It is widely accepted that after long-course neoadjuvant chemoradiotherapy (nCRT) for rectal cancer, the surgical specimen will have a lower lymph node (LN) number than surgical specimens from patients without nCRT.1–3 Although the TNM staging system4 recommends that at least 12 LNs are examined for correct staging, the prognostic significance of this threshold in those who have undergone nCRT is controversial. Some question this cutoff and associate the low LN count with better response to the nCRT1 , 5–8; others contend that fewer LNs increase the risk for understaging and correlate with worse survival.9–11 In addition, several studies have shown that thorough specimen analysis increases LN yield, suggesting that nCRT does not decrease LN count.12–16
LN-revealing solutions (LNRS) are an option to increase LN count.13 , 17–20 Carnoy’s solution (CS) is a tissue fixative that efficiently preserves tissue morphology and the abilities to perform immunohistochemical analysis and DNA extraction.21 It consists of ethanol (60%), chloroform (30%), and glacial acetic acid (10%), allowing an easier identification of the LNs and greatly facilitating the identification of small LNs (Fig. 1).22 , 23 Compared with 10% neutral buffered formalin (NBF), CS requires the same workplace and worker safety measures, has lower occupational risk, and is slightly more expensive ($1.22 per specimen).22
We hypothesized that resected LNs are missed by routine histopathological analysis after nCRT for rectal cancer. The main objective was to compare CS and NBF concerning the number of cases with <12 LNs and the total LN count. Our secondary objectives were:
- Check if surgically harvested LNs are missed following routine histopathological analysis and if this is clinically relevant.
- Evaluate the relationship between tumor response to nCRT and LN count.
- Observe if the research protocol influenced LN retrieval.
MATERIALS AND METHODS
One hundred thirty specimens from patients who underwent low anterior resection (LAR) with total mesorectal excision (TME) after nCRT for rectal adenocarcinoma were randomly assigned for fixation with CS (CS group) or NBF (NBF group). The randomization ratio was 1:1.
Specimens were fixed for 24 to 48 hours; then, the pericolic fat was measured and manually dissected for LNs. All histological blocks were submerged in formalin for processing. Pathological analysis was performed according to the College of American Pathologists protocol.24 Pathological response to nCRT was determined by the Tumor Regression Grade (TRG) system.24 , 25 It ranges from 0 to 3, where TRG0 is equivalent to complete response.
After dissection, the residual fat from the NBF group was immersed in CS for another 24 to 48 hours and dissected in search for missed LNs (Revision group). Data from this analysis were added to the NBF group providing the NBF+Revision group (Fig. 2).
All patients scheduled for surgical treatment after nCRT for rectal adenocarcinoma were considered for the study. Complete long-course neoadjuvant therapy with 50.4 Gy pelvic irradiation and 5-fluorouracil plus leucovorin was required.26 , 27
To homogenize the specimens, only standard LAR (ligation of the inferior mesenteric artery at its origin) with TME were included. Exclusion criteria were: multivisceral resections, abdominoperineal resection of the rectum, and previous surgery on the left colon/rectum (including transanal endoscopic microsurgery).28
Considering the same eligibility criteria, data from the last 65 patients operated on before the study were collected (Retrospective group).
This study was performed in a reference cancer center in Sao Paulo, Brazil, during the period of May 2012 through March 2015. It was approved by our ethics committee and is registered at clinicaltrials.gov as NCT02629315: Carnoy’s Solution Versus Formalin in Rectal Cancer Specimens Following Neoadjuvant Chemoradiation: a Randomized Trial.
Five specialized surgeons operated on all patients and 1 pathologist was responsible for handling all specimens.
The primary outcome was the number of cases with ≥12 LNs. To avoid count errors, LNs were counted by 2 independent pathologists (M.A.P. and E.S.M.) and discrepancies were verified by a third pathologist (S.F.F.).
As for the secondary end points, the pathologist responsible for dissecting was also accountable for collecting the following data: pericolic fat volume, duration of the dissection and number, and size and status of the LNs.
To calculate the sample size, a pilot study with 50 cases was performed. The number of cases with <12 LNs examined was 9 (36%) and 4 (16%) for NBF and CS groups. To confirm these findings with statistical power of 80% (β = 0.2) at a 2-sided α of 5%, a sample size of 62 per arm would be required. This was rounded to 65 per arm. Because methodology was exactly the same, the pilot sample was included in the study.
Randomization was computer generated using SAS Enterprise Guide version 4.3. After each surgery, the statistician informed which group the specimen was assigned. Because allocation occurred after all eligibility criteria were matched, no case was lost. Surgeons remained blind during the procedure. The pathologist could not be blinded because both solutions have distinct scents.
Statistical analysis was performed using SPSS for Windows, version 18.0 (SPSS Inc, Chicago, IL). The χ2 test was used for nominal variables, whereas continuous variables were analyzed with the t test or the ANOVA test. Pearson correlation coefficient (r) was used to measure the linear relationship between 2 variables. The OR was calculated with 95% CI. All tests were 2-sided, and p < 0.05 was regarded as significant.
All 130 patients included in the trial received the intended intervention and were analyzed according to the randomization scheme. None of the patients were excluded.
CS and NBF Groups
Carnoy’s solution and NBF groups were similar regarding age (p = 0.21), BMI (p = 0.72), and interval between nCRT and surgery (median of 14 weeks for both groups, p = 0.08). Women were predominant in the NBF group (p = 0.02). Laparoscopic access was performed in 22 cases in the CS group (3 conversions) and 28 cases in the NBF group (4 conversions). Dissection duration was 45.7 and 49.0 minutes for the CS and NBF groups (p = 0.02). LN count was 24.0 and 16.3 in groups CS and NBF (p < 0.01). Lymph node sizes ranged from 1 to 25 mm (CS group) and 1 to 24 mm (NBF group). The mean number of LNs <5 mm was higher in the CS group (21.6 vs 14.6, p < 0.01). Six cases in the CS and 22 in the NBF group had <12 LNs (p = 0.001) (Table 1). The OR of having <12 LNs in the NBF group compared with the CS group was 6.42 (95% CI, 2.73–15.07; p < 0.001).
A total of 722 LNs were found after treating the residual fat from the NBF group with CS (mean, 11.1 LNs). Dissection duration was 31.5 minutes (mean), and LN size ranged from 1 to 4 mm. Eight metastatic LNs were found in 6 cases (9.2%); they measured 1 to 3 mm.
Adding to the NBF group, the findings from the Revision group reduced the number of cases with <12 LNs from 22 (33.8%) to 3 (4.6%) (Table 2). It also changed the N status of 4 patients (N0 to N1a, N1a to N1b, N1b to N2a, and N2a to N2b) and upstaged 2 of them (Table 2).
Lymph nodes in the NBF+Revision group were <5 mm in 93.6% of the cases, and 55.4% of the positive nodes were <5mm.
Data from the Revision group increased the LN count 84% in TRG0 cases (from 12.0 to 22.1, p < 0.001), 63.3% in TRG1 (from 19.1 to 31.2, p < 0.001), 70.4% in TRG2 (from 15.0 to 26.7, p < 0.001), and 55.8% in TRG3 (from 19.2 to 30.0, p < 0.001).
CS and NBF+Revision Groups
Comparing groups CS and NBF+Revision groups, dissection duration was 45.7 and 80.9 minutes (p < 0.001), LN count was 24.0 and 27.4 (p = 1.0), and the mean number of LNs <5 mm was 21.6 and 25.7 (p = 0.04). Six cases in the CS group and 3 in the NBF+Revision group had <12 LNs dissected (p = 0.49) (Table 1).
Pathological complete response (pCR) was observed in 25 cases (11 and 14 cases in groups CS and NBF+Revision). In 3 pCR patients, <12 LNs were dissected (2 from the CS group). Two cases in the NBF+Revision group were ypT0ypN1.
All 130 cases were analyzed for factors influencing LN count (Table 3). There was no linear relationship between LN number and the following variables: age (r = −0.162, p = 0.06), BMI (r = −0.069, p = 0.43), interval between nCRT and surgery (r = −0.159, p=0.07), pre- and post-nCRT CEA level (r = 0.165, p = 0.06 and r = 0.116, p = 0.21), percentage of tumor reduction (r = −0.096, p = 0.27), and pericolic fat volume (r = 0.043, p = 0.62). Tumor sizes pre- and post-nCRT were associated with LN number (Fig. 3).
Pathological T status and TRG did not correlate with LN count. Higher pT status and higher TRGs were both associated with increased number of positive LNs (Table 3). In addition, pCR and non-pCR patients had an equivalent number of cases with <12 LNs (3/25 and 6/105; p = 0.26) and similar mean LN count (Table 3). Considering only the NBF group, pCR and non-pCR patients had an equivalent number of patients with <12 LNs (7/14 for pCR and 15/51 for non-pCR, p = 0.14), with a lower LN count for the pCR patients (12.3 vs 17.4, p = 0.04).
The Retrospective group and the NBF group were similar concerning age (60.5 and 58.5, p = 0.37), sex (29 and 36 women, p = 0.21) and BMI (25.9 and 26.3, p = 0.66). The mean number of LNs (16.2 and 16.3, p = 0.94) and the number of cases with <12 LNs (20 and 22, p = 0.85) were also comparable. The interval between the end of nCRT and surgery was longer in the Retrospective group (median, 17 vs 14 weeks, p < 0.01). Laparoscopic access was performed in 33 cases from the Retrospective group (4 conversions). Ten cases in the Retrospective group had pCR and 1 was ypT0ypN2a.
Surgical specimens from patients who underwent nCRT for rectal cancer may have fewer LNs in comparison with those without this therapy. Literature about this fact is overwhelming, but focused on its clinical significance, disregarding the reason for this finding, and neglecting what happens with the LNs. Also, there is no consensus about the meaning of finding <12 LNs in these patients. Some advocate that the cutoff of 12 LNs is inadequate in this setting, that fewer LNs indicate a good response to nCRT and good prognosis.1 , 5–7 But these findings may be due to the limitations and bias in those studies, and others present an opposite view, demonstrating that increased LN count reduces the chance for understaging and correlates with improved survival.9–11 A large observational study noticed that high LN count was associated with better survival.10 In addition, some have shown that a zealousness analysis of the specimen substantially increases LN count.12–16 Scheel et al12 reported a significant increase in LN count with acetone compression and whole mesorectal embedding compared with manual dissection (MD). Interestingly, the mean number of LNs retrieved by acetone compression was similar for patients with and without nCRT (27 and 30, p = 0.13), but after MD, nCRT patients had significantly fewer LNs (p = 0.003). This highlights that LN count depends more on the specimen preparation than on nCRT. Wang et al13 reported that, after nCRT, LN retrieval (5.2 vs.20.4) and the number of positive LNs (0.4 vs 1.2) were both significantly increased by using 100% alcohol as fixative solution. In a randomized study, the injection of methylene blue in colorectal specimens significantly reduced the cases with <12 LNs.14 Also, an alcohol protocol significantly increased LN yield in patients with nCRT (15 vs. 23).15
So, it may be inferred that finding LNs after nCRT is a troublesome task. In comparison with patients with surgery upfront, LNs in those who underwent nCRT are significantly smaller, and this difference is even more pronounced when MD is performed.12 In a large series, the addition of 1 Gy decreased the number of LNs retrieved by 0.21%.3 This suggests that radiotherapy plays a main role, and local fibrosis along with lymphoid tissue radiosensitivity may justify this.3 , 13 , 29 In addition, routine specimen analysis has low sensitivity for LNs ≤3mm,13 , 22 and analyzing them is extremely relevant because micrometastases (0.2–2 mm) are a significant percentage of the positive LNs and negatively impact survival.12 , 30 , 31 Bhangu et al11 observed in a large database that each additional LN examined increased the chance for stage III disease (N+) in 3.9%.
The present study was designed to rigorously verify if <12 LNs can be expected in surgical specimens from patients with rectal cancer after nCRT. Carnoy’s solution was chosen because it allows an easier identification of small LNs. In a randomized trial, CS significantly increased LN yield following D2 gastrectomy in patients with gastric cancer when compared with NBF (50.4 vs 34.8, p < 0.001).22
Concerning our primary outcome, CS was superior to NBF with 9.2% vs 33.8% (p = 0.001) of the cases with <12 LNs. This is similar to the 9.1% observed by Scheel et al12 in their cases with nCRT and the 9.2% from Luz et al17 in colorectal cancer specimens following the use of a LNRS. Compared with CS, NBF has a chance 6.42 times higher of retrieving <12 LNs.
The mean number of LNs per specimen was superior in the CS group (24.0 vs 16.3) mainly because of the finding of small LNs (mean number of LNs <5 mm: 21.6 vs 14.6). Carnoy’s solution also allowed for faster dissection, gaining 4minutes (mean) per specimen. The mean LNs/min was 0.53 and 0.33 for CS and NBF.
Groups were homogeneous except for sex, which was not relevant because it did not correlate with LN count in the prospective analysis (130 patients). The Revision group was created to verify if LNs are missed following routine NBF fixation. The 722 LNs found in the residual fat accounted for 40.4% of all LNs in the NBF+Revision group and reduced the number of specimens <12 LNs from 33.8% to 4.61%. LNs ≤3 mm were being missed in all cases in the NBF group, and this fact was clinically relevant because 9.2% of the patients had metastatic LNs being discarded. This modified the N status in 4 cases and upstaged 2 of them. One case deserves special attention, the patient had stage IIA disease (ypT3ypN0, 0+/36 LNs), and after a 2-mm compromised LN was found in the revision (Fig. 4), was upstaged to stage IIIB (ypT3ypN1a, 1+/42 LNs). This highlights the importance for adequate analysis of all resected LNs and demonstrates that a cutoff number does not necessarily accurately predict stage. Lymph node count in N+ patients also underlines this (28.5, range 13–66). So, the cutoff should not be reduced; it serves to emphasize the importance of examining at least 12 LNs. After adequate surgery and dedicated histopathological analysis, only a few patients will be below this threshold (<10%).
Groups CS and NBF+Revision were similar concerning the number of LNs and cases with <12 LNs. NBF+Revision group had more LNs <5 mm, probably because it underwent 2 dissections. As expected, the time dedicated for dissection was much longer in this last group (80.9 vs 45.7 minutes).
All 130 prospective cases were analyzed for factors that might influence LN yield. This was possible because groups CS and NBF+Revision had an equivalent number of LNs. Larger tumors were mildly associated with increased LN number. Higher pT status and TRGs both correlated with an increased number of positive LNs.
Some groups reported that low LN yield is sign of good response to nCRT and of good prognosis.1 , 2 , 5 , 7 , 32–35 This was not observed in the present study, and the mean number of LNs was way above the minimum recommended even in TGR0 cases. In addition, 5 of the 9 cases with <12 LNs were bad responders. To further investigate this matter, we compared the number of cases with <12 LNs and the mean LN count in pCR and non-pCR patients in the 130 cases (which reflects what happens when CS is used) and in the NBF group only (which is expected to be similar to what the literature reports). The number of cases with <12 LNs was similar for pCR and non-pCR in both situations. As for LN yield, this was equivalent in the prospective cases, but, in the NBF group, pCR patients had a significantly lower LN count compared with the non-pCR patients (12.3 vs 17.4). This demonstrates that LNs are harder to find in the best responders when routine NBF is used. It also explains why some mistakenly correlate finding fewer LNs with good response to nCRT.1 , 7 , 33 , 34 We also observed that the Revision group data increased LN count more expressively in TRG0 patients, confirming that LN dissection is more burdensome in the best responders.
The incidence of pCR in the 195 cases was 18.4% (25 and 11 cases in the prospective and retrospective groups). As discussed elsewhere, preoperative sensitivity for identifying pCR patients is low.27 , 36 , 37
The Retrospective group was analyzed to verify if the study influenced the results in the NBF group. These groups had equivalent LN count demonstrating that the NBF group was neither over- nor underdissected.
Limitations and Validity
Among the limitations of the study are that it was unicentric and a second dissection was not performed in the CS group.
To ensure internal validity, rigorous measures were taken: standard LAR with TME after complete nCRT was required, patients with previous procedures that might compromise LN harvest were excluded, patients were operated on by experienced surgeons blind to the allocation, the same pathologist dissected all specimens, and the groups were randomly distributed.
As for the external validity, NBF group results match literature reports from reference centers worldwide and are also equivalent to the retrospective cases when noncontrolled variables were in place.
Other strong points are: its design allowed no specimen loss; LN count was high in groups NBF and Retrospective, certifying the high standard of the routine treatment in our institution; and there was no ethical issue from submitting the specimens to an inferior fixative, because patients from the NBF group also benefited from the CS. In addition, this is the first randomized trial with a large number to test the efficacy of LNRS in patients with nCRT for rectal cancer, and the first time that pCR and non-pCR patients were consistently investigated concerning LN count.
In patients with rectal cancer who underwent nCRT, specimen fixation with CS reduces the number of cases with <12 LNs compared with routine NBF. Surgically harvested LNs are missed following routine histopathological analysis, and this is clinically relevant. Carnoy’s solution speeds up dissection and increases LN count due to the finding of small LNs. The research protocol did not influence LN retrieval. Finding <12 LNs is not a sign of good tumor response to nCRT.
1. Persiani R, Biondi A, Gambacorta MA, et alPrognostic implications of the lymph node count after neoadjuvant treatment for rectal cancer
. Br J Surg. 2014;101:133–142.
2. Awwad GE, Tou SI, Rieger NAPrognostic significance of lymph node yield after long-course preoperative radiotherapy in patients with rectal cancer
: a systematic review. Colorectal Dis. 2013;15:394–403.
3. Rullier A, Laurent C, Capdepont M, et alLymph nodes after preoperative chemoradiotherapy for rectal carcinoma: number, status, and impact on survival. Am J Surg Pathol. 2008;32:45–50.
4. Edge SB, Byrd DR, Carducci MA, et alAJCC Cancer Staging Manual. 2010.7th ed. New York, NY: Springer-Verlag;
5. Kim WR, Han YD, Cho MS, et alOncologic impact of fewer than 12 lymph nodes in patients who underwent neoadjuvant chemoradiation followed by total mesorectal excision for locally advanced rectal cancer
. Medicine (Baltimore). 2015;94:e1133.
6. Hall MD, Schultheiss TE, Smith DD, et alImpact of total lymph node count on staging and survival after neoadjuvant chemoradiation therapy for rectal cancer
. Ann Surg Oncol. 2015;22(suppl 3):S580–S587.
7. Damin DC, Rosito MA, Contu PC, et alLymph node retrieval after preoperative chemoradiotherapy for rectal cancer
. J Gastrointest Surg. 2012;16:1573–1580.
8. Ha YH, Jeong SY, Lim SB, et alInfluence of preoperative chemoradiotherapy on the number of lymph nodes retrieved in rectal cancer
. Ann Surg. 2010;252:336–340.
9. Lykke J, Jess P, Roikjaer ODanish Colorectal Cancer Group. A minimum yield of twelve lymph nodes in rectal cancer
remains valid in the era of neo-adjuvant treatment: results from a national cohort study. Int J Colorectal Dis. 2015;30:347–351.
10. Lykke J, Jess P, Roikjaer ODanish Colorectal Cancer Group. Increased lymph node yield is associated with improved survival in rectal cancer
irrespective of neoadjuvant treatment: results from a national cohort study. Dis Colon Rectum. 2015;58:823–830.
11. Bhangu A, Kiran RP, Brown G, Goldin R, Tekkis PEstablishing the optimum lymph node yield for diagnosis of stage III rectal cancer
. Tech Coloproctol. 2014;18:709–717.
12. Scheel AH, Reineke RA, Sprenger T, et alComprehensive lymph node morphometry in rectal cancer
using acetone compression. J Clin Pathol. 2015;68:458–464.
13. Wang H, Safar B, Wexner SD, Denoya P, Berho MThe clinical significance of fat clearance lymph node harvest for invasive rectal adenocarcinoma
following neoadjuvant therapy. Dis Colon Rectum. 2009;52:1767–1773.
14. Borowski DW, Banky B, Banerjee AK, et alIntra-arterial methylene blue injection into ex vivo colorectal cancer specimens improves lymph node staging accuracy: a randomized controlled trial. Colorectal Dis. 2014;16:681–689.
15. Duldulao M, Booth C, Denham L, Choi A, Friedman G, Kazanjian KAlcohol fat clearing increases lymph node yield after surgery for colorectal cancer. Am Surg. 2014;80:1054–1058.
16. Sprenger T, Rothe H, Homayounfar K, et alPreoperative chemoradiotherapy does not necessarily reduce lymph node retrieval in rectal cancer
specimens–results from a prospective evaluation with extensive pathological work-up. J Gastrointest Surg. 2010;14:96–103.
17. Profeta da Luz MM, Lacerda-Filho A, Demas Alvares Cabral MM, et alThe role of lymph node revealing solution
on the improvement of lymph node harvest in colorectal cancer specimens. Colorectal Dis. 2016;18:247–254.
18. Ma XL, Ye JX, Su J, Qi FF, Meng QY, Shi XYA modified GEWF solution is cost-saving and effective for lymph node retrieval in resected colorectal carcinoma specimens. Pathol Res Pract. 2014;210:543–547.
19. Iversen LH, Laurberg S, Hagemann-Madsen R, Dybdahl HIncreased lymph node harvest from colorectal cancer resections using GEWF solution: a randomised study. J Clin Pathol. 2008;61:1203–1208.
20. Koren R, Siegal A, Klein B, et alLymph node-revealing solution: simple new method for detecting minute lymph nodes in colon carcinoma. Dis Colon Rectum. 1997;40:407–410.
21. Pereira MA, Dias AR, Faraj SF, et alCarnoy’s solution is an adequate tissue fixative for routine surgical pathology
, preserving cell morphology and molecular integrity. Histopathology. 2015;66:388–397.
22. Dias AR, Pereira MA, Mello ES, Zilberstein B, Cecconello I, Ribeiro Junior UCarnoy’s solution increases the number of examined lymph nodes following gastrectomy for adenocarcinoma: a randomized trial. Gastric Cancer. 2016;19:136–142.
23. Luz DA, Ribeiro U Jr, Chassot C, Collet E Silva Fde S, Cecconello I, Corbett CECarnoy’s solution enhances lymph node detection: an anatomical dissection study in cadavers. Histopathology. 2008;53:740–742.
25. Ryan R, Gibbons D, Hyland JM, et alPathological response following long-course neoadjuvant chemoradiotherapy
for locally advanced rectal cancer
. Histopathology. 2005;47:141–146.
26. Sauer R, Becker H, Hohenberger W, et alGerman Rectal Cancer
Study Group. Preoperative versus postoperative chemoradiotherapy for rectal cancer
. N Engl J Med. 2004;351:1731–1740.
27. Habr-Gama A, Perez RO, Nadalin W, et alOperative versus nonoperative treatment for stage 0 distal rectal cancer
following chemoradiation therapy: long-term results. Ann Surg. 2004;240:711–717.
28. Dias AR, Nahas CS, Marques CF, Nahas SC, Cecconello ITransanal endoscopic microsurgery: indications, results and controversies. Tech Coloproctol. 2009;13:105–111.
29. Marijnen CA, Nagtegaal ID, Klein Kranenbarg E, et alPathology Review Committee and the Cooperative Clinical Investigators. No downstaging after short-term preoperative radiotherapy in rectal cancer
patients. J Clin Oncol. 2001;19:1976–1984.
30. Märkl B, Herbst C, Cacchi C, et alPrognostic significance of histologically detected lymph node micrometastases of sizes between 0.2 and 2 mm in colorectal cancer. Int J Colorectal Dis. 2013;28:977–983.
31. Sirop S, Kanaan M, Korant A, et alDetection and prognostic impact of micrometastasis in colorectal cancer. J Surg Oncol. 2011;103:534–537.
32. Li Q, Zhuo C, Liang L, Zheng H, Li D, Cai SLymph node count after preoperative radiotherapy is an independently prognostic factor for pathologically lymph node-negative patients with rectal cancer
. Medicine (Baltimore). 2015;94:e395.
33. Ishihara S, Fukushima Y, Akahane T, et alNumber of lymph nodes in rectal cancer
is correlated with response to preoperative chemoradiotherapy but is not associated with patient survival. Hepatogastroenterology. 2014;61:1000–1007.
34. de Campos-Lobato LF, Stocchi L, de Sousa JB, et alLess than 12 nodes in the surgical specimen after total mesorectal excision following neoadjuvant chemoradiation: it means more than you think! Ann Surg Oncol. 2013;20:3398–3406.
35. La Torre M, Mazzuca F, Ferri M, et alThe importance of lymph node retrieval and lymph node ratio following preoperative chemoradiation of rectal cancer
. Colorectal Dis. 2013;15:e382–e388.
36. Nahas SC, Rizkallah Nahas CS, Sparapan Marques CF, et alPathologic complete response in rectal cancer
: can we detect it? lessons learned from a proposed randomized trial of watch-and-wait treatment of rectal cancer
. Dis Colon Rectum. 2016;59:255–263.
37. Pereira MA, Dias AR, Faraj SF, et alOne-level step section histological analysis is insufficient to confirm complete pathological response after neoadjuvant chemoradiation for rectal cancer
. Tech Coloproctol. 2017;21:745–754.
Carnoy’s solution; Combined modality treatment; Fat clearing solution; Lymph node excision; Lymph node revealing solution; Neoadjuvant chemoradiotherapy; Pathology; Random allocation; Rectal adenocarcinoma; Rectal cancer
Supplemental Digital Content
© 2018 The American Society of Colon and Rectal Surgeons