The Value of Lymphadenectomy Post-Neoadjuvant Therapy in Carcinoma Esophagus: a Review

Abstract

Lymph nodal metastasis is one of the most important prognostic factors determining survival in patients with carcinoma esophagus. Radical esophagectomy, with the resection of surrounding lymph nodes, is considered the prime treatment of carcinoma esophagus. An extensive lymphadenectomy improves the accuracy of staging and betters locoregional control, but its effect on survival is still not apparent and carries the disadvantage of increased morbidity. The extent of lymphadenectomy during esophagectomy also remains debatable, with many studies revealing contradictory results, especially in the era of neoadjuvant therapy. The pattern of distribution and the number of nodal metastasis are modified by neoadjuvant therapy. The paper reviews the existing evidence to determine whether increased lymph node yield improves oncological outcomes in patients undergoing esophagectomy with particular attention to those patients receiving neoadjuvant therapy.

Introduction

Esophageal cancer is the eighth most common cancer with an annual incidence of 572,000 new cases, constituting 3.2% of all cases globally and the seventh leading cause of cancer-related death with 500,000 death annually [1]. Esophageal cancer is an aggressive cancer with early lymphatic dissemination to the neck, chest, and abdomen. Traditionally, the prognosis of patients with esophageal cancer has remained dismal, especially in the presence of lymph node metastases. Esophagectomy is considered a prime modality of treatment but carries high morbidity and mortality. However, operative mortality for esophagectomy has significantly declined to less than 5% in high-volume centers [2–4], owing to better postoperative care, better surgical skills, usage of minimally invasive surgery, and the introduction of neoadjuvant therapy. With declining operative mortality, the primary focus has been switched to prolonging long-term survival. En bloc resection principles have been applied to esophageal surgery for prolonging survival [2]. Since the extent of lymphadenectomy and removal of the primary tumor is a factor modifiable by the surgeon, the emphasis of treatment is tilted towards exploring the role of extended lymphadenectomy. Nonetheless, the efficacy of en bloc resection and the extended nodal dissection is yet to be proved [5].

One of the prognostic parameters determining survival in carcinoma esophagus is the nodal involvement. Lymphadenectomy improves the accuracy of staging and betters locoregional control, but its effect on survival is still not apparent. For limited nodal disease, an extensive lymphadenectomy potentially increases the chance of an R0 resection, more accurate disease staging, better locoregional disease control, and perhaps even may improve long-term survival [2, 6, 7]. For staging purposes, an extended lymphadenectomy is superior to a limited nodal dissection. However, the benefit of survival looks less likely with extensive nodal involvement, which increases the risk of systemic dissemination. Surgical series comparing more extensive with limited lymph node dissection reported higher survival in the radical surgery group than in limited lymphadenectomy in lesser nodal disease [2, 6, 7]. However, three major meta-analyses failed to show any survival benefit of transthoracic esophagectomy (TTE) with extended lymphadenectomy over transhiatal esophagectomy (THE) with limited nodal dissection [8–10]. In all these meta-analyses comparing TTE and THE, transthoracic resection was mostly Ivor-Lewis approach, in which case extended lymphadenectomy in the upper mediastinum may not be possible making interpretation of the data difficult.

The spread and distribution of nodal dissemination depends on histology, tumor location, depth of invasion, and neoadjuvant therapy. The neoadjuvant treatment potentially modifies the distribution pattern of nodal metastases [11]. Neoadjuvant chemoradiotherapy (nCRT) has become a standard of care for esophageal cancer patients globally after the CROSS trial [12]. As neoadjuvant therapy is known to frequently “sterilize” regional nodes, the therapeutic and prognostic implications of an extensive lymphadenectomy are still not clear, with several studies publishing the conflicting results. The present review attempts to examine the existing evidence to know if the increasing nodal harvest has a therapeutic and prognostic bearing on patients with esophageal cancer undergoing esophagectomy, with a specific focus on neoadjuvant therapy settings.

Definition, Types, and Extent of Lymphadenectomy

The 1994 Munich Consensus Conference of the International Society for Diseases of the Esophagus (ISDE) defined the terms and types of lymphadenectomy [13]. The regions of lymph nodal dissection during esophagectomy are stratified into three fields, the abdominal (field-I), the thoracic (field-II), and the cervical (field-III) [13, 14].

Standard lymphadenectomy incorporates the entire thoracic esophagus, paraesophageal, subcarinal, and right and left parabronchial lymph nodes without lymphadenectomy of upper mediastinum. Extended lymphadenectomy encompasses the standard lymphadenectomy template with nodal dissection of the upper mediastinum on the right side (right-sided paratracheal, recurrent nerve, and apical nodes). Total lymphadenectomy comprises the standard lymphadenectomy with lymphadenectomy in the bilateral upper mediastinum (bilateral paratracheal, recurrent nerve, and apical nodes). Three-field lymphadenectomy combines nodal dissection in field-I and total lymphadenectomy in field-II and field-III. Field-III includes cervical lymphadenectomy. It refers to clearance in the omohyoid triangle with sparing of the jugular vein and sternocleidomastoid muscle, caudally till the superior margin of the clavicle and cranially extending till cricoid cartilage. Field-I (abdominal field) is also defined by anatomical boundaries. Hiatus forms the cephalad border while the superior border of the pancreas caudal. The right border is delineated by the hepatoduodenal ligament and origin of the right gastric artery from the common hepatic artery, while the left margin is formed by splenic hilum. The anterior aspect of the aorta constitutes the posterior limit [13].

Prevalence of Nodal Involvement in Esophageal Cancer

Lymph nodal dissemination occurs at the outset in the course of esophageal cancer owing to the widespread lymphatic drainage system in the submucosa. The nodal involvement proportionately increases with the depth of tumor infiltration. Metastases to lymph nodes are found in up to 6.5% of T1a tumors, 35–50% of T1b tumors, 43–80% of T2 tumors, and 78–85% of T3 tumors [6, 7, 15, 16]. Lymph node involvement may occur in the cervical, thoracic, or abdominal nodes irrespective of whether the primary tumor is located at the upper, mid, distal esophagus, or gastro-esophageal junction [5]. Deposits in neck nodes are detected in about 23% of the third distal and 17% of gastro-esophageal junction tumors [4, 6, 16].

T2 tumors had celiac nodal and distant nodal metastasis in 23 and 31% of cases [7]. Nodal metastasis occurred in 85% of T3 tumors, of which 45% had more than four nodal involvement with a prevalence of distal nodal metastasis in 40% and celiac nodal involvement in 27% of the patients. Up to 40% of the subjects have distant nodal involvement (neck nodes for lower third and celiac nodes for third upper tumors) [6, 7]. So, as the depth of tumor infiltration increases, the nodal dissemination grows both in the number of nodes involved and in distant spread too [6, 7, 10, 16].

The Extent of Lymphadenectomy and Accuracy of Staging

Transthoracic esophagectomy (TTE) with a more comprehensive nodal dissection provides better insights into the lymphatic spread of tumor cells. The chance of detecting a positive tumor node increases with dissecting more lymph nodes, which has a bearing on the pathological staging of the tumor (stage migration). The higher nodal yield was seen after transthoracic approach, in studies comparing transthoracic and the transhiatal approach. Superior prognosis with good survival was found with extended lymphadenectomy and examination of more nodes than with limited lymphadenectomy in node-negative patients.

Kutup et al. noted a higher median lymph node yield with transthoracic than by transhiatal approach (27.0 vs 17.0) [17]. HIVEX, a randomized controlled trial comparing the effectiveness of a limited transhiatal resection against an extended transthoracic resection in patients with adenocarcinoma of the gastro-esophageal junction and lower esophagus, reported an average harvest of 31 nodes post-transthoracic resection and 16 nodes after transhiatal resection [18]. Bollschweiler et al. noted that node-negative patients and those where more than 15 nodes examined had a significantly better prognosis than those with fewer nodes examined [19]. Likewise, Kang et al. also noted a higher nodal involvement with an increased number of nodal stations and fields dissected and examined. Nodal positivity was seen in 53% of patients undergoing a three-field dissection with a median of 33 nodes excised, as against 33% having nodal dissemination in patients having a 1-field dissection with a median of 17 nodes dissected [20].

Relationship Between the Extent of Lymphadenectomy and Prognosis

Several studies had observed an improved survival with more extensive surgery [6, 7, 15]. In Japan, 3-field lymphadenectomy has been standard practice since the 1980s. Nishimaki et al. reported a 5-year survival of 68% after 3-field lymphadenectomy [21]. The seminal paper of Akiyama and colleagues cited an occurrence of cervical nodal metastases of 46%, 29%, and 27% for primary squamous cell carcinomas of the upper, middle, and lower thoracic esophagus, respectively [22]. The reported 5-year survival difference of 55% for 3-field lymphadenectomy versus 38% for 2-field lymphadenectomy (P = 0.0013) was attributed to lymphadenectomy extent.

Similarly, a small randomized trial by Nishihira et al. comparing 3-field lymphadenectomy with cervical and upper mediastinal lymph node dissection versus standard 2-field nodal dissection for squamous cell carcinoma of the thoracic esophagus reported an overall survival of 66% at 5 years for the extended dissection with a mean harvest of 82 nodes versus 48% for standard dissection with a harvest of 43 nodes. However, the survival difference at 5 years was not significant [23]. Isono et al., in a classical study of 3-field nodal dissection, noted the incidence of lymph nodal metastases of 45.8% in the neck, 37.5% in the mediastinum, and 16.7% in the abdomen for upper esophageal cancer. When the tumor was located in the abdominal esophagus, the rate of cervical nodal metastasis was 33.3%; hence, they recommended 3-field dissection for all subsites [24].

Kato et al. divided 150 patients treated for thoracic esophageal cancer into two matched groups, with one group receiving 3-field and another 2-field lymphadenectomy [25]. The 3-field patients in comparison with the 2-field group had a better 3-year survival (48% versus 34%; P < 0.01). The extent of improved lymph node staging was also observed. In the 3-field group, 49 patients were node-positive, a mean of 69 nodes were resected and 3.7 nodes were positive; in 2-field patients, these numbers were 48, 36.4, and 3.2, respectively. The Japanese Society for Esophageal Diseases conveyed a nationwide survey, in which 1791 patients had 3-field lymphadenectomy and 2799 2-field. Survival analysis indicated a significant 5-year survival benefit of 34% in the 3-field group versus 27% in the 2-field group respectively. However, this survival difference was not noted in early- or advanced-stage cases [26].

In another series of 3-field lymph nodal dissection done for carcinoma esophagus, patients with metastases to more than five nodes; simultaneous cervical, mediastinal, and abdominal nodal dissemination; or cervical nodal involvement from lower thoracic esophageal tumors did not have a survival advantage from 3-field lymphadenectomy [21]. In a study examining the effect of 1-, 2-, or 3-field lymphadenectomy, increased survival was reported with an increasing extent of lymph node resection. Five-year survivals in patients undergoing a 1-, 2-, or 3-field lymphadenectomy were 21.2%, 36.3%, and 53.7% (P = 0.019). Nonetheless, the benefit of extended nodal clearance was observed mostly in the node-negative group [20], suggesting that the benefit is probably related to stage migration.

Fujita et al. evaluated the outcomes of 4 different types of nodal dissections in patients with squamous cell carcinoma of thoracic esophagus and found no significant difference in 5-year survival [14]. Nevertheless, the usefulness of 3-field nodal dissection with more prolonged survival was observed only in patients with node-positive upper or middle thoracic esophagus cancer. In patients with lower thoracic esophagus cancer or node-negative disease, no difference in survival was observed among the four different types of nodal dissections.

The purported therapeutic advantage of 3-field nodal dissection may be as a result of improved staging, surgeon experience, and patient selection. No adequately powered randomized trials till to date have been conducted to provide high-level evidence for performing 3-field lymphadenectomy (see Table 1).

Table 1.

Studies comparing more extensive nodal dissection (3-field or TTE) with less extensive nodal dissection (2-field or THE) in primary setting without neoadjuvant therapy

Authors, year	Comparison	Study design	Histology	n	Sites	5-year OS	P value	5-year DFS	LRR	R0 resection	Neck node positivity
Kato et al. 1991 [25]	3 FLND 2 FLND	RCT	SCC	77 73	All	49% 34%	< 0.01				26%
Nishihira et al. 1998 [23]	3 FLND 2 FLND	RCT	SCC	32 30	M, L	66% 48%	NS	20% 24%
Akiyama et al. 1994 [22]	3 FLND 2 FLND	Retrospective	SCC	398 324	All	55% 38%	0.0013			78% (in entire series)	27–46%
Isono et al. 1994 [24]	3 FLND 2 FLND	Retrospective	SCC	174 99	All	44% 26%	0.001				33%
Japanese national wide survey 1991 [26]	3 FLND 2 FLND	Retrospective	SCC	1791 2799	All	34% 26.7%	< 0.001				26%
Kang et al. 2007 [20]	1-field 2-field 3-field	Retrospective	SCC	67 102 64	All	21%, 36% 53.7%	0.001		52% 19% 38%	94.8% (in entire series)
Fujita et al. 2003 [14]	3 FLND Total ND Extended ND Standard ND	Retrospective	SCC	176 65 21 40	All	49% 48% 38% 45%	NS		23% 26% 33% 23%	75% (in entire series)	24%
Goldminc et al. 1993 [27]	TTE THE	RCT	SCC	35 32	M, L	< 20% < 40% (3-year OS)	NS
Chu et al. 1997 [28]	TTE THE	RCT	SCC	19 20	All	13.5 vs 16 months (Median OS)	NS
Omloo et al. 2007 [29]	TTE THE	RCT	A	110 95	Type1 Type2	36% 34%	NS	64% 23%	25% 45%	72% 72%
Boshier et al. 2011 [8]	TTE THE	Metanalysis	Both	3389 2516	All	26.6% 25.8%	NS
Hulscher et al. 2001 [10]	TTE THE	Metanalysis	Both	2677 2264	All	23% 21.7%	NS
Rindani et al. 1999 [9]	TTE THE	Metanalysis	Both	2808 2675	All	26% 24%	NS

RCT randomized controlled trial, 3FLND 3-field nodal dissection, 2FLND 2 field nodal dissection, ND nodal dissection, Type 1 Siewart type tumor, Type I- Siewart type tumor, THE transhiatal esophagectomy, TTE transthoracic esophagectomy. A adenocarcinoma, SCC squamous cell carcinoma, M middle third esophageal tumor, L lower third esophageal tumor, NS not significant, OS overall survival, DFS disease-free survival, LRR locoregional relapse. Note the three fields of Kang et al. corresponds to nodal stations in paraesophageal, upper thoracic, and abdominal regions and does not include neck nodes

Transhiatal Versus Transthoracic Resection

Transhiatal esophagectomy with a limited nodal clearance is one extreme of lymphadenectomy. A formal lymphadenectomy is not carried out, and nodal sampling is restricted to a limited area, i.e., abdominal region. However, three randomized trials failed to demonstrate a clear superiority of transthoracic resection with extended lymphadenectomy over a limited transhiatal resection [18, 26, 28]. In the trial by Hulscher et al., there was no significant survival benefit for either of the approach [18]. Furthermore, there was a trend towards better 5-year survival in extended transthoracic esophagectomy compared with limited transhiatal resection for type I esophageal adenocarcinoma [29].

Furthermore, patients with a limited positive lymph node seem to derive an advantage from an extended transthoracic esophagectomy [29]. Again, the improved staging of pN1 and pN2 patients can clarify this effect ascribed to therapy. Three major meta-analyses comparing THE against TTE failed to demonstrate any survival benefit of transthoracic resection with extended lymphadenectomy over limited transhiatal resection [8–10]. The caution must be exercised while interpreting the results of these studies since the majority of transthoracic resections were by the Ivor-Lewis approach where upper mediastinal nodal dissection is difficult and may not have been carried out.

In a matched comparative study, Johansson et al. noted en bloc esophagectomy confers a better survival than THE in patients with T3N1 disease and fewer than nine lymph node metastases [30]. However, with the increasing numbers of lymph node positivity, the probability of systemic dissemination is increased, and consequently, more radical and aggressive surgery is unlikely to add any survival advantage [31].

Lymph Node Ratio

Lymph node ratio, the ratio of involved to total lymph node, is a better prognostic parameter in esophageal carcinoma. The data reveals only lower values of LNR value are prognostic, for which more lymph nodes need to be examined. However, if inadequate nodes are surveyed, the LNR halts to be a prognostic marker [5]. If 15 or more nodes are excised, LNR serves as a powerful predictor of survival, with LNR > 0.3 associated with the worst survival; however, if < 15 nodes are resected, then only the nodal metastases but not the LNR predict survival [5].

Bhamidipati et al. showed that patients with LNR greater than 0.2 have significantly poorer DFS; besides, LNR was also recognized as an independent prognostic parameter for both DFS and OS in esophageal cancer [32]. Also, LNR was shown to be a better prognostic marker than the N stage for esophageal cancer [33]. Chen et al. have suggested a staging system based on LNR may have better prognostic stratification of patients with esophageal squamous cell carcinoma (ESCC) than the current TNM system, especially for those undergoing limited lymphadenectomy [34]. LNR has also been shown to be an independent prognostic factor for predicting postoperative distant metastasis and prognosis in ESCC [35].

Role of Neoadjuvant Chemoradiation

Two large trials have established the role of neoadjuvant chemoradiation in carcinoma esophagus and gastro-esophageal junction. CROSS trial evaluated 368 patients with clinically resectable, locally advanced cancer of the esophagus or esophagogastric junction were randomly assigned to receive either weekly administration of 5 cycles of intravenous carboplatin and paclitaxel with concurrent radiotherapy (41.4 Gy, given in 23 fractions of 1.8 Gy on 5 days per week) followed by surgery or surgery alone. Seventy-five percent of the patients had adenocarcinoma and 23% squamous. After a median follow-up period of 84 months, median OS was 48.6 months in the nCRT plus surgery arm and 24.0 months in the surgery alone arm (HR 0·68 [95% CI 0·53–0·88]; P = 0.003). Thus, confirming the long-term oncological benefits in favor of nCRT plus surgery arm [12, 36].

NEOCRTEC trial, a large multicenter randomized controlled trial from China, evaluated nCRT plus surgery with surgery alone in 451 patients with locally advanced resectable esophageal squamous cell carcinoma. nCRT with 2 cycles of cisplatin and vinorelbine administered concomitantly with 40 Gy of radiotherapy was well tolerated. nCRT arm had a higher R0 resection rate (98.4% vs 91.2%; P = 0.002), a prolonged disease-free survival (100.1 months vs 41.7 months; hazard ratio, 0.58; 95% CI, 0.43 to 0.78; P < 0.001), and a better median overall survival (100.1 months vs 66.5 months; hazard ratio, 0.71; 95% CI, 0.53 to 0.96; P = 0.025) [37].

Low Nodal Yield After Neoadjuvant Therapy

A large nationwide Swedish study addressed the query of nodal yield after neoadjuvant therapy. The results of this study suggest that any neoadjuvant therapy (including neoadjuvant chemoradiotherapy) effectively downsizes the tumor and nodal size and reduces the lymph node yield, number of removed and examined metastatic lymph nodes, and also the proportion of tumor-involved resection margins [38]. The detection of smaller lymph nodes in the specimen is challenging to the pathologist, who also plays an essential role in the final lymph node count beside a surgeon. The shrinking of the primary tumor and nodes appears to mediate this effect [38].

The current literature on this subject is sparse. Randomized clinical trials on neoadjuvant therapy have shown to increase the resectability with effectual downsizing of the tumor but did not disclose data on lymph node yield. Two post hoc analyses of Dutch and French randomized clinical trials suggested a reduced lymph node yield after neoadjuvant therapy compared with surgery alone for esophageal cancer [39, 40].

Increasing Nodal Yield

Immunohistochemical analysis for cytokeratin can detect micrometastatic involvement of lymph nodes that are missed on routine pathologic examination. In a study examining nodal biopsy obtained by thoracoscopic/laparoscopic pre-treatment staging for esophageal cancer, immunohistochemical analysis (IHC) using cytokeratin (CK) of AE1/AE3 identified micrometastases in 3.9% of 280 histologically negative lymph nodes, and another 4.3% were upstaged to N1 from previous N0 status [41]. IHC analysis for apoptosis marker, p53, has also been studied.

IHC staining techniques can identify nodal metastases missed by routine Hemotoxylin and eosin (H&E) examination in a significant number of patients. The extent of nodal burden identified by both IHC and H&E analysis can more precisely predict survival than either technique alone [42]. Regular H&E pathological examination can underscore esophageal cancer in these cases. Identifying CK-positive cells in resected nodes is an independent prognostic factor in otherwise node-negative esophageal cancer [43]. Further studies are required to elucidate the role of IHC in the neoadjuvant setting.

Prognostic Markers

Lymphovascular invasion (LVI), extracapsular invasion (ECI), tumor regression grade (TRG), and perineural invasion (PNI) are considered robust prognostic parameters for determining distant recurrence and could be utilized for detecting high-risk individuals with node-negative esophageal carcinoma and an indicator for sub-stratifying in further stage classification. The presence of perineural infiltration and lymphovascular invasion post-neoadjuvant therapy in esophagectomy specimens identifies patients at high risk for recurrence [44].

Tumor deposits are isolated tumor deposits found in the peri-adventitial or perinodal connective tissue, away from the tumor’s leading edge and with no evidence of residual nodal tissue. Patients with tumor deposits are associated with more advanced esophageal cancer and have a poor prognosis in comparison with those without tumor deposits [45]. Tumor deposits were identified in 32% of node-negative patients in a study of more than 1000 patients undergoing curative esophageal resection for esophageal cancer. It might be more reasonable to be regarded as an indicator of stage migration in esophageal cancer [45].

Effect of Neoadjuvant Therapy on Lymphadenectomy

The number of nodes excised is a known predictor of survival post esophagectomy in patients who did not undergo neoadjuvant treatment. Resection of a minimum of 23 regional lymph nodes has been recommended to maximize this survival benefit [30]. The effect of neoadjuvant therapy on dissemination and the pattern of lymph node distribution is a matter of debate. Neoadjuvant chemoradiotherapy influences both the number and distribution pattern of nodal metastasis [11]. The effect of chemoradiotherapy on lymph nodes may differ between the inside and outside radiation field. Besides, neoadjuvant systemic chemotherapy alone may also affect locoregional metastatic lymph nodes.

Neoadjuvant chemoradiation treat involved lymph nodes. Lymph nodes inside the radiation field are sterilized by radiotherapy and concurrent chemotherapy, whereas nodes beyond the field of radiation are only affected by chemotherapy. Fibrosis or sterile nodes in previously affected lymph nodes are seen on pathology after successful chemoradiation or systemic chemotherapy. Tumor location and depth of invasion decide the extent of the neoadjuvant radiation field [46]. For the bulkier, deeply infiltrative, longer tumors, as well as poorly differentiated squamous cell carcinoma, the radiation field should be wide enough to encompass the region at high risk of nodal metastasis [46].

In a study of 402 patients with esophageal and esophagogastric junctional carcinoma undergoing nCRT, Castoro et al. found not only the frequency of nodal metastasis decreased but also the nodal localization and patterns were modified significantly with neoadjuvant therapy [11]. There were lesser paracardial nodal metastases in patients with adenocarcinoma who underwent neoadjuvant chemoradiotherapy. Similarly, significantly lesser paraesophageal, paracardial, and subcarinal node metastases were found in patients with squamous cell carcinoma (SCC). These data question the necessity for maximal lymphadenectomy after nCRT (see Table 2).

Table 2.

Studies not supporting the role of extended lymphadenectomy in post-neoadjuvant therapy settings

Author, year	n	Histology	Study design	Conclusion
Noordman et al. 2018 [49]	701 (318 treated by TTE, 383 by THE)	Adenocarcinoma	Post hoc evaluation of patients from CROSS 1 and 2 trials (phase 2 and 3 trials)	TTE over THE had favorable prognostic effect in patients who were treated with surgery alone, but absent in the nCRT and surgery group
Robb et al. 2015 [40]	195 (97 treated by nCRT + surgery, 98 by surgery)	Both subtypes	Post hoc analysis of FFCD 9901	nCRT resulted in tumoral and nodal downstaging and reduction of proportion of resected (NLNr) and involved nodes (NLNi). After adjustment by treatment, NLNi correlated with prognosis, whereas NLNr did not
Talsma et al. 2014 [39]	368 (180 by nCRT + surgery, 188 by surgery alone)	Both subtypes	Post hoc analysis of CROSS trial	Number of resected nodes were significantly associated with survival for patients in the surgery-alone arm, but not in the nCRT arm
Shridhar et al. 2013 [47]	358 treated by nCRT	Both subtypes (predominantly adenocarcinoma)	Single institutional retrospective study	The number of nodes removed after nCRT did not impact OS or DFS
Vallbohmer et al. 2010 [48]	299 (284 treated by nCRT, 15 by nCT)	Both subtypes	Multicentric retrospective study	Cox regression analysis identified age as the only independent predictor of survival, whereas gender, histology, type of esophagectomy, type of neoadjuvant therapy, and the number of resected lymph nodes had no prognostic impact
Markar et al. 2017 [50]	608 (301 by nCRT + surgery, 307 by nCT+ Surgery)	Adenocarcinoma	Multicentric retrospective study	In nCRT group lymph node harvest did not affect survival or recurrence by RA-CUSUM analysis. Node harvest significantly influenced survival and recurrence in the nCT group

nCRT neoadjuvant chemoradiation, nCT neoadjuvant chemotherapy, THE transhiatal esophagectomy, TTE transthoracic esophagectomy, OS overall survival, DFS disease-free survival

The post hoc analysis of 2 randomized controlled trials Dutch CROSS trial and French FFCD 9901 provided a critical observation that the number of resected nodes has prognostic implications on survival in patients after surgery alone, but loses its significance in patients who receive neoadjuvant chemoradiotherapy [39, 40]. Not only does the number of dissected and positive nodes decrease after nCRT, but also, the upstaging effect of the number of resected nodes on the number of positive nodes disappears. Several retrospective studies also corroborated this finding.

Shridhar et al., in a series of 365 patients with carcinoma esophagus, found the number of nodes resected did not affect DFS or overall survival in patients who underwent neoadjuvant chemoradiotherapy. Pathologic responses and stage were found to be the powerful determinant of survival after nCRT in esophageal cancer [47]. A multicenter German study evaluated the survival of patients with locally advanced esophageal cancer who have undergone multimodality treatment with a complete histopathologic response. Age was identified as the only independent predictor of survival, whereas histology, the number of resected nodes, and type of neoadjuvant therapy did not influence survival [48].

A multicentric Dutch study studied the effect of surgical approach on long-term survival in patients with esophageal adenocarcinoma with or without neoadjuvant chemoradiotherapy [49]. Of 701 patients analyzed, 318 underwent TTE with extended lymphadenectomy and 383 THE with limited lymphadenectomy. It was noted that the addition of nCRT to surgery decreases the necessity for TTE with extended lymphadenectomy to improve long-term survival in esophageal adenocarcinoma patients. TTE with radical lymphadenectomy had better prognostic outcomes in patients who were treated with surgery alone (hazard ratio (HR) 0.77, 95% confidence interval (CI) 0.58–1.03). This favorable prognostic effect of TTE over THE was lost with the addition of nCRT to surgery (HR 1.16, 95% CI 0.80–1.66) and in node-positive patients.

A multicenter propensity score–matched study evaluated the data of patients undergoing multimodality treatment for esophageal adenocarcinoma from 10 European centers with prospectively maintained databases from 2001 to 2012 [50]. Risk-adjusted curve analysis (RA-CUSUM) was used to determine a lymph node harvest threshold that affected overall survival in each of neoadjuvant chemoradiotherapy followed by surgery (NCRS) and neoadjuvant chemotherapy followed by surgery (NCS) groups. In the NCRS group, RA-CUSUM analysis showed that lymph node harvest did not affect survival or recurrence with no identifiable change-point in the RA-CUSUM curve. Node harvest significantly influenced survival and recurrence in the NCS group. The mean change-point in overall survival was observed to be situated between 22 and 52 lymph nodes. Beyond the nodal yield threshold of 52 lymph nodes, there were no meaningful improvements in disease-free survival and overall recurrence [50].

Studies Supporting the Role of Extended Lymphadenectomy in Neoadjuvant Therapy Settings

Furthermore, studies have shown that high lymph node yields are possible after transthoracic esophagectomy with en bloc 2-field lymphadenectomy in patients post neoadjuvant chemotherapy. TTE allows for excellent postoperative staging. Furthermore, the omission of any of the upper mediastinal, thoracic, or abdominal field lymphadenectomy can negatively impact survival. A minimal lymphadenectomy was projected to lead to a 23% reduction in survival in patients with N1 or N2 disease [51] (see Table 3).

Table 3.

Studies supporting the role of extended lymphadenectomy post-neoadjuvant therapy settings

Author, year	n	Histology	Study design	Conclusion
Phillips et al. 2016 [51]	305 treated by nCT and TTE	Adenocarcinoma	Single institutional retrospective study	An estimated additional 4 cancer-related deaths were projected if proximal thoracic nodes were omitted, 2 additional deaths if minimal abdominal nodes were omitted, and 9 extra deaths if abdominal nodes and thoracic were omitted
Solomon et al. 2009 [53]	4224 (of which 1230 treated by nRT)	Both subtypes	SEER data base of 4224 who underwent surgical extirpation	Adequate lymphadenectomy and neoadjuvant radiotherapy to be independent predictors of improved survival
Torgersen et al. 2011 [54]	88 treated by neoadjuvant therapy and surgery	Both subtypes (predominantly adenocarcinoma)	Single institutional retrospective study	Patients with ≥18 nodes resected had a significantly longer median OS than those with <18 nodes resected
Hagens et al. 2019 [52]	50 treated by nCRT and TTE	Adenocarcinoma	Single institutional retrospective study	> 50% had nodal metastases within the radiation field. nCRT should not be a reason to minimize lymphadenectomy in patients with esophageal adenocarcinoma
Miyata et al. 2018 [55]	584 treated by nCT and TTE	Both subtypes but predominantly squamous	Retrospective study from two high-volume Japanese centers	Location and number of nodal metastases have a prognostic impact in patients with esophageal cancer undergoing neoadjuvant chemotherapy
Visser et al. 2018 [56]	10 studies- primary surgery, 6 – neoadjuvant therapy + surgery, 9-both arms	Both subtypes	Meta-analysis of 26 studies	OS, DFS significantly improved in the high lymph node yield group. Subset analysis of neoadjuvant-treated patients demonstrated a survival benefit of high lymph node yield on OS

nRT neoadjuvant radiation, nCRT neoadjuvant chemoradiation, nCT neoadjuvant chemotherapy, OS overall survival, DFS disease-free survival

Esophageal adenocarcinoma frequently metastasizes to both the mediastinal and abdominal lymph node stations. In a study by Hagens et al., more than half of the patients had lymph node metastases within the radiation field. nCRT was, therefore, not a substitute to decrease lymphadenectomy in patients with esophageal adenocarcinoma [52].

Solomon et al. analyzed the SEER database of 4224 patients and noted an additive survival benefit when both neoadjuvant radiation and adequate lymphadenectomy were combined in node-positive esophageal adenocarcinoma patients as both were independent predictors of improved survival [53]. Another study by Torgersen et al. found that the extent of lymphadenectomy, percentage of positive nodes, and pathological lymph node status were significant prognostic markers in patients who undergo esophagectomy after neoadjuvant therapy [54]. Patients with 18 nodes excised had significantly better survival than those with < 18 nodes resected. Likewise, patients with no nodal metastasis had significantly better median OS (51.4 vs 27.4 months; P = 0.025) and also DFS (45.3 vs 12.9 months; P = 0.03) against lymph node–positive patients. A recent study from Japan indicated that the number and location of nodal metastases in patients of esophageal cancer have a prognostic significance in patients undergoing preoperative chemotherapy [55]. Inadequate lymphadenectomy done as per response to neoadjuvant therapy cannot be rationalized [55].

Finally, a meta-analysis of 26 studies analyzed the effect of lymph nodal harvest on survival in patients with esophageal cancer undergoing esophagectomy with or without neoadjuvant therapy [56]. Significantly better overall survival was observed in the high lymph node yield group (hazard ratio (HR) = 0.81; 95% CI = 0.74–0.87; P < 0.01). Also, disease-free survival was significantly better in the high lymph node yield group (HR = 0.72; 95% CI = 0.62–0.84; P < 0.01). Subset analysis showed with the inclusion of neoadjuvant therapy, there was a significant improvement in overall survival with high lymph node yield (HR = 0.82; 95% CI = 0.73–0.92; P < 0.01).

Conclusions

The relationship between survival and lymph nodal retrieval in the era of neoadjuvant therapy is much more complex to establish. Furthermore, the performance of a transthoracic resection with extended lymphadenectomy is not associated with unacceptable morbidity in high-volume centers with the expertise and allows for optimal staging, locoregional disease control, and survival. As the surgical radicality decreases, the opportunity to remove diseased nodes is missed out and also leads to the erroneous assignment of post-therapy nodal status.

The post hoc analysis of randomized controlled trials has suggested that the number of resected nodes has a prognostic impact on patients’ survival after surgery alone and loses its significance in patients who receive neoadjuvant chemoradiation. Nonetheless, neoadjuvant therapy in esophageal cancer should not be taken as a proxy to conceal the deficiencies of an inadequate lymphadenectomy; instead, both these modalities should be considered collaborative to broaden the survival benefit as shown in many studies.

Finally, to provide a conclusive answer to this heated debate, adequately powered randomized trials need to be conducted on patients treated with neoadjuvant chemoradiotherapy and randomized to transhiatal resection with limited nodal dissection and transthoracic approach with extended nodal dissection and stratified further on histology and level of cancer.

Compliance with Ethical Standards

Conflict of Interest

The authors declare no conflicts of interest.