UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Carcinogenesis and treatment of adenocarcinoma of the oesophagus and gastric cardia

Hulscher, J.B.F.

Link to publication

Citation for published version (APA):Hulscher, J. B. F. (2002). Carcinogenesis and treatment of adenocarcinoma of the oesophagus and gastriccardia.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 24 Mar 2021

Chapter 4

Extended transthoracic resection versus limited transhiatal resection for

adenocarcinoma of the mid-/distal oesophagus and gastric cardia: results of

a randomised study

Jan B.F. Hulscher1, Johanna W. van Sandick', Angela G.E.M, de Boer2, Bas P.L.

Wijnhoven3, Jan G.P. Tijssen4, Paul Fockens5, Peep F.M. Stalmeier6, Fiebo J.W. Ten Kate7,

Herman van Dekken8, FJuug Obertop', Hugo W. Tilanus3, J. Jan B. van Lanschot'

Academic Medical Center at the University of Amsterdam: Departments of Surgery , Medical

Psychology2, Cardiology4, Gastro-enterology5, and Pathology

Erasmus University Hospital Rotterdam: Departments of Surgery" and Pathology8

RADIAN and Medical Technology Assessment, Nijmegen6

Accepted for publication N Engl J Med.

Abstract

Background. Controversy still exists about the optimal surgical approach for oesophageal carcinoma.

Methods. Between April 1994 and February 2000, 220 patients with adenocarcinoma of the mid-

/distal oesophagus or gastric cardia involving the distal oesophagus were randomised for transhiatal

oesophagectomy or transthoracic oesophagectomy with extended en-bloc lymphadenectomy. Main

end-points were overall survival and disease-free survival. Early morbidity and mortality, quality

adjusted life years and cost-effectiveness were determined.

Results. 106 patients were randomised for transhiatal oesophagectomy, 114 for transthoracic

oesophagectomy. Demographic and tumour characteristics were comparable. Peri-operative morbidity

was higher after transthoracic oesophagectomy, leading to prolonged ventilation time, ICU/MCU-stay

and hospital-stay. Hospital mortality was two percent and four percent resp.(p=0.45).

Radicality of surgery and pTNM stages were comparable. At the end of follow-up 142 patients had

died, 74 (69 percent) after transhiatal resection and 68 (60 percent) after transthoracic resection.

(p=0.12) Although the difference in survival was not statistically significant, there was a trend

towards a survival benefit of the extended approach at five years: disease-free survival was 27 percent

versus 39 percent, while overall survival was 29 percent versus 39 percent. Costs of treatment were

23,809 euro and 37,099 euro resp.

Conclusion. Transhiatal oesophagectomy carries lower morbidity than transthoracic oesophagectomy

with extended en-bloc lymphadenectomy. Although median overall, disease-free, and quality-adjusted

survival are not statistically different, there is a trend towards an improved long-term survival at five

years favouring the extended approach. Further follow-up of the cohort will possibly demonstrate

whether there is a long-term survival benefit to the transthoracic approach.

74

Introduction

The incidence of adenocarcinoma of the distal oesophagus and gastric cardia is rising, while long-term

survival after 'curative' surgery is only 20 percent.1,2 Whereas surgery is generally considered as

offering the best chance for cure, opinions on how to improve survival rates by surgeiy are conflicting.

One strategy aims at decreasing early post-operative risks by using a limited cervico-abdominal

(transhiatal) oesophagectomy without formal lymphadenectomy. The other intends to improve long-

term survival by performing a combined cervico-thoraco-abdominal resection, with wide excision of

the tumour and peri-tumoural tissues, and extended lymph node dissection of the posterior

mediastinum and the upper abdomen (transthoracic oesophagectomy with extended en-bloc

lymphadenectomy). "

Aim of the study was to determine whether transthoracic oesophagectomy with extended en-bloc

lymphadenectomy sufficiently improves overall, disease-free and quality-adjusted survival in

comparison to transhiatal oesophagectomy, to compensate for the possible increase in peri-operative

morbidity, mortality and costs of treatment.

Methods

The study was performed in two academic centres both performing over 50 oesophagectomies per

year. Patients with histologically confirmed adenocarcinoma of the mid-/distal oesophagus or gastric

cardia involving the distal oesophagus without evidence of distant metastases (including histologically

proven tumour positive cervical nodes or irresectable celiac lymph nodes) and/or local irresectability

were randomised for transhiatal oesophagectomy or transthoracic oesophagectomy with extended en-

bloc lymphadenectomy between April 1994 and February 2000.

Patients were older than 18 years and in adequate physical condition (American Society of

Anesthesists Class I or II). Exclusion criteria were previous/coexisting cancer, previous

gastric/oesophageal surgery, neoadjuvant chemo- or radiation therapy, recurrent laryngeal nerve palsy

and the impossibility to construct a gastric tube.

Pre-operative diagnostic work-up consisted of endoscopy with histological biopsy, endosonography,

external sonography of abdomen and neck (with cytological biopsy if indicated), chest radiograph,

indirect laryngoscopy, and bronchoscopy if tumour ingrowth in the upper airway was suspected. CT-

scans were performed on indication only. Patients with cardia carcinoma underwent laparoscopy with

laparoscopic ultrasonography.6 PET-scans were not performed. After written informed consent,

patients were randomised two to four weeks before surgery. The randomisation was stratified for

75

hospital and tumour localisation (oesophagus versus cardia by endoscopy). No forms of blocking

were used within each of the four strata. Oesophageal carcinoma was defined as bulk of the tumour

situated in the oesophagus and/or the presence of Barrett's mucosa, while cardia carcinoma was

defined as bulk of the tumour in the cardia without Barrett's mucosa being present.

Surgery and Pathology

Operations were performed by or under direct supervision of a surgeon experienced in oesophageal

surgery (HO, HWT, JJBvL). During transhiatal oesophagectomy the oesophagus was dissected

under direct vision through the widened hiatus of the diaphragm, up to the inferior pulmonary vein.

The tumour and its adjacent lymph nodes were dissected en-bloc. A three cm wide gastric tube was

constructed. The left gastric artery was transected at its origin with resection of local lymph nodes.

Celiac lymph nodes were dissected only when clinically suspicious.

After right-sided mobilisation of the cervical oesophagus, the intra-thoracic, normal oesophagus was

bluntly resected distal-wards. Oesophago-gastrostomy was performed in the neck, without cervical

lymphadenectomy.

Postero-lateral thoracotomy was the first step in transthoracic resection with extended en-bloc

lymphadenectomy. Thoracic duct, azygos vein, ipsilateral pleura and all peri-oesophageal tissue in the

mediastinum were dissected en-bloc. The specimen included the following lymph nodes (en-bloc):

lower and middle mediastinal, subcarinal, and right-sided paratracheal. The aorto-pulmonary window

nodes were dissected separately. Via a midline laparotomy paracardiac, lesser curvature, left gastric

artery (along with lesser curvature), coeliac trunc, common hepatic -, and splenic artery nodes were

dissected, and a gastric tube was constructed. The cervical phase was identical to the transhiatal

procedure, but a left-sided approach was used.

In both procedures the origin of the left gastric artery was marked. Subcarinal nodes were marked

separately in case of a transthoracic resection. Peri-oesophageal tissue and lesser omentum were

palpated on the presence of lymph nodes, and subsequently dissected. Separate lymph nodes were

marked by location in separate boxes, cut in two with both sides stained with hematoxin-eosin.

Pathologic grading was performed by or under supervision of a senior gastroenterological pathologist.

(FJWtK, HvD, GJAO) Tumours were staged according to the UICC '97 pTNM classification. Cardia

carcinoma and distal oesophageal carcinoma were considered one clinical entity.7,8,9

Early post-operative complications were prospectively scored by the study-coordinators.(JBFH, JWvS,

BPLW) Epidural anaesthesia was applied postoperatively to minimise pulmonary complications.

76

Follow-up and Quality-Adjusted Life Years

All patients were seen at the outpatient clinic at three to four months' intervals during the first two

years, and every six months thereafter for three more years. After five years, follow-up was obtained

by telephone from the patient or family practitioner. Recurrence of disease was diagnosed on clinical

grounds. However, when a relapse was surmised, radiologic, endoscopic or histologic confirmation

was sought for. Recurrent disease was divided into loco-regional (upper abdominal or mediastinal) and

distant recurrent disease (including cervical recurrence). Overall survival and disease-free survival

were the main endpoints of the study.

Survival was quality-adjusted by calculating quality-adjusted life years (QALY's).10 To calculate a

QALY, for all 220 patients, the duration of living in a certain state of health was obtained from the

clinical data file and multiplied by a factor representing the quality or 'utility' of that health state.

Seven health states were identified: 1) hospitalisation immediately after oesophagectomy without

complications, 2) hospitalisation immediately after oesophagectomy with early post-operative

pneumonia, 3) early recovery at home, 4) living without recurrent disease, 5) living with loco-regional

recurrent disease, 6) living with distant recurrent disease, and 7) living after surgery for an irresectable

tumour. In a single interview 3 to 12 months post-operatively, utilities for these seven hypothetical

health states were obtained in a subsample of 48 out of 59 eligible consecutive recurrence-free

patients, interviewed between January 1997 and March 1998. The utilities were elicited with the

choice based Standard Gamble method, using a probability wheel.10 The whole study cohort was

subsequently used for the QALY analysis.

Cost-Effectiveness Analysis

Costs were defined as the volumes of resources used multiplied by the prices per unit of resource.

They consisted of direct medical costs, direct non-medical costs and indirect costs. Direct medical

costs included pre-operative costs, in-hospital costs of primary surgical treatment and medical costs

during follow-up. Direct non-medical costs included expenses for the patient, indirect costs included

absenteeism from work.

Data on resource use during hospital treatment were collected prospectively from the hospital

information system. Other volumes were assessed by a self-report cost-questionnaire, which was sent

at baseline and three months after surgery.

The (1998) prices (in euros, 1 euro = 1 US dollar) of the operation per minute, Intensive Care

Unit/Medium Care Unit (ICU/MCU) stay per day, and stay at the surgical ward per day were assessed

on the basis of real-cost calculations. Other prices were assessed following the Netherlands

guideline for cost research. ~ Finally, incremental costs per incremental QALY were computed.

77

Statistics

To detect an estimated improvement of the median survival from 14 months to 22 months,

corresponding with an increase of the two-years survival rates from 30% to 45% for patients

undergoing transthoracic resection with extended en-bloc lymphadenectomy, 220 patients had to be

enrolled using a significance level of 0.05 (two-sided) and a power of 0.90.

Chi-square or Fisher Exact tests were used to compare categorical data, while Student's t-test or Mann

Whitney U-tests were used for continuous data. All reported P-values are two-sided. P-values of <

0.05 were considered statistically significant. Survival curves were calculated from the time of

randomisation until death (of any cause) or until the time of the last visit (censored). Disease-free

survival was related to the time of first relapse or death of any cause, or the time of the last visit

without previous relapse (censored). Survival curves were constructed using the Kaplan-Meier

method, and the log-rank was used to test for significance.

Patients with distant metastases and/or locally irresectable tumour detected during operation were

included in the (disease-free) survival analysis even when the surgeon refrained from resection or

performed a different resection (intention to treat principle). In a separate analysis the disease-free

interval was studied only in patients undergoing R0-resection and leaving the hospital alive, because

in patients in whom there is no (microscopic) tumour residue the possible benefit of an extended

lymphadenectomy might be greater.

Results

Between April 1994 and February 2000, 220 of 263 eligible patients were randomised. There were no

baseline differences between the groups regarding demographic or tumour characteristics (table I).

Complete endosonography was possible in 88 percent of patients. The mean time between

randomisation and surgery was three weeks in both groups.

Ninety-three patients in the transhiatal group (88 percent) and 109 patients in the transthoracic group

(96 percent) underwent the planned procedure. (p=0.08) One patient did not undergo resection after

massive aspiration, while local irresectability and/or distant metastases (detected during the operation)

precluded resection in another 11 patients. Total gastrectomy was performed in three patients, while

conversion took place in three more patients. Transhiatal resection was associated with a shorter

median operative time (3.5 his versus 6 hrs resp, p<0.001) and a lower median per-operative blood-

loss ( 1.0 L versus 1.9 L, pO.OO 1 ). There was no per-operative mortality. Transhiatal resection was

associated with less pulmonary complications and chylous leakage, reflected in a shorter ventilation

78

time, ICU/MCU-stay and hospital-stay, (table 2). Overall in-hospital mortality was 3 percent: 2

percent (2 patients) in the transhiatal group and 4 percent (5 patients) in the transthoracic group.

pTNM stages were comparable, with a tendency for more stage IV tumours in the transthoracic

group: 15 percent versus 7 percent (table 3). A mean of 16 ± 9 nodes was identified after transhiatal

resection versus 31 ± 14 after transthoracic resection. One hundred patients undergoing transthoracic

resection (87.8 percent) had 15 or more lymph nodes identified in the resection specimen. There was

no difference in radicality of surgery (residual tumour classification) between the two groups.

Follow-up continued until July 2002, ensuring a minimum potential follow-up of two and a half years.

Follow-up was complete in all patients. The median potential follow-up of all patients was 4.7 years

(range: 2.5 - 8.3). Recurrent disease developed in 62 (58 percent) and 57 (50 percent) patients after

transhiatal and transthoracic resection respectively. Loco-regional recurrence occurred in 14 percent

and 12 percent resp. of patients, distant recurrence in 25 percent and 18 percent resp., and a

combination of both in 18 percent and 19 percent resp.(p=0.6). Median disease-free interval was 1.4

years (95 percent confidence interval 0.8 to 2.0) versus 1.7 years (95 percent confidence interval 0.7 to

2.7) (p= 0.15, figure 1). Estimated five-year disease-free survival was 27 percent (95 percent

confidence interval 19 percent to 38 percent) after transhiatal resection versus 39 percent (95 percent

confidence interval 30 percent to 48 percent) after transthoracic resection. The 95 percent confidence

interval of the difference ranges from -1 percent to 24 percent.

At the end of follow-up 142 patients had died, 74 (69 percent) in the transhiatal group and 68 (60

percent) in the transthoracic group (p=0.12). Thirteen patients died from not cancer-related causes.

Median survival after transhiatal resection was 1.8 years (95 percent confidence interval 1.2 to 2.4),

versus 2.0 years (95 percent confidence interval 1.1 to 2.8) after transthoracic resection with extended

en-bloc lymphadenectomy (p=0.38, figure 2). Estimated overall five-year survival is 29 percent (95

percent confidence interval 20 percent to 38 percent) after transhiatal resection versus 39 percent (95

percent confidence interval 30 percent to 48 percent) after transthoracic resection. The 95 percent

confidence interval of the difference ranges from -3 percent to 23 percent.

Median QALY after transhiatal resection was 1.5 (95 percent confidence interval 0.8 to 2.1) versus 1.8

(95 percent confidence interval 1.1 to 2.4) after transthoracic resection with extended en-bloc

lymphadenectomy (p=0.26, figure 3).

Mean total costs were 23,809 euros for transhiatal resection and 37,099 euros for transthoracic

resection with extended en-bloc lymphadenectomy (table 4). Therefore, costs of treatment for

transthoracic resection were 56 percent higher. Costs of surgery, costs of ICU/MCU stay, and costs of

in-hospital stay at the regular surgical ward were the largest contributors to the overall costs (table 4).

Incremental costs were 41,531 euros per QALY gained.

79

Discussion

The present study was performed to investigate whether transthoracic oesophagectomy with extended

en-bloc lymphadenectomy could sufficiently improve the overall and disease-free survival and the

quality-adjusted survival in comparison to transhiatal oesophagectomy to compensate for the possible

increase in peri-operative morbidity, mortality and costs of treatment. Hospital morbidity (but not

mortality) was significantly higher after transthoracic resection, leading to a prolonged MCU/1CU and

hospital stay and increased costs. The disease-free and the overall survival curves were comparable

early on but diverged (without reaching statistical significance) after three years, favouring the

extended resection.

In the present study, almost 90 percent of patients undergoing an extended en-bloc lymphadenectomy

had 15 or more lymph nodes removed and identified, indicating that the lymphadenectomy had been

adequate.15 Postero-lateral thoracotomy gives wide access to the mediastinum, which offers, besides

the possibility of performing an extended lymphadenectomy, the theoretical advantage of an extended

en-bloc dissection of all peri-tumoural tissues, thus improving local control. However, radicality of

surgery was comparable in both groups, reflecting the possibility of obtaining adequate local control

by transhiatal resection. pTNM stages were also comparable. There were slightly more patients with

celiac node involvement (Ml) in the extended en-bloc group, probably due to the lymphadenectomy in

the upper abdomen, leading to upgrading of the tumour when positive nodes were found. Staging was

therefore improved after an extended en-bloc lymphadenectomy, as has been shown for gastric

cancer.16'17 Stage migration might slightly influence the stage-by-stage comparison, as positive nodes

were found in the extended fields in ca. 20 percent of the patients.16 However, this phenomenon does

not affect comparison of overall survival rates.

An extended resection is believed to reduce the rate of loco-regional recurrence, thereby increasing

quality of life and prolonging disease-free and overall survival. In the present series the pattern of

recurrence was comparable after both resection forms. We observed virtually identical disease-free

and overall survival curves in the first two years of follow-up. Later during the follow-up both the

disease-free and the overall survival curve diverged, showing a trend favouring the extended

transthoracic approach. Estimated five-year disease-free survival rates were 27 percent and 39

percent resp., while five-year overall survival rates were 29 percent and 39 percent resp. Overall

survival rates were comparable with those mentioned in recent reports on en-bloc resection while

demographic and tumour characteristics were also comparable with the population in the other

studies.18'19 Given the present data one can be reasonably certain that the difference in survival at two

years is small. The original hypothesis of our study that transthoracic resection with extended en-bloc

lymphadenectomy significantly improves survival at two years is thus refuted. However, there was a

80

strong tendency (albeit not statistically significant) towards a better overall and disease-free survival at

five years, in favour of the extended approach. One can not discard the possibility that when an even

longer follow-up has been completed for all patients, this difference in favour of the extended

approach will become statistically significant.

Early morbidity, but not mortality, is significantly increased after an extended transthoracic resection,

leading to prolonged MCU/ICU and hospital stay, and increased costs. However, improvement of the

peri-operative care might be able to lower these early morbidity (and mortality) rates, thereby

decreasing the early benefits of a limited transhiatal resection. At present, faced with the increased

early morbidity and uncertain long-term benefit, we can not recommend the routine use of

transthoracic resection with extended en-bloc lymphadenectomy for adenocarcinoma of the mid-/distal

oesophagus or gastric cardia. Further follow-up of the patients in this study might clarify whether the

long-term survival benefits of the extended approach will outweigh the increase in early morbidity and

associated costs, and thereby more clearly define the role of transthoracic resection with extended en-

bloc lymphadenectomy in this population.

References

1) Müller JM, Erasmi H, Zieren U, Pichlmaier H. Surgical therapy of ooesophageal carcinoma. Br J

Surg 1990; 77: 845-7

2) Hulscher JBF, Tijssen JGP, Obertop H, Van Lanschot JJB. Transthoracic versus transhiatal

resection for carcinoma of the oesophagus: a meta-analysis. Ann Thorac Surg 2001; 72: 306-13

3) Goldmine M, Maddern G, LePrise E, Meunier B, Champion JP, Launois B. Oesophagectomy by

transhiatal approach or thoracotomy: a prospective randomized trial. Br J Surg 1993; 80: 367-70

4) Chu KM, Law SY, Fok M, Wong J. A prospective randomized comparison of transhiatal and

transthoracic resection for lower-third esophageal carcinoma. Am J Surg 1997; 174: 320-4

5) Jacobi CA, Zieren HU, Müller JM, Pichlmaier H. Surgical therapy of oesophageal carcinoma: the

influence of surgical approach and oesophageal resection on cardiopulmonary function. Eur J

Cardiothoracic Surg 1997; 11: 32-7

6) Hulscher JBF, Nieveen van Dijkum EJM, De Wit LT, et al. Laparoscopy and laparoscopic

ultrasonography in staging carcinoma of the gastric cardia. Eur J Surg 2000; 166: 862-5

7) UICC TNM Classification of malignant tumours. Sobin LH, Wittekind Ch eds., Wiley & Sons,

New York, 1997

8) Wijnhoven BPL, Siersema PD, Hop WC.I, Van Dekken H, Tilanus HW. Adenocarcinomas of the

distal oesophagus and gastric cardia are one clinical entity. Br J Surg 1999; 86: 529-35

81

9) Steup WH, De Leyn P, Deneffe G, Van Raemdonck D, Coosemans W, Lernt T. Tumours of the

oesophagogastric junction. Long-term survival in relation to the pattern of lymph node metastasis

and a critical analysis of the accuracy of pTNM classification. J Thorac Cardiovasc Surg 1996;

111:85-95

10) Gold MR, Siegel JE, Russell LB, Weinstein MC. Cost-effectiveness in health and medicine. New

York: Oxford University Press, 1996

11) Jansen SJ, Stiggelbout AM, Wkker PP, Nooij MA, Noordijk EM, Kievit J. Unstable preferences:

a shift in valuation or an effect of the elicitation procedure? Med Decis Making 2000; 20: 62-71

12) Oostenbrink JB, Koopmanschap MA, Rutten FFH. Handleiding kostenonderzoek: methoden en

richtlijnen prijzen voor economische evaluaties in de gezondheidszorg. Amstelveen, the

Netherlands: College voor Zorgverzekeringen, 2000

13) Hagen JA, DeMeester SR, Peters JH, Chandrasoma P, DeMeester TR. Curative resection for

oesophageal adenocarcinoma. Ann Surg 2001;234:520-31

14) Altorki N, Skinner D. Should en bloc oesophagectomy be the standard of care for oesophageal

carcinoma? Ann Surg 2001;234: 581-7

15) Fumagalli U. Resective surgery for cancer of the thoracic oesophagus. Results of a Consensus

Conference held at the Vlth World Congress of the International Society for Diseases of the

Oesophagus. Dis Oesoph 1996; 9: S30-38

16) Huisdier JBF, Van Sandick JW, Offerhaus GJA, Tilanus HW, Obertop H, Van Lanschot JJB. A

prospective analysis of the diagnostic yield of extended en bloc resection for adenocarcinoma of

the oesophagus or gastric cardia. Br J Surg 2001; 88: 715-9

17) Bunt AMG, Hermans J, Smit VTHBM, Sasako M, Hoefsloot FA, Fleuren G, Bruijn JA.

Surgical/pathological-stage migration confounds comparisons of gastric cancer survival rates

between Japan and western countries. J Clin Oncol 1995; 13: 19-25

18) Bonenkamp JJ, Hermans J, Sasako M, et al. Randomised comparison of morbidity after Dl and

D2 dissection for gastric cancer in 996 Dutch patients. Lancet 1995; 345: 745-8

19) Bonenkamp JJ, Hermans J. Sasako M, Van de Velde CJH. Extended lymph-node dissection for

gastric cancer. N Engl J Med 1999; 340: 908-14

Acknowledgements

The authors wish to express their gratitude to Professor T. Lernt of the Department of Surgery of the University Hospital Gasthuisberg Leuven for his assistance in designing the study and in standardising the transthoracic surgical technique. Finally the authors want to thank C. Manshanden from the Department of Surgery of the Academic Medical Center at the University of Amsterdam, who has acted as study-coordinator in the beginning ot the study. & 6

This study was supported by a grant from the Dutch Health Care Insurance Funds Council (grant # 1996-041).

82

Table 1: Characteristics of 220 patients randomised for either transhiatal oesophagectomy (THO) (n=106) or transthoracic oesophagectomy with extended en-bloc lymphadenectomy (TTO)(n=114)

Variable THO TTO

Age (years)* Gender (male/female) Weight loss (kilos)1

ASAI/IIf

69 (23-79) 92/14

4 (0-23) 39/67

64 (35-78) 97/17

4 (0-27) 40/74

n.s. n.s. n.s. n.s.

Oesophagus/cardia 87/19 93/21 n.s.

* values depicted are mean (range) values depicted are median (range) American Society of Anesthesists classification

*cardia: tumour bulk at or distal from gastro-oesophageal junction as seen on endoscopy/endosonography in the absence of Barrett remnants, ingrowth in oesophagus present.

83

Table 2: Early post-operative complications including ventilation-time, Intensive Care Unit/Medium Care Unit - stay (ICU/MCU-stay) and total hospital stay in days (median; range) in 220 patients randomised for either transhiatal oesophagectomy (THO) (n=106) or transthoracic oesophagectomy with extended en-bloc lymphadenectomy (TTO) (n=l 14)

V a r i a b l e THO TTO P-value

Post-operative complications* Pulmonary complications^ Cardiac complications Anastomotic leakage*

Subclinical Clinical

Temporary vocal cord paralysis Chylous leakage Wound infection

Ventilation time ICU/MCU-stay Hospital stay5

Hospital mortality

29 (27) 65 (57) <0.001 17 (16) 30 (26) 0.10 15 (14) 18 (16) 0.85 9 (9) 8 (7) 6 (6) 10 (9)

14 (13) 24 (21) 0.15 2 (2) 11 (10) 0.02 8 (8) 11 (10) 0.53

1 (0-19) 2 (0-76) <0.001 2 (0-38) 6 (0-79) O.001

15 (4-63) 19 (7-154) O.001

2 (2) 5 (4) 0.45

* Values in parentheses are the percentages based on the total number of patients randomised for either THO or TTO. For ventilation time, ICU/MCU time and hospital time the values in parentheses denote the range.

Pulmonary complications include among others pneumonia (isolation of pathogen from sputum culture and a new or progressive infiltrate on chest X-ray) and atelectasis (lobar collapse on chest X-ray)

Subclinical anastomotic leakage is anastomotic leakage seen only on the contrast radiograph, clinical anastomotic leakage is anastomotic leakage resulting in a cervical salivary fistula! Reintervention was needed in two patients, both after TTO. ^Hospital stay: stay in hospital from day of operation to discharge. Patients are generally admitted two days prior to surgery.

84

Table 3: Histology, pTNM-stage and radically of surgery in patients undergoing subtotal oesophagectomy by either transhiatal resection (THO) (n=94) or transthoracic resection with extended en-bloc lymphadenectomy (TTO) (n=lll). Patients scheduled for THO or TTO who underwent a total gastrectomy or were irresectable, are left out of the analysis; patients in whom conversion to TTO resp. THO took place are included under the intended surgical approach.

Variable THO TTO P-value M

Histology* Adenocarcinoma 90 (96) Other1 4 (4)

Stage1

0 2 (2) I 10 (11) Ha 18 (19) IIb 10 (11) III 47 (50) IV 7 (7)

Radicality achieved5

RO 68 (73) Rl 23 (24) R2 1 (1) uncertain 2 (2)

107 (96) 4 (4)

0.99

2 (2) 15 (14) 10 (9) 7 (6)

60 (54) 17 (15)

79 (71) 28 (25) 4 (4)

0.15

0.28

lymph nodes dissected 16±! 31 ± 14 <0.001

* Values in parentheses are the percentages of all patients undergoing subtotal oesophagectomy by either THE (94) or TTE (111) : There were erroneously two patients with high grade dysplasia (stage O), one with squamous cell carcinoma and one with adeno-squamous carcinoma, in each group ' Three of the four patients with R2 resections after TTE became R2 due to distant

metastases found during the abdominal phase 'R0: no residual tumour remaining; Rl: microscopically residual tumour remaining; R2: macroscopically residual tumour remaining 'P-values are based on all stages combined and all resections achieved

85

Table 4. Mean total costs (in euros) of treatment per patient for patients scheduled for transhiatal oesophagectomy (THO) (n=106) and patients scheduled for transthoracic oesophagectomy with extended en-bloc lymphadenectomy (TTO) (n=l 14)

Variable THO TTO

Pre-operative costs Outpatient visits 419 382 Diagnostics 1206 1345 GP visits* 37 42 Sub-total 1662 1768

In-hospital treatment costs Oesophagectomy 3551 5389 ICU/MCU stayt 5563 13459 Stay at surgical ward 5934 7409 Reinterventions 101 340 Pathology 473 633 Diagnostics 1114 1688 Sub-total 16736 28918

Medical Costs during follow-up Outpatient visits 754 737 Readmissions 839 1399 GP visits 46 57 Aids 439 438 Other 18 15 Sub-total 2096 2646

Non-medical costs Travel costs 143 124 Other costs for patient 78 116 Sub-total 221 240

Indirect costs Absenteeism from work 3094 3527

Total costs (euros) 23,809 37,099

*GP: General Practitioner 11CU/MCU: Intensive Care Unit/Medium Care Unit

86

Figure 1: Kaplan Meier curves of the disease-free interval of patients randomised for transhiatal oesophagectomy (THO) (n=106) or transthoracic oesophagectomy with extended en-bloc iymphadenectomy (TTO) (n=l 14).

3

Ü o.o

TTO

ii i 11 i—H—M i in—m+

-H ,

THO

0 1 10

Time (years)

.t risk THO 106 68 47 32 20 15 11 4

TTO 114 69 53 39 31 20 13 7

87

Figure 2: Kaplan Meier curves of the overall survival of patients randomised for transhiatal oesophagectomy (THO) (n=106) or transthoracic oesophagectomy with extended en-bloc lymphadenectomy (TTO) (n=114).

1.0

.8

Co > > i _

3 a> .4 05 1

CD > O CD > .2

^—» Co 13

E ^

Ü 0.0

TTO

- i — H — H — h m — m -

• -HH-+ -~h

THO

0 1 4 5 6 7 8

Time (years) risk THO 106 68 47 32 20 15 11 4

TTO 114 69 53 39 31 20 13 7

88

Figure 3: Kaplan Meier curves of the Quality of Life Adjusted Survival (QALY) based on the Standard Gamble (SG) of patients randomised for transhiatal oesophagectomy (THO) (n=106) or transthoracic oesophagectomy with extended en-bloc lymphadenectomy (TTO) (n=l 14).

1-01

•HH—I 111 I I I I h -H H-H

TTO

-H-H h

"4-h, I I H >

-+-H--H h-'H—H+H-K

THO

Time (years)

IriskTHO 106

TTO 114

68

69

47

53

32

39

20

31

15

20

11

13

89