Treatment response and prognosis of patients after recurrence of esophageal cancer☆☆☆★

Patients and treatment for primary tumor 

For inclusion in this study, patients were required to have presented at the Department of Surgery, Chiba University Hospital, between January 1990 and December 1999 with histologically proven primary esophageal SCC and to have been treated by potentially curative operation with extended lymphadenectomy. A review of clinical records identified 258 patients who satisfied entry criteria. Radical esophagectomy was performed with 3-field lymphadenectomy (neck, mediastinal, and abdominal lymphadenectomy)2 for 144 patients and 2-field lymphadenectomy (mediastinal and abdominal lymphadenectomy) for 114 patients. At the time of operation, TNM classification16 was determined by pathologic examination (Table I). A total of 97 patients received perioperative adjuvant therapy, including chemoradiotherapy in 51 patients, radiotherapy in 19, and chemotherapy in 27. Among 258 patients, 95 patients had recurrent tumors by the end of 2000 (mean follow-up was 22 months, range: 2-113) including 83 males (87%) and 12 females (13%), with a mean age of 61 years (range: 41-79). At the time of diagnosis of recurrent tumors, 89 patients (94%) had a performance status (PS) on the Zubrod scale of 0 to 1 and 6 patients (6%) had a PS 2. Patients who had a new secondary primary esophageal cancer in the remnant esophagus were excluded from this study. Serum samples were obtained at the time of recurrence and stored at −80°C until assay. Patient recruitment and sample collection were performed within the guidelines of protocols approved by the institutional review boards. Informed consent was obtained from all patients.

Table I. Comparison of patient characteristics with or without recurrence

		Variables	No. of patients without recurrence (n = 163)	No. of patients with recurrence (n = 95)	P*
	Location	Upper	24	25	.018
		Lower	139	70
	Tumor depth	T1-T2	124	27	<.001
		T3-T4	39	68
	N factor	N0	59	22	<.001
		N1	104	84
	M factor	M0	126	61	.017
		M1	37	34
	*Fisher exact probability.

Follow-up schedule and diagnosis of recurrence 

After the initial esophagectomy, all patients underwent clinical examination and imaging studies on a regular basis. Eighty-four patients died during the course of this study, which included a median follow up period for survivors of 36 months. Postoperative recurrence was identified from positive findings on successive 3-month examinations of tumor markers (carcinoembryonic antigen, squamous cell carcinoma antigen, and cytokeratin 19 fragment), 6-month examinations of ultrasonography, and annual computed tomography. All recurrent tumors were evaluated by imaging studies and were defined as “recurrent” if tumor sizes measured greater than 1 cm.

Treatment plan for recurrent tumors 

Treatment strategies in the management of recurrent tumors were as follows: (1) patients who had not received perioperative radiation therapy or nonradical radiation therapy (less than 40 Gy) underwent chemoradiation therapy; (2) patients who had impaired renal function (creatinine clearance < 40 L/day) or leucopenia (<3000 cells/mm3) underwent radiation therapy alone; (3) patients who had received perioperative radical radiation therapy (60 Gy) underwent chemotherapy alone; (4) patients who had distant organ metastasis underwent chemotherapy alone; (5) patients who had recurrent tumors in a limited region underwent salvage operation or chemoradiotherapy; and (6) patients who rejected treatment received no treatment. Among 95 patients with recurrent tumors, 76 underwent nonsurgical salvage treatment, 7 had reoperation, and 12 received no treatment. Of the nonsurgical treatment group, 47 patients received chemoradiotherapy; 17, chemotherapy alone; and 12, only radiotherapy. The chemotherapy was performed in 2 cycles with a 3-week interval between cycles, consisted of cisplatin 70 mg/m2/day intravenous administration on day 1, 5-FU 700 mg/m2/day in continuous intravenous infusion for 5 days (day 1 to day 5), and leucovorin 27 mg/body/day in intravenous infusion for 5 days (day 1 to day 5). The chemoradiotherapy consisted of a radiotherapy dose of 2 Gy/day, was initiated on day 1 of chemotherapy, and continued daily for 5 days/wk for 4 or 6 weeks, totaling 40 or 60 Gy. Patients who had received adjuvant radiotherapy underwent 40 Gy of radiotherapy. Concurrent chemotherapy consisted of cisplatin 10 mg/m2/day intravenous administration and 5-FU 500 mg/m2/day in continuous intravenous infusion for 5 days (day 1 to day 5). The radiation therapy consisted of a dose of 2 Gy/day and continued daily for 5 days/wk for 4 or 6 weeks, totaling 40 or 60 Gy.

Definition of response to treatment for recurrent tumor 

Reevaluation of the recurrent tumor was performed by imaging examinations (such as computed tomography and upper gastrointestinal contrast studies) 2 weeks after completion of treatment. The response to treatment was categorized as either a complete response, partial response, stable disease, or progressive disease. This evaluation was on the basis of a comparison of initial and pretreatment imaging studies, and posttreatment imaging findings. Complete response was defined as “the disappearance of the tumor.” Partial response was defined as “a reduction of 50% or more of the tumor volume.” The sum of the perpendicular diameter of the lesion was used to calculate tumor volume. Stable disease was defined as “less than 50% decrease or less than 25% increase of tumor volume.” Progressive disease was defined as “more than 25% increase of tumor volume.”

Enzyme immunoassay for S-p53-Abs and S-CRP 

S-p53-Abs were assessed by enzyme-linked immunoabsorbent assay with the anti-p53 EIA kit (Pharmacell, Paris, France). In brief, the samples were added to the wells of a microtiter plate coated with either wild-type human p53 or a control protein and incubated for 1 hour. A conjugated second antibody was added, incubated for another hour, and then the substrate solution was added. After addition of the stop solution, the color reaction was immediately measured by the absorption at 450 nm using a photo spectrometer. All samples were considered positive at an optic density above the lowest positive control sample. A p53 index [(specific signal of the sample)/(specific signal of low positive control)] of 1.1 was selected as the cut-off level. The specificity of this assay was higher than 99% (206 of 208 healthy persons were seronegative), according to the manufacturer's instruction.

S-CRP was measured before treatment with a latex CRP immunodetection kit (CRP-Latex Seiken, Tokyo, Japan). Albumin concentration (Alb) was also measured before treatment. The cut-off values were 1.0 mg/dL for S-CRP and 3.5 mg/dL for Alb. The specificity of the S-CRP value of 1.0 mg/mL was 99% (624 of 632 healthy persons had levels less than 1.0 mg/dL).

Statistical analyses 

Fisher exact probability test was applied to determine the significance between the 2 groups. Survival probabilities were calculated by the product limit method of Kaplan and Meier. Differences between groups were tested using the log-rank test. The influence of each clinicopathologic variable on survival was assessed by Cox proportional hazards model. All statistical analyses were carried out using software (Statview v. 5.0 for Macintosh, SAS Institute, Inc, Cary, NC) and all P values were considered to be statistically significant if < .05.

Pattern and time of recurrence 

Among all 258 patients, the overall hospital mortality was 3% and the overall 5-year survival was 48%. Follow-up evaluations revealed tumor recurrence in 95 patients. Among all clinicopathologic variables, tumor location, tumor depth, nodal status, and distant metastases were significantly associated with the rate of recurrence (Table I). Among patients with recurrent disease, the median disease-free survival was 7 months (range, 1-57). Of the patients with recurrent disease, 39% had the recurrence within 6 months, whereas 31% had a recurrence between 7 and 12 months postoperatively. The most common recurrence occurred in lymph nodes (45 patients, 47%) followed by distant organ metastasis (35 patients, 37%) that included lung, liver, or bone. Among the 45 patients who had lymph node recurrence, sites of recurrence were as follows: neck (n = 18), mediastinal (n = 21), and abdominal (n = 6). Within 6 months after operation, 11 of 43 lymph node recurrences and 17 of 35 organ metastases were detected (P = .031). After 2 years, lymph node recurrences were detected in 7 patients and organ metastasis in only 2 patients. Recurrences were diagnosed by computed tomography in 60 patients (63%) and by ultrasonography in 17 patients (18%).

Status of S-p53-Abs and S-CRP 

A total of 83 patients were evaluated for prognosis after treatment for recurrent tumors. Status of S-p53-Abs was able to be analyzed in 42 of 83 patients; 19 patients (45%) were negative and 23 patients (55%) were positive. The mean and median values of pretreatment S-CRP were 3.1 mg/dL and 1.0 mg/dL (range, 0.1 to 28). When using a cut-off value of 1.0 mg/dL for S-CRP, 39 patients (47%) revealed low levels and 44 patients (53%) revealed high levels.

Response to treatment and prognosis 

Of 59 patients who underwent either radiotherapy or chemoradiotherapy, 53 received a total radiation dose of 40 Gy or more. Among the 47 patients who underwent chemoradiotherapy, 6 patients received 2 cycles of chemotherapy and 41 patients received 1 cycle. Leucopenia (less than 3000 cells/mm3) was observed in 31 patients and renal dysfunction (creatinine more than 2.0 mg/mL) was observed in 3 patients. No treatment-related deaths were observed. The median survival in patients after recurrence was 8 months (range, 2-77). The overall 1-year survival after recurrence was higher in the treatment group versus the nontreated group (31% vs 0%, P < .001) (Fig, A).

View Large Image Download to PowerPoint

Fig. Survival curves of 95 patients after recurrence of esophageal carcinoma. A, Treatment group (n = 83) and nontreatment group (n = 12). B, Seventy-six patients with recurrent tumors who underwent nonsurgical treatment. The responder group included complete responses (n = 5) and partial responses (n = 21). The nonresponder group included stable disease (n = 26) and progressive disease (n = 24). P value was calculated by the log-rank test.

Table II. Predictive variables for nonsurgical treatment response in the 76 patients with recurrent esophageal cancer

	Variable		No. of nonresponders (n = 50)	No. of responders (n = 26)	Pd
	Type of recurrencea	Lymph node	17	19	.014
		Organ	23	7
	Type of treatmentb	CRT or RT	33	20	.238
		CT	17	6
	Perioperative adjuvant therapy	(−)	23	21	.020
		(+)	20	5
	Time of recurrence	<1 y	37	9	.001
		≥1 y	13	17
	No. of recurrent tumors	1	13	22	<.001
		≥1	37	4
	Albumin	<3.5 mg/dL	13	2	.049
		≥3.5 mg/dL	37	24
	CRP	<1.0 mg/dL	20	19	.006
		≥1.0 mg/dL	30	7
	S-p53-Absc	(−)	10	11	.010
		(+)	15	2

aPleural and local recurrence were excluded; bCRT, chemoradiotherapy; RT, radiotherapy; CT, chemotherapy; cS-p53-Abs were evaluated in patients who underwent operation between 1997 and 2000; dFisher exact probability.

A high level of S-CRP (>1.0 mg/dL) was useful to identify the nonresponder group with a sensitivity of 60% and specificity of 73%. The positive predictive value was 81%, and the negative predictive value was 49%.

Univariate and multivariate analysis for prognostic variables 

Using univariate analysis, the type of recurrence, time of recurrence, number of recurrent tumors, Alb, S-CRP, and S-p53-Abs all yielded a significant estimate of prognosis (Table III).

Table III. Univariate analysis of the survival of the 83 patients with recurrent esophageal cancer after treatment

	Variable		No. of patients (n = 83)	1-y survival (%)	Pd
	Type of recurrencea	Lymph node	38	35	.038
		Organ	33	23
	Type of treatmentb	CRT, RT, CT	76	32	.065
		Operation + CRT	7	69
	Perioperative adjuvant therapy	(−)	44	26	.727
		(+)	51	21
	Time of recurrence	<1 y	49	15	.010
		≥1 y	34	41
	No. of recurrent tumors	1	38	45	<.001
		≥2	45	9
	Albumin	<3.5 mg/dL	20	11	.026
		≥3.5 mg/dL	63	28
	CRP	<1.0 mg/dL	39	33	.002
		≥1.0 mg/dL	44	16
	S-p53-Absc	(−)	19	65	.003
		(+)	23	22

aPleural and local recurrence were excluded; bCRT, chemoradiotherapy; RT, radiotherapy; CT, chemotherapy; cS-p53-Abs wee evaluated in patients who had recurrent tumor between 1997 and 2000; dlog-rank test.

Table IV. Multivariate analysis of survival of patients after recurrence of esophageal cancer

	Model A
	Variables	P	Hazard ratio	95% CI
	Type of recurrence	.806	1.07	0.62-1.87
	Time of recurrence	.015	2.02	1.15-3.55
	No. of recurrent tumors	<.001	2.82	1.55-5.13

	Model B
	Variables	P	Hazard ratio	95% CI
	Time of recurrence	.089	2.51	0.87-7.25
	No. of recurrent tumors	.262	1.76	0.66-4.74
	S-p53-Abs	<.001	10.6	2.76-40.00
	Serum albumin	.663	1.33	0.37-4.71
	CRP	.007	5.05	1.56-16.39

CI, confidence interval.

Salvage operation for recurrent tumors 

Salvage operation was performed in 7 patients who had recurrent tumors in a limited region. Of these patients, 6 had lymph node recurrence in the neck region and 1 had mediastinal lymph node recurrence. All 7 patient received chemoradiotherapy after resection of recurrent tumors and survived with an acceptable quality of life. Survival was greater than 1 year for 5 patients, including 3 who survived more than 3 years, but none were ultimately cured.