Is T classification still correlated with lymph node status after preoperative chemoradiotherapy for rectal cancer?
It is well known that the risk of lymph node involvement increases according to pathologic T classification in rectal cancers, but to the authors' knowledge, the correlation between risk of lymph node involvement and ypT classification in rectal cancers treated with preoperative chemoradiotherapy (CRT) remains unclear. The current study investigated the correlation between tumor involvement in regional lymph nodes and rectal mural tumor status in patients who underwent preoperative CRT for rectal cancer.
Between October 2001 and February 2005, 282 patients underwent preoperative CRT followed by proctectomy for locally advanced rectal adenocarcinoma. Correlations between lymph node status and ypT classification, Dworak regression grade, and magnetic resonance (MR) volumetry findings were explored.
Lymph nodes harboring tumors were found in 87 of 282 (30.9%) patients. The rate of lymph node involvement was found to be correlated with ypT-classification (P < .001); positive lymph nodes were detected in 1 of 45 (2.2%) ypT0 patients, 1 of 13 (7.7%) ypT1 patients, 13 of 77 (16.9%) ypT2 patients, 69 of 140 (49.3%) ypT3 patients, and 3 of 7 (42.9%) ypT4 patients. The rate of lymph node involvement decreased as Dworak regression grade increased (P < .001); tumor-harboring lymph nodes were found in 62.3% of Grade 1 patients, 31.4% of Grade 2 patients, 16.1% of Grade 3 patients, and 2.2% of Grade 4 patients. There were no differences noted with regard to MR volumetry findings, including mean volume of pre- or post-CRT tumor and the tumor volume reduction rate between lymph node-negative and lymph node-positive patients.
Pathologic T classification is still the most reliable predictor of lymph node metastasis in rectal cancer patients who have undergone preoperative CRT. The risk of lymph node metastasis was found to be 3.4% in rectal cancer that had regressed to ypT0 or ypT1. Cancer 2006. © 2006 American Cancer Society.
The current conventional treatment for locally advanced rectal cancer is surgery followed by adjuvant chemoradiotherapy (CRT), which has significantly improved local control and overall survival compared with surgery alone.1 Since the late 1990s, several randomized studies have reported advantages of preoperative CRT followed by surgery in terms of local control and overall survival.2–4 Recently, a prospective randomized study with a large sample size and long-term follow-up reported that, compared with postoperative CRT, preoperative CRT improved local control and was associated with reduced toxicity, increased sphincter preservation, and comparative overall survival.5 As preoperative CRT has become more popular, several prognostic factors have been proposed, and lymph node status has been shown to be a very important prognostic factor associated with disease-free survival.6
The most likely explanation for the long-term efficacy of preoperative CRT is the subsequent downstaging of both mesorectal lymph nodes and the primary rectal tumor. With this potential benefit of preoperative CRT, some studies reported that preoperative CRT followed by local excision could be performed for selected patients not considered to be suitable for major abdominal surgery,7–9 and reported the rate of local failure as 4% to 20%. Local excision can provide cure with less morbidity for patients with tumors confined to the rectal wall and not involving mesorectal lymph nodes. Therefore, lymph node status should be accurately assessed in order for local excision to be an acceptable treatment after preoperative CRT. Whereas several methods such as transrectal ultrasonography (TRUS) and magnetic resonance imaging (MRI) have been used in attempts to identify tumor involvement in mesorectal lymph nodes, such approaches have provided disappointing results.10, 11
It is well known that the risk of lymph node involvement increases according to pathologic T (pT) classification in rectal cancers.12, 13 For rectal cancers treated with preoperative CRT, post-CRT pathologic T (ypT) classification has also been reported to be correlated with lymph node status, and some authors suggested using transanally resected tumor pathology results as determinants for whether additional radical proctectomy should be performed.14, 15 However, other authors have disagreed with the use of ypT-classification as a predictor of lymph node status.16 To our knowledge, the clinical implications for using ypT-classification as a predictor of lymph node status remain largely unknown. The current study evaluated the relation between ypT-classification and lymph node status in patients who underwent preoperative CRT for rectal cancer.
MATERIALS AND METHODS
Between October 2001 and February 2005, 341 patients with rectal cancer underwent preoperative CRT at the National Cancer Center, Republic of Korea. The inclusion criteria used during retrospective reviewing of clinical data were: 1) histologically confirmed adenocarcinoma of the middle or distal rectum (< 8 cm above the anal verge); 2) locally advanced stage of disease (cT3 or cT4: tumor through the muscularis propria into the subserosa [T3], tumor directly invades other organs or structures, and/or perforates the visceral peritoneum [T4]) evaluated by MRI with or without TRUS; 3) no distant metastasis in staging workup; 4) no previous or concurrent malignancy; and 5) available surgical record and pathology report. The study was performed in accordance with the guidelines of our institutional review board and informed consent was obtained from each patient before preoperative CRT.
Of the 294 patients who met those criteria, 7 were treated with transanal excision because their comorbidities made general endotracheal anesthesia and radical proctectomy impossible. Distant metastases that were undetected in initial staging workups were discovered in 5 patients during open surgery. These 12 patients were excluded from analysis, leaving a study population of 282 patients who had received preoperative CRT followed by radical proctectomy.
Preoperative radiotherapy of 45 grays (Gy) given in 25 fractions was delivered to the pelvis, followed by a boost of 5.4 Gy given in 3 fractions to the primary tumor over 5.5 weeks. All patients underwent computed tomography (CT) simulation for three-dimensional conformal radiotherapy, and the target volume for radiotherapy encompassed all detectable tumors observed on CT and MRI. The superior border was placed at the level of L5/S1, and the inferior border at >3 cm caudal to the macroscopic tumor. All patients underwent three-dimensional treatment planning and the prescription dose was specified at the isocenter. A 3-field treatment plan was composed of a 6-megavolt (MV) photon posterior-anterior field and 15-MV photon opposed lateral fields with wedges of 45°. The beam weights of the 3-field plan were optimized to minimize the maximum dose within the target volume.
Chemotherapy was initiated on Day 1 of pelvic radiotherapy. One of 2 chemotherapy regimens were used: 5-fluorouracil and leucovorin (FL) for 143 (50.7%) patients (from October 2001 to June 2004), or capecitabine for 139 (49.3%) patients (from April 2003 to February 2005). In the FL group, 2 cycles of an intravenous bolus injection of 5-fluorouracil (at a dose of 400 mg/m2/day) and leucovorin (at a dose of 20 mg/m2/day) for 3 days in the first and fifth weeks of radiotherapy were administered. Capecitabine was delivered orally at a dose of 825 mg/m2 twice daily for the duration of radiotherapy without weekend breaks. Because capecitabine is supplied as 150-mg and 500-mg tablets, the administered dose was rounded up.
Approximately 4 to 8 weeks (median, 6 wks) after the completion of preoperative CRT, all patients underwent open radical proctectomy with total mesorectal excision. Anastomoses were performed using either the double-stapled (mainly for colorectal anastomosis) or hand-sewn (mainly for coloanal anastomosis) technique.
Postoperative chemotherapy was initiated at approximately 3-6 weeks after curative resection for all patients regardless of their pathologic stages. In the FL group, 4 cycles of a monthly intravenous bolus injection of 5-fluorouracil (at a dose of 400 mg/m2/day on Days 1-5) and leucovorin (at a dose of 20 mg/m2/day on Days 1-5) were delivered. In the capecitabine group, 6 cycles of capecitabine (at a dose of 1250 mg/m2 twice daily for 14 days, followed by 7 days rest at each cycle) were administered.
Before preoperative CRT, staging workups were performed on all patients and involved digital rectal examination, complete blood count, liver function test, level of carcinoembryonic antigen, video colonoscopy, chest radiography, CT scanning of the abdomen and pelvis, and MRI scanning with or without TRUS. Tumor staging was determined by MRI. MR T3 lesions were defined as those in which the tumor signal intensity extended through the muscle layer into the perirectal fat, with obliteration of the interface between the muscle and perirectal fat. MR T4 lesions were defined as those in which the tumor signal intensity extended into an adjacent structure or viscus.
To measure tumor response quantitatively, MR volumetry was performed on 2 occasions: at the initial workup and just before surgery after completion of CRT. Cross-sectional areas of lesions were measured using axial T2-weighted images by tracing the lesion boundary. On T2-weighted images, cross-sectional lesion areas were defined as intermediate signal intensity areas that differed from the normal adjacent rectal wall in terms of signal intensity and contour. MR-based tumor volume was calculated for each MR scan to determine volume reduction rate. Volume reduction rate was determined using the equation R (%) = (VpreCRT − VpostCRT) × 100/VpreCRT, in which the tumor volume reduction rate (R) equals the post-CRT tumor volume (VpostCRT) subtracted from the pre-CRT tumor volume (VpreCRT) divided by the pre-CRT tumor volume.
After surgery, the post-CRT tumor stage was determined according to the TNM classification system recommended by International Union Against Cancer and American Joint Committee on Cancer.17 Pathologic therapeutic response to preoperative CRT was also evaluated in all resected specimens. In each case, the entire tumor plus the pericolic and mesenteric fat was serially sliced into 4-mm thick sections and embedded in paraffin. The tumor regression grade was microscopically evaluated using the scale proposed by Dworak et al.18 Regression was graded as follows: Grade 0: no regression; Grade 1: dominant tumor mass with obvious fibrosis and/or vasculopathy; Grade 2: dominant fibrotic changes with some obvious tumor cells or groups of cells; Grade 3: fibrotic tissue with or without mucous substance that contained tumor cells that were very few in number and were difficult to detect microscopically; and Grade 4: fibrotic mass or acellular mucin pools only with no detectable tumor cells (i.e., complete response [CR]).
Gender (male vs. female), chemotherapy regimen (FL vs. capecitabine), lymph node status (positive vs. negative), and primary tumor response (CR vs. non-CR) were considered binary variables. The ypT-classification (ypT0: no tumor cell found; ypT1: tumor confined to the mucosa or submucosa; ypT2: tumor involves muscularis propria but not beyond; ypT3; and ypT4) and Dworak regression grade were considered categoric variables. Age, VpreCRT, VpostCRT, and tumor volume reduction rate were analyzed as continuous variables. Student t tests, 1-way analysis of variance (ANOVA), and the Fisher exact test were used to compare various parameters between chemotherapy groups (FL vs. capecitabine) and for determining the probability of lymph node positivity. Correlations between ypN classification and ypT-classification and between ypN classification and Dworak regression grade were assessed using a Spearman correlation coefficient test. For multivariate analysis to evaluate the correlations between lymph node positivity and various parameters, a stepwise procedure was performed using a logistic regression model. All statistical tests were 2-sided and were performed using SAS software (version 8.0.1; SAS Institute, Inc., Cary, NC). A P-value of.05 was considered a significant difference.
The study population had a mean age of 56 years (range, 27 yrs-82 yrs), and was mostly male (male:female ratio of 192:90). The mean distance from the anal verge to the caudal tumor edge was 4.8 cm (range, 1 cm-8 cm). The study population characteristics are summarized in Table 1. Analysis of the 2 different chemotherapy regimen groups showed that there were no significant differences between the FL and capecitabine groups in terms of MR tumor volume, tumor volume reduction rate, ypT classification, ypN classification, and Dworak regression grade.
Table 1. Characteristics of Patients Treated with FL or Capecitabine after Preoperative CRT
|Mean age in yrs (range)||57 (27–75)||56 (31–82)||.758*|
|Gender (no.)|| || ||.525†|
| Male||100||92|| |
| Female||43||47|| |
|Mean distance from anal verge in cm (range)||4.6 (0–9)||5.0 (0.5–9)||.119*|
|Pre-CRT MR tumor volume in cm3 (mean ± SD)||19.1 ± 17.3||18.5 ± 21.4||.780*|
|Post-CRT MR tumor volume in cm3(mean ± SD)||6.1 ± 8.2||5.1 ± 6.1||.275*|
|Tumor volume reduction rate (%, mean ± SD)‡||67.6 ± 20.1||67.8 ± 22.1||.950*|
|ypT classification (%)|| || ||.588†|
| ypT0||19 (13.3)||26 (18.7)|| |
| ypT1||7 (4.9)||6 (4.3)|| |
| ypT2||37 (25.9)||40 (28.8)|| |
| ypT3||77 (53.8)||63 (45.3)|| |
| ypT4||3 (2.1)||4 (2.9)|| |
|ypN classification (%)|| || ||.095†|
| ypN0||106 (74.1)||89 (64.0)|| |
| ypN1||22 (15.4)||37 (26.6)|| |
| ypN2||15 (10.5)||13 (9.4)|| |
|Dworak regression grade (%)|| || ||.396†|
| Grade 1||31 (21.7)||22 (15.8)|| |
| Grade 2||79 (55.2)||74 (53.2)|| |
| Grade 3||14 (9.8)||17 (12.2)|| |
| Grade 4||19 (13.3)||26 (18.7)|| |
One or more regional lymph nodes were retrieved from each patient and underwent pathology examination. The mean number of harvested lymph nodes for the total study population was 15.0 (range, 1 lymph node-42 lymph nodes), and were 15.5 and 14.4, respectively, for the FL and capecitabine subgroups (P = .269). The mean number of harvested lymph nodes according to ypT classification was 12.1 for ypT0, 14.0 for ypT1, 13.9 for ypT2, 16.3 for ypT3, and 20.1 for ypT4 (P = .006).
Lymph Node Status According to ypT-classification
Pathology analysis of surgical specimens showed that 87 (30.9%) patients had lymph nodes harboring tumors. Of the 45 patients with no residual primary tumor after preoperative CRT, only 1 patient had a tumor-positive lymph node (2.2%). The risk for lymph node involvement increased as the remaining mural tumor advanced; positive lymph nodes were detected in 1 of the 13 (7.7%) ypT1 patients, 13 of the 77 (16.9%) ypT2 patients, 69 of the 140 (49.3%) ypT3 patients, and in 3 of the 7 (42.9%) ypT4 patients. The difference in lymph node status between ypT subgroups was found to be statistically significant (P < .001). Overall, lymph node positivity was found to be correlated with ypT classification (correlation coefficent [r] = 0.407; P < .001). The FL and capecitabine subgroup analysis also demonstrated that lymph node positivity to be correlated with ypT classification in each subgroup (P < .001 for both) (Table 2).
Table 2. Lymph Node Positivity According to Primary Tumor Stage for FL and Capecitabine Patients after Preoperative CRT
|ypT0||19||0||0||0/19 (0)||< .001||25||1||0||1/26 (3.8)||< .001||1/45 (2.2)||< .001|
|YpT1||7||0||0||0/7 (0)|| ||5||1||0||1/6 (16.7)|| ||1/13 (7.7)|| |
|ypT2||31||4||2||6/37 (16.2)|| ||33||5||2||7/40 (17.5)|| ||13/77 (16.9)|| |
|ypT3||48||18||11||29/77 (37.7)|| ||23||29||11||40/63 (63.5)|| ||69/140 (49.3)|| |
|ypT4||1||0||2||2/3 (66.7)|| ||3||1||0||1/4 (25.0)|| ||3/7 (42.9)|| |
Lymph Node Status According to Dworak Regression Grade and MR Volumetry
We found that tumor involvement in regional lymph nodes after preoperative CRT decreased as the Dworak regression grade increased; tumor-harboring lymph nodes were found in 33 of 53 (62.3%) Grade 1 patients, 48 of 153 (31.4%) Grade 2 patients, 5 of 31 (16.1%) Grade 3 patients, and 1 of 45 (2.2%) Grade 4 patients. The difference in lymph node status was found to be correlated with Dworak regression grade (P < .001). Lymph node positivity also was found to be correlated with Dworak regression grade (r = 0.395; P < .001).
MR volumetry was performed in 255 patients (90.4%). Lymph node-negative and lymph node-positive groups showed no significant differences in terms of VpreCRT (18.7 ± 20.7 cm3 vs. 18.7 ± 16.4 cm3; P = .993), mean VpostCRT (5.5 ± 7.6 cm3 vs. 5.7 ± 5.7 cm3; P = .845), and tumor volume reduction rate (68.8 ± 21.9% vs. 64.6 ± 20.4%; P = .149).
Multivariate analysis (Table 3) showed that ypT classification and Dworak regression grade were associated with lymph node status (P = .004 and P = .002, respectively). In contrast, age, gender, VpreCRT, VpostCRT, tumor volume reduction rate, and chemotherapy regimen (FL or capecitabine) were not found to be correlated with lymph node status (P >.05 for each).
Table 3. Correlations between Lymph Node Status and Various Factors Using Univariate and Multivariate Analysis
|Mean age in yrs ± SD||56.5 ± 11.4||56.4 ± 10.6||.932||.691|
|Gender (no.) (M/F)||136/58||56/32||.242||.473|
|Mean distance from anal verge in cm ±SD||4.8 ± 2.1||5.0 ± 2.2||.496||.369|
|Mean pre-CRT MR tumor volume in cm3 ± SD||18.7 ± 20.7||18.7 ± 16.4||.993||.270|
|Mean post-CRT MR tumor volume in cm3 ± SD||5.5 ± 7.6||5.7 ± 5.7||.845||.623|
|Mean reduction rate of tumor volume (%) ± SD†||68.8 ± 21.9||64.6 ± 20.4||.149||.937|
|ypT classification (0/1/2/3/4)||44/12/64/71/4||1/1/13/69/3||< .001||.004|
|Dworak regression grade (1/2/3/4)||20/105/26/44||33/48/5/1||< .001||.002|
Lymph node metastasis is one of the most important prognostic factors in rectal cancer. The presence of metastatic lymph nodes in nearby mesorectum or along the iliac vessels greatly influences treatment, including the extent of surgical removal, the possibility of sphincter preservation, and the need for adjuvant therapy. The prognostic importance of the ypN classification after preoperative CRT is also well known.10, 19 For rectal cancer patients undergoing radical surgery without preoperative CRT, the most important predictor determining the risk of lymph node involvement is reported to be pT classification.12, 13 However, in rectal cancer patients treated with preoperative CRT, there is debate regarding whether ypT classification is still correlated with lymph node status.14–16
Read et al.14 reported the risk for lymph node metastasis as 3.4% (3 of 87 patients) in ypT0-1 patients, 23% (43 of 186) in ypT2 patients, and 51.5% (174 of 371) in ypT3-4 patients. They concluded that ypT classification could predict lymph node status after neoadjuvant therapy, leading to a more accurate selection of patients for transanal full-thickness resection. However, Zmora et al.20 reported a higher risk of lymph node metastasis (12.1%) in ypT0-1 patients, and that there was 11.8% chance of lymph node positivity in patients with totally regressed mural tumors, or ypT0N1 disease. Medich et al.16 insisted that ypT classification could not reliably predict lymph node status, even if preoperative CRT often downsized and downstaged locally advanced rectal cancer. The data from the current study indicated that the risk for lymph node metastasis was closely associated with the status of the remnant rectal tumor. The risk for lymph node metastasis in patients with ypT0-1 disease after preoperative CRT was found to be only 3.4% (2 of 58 patients), which is similar to the result of Read et al.14 We found that there was a sharp increase in lymph node metastasis between ypT2 and ypT3-4 tumors in addition to ypT0-1 and ypT2 (16.9% in patients with ypT2 disease and 49.0% in patients with ypT3-4 disease).
The ypT-ypN correlation after preoperative CRT has a potential for use in clinical practice. The mainstay of surgery for patients with locally advanced rectal cancer remains low anterior resection or abdominoperineal resection (APR). These surgeries continue to present significant risks for perioperative morbidity and mortality, at rates of approximately 30% and 3%, respectively.21 To avoid the risks, several studies have investigated transanal local excision as a primary treatment for some selected patients with pT1 or pT2 cancers.22, 23 Because preoperative CRT has been reported to provide clear downstaging of tumors, resulting in improved disease-free survival, local control, and sphincter preservation, local excision after preoperative CRT was performed for locally advanced rectal cancers including cT2 and cT3.7, 8 Previous studies regarding local excision after preoperative CRT had limitations such as small patient number,8 heterogeneity of indication such as refusal of APR, and severe comorbidity.7, 9 To our knowledge to date, there has been no prospective study comparing results between transanal excision and conventional open resection after preoperative CRT in selected patients with ypT0-1 disease. From an oncologic point of view, the problem with local excision after preoperative CRT is the nonremoval of the adjacent mesorectum containing regional lymph nodes in which metastatic tumor cells may reside. Although some authors suggest that patients with local recurrence after local excision still could undergo en bloc resection because transanal resection does not violate the Waldeyer fascia,8 it is not easy to discriminate surgical planes in secondary surgery for recurring rectal cancer. Therefore, before or immediately after local excision, it is important to be able to predict the likelihood of rectal tumor metastases to regional lymph nodes, and whether the tumor is confined to the rectal wall, because such predictions will affect the decision regarding whether to perform radical proctectomy.
Other than ypT classification, TRUS and MRI studies have been mainly used to define the extent of local tumor spread through the rectal wall and to identify regional lymph node metastases after preoperative CRT.24 However, the use of such imaging was found to be inaccurate for preoperative staging of tumors regressed by CRT, especially for evaluating lymph status.11, 25 The suggested reasons for such inaccuracy are radiation-induced inflammation, fibrotic change, and residual rectal ulcer after the completion of preoperative CRT.10, 11, 26 In the current study, we were unable to identify a correlation between lymph node status and MR volumetry findings, including mean volume of pre-CRT tumor, mean volume of post-CRT tumor, and tumor volume reduction rate. The size of the primary tumor in colorectal cancer patients was proven to be of no prognostic significance in a large randomized National Surgical Adjuvant Breast and Bowel Project (NSABP) study.27 In that study, the mean dimension of Dukes B2 tumors was greater than the mean dimension of lymph node-positive Dukes C2 tumors. In the current study, we also found that there was no correlation between tumor volume measured by MR and lymph node status, although we did not measure tumor volume directly in resected specimens.
Tumor regression grade was reported to be a predictor of lymph node metastasis. Vecchio et al.28 reported that 4 (8.9%) patients with complete regression or near-complete regression had ypN0 disease, and these demonstrated a low local failure rate (2%). We also examined any correlation between lymph node status and tumor regression grade, but found that only patients with complete regression (Dworak regression Grade 4) were associated with a low incidence of lymph node metastasis and that there was sharp demarcation (2.2% vs. 16.1%) between the complete regression group (Grade 4) and the near-complete regression group (Grade 3). Therefore, the pathologic regression grade did not demonstrate any additional benefit over the ypT classification.
A benefit of preoperative CRT may be eradication of all viable tumor cells. The judgment of CR has to be based on thorough pathologic examination of the entire lesion in the resected rectum. In the current study, the resected rectum was sliced into 4-mm thick serial sections and the entire lesion and the surrounding nonneoplastic area was examined microscopically. We found that 45 of 282 (16.0%) patients showed primary tumor CR. The likelihood of CR after preoperative CRT has been reported previously at rates of 5% to 27%.14–16 The large range of the CR rate appears to be due to differences in preoperative CRT regimens, including chemotherapeutic agents, radiation doses, and the interval between preoperative CRT and surgery.29, 30
The data from the current study indicate that pathologic T classification may be a reliable predictor of lymph node metastasis after preoperative CRT. In the current study, the risk of lymph node metastasis was found to be 3.4% in patients with rectal cancers that had regressed to ypT0 or ypT1. This result is relevant for clinical practice in terms of identifying those patients for whom local excision may be a curative treatment option; however, validation in a prospective trial is needed.