Abstract
Background
Endoscopic resection (ER) has widely been established as a curative treatment approach in early-stage adenocarcinoma of the esophagus (EAC) limited to the mucosa (pT1a). The first report dates back to the year 2000. 1 In the meantime, several studies have been able to show that ER is effective and safe also in the long-term follow-up (FU).2–4 Esophagectomy has been the previous gold standard for EAC. In contrast to the organ-preserving approach of ER, it carries a mortality risk of at least 2%–5%. Morbidity of surgery ranges from 20% to 50%. That is why ER has gained importance during the last years.
There are two prerequisites for any kind of endoscopic treatment with a curative intent in EAC: First, the target lesion should have a very low risk of lymph-node (LN) metastasis. In contrast to esophagectomy, the LNs remain in place during endoscopic treatment. Second, endoscopic therapy should be effective and safe also in the long term. Only by long-term FU can extraesophageal tumor recurrence be ruled out. According to studies, both conditions are fulfilled for mucosal EAC.2,5,6
But what is the situation regarding EAC with invasion beyond the muscularis mucosae into the submucosal (sm) layer (pT1b EAC)? It is known that the overall risk of LN metastasis in pT1b lesions is higher than in mucosal EAC. However, the LN risk needs to be looked at from a more differentiated aspect: If the invasion depth into the submucosa is pragmatically divided in equal thirds (sm1/2/3), the rate of LN metastasis has been shown to depend on the depth of invasion.5,7 The risk of LN metastasis ranges from 0 to 20% in case of incipient sm invasion (pT1b sm1).5–14 Getting into even more detail, it has been reported that among pT1b sm1 EAC lesions there is a favorable histological pattern along with a very low risk of LN metastasis in accordance with the LN rates known from pT1a EAC.5–8,15
Endoscopic therapy for pT1b EAC
Recently published studies were able to show that the treatment strategy may be changed from surgery to endoscopic therapy also in pT1b EAC patients with incipient sm invasion (pT1b sm1), if a favorable histological pattern is present (“low-risk sm invasion”).16–18 In this group of patients, ER can be used as an alternative to surgery in experienced hands. This derives from the fact that long-term results of ER for low-risk pT1b EAC are very good, and the risk of LN metastasis (2%) is relatively lower than the mortality risk of surgery (2%–5%).16,17 A favorable histopathological pattern is defined by a good to moderate tumor differentiation (G1–2), and the absence of tumor invasion into lymphatic vessels (L0) or veins (V0).
The pragmatic measurement approach (sm1/2/3)
Despite all the enthusiasm for the advances in endoscopic therapy for EAC, the issue of “sm invasion” encourages discussions. The key question is whether the pragmatic approach of measuring the tumor invasion depth into the submucosa in equal thirds—with sm1 being the upper third—should be replaced by a standardized cut-off line in numbers. 19
Where do we stand? The “traditional” pragmatic measurement approach has been the basis for the evaluation of the risk of LN metastasis of pT1b lesions in large surgical series, as well as for the clinical series on efficacy and safety of ER for low-risk pT1b sm1 lesions. The use of the pragmatic classification has been proven to be safe for defining the indication for endoscopic therapy.16,17
The question has been raised whether the pragmatic classification generally makes sense for the measurement of the sm invasion depth in ER specimens. Why that? In contrast to surgical resection specimens containing all layers of the resected esophagus, the submucosa is not completely resected during ER. There may therefore be great variation in the absolute diameter of the sm thirds. It must be noted, however, that during ER, by far the largest part of the submucosa is within the resection specimen. This is well illustrated by the fact that the proper muscle layer can clearly be seen after ER.
The endoscopist should not have concerns about not having resected the whole amount of submucosa during ER, if the basal margins of the ER are tumor free and if the pragmatic classification for measuring the invasion depth is used. Subtotal resection may lead to a virtual overstaging, but not to an understaging of the sm invasion depth: The smaller the amount of submucosa, the relatively deeper the relative invasion depth measured by the pathologist. In case of an overstaging (e.g. pT1b sm2 instead of sm1), the indication for esophagectomy will be made.
Several factors may influence the amount of submucosa that may be resected during ER. Biopsies prior to ER, inflammatory changes within the Barrett’s segment, as well as fibrosis after previous endoscopic treatment may decrease the thickness of the sm layer. On the other hand, sm injection prior to ER may increase the diameter of the submucosa. After resection, there is shrinking of the ER specimen. There is therefore great variation in the thickness of the sm layer in ER specimens. This argues in favor of a pragmatic measurement approach for the invasion depth.
Measurement in microns (µm)
In other early neoplasias of the gastrointestinal (GI) tract, such as early colorectal or gastric cancer, the measurement of the invasion depth into the submucosa is based on another standard: The absolute measurement approach in numbers (microns, µm). The Paris Classification has proposed to define sm1 invasion in columnar epithelial tumors of the esophagus not to extend beyond the first 500 µm of the sm layer.19,20
Now the question arises whether this approach may be able to replace the pragmatic measurement approach in pTb EAC, and through this, to be the new standard. Fotis et al. examined this question in their interesting study published in this issue of the UEG journal. 20
Comparison of classification systems for pT1b invasion
The primary aim of the study by Fotis et al. 20 was to assess the prevalence of LN metastases in pT1b EAC in relation to the absolute depth of sm invasion (Paris Classification). As a second aim, they evaluated the efficacy of ER for low-risk sm1 EAC, as well as the impact of the two different classification systems on the subsequent clinical management and LN involvement. 20
Pathology reports of all patients who had undergone surgery or ER for EAC at Erasmus Medical Center during a period of 20 years were retrospectively reviewed, and only patients with pT1b lesions were included in the study. 20 In the ER patients, median FU was 18 months. No patient was diagnosed with cancer recurrence. However, the group of ER patients was relatively small, and FU duration was short, which may limit the validity of the study. In the surgical group there is one detail that leaps to the eye: The width of the sm layer varied widely from 1030 to 7400 µm. Also in the surgical group, there were seven sm1 patients according to the pragmatic classification. Two of these patients had LN involvement, with an sm invasion depth >500 µm. According to the Paris Classification, these patients would have been staged as sm2/3 patients and would have been primary candidates for surgery. These results confirm that in the individual patient, the upper third of the submucosa may be wider than 500 µm.
Now the question arises how many patients from the different groups belonged to the pragmatically measured sm1 group with a favorable histological pattern (low risk). Second, it may be interesting to know how many of them had an invasion depth beyond 500 µm. From a total of 54 patients, there were 14 pragmatically measured pT1b sm1 lesions with a histological low-risk pattern. One LN event was observed in the surgical group in a patient with an invasion depth of 700 µm. This patient would have been defined as having an sm2 lesion according to the Paris Classification. Therefore, this lesion may have been underestimated regarding its risk of LN metastasis in the clinical setting, if only the pragmatic classification would have been used for risk stratification. A total of three of the 14 sm1 LR patients had an invasion beyond 500 µm (21% discrepancy between the two classification systems).
The authors 20 rightly conclude that the different classification systems may lead to different LN risks and treatment strategies. The clinical relevance may, however, be limited: It may in general be critical to compare primarily endoscopically treated patients to surgical patients, because of the differences in the amount of submucosa resected and the histopathological work-up. It may be difficult for the pathologist to find the deepest point of tumor invasion in a relatively large surgical specimen. Tumor invasion may therefore be understaged with a relatively higher rate of LN involvement in comparison with ER specimens. In addition, surgical specimens are routinely cut in 5-mm slices, whereas ER specimens are cut in 2 mm slices. 20 The latter technique therefore allows a more detailed view of the lesion and its invasion depth. The authors also conclude that patients with low-risk sm1 EAC according to the Paris Classification are suitable candidates for endoscopic therapy. Nevertheless, the number of ER patients in the present study was relatively too small to draw final conclusions.
The other side of the coin
Let’s change the view: How about lesions that invade less than 500 µm into the submucosa (Paris Classification sm1), but that invade into the mid- (sm2) or even lower third (sm3) of the submucosa if using the pragmatic classification? Would it not be dangerous to look at the Paris Classification only when performing ER with a curative intent? Using microns only may in the worst case give a false sense of security to the endoscopist, if the amount of submucosa resected is relatively thin.
How can the endoscopist get out of this dilemma? It can be suggested to use both measurement approaches to evaluate tumor invasion depth and to define the treatment strategy. Patients may be good candidates for ER with a curative intent if they are “double sm1” according to both classifications. In case of discrepancy, the decision should be made on the basis of the relatively worse staging result.
Conclusions
The traditional pragmatic measurement approach of dividing the submucosa into equal thirds remains the basis for endoscopic therapy with a curative intent in low-risk sm1 EAC, even though there may be great variation in the absolute invasion depth.
The Paris Classification offers a well-defined cut-off line of 500 µm and has therefore created the impression of being the new standard. However, no recommendation can be made to use the Paris Classification only for clinical decision making because it may give a false sense of security to the endoscopist. It can therefore be suggested to use both classifications for clinical decision making (“double sm1”).