|
Review Article
Clinical T2N0M0 carcinoma of thoracic esophagus
1Radiation Oncology Service, Department of Radiology, Brooke Army Medical Center, 3851 Roger Brooke Drive, Fort Sam Houston, TX 78234, United States.
2Deparment of Medical Oncology, Roswell Park Cancer Institute, Elm & Carlton Streets, Buffalo, NY 14263, United States.
3Deparment of Radiation Medicine, Roswell Park Cancer Institute, Elm & Carlton Streets, Buffalo, NY 14263, United States.
|
|
Abstract
The incidence of esophageal cancer continues to rise. Historically, the majority of patients had been diagnosed with more advanced stages
of disease. However, with screening programs, like those in place for patients with Barrett's esophagus, many patients are being diagnosed
earlier in their disease course. Results of randomized trials and meta-analyses have led to emerging guidelines advocating for neoadjuvant
therapy as part of the treatment algorithm for esophageal cancer, particularly for T3 or greater T-stage and/or node-positive disease. And for
those patients that present with disease localized to the esophageal mucosa, endoscopic mucosal resection has proved to be an effective and
less morbid alternative to radical resection. However, there is a group of patients that lie between these two extremes for which evidence is
sparse and treatment recommendations vary. In this manuscript, recent data regarding work-up and treatment are reviewed along with a
closer look at those patients clinically staged with T2N0M0 (Stage IIA) esophageal cancer. Pertinent data in this subset of patients is analyzed
as potential treatment algorithms are discussed.
Key words
T2 esophageal cancer; endoscopic ultrasound; esophagectomy; radiation therapy; chemotherapy trimodality therapy
J Thorac Dis 2010;2:36-42. DOI: 10.3978/j.issn.2072-1439.2010.02.01.013
|
|
Esophageal cancer is one of the ten most common malignancies
worldwide, and its incidence is rising in the United States, where
there were an estimated 16,470 new cases and 14,530 deaths from
esophageal cancer in 2009 (1,2). While squamous cell carcinoma
is the most common histology of esophageal cancer worldwide,
there have been substantial increases in the rates of adenocarcinoma
in both the United States and Europe over the last several years.
Adenocarinoma is now the most common histology in these countries
(3). Analysis of this increase of adenocarcinoma has found
rising incidence across stages, ages, and gender indicating that it is
not artificially increased because of increased surveillance and earlier
diagnosis (4). Esophageal cancer is a significant health concern
throughout the world not only because of its prevalence but
also because of its substantial mortality, with 5-year overall survival
rates of <20% overall (5).
Historically, most patients have been diagnosed with locally advanced
or metastatic disease, both of which carry a dismal prognosis
even with aggressive multimodality therapy. In recent years,
screening programs have been developed, particularly for patients
with a history of Barrett’s esophagus. A recent a recent meta-analysis reported an estimated cancer incidence in Barrett’s esophagus
of 4.1-6.1/1,000 person-years (6). This screening has led to an increase
in early stage esophageal detection and treatment. With
many factors to account for, including stage, location, and histology,
treatment of esophageal cancer should be individualized for
each patient to optimize outcome while minimizing morbidity. In
this review, the evolution of esophageal cancer staging and management
is examined with special attention to the unique case of
clinical (c) T2N0M0 cancer of the thoracic esophagus. This optimal
management strategy in lesions felt to be invading into, but not
through, the muscle wall of the esophagus (T2) is a topic of debate.
In this manuscript, the relevant prospective and retrospective data
for treating this stage of esophageal cancer are evaluated in detail
as potential management algorithms are discussed.
|
|
Work-Up
Differentiating between potentially curable and incurable disease
has been difficult historically. Secondary to inaccurate staging
and aggressive pathology, local and distant recurrences have
been common despite aggressive treatment that often utilizes combinations
of surgery, radiation therapy, and chemotherapy. As patterns
of disease failure are being better elucidated, accurate staging
is critical for both prognosis and for guiding management. Proper
staging allows for the selection of patients who can be treated with
monotherapy or combined modality therapy, while also identifying
those patients with metastatic disease and sparing them from undergoing
intensive local therapy (7,8).
Pre-therapy staging of esophageal cancer should include esophageal ultrasonograpy (EUS). An echoendoscope is passed into
the esophagus where there is direct visualization of the histological
layers of the esophagus potentially involved with tumor (9,10).
It has been shown to be relatively accurate in evaluating the depth
of tumor invasion and also appears to be useful in assessing involvement
of regional lymph nodes (11). There have been several
studies completed suggesting that EUS is superior to other imaging
modalities, such as computed tomography (CT) and fluorodeoxyglucose
positron emission tomography (FDG-PET), for assessing
T-stage since these other modalities lack the ability to differentiate
between layers of the esophagus (12-17). In addition,
EUS has been found to be more accurate and has a higher sensitivity
than CT and PET for determining lymph node involvement
(17-18). There is however, some discrepancy between studies as
to how accurate EUS is in determining depth through the esophagus
and lymph node involvement, particularly for early T-stages (9,19-22). An early meta-analysis found EUS to have 89% accuracy
for determining T-stage, and a more recent meta-analysis by Kelly
et al. found EUS to have 79% accuracy for determining N-stage
(12,14). Puli et al. also published a meta-analysis that included 49
studies and 2558 patients which showed a pooled sensitivity of
EUS to determine T-stage of approximately 81-90%, with a higher
sensitivity for advanced (T3-T4) disease (9). In this study, the
pooled sensitivity and specificity of EUS was 81.6% (95% Confidence
Interval [CI]: 77.8-84.9) and 99.4% (95% CI: 99.0-99.7), respectively
for T1 disease, 81.4% (95% CI: 77.5-84.8) and 96.3%
(95% CI: 95.4-97.1), respectively for T2 disease, 91.4% (95% CI:
89.5-93.0) and 94.4% (95% CI: 93.1-95.5), respectively for T3 disease,
and 92.4% (95% CI: 89.2-95.0) and 97.4% (95% CI:
96.6-98.0), respectively for T4 disease (9). In the same study, the
sensitivity and specificity of EUS to diagnose nodal (N) stage disease
was 84.7% (95% CI: 82.9-86.4) and 84.6% (95% CI:
83.2-85.9), respectively. These modest numbers for nodal diagnosis
were improved with the addition of fine-needle aspiration
(FNA), a procedure that allows for tissue diagnosis and carries little
additional risk of morbidity or complication (9,23). With the
addition of FNA to EUS, sensitivity improved to 96.7% (95% CI:
92.4-98.9), and specificity improved to 95.5% (95% CI:
91.0-98.2).
In more recent years, FDG-PET has been used increasingly for
initial staging in esophageal cancer, as malignant cells have in
creased uptake of FDG and following phosphorylation are temporarily
retained in the cell (7). These areas of increased FDG
concentration are then evaluated with PET (24). FDG-PET has
been shown to be more specific but often less sensitive than CT for
detecting locoregional disease, and both have been found to be less
sensitive and accurate that EUS for determining T and N stage
(17-18,25). In initial staging of esophageal cancer, FDG-PET has
been found to have the most utility in detecting distant metastases.
Lerut et al. compared FDG-PET to the combination of EUS and
CT for the detection of distant disease, and found FDG-PET to
have superior accuracy 86% versus (vs.) 62% in favor of FDG-PET (26). In the United States, current guidelines recommend
FDG-PET or PET/CT, preferably, as part of the initial staging algorithm
in esophageal cancer management (27).
|
|
Treatment
Endoscopic Mucosal Resection (EMR)
With earlier detection, particularly in patients undergoing
screening for Barrett’s esophagus, along with improved staging
with EUS, there are large populations of esophageal patients with
superficial disease (Tis or T1a). In these patients, the risk of lymph
node involvement in surgical series is typically on the order of 5%
or less (28-29), and EUS has been shown to have an accuracy in
detecting T1a disease of over 90% in most series (30-32).
Esophagectomy is the standard surgical treatment in esophageal
cancer but carries significant morbidity and risk of mortality (29,33). For these patients with superficial disease, endoscopic therapy,
typically in the form of endoscopic mucosal resection (EMR),
has been gaining ac ceptance worldwide. A study evaluating the
Surveillance Epidemiology and End Results (SEER) database of
the National Cancer Institute (NCI) from 1988-2003 found that for
patients with superficial esophageal tumors treated endoscopically
or with surgical resection experienced equivalent long-term survival
(34). In fact, recent studies have shown 5-year OS rates of
over 95% and relatively low recurrence rates for patients treated
with EMR (35-36). For those patients treated with EMR found to
have T1b or greater depth of invasion on pathological assessment,
EMR acts as a diagnostic tool, with further therapy recommended
because of the increased risk of lymph node metastases (35,37-39).
Esophagectomy
While the rate of lymph node metastasis with mucosal cancers
is typically 5% or less, submucosal invasion (T1b) carries a much
higher risk of LVI and lymph node involvement (28). Kim et al.
found that the rate of lymph node involvement increased from 6%
(4/64) in mucosal cancers to 29% (39/133) in submucosal cancers
(P =0.001), while Cen et al. reported an increase from 4% in T1a
patients to 23% in patients with T1b tumors (28-29). Other series
have rates of lymph node metastasis as high as 40-50% in patients
with submucosal extension of disease and clinical T2 disease can
carry an approximately 50% risk of spread to lymph nodes (40-41).
For these patients, surgical management should include primary resection
and lymph node dissection (41-43). There is some debate
over the extent of lymph node dissection that should be undertaken
at resection, but there is also data suggesting that the number of
nodes removed at resection is associated with survival (44). In addition,
at least a 2-field abdominal and thoracic lymphadenectomy
has been associated with improved local control (45). Less invasive
techniques such as laparoscopic or thoracoscopic approaches have shown some promise but their routine use is still considered
investigational (46).
Neoadjuvant Concurrent Chemotherapy and Radiation Therapy
While neither on its own has been highly efficacious in the
neoadjuvant setting, the combination of concurrent chemotherapy
and radiation therapy has some theoretical advantages. With approximately
one-third of patients recurring distantly after surgical
management alone, additional therapy continues to be investigated
(47). Microscopic disease outside the typical radiation field or surgical
resection can be treated with the systemic therapy. In addition,
chemotherapy can sensitize the local-regional disease to the
radiation therapy, thus increasing its efficacy. Ideally, increased efficacy
would allow for tumor downsizing and allow for a more
complete surgical resection (48). And while acute toxicity may potentially
be increased, with such poor prognosis for esophageal
cancer patients, several studies have been undertaken to evaluate
the benefit of neoadjuvant concurrent chemoradiation. While
neoadjuvant chemoradiation should be of theoretical benefit, randomized
trials evaluating its efficacy have had conflicting results,
with very few of the studies reaching target accrual allowing for
complete analysis (49-54). The recently published Cancer and
Leukemia Group B (CALGB) 9781 trial assessed OS and treatment
response of neoadjuvant chemoradiation followed by surgery compared
to surgery alone (54). And while the study failed to reach
accrual, the results are intriguing. Neoadjuvant therapy consisted
of cisplatin 100 mg/m2 and fluorouracil 1,000 mg/m2/d for 4 days
during weeks 1 and 5 concurrent with radiotherapy (50.4 Gray),
followed by surgery. Neoadjuvant therapy was well tolerated with
no increase in increase in perioperative mortality. Pathologic CR
was 40% in the neoadjuvant arm and median survival was 4.48
years vs. 1.79 years (P=0.02) in favor of neoadjuvant chemoradiation.
In addition, 5-year OS was 39% for patients receiving neoadjuvant
therapy vs. 16% in the surgery alone arm. With so many of
the trials having results that are difficult to interpret because of
study design flaws or failure to reach targeted accrual, meta-analyses
are very important to determine what, if any, role neoadjuvant
chemoradiation has in esophageal cancer treatment. In a recent
meta-analysis by Gebski et al. 10 randomized trials comparing
neoadjuvant chemotherapy and radiation therapy followed by
surgery vs. surgery alone are analyzed (55). The hazard ratio (HR)
for mortality with neoadjuvant chemoradiation (either sequential or
concurrent) vs. surgery alone was found to be 0.81 (95% CI 0.70–
0.93; P=0.002). This corresponded to a 13% absolute difference in
OS at 2 years. The results were similar for both adenocarcinoma
and squamous cell carcinoma. Another recent meta-analyses presented
by Thirion et al. also demonstrated an OS benefit to neoadjuvant
chemoradiation, HR = 0.82 (95% CI: 0.72-0.93, P = 0.002)
translating into 2 and 5-year absolute benefits of 7% (from 37% to
44%) and 7% (from 18% to 25%), respectively (56). This study also
found a significant disease-free survival (DFS) benefit for neoadjuvant therapy, HR= 0.85 (95% CI: 0.72-0.99, P= 0.036),
yielding a 2 and 5-year absolute DFS increase of 6% (from 26% to
32%) and 4% (from 11% to 15%), respectively. R0 resections were
also increased by 6% (73% vs. 67%, P=0.03) with the addition of
neoadjuvant therapy and there was no difference in perioperative
mortality among the two groups.
|
|
Clinical T2N0
Based on individual randomized trials and meta-analyses, there
is an emerging agreement that either chemoradiation or systemic
therapy alone should be considered as part of the treatment algorithm
for esophageal cancer, particularly for T3 or greater T-stage
and/or node-positive disease (2,57). And in those patients that
present with disease localized to the esophageal mucosa only,
EMR is proving to be an efficacious and less morbid alternative to
radical resection. But what treatment algorithm should be considered
for the medically fit patient with intermediate risk disease,
such as clinical T2N0M0 (Stage IIA), who have clinical evidence
of disease invading into the muscularis propria after initial staging
is complete? Killinger et al. reported on their experience with
pathologic T2N0M0 disease treated with esophagectomy alone
(58). In this retrospective analysis, patients with pT2N0M0 disease
had a survival rate on par with patients treated with pT1N0M0 (approximately
50% at 5-years; P=0.83), and trended toward improved
survival over patients with pT3N0M0 disease (p=0.06).
Tachibana et al. published their results of patients with squamous
cell carcinoma of the esophagus treated with upfront esophagectomy
and found to have pT1 or pT2 disease (59). Among patients
that were node negative, patients with pT1 and pT2 tumors had
similar cancer specific survival (CSS).
Clinical T2N0M0 is a relatively unusual entity, with the majority
of esophageal cancers presenting with more advanced disease,
and many of those detected as part of screening programs being diagnosed
with disease confined to the mucosa (24). In addition, by
the time esophageal cancer has extended to the muscularis propria,
there is an approximately 50% risk of nodal involvement (41). Despite
its relative rarity, consideration should be made for a potential
treatment algorithm for this stage of disease. In the United States,
the National Comprehensive Cancer Network (NCCN) guidelines
advocate for a multimodality approach for any clinical T-stage that
is T2 or greater regardless of nodal status (27). For either adenocarcinoma
or squamous cell carcinoma, at least concurrent
chemoradiation is recommended, and in adenocarcinoma the addition
of surgery after concurrent therapy is favored.
Is there a basis for this recommendation, particularly for
cT2N0M0 patients, that we can draw from the published randomized
trials? As previously described, many of the randomized trials
evaluating the role of neoadjuvant chemoradiation are flawed in
some way, whether it is due to study design, failure to accrue, or
outdated staging techniques and treatment. In reviewing the major
randomized phase III studies published over the last 15 years, cT2N0M0 patients were typically eligible for study but the data for
these patients is sparse (49-54). In the study by Burmeister et al.,
neither EUS nor PET was used for initial staging and the published
results did not include stratification according to depth of invasion
(53). Similarly, the trial published by Urba et al. did not utilize
EUS for initial staging and did not include results based on
esophageal invasion (52). Walsh et al. did not employ EUS or
PET, and they did not routinely employ CT as part of the initial
staging (51). The published results of this study do include some
data on pathologic staging, noting that of the 55 patients who underwent
surgical resection alone, 8 (15%) had stage IIA (pT2N0 or
pT3N0 by current staging) disease. The European Organization for
Research and Treatment of Cancer (EORTC) randomized trial relied
on CT scan for clinical staging (49). They utilized a clinical
staging system where the T stage was defined by the maximal
transverse diameter of the esophageal tumor, with < 1 cm being
T1, between 1 and 3 cm being T2, and > 3 cm being T3. Lymph
nodes were considered involved if the maximal transverse diameter
was > 1 cm. While 92/282 (33% ) of patients enrolled were
cT2N0M0, the positive predictive value (PPV) of the disease-stage
classification based on the CT scan for T2 lesions was only 22%
and for N0 was 49% (49). Thus, very few of those felt to be
cT2N0M0 turned out to be pT2N0M0, and very few of these patients
would have been similarly staged with modern staging procedures.
The recently published CALGB 9781 study recommended
EUS or laparoscopy/thoracoscopy as part of initial staging and one
or the other was completed in 43/56 (77%) of patients (54). With
these more modern staging techniques, only 3/56 (5%) of enrolled
patients were felt to have cT2N0M0 disease, so drawing meaningful
conclusions from this trial for these 3 patients is difficult. The
EORTC 40001-22001 study comparing neoadjuvant chemoradiation
followed by surgery vs. surgery alone for clinical stages I-II
closed to accrual in 2004, and may potentially aid in answering this
clinical question (60).
With such a dearth of randomized data for these patients, examination
of single institution retrospective experiences may be of
more value. The M. D. Anderson experience with cT2N0M0 disease
was presented at the International Society of Gastrointestinal
Oncology 2009 Gastrointestinal Oncology Conference (61). They
presented data on 272 patients with cT2N0M0 esophageal cancer,
186 (68%) of which underwent surgery first and 86 (32%) who
first underwent neoadjuvant chemoradiation. They found that patients
receiving preoperative chemoradiation had superior OS when
compared to the group having surgery alone (65.27 months vs.
25.9 months, P=0.006). In addition there was increased time to recurrence
(TTR) among the group receiving neoadjuvant chemoradiation
(52.87 months vs. 18.67 months, P=0.006). Postoperative
complications were more frequently associated with neoadjuvant
chemoradiation but this did not impact on treatment-related mortality.
Of the 186 patients who underwent surgery upfront, 147 (79%)
had a change in their T-stage, 47 (25%) of which were initially
overstaged and 100 (54%) of which were understaged. In addition, of the 186 felt to have N0 and M0 disease clinically, 101 (54%)
have lymph node involvement and 25 (13%) were found to have
M1 disease. They concluded that there is significant stage migration
in patients with cT2N0M0 esophageal cancer treated with upfront
surgery, and that preoperative neoadjuvant chemoradiation in
this population improved outcome (61).
The Cleveland Clinic also recently published their single institution
experience with T2N0M0 esophageal cancer treated
from1987-2005 (62). Out of 742 patients diagnosed with
esophageal cancer during that timeframe, 61 (8%) were determined
to have cT2N0M0 disease. Clinical staging for primary and regional
nodes was accomplished with EUS and designation of cT2
was invasion into the 4th ultrasound (muscularis propria) layer of
the esophagus (62-64), while nodal assessment utilized size, shape,
border, and texture to determine involvement. EUS guided FNA of
suspected lymph nodes was performed in 58 patients. Metastatic
assessment was by CT, with approximately one-third of patients also
undergoing FDG-PET. Of the 61 patients with cT2N0M0 disease,
45 underwent surgery alone, 8 had surgery and postoperative
adjuvant therapy, and 8 underwent some form of neoadjuvant therapy.
Of 53 patients with cT2N0M0 disease who underwent
surgery upfront, on 7 (13%) were found to have pT2N0M0 disease,
while 29 (55%) were initially overstaged and 17 (32%) were understaged.
Of those patients that were overstaged, 3/29 patients
(10%) were found to have in situ disease, 11/29 (38%) had invasive
disease confined to the mucosa (T1a), and 15/29 (52%) had
extension into the submucosa (T1b). Of those patients that were
understaged, 13/17 (76%) had lymph node involvement. In fact,
EUS was found to have 29% sensitivity and 89% specificity for determining
T2N0M0 disease. For those patients overstaged as
cT2N0M0 and undergoing surgery alone, their 5-year OS was similar
to matched controls (69% vs. 63%, P=0.8). For those patients
that were understaged and treated with surgery alone, their survival
was similar to matched patients with >pT2N0M0 treated with
surgery alone (P=.4). And though the numbers were small, patients
understaged as cT2N0M0 had a trend toward improved survival if
they received adjuvant therapy when compared to similar pTNM
staged patients who underwent surgery alone (43% vs. 10%, P=0.
17). For the 8 patients with cT2N0M0 disease receiving induction
therapy, they had a decreased 5-year OS compared to the other
cT2N0M0 patients (13% vs. 52%, P=0.05). Survival of patients
with cT2N0M0 and pT2N0M0 treated with upfront surgery was
similar, approximately 50% at 5 years, but the authors were quick
to caution that this does not imply that clinical staging accurately
reflects pathologic staging (62). In fact, they felt that it was relatively
useless and that the relatively high survival in the overstaged
group balanced out for the poor survival among the understaged
group. The authors recommended that patients with cT2N0M0 tumors
should undergo surgical resection with lymphadenectomy upfront.
And those patient clinically understaged should go on to receive
adjuvant therapy, and for those patients overstaged or appropriately
staged, surgery should serve as definitive therapy (62,65).
|
|
Discussion
With evolving diagnostic tools and techniques such as EUS,
EUS-guided FNA, and FDG-PET, initial clinical staging for
esophageal cancer is becoming increasingly accurate. In recent
years, there has been a movement for initial stage to guide management.
Those patients with disease clinically confined to the mucosa
appear to be adequately treated with EMR or potentially other
endoscopic techniques with diminished morbidity when compared
to esophagectomy. And in those patients known to have more advanced
disease (cT3 and/or N1) multimodality therapy such as
neoadjuvant chemoradiation should be strongly considered based
meta-analyses of randomized data.
The optimal strategy for cT2N0M0, an intermediate risk group
of esophageal cancer patients, remains undefined. As shown, there
is very little randomized data evaluating these patients, and single
institution retrospective data conflict both in terms of results and
treatment recommendations. In evaluating the M. D. Anderson experience,
the PPV of their initial staging in predicting final pathologic
staging was very poor, with the majority of patients being upstaged
both in terms of T-stage and N-stage following surgery. So
it is not surprising that they saw an increase in survival with the addition
of neoadjuvant chemoradiation since the majority of patients
in their analysis likely had T3 and/or N1 at diagnosis. Staging
techniques were not fully addressed, and it would be of interest to
know if EUS, FNA, and FDG-PET were routinely employed. The
group at the Cleveland Clinic had similarly poor PPV for T2N0M0,
with the majority of patients being either under or overstaged.
This included 14/53 (26%) patients treated with surgery upfront
who had either pTisN0M0 or pT1aN0M0 disease. These patients
could have potentially been candidates for less invasive definitive
therapy, such as EMR, if they had been staged properly. In addition,
another 17/53 (32%) treated with upfront surgery were initially
understaged. And while the authors of the study advocated for
adjuvant therapy for patients in this population based on their institutional
experience, neither adjuvant radiation nor chemotherapy
have been shown to definitively improve outcomes in esophageal
cancer in the phase III setting (62,65). And though understaged
cT2N0M0 had similar outcomes with correctly staged more advanced
patients, ideally these patients should have been considered
for neoadjuvant treatment.
The main challenge faced by both of these retrospective series
looking at cT2N0M0 disease was the poor PPV of initial staging.
Initial staging for this subset of patients was futile as final pathologic
stage ran the gamut from in situ to metastatic disease. It is
therefore difficult to draw meaningful conclusions from their results.
Should all patients staged as cT2N0M0 receive neoadjuvant
chemoradiation since there is a substantial risk of understaging?
Or should these patients undergo upfront surgery and await final
pathologic staging to guide any adjuvant therapy, where data for its
utility is limited? Neither strategy seems optimal, and the key may
lie in improved initial staging. As previously reviewed EUS continues to evolve as a staging tool and with more recent incorporation
of FNA, the accuracy of EUS is improving. This along with
advancements such as FDG-PET/CT should allow for more reliable
initial staging. That in combination with publication of modern
clinical trials, such as the EORTC 40001-22001 study evaluating
the utility of neoadjuvant chemoradiation for clinical stages I-II
esophageal cancer, may potentially allow for a more universal algorithm
for this relatively uncommon but clinically significant
group of patients.
|
|
References
Cite this article as: Wagner TD, Khushalani N, Yang GY. Clinical T2N0M0 carcinoma of thoracic esophagus. J Thorac Dis 2010;2:36-42. doi: 10.3978/j.issn.2072-1439.2010.02.01.013
|
