Early Oesophageal Cancer Treatment: EMR, ESD or Surgery?

Early oesophageal cancer (clinically T1a or T1b with no lymph node involvement) can sometimes be treated through the endoscope instead of by oesophagectomy. Endoscopic mucosal resection (EMR), endoscopic submucosal dissection (ESD) or ablation is best suited to small, well or moderately differentiated tumours confined to the mucosa, in patients who accept close endoscopic surveillance. In the practice of Dr. Preeti Vijayakumaran, Consultant, Oncology, Artemis Hospitals, Gurugram, India, most other early tumours are treated surgically.

Dr. Vijayakumaran set out the criteria she uses to decide between endoscopic treatment and surgery, the size and depth thresholds that guide the choice between EMR and ESD, and why ablation alone can be the wrong answer. This article also covers how Barrett's oesophagus and dysplasia are managed, since that is where most genuinely early cancers are found. The broader picture of the disease is in the complete guide to oesophageal cancer diagnosis and treatment.

What counts as early oesophageal cancer

Early oesophageal cancer is clinically T1 N0. It has two subgroups, defined by how deep the tumour has grown into the wall:

• T1a: the tumour extends no deeper than the muscularis mucosae, the thin muscle layer at the base of the mucosa.
• T1b: the tumour reaches the submucosa but has not reached the muscularis propria, the main muscle layer.

Both are node negative. The distinction between T1a and T1b matters because the risk of lymph node spread rises sharply once a tumour enters the submucosa, and endoscopic treatment removes only the tumour, never the lymph nodes.

Accurate staging is therefore the precondition for any endoscopic approach. In early disease, Dr. Vijayakumaran adds endoscopic ultrasound after the PET scan for a more precise T and N stage; the full workup is described in the article on how oesophageal cancer is diagnosed and staged.

Endoscopic treatment of early oesophageal cancer: EMR, ESD and ablation

Three endoscopic techniques are available, and they differ in whether they return tissue for the pathologist.

• Endoscopic mucosal resection (EMR) removes the lesion together with the surrounding mucosa. It suits smaller lesions.
• Endoscopic submucosal dissection (ESD) dissects the lesion out through the submucosal layer, allowing larger lesions to be removed in one piece.
• Ablation destroys abnormal tissue in place, without removing it, by methods such as cryotherapy (freezing). Because no specimen is retrieved, ablation provides no tissue to confirm the depth of the lesion.

Who is not a candidate for endoscopic treatment

Dr. Vijayakumaran identifies three features that rule endoscopic management out:

1. Poorly differentiated histology, whether adenocarcinoma or squamous cell carcinoma.
2. T2 disease, meaning invasion into the muscularis propria.
3. Node-positive disease.

The size of the lesion and its histology are the other key factors in the decision.

EMR or ESD: matching the lesion to the technique

Within those limits, the choice depends on how the lesion looks, how deep it goes and how large it is.

• Intramucosal lesions (carcinoma in situ, or cancer extending only into the lamina propria) that are well or moderately differentiated can be treated endoscopically.
• Flat lesions, whether high-grade dysplasia or an intramucosal T1a carcinoma limited to the lamina propria, can be treated by ablation or by endoscopic resection.
• Nodular lesions must be resected by EMR. Ablation alone is not sufficient.
• T1a lesions reaching the muscularis mucosae should be resected by EMR or ESD. For lesions smaller than 2 cm, EMR is preferred; for lesions larger than 2 cm, ESD is preferred.

Squamous cell carcinoma: depth and the 15 mm threshold

For early squamous cell carcinoma confined to the epithelium or lamina propria (often written M1 and M2), Dr. Vijayakumaran uses a size threshold of 15 mm: EMR for lesions smaller than 15 mm, ESD for larger ones. Squamous tumours that invade deeper carry a higher risk of nodal recurrence, so endoscopic management is usually not preferred for them.

Superficial submucosal invasion: possible, but high risk

Some T1b tumours with only superficial submucosal invasion (SM1) can still be offered endoscopic treatment. The depth limits are less than 500 microns of submucosal invasion for adenocarcinoma and less than 200 microns for squamous cell carcinoma. Dr. Vijayakumaran stresses that these patients remain at high risk of lymph node metastasis, and the decision needs to be made with that risk in view.

The 1.5 cm decision point for T1b disease

For lesions larger than 1.5 cm, particularly T1b disease, the decision is between ESD and oesophagectomy. Lesions smaller than 1.5 cm can be removed by EMR, after which a multidisciplinary group decides whether further treatment is needed, based on the pathology of the resected specimen.

Dr. Vijayakumaran's practice: when endoscopy, when surgery

Her hospital's practice condenses these criteria into a clear rule. For a T1a lesion smaller than 2 cm that is well or moderately differentiated, her team offers endoscopic treatment, provided the patient understands the need for surveillance afterwards. For all other early cases, she prefers to advocate oesophagectomy, unless the patient is not fit to undergo it.

The logic is that endoscopic treatment trades a smaller operation for a greater dependence on follow-up. A patient who will not or cannot return for regular endoscopies loses the safety net that makes the approach reasonable.

The limits of cryotherapy and other ablation

Cryotherapy and other ablative therapies apply only to very early oesophageal cancer, especially lesions detected by screening. Their weakness, in Dr. Vijayakumaran's explanation, is twofold.

First, ablation demands extremely accurate staging. It is reasonable only when endoscopy and endoscopic ultrasound show a flat high-grade dysplastic lesion or an early intramucosal carcinoma.

Second, ablation leaves structural abnormalities at the treated site, and those changes make later surveillance difficult. A recurrence can hide in scarred, distorted tissue.

For both reasons, if there is any suspicion of nodular disease or of deeper invasion, she recommends endoscopic resection rather than an ablative procedure such as cryotherapy alone. Resection removes the lesion and returns a specimen the pathologist can stage. She gave this explanation in answer to a doctor's question, recorded in the Live Discussion from this masterclass.

Managing Barrett's oesophagus and dysplasia: endoscopic mapping

Most real early-stage oesophageal cancers are found by screening. In Dr. Vijayakumaran's setting, screening is recommended from the age of 35 to 40 years for patients with reflux symptoms or a risk of Barrett's oesophagus, the change in the lining of the lower oesophagus caused by chronic acid reflux that predisposes to adenocarcinoma. The age to start varies between institutional protocols and guidelines.

Screened patients undergo endoscopic mapping, in which every area showing dysplastic change is biopsied. Management then follows the worst finding:

The same principle runs through this table and the criteria for early cancer: flat, small, superficial lesions can be ablated; anything nodular or larger needs to be resected, so that the pathologist can confirm what was there.

Surveillance after endoscopic treatment

Endoscopic treatment preserves the oesophagus, but it commits the patient to structured endoscopic follow-up that is most intensive in the first two years, with imaging added for T1b disease because of its higher risk of nodal spread. The full schedule is set out in the complete guide's section on follow-up.

When early disease still needs surgery

Not every early tumour fits the endoscopic criteria, and an endoscopic resection specimen can reveal deeper invasion or adverse features than staging suggested. For these patients, and for intermediate T2 N0 disease, surgery is the main option, either upfront or after preoperative treatment. The operation, its minimally invasive and robotic forms, and the evidence for combining it with chemotherapy and radiotherapy are covered in the article on locally advanced oesophageal cancer and the major trials.