneous evaluation for metastatic disease, local recurrence, and metachronous neoplasia in colorectal carcinoma. AJR Am J Roentgenol 2002; 178:283–90.[Abstract/Free Full Text]
  32. Gastrointestinal Tumor Study Group Prolongation of the disease-free interval in surgically treated rectal carcinoma. N Engl J Med 1985; 312:1465–72.[Abstract]
  33. Påhlman L. Improved survival with preoperative radiotherapy in resectable rectal cancer. Swedish Rectal Cancer Trial. N Engl J Med 1997; 336:980–7.[Abstract/Free Full Text]
  34. Marijnen CA, Nagtegaal ID, Kapiteijn E, et al. Radiotherapy does not compensate for positive resection margins in rectal cancer patients: report of a multicenter randomized trial. Int J Radiat Oncol Biol Phys 2003; 55:1311–20.[CrossRef][Medline]
  35. Shoup M, Guillem JG, Alektiar KM, et al. Predictors of survival in recurrent rectal cancer after resection and intraoperative radiotherapy. Dis Colon Rectum 2002; 45:585–92.[CrossRef][Medline]
  36. Maetani S, Nishikawa T, Iijima Y, et al. Extensive en bloc resection of regionally recurrent carcinoma of the rectum. Cancer 1992; 69:2876–83.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Clin. Cancer Res.Home page
K. Ezumi, H. Yamamoto, K. Murata, M. Higashiyama, B. Damdinsuren, Y. Nakamura, N. Kyo, J. Okami, C. Y. Ngan, I. Takemasa, et al.
Aberrant Expression of Connexin 26 Is Associated with Lung Metastasis of Colorectal Cancer
Clin. Cancer Res., February 1, 2008; 14(3): 677 - 684.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Akasu, T.
Right arrow Articles by Moriya, Y.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Akasu, T.
Right arrow Articles by Moriya, Y.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS