Annals of Surgical Oncology Cite Track
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
10.1245/s10434-006-9197-3
Annals of Surgical Oncology 14:396-404 (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 Google Scholar
Google Scholar
Right arrow Articles by Fahy, B. N.
Right arrow Articles by Weiser, M. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fahy, B. N.
Right arrow Articles by Weiser, M. R.

Original Article

Carcinoid of the Rectum Risk Stratification (CaRRS): A Strategy for Preoperative Outcome Assessment

Bridget N. Fahy, MD1, Laura H. Tang, MD2, David Klimstra, MD2, W. Douglas Wong, MD1, Jose G. Guillem, MD1, Philip B. Paty, MD1, Larissa K. F. Temple, MD1, Jinru Shia, MD2 and Martin R. Weiser, MD1

1 Department of Surgery, Division of Colorectal Surgery, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021, USA
2 Department of Pathology, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021, USA

Correspondence: Address correspondence and reprint requests to: Martin R. Weiser, MD; E-mail: weiser1{at}mskcc.org


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Background: Predicting rectal carcinoid behavior exclusively on the basis of tumor size is imprecise. We sought to identify factors associated with outcome and incorporate them into a preoperative risk stratification scheme.

Methods: Seventy patients with rectal carcinoid evaluated at our institution were identified. Demographic, clinical, and histopathologic data were collected and correlated with recurrence and survival.

Results: The mean age of our cohort was 53.6 years. Fifty-seven percent of patients were women. The mean tumor size was 1.3 cm (range, .1–5 cm). Twenty-five percent of patients had deeply invasive tumors (into the muscularis propria or deeper); an equal percentage had tumors with lymphovascular invasion (LVI) or an increased mitotic rate (two or more mitoses per 50 high-power fields). Eleven patients (17%) had distant metastases at presentation. Sixty-one patients were followed for a median of 22 months (range, 2–308 months), during which seven patients developed recurrence and seven died of disease (including two of seven whose disease recurred). Poor outcome was associated with large tumor size, deep invasion, presence of LVI, and increased mitotic rate. These factors were incorporated into a Carcinoid of the Rectum Risk Stratification (CaRRS) score. CaRRS predicted recurrence-free and disease-specific survival better than any single factor alone.

Conclusions: Poor prognostic features of rectal carcinoids include large size, deep invasion, LVI, and increased mitotic rate. The CaRRS score incorporates these features and accurately predicts outcome. Because the CaRRS score is based on values available by preoperative biopsy, it can identify patients with favorable prognosis and those with poor prognosis who may benefit from additional staging or surveillance.

Key Words: Carcinoid • Rectum • Outcome • Risk assessment


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Rectal carcinoids are rare, accounting for approximately 15% of all gastrointestinal tract carcinoids and 1.3% of all rectal tumors. Rectal carcinoids tend to exhibit more indolent behavior when compared with carcinoids at other sites. A recent study by Rorstad1 reported that 5-year survival rates for rectal carcinoids varied from 62% to 100%, whereas those for small intestine were 52% to 77% and for colon (excluding appendix) were 33% to 75%.

The wide variability in survival for patients with rectal carcinoids challenges the assumption that these tumors behave in an indolent fashion. Available reports are conflicting regarding what constitutes a virulent carcinoid tumor. Several factors have been studied as possible predictors of behavior in rectal carcinoids. Size of the primary tumor has been most thoroughly examined. The risk of metastasis has been shown by several authors to correlate with size.2–4 According to a report by Mani et al.,3 metastases were found in 2% of patients with tumors <1.0 cm, 10% to 15% in patients with tumors measuring 1.0 to 1.9 cm, and 60% to 80% in patients with tumors measuring >2 cm. Similar findings were reported by Schindl et al.,4 where size of the rectal carcinoid was found to correlate with the presence of regional (lymph node) or distant metastases. However, Jetmore et al.5 found that for 1- to 2-cm tumors, size is not as predictive of behavior.

Depth of invasion has also been implicated as an important prognostic factor in rectal carcinoids. Naunheim et al.6 reported that the interaction between size and depth of invasion was important in predicting the behavior of rectal carcinoids. For small tumors, <2 cm, metastases were seen in 2% of patients if the tumor was contained within the submucosa, and the risk of metastases rose to 48% if the tumor invaded the muscularis propria. These findings were corroborated by Soga,2 who reported a stepwise increase in the risk of metastases with increasing depth of invasion.

Eleven percent to 36% of patients with rectal carcinoids present with distant metastases at the time of diagnosis.7,8 The existing literature is consistent regarding the significance of distant metastases and outcome in patients with rectal carcinoids; no single criterion other than existing distant metastases accurately predicts the behavior of these tumors.

Pathologic features believed to be associated with aggressive behavior include DNA ploidy, invasion of the muscularis propria, increased mitotic figures, vascular or lymphatic invasion, perineural invasion (PNI), and mucin production. Two authors reported that atypical histopathology was correlated with increased tumor size and/or depth of invasion, which was associated with decreased disease-free survival and increased risk of distant metastases.4,8 Rectal carcinoid tumors were classified as atypical if they manifested one of the following features, according to Koura et al.8: lymphatic invasion, anaplasia, frequent mitotic cells, cellular pleomorphism, and mucin production. The relative importance of these individual factors and their impact on clinical outcome has not been studied.

The aims of the current study were twofold: first, to identify clinical or pathologic variables associated with distant disease, recurrence, and disease-related outcome in rectal carcinoids; and second, to develop a preoperative strategy for stratifying rectal carcinoids into low, intermediate, and high risk on the basis of the above variables.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Patients and Methods
All patients evaluated at our institution from 1957 to 2004 with pathologically confirmed diagnosis of rectal carcinoid were retrospectively reviewed. A total of 70 cases of rectal carcinoid tumors were identified during this period. The study design was reviewed and approved by the Institutional Review Board at Memorial Sloan-Kettering Cancer Center. Specific data extracted from the charts included patient age, medical history, family history, presenting signs or symptoms, clinical and pathologic features of the tumor, radiographic and laboratory evaluation, methods of treatment, and outcomes.

Pathologic Evaluation
All cases were reviewed by a single pathologist (L.T.). Carcinoid tumors were defined as having one or more of the typical organoid growth patterns characteristic of well-differentiated endocrine neoplasms along with relatively uniform nuclei having coarsely clumped chromatin. The presence of neuroendocrine differentiation was evidenced by positive immunohistochemical staining for chromogranin or synaptophysin. Importantly, the mitotic rate was consistently less than 10 mitoses per 10 high-power fields (HPF) (or 50/50 HPF), and neuroendocrine neoplasms with a higher mitotic rate were regarded as representing high grade neuroendocrine carcinomas. For each available case, the size of the primary lesion, depth of invasion, presence of lymphovascular invasion (LVI) or PNI, and mitotic rate were evaluated. The size of the tumor was determined from chart review and was not confirmed. The margin status after excision or resection was determined from review of the pathology report, but was not confirmed. Seven patients required an additional procedure because of close or indeterminate excision margin status; all resection margins were reported as negative.

Surgical Management
The method of diagnosing and/or surgically treating rectal carcinoids was noted for each patient. Patients whose tumors were removed either endoscopically or by transanal excision were grouped together and classified as having undergone local excision. Radical resections included either low anterior resection or abdominoperineal resection. Because of the retrospective nature of our study, we were unable to definitively determine why a certain method of surgical management was chosen.

Additional surgical procedures required to treat the primary rectal carcinoid were recorded for each patient, when possible. The frequency and types of nonsurgical treatments provided after surgical biopsy or treatment were noted. Because of the small number of patients who received adjuvant therapies, specific details regarding the type of chemotherapy or dosage of radiation are not presented.

Follow-up
Sixty-one patients were available for follow-up and were followed for a median of 22 months, with a range of 2 to 308 months. The method used to follow up after initial excision or resection was noted for each patient (e.g., physical examination, endoscopy, computed tomography, octreotide scan, endorectal ultrasound). Disease recurrence was defined as local or distant disease diagnosed more than 3 months after the date of the initial excision or resection. In most cases, the diagnosis of recurrent or metastatic disease was made on the basis of the results of biopsy findings of the new lesion or radiographic findings consistent with metastases.

Statistical Analysis
Categorical groups were compared by {chi}2 test. Kaplan–Meier methodology was used for time-to-event analysis and comparisons made with the log rank test. Outcome was measured as recurrence-free survival (RFS) and disease-specific survival (DSS). Differences of P < .05 were considered significant. Time to recurrence or death was calculated from the date of surgery to date of first recurrence, death, or last follow-up. Statistical analysis was performed by SPSS 12.0 software (SPSS, Chicago, IL).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
A total of 70 patients with rectal carcinoids were evaluated at our institution between 1957 and 2004. The demographic, clinical, and pathologic characteristics of the patients are listed in Table 1Go. There was a slight predominance of females in our population. Most tumors were located in the midrectum. Metastatic disease was present in 17.5% of patients at the time of initial presentation; the liver was the most common site of distant metastasis. A slight majority of patients presented with symptoms possibly referable to their rectal carcinoids, with bright red blood per rectum being the most common symptom. No patient presented with symptoms consistent with carcinoid syndrome. Most of the tumors were small; the average tumor size was 1.3 cm. In those for whom depth of invasion data was available (n = 63), most tumors were limited to the submucosa.


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

  		TABLE 1. Characteristics of 70 patients with rectal carcinoids
 
Information regarding LVI was available for 58 of the tumors, of which 24% had evidence of LVI. Twelve patients underwent radical resection; 11 of 12 patients had information available regarding lymph node involvement by the rectal carcinoid. Nine of 11 patients had lymph node metastases. The number of involved nodes ranged from two to eight, with a median of three nodes involved. Data regarding PNI was available for 59 of the tumors; 12% were found to have PNI. Tumors were divided into those with low (<2/50 HPF) and increased (≥2/50 HPF) mitotic rate. Mitotic rate data were available for 60 of the tumors; 28% had an increased mitotic rate. Tumors with an increased mitotic rate had a mean rate of 10.4/50 HPF (range, 2–50/50 HPF).

The method of treatment is summarized in Fig. 1Go. Surgical treatment information was available for 69 patients. Most patients, 70%, received local excision of their tumors. Additional operative procedures were performed in 11 patients, and 10 patients received some form of additional nonsurgical treatment (chemotherapy or radiation).


Figure 1
View larger version (14K):
[in this window]
[in a new window]

  		FIG. 1. Treatment strategies for all patients.

 
Nine patients were not available for follow-up and were excluded from the survival analysis. Seven patients developed recurrent disease at a median of 29 months after diagnosis. Two patients developed local recurrence, and five patients developed hepatic recurrence. At the time of last follow-up, 75% of patients were without evidence of disease, and an equal percentage of patients were either alive with disease or dead of disease (11.5%).

Distant Metastasis at Presentation
The key clinical and pathologic characteristics of rectal carcinoids in patients with and without distant metastases at presentation are listed in Table 2Go. Patients with metastatic disease were far more likely to have larger tumors, deeper tumors, tumors with LVI, and an increased mitotic rate. Although most patients who presented with metastatic disease were dead of disease at last follow-up, 40% of patients who initially presented with metastatic disease were still alive after a median follow-up of 20 months.


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

  		TABLE 2. Clinicopathologic and outcomes in patients with and without distant metastases at initial presentation
 
Patients Who Developed Recurrent Disease
A total of seven patients developed recurrent disease during the course of follow-up; the clinical and pathological features of their tumors are listed in Table 3Go. Two patients had tumors <2 cm; these same two patients had tumors confined to the submucosa. Three of five patients with available data had LVI present in their tumors. PNI was present in one of five patients for whom these data were available. An increased mitotic rate was found in five of six tumors. Two patients treated by local excision experienced local recurrence, and one of these patients developed a distant recurrence. The site of recurrence in patients initially managed by radical resection was the liver. Recurrence occurred between 10 and 92 months after initial resection, and tended to recur longer after radical resection (range, 10–92 months; average, 43.3 months) than after local excision (range, 14–29 months; average, 20.3 months). Most patients were alive with disease at last follow-up (n = 4); only one patient was without disease, and two patients had died of their disease.


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

  		TABLE 3. Clinical and pathological profile of patients who developed recurrent disease
 
Size of Primary
Because size of the primary tumor has historically been considered the most important factor predicting the behavior of rectal carcinoids, we sought to examine the interrelationship between size and the other clinical and pathologic variables included in the current study. Patients were divided on the basis of the size of their primary lesion (<1 cm, 1 to 1.9 cm, ≥2 cm) and then analyzed for differences in demographic, clinical, and pathologic factors (Table 4Go). There was no association between size of the primary tumor and any of the demographic variables. Only one patient with a tumor <2 cm presented with metastatic disease. The pathologic variables statistically significantly associated with the size of the primary lesion included depth of invasion, presence of LVI, and mitotic rate. Two patients (4.5%) with tumors <2 cm developed recurrent disease during the period of follow-up, whereas 4 (36%) of 11 of patients with tumors ≥2 cm developed recurrent disease. No patient with a tumor <2 cm died of disease during the period of follow-up, whereas 46% of patients with tumors ≥2 cm were dead of disease at last follow-up.


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

  		TABLE 4. Clinicopathologic findings and outcomes in patients based on size of tumor
 
Factors Associated With RFS and DSS
Data relating to univariate analysis of RFS and DSS are listed in Table 5Go. Five-year RFS was associated with tumor size, depth of penetration into the intestinal wall, and mitotic rate. Five-year DSS was associated with the above factors as well as LVI. Because of the small numbers in the study, multivariate analysis was not performed.


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

  		TABLE 5. Factors associated with recurrence-free survival (RFS) and disease-specific survival (DSS)
 
Carcinoid of Rectum Risk Stratification (CaRRS)
The results of the analyses above demonstrated that the factors that correlated most closely and consistently with poor outcome included size, depth of invasion, presence of LVI, and mitotic rate. In addition to being reliably associated with virulent behavior or poor outcome, these factors have the advantage of being determined by initial biopsy and can therefore be used as part of a preoperative risk stratification. We incorporated these four factors into a risk score by assigning points to the various features, as listed in Table 6Go. Features assigned 0 points were: size <1 cm, invasion confined to the mucosa/submucosa, no LVI, mitotic rate <2/50 HPF. Features assigned 1 point were: size between 1 and 1.9 cm, invasion into the muscularis propria or deeper, presence of LVI, and mitotic rate ≥2/50 HPF. Two points were assigned to tumors ≥2 cm. The points were then totaled to yield an overall risk score. Patients were then categorized into low, medium, and high risk on the basis of total scores of 0, 1 to 2, and ≥3 points, respectively. The CaRRS score accurately predicted RFS and DSS outcome. As seen in Fig. 2Go, patients with low risk tumors had a much higher 5-year RFS than patients with either intermediate or high-risk tumors. Figure 3Go illustrates the 5-year DSS curve based on the CaRRS groups and shows a marked increased 5-year DSS for patients with low-risk tumors compared with those with high-risk tumors. Patients with intermediate-risk tumors did not greatly differ from other two groups in terms of 5-year DSS.


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

  		TABLE 6. Carcinoid of the rectum risk stratification (CaRRS)
 

Figure 2
View larger version (12K):
[in this window]
[in a new window]

  		FIG. 2. Recurrence-free survival for patients based on Carcinoid of the Rectum Risk Stratification (CaRRS) Score.

 

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

  		FIG. 3. Disease-specific survival for patients based on Carcinoid of the Rectum Risk Stratification (CaRRS) score.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Not all rectal carcinoids behave in an indolent fashion; there seems to be a subset of tumors that behaves aggressively, either because of distant metastasis or a tendency toward recurrence. The aim of the current study was to determine which demographic, clinical, and histopathologic features are associated with distant metastases, recurrence, or death. Our secondary aim was to combine these features into a preoperative risk assessment score that could identify patients with aggressive tumors so they may be considered for more extensive staging evaluation and/or follow-up.

No demographic factors were able to identify patients with aggressive rectal carcinoids. In the current study, the four factors that correlated most closely with aggressive behavior and/or poor outcome were: size of the primary tumor, depth of invasion, presence of LVI, and mitotic rate. Tumor size correlated with the presence of metastatic disease; only one patient with a tumor <2 cm (n = 56) presented with distant metastasis, whereas 64% of patients with tumors ≥2 cm presented with metastatic disease. This trend is consistent with the frequency of metastatic disease reported by Soga9 in his analysis of 849 rectal carcinoids. The risk of metastases for rectal carcinoids ≤1 cm was 10%, 18% for tumors 1.1 to 2.0 cm, and 57% for tumors ≥3 cm. Of note, however, both in our study and the reports by Soga9 and Naunheim et al.,6 metastases were occasionally seen in tumors <2 cm, illustrating that the risk of distant metastases is not governed by size of the primary tumor alone.

Size of the rectal carcinoid was also statistically significantly associated with 5-year RFS and DSS. Patients with tumors <1 cm had a markedly higher 5-year RFS (92%) compared with patients with tumors ≥2 cm (30%). Five-year DSS was also significantly higher in patients with tumors <1 or 1 to 1.9 cm compared with those with tumors ≥2 cm. Our findings are consistent with those of Koura et al.,8 who reported a statistically significant correlation between tumor size and metastasis-free survival. Five-year metastasis-free survival was 100% in patients with tumors <1 cm, 65% in patients with tumors 1 to 1.9 cm, and 25% in patients whose tumors were ≥2 cm.

Depth of tumor invasion was also statistically significantly associated with the presence of distant metastases at presentation and both RFS and DSS. Seven of 16 patients with deep lesions presented with metastatic disease, compared with only 1 of 39 patients with superficial invasion. The 5-year RFS in patients whose tumors invaded into the muscularis propria or deeper was 30%, compared with an 88% 5-year RFS in those whose tumors were limited to the mucosa/submucosa. A 68% decrease in 5-year DSS was seen in patients with deeper tumors compared with those whose tumors were limited to the mucosa or submucosa (32% vs. 100%, respectively). Several other authors have also noted the importance of depth of invasion in predicting the behavior of rectal carcinoids.3,7,10 Soga9 reported a 5-year overall survival for all rectal carcinoids of 88%, while that for patients whose tumors were confined to the submucosa was 98%.

Aside from size and depth of invasion, few other features of rectal carcinoids have been systematically examined for their effect on outcome. The results of the current study showed that LVI and mitotic rate were consistently associated with poor outcomes, such as distant metastases, recurrence, and diminished survival. All patients who presented with distant metastases had LVI in their tumors, in contrast to the 18% incidence of LVI found among patients who did not present with distant metastases. The presence of LVI was also associated with a 30% reduction in RFS and DSS. Koura et al.8 is among the few authors who has studied the importance of features such as LVI and PNI on behavior of rectal carcinoids. "Typical" carcinoids, in their study, had no or rare mitotic cells and did not invade the lymphatics, blood vessels, perineurium, or muscularis propria. "Atypical" carcinoids, in contrast, had at least one of these features. Five-year metastasis-free survival was 100% in 16 patients with "typical" carcinoids versus 50% in four patients with "atypical" carcinoids.

We chose to study the mitotic rate of rectal carcinoid tumors because previous studies have demonstrated the prognostic significance of mitotic rate in distinguishing low-grade from intermediate-grade pancreatic endocrine neoplasms11 and pulmonary carcinoids.12 Additionally, mitotic rate is widely considered to be one of the most important measures of aggressiveness in well-differentiated endocrine neoplasms. We found that six of seven patients who presented with distant metastases had a mitotic rate of ≥2/50 HPF, in contrast to only 24% (11 of 46) of patients who did not present with metastases. Similarly, five of six patients who developed recurrence had tumors with an increased mitotic rate. Decreased 5-year RFS and tumors with DSS were seen in patients with an increased mitotic rate (57% and 70%, respectively).

Federspiel et al.7 were among the first to evaluate mitotic rate in colorectal carcinoids from a cohort of 35 patients. Only three of their patients (9%) had a mitotic rate >0/10 HPF; all patients who did not present with metastases and were alive at last follow-up had amitotic carcinoids. Although other authors have included mitotic rate under a general categorization of "atypical" versus "typical" carcinoids,4,8 to our knowledge, we are the first to independently evaluate the significance of mitotic rate on outcome in patients with rectal carcinoid, and to show its consistent and statistically significant association with distant metastases, RFS, and DSS.

Our small sample size prevented us from performing multivariate analyses, which could determine the independent risk factors for poor outcome in rectal carcinoids. Instead, we examined the frequency of adverse clinical and histopathologic features in relationship to tumor size, which has traditionally been considered the primary governor of rectal carcinoid behavior. Our results are consistent with the existing literature, which notes that although size is an important factor in predicting the behavior of rectal carcinoids, it does not tell the whole story.2,8,13 We found several instances in which poor histopathologic features were seen in small tumors: three of seven tumors that recurred were <2 cm, 25% of patients with tumors that invaded into the muscularis propria or deeper were <2 cm, 43% of tumors with LVI were <2 cm, and 50% of tumors with a mitotic rate ≥2/50 HPF were <2 cm in size. Additionally, there was generally a stepwise increase in the frequency of poor prognostic features with increasing tumor size, such that tumors 1 to 1.9 cm had an increased frequency of poor characteristics compared with tumors <1 cm, and a lower frequency of these characteristics compared with tumors ≥2 cm. Other investigators have also found that factors other than size alone affects outcome and behavior in rectal carcinoids. Naunheim et al.6 reported that the risk of metastases for tumors <2 cm was 6% overall, and ranged from 2% if the tumor was contained within the submucosa to 48% if the tumor invaded into the muscularis. Koura et al.8 found that the frequency of atypical histopathology increased with increasing size of the tumor. Twenty-five percent (4 of 16) of patients with tumors <1 cm had atypical carcinoids compared with 50% (4 of 8) of patients with tumors 1 to 1.9 cm, and a 75% incidence of "atypical" carcinoids in patients with tumors ≥2 cm (1 of 4).

The second aim of our study was to develop a strategy for stratifying rectal carcinoids into low-, intermediate-, and high-risk groups on the basis of the factors found to be associated with distant metastases, recurrence, and survival. The CaRRS score was determined by assigning points to the four variables identified as important in determining the behavior of rectal carcinoids: size, depth of invasion, LVI, and mitotic rate. We found that our risk groups were able to distinguish patients in terms of their RFS and DSS (Figs. 2Go and 3Go). We were able to show that patients in the low-risk group had essentially no risk of recurrence and a 100% 5-year DSS.

Schindl et al.4 were among the few investigators to attempt to stratify rectal carcinoids by more than a single factor (e.g., size, depth of invasion). They classified rectal carcinoids as either benign (<2 cm, had a typical growth pattern, local invasion, and no regional or distant metastases) or malignant (>2 cm, or had atypical histology, or regional or distant metastases). Sixteen patients with benign tumors underwent local excision with no local recurrence or death from disease after a mean follow-up of 98 months. In contrast, 4 (28.5%) of 14 patients with malignant tumors developed local recurrence, and 5 (36%) of 14 died of disease.

The universally good outcome of our low-risk patients has implications for preoperative staging and follow-up. In patients with tumors <1 cm, limited to the mucosa/submucosa, without LVI, and a mitotic rate <2/50 HPF, no additional staging seems necessary because the risk of distant disease is negligible. Additionally, no unique follow-up is required in these patients because they are at no appreciable risk for recurrence or disease-related death. For patients with intermediate-risk tumors, closer follow-up seems justified and should be continued for a minimum of 10 years, because recurrence was seen in our patients up to 7.5 years after initial resection. In addition, these patients should be considered for additional staging tests (e.g., computed tomography, endorectal ultrasound, scintigraphy) to identify occult metastases, as was seen in one of our patients with an intermediate-risk tumor. Finally, in patients with high-risk lesions, both rigorous pretreatment staging evaluation and close follow-up are warranted, given their great risk for distant metastases (7 of 15, 47%), recurrence (4 of 13, 31%), and disease-related death (4 of 15, 27%) (Fig. 3Go).

We are unable to make specific surgical treatment recommendations based on the results of the current study because of its retrospective nature. However, given the excellent outcome in patients with low-risk lesions (as defined above) after local excision, these patients can likely be safely managed by local excision alone. The optimal method of surgical management is a more challenging question for patients with intermediate- and high-risk lesions. Of the 14 patients with intermediate-risk lesions who underwent local excision, two patients developed local recurrence (14%). Although radical treatment would result in overtreatment if all patients with intermediate-risk lesions were subjected to such procedures, consideration of a radical procedure seems justified given the demonstrated risk of local recurrence in this group after local excision. Of the 15 patients with high-risk lesions, three were treated by local excision, and disease in one of these patients recurred distantly. The ability of a radical resection to reduce this patient’s risk of distant recurrence is unknown; therefore radical resection cannot be routinely recommended in all patients with high-risk rectal carcinoids.

The optimal method for following these patients after excision or resection is currently unknown. Patients in the current study were followed by a variety of modalities: physical examination, computed tomography, lower endoscopy, octreotide scan, and endorectal ultrasound. The most common surveillance methods were lower endoscopy, followed by computed tomography and physical examination alone. Endorectal ultrasound has only recently been incorporated into our surveillance program for these patients, but it is currently being used by most of the surgeons in our group. Only eight patients in the current study have been followed by endorectal ultrasound, and so we are unable to comment on its usefulness in following patients with rectal carcinoids. Serum chromogranin A has been found by some investigators to be a useful tumor marker for following patients with gastrointestinal carcinoids,14 but this was not routinely assessed in our patients.

The findings of the current study highlight several important features of rectal carcinoids: (1) not all rectal carcinoids behave in an indolent fashion; (2) size alone is an inadequate indicator for predicting the behavior of rectal carcinoids; and (3) depth of invasion, LVI, and mitotic rate are important histopathologic features in these tumors. Additionally, we have proposed a simple risk stratification scheme that can help predict the risk of RFS and DSS on the basis of four clinical and histopathological variables readily available from analysis of biopsy samples or endoscopic excisional material.

Received for publication June 20, 2006. Accepted for publication July 13, 2006.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Rorstad O. Prognostic indicators for carcinoid neuroendocrine tumors of the gastrointestinal tract. J Surg Oncol 2005; 89:151–60.[CrossRef][Medline]
  2. Soga J. Carcinoids of the rectum: an evaluation of 1271 reported cases. Surg Today 1997; 27:112–9.[CrossRef][Medline]
  3. Mani S, Modlin IM, Ballantyne G, Ahlman H, West B. Carcinoids of the rectum. J Am Coll Surg 1994; 179:231–48.[Medline]
  4. Schindl M, Niederle B, Hafner M, et al. Stage-dependent therapy of rectal carcinoid tumors. World J Surg 1998; 22:628–33; discussion 634.[CrossRef][Medline]
  5. Jetmore AB, Ray JE, Gathright JB Jr, McMullen KM, Hicks TC, Timmcke AE. Rectal carcinoids: the most frequent carcinoid tumor. Dis Colon Rectum 1992; 35:717–25.[CrossRef][Medline]
  6. Naunheim KS, Zeitelrs J, Kaplan EL, et al. Rectal carcinoid tumors—treatment and prognosis. Surgery 1983; 94:670–6.[Medline]
  7. Federspiel BH, Burke AP, Sobin LH, Shekitka KM. Rectal and colonic carcinoids. A clinicopathologic study of 84 cases. Cancer 1990; 65:135–40.[CrossRef][Medline]
  8. Koura AN, Giacco GG, Curley SA, Skibber JM, Feig BW, Ellis LM. Carcinoid tumors of the rectum: effect of size, histopathology, and surgical treatment on metastasis free survival. Cancer 1997; 79:1294–8.[CrossRef][Medline]
  9. Soga J. Early-stage carcinoids of the gastrointestinal tract: an analysis of 1914 reported cases. Cancer 2005; 103:1587–95.[CrossRef][Medline]
  10. Shirouzu K, Isomoto H, Kakegawa T, Morimatsu M. Treatment of rectal carcinoid tumors. Am J Surg 1990; 160:262–5.[Medline]
  11. Hochwald SN, Zee S, Conlon KC, et al. Prognostic factors in pancreatic endocrine neoplasms: an analysis of 136 cases with a proposal for low-grade and intermediate-grade groups. J Clin Oncol 2002; 20:2633–42.[Abstract/Free Full Text]
  12. Travis WD, Rush W, Flieder DB, et al. Survival analysis of 200 pulmonary neuroendocrine tumors with clarification of criteria for atypical carcinoid and its separation from typical carcinoid. Am J Surg Pathol 1998; 22:934–44.[CrossRef][Medline]
  13. Heah SM, Eu KW, Ooi BS, Ho YH, Seow-Choen F. Tumor size is irrelevant in predicting malignant potential of carcinoid tumors of the rectum. Tech Coloproctol 2001; 5:73–7.[CrossRef][Medline]
  14. Nikou GC, Lygidakis NJ, Toubanakis C, et al. Current diagnosis and treatment of gastrointestinal carcinoids in a series of 101 patients: the significance of serum chromogranin-A, somatostatin receptor scintigraphy and somatostatin analogues. Hepatogastroenterology 2005; 52:731–41.[Medline]




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 Google Scholar
Google Scholar
Right arrow Articles by Fahy, B. N.
Right arrow Articles by Weiser, M. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fahy, B. N.
Right arrow Articles by Weiser, M. R.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS