doi:10.1111/codi.12682
Original article
Lymph node status as a prognostic indicator after preoperative neoadjuvant chemoradiotherapy of rectal cancer N. Mirbagheri*, B. Kumar†, S. Deb‡, B. R. Poh*, J. G. Dark§, C. C. Leow* and W. M. K. Teoh* *Department of Academic Surgery, Dandenong Hospital, Melbourne, Victoria, Australia, †Department of Anatomical Pathology, Monash Medical Centre, Southern Health, Clayton, Victoria, Australia, ‡Department of Pathology, Peter MacCallum Cancer Centre, East Melbourne, Victoria,Australia and §Faculty of Business and Economics, University of Melbourne, Melbourne, Victoria, Australia Received 16 January 2014; accepted 29 March 2014; Accepted Article online 10 June 2014
Abstract Aim The primary aim of this study was to examine lymph node status after neoadjuvant chemoradiotherapy (CRT) using a novel scoring system describing the pathological lymph node regression grade. The proposed scoring system was based on the percentage of fibrosis and the presence of residual tumour amount. The secondary aim of the study was to assess the oncological impact of this scoring system. Method The project was a retrospective cohort study over a 10-year period. Two hundred and two patients with rectal cancer who had received CRT followed by curative surgery were included. A histopathologist prospectively scored each specimen and the impact of the scoring system on survival and recurrence was analysed. Results One hundred and ninety patients completed long-course preoperative CRT and formed the basis of the study. Overall, 40 recurrences (local and distant)
Introduction Total mesorectal excision and preoperative neoadjuvant chemoradiotherapy (CRT) have revolutionized the management of rectal cancer. Preoperative CRT has enabled the assessment of the biology of rectal tumours. In some cases where the tumour has completely responded to CRT, it has enabled less aggressive surgical treatment [1]. Lymph node status is an essential part of the TNM staging system, but attempts at preoperative staging have been unreliable. At best, the accuracy of N staging by endorectal ultrasound or pelvic MRI ranges from 60 to 80% [2,3]. There are therefore significant limitations to the evaluation of the response of lymph nodes to preoperative CRT using conventional imaging. Furthermore, histopath-
were observed over a median follow-up of 36 months. The lymph node regression score was a significant predictor of tumour recurrence (hazard ratio 1.273, 95% CI 1.048–1.548; P = 0.015). The overall mortality rate was 21%, and a lower lymph node regression score was correlated with an improved survival curve (P = 0.01). Conclusion The results demonstrate that lymph node response to neoadjuvant CRT based on a nodal regression scoring system is related to recurrence. Keywords Rectal neoplasms, chemoradiotherapy, lymph nodes, neoadjuvant therapy What does this paper add to the literature? This is the first study to examine the clinical impact of lymph node regression after neoadjuvant chemoradiotherapy for rectal cancer. This assessment may allow clinicians to tailor adjuvant treatment.
ological examination of a resected irradiated rectal cancer specimen provides limited evidence concerning the pretreatment cancer status. This makes longitudinal pre- and post-treatment assessment of lymph node status difficult. In the present study lymph nodes obtained from curative surgical specimens of irradiated rectal cancer were studied to determine the presence of any residual cancer cells and fibrosis in order to provide a nodal regression score. This was then related to the clinical cancer-specific outcome to assess whether the grading system could be of prognostic significance.
Method Patient recruitment and eligibility
Correspondence to: Dr Naseem Mirbagheri, David Street, Dandenong Hospital, Dandenong, Victoria 3175, Australia. E-mail: [email protected]
Patients with biopsy-proven rectal cancer with no evidence of metastasis who underwent long-course neoad-
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
O339
N. Mirbagheri et al.
Lymph node status after CRT for rectal cancer
juvant CRT treatment before a curative resection were eligible for the study. All patients with a preoperative or intra-operative diagnosis of metastasis were excluded. Patients who died within 4 weeks of surgery were also excluded. Eligible patients treated over a 10-year period (2001–11) at a tertiary centre (Southern Health, Victoria, Australia) were studied retrospectively. Patients were identified based on the operation and diagnostic coding and medical records were studied to obtain clinical information. The study was approved by the Ethics Board of the Research Directorate at Southern Health (application no. 12082Q). Follow-up data were obtained through hospital medical records, the primary surgeon’s notes and by direct contact with the treating oncologist or the local medical practitioner. All patients had been seen for
Table 1 The LRG score. Score
Description
0 1 2 3 4 5
Normal lymph node 100% fibrosis, no residual cancer 75–100% fibrosis, 0–25% cancer 50–75% fibrosis, 25–50% cancer 25–50% fibrosis, 50–75% cancer 0–25% fibrosis, 75–100% cancer
follow up every 3 months during the first 2 years after treatment and thereafter at 6-monthly intervals for the next 3 years. The modality and frequency of surveillance imaging and endoscopic examination was
Normal Lymph Node (LRG 0)
100% Fibrosis (LRG 1)
50% Cancer Cells (LRG 4)
Complete replacement with mucinous adenocarcinoma cells (LRG 5)
Figure 1 Lymph node histopathological slides representing lymph node regression (LRG) scores.
O340
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
N. Mirbagheri et al.
at the discretion of the treating physician or surgeon. Most patients had a colonoscopy at 12 months after surgery and then at every 2 or 3 years. CT scans and serum carcinoembryonic antigen testing was performed at 6–12-monthly intervals. Local recurrence was defined by the presence of a suspicious mass on postoperative pelvic imaging (CT, MRI and/or positron emission tomography [PET] scan) which was considered to be a recurrence by radiologist and treating clinician. Distant recurrence was defined as the presence of a new lesion in a distant organ detected by CT or PET. The time to recurrence, or the disease-free survival (DFS), was measured in months from the date of surgery to the date of recurrence detected radiologically. Overall survival (OS) was defined as the time in months from the date of surgery to the date of death. Pathological data included the application of a lymph node regression score to all specimens, as discussed below. Other data included circumferential resection margin (CRM) status (positive if tumour within 1 mm), lymphovascular invasion (LVI), perineural invasion (PNI) and primary tumour regression grade (TRG) was applied using Mandard’s tumour regression scoring system [4]. For both the LRG and TRG scoring the pathologist was blinded to the patients’ clinical outcome. Lymph node regression scoring
A senior pathologist (BK) unaware of the clinical outcome reviewed all slides for all patients. To assess for interobserver variability 35 random specimens were also scored by a second pathologist (SD). Routine haematoxylin and eosin (H&E) sections of lymph nodes were assessed for the presence of tumour cells or evidence of regression determined by the presence of fibrosis. The lymph node regression (LRG) score was based on the percentage of residual tumour cells and the degree of fibrosis in each lymph node found in the specimen (Table 1, Fig. 1). Given the variable response in the lymph nodes in any single resection specimen, the final LRG score assigned was assessed in two different ways: 1 LRG maximum (LRGmax): as each specimen has a variable number of lymph nodes and each lymph node may have a different regression score based on its response to treatment, an overall score for each patient (specimen) was given based on the worst score for that specimen. For example, the LRGmax would be 3 for a specimen containing two lymph nodes with scores of 2 and 3.
Lymph node status after CRT for rectal cancer
2 LRG sum (LRGsum): this reflects the aggregate tumour burden in all the lymph nodes in the specimen. For example, the LRGsum would be 5 for a specimen containing two lymph nodes with scores of 2 and 3. Statistical analysis
Statistical analysis was performed using IBM SPSS STATIS21.0 (IBM Corporation, New York, NY, USA). A P-value of < 0.05 was considered clinically significant. Kaplan–Meier survival curves were plotted using recurrence and death as the end-points and were compared using the log-rank test. For the purpose of survival curve analysis for LRGsum, patients were grouped into LRGsum 0, LRGsum 1–9 and LRGsum 10 or more. TICS
Table 2 Patient characteristics (n = 190), n(%). Age (years) Mean SD Gender Male Female Tumour location Upper Middle Lower Unknown Preoperative staging (AJCC 7th edn) Stage I Stage IIA Stage IIB/C Stage IIIA Stage IIIB Stage IIIC Stage IV Unknown Time to surgery (weeks)* Median Range Surgery type AR APR Adjuvant treatment Received Not received Follow-up (months) Median Range
63.4 11.1 129 (67.9) 61 (32.1) 29 94 58 9 0 62 6 8 85 6 0 23
(15.2) (49.5) (30.5) (4.7)
(32.6) (3.1) (4.2) (44.7) (3.1) (12.1)
6 2–24 113 (59.5) 77 (40.5) 126 (66.3) 64 (33.7) 36 3–148
AJCC, American Joint Committee on Cancer; AR, anterior resection; APR, abdominoperineal resection; SD, Standard Deviation. *Times to surgery for 30 patients were unknown.
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
O341
N. Mirbagheri et al.
Lymph node status after CRT for rectal cancer
Multivariate analysis estimated Cox proportional hazard models for mortality and recurrence. Ordered logit and ordinary least squares models were used to determine the predictors for LRGmax and LRGsum, respectively.
Results Patient characteristics
Two hundred and twenty-seven patients were identified as eligible for the study. Of these, 12 had received short-course CRT and were thus excluded, seven were lost to follow up, 15 patients had metastatic disease identified intra-operatively and three died within 4 weeks of operation from surgical complications. Thus baseline data were obtained from 190 patients. The mean age was 63.4 years (range: 32–87) and 129 of the patients were men. Other patient characteristics are shown in Table 2. Long-course CRT was administered in the form of 50.4 Gy over 28 fractions with 5-fluorouracil over a period of 5.5 weeks. The median time to surgery was 6 weeks and the most common operation performed was anterior resection with a covering loop ileostomy. Postoperative chemotherapy was offered to 66.3% of patients; however, details of treatment protocols and completion rates were not available. The median followup period was 36 months. Pathological characteristics
The primary tumour of all but two patients showed response to treatment. In 34 (17.9%) patients there was evidence of complete primary tumour response (TRG 1), but the most common score was TRG 3 in 85 (42.1%) specimens. The average number of lymph nodes retrieved per specimen was 11. Amongst the positive lymph nodes the most common score was LRG 2 (33.7%). The number of patients with pathological evidence of lymph node metastasis (ypN+ve) was 85 (44.7%), thus the majority (55.3%) had no evidence of tumour in the lymph nodes. The presence of capsular fibrosis and mucin lakes was noted in some lymph nodes. Mucin lakes were considered in a similar fashion to the Mandard scoring system and were considered as a previous tumour but with no residual tumour cells. The presence of capsular fibrosis was considered as exposure of that lymph node to radiation, therefore the presence of capsular fibrosis in the absence of tumour cells was considered LRG 0. The scoring system was found to be simple and repeatable with a Pearson correlation
O342
ratio of 0.9 for the 35 specimens scored by the two pathologists. The median LRGmax and LRGsum was 3, and LRGSum ranged from 0 to 78. In the 34 patients with complete primary tumour response (ypT0, 17.9%), five patients also had evidence of complete lymph node response (ypT0ypN0), 26 patients had no evidence of tumour in the lymph nodes (ypT0N0) and three patients had residual tumour in the lymph nodes despite complete primary tumour response (ypT0ypN+). In 13 of 85 patients with positive lymph nodes (ypN+), there was evidence of complete response in the lymph nodes but not in the primary tumour (ypT+ypN0, 15.2%). Of
Table 3 Pathological characteristics of 190 specimens, n(%). TRG score TRG 1 TRG 2 TRG 3 TRG 4 TRG 5 LRG scoring system, median (range) LRGmax LRGsum LRG score* LRG 1 LRG 2 LRG 3 LRG 4 LRG 5 LVI Present Absent CRM Involved Not involved No. of lymph nodes retrieved per specimen Median Range Pathological lymph node status per specimen ypN+ve ypN–ve LNR† Average PNI Present Absent
34 52 80 22 2
(17.9) (27.4) (42.1) (11.6) (1.0)
3 (0–5) 3 (0–78) 41 100 45 72 41
(13.7) (33.7) (15.0) (24.0) (13.6)
27 (14.2) 163 (85.8) 11 (5.8) 179 (94.2) 11.5 2–39 85 (44.7) 105 (55.3) 0.1 4 (2.1) 186 (97.9)
TRG, tumour regression grade; LRG, lymph node regression; LVI, lymphovascular invasion; CRM, circumferential resection margin; PNI, perineural invasion. *The LRG scores are for all lymph nodes scored in every specimen. †Lymph node ratio = number of metastatic lymph nodes/total number of lymph nodes retrieved.
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
N. Mirbagheri et al.
Lymph node status after CRT for rectal cancer
1.0 LRGmax 1 0.8
Cum survival
LRGmax 2 LRGmax 0 0.6 LRGmax 3 LRGmax 5
0.4
LRGmax 4 0.2 P = 0.01 0.0 0
20
LRGmax score 0 1 2 3 4 5
40
60 80 100 Overall survival
120
Mean overall survival (months) 106.9 135.8 67.1 80.4 50.4 60.8
140
95% Cl 95.3-118.5 113.1-158.6 36.6-97.6 54.4-106.5 42.7-58.1 37.8-83.9
1.0 LRGmax 1 LRGmax 0 LRGmax 2
Cum survival
0.8
0.6
LRGmax 3 LRGmax 5
0.4 LRGmax 4 0.2
P = 0.009
0.0 0
Figure 2 Survival curves as a function of LRGmax.
20
40
60 80 100 Disease free survival
120
140
LRGmax score Mean disease-free survival (months) 95% Cl 0 107.5 95.7-119.3 1 136.3 114.4-158.2 2 68.1 36.2-99.9 3 45.5-105.5 75.5 4 48.6 39.7-57.5 5 56.6 29.9-83.1
note, in Table 3 specimens with maximum LRG score of 0.9 (i.e. previous tumour but complete response in all involved lymph nodes) were listed as yN+ve to
differentiate them from those specimens without any evidence of metastasis in the lymph nodes, i.e. LRG0 (i.e. yN–ve).
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
O343
N. Mirbagheri et al.
Lymph node status after CRT for rectal cancer
Clinical outcomes, prognostic factors and survival curves
Discussion To our knowledge, this is the first study to show the clinical significance of pathological lymph node response to preoperative treatment as a regression grade. It demonstrates that oncological outcome is better in patients with lymph node metastasis where a pathological response to preoperative treatment can be observed. Factors such as TRG score, location of the tumour and the presence of lymphovascular invasion were also shown to correlate with the regression grade. Lymph node regression scoring in the mesorectal nodes was first reported by Caricato et al. [5] in 2007, who showed a complete lymph node response (LRG 1)
1.0
1.0
0.8
0.8 LRGsum 0
0.6 LRGsum 1–9 LRGsum ≥10
0.4 0.2
LRGsum 0 Cum survival
Cum survival
The 5-year OS and DFS for the cohort was 72.6% and 75.8%, respectively, with an overall recurrence and mortality rate of 21%. Of the 40 recurrences, eight were local, 24 distant and eight both local and distant. The average OS and DFS were 41.9 and 38.1 months, respectively. The OS and DFS survival curve analysis (Figs 2 and 3) categorized the patients into groups based on their LRG score type. LRGsum was the strongest predictor of OS and DFS. The survival curve for LRGmax illustrates that patients with complete lymph node regression (LRG 1) have an improved survival curve in both OS and DFS. LRGsum has an advantage over LRGmax in that it largely relies on more nodes being sampled, which is reflected in the more significant P-values. Using multivariate Cox proportional hazard regression analysis, LRG scoring was not found to be a significant factor for mortality (Table 4). This is probably because the LRG score is highly correlated to other significant variables such as LVI and TRG. Thus, significant regressors for mortality using the LRGsum analysis included LVI and TRG (Table 4). Importantly, the LRG score was a significant predictor for recurrence; other significant regressors for recurrence were LVI and tumour location. The results using LRGmax analysis were similar to that of LRGsum. It should be noted that CRM was positively correlated with LVI. Whilst CRM was a statistically significant factor predicting recurrence, when both variables (LVI and CRM) were included
CRM was not statistically significant. Further, multivariate regression analysis demonstrated that higher TRG scores (P = 0.005), tumours located in the middle rectum (P = 0.013) and LVI (P = 0.029) were associated with higher LRG scores. Patients with LRG = 0 (i.e. no tumour in the lymph nodes) were included in the analysis as they acted as a control group, and we hypothesized that patients with complete tumour response in the lymph nodes (LRG 1) may fare even better than those with LRG 0. But given the small sample size for LRG 1 this hypothesis was not adequately tested in this study. When a subgroup analysis is performed on patients with lymph node metastasis (yN+ve, n = 85), only LRGsum remained a significant predictor of recurrence [hazard ratio (HR) 1.117, 95% CI 1.023–1.220; P = 0.014].
0.6
LRGsum 1–9 LRGsum ≥10
0.4 0.2
P = 0.000
0.0 0
20
40
100 60 80 Overall survival
120
140
LRGsum score Mean overall survival (months) 95% Cl
0 1–9 ≥10
117.7 85.9 51
103.8-131.7 66.2-105.7 35.9-66.2
P = 0.000
0.0 0
20
40
60 80 100 Overall survival
120
LRGsum score Mean overall survival (months)
0 1–9 ≥10
119.3 87.3 46.2
140
95% Cl
105.3-133.2 66.4-108.2 27.9-64.4
Figure 3 Survival curves as a function of LRGsum.
O344
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
N. Mirbagheri et al.
Lymph node status after CRT for rectal cancer
Table 4 Multivariate Cox proportional hazards regression analysis for oncological outcomes as a function of lymph node regression (LRG) scores (n = 190).
Mortality (n = 39) LRGsum LRGmax* Female sex Age Time to surgery Adjuvant therapy Tumour stage Total LN retrieved CRM LVI PNI TRG LNR Middle rectum Lower rectum APR Recurrence (n = 40) LRGsum LRGmax* Female sex Age Time to surgery Adjuvant therapy Tumour stage Total LN retrieved CRM LVI PNI TRG LNR Middle rectum† Lower rectum† APR‡
HR
95% CI
P-value
1.587 1.074 0.380 0.991 0.998 0.514 1.001 0.993 1.492 2.093 4.847 2.669 1.182 0.713 0.477 1.515
0.665–3.789 0.863–1.337 0.113–1.279 0.950–1.032 0.877–1.136 0.200–1.319 0.999–1.003 0.911–1.082 0.355–6.262 1.098–7.676 0.499–47.068 1.431–4.979 0.175–7.988 0.192–2.649 0.109–2.093 0.603–3.808
0.575 0.521 0.118 0.656 0.971 0.166 0.085 0.871 0.585 0.032 0.174 0.002 0.864 0.613 0.326 0.377
1.061 1.273 0.649 1.009 0.940 0.594 1.001 0.412 1.573 3.312 4.810 1.186 0.576 0.253 0.439 2.125
1.028–1.095 1.048–1.548 0.299–1.836 0.970–1.049 0.815–1.084 0.231–1.526 0.999–1.003 0.887–1.049 0.467–5.303 1.216–9.012 0.559–41.392 0.722–1.946 0.088–3.765 0.082–0.786 0.132–1.457 0.841–5.371
0.000 0.015 0.649 0.659 0.397 0.279 0.186 0.412 0.465 0.019 0.153 0.500 0.565 0.017 0.179 0.111
Bold represents significant values. HR, hazard ratio; TRG, tumour regression grade; LRG, lymph node regression; LVI, lymphovascular invasion; LN, lymph nodes; LNR, lymph node ratio; CRM, circumferential resection margin; PNI, perineural invasion; APR, abdominoperineal resection. *The results for LRGmax are from a separate Cox regression restricted model analysis. †Compared with the upper rectum. ‡Compared with anterior resection.
in 51% of specimens compared with 13.7% in this study. Their study, however, was limited by a small sample size (n = 35) and no clinical follow up. The importance of regression score in lymph nodes as well the primary tumour is based on the underlying pathophysiological
mechanism of lymph node metastasis. For tumour cells to metastasize effectively, rather than spill over, they need to acquire the ability to express cytokines such as vascular endothelial growth factor and adopt the normal mechanism of lymph node homing during metastasis (i.e. chemokine receptor expression) [6]. These ‘upgraded’ cancer cells may be more robust than the primary cancer cells, more resilient to CRT and thus have a differential response to treatment. In this study, we have illustrated that within the same specimen lymph nodes can have varied responses to CRT. Pathological complete response (pCR) had been traditionally described as complete eradication of all viable tumour cells from the rectal wall and mesorectum after preoperative CRT [7]. Complete clinical responders (cCR) are those patients in whom the tumour has been eradicated clinically and radiologically. This is supported by complete tumour regression in the primary tumour upon local excision irrespective of lymph nodes status. The rates of pCR in the literature vary from 10 to 20% compared with our observed rate of 17.9% [2,7,8]. The risk of lymph node metastasis in cCR varies from 0 to 14%, consistent with our observed rate of 8.8% [9,10]. Of note, in our study out of those patients with cCR in the primary tumour and lymph node metastasis in the specimen, 63% (5/8 specimens) also showed a pCR in the lymph nodes (LRG1). We observed that the LRG1 cohorts do better than those with LRG > 1 and even fare better than LRG0 in the later stages of follow up as demonstrated by the flattening out of the survival curve. Furthermore, Kaplan–Meier graphs also show better survival curves and disease-free status for patients with lower LRG scores. A multicentre retrospective study by Yeo et al. [8] on 333 locally advanced rectal cancers post-CRT, showed that ypN status was the most relevant independent prognostic factor for both OS and DFS. Therefore, it is not an unexpected finding that LRG scores have a similar oncological significance. Of interest, three clinicopathological factors (tumour location, TRG score and lymphovascular status) correlated significantly with LRG scores. In other studies it has been shown that in irradiated rectal cancer, ypT status (i.e. TRG score) correlates with yN status, thus supporting our result [11]. To our knowledge this is the only study that has shown tumour location to be a significant predictor of lymph node response posttreatment, but the underlying mechanism is uncertain. This study is inherently limited by its retrospective nature (and thus the biases involved in the clinical information obtained), the heterogeneity of surgical techniques employed and the pathological preparation of lymph nodes. The LRG score was made on lymph
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
O345
N. Mirbagheri et al.
Lymph node status after CRT for rectal cancer
nodes that were prepared at the same time as the specimen was resected and the specimens were no longer available for further lymph node sampling. Moreover lymph nodes completely destroyed by radiation may have been missed. Notably, we observed that higher sampling of lymph nodes (especially when 11 or more nodes were sampled) markedly improved the prognostic power of LRG analysis (data not shown). While no current guidelines are provided to define adequate nodal sampling in these posttreatment specimens, the suggestion is that adequate lymph node sampling may be a barrier to optimizing nodal assessment and should be addressed in future prospective studies. In conclusion, lymph node response in the form of pathological regression grade appears to be an important prognostic indicator for oncological outcomes. Therefore more tailored adjuvant treatment can be directed to patients with poorer lymph node response. We hope that this pilot study creates a platform for further supporting studies in which a lymph node grading system can be implemented and patients followed up prospectively. As we develop further insights into the pathological response of irradiated lymph nodes, we may enter a new era of rectal cancer treatment where surrogate markers for lymph node response may enable a more conservative operative approach.
Acknowledgement We would like to thank Zdenka Prodanovic, Tissue Bank Manager at Southern Health, for her invaluable assistance in facilitating this project.
Author’s contributions Naseem Mirbagheri: Study concept, design, acquisition of data, data analysis and interpretation, manuscript writing and final approval. Beena Kumar: Study design, acquisition of data, manuscript writing and final approval. Siddhartha Deb: Study design, acquisition of data, data analysis and interpretation, manuscript writing and final approval. Ben Poh: Study design, acquisition of data, data analysis, manuscript writing and final approval. Jonathan Dark: Study design, data analysis and interpretation, manuscript writing and final approval. Christopher Leow: Acquisition of data, manuscript writing and final approval. William Teoh: Study
O346
design, interpretation of data, manuscript writing and final approval.
References 1 Habr-Gama A. Non-operative management of rectal cancer after neoadjuvant chemoradiation. Br J Surg 2009; 96: 125–7. 2 Kosinski L, Habr-Gama A, Ludwig K, Perez R. Shifting concepts in rectal cancer management: a review of contemporary primary rectal cancer treatment strategies. CA Cancer J Clin 2012; 62: 173–202. 3 Kahn H, Alexander A, Rakinic J, Nagle D, Fry R. Preoperative staging of irradiated rectal cancers using digital rectal examination, computed tomography, endorectal ultrasound, and magnetic resonance imaging does not accurately predict T0, N0 pathology. Dis Colon Rectum 1997; 40: 140–4. 4 Mandard AM, Dalibard F, Mandard JC, Marnay J. Pathologic assessment of tumor regression after preoperative chemoradiotherapy of esophageal carcinoma. Clinicopathologic correlations. Cancer 1994; 73: 2680–6. 5 Caricato M, De Dominicis E, Vincenzi B, Rabitti C. Tumor regression in mesorectal lymphnodes after neoadjuvant chemoradiation for rectal cancer. Eur J Surg Oncol 2007; 33: 724–8. 6 Nathanson SD. Insights into the mechanisms of lymph node metastasis. Cancer 2003; 98: 413–23. 7 Jang TY, Yu CS, Yoon YS et al. Oncologic outcome after preoperative chemoradiotherapy in patients with pathologic T0 (ypT0) rectal cancer. Dis Colon Rectum 2012; 55: 1024–31. 8 Yeo S-G, Kim DY, Kim TH et al. Pathologic complete response of primary tumor following preoperative chemoradiotherapy for locally advanced rectal cancer. Ann Surg 2010; 252: 998–1004. 9 Lindebjerg J, Spindler KLG, Ploen J, Jakobsen A. The prognostic value of lymph node metastases and tumour regression grade in rectal cancer patients treated with long-course preoperative chemoradiotherapy. Colorectal Dis 2009; 11: 264–9. 10 Habr-Gama A, Perez RO, Proscurshim I et al. Absence of lymph nodes in the resected specimen after radical surgery for distal rectal cancer and neoadjuvant chemoradiation therapy: what does it mean? Dis Colon Rectum 2008; 51: 277–83. 11 Read TE, Andujar JE, Caushaj PF et al. Neoadjuvant therapy for rectal cancer: histologic response of the primary tumor predicts nodal status. Dis Colon Rectum 2004; 47: 825–31.
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
Original article
Lymph node status as a prognostic indicator after preoperative neoadjuvant chemoradiotherapy of rectal cancer N. Mirbagheri*, B. Kumar†, S. Deb‡, B. R. Poh*, J. G. Dark§, C. C. Leow* and W. M. K. Teoh* *Department of Academic Surgery, Dandenong Hospital, Melbourne, Victoria, Australia, †Department of Anatomical Pathology, Monash Medical Centre, Southern Health, Clayton, Victoria, Australia, ‡Department of Pathology, Peter MacCallum Cancer Centre, East Melbourne, Victoria,Australia and §Faculty of Business and Economics, University of Melbourne, Melbourne, Victoria, Australia Received 16 January 2014; accepted 29 March 2014; Accepted Article online 10 June 2014
Abstract Aim The primary aim of this study was to examine lymph node status after neoadjuvant chemoradiotherapy (CRT) using a novel scoring system describing the pathological lymph node regression grade. The proposed scoring system was based on the percentage of fibrosis and the presence of residual tumour amount. The secondary aim of the study was to assess the oncological impact of this scoring system. Method The project was a retrospective cohort study over a 10-year period. Two hundred and two patients with rectal cancer who had received CRT followed by curative surgery were included. A histopathologist prospectively scored each specimen and the impact of the scoring system on survival and recurrence was analysed. Results One hundred and ninety patients completed long-course preoperative CRT and formed the basis of the study. Overall, 40 recurrences (local and distant)
Introduction Total mesorectal excision and preoperative neoadjuvant chemoradiotherapy (CRT) have revolutionized the management of rectal cancer. Preoperative CRT has enabled the assessment of the biology of rectal tumours. In some cases where the tumour has completely responded to CRT, it has enabled less aggressive surgical treatment [1]. Lymph node status is an essential part of the TNM staging system, but attempts at preoperative staging have been unreliable. At best, the accuracy of N staging by endorectal ultrasound or pelvic MRI ranges from 60 to 80% [2,3]. There are therefore significant limitations to the evaluation of the response of lymph nodes to preoperative CRT using conventional imaging. Furthermore, histopath-
were observed over a median follow-up of 36 months. The lymph node regression score was a significant predictor of tumour recurrence (hazard ratio 1.273, 95% CI 1.048–1.548; P = 0.015). The overall mortality rate was 21%, and a lower lymph node regression score was correlated with an improved survival curve (P = 0.01). Conclusion The results demonstrate that lymph node response to neoadjuvant CRT based on a nodal regression scoring system is related to recurrence. Keywords Rectal neoplasms, chemoradiotherapy, lymph nodes, neoadjuvant therapy What does this paper add to the literature? This is the first study to examine the clinical impact of lymph node regression after neoadjuvant chemoradiotherapy for rectal cancer. This assessment may allow clinicians to tailor adjuvant treatment.
ological examination of a resected irradiated rectal cancer specimen provides limited evidence concerning the pretreatment cancer status. This makes longitudinal pre- and post-treatment assessment of lymph node status difficult. In the present study lymph nodes obtained from curative surgical specimens of irradiated rectal cancer were studied to determine the presence of any residual cancer cells and fibrosis in order to provide a nodal regression score. This was then related to the clinical cancer-specific outcome to assess whether the grading system could be of prognostic significance.
Method Patient recruitment and eligibility
Correspondence to: Dr Naseem Mirbagheri, David Street, Dandenong Hospital, Dandenong, Victoria 3175, Australia. E-mail: [email protected]
Patients with biopsy-proven rectal cancer with no evidence of metastasis who underwent long-course neoad-
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
O339
N. Mirbagheri et al.
Lymph node status after CRT for rectal cancer
juvant CRT treatment before a curative resection were eligible for the study. All patients with a preoperative or intra-operative diagnosis of metastasis were excluded. Patients who died within 4 weeks of surgery were also excluded. Eligible patients treated over a 10-year period (2001–11) at a tertiary centre (Southern Health, Victoria, Australia) were studied retrospectively. Patients were identified based on the operation and diagnostic coding and medical records were studied to obtain clinical information. The study was approved by the Ethics Board of the Research Directorate at Southern Health (application no. 12082Q). Follow-up data were obtained through hospital medical records, the primary surgeon’s notes and by direct contact with the treating oncologist or the local medical practitioner. All patients had been seen for
Table 1 The LRG score. Score
Description
0 1 2 3 4 5
Normal lymph node 100% fibrosis, no residual cancer 75–100% fibrosis, 0–25% cancer 50–75% fibrosis, 25–50% cancer 25–50% fibrosis, 50–75% cancer 0–25% fibrosis, 75–100% cancer
follow up every 3 months during the first 2 years after treatment and thereafter at 6-monthly intervals for the next 3 years. The modality and frequency of surveillance imaging and endoscopic examination was
Normal Lymph Node (LRG 0)
100% Fibrosis (LRG 1)
50% Cancer Cells (LRG 4)
Complete replacement with mucinous adenocarcinoma cells (LRG 5)
Figure 1 Lymph node histopathological slides representing lymph node regression (LRG) scores.
O340
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
N. Mirbagheri et al.
at the discretion of the treating physician or surgeon. Most patients had a colonoscopy at 12 months after surgery and then at every 2 or 3 years. CT scans and serum carcinoembryonic antigen testing was performed at 6–12-monthly intervals. Local recurrence was defined by the presence of a suspicious mass on postoperative pelvic imaging (CT, MRI and/or positron emission tomography [PET] scan) which was considered to be a recurrence by radiologist and treating clinician. Distant recurrence was defined as the presence of a new lesion in a distant organ detected by CT or PET. The time to recurrence, or the disease-free survival (DFS), was measured in months from the date of surgery to the date of recurrence detected radiologically. Overall survival (OS) was defined as the time in months from the date of surgery to the date of death. Pathological data included the application of a lymph node regression score to all specimens, as discussed below. Other data included circumferential resection margin (CRM) status (positive if tumour within 1 mm), lymphovascular invasion (LVI), perineural invasion (PNI) and primary tumour regression grade (TRG) was applied using Mandard’s tumour regression scoring system [4]. For both the LRG and TRG scoring the pathologist was blinded to the patients’ clinical outcome. Lymph node regression scoring
A senior pathologist (BK) unaware of the clinical outcome reviewed all slides for all patients. To assess for interobserver variability 35 random specimens were also scored by a second pathologist (SD). Routine haematoxylin and eosin (H&E) sections of lymph nodes were assessed for the presence of tumour cells or evidence of regression determined by the presence of fibrosis. The lymph node regression (LRG) score was based on the percentage of residual tumour cells and the degree of fibrosis in each lymph node found in the specimen (Table 1, Fig. 1). Given the variable response in the lymph nodes in any single resection specimen, the final LRG score assigned was assessed in two different ways: 1 LRG maximum (LRGmax): as each specimen has a variable number of lymph nodes and each lymph node may have a different regression score based on its response to treatment, an overall score for each patient (specimen) was given based on the worst score for that specimen. For example, the LRGmax would be 3 for a specimen containing two lymph nodes with scores of 2 and 3.
Lymph node status after CRT for rectal cancer
2 LRG sum (LRGsum): this reflects the aggregate tumour burden in all the lymph nodes in the specimen. For example, the LRGsum would be 5 for a specimen containing two lymph nodes with scores of 2 and 3. Statistical analysis
Statistical analysis was performed using IBM SPSS STATIS21.0 (IBM Corporation, New York, NY, USA). A P-value of < 0.05 was considered clinically significant. Kaplan–Meier survival curves were plotted using recurrence and death as the end-points and were compared using the log-rank test. For the purpose of survival curve analysis for LRGsum, patients were grouped into LRGsum 0, LRGsum 1–9 and LRGsum 10 or more. TICS
Table 2 Patient characteristics (n = 190), n(%). Age (years) Mean SD Gender Male Female Tumour location Upper Middle Lower Unknown Preoperative staging (AJCC 7th edn) Stage I Stage IIA Stage IIB/C Stage IIIA Stage IIIB Stage IIIC Stage IV Unknown Time to surgery (weeks)* Median Range Surgery type AR APR Adjuvant treatment Received Not received Follow-up (months) Median Range
63.4 11.1 129 (67.9) 61 (32.1) 29 94 58 9 0 62 6 8 85 6 0 23
(15.2) (49.5) (30.5) (4.7)
(32.6) (3.1) (4.2) (44.7) (3.1) (12.1)
6 2–24 113 (59.5) 77 (40.5) 126 (66.3) 64 (33.7) 36 3–148
AJCC, American Joint Committee on Cancer; AR, anterior resection; APR, abdominoperineal resection; SD, Standard Deviation. *Times to surgery for 30 patients were unknown.
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
O341
N. Mirbagheri et al.
Lymph node status after CRT for rectal cancer
Multivariate analysis estimated Cox proportional hazard models for mortality and recurrence. Ordered logit and ordinary least squares models were used to determine the predictors for LRGmax and LRGsum, respectively.
Results Patient characteristics
Two hundred and twenty-seven patients were identified as eligible for the study. Of these, 12 had received short-course CRT and were thus excluded, seven were lost to follow up, 15 patients had metastatic disease identified intra-operatively and three died within 4 weeks of operation from surgical complications. Thus baseline data were obtained from 190 patients. The mean age was 63.4 years (range: 32–87) and 129 of the patients were men. Other patient characteristics are shown in Table 2. Long-course CRT was administered in the form of 50.4 Gy over 28 fractions with 5-fluorouracil over a period of 5.5 weeks. The median time to surgery was 6 weeks and the most common operation performed was anterior resection with a covering loop ileostomy. Postoperative chemotherapy was offered to 66.3% of patients; however, details of treatment protocols and completion rates were not available. The median followup period was 36 months. Pathological characteristics
The primary tumour of all but two patients showed response to treatment. In 34 (17.9%) patients there was evidence of complete primary tumour response (TRG 1), but the most common score was TRG 3 in 85 (42.1%) specimens. The average number of lymph nodes retrieved per specimen was 11. Amongst the positive lymph nodes the most common score was LRG 2 (33.7%). The number of patients with pathological evidence of lymph node metastasis (ypN+ve) was 85 (44.7%), thus the majority (55.3%) had no evidence of tumour in the lymph nodes. The presence of capsular fibrosis and mucin lakes was noted in some lymph nodes. Mucin lakes were considered in a similar fashion to the Mandard scoring system and were considered as a previous tumour but with no residual tumour cells. The presence of capsular fibrosis was considered as exposure of that lymph node to radiation, therefore the presence of capsular fibrosis in the absence of tumour cells was considered LRG 0. The scoring system was found to be simple and repeatable with a Pearson correlation
O342
ratio of 0.9 for the 35 specimens scored by the two pathologists. The median LRGmax and LRGsum was 3, and LRGSum ranged from 0 to 78. In the 34 patients with complete primary tumour response (ypT0, 17.9%), five patients also had evidence of complete lymph node response (ypT0ypN0), 26 patients had no evidence of tumour in the lymph nodes (ypT0N0) and three patients had residual tumour in the lymph nodes despite complete primary tumour response (ypT0ypN+). In 13 of 85 patients with positive lymph nodes (ypN+), there was evidence of complete response in the lymph nodes but not in the primary tumour (ypT+ypN0, 15.2%). Of
Table 3 Pathological characteristics of 190 specimens, n(%). TRG score TRG 1 TRG 2 TRG 3 TRG 4 TRG 5 LRG scoring system, median (range) LRGmax LRGsum LRG score* LRG 1 LRG 2 LRG 3 LRG 4 LRG 5 LVI Present Absent CRM Involved Not involved No. of lymph nodes retrieved per specimen Median Range Pathological lymph node status per specimen ypN+ve ypN–ve LNR† Average PNI Present Absent
34 52 80 22 2
(17.9) (27.4) (42.1) (11.6) (1.0)
3 (0–5) 3 (0–78) 41 100 45 72 41
(13.7) (33.7) (15.0) (24.0) (13.6)
27 (14.2) 163 (85.8) 11 (5.8) 179 (94.2) 11.5 2–39 85 (44.7) 105 (55.3) 0.1 4 (2.1) 186 (97.9)
TRG, tumour regression grade; LRG, lymph node regression; LVI, lymphovascular invasion; CRM, circumferential resection margin; PNI, perineural invasion. *The LRG scores are for all lymph nodes scored in every specimen. †Lymph node ratio = number of metastatic lymph nodes/total number of lymph nodes retrieved.
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
N. Mirbagheri et al.
Lymph node status after CRT for rectal cancer
1.0 LRGmax 1 0.8
Cum survival
LRGmax 2 LRGmax 0 0.6 LRGmax 3 LRGmax 5
0.4
LRGmax 4 0.2 P = 0.01 0.0 0
20
LRGmax score 0 1 2 3 4 5
40
60 80 100 Overall survival
120
Mean overall survival (months) 106.9 135.8 67.1 80.4 50.4 60.8
140
95% Cl 95.3-118.5 113.1-158.6 36.6-97.6 54.4-106.5 42.7-58.1 37.8-83.9
1.0 LRGmax 1 LRGmax 0 LRGmax 2
Cum survival
0.8
0.6
LRGmax 3 LRGmax 5
0.4 LRGmax 4 0.2
P = 0.009
0.0 0
Figure 2 Survival curves as a function of LRGmax.
20
40
60 80 100 Disease free survival
120
140
LRGmax score Mean disease-free survival (months) 95% Cl 0 107.5 95.7-119.3 1 136.3 114.4-158.2 2 68.1 36.2-99.9 3 45.5-105.5 75.5 4 48.6 39.7-57.5 5 56.6 29.9-83.1
note, in Table 3 specimens with maximum LRG score of 0.9 (i.e. previous tumour but complete response in all involved lymph nodes) were listed as yN+ve to
differentiate them from those specimens without any evidence of metastasis in the lymph nodes, i.e. LRG0 (i.e. yN–ve).
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
O343
N. Mirbagheri et al.
Lymph node status after CRT for rectal cancer
Clinical outcomes, prognostic factors and survival curves
Discussion To our knowledge, this is the first study to show the clinical significance of pathological lymph node response to preoperative treatment as a regression grade. It demonstrates that oncological outcome is better in patients with lymph node metastasis where a pathological response to preoperative treatment can be observed. Factors such as TRG score, location of the tumour and the presence of lymphovascular invasion were also shown to correlate with the regression grade. Lymph node regression scoring in the mesorectal nodes was first reported by Caricato et al. [5] in 2007, who showed a complete lymph node response (LRG 1)
1.0
1.0
0.8
0.8 LRGsum 0
0.6 LRGsum 1–9 LRGsum ≥10
0.4 0.2
LRGsum 0 Cum survival
Cum survival
The 5-year OS and DFS for the cohort was 72.6% and 75.8%, respectively, with an overall recurrence and mortality rate of 21%. Of the 40 recurrences, eight were local, 24 distant and eight both local and distant. The average OS and DFS were 41.9 and 38.1 months, respectively. The OS and DFS survival curve analysis (Figs 2 and 3) categorized the patients into groups based on their LRG score type. LRGsum was the strongest predictor of OS and DFS. The survival curve for LRGmax illustrates that patients with complete lymph node regression (LRG 1) have an improved survival curve in both OS and DFS. LRGsum has an advantage over LRGmax in that it largely relies on more nodes being sampled, which is reflected in the more significant P-values. Using multivariate Cox proportional hazard regression analysis, LRG scoring was not found to be a significant factor for mortality (Table 4). This is probably because the LRG score is highly correlated to other significant variables such as LVI and TRG. Thus, significant regressors for mortality using the LRGsum analysis included LVI and TRG (Table 4). Importantly, the LRG score was a significant predictor for recurrence; other significant regressors for recurrence were LVI and tumour location. The results using LRGmax analysis were similar to that of LRGsum. It should be noted that CRM was positively correlated with LVI. Whilst CRM was a statistically significant factor predicting recurrence, when both variables (LVI and CRM) were included
CRM was not statistically significant. Further, multivariate regression analysis demonstrated that higher TRG scores (P = 0.005), tumours located in the middle rectum (P = 0.013) and LVI (P = 0.029) were associated with higher LRG scores. Patients with LRG = 0 (i.e. no tumour in the lymph nodes) were included in the analysis as they acted as a control group, and we hypothesized that patients with complete tumour response in the lymph nodes (LRG 1) may fare even better than those with LRG 0. But given the small sample size for LRG 1 this hypothesis was not adequately tested in this study. When a subgroup analysis is performed on patients with lymph node metastasis (yN+ve, n = 85), only LRGsum remained a significant predictor of recurrence [hazard ratio (HR) 1.117, 95% CI 1.023–1.220; P = 0.014].
0.6
LRGsum 1–9 LRGsum ≥10
0.4 0.2
P = 0.000
0.0 0
20
40
100 60 80 Overall survival
120
140
LRGsum score Mean overall survival (months) 95% Cl
0 1–9 ≥10
117.7 85.9 51
103.8-131.7 66.2-105.7 35.9-66.2
P = 0.000
0.0 0
20
40
60 80 100 Overall survival
120
LRGsum score Mean overall survival (months)
0 1–9 ≥10
119.3 87.3 46.2
140
95% Cl
105.3-133.2 66.4-108.2 27.9-64.4
Figure 3 Survival curves as a function of LRGsum.
O344
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
N. Mirbagheri et al.
Lymph node status after CRT for rectal cancer
Table 4 Multivariate Cox proportional hazards regression analysis for oncological outcomes as a function of lymph node regression (LRG) scores (n = 190).
Mortality (n = 39) LRGsum LRGmax* Female sex Age Time to surgery Adjuvant therapy Tumour stage Total LN retrieved CRM LVI PNI TRG LNR Middle rectum Lower rectum APR Recurrence (n = 40) LRGsum LRGmax* Female sex Age Time to surgery Adjuvant therapy Tumour stage Total LN retrieved CRM LVI PNI TRG LNR Middle rectum† Lower rectum† APR‡
HR
95% CI
P-value
1.587 1.074 0.380 0.991 0.998 0.514 1.001 0.993 1.492 2.093 4.847 2.669 1.182 0.713 0.477 1.515
0.665–3.789 0.863–1.337 0.113–1.279 0.950–1.032 0.877–1.136 0.200–1.319 0.999–1.003 0.911–1.082 0.355–6.262 1.098–7.676 0.499–47.068 1.431–4.979 0.175–7.988 0.192–2.649 0.109–2.093 0.603–3.808
0.575 0.521 0.118 0.656 0.971 0.166 0.085 0.871 0.585 0.032 0.174 0.002 0.864 0.613 0.326 0.377
1.061 1.273 0.649 1.009 0.940 0.594 1.001 0.412 1.573 3.312 4.810 1.186 0.576 0.253 0.439 2.125
1.028–1.095 1.048–1.548 0.299–1.836 0.970–1.049 0.815–1.084 0.231–1.526 0.999–1.003 0.887–1.049 0.467–5.303 1.216–9.012 0.559–41.392 0.722–1.946 0.088–3.765 0.082–0.786 0.132–1.457 0.841–5.371
0.000 0.015 0.649 0.659 0.397 0.279 0.186 0.412 0.465 0.019 0.153 0.500 0.565 0.017 0.179 0.111
Bold represents significant values. HR, hazard ratio; TRG, tumour regression grade; LRG, lymph node regression; LVI, lymphovascular invasion; LN, lymph nodes; LNR, lymph node ratio; CRM, circumferential resection margin; PNI, perineural invasion; APR, abdominoperineal resection. *The results for LRGmax are from a separate Cox regression restricted model analysis. †Compared with the upper rectum. ‡Compared with anterior resection.
in 51% of specimens compared with 13.7% in this study. Their study, however, was limited by a small sample size (n = 35) and no clinical follow up. The importance of regression score in lymph nodes as well the primary tumour is based on the underlying pathophysiological
mechanism of lymph node metastasis. For tumour cells to metastasize effectively, rather than spill over, they need to acquire the ability to express cytokines such as vascular endothelial growth factor and adopt the normal mechanism of lymph node homing during metastasis (i.e. chemokine receptor expression) [6]. These ‘upgraded’ cancer cells may be more robust than the primary cancer cells, more resilient to CRT and thus have a differential response to treatment. In this study, we have illustrated that within the same specimen lymph nodes can have varied responses to CRT. Pathological complete response (pCR) had been traditionally described as complete eradication of all viable tumour cells from the rectal wall and mesorectum after preoperative CRT [7]. Complete clinical responders (cCR) are those patients in whom the tumour has been eradicated clinically and radiologically. This is supported by complete tumour regression in the primary tumour upon local excision irrespective of lymph nodes status. The rates of pCR in the literature vary from 10 to 20% compared with our observed rate of 17.9% [2,7,8]. The risk of lymph node metastasis in cCR varies from 0 to 14%, consistent with our observed rate of 8.8% [9,10]. Of note, in our study out of those patients with cCR in the primary tumour and lymph node metastasis in the specimen, 63% (5/8 specimens) also showed a pCR in the lymph nodes (LRG1). We observed that the LRG1 cohorts do better than those with LRG > 1 and even fare better than LRG0 in the later stages of follow up as demonstrated by the flattening out of the survival curve. Furthermore, Kaplan–Meier graphs also show better survival curves and disease-free status for patients with lower LRG scores. A multicentre retrospective study by Yeo et al. [8] on 333 locally advanced rectal cancers post-CRT, showed that ypN status was the most relevant independent prognostic factor for both OS and DFS. Therefore, it is not an unexpected finding that LRG scores have a similar oncological significance. Of interest, three clinicopathological factors (tumour location, TRG score and lymphovascular status) correlated significantly with LRG scores. In other studies it has been shown that in irradiated rectal cancer, ypT status (i.e. TRG score) correlates with yN status, thus supporting our result [11]. To our knowledge this is the only study that has shown tumour location to be a significant predictor of lymph node response posttreatment, but the underlying mechanism is uncertain. This study is inherently limited by its retrospective nature (and thus the biases involved in the clinical information obtained), the heterogeneity of surgical techniques employed and the pathological preparation of lymph nodes. The LRG score was made on lymph
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
O345
N. Mirbagheri et al.
Lymph node status after CRT for rectal cancer
nodes that were prepared at the same time as the specimen was resected and the specimens were no longer available for further lymph node sampling. Moreover lymph nodes completely destroyed by radiation may have been missed. Notably, we observed that higher sampling of lymph nodes (especially when 11 or more nodes were sampled) markedly improved the prognostic power of LRG analysis (data not shown). While no current guidelines are provided to define adequate nodal sampling in these posttreatment specimens, the suggestion is that adequate lymph node sampling may be a barrier to optimizing nodal assessment and should be addressed in future prospective studies. In conclusion, lymph node response in the form of pathological regression grade appears to be an important prognostic indicator for oncological outcomes. Therefore more tailored adjuvant treatment can be directed to patients with poorer lymph node response. We hope that this pilot study creates a platform for further supporting studies in which a lymph node grading system can be implemented and patients followed up prospectively. As we develop further insights into the pathological response of irradiated lymph nodes, we may enter a new era of rectal cancer treatment where surrogate markers for lymph node response may enable a more conservative operative approach.
Acknowledgement We would like to thank Zdenka Prodanovic, Tissue Bank Manager at Southern Health, for her invaluable assistance in facilitating this project.
Author’s contributions Naseem Mirbagheri: Study concept, design, acquisition of data, data analysis and interpretation, manuscript writing and final approval. Beena Kumar: Study design, acquisition of data, manuscript writing and final approval. Siddhartha Deb: Study design, acquisition of data, data analysis and interpretation, manuscript writing and final approval. Ben Poh: Study design, acquisition of data, data analysis, manuscript writing and final approval. Jonathan Dark: Study design, data analysis and interpretation, manuscript writing and final approval. Christopher Leow: Acquisition of data, manuscript writing and final approval. William Teoh: Study
O346
design, interpretation of data, manuscript writing and final approval.
References 1 Habr-Gama A. Non-operative management of rectal cancer after neoadjuvant chemoradiation. Br J Surg 2009; 96: 125–7. 2 Kosinski L, Habr-Gama A, Ludwig K, Perez R. Shifting concepts in rectal cancer management: a review of contemporary primary rectal cancer treatment strategies. CA Cancer J Clin 2012; 62: 173–202. 3 Kahn H, Alexander A, Rakinic J, Nagle D, Fry R. Preoperative staging of irradiated rectal cancers using digital rectal examination, computed tomography, endorectal ultrasound, and magnetic resonance imaging does not accurately predict T0, N0 pathology. Dis Colon Rectum 1997; 40: 140–4. 4 Mandard AM, Dalibard F, Mandard JC, Marnay J. Pathologic assessment of tumor regression after preoperative chemoradiotherapy of esophageal carcinoma. Clinicopathologic correlations. Cancer 1994; 73: 2680–6. 5 Caricato M, De Dominicis E, Vincenzi B, Rabitti C. Tumor regression in mesorectal lymphnodes after neoadjuvant chemoradiation for rectal cancer. Eur J Surg Oncol 2007; 33: 724–8. 6 Nathanson SD. Insights into the mechanisms of lymph node metastasis. Cancer 2003; 98: 413–23. 7 Jang TY, Yu CS, Yoon YS et al. Oncologic outcome after preoperative chemoradiotherapy in patients with pathologic T0 (ypT0) rectal cancer. Dis Colon Rectum 2012; 55: 1024–31. 8 Yeo S-G, Kim DY, Kim TH et al. Pathologic complete response of primary tumor following preoperative chemoradiotherapy for locally advanced rectal cancer. Ann Surg 2010; 252: 998–1004. 9 Lindebjerg J, Spindler KLG, Ploen J, Jakobsen A. The prognostic value of lymph node metastases and tumour regression grade in rectal cancer patients treated with long-course preoperative chemoradiotherapy. Colorectal Dis 2009; 11: 264–9. 10 Habr-Gama A, Perez RO, Proscurshim I et al. Absence of lymph nodes in the resected specimen after radical surgery for distal rectal cancer and neoadjuvant chemoradiation therapy: what does it mean? Dis Colon Rectum 2008; 51: 277–83. 11 Read TE, Andujar JE, Caushaj PF et al. Neoadjuvant therapy for rectal cancer: histologic response of the primary tumor predicts nodal status. Dis Colon Rectum 2004; 47: 825–31.
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, O339–O346
