Superficial esophageal cancer: endoscopic resection or radical surgery?

Keywords::

• endoscopic resection
• endoscopic submucosal dissection
• radical oncologic surgery
• superficial esophageal cancer

In recent years, multiple new endoscopic therapies, which are far less invasive compared with radical surgery with lymphadenectomy, have become available for the treatment of Barrett’s esophagus and superficial esophageal cancer. These include resection techniques, such as endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD). In these procedures, the resected tissue specimens can be used to stage tumor infiltration depth, both in mucosal and submucosal esophageal cancer. By contrast, oncologic esophagectomy with systematic lymphadenectomy enables the complete resection of the lesion, including potentially affected lymph nodes, and thus results in accurate staging of the cancer.

In cases of superficial esophageal cancer, the decision regarding endoscopic therapy versus radical surgery requires weighing the benefits of the less invasive and less risky procedure against the potentially curative, although more invasive procedure, in patients with lymphatic involvement. The current dilemma is not only the unavailability of suitable staging devices with high accuracy, but also the lack of markers predictive of lymphatic spread.

With the emergence of ESD techniques for submucosal esophageal cancer, it has become a matter of debate if these are oncologically adequate in view of the increased risk of lymph node metastases with tumor infiltration at submucosal (sm) levels 1–3 (1 = upper sm third; 2 = midle sm third; 3 = lower sm third). Although there is no doubt that mucosal adenocarcinoma (ADC) of the esophagus with a low risk for lymph node metastases should be treated using EMR [1], data on squamous cell carcinoma (SCC) are sparse, and data on neoplasia, mucosal and submucosal cancer have been reported primarily for mixed patient populations with high-grade intraepithelial neoplasia, thus rendering a definite conclusion on long-term results of endoscopic techniques for this entity difficult [2]. Due to the more aggressive tumor spread in SCC compared with superficial Barrett’s cancer, the use of local ablation methods has to be carefully considered, and the question as to whether mucosal SCC – and submucosal tumors in general – should be resected endoscopically has not yet been answered conclusively. Considering the frequent undermining submucosal tumor spread in SCC, endoscopic procedures do not appear to be oncologically justifiable for sm invasion. Surgical studies of oncologic resectates of superficial SCCs have shown that a lymph node metastasis rate of 8% can be anticipated from mucosa (m) infiltration level 3 onwards [3]. Significantly higher rates of 18 [4] and 23% [5] have been reported by other studies from Japan in cases of mucosal level m3 infiltration. This leads to the conclusion that, based on the m1–3 subclassification, only levels m1 and m2 are more likely not to be associated with lymphatic involvement in SCC, and are thus oncologically adequate in terms of endoscopic resection. Indications for endoscopic resection in superficial SCC are low- and high-grade intraepithelial neoplasia, tumor infiltration levels m1 and m2 (except for risk factors L+, V+ and G3), as well as the macroscopic types 0–IIa, 0–IIb and 0–IIc. By contrast, tumor infiltration levels of m3 or higher, tumor diameter greater than 20 mm, and multifocal carcinoma and macroscopic types 0–I and 0–III imply radical esophagectomy with systematic lymphadenectomy.

A systematic review of surgically resected SCC specimens revealed an increasing rate of lymph node metastasis for tumor infiltration levels sm1–3: 27% in the upper submucosal third (sm1), 36% in the middle submucosal third (sm2), and as high as 55% in the lower submucosal third (sm3) [Gockel I, Sgourakis G, Lyros O et al. Risk of lymph node metastasis in submucosal esophageal cancer: a review of surgically resected patients (2010); Submitted Manuscript]. At these depths of infiltration, the percentage of lymphatic channel invasion was even higher, at 52 (sm1), 65 (sm2) and 64% (sm3) [Gockel I, Sgourakis G, Lyros O et al. Risk of lymph node metastasis in submucosal esophageal cancer: a review of surgically resected patients (2010); Submitted Manuscript]. Every type of submucosal SCC therefore represents a clear indication for transthoracic esophagectomy with at least two-field lymphadenectomy, on the precondition that the patient’s operative risk is tolerable.

Although the risk of lymphatic spread in mucosal ADC is neither negligible nor approximating zero – and may even be as high as 7.1%, as in our own patient population [6] – endoscopic therapy has become a favorable alternative to radical surgery. If lymph node metastases can be excluded by CT scan, endosonography and PET-CT, it should be offered as a reasonable esophagus-sparing treatment option. A recently published surgical series demonstrated that, even in the presence of sm1 submucosal invasion, the risk of lymph node metastasis is relatively low [7]. A systematic review of oncologic esophagectomies revealed a rate of 6% for all Barrett’s carcinomas sm1, and a steadily increasing rate of 23% in sm2 and 58% in sm3 [Gockel I, Sgourakis G, Lyros O et al. Risk of lymph node metastasis in submucosal esophageal cancer: a review of surgically resected patients (2010); Submitted Manuscript]. Bearing in mind that the operative mortality reported even by high-volume centers ranges from 2–5% – although rates as low as 1.4% were cited for minimally invasive esophagectomy [8] – the 6% risk of lymph node metastasis following endoscopic therapy can be regarded as justified.

The study by Manner et al. reports the long-term results of endoscopic resection with a curative intent in early Barrett’s carcinoma with ‘low-risk’ submucosal invasion in the largest series of patients to date [9]. Low-risk submucosal cancer was defined as invasion of sm1, absence of infiltration into lymphatic vessels/veins, histological grade G1/2, and macroscopic type I/II. The calculated 5-year survival rate of all 21 patients was 66% and no tumor-related deaths occurred [9]. However, a possible point of criticism may be that even using advanced techniques, it remains difficult to predict a ‘high-risk’ status and/or to detect lymph node metastasis in submucosal carcinomas with a sufficiently high accuracy. Thus, as the depth of infiltration in ‘diagnostic’ endoscopic resection (ER) can only be estimated by dividing the respective esophageal specimen into sm levels 1–3, in contrast to the sm classification in surgical resectates, the identification of sm1 invasion in ER specimens must be regarded with caution. Using ER, the specimen may easily be damaged by injection or manipulation, and it frequently does not contain the full thickness of the submucosa. Furthermore, the base of the endoscopically resected sample shrinks as a result of electrocautery. The actual depth of tumor extension into the submucosal layer may not be reliably assessed owing to the absence of the muscularis propria. Therefore, endoscopic therapy for sm1 Barrett’s carcinoma should be cautiously considered in patients who are poor surgical candidates, such as elderly patients with a high comorbidity or those who refuse surgical therapy. However, it must be emphasized that no long-term results comparing endoscopic treatment with radical surgery are available for this special histologic entity and tumor stage. A study by Pohl et al. compared the costs and quality-adjusted life years following endoscopic versus surgical therapy for early cancer in Barrett’s esophagus, including superficial submucosal invasion (sm1) [10]. During the 5-year study period, the calculated endoscopy costs were US$17,000 and yielded 4.88 quality-adjusted life years, in contrast to $28,000 and 4.59 quality-adjusted life years for esophagectomy.

Due to the lack of prospective, randomized trials evaluating the optimal treatment strategy for patients with superficial esophageal cancer, the following recommendation can be made based on an analysis of the currently available literature:

• • SCC – only m1 and m2 depths of infiltration and ‘low-risk’ indications should be resected endoscopically in poor surgical candidates. From m3 onwards, radical surgery with at least two-field lymphadenectomy is mandatory;

• • ADC – due to the low risk of lymph node metastasis in mucosal cancer, stages m1–m3 are suitable for endoscopic therapy in the presence of a ‘low-risk’ situation. Involvement of the superficial sm level (sm1) might be treated by ESD in patients with comorbidity and high operative risk – but in cautious surveillance programs only. ADC sm2 and sm3 continue to require surgical resection with systematic lymphadenectomy.

In particular, patients with increased operative risk and superficial esophageal cancer can be recommended esophagectomy-sparing techniques after having been evaluated by a surgeon – on the precondition of meticulous staging and accurate follow-up in endoscopic resections.

Future randomized trials comparing both treatment options in patients with superficial esophageal cancer should not only assess long-term survival and tumor recurrences, but also functional results and the disease-specific quality of life.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.