Improving the quality of pretreatment staging in patients with esophageal carcinoma – a fast track study

Abstract

Background. Current guidelines for esophageal cancer recommend series of diagnostic investigations to determine pretreatment TNM stage. When investigations are done sequentially, diagnostic work-up time may be prolonged considerably. Aim of the study was to determine the feasibility and efficacy of a fast track staging strategy within five days after the first consultation. Material and methods. Between 2007 and 2010 all patients presenting with esophageal cancer at the Department of Gastroenterology in a tertiary referral center were prospectively analyzed. At Day 1 all patients underwent computed tomography (CT), endoscopic ultrasound (EUS) and ultrasonography of the neck (US). Results and treatment implications were discussed within a multidisciplinary meeting. This fast track strategy was considered completed successfully if pre-treatment TNM classification was achieved and therapy was proposed to the patient at the outpatient clinic at Day five. In those cases where staging period time was prolonged, the number and type of additional tests were documented including the ensuing time delay. Results. In 111 patients CT, EUS and US were performed in 100%, 88.3% and 97.3% respectively. A final TNM stage and treatment proposal was reached at Day 5 in 60% of the patients. Additional tests were diverse and mainly used to prove local irresectabilty or presence of distant metastasis. Multivariate analysis identified presence of lymphadenopathy (HR 0.25 p = 0.03) and metastasis (HR 0.27 p = 0.03) as significant predictors of not completing the staging period within five days. In 18% of patients overuse of at least one test occurred, most commonly because CT already revealed distant metastasis. Conclusion. Employment of a fast track five day staging strategy in patients with esophageal carcinoma is feasible. Definite TNM stage and treatment proposal can be achieved in 60% of cases, but comes at the expense of test overuse in about one fifth of patients.

After esophageal or gastric cardia cancer has been diagnosed with upper endoscopy (UE) and biopsy staging of the tumor is the next step [1]. An accurate diagnostic work-up is essential to determine further treatment possibilities, including surgical resection, neoadjuvant therapy, curative chemoradiation therapy and various palliative treatment options [2].

Pretreatment TNM-staging is accomplished by a combination of several investigations. The most commonly used staging strategy in esophageal cancer is a sequential approach performing a computed tomography (CT) first. The main strength of this diagnostic modality is its ability to detect distant metastasis, in which case patient management changes radically from a curative to a palliative intent [3,4]. However, the accuracy in staging the extent of local tumor invasion is limited. Endoscopic ultrasound (EUS) has proven its superiority in determining the depth of tumor invasion and loco-regional adenopathy [5–8]. In addition, external ultrasonography of the neck (US) has shown to be superior to CT in detecting supraclavicular lymphadenopathy in esophageal cancer [9]. For these reasons, the current diagnostic guidelines in the Netherlands suggest that a combination of these three investigations (CT, EUS, US) should be applied in all patients [10]. However, data on the feasibility and yield of performing these three diagnostic modalities concurrently rather than sequentially is lacking.

The pretreatment staging period is associated with much uncertainty and distress for patients and their family. For this reason it is highly desirable that staging should be completed as soon as possible.

In this prospective study we determined the efficacy of a fast track staging strategy and the feasibility to perform this within five days in patients with esophageal and gastric cardia cancer.

Material and methods

Design of the study

All consecutive patients with histology proven esophageal, junctional or gastric cardia carcinoma seen at the gastroenterology department of our tertiary referral center from January 2007 to February 2010 were included. Pretreatment staging was commenced within one week after the diagnosis esophageal cancer was established and included CT-scan of thorax and upper abdomen, US of the neck and EUS. In case of suspected lymph nodes seen on US, fine needle aspiration was performed. Tumor was staged according to the sixth edition of TNM classification. All diagnostic tests were performed sequentially at Day 1. At Day 5 pretreatment staging results were discussed in multidisciplinary meeting and finally presented to the patient at the outpatient clinic (Figure 1). Participants of the multidisciplinary meeting were gastroenterologists, surgeons, oncologists, radiotherapists and pathologists. This fast track was successfully completed if definitive TNM classification and therapy was proposed to the patient at the outpatient clinic at Day 5. Within one week patients were seen by their allocated physician to initiate further therapy. In those cases where the staging period time was prolonged by performing additional tests to come to a final TNM classification, the number, kind of additional test, and time of delay were documented.

Figure 1. Fast track staging strategy of esophageal carcinoma.

Data collection and statistical analysis

Start and end date of the staging period, date and results of all diagnostic modalities were prospectively recorded. Final TNM-classification as well as treatment advices were noted. Univariate (χ²) and multivariate analysis were carried out to determine predictive factors for completion of staging after CT, EUS and US. P-value of < 0.05 was considered as significant. Statistical analysis was performed with the Statistical Package for Social Sciences for Windows, version 16.0 (SPSS, Chicago, IL, USA).

Results

Patients

A total of 111 patients were included, 79 men and 32 females with a median age of 63 years. Baseline characteristics are summarized in Table I. Esophageal cancer was demonstrated in 84 patients. Twelve patients suffered from esophagogastric junction cancer and 15 from cardia carcinoma. CT-scan of thorax and upper abdomen was assessed in 111 patients (100%). A total of 98 patients had undergone EUS (88%). Reasons for not performing EUS were: inability to pass the EUS scope due to a severe esophageal stenosis (n = 11), unsuccessful introduction of the EUS-scope (n = 1) and patient refusal (n = 1). US were performed in 108 patients (97%). In two patients CT-scan of the neck was applied instead and in one patient US was not performed for an unknown reason.

Table I. Patient characteristics.

	Number of patients (n = 111)
Median age (range)	63 (41–84) years
Male : Female	79 : 32
Location tumor (%)
Proximal esophagus	12 (11%)
Mid esophagus	23 (21%)
Distal esophagus	49 (44%)
Junction	12 (11%)
Cardia	15 (13%)
Histology (%)
Adenocarcinoma	71 (64%)
Squamous cell carcinoma	40 (36%)

After CT, EUS and US were performed TNM stage was complete in 66 patients (60%). In one patient staging could not be completed because of death due to a massive upper gastrointestinal bleeding, which was not related to any of the investigations. Additional investigations were required in 44 patients. Patients needed either one (n = 29) or two (n = 15) complementary tests. These additional tests are shown in Table II.

Table II. Type and number of additional tests required after CT, EUS and US. Result of tests and median time interval from the fifth day until result of test is described.

	Total No.	Diagnostic yield T4 or M1		Diagnostic yield not T4 or M1
Additional diagnostic test after 5 days in 44 patients	Total No.	No.	Median time until test (range)	No.	Median time until test (range)
EUS with FNA	12	4	17 days (10–39)	8	16 days (10–32)
Bronchoscopy	7	0		7	5 days (1–31)
Ultrasonography of liver	4	0		4	7 days (5–12)
CE ultrasonography of liver	7	4	9 days (4–18)	3	6 days (4–18)
Biopsy of liver lesion	4	3	16 days (7–31)	1	27 days
CT of adrenal gland	7	2	10 days (7–13)	5	6 days (5–18)
Biopsy of adrenal lesion	2	2	25 days (24–25)	0
Repeated cytological puncture of cervical LN	6	0		6	10 days (3–19)
PET-scan	5	1	12 days	4	13 days (6–14)
Diagnostic laparoscopy	1	1	18 days	0
MRI liver	1	0		1	6 days
Pleural fluid aspiration	2	0		2	7 days (7–7)
EBUS with FNA	1	0		1	14 days
Total	59	17	14 days (4–39)	42	8.5 days (1–32)

CE ultrasonography, contrast enhanced ultrasonography; EBUS, endobronchial ultrasonography; FNA, fine needle aspiration; LN, lymph node.

Bronchoscopy was indicated in seven patients because of signs of bronchial ingrowth on CT. Assessment by EUS was incomplete in five of these patients due to severe esophageal stenosis and in the other two patients no signs of bronchial invasion were seen during EUS. Bronchoscopy revealed no ingrowth in all patients.

Outcome

The final TNM stage after pretreatment staging is shown in Table III. T3 carcinomas were observed in 61% of patients, 75% of patients had N1 and distant metastases (M1) were present in 20% of patients. Median time from start of pretreatment staging until definitive TNM stage and treatment proposal at the outpatient clinic was five days (range 5–47 days). In 44 (40%) patients the pretreatment staging period could not be completed at Day 5. In these cases pretreatment staging time was extended with a median of 13 days (range 1–39 days). Total median staging time was significantly longer in patients with M1b disease vs. M0/1a, 17 vs. 5 days (p = 0.002). In uni- and multivariate analysis, besides the presence of distant metastasis (M1b vs. M0/1a, HR 0.38, p = 0.03) and N-stage (N1 vs. N0, HR 0.30, p = 0.04), no other significant predictors of delaying the pretreatment staging period were identified (Table IV).

Table III. Clinical TNM-stage of esophageal and gastric cardia carcinoma.

	Esophageal or gastric cardia carcinoma (n = 111)
cT stage
1	4
2	20
3	68
4	12
x	7*
cN stage
0	25
1	83
2	3
x	0
cM stage
0	82
1a	7
1	22
x	0

*EUS not performed.

Table IV. Univariate and multivariate analysis of patient and tumor related risk factors of completion of staging after CT, EUS and US.

	Total	Univariate analysis HR (95% CI)	P	Multivariate analysis HR (95% CI)	P
Age (yr)			0.57		0.52
< 60	41	1.00		1.00
≥ 60	69	0.80 (0.36–1.76)		0.73 (0.28–1.88)
Gender			0.61		0.30
Male	78	1.00		1.00
Female	32	0.80 (0.35–1.85)		0.58 (0.21–1.62)
Histology			0.81		0.28
AC	71	1.00		1.00
SCC	39	1.11 (0.50–2.46)		2.21 (0.52–9.42)
Location tumor			0.60		0.27
Proximal esophagus	12	1.24 (0.35–4.44)		0.70 (0.11–4.42)
Mid esophagus	22	1.90 (0.66–6.46)		1.40 (0.31–6.22)
Distal esophagus	49	1.00		1.00
Junction	12	1.23 (0.35–4.44)		3.01 (0.66–13.67)
Cardia	15	2.43 (0.68–9.70)		4.27 (0.99–18.30)
T-stage			0.14		0.19
T 1/2	24	1.00		1.00
T 3/4	80	0.48 (0.18–1.29)		0.47 (0.15–1.44)
N-stage			0.02		0.04
N 0	25	1.00		1.00
N 1	85	0.30 (0.10–0.86)		0.27 (0.08–0.95)
M-stage			0.04		0.03
M 0/1a	88	1.00		1.00
M 1	22	0.38 (0.15–0.98)		0.25 (0.07–0.90)

AC, adenocarcinoma; SCC, squamous cell carcinoma.

Accuracy of pretreatment TNM staging

After completion of staging, curative approaches were attainable and proposed in 67 patients (61%). In this subgroup CT, EUS and US were sufficient for adequate staging in 40 patients. Complementary tests were required in 27 patients. Definitive chemoradiotherapy was indicated in 12 patients. Fifty-five patients were scheduled for surgical treatment, including primary resection (n = 35) and preoperative chemoradiotherapy followed by surgery (n = 20). Primary surgical resection could not be carried out in four patients. In two patients hepatic metastases were found during laparotomy. This was not observed on preoperative CT-, PET-scan and ultrasound. In the third patient surgery could not be performed because of extensive tumor fixation to the trachea. In this case adequate preoperative endosonographic determination of the T stage was not possible because of a luminal stenosis which could not be passed. CT-scan suggested tracheal ingrowth, however bronchoscopy was normal. In the fourth patient ingrowth into aorta and visceral pleura made resection impossible. This observation was not visible on preoperative CT-scan and EUS was not completed because of tumor related stenosis.

Surgery was cancelled in four of 20 patients receiving neoadjuvant therapy, because of the recognition of hepatic metastases intraoperatively (n = 1, not detected on preoperative CT-scan) and the development of cutaneous (n = 2) and leptomeningeal (n = 1) metastases during chemoradiotherapy.

Overuse

As a consequence of performing staging investigations concurrently rather than sequentially, some investigations may have been redundant. In total, overuse of at least one test did occur in 20 patients (18%) (Table V). The reason in most cases was CT-scan (n = 13) revealing distant metastases, with overuse of EUS (n = 10) and US (n = 13). In one subject EUS showed minimal traces of ascites, followed by a positive diagnostic laparoscopy. In another two patients CT and EUS were not of significant clinical use, because distant cervical lymphogenic metastasis was already proven by US-guided cytological puncture. In one patient a histological proven esophageal metastasis was demonstrated proximal from the primary gastric cardia tumor. Theoretically, the diagnostic value and therapeutic consequence of CT, EUS and US are limited in this case. Furthermore, two patients underwent EUS, while CT and US guided cytological puncture showed distant metastasis. Lastly, US was unnecessary because both CT and EUS revealed metastasized cancer in one patient.

Table V. Overuse of diagnostic investigations because of signs of distant metastasis demonstrated by other modalities.

		Overuse
Diagnostic modalities leading to M1 stage	Number of patients	CT (n = 4)	EUS (n = 15)	US (n = 16)
CT	13	–	10	13
EUS	1	1	–	1
US	2	2	2	–
UE	1	1	1	1
CT and US	2	–	2	–
CT and EUS	1	–	–	1
Total	20	4	15	16

CT, computed tomography; EUS, endoscopic ultrasound; UE, upper endoscopy; US, ultrasonography of the neck.

Discussion

Esophageal carcinoma is associated with a high disease and treatment related morbidity and mortality. To commence adequate therapy in a timely fashion both accurate pretreatment staging as well as reduction of pretreatment staging time are essential. Current guidelines for pretreatment staging of patients with esophageal or cardiacarcinoma recommend US of the neck, EUS and CT of thorax and abdomen. We found that in 60% of the patients a final TNM classification can be achieved within five days by performing these three diagnostic modalities concurrently instead of sequentially. Prolongation of the pretreatment staging time was significantly associated with the suspicion of distant metastasis. In 40% of the patients additional tests were needed. In about one third of these patients metastatic disease was proven with a median extension of the pretreatment staging time of 13 days. In two-thirds of the patients in whom additional tests were negative, pretreatment staging time was delayed with a week.

The outcome of this fast-track protocol could be further optimized if EUS-FNA would have been performed at the time of the diagnostic EUS. However, with tight slots to perform EUS this would disrupt the program and appeared only necessary in 10% of the patients. From our data it seems that the most profit can be achieved by restricting the indication to perform a bronchoscopy for suspicion of tracheal ingrowth and repetition of cervical lymph node punctures. In none of our patients these particular tests led to a change in TNM classification and only caused a delay in the start of therapy. In our series the indication to perform bronchoscopy was compression of the tracheobronchial tree as demonstrated by CT-scan. It is well known that this particular feature is less often a true sign of infiltration by esophageal tumor growth [11,12]. Tracheal infiltration is mainly seen in those patients with a fixed pars membranacea.

The preoperative diagnostic EUS, CT- and PET-CT failed to predict irresectability in four patients, where hepatic metastases and a fixed tumor were found during laparotomy. Unfortunately these surgical operations could not have be prevented because accurate diagnostic modalities for small metastatic lesions and ingrowth in surrounding anatomical structures organs, in particular in those cases in which the EUS-scope cannot pass, are lacking. Although PET is commonly used by some, mainly for detecting metastasis, results from our own center have indicated that PET is not of additional value for pretreatment TNM staging [13]. However, for the ultimate verdict about the value of PET-CT in the preoperative work-up of esophageal cancer patients additional randomized, prospective studies evaluating its value and cost effectiveness are required.

By performing all three investigations simultaneously (CT, US, and EUS), overuse of some tests occurred in 20% of the patients. This main reason was that the initial CT-scan already showed metastases. This overuse could be reduced by adopting a conditional staging strategy with a specific sequential order of diagnostic tests (CT, US, and EUS) [14]. However, the actual reduction in the number of tests is relatively low and does not seem to outweigh the disadvantages of logistic complexity and associated delay of starting therapy.

Total staging time was significantly delayed in those patients with more progressive disease. Although an extended period of uncertainty is inconvenient to patients and should be avoided when possible, in patients with a higher likelihood of disseminated disease it seems less crucial as therapy with a potentially curative intent is not delayed. Nevertheless, as already eluded to previously it should not be forgotten that of all patients with a suspicion of advanced disease in whom additional investigations were performed, advanced disease was only proved in 30%. This clearly calls for appropriate indications when conducting additional investigations and careful planning without losing too much time.

There are several limitations of this staging protocol. In our study population nearly 80% of patients did not have distant metastasis, an observation which is comparable to data from a recent Dutch cohort study [4]. It should be emphasized that this fact could have been influenced by selection bias. Our institution is a tertiary referral hospital, a high volume center for esophageal cancer surgery. It is imaginable that patients with established metastatic disease were not referred to our center for further analysis. This may be a reason for the high proportion of patients with only local disease in our cohort.

As to the feasibility of this staging strategy, CT and US were accomplished in nearly all patients. EUS however, could not be completed in all patients. In most cases, this was caused by a proximally located stricturing tumor that could not be traversed during EUS. Although some authors have suggested in such cases to dilate the stenosis in order to pass a EUS-scope, this policy has not been routinely adopted in our institution because of the increased risk of perforation [15–18].

Ideally, this kind of fast-track staging protocol should be part of an integrated process entailing diagnostics, multidisciplinary discussion, patient consultation and information, (multimodality) therapy and careful post-treatment monitoring with each subsequent step swiftly following the preceding step. This includes reduction of waiting time for surgery, chemo- and/or radiotherapy, as well as standardized pathways for postoperative and palliative care.

In conclusion, in patients with esophageal and cardia carcinoma the employment of a fast track five day staging strategy which includes a set of combined diagnostic investigations, achieves a definite TNM stage and treatment proposal in 60% of cases. In particular, patients with local disease benefit from this strategy. In our opinion, the benefit outweighs the overuse of at least one of these tests in 18% of patients. We believe that implementation of this fast track staging strategy contributes to a further optimization of the medical care of patients with esophageal or gastric cardia cancer.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.