|
Editorial
Low prevalence of invasive adenocarcinoma and occult cancer
on esophageal resection for Barrett’s esophagus with
high-grade dysplasia: Evidence for conservative management
Department of Anatomic Pathology, Cleveland Clinic, Cleveland, OH
J Gastrointest Oncol 2011; 2: 5-7. DOI: 10.3978/j.issn.2078-6891.2011.003
|
|
Due to the variety of therapeutic options that are currently
available for patients diagnosed with Barrett’s-related
high-grade dysplasia (BE-HGD), the choice of optimal
management continues to be a topic of discussion among
gastroenterologists, surgeons, oncologists, pathologists,
and patients. Per the current American College of
Gastroenterology guidelines, HGD is considered a
threshold for therapeutic intervention (1). The choice of
management ranges from the most conservative approach
- continued endoscopic surveillance (2,3) to the most
aggressive option - esophagectomy, with endoscopic
therapies such as endoscopic mucosal resection (4) and
ablation therapy somewhere in the middle (5).
The potential to completely eradicate the diseased
segment as well as the fact that 12.7% - 75% (mean-
39.3%) of patients with a pre-operative diagnosis of
HGD will harbor adenocarcinoma on esophagea l
resection (6) are the most compelling reasons in favor of
esophagectomy. Esophagectomy, however, is associated
with significant mortality and morbidity, with estimates
of mortality ranging from 0% - 2% at high-volume centers
to 8%-10% at low volume centers (7). On the contrary,
with significant advances in endoscopic techniques, the
role of esophagectomy is becoming restricted to patients
diagnosed with multifocal dysplasia who have failed
endoscopic therapy and patients in whom pre-operative
imaging modalities such as endoscopic ultrasound staging (EUS) suggest the presence of at least submucosal disease.
Compared to that which had been found in earlier
studies, where up to 40% of patients with a pre-operative
diagnosis of HGD demonstrated adenocarcinoma on
resection, Konda et al in a meta-analysis of 23 studies
found that only 12.7% cases of HGD showed evidence of
underlying invasive adenocarcinoma in esophagectomy
specimens (8). In this analysis, invasive adenocarcinoma
was defined as tumor invading the submucosa (submucosal
adenocarcinoma, SMC) and beyond. This definition
was specifically adopted for the study as the risk of
lymph node metastasis is much lower with intramucosal
adenocarcinoma (IMC, 0%-8%) (9) as compared to
submucosal invasion (8%-33%) (10). In the study by Nasr
and Schoen (11) published in this edition of the journal,
using the same rationale, the authors provide compelling
evidence that the rate of invasive adenocarcinoma (IMC
and SMC) is 17.6%, much lower than the reported average
rate of approximately 40%. In a retrospective analysis of
68 patients undergoing esophagectomy for a pre-operative
diagnosis of HGD, they identified 4 cases of IMC and 8
cases of SMC on esophageal resection, with an overall rate
of SMC of 11.7%. There was no statistical difference in the
average size of tumors in the IMC vs invasive carcinoma
categories (0.61 cm vs 1.86 cm). Of the 8 cases of invasive
adenocarcinoma, the incidence rate of occult SMC was 4/68
(5.9%). A time-based analysis of two groups (1993-2000
and 2000-2007) showed no difference in the detection rate
of adenocarcinoma associated with HGD.
In an attempt to predict which cases of HGD will harbor
concurrent adenocarcinoma, several pre-operative factors
including pre-operative biopsy protocols, endoscopic
findings as well as histologic features have been the focus
of attention of many recent studies. Significant variability
in pre-operative sampling protocols, endoscopic evaluation
techniques, histologic assessment, as well as potential selection bias in the cohorts may have contributed to the
relatively high estimated rate of occult adenocarcinoma
in some of the previous studies. One of the limitations
of the study by Nasr and Schoen, which according to the
authors may have led to a higher rate of occult cancer, is
the lack of standardized pre-operative testing including
imaging studies and presumably endoscopic evaluation.
The Seattle biopsy-based endoscopic surveillance protocol,
consisting of serial 4-quadrant biopsies at 1-cm intervals
with jumbo biopsy forceps, along with aggressive targeting
of endoscopically visible lesions has been advocated as a
technique that can improve the rate of detecting carcinoma
(2,12). In a recent study, Kariv et al demonstrated that even
this extensive tissue sampling protocol misses a substantial
percentage of cancers detected at esophagectomy (13).
One needs to however bear in mind that this study was
a cross-sectional study that analyzed data at one specific
time point. In fact, Kariv et al have recommended that
more serial endoscopies may be more important than one
rigorous protocol, possibly because prevalent dysplasia,
which is known to harbor higher rates of adenocarcinoma,
is screened out.
Of the 8 cases of invasive adenocarcinoma in this study, 4
(50%) had evidence of an endoscopic abnormality (erosion,
nodules or stricture). There is sufficient evidence to support
the contention that endoscopically visible lesions in a
Barrett’s segment are often associated with adenocarcinoma
on esophagectomy and therefore must be ta rgeted
aggressively, particularly with a biopsy diagnosis of HGD
(13-16).
The more conservative treatment options demand better
distinction between HGD, IMC, and SMC on mucosal
biopsies. This large surgical series further provides evidence
that it is important to separate IMC from SMC, as it may
inf luence the choice of therapeutic intervention. Given
the clear prognostic difference between HGD, IMC, and
SMC, pathologists are often expected to reliably make this
distinction on small biopsy material. The approximately
40% adenocarcinoma rate in patients with a pre-operative
diagnosis of HGD highlights the fact that it is not always
possible for pathologists to make this distinction. The two
main problems are: 1) sampling error – e.g., do more biopsies
help pathologists distinguish HGD from IMC from SMC?
and 2) interobserver variability – e.g., can pathologists
reliably distinguish the higher end of Barrett’s neoplasia
spectrum? In an attempt to assess histologic features on preoperative
biopsies that would be associated with a higher
risk of concurrent adenocarcinoma on resection, two recent
studies performed at the University of Michigan (UM)
(17) and Cleveland Clinic (CCF) (18) identified categories
of HGD suspicious for adenocarcinoma (UM) and HGD with marked glandular architectural distortion (CCF).
Compared to HGD alone, both categories were significantly
associated with IMC or SMC. Nevertheless, pathologists
are relatively poor at separating HGD from IMC and even
SMC (18). In addition, pathologists rarely find diagnostic
evidence of SMC in biopsy material. In another study
performed at the Cleveland Clinic, the overall rate of SMC
on esophageal resections from patients diagnosed with
Barrett’s-related HGD or worse was 21.4% (24/112). Of
these cases, only 3 cases (2.7%) had unequivocal evidence
of submucosal invasion on biopsy (19). Pathologists also
struggle with the distinction between SMC and IMC
because of the well-recognized split muscularis mucosae
(20,21). On superficial biopsies, or even endoscopic mucosal
resection (EMR) specimens, it is often difficult to decide
whether neoplasm below one layer of muscularis mucosae is
within the submucosa or “pseudo-submucosa.”
While all the available modalities of risk assessment
including endoscopy, imaging, and histology do allow us
to guide clinical intervention, none are perfect. Although
EMR and ablation therapy are emerging as popular
choices for management of Barrett’s-related HGD and
IMC, recurrence of neoplasia at the rate of 11%-21% has
been reported in these patients (22,23). The advantages
of these approaches, specifically EMR, are larger tissue
samples that not only allow better evaluation of histologic
landmarks, but also improve diagnostic accuracy and
staging (24). This approach does sound reasonable if there
is an endoscopically visible lesion. In a series of 78 patients
undergoing esophagectomy, Oh DS et al demonstrated
that nearly a third of patients with IMC did not have any
visible lesions on endoscopic evaluation, thus concluding
that some cases of IMC may not be amenable to endoscopic
therapies (25). The current study does, however, caution
about overestimating the rate of occult adenocarcinoma,
suggesting that esophagectomy is not indicated in all
patients diagnosed with HGD; others may examine this
same data and argue that 6% risk of unsuspected (deeply)
invasive adenocarcinoma is too high to justify carte blanche
conservative therapy. In fact, this series highlights the
difficult decisions that patients and their doctors must make
when faced with a diagnosis of HGD. Unquestionably,
there is a risk of unsuspected adenocarcinoma and lymph
node metastasis in patients with Barrett’s-related HGD.
This risk is dependent on numerous factors including,
the r igor of the sampl ing protocol, the endoscopic
appearance, the reliability of the pathologic interpretation,
the multifocality of the neoplasia, whether the patient is
actively under endoscopic Barrett’s surveillance, and the
results of additional staging modalities such that there is no
“cookbook” answer for the treatment of HGD. In reality, the ultimate choice of therapy must be individualized by
taking into consideration all of the variables in addition to
patient’s individual profile to come to a consensus decision
for therapeutic intervention.
|
|
References
Cite this article as:
Patil D, Plesec T, Goldblum J. Low prevalence of invasive adenocarcinoma and occult cancer
on esophageal resection for Barrett’s esophagus with
high-grade dysplasia: Evidence for conservative management. J Gastrointest Oncol. 2011;2(1):5-7. DOI:10.3978/j.issn.2078-6891.2011.003
|
