Preoperative evaluation of pelvic lateral lymph node of patients with lower rectal cancer: comparison study of MR imaging and CT in 53 patients

Background Preoperative assessment of the lateral pelvic lymph nodes is important for treatment strategy to patients with lower rectal cancer.Materials and methods Fifty-three patients with primary lower rectal cancer were preoperatively assessed by spiral computed tomography (CT) and magnetic resonance imaging (MRI) at 1.5 T with a phased-array coil. Preoperative tumor and lymph node stages were compared with the final histological findings.Results The MRI tumor stage coincided with the histological stage in 36 of 53 patients (68%). The MRI and CT lymph node stage coincided with the histological stage in 33 (62%) and 26 (49%) of 53 patients, respectively. However the accuracy of MRI in detecting the lateral pelvic lymph node involvement was 83%, compared to 77% of CT (p<0.05).Conclusions With the use of MRI, the lateral pelvic lymph node involvement can be predicted with high accuracy, allowing preoperative identification of patients who need radiotherapy or extensive surgery to escape recurrence.     

MRI在直肠癌术前分期中的作用

目的:评估盆腔MRI检查在直肠癌术前分期和治疗决策中的作用。方法:对2009年4月至2010年6月手术治疗的60例直肠癌病例的术前盆腔MRI检查结果与术后组织病理学诊断结果进行比较,分析MRI对直肠癌术前分期的准确率。结果:MRI对直肠癌浸润深度(T分期)的诊断准确率为75%,对T2期肿瘤的诊断准确率为73.1%,对T3期肿瘤的诊断准确率为86.7%;对淋巴结转移的诊断准确率为32.4%。在病理确诊淋巴结转移的16例病人中,MRI检出淋巴结平均数为5.8枚;在淋巴结转移阴性的44例病人中,MRI检出淋巴结平均数为2.4枚;两组淋巴结数有显著差异(P0.05)。结论:术前MRI检查可较准确地判断肿瘤在直肠壁的浸润深度,但对淋巴转移的诊断准确率较低,故MRI可作为直肠癌术前分期的方法,为新辅助治疗提供依据,为术后辅助化疗提供信息。     

462例中下段直肠癌淋巴转移规律与淋巴清扫范围的分析

目的 探讨中下段直肠癌的淋巴转移规律和淋巴清扫范围。方法 对1990－1999年行传统直肠癌根治术的373例和行传统直肠癌根治术加盆腔侧方淋巴清扫术（简称侧方清扫术）的89例中下段直肠癌患者进行回顾性分析。结果 全组淋巴转移率为41．8％,患者年龄、癌灶浸润深度、大体分型、癌灶大小是影响淋巴转移率的重要因素（P＜0．05）。89例侧方清扫术的盆腔侧方淋巴转移率为15．7％,其中85．7％位于癌灶同侧。有盆腔侧方淋巴结转移者均为浸润深度T3、T4者;癌灶＞3cm、溃疡型或浸润型、年龄＜60岁者盆腔侧方淋巴结转移较高。侧方清扫术组的盆腔复发率为5．6％,明显低于传统直肠癌根治术组的17．7％（P＜0．05）;侧方清扫术组和传统直肠癌根治术组的5年生存率分别为46．7％和47．9％（P＞0．05）。结论 应提高对中下段直肠癌淋巴转移规律的认识,对怀疑或证实有淋巴结转移、癌灶侵犯浆膜或穿透肠壁、癌灶＞3cm、溃疡型或浸润型、年龄＜60岁者建议行侧方清扫术。     

Preoperative staging by multidetector-row computed tomography in patients with rectal carcinoma

BACKGROUND: Multidetector-row computed tomography (MDCT, or multi-slice CT) has been introduced in 2000. So far, there has been no published study on this modality in patients with rectal carcinoma. METHODS: Twenty patients with rectal carcinoma were preoperatively examined by MDCT and conventional CT (CCT). Diagnostic accuracies of both modalities were compared regarding the evaluation of depth of tumor invasion (Tis/T1/T2, T3, T4) and lymph node metastasis based on the pathologic findings. RESULTS: Although CCT detected a tumor in 13 (65%) of 20 patients, MDCT revealed a tumor in all 20 patients (P = 0.004). Regarding depth of tumor invasion, the concordance rate was significantly higher for MDCT (20/20: 100%) than for CCT (12 of 20: 60%; P = 0.002). Regarding lymph node metastasis, the overall accuracy was 70.0% in CCT, and also 70.0% in MDCT. CONCLUSIONS: MDCT was superior to CCT in the evaluation of depth of tumor invasion, but was equal to CCT in the evaluation of lymph node metastasis.     

Focal levator ani eventrations: detection and characterization by magnetic resonance in patients with pelvic floor dysfunction

The aim of this study was to assess the frequency and imaging characteristics of focal levator eventrations in patients with pelvic floor dysfunction on magnetic resonance (MRI). A review of 81 dynamic MR pelvic examinations in patients with pelvic floor dysfunction was carried out to detect and characterize focal eventrations in the levator ani muscle. These were defined as muscle outpouchings which made an angle of >180 degrees with the remainder of the muscle and had a depth of >or=1 cm. Of 81 patients 11(13.5%) had focal eventrations in the levator muscle on MRI: bilateral in 2 cases, right in 5 and left in 4. There was protrusion of pelvic viscera into the eventration in 5 cases, fat in 7 and fluid in 1. Focal levator ani muscle abnormalities are not uncommon on MRI in patients with pelvic floor dysfunction. Characterization of levator muscle morphology can be useful as a research tool in this population.     

A prospective comparison between multidetector-row computed tomography and magnetic resonance imaging in the preoperative evaluation of rectal carcinoma

BACKGROUND: Multidetector-row computed tomography (MDCT, or multislice CT) is a new modality with four detectors, which makes examination time shorter and produces higher resolution and multiplanar reformation of the images. Its diagnostic role in patients with rectal carcinoma has not been determined. METHODS: Twenty-one patients with rectal carcinoma were preoperatively examined by both MDCT and magnetic resonance imaging (MRI). Diagnostic accuracies of both modalities were compared regarding depth of tumor invasion and lymph node metastasis based on the pathologic findings. RESULTS: Both examinations detected all tumors. Regarding depth of tumor invasion, the concordance was 95.2% (20 of 21) for MDCT and 100% (21 of 21) for MRI. Regarding lymph node metastasis, the overall accuracy was 61.9% for MDCT and 70.0% for MRI. CONCLUSIONS: Multidetector-row computed tomography was equal to MRI in the preoperative local staging of rectal carcinoma.     

仿真内窥镜CT和多平面重建对直肠癌术前分期的判断

目的：探讨CT仿真内镜（CTVE）和多平面重建（MPR）对直肠癌术前分期的判断。方法：通过系统地采用CTVE和MPR与普通CT对比，对45例直肠癌患者的分期进行了评估。结果：CTVE和 MPR术前分期的准确率为86.7%（39/45），普通盆腔CT为66.7%（30/45），两者间差异有显著性（P<0.05）。有淋巴结转移的术前准确判断敏感性CTVE和 MPR为76.9%（20/26），普通盆腔CT为50.0%（13/26），但差异有显著性（P<0.05）。结论：CTVE和 MPR对直肠癌进行临床分期的准确性较普通CT高，对临床治疗具有指导意义。     

3.0T MR扩散加权成像对直肠癌盆腔淋巴结转移的诊断价值

目的探讨3.0T MR扩散加权成像对直肠癌盆腔淋巴结转移的诊断价值。 方法回顾性分析经手术病理证实的53例直肠癌患者的临床资料、术前常规MRI及扩散加权成像(DWI)资料,以病理结果为参照,将淋巴结分为转移性和非转移性两组,分别测量定量参数ADC值,采用SPSS19.0软件包进行统计学处理,诊断效能采用ROC曲线分析,确定诊断转移性淋巴结的最佳临界点,计算诊断的敏感度、特异度和准确度。以P<0.05认为差异具有统计学意义。 结果116枚淋巴结中,转移性淋巴结47枚(40.5%),非转移性淋巴结69枚(59.5%)。转移性淋巴结的ADC值为(0.81±0.10)×10^－3mm²/s明显低于非转移性淋巴结ADC值(1.06±0.13)×10^－3mm²/s,差异具有显著统计学意义(t＝10.47, P<0.01);ROC曲线分析显示ADC值用于诊断转移性淋巴结的最佳临界点为0.95×10^－3mm²/s,其敏感度为91.5%,特异度为78.3%,准确度为83.6%。 结论3.0T MR扩散加权成像ADC值对直肠癌盆腔淋巴结转移具有较高的鉴别诊断价值,可为患者手术方案的选择和预后的评估提供更多的参考。     

盆腔侧方淋巴结转移对低位直肠癌预后的影响

目的 探讨盆腔侧方淋巴结转移对低位直肠癌预后的影响.方法 对1994年至2005年行根治性切除联合盆腔侧方淋巴结清扫的176例低位直肠癌患者的资料进行回顾性分析.探讨低位直肠癌患者盆腔侧方淋巴结转移对其预后的影响.结果 全组盆腔侧方淋巴结转移33例(18.8%),其中髂内及直肠中动脉根部淋巴结转移占51.5%,闭孔淋巴结转移占39.4%.年龄≤40岁、浸润型癌、T3-4期、上方淋巴结转移患者的盆腔侧方淋巴结转移率较高(P＜0.05).全组5年生存率为64.1%,TNM分期Ⅰ、Ⅱ、Ⅲ期患者5年生存率分别为94.1%、79.1%、42.1%.癌灶大小、浸润深度、上方淋巴结转移、盆腔侧方淋巴结转移是影响低位直肠癌患者预后的重要因素(P＜0.05).盆腔侧方淋巴结阴性患者5年生存率为73.6%,而侧方淋巴结转移患者为21.4%,两组差异具有统计学意义(P＜0.05).结论 盆腔侧方淋巴结转移是影响低位直肠癌预后的重要因素.     

腔内超声对直肠癌术前分期的诊断价值

目的探讨腔内超声(Endoluminal ultrasound,ELUS)对术前直肠癌分期评价的临床应用价值。方法对69例直肠癌患者术前进行ELUS检查进行术前分期诊断,并且与术后病理诊断进行对照。结果术前ELUS检查T1期诊断正确率为93.33%(14/15),T2期为83.33%(15/18),T3期为74.07%(20/27),T4期为66.67%(6/9)。ELUS诊断正确率为79.71%(55/69)。ELUS对转移淋巴结诊断总正确率为65.22%(45/69),灵敏度为76.19%,特异度为61.71%。结论 ELUS对于直肠癌癌组织的浸润程度以及周围转移淋巴结诊断与术后病理诊断的正确率较高,有希望成为直肠癌术前分期的有效方法。     

Accuracy of high-resolution rectal magnetic resonance imaging re-staging with histopathology in locally advanced rectal cancer after neoadjuvant chemoradiotherapy

Background/objectiveRe-staging of locally advanced rectal cancer (LARC) following neoadjuvant chemoradiotherapy (NCRT) is a crucial step in surgical decision-making. Currently, MRI is the imaging of choice for evaluation of LARCs, however, the diagnostic accuracy of this modality is inconsistent. In this study, we evaluated the diagnostic accuracy of MRI in LARC and analyzed the factors that influenced the accuracy.MethodsThe records of 133 patients diagnosed with LARC who were operated on during 2011–2018 were retrospectively reviewed. All patients received NCRT followed by re-staging based on high-resolution rectal MRI. The MRI results were analyzed for their yT and yN accuracy and anal sphincter involvement and compared with the related histopathological studies after definitive surgery.ResultsRe-staging MRIs gave overall accuracy in both the yT stage and yN evaluation of 85% (K 0.45 and 0.21, respectively). The MRI tended to overstaging for tumor invasion and understaging for lymph node involvement (sign test p-values = 0.017 and 0.022, respectively.) The highest accuracy of the yT stage was yT4b (93%, K 0.71). The study found that larger tumors (>3 cm) were associated with significantly higher accuracy in the yT readings while lack of lymphovascular invasion was associated with higher accuracy in the yN readings. The negative predictive value for anal sphincter involvement was 100%.ConclusionMRI has limited accuracy in post-NCRT re-staging in LARC, tending to give overstaged yT readings and understaged yN readings. An MRI exclusion of sphincteric involvement is highly reliable.     

Is Total Mesorectal Excision Always Necessary for T1–T2 Lower Rectal Cancer?

Background

The goal of this multicenter study was to clarify the determinants of local excision for patients with T1–T2 lower rectal cancer.

Methods

Data from 567 consecutive patients who underwent radical resection for T1–T2 lower rectal cancer at 12 institutions between 1991 and 1998 were reviewed. Rates of lymph node metastasis were investigated using a tree analysis, which was hierarchized using independent risk factors for nodal involvement.

Results

The independent risk factors for lymph node metastasis were female gender, depth of tumor invasion, histology other than well-differentiated adenocarcinoma, and lymphatic invasion. According to the first three parameters that can be obtained preoperatively, only 0.99% of the patients without risk factors had lymph node metastasis. On the other hand, even if the lower rectal cancer was at stage T1, women with histological types other than well-differentiated adenocarcinoma had an approximately 30% probability of having lymph node metastasis. Lymphatic invasion was most useful to predict nodal involvement among patients with T2 lower rectal cancer. The rates of lymph node metastasis in T2 patients with and without lymphatic invasion were 32.9% and 9.1%, respectively.

Conclusions

Gender is one of the most important predictors for lymph node metastasis in patients with early distal rectal cancer. Three parameters, including depth of tumor invasion, histology, and gender, are useful determinants for local excision. Additional studies are required to establish the minimum optimal treatment for T2 lower rectal cancer.     

肛提肌标识在直肠癌腹会阴联合切除术中的作用

目的 探讨肛提肌标识在直肠癌腹会阴联合切除术中的作用.方法 回顾性分析2001年1月至2008年1月南京医科大学第一附属医院收治的109例直肠癌患者的临床资料,其中55例采用传统方法手术（传统法组）,54例采用肛提肌标识法进行手术（肛提肌标识法组）.手术遵循直肠癌全系膜切除术原则,锐性分离直肠系膜,整块切除.两组患者术前肠道准备、麻醉选择、患者体位、腹部切口、会阴部切口、会阴部缝合与Miles术相同.传统法组用电刀或超声刀切开会阴部脂肪组织,自尾骨的前方进入盆腔,与腹部手术医师会合,靠近盆壁切断两侧肛提肌未进行标识则进行后续手术操作.肛提肌标识法组采用电刀切开肛门周围间隙脂肪组织,分离两侧坐骨肛管间隙脂肪组织,切断后方肛尾韧带,直达肛提肌平面,标识肛提肌后进行后续手术操作.术后病理检查为Ⅰ期者进行随访观察;术后病理检查为Ⅱ期者,如组织学分化差、T4期、血管淋巴管浸润、检出淋巴结数目＜12枚,则行辅助化疗,如无则进行随访观察;术后病理检查为Ⅲ、Ⅳ期者,行术后化疗.术后第1年,每3个月复查1次血常规、肝肾功能、胸部X线片和肝胆B超.1年后每6个月复查1次上述检查;每年复查1次CT和肠镜检查.随访时间截至2012年12月.计数资料采用x2检验,计量资料采用t检验,Kaplan-Meier法绘制生存曲线,生存率比较采用Log-rank检验.结果 两组患者顺利完成手术,传统法组患者和肛提肌标识法组患者的会阴手术时间分别为（60±15） min和（30±10） min,术中出血量分别为（300 ±60） mL和（30±20） mL,两组比较,差异有统计学意义（t=3.936,5.687,P＜0.05）.传统法组患者中,3例直肠破损,2例尿道（阴道）破损,10例切口感染;而肛提肌标识法组患者中,只有9例切口感染.109例患者中,术后化疗周期少于12个疗程者30例,6个疗程及以上者41例.中位随访时?     

T2期直肠癌淋巴结转移相关影响因素

目的探讨影响T2期直肠癌淋巴结转移的临床病理因素。方法回顾分析福建医科大学附属第一医院2006年3月至2011年1月间行根治性切除的122例T2期直肠癌患者的临床资料,分析影响其淋巴结转移的相关临床病理因素。结果122例T2期直肠癌患者中有26例（21．3％）发生淋巴结转移。单因素分析显示,肿瘤距肛缘距离（P〈0.05）、大体类型（P〈0.05）、组织类型（P〈0．01）、分化程度（P〈0．05）及肿瘤浸润深度（P〈0．05）与T2期直肠癌淋巴结转移有关。多因素分析显示,肿瘤浸润深度是影响T2期直肠癌淋巴结转移的独立因素（P〈0.05）;直肠癌浸润浅肌层和深肌层者淋巴结转移率分别为13．0％（7／54）和27．9％（19／68）。结论对于局限于浅肌层的L期直肠癌,因其淋巴结转移率较低,可考虑行经肛局部切除手术。     

Feasibility and functional outcome of laparoscopic intersphincteric rectal resection for ultra-low rectal cancer

Aim-Background

The Aim of this study is to evaluate the feasibility and analyze the functional outcome of laparoscopic intersphincteric resection (LISR) in ultra-low rectal cancer. The preservation of anal function following curative operations for low rectal cancer is becoming increasingly important. Laparoscopic intersphincteric resection of the rectum is the utmost sphincter saving operation for rectal cancer. The rectum is laparoscopically resected along with the internal anal sphincter, providing an adequate distal margin for even the ultra-low tumours of the rectum.

Methods

Between 2008 and 2012, nine patients, 2 with a T3 tumour that received preoperative chemoradiotherapy and 7 patients with a non-fixed T2 rectal adenocarcinoma, underwent LISR by a single surgeon. Preoperative tumour staging included endorectal ultrasonography (ERUS) and pelvic MRI. Patients with multiple distant metastases, tumour invasion into adjacent organs and invasion into the external anal sphincter and/or levator ani, were excluded from LISR. Covering ileostomy in seven patients was reversed with a satisfactory functional outcome in each case.

Results

All patients underwent LISR with curative intent. There was no postoperative mortality. Complications included anal stenosis, prolapse of the neorectum and pelvic hematoma. The overall quality of life and functional outcome were deemed satisfactory.

Conclusion

In selected patients, intersphincteric rectal resection may provide an acceptable functional outcome for ultra-low rectal cancer patients, without a permanent stoma.     

Lateral pelvic node dissection after neoadjuvant chemoradiation in rectal cancer with right internal iliac node metastasis but without regional node metastasis: A case report

Introduction and importanceTotal mesorectal excision (TME) with lateral pelvic node dissection was routinely done in low clinical T3 rectal tumors below the peritoneal reflection as stated in the Japanese guidelines for colorectal cancer. Our institution follows the same practice in selected patients. This is our first reported case wherein a patient with rectal cancer underwent total mesorectal excision with lateral lymphadenectomy after neoadjuvant treatment with a positive lateral node on histopathology.Case presentationA 49 year old female rectal had rectal adenocarcinoma 4 cm FAV. Pelvic MRI revealed a low rectal tumor abutting the mesorectal fascia anteriorly, anal sphincters not involved, and confluent enlarged right iliac nodes. After neoadjuvant treatment, interval decrease in size of the rectal lesion and the right iliac nodes were noted. Patient underwent partial intersphincteric resection, lateral pelvic node dissection and protective loop ileostomy.Clinical discussionHistopathology revealed a rectal adenocarcinoma with one right internal iliac lymph node was positive for tumor involvement. Circumferential resection margin was 4.0 mm. Patient is currently on 4th cycle of adjuvant chemotherapy. Preoperative chemoradiation could not completely eradicate lateral pelvic node metastasis. Therefore, lateral pelvic node dissection should be considered if lateral pelvic lymph node metastasis is suspected even after neoadjuvant therapy.ConclusionUnlike TME, performance of a routine lateral lymphadenectomy in rectal cancer surgery varies by geographic location. Reports from Asian countries and our practice in our institution shows that it can be performed safely. This could improve the oncologic outcomes of patients especially if combined with neoadjuvant chemoradiotherapy.     

Distribution of Metastatic Cancer Cells in Colorectal Mesentery

Background

To investigate the distribution of metastatic cancer cells in the mesentery (referred to as metastasis V) and enrich the understanding of the metastasis of colorectal cancer.

Methods

A total of two hundred ninety-nine patients who received colorectal operations at the Department of Gastrointestinal Surgery, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology between April 2018 and December 2018 were included. Samples were acquired from the dissected mesentery after the operation, and hematoxylin–eosin staining or immunohistochemistry was used to detect metastatic cancer cells. Pathological factors, including tumor position, tumor size, invasion depth, tumor differentiation, lymph node involvement, local vessel invasion, and perineural invasion, were recorded.

Results

Metastatic cancer cells in the colorectal mesentery (metastasis V) were detected in 62 of 299 patients. Metastasis V was closely correlated with tumor invasion depth, lymph node metastasis, tumor differentiation, and perineural and vessel invasion by cancer cells. Metastasis V occurred more frequently in patients with T3 stage (26.27%) and T4 stage (40.00%) than in patients with T1 and T2 stages (0% and 2%, respectively). Metastasis V was frequently detected in patients with N2a and N2b stage tumors (51.72% and 61.54%, respectively). Metastasis V was more frequently detected in patients with perineural metastasis and local vessel invasion. In addition, metastasis V incidences in colon and rectal cancer were similar.

Conclusion

The incidence rate of metastasis V is correlated with tumor staging factors and occurs more frequently in advanced-stage patients.

     

Magnetic resonance imaging of the low rectum: defining the radiological anatomy

Low rectal cancer provides a particular surgical challenge of local tumour control and sphincter preservation. Histopathological studies have shown that an involved circumferential resection margin (CRM) and depth of extramural invasion are independent markers of poor prognosis and correlate with high local recurrence rates due to residual microscopic disease [ 1 ]. Recent data suggests that a CRM at risk of tumour involvement can be reliably seen on the pre‐operative magnetic resonance imaging (MRI) scan with good correlation with the histological specimen [ 2 - 5 ]. In published series, low rectal cancers have a higher incidence of involved resection margins, with rates up to 30% for abdomino‐perineal excision (APE) vs 10% for low anterior resection (LAR) [ 6 - 9 ]. This has been attributed to narrow surgical planes deep within the pelvis as the mesorectum becomes narrowed and tapered, forming a bare muscle tube at the level of the anal sphincter complex. The challenge for the surgeon is to undertake careful removal of a cylinder of tissue beyond the rectal wall without perforating the tumour. An overall local recurrence rate of 10% after APE for all stages of rectal cancer has been reported and this low rate was attributed to the surgical technique that included a wide peri‐anal dissection and lateral division of the levator ani. The abdominal dissection was stopped above the tumour, taking care to avoid separation of the tumour from the levator ani to reduce the risk of inadvertent tumour cell spillage [ 8 ]. Therefore, rates of involved surgical margins from APE specimens may be reduced when a cuff of levators is taken compared with standard resection. In this review, we will discuss how MRI of the low rectum can aid in the staging and optimization of the best treatment strategy for low rectal cancer.     

Residents entering Complex General Surgical Oncology fellowship lack confidence with pelvic MRI for rectal cancer: Results of a needs assessment survey

BackgroundMagnetic resonance imaging (MRI) is essential for the multidisciplinary treatment of rectal cancer. However, baseline experience of surgical residents with MRI is unknown. Therefore, a needs assessment survey was conducted to examine confidence with pelvic MRI for residents entering Complex General Surgical Oncology (CSGO) fellowships.MethodsA multi-institutional survey evaluated incoming CGSO fellows’ experience with pelvic MRI for rectal cancer in residency. Additionally, confidence was assessed for essential components of pelvic MRI including T- and N-stage, circumferential resection margin (CRM), extramural venous invasion (EMVI), and pelvic anatomy.ResultsOf the twenty-four incoming fellows who completed the survey (response rate = 44%), 20 reported frequent use of pelvic MRI for rectal cancer in residency, but 16 reported rarely/never interpreting images themselves for staging or operative planning. Most respondents reported low confidence for T-stage, N-stage, CRM, EMVI, as well as pelvic anatomy, particularly for lateral and posterior pelvis.ConclusionsThe development of a pelvic MRI curriculum for residents entering CGSO fellowships could enhance their clinical training in the multidisciplinary management of patients with rectal cancer.     

Prediction of Lymph Node Status in Superficial Esophageal Carcinoma

Background  Esophageal carcinoma is among the cancers with the worst prognosis. Real chances for cure depend on both early recognition and early treatment. The ability to predict lymph node involvement allows early curative treatment with less invasive approaches. Aims  To determine clinicohistopathological criteria correlated with lymph node involvement in patients with early esophageal cancer (T1) and to identify the best candidate patients for local endoscopic or less invasive surgical treatments. Methods  A total of 98 patients with pT1 esophageal cancer [67 with squamous cell carcinomas (SCC) and 31 with adenocarcinomas (ADK)] underwent Ivor–Lewis or McKeown esophagectomy in the period between 1980 and 2006 at our institution. Based on the depth of invasion, lesions were classified as m1, m2, or m3 if mucosal, and sm1, sm2, or sm3 if submucosal. Results  The rates of lymph node metastasis were 0% for the 27 mucosal carcinomas (T1m) and 28% for the 71 submucosal (T1sm) carcinomas (P < 0.001). Sm1 carcinomas were associated with a lower rate of lymph-node metastasis (8.3% versus 49 % sm2/3, P  = 0.003). As for histotype, the rates of lymph node metastasis for sm1 were 0% for ADK and 12.5% for SCC; for sm2/3 there were no significant differences. On multivariate analysis, depth of infiltration, lymphocytic infiltrate, angiolymphatic and neural invasion were significantly associated with lymph node involvement. Neural invasion was the single parameter with the greatest accuracy (82%); depth of infiltration and angiolymphatic invasion had 75% accuracy. Altogether these three parameters had an accuracy of 97%. Five-year survival rate was 56.7% overall: 77.7% for T1m and 53.3% for T1sm (P  = 0.048). Conclusions  The most important factors for predicting lymph node metastasis in early esophageal cancer are depth of tumor infiltration, angiolymphatic invasion, neural invasion and grade of lymphocytic infiltration. The best candidates for endoscopic therapy are tumors with high-grade lymphocytic infiltration, no angiolymphatic or neural invasion, mucosal infiltration or sm1 (only for ADK), and tumor <1 cm in size. For sm SCC and sm2/3 ADK the treatment of choice remains esophagectomy with standard lymphadenectomy.