Multidisciplinary team management is associated with improved outcomes after surgery for esophageal cancer

Introduction

Cancer of the esophagus is now the ninth most common cancer worldwide, accounting for 7000 new diagnoses and 6700 deaths per year in the UK alone.¹ Despite sporadic reports to the contrary the overall outlook for patients diagnosed with this tumor remains bleak and there remains a prevailing pessimism among clinicians regarding the treatment and prognosis of patients with cancer of the esophagus.

Much of this nihilistic attitude originated from a critical surgical review conducted over 20 years ago of 84 000 patients who had received treatment between 1953 and 1979.² This suggested that of every 100 patients with esophageal cancer, only 39 would have their tumors resected, of whom 26 would leave hospital, and only four would survive 5 years. Despite many advances in preoperative radiological staging, improved anesthetic technique and more radical surgical procedures over the next 10 years, a report from Leeds in 1994 encompassing the years 1975–1994 showed no improvement.³ Moreover, the recent regional audit of the management of esophagogastric cancer in Wales demonstrated that many individual surgeons' caseloads remained small, staging strategies were idiosyncratic, and open and closed operations were performed in as many as 23% of cases.⁴

Findings such as these have led to the guidance on commissioning cancer services.⁵ These guidelines stipulate that specialist teams should be established at appropriate cancer centers or units, and that these should draw patients from populations of more than one million.

The aim of this study was to compare the outcomes of patients undergoing R0 esophagectomy by one consultant anesthetist and two specialist surgeons, working as part of a multidisciplinary team (offering specialist radiological, and anesthetic support, and neoadjuvant chemoradiotherapy where indicated) with outcomes after surgery alone performed by six general surgeons working independently in a large UK cancer unit serving a population of 560 000.

Patients and Methods

Two groups of patients were studied. Between January 1, 1991 and December 31, 1997 342 consecutive patients with esophageal carcinoma were diagnosed at the Royal Gwent Hospital, Wales. Clinical and pathological information on this group of patients between January 1, 1991 and September 30, 1995 was obtained by retrospective review of patients' case notes and was obtained prospectively from October 1, 1995 until December 31, 1997. Seventy-seven of these patients (median age 60 years, 50 male) underwent surgery with curative intent by six consultant general surgeons and potentially curative resection was possible in 54 (70%) patients. This comprises our historical control group. Between January 1, 1998 and December 31, 2003 185 patients out of a total of 303 patients (61%) diagnosed with esophageal cancer were referred to and treated by a multidisciplinary team (MDT) comprising two consultant surgeons (WGL and KS, who operated together as a team), two radiologists, one oncologist, two gastroenterologists, one anesthetist, one pathologist, and one specialist nurse. Sixty-seven of these patients (median age 61 years, 47 male) underwent anesthesia and surgery with curative intent by one consultant anesthetist and the two MDT consultant surgeons. Potentially curative resection was possible in 53 (79%) patients. This comprises our MDT group. The demographic details of the patients undergoing R0 esophagectomy are shown in Table 1. Preoperative staging was done with the aid of computerized tomography (CT), endoluminal ultrasound (EUS, after October 1, 1995) and laparoscopy if considered appropriate (tumors of perceived preoperative stage of T3 N1 MX of the lower third of the esophagus). The CT examinations undertaken before June 23, 1997 were performed using an incremental General Electric CT scan system. Between June 23, 1997 and April 10, 2002 the CT examinations were performed using a helical Siemens Somatom +4 CT system; and after April 10, 2002 the CT examinations were performed using a multislice Toshiba Aquilon volume acquisition spiral CT system. EUS was performed using an MH-908 esophagoprobe.

The definition of a potentially curative resection was that all visible tumor was removed and that both proximal and distal resection margins were free of tumor on histological examination. Involvement of the circumferential resection margin (CRM) was defined as the presence of tumor less than 1 mm from the circumferential margin.

Surgical treatment

The details of the surgery performed are shown in Table 2. Transhiatal resection was performed for patients with tumors of the lower third of the esophagus deemed to be T1-2 N0 on staging, or if the patient was assessed as American Society of Anesthesiologists (ASA) grade III. All tumors were staged in accordance with the TNM Classification of Malignant Tumors.^6,7

Neoadjuvant treatment

Following the inception of the MDT, all patients with a diagnosis of esophageal cancer were discussed weekly, and treatment strategies were developed and tailored for individual patients based on the treatment algorithm shown in Fig. 1. Patients with a perceived preoperative stage of T3 NX M0 and ASA grades I–III were considered for multi modality treatment. Twenty-six patients managed by the MDT received multimodal therapy. Patients treated prior to 2002 were offered chemoradiotherapy as the neoadjuvant treatment of choice, but this policy was changed following the publication of the results of the large UK randomized trial of neoadjuvant chemotherapy (OE02). Patients were subsequently offered neoadjuvant chemotherapy alone. Thirteen patients received neoadjuvant chemotherapy (median age 58 years [47–70], 11 male) and 13 received chemoradiotherapy (56 years [38–76], 7 male). Neoadjuvant chemotherapy consisted of two cycles of cisplatin (80 mg/m²) and 5FU for 4 days. Neoadjuvant chemoradiotherapy consisted of 2 cycles of cisplatin (60 mg/m²) with 225 mg/m² per day of infusional 5FU prior to 45 Gy in 25 Fractions in 5 weeks concurrent with 3-weekly cisplatin (60 mg/m²) and infusional 5FU (200 mg/m²/day). Radiotherapy was administered with a two-phase technique starting with parallel-opposed fields followed by a three field 3-dimensional plan.

Follow up

Patients undergoing esophagectomy were reviewed every 3 months for the first year and every 6 months thereafter. All of the control patients were followed up for at least 5 years or until death. The median duration of follow-up for MDT patients was 31 months (5–65), and 30 patients (57%) were followed up for at least 5 years or until death. Endoscopy and computed tomography were arranged if recurrent disease was suspected. Causes of death were sought from case notes, pathology records, and general practitioners' records.

Statistical analysis

Statistical analysis appropriate for non-parametric data was used. Grouped data were expressed as median (range). Grouped data was compared using the Mann–Whitney U-test for unpaired data.⁸ Cumulative survival was calculated by the life table method of Kaplan and Meier.⁹ Differences in survival times between groups of patients were analysed by the log rank method. Cox's proportional hazards model was used to assess the prognostic value of individual variables. Data analysis was carried out with the Statistical Package for Social Sciences (SPSS) version 12 (SPSS, Chicago, Illinois, USA).

Results

Patterns of referral of patients diagnosed with esophageal cancer

Between January 1, 1991 and September 30, 1995 we have no data for the total number of patients referred for a surgical opinion, but a median of seven (3–13) R0 esophagectomies were performed each year. Between 1996 and 2003, the number of patients referred per year for a surgical opinion increased from 11 (23%) patients in 1996 to 44 (76%) patients in 2003 (χ² = 32.13, DF = 1, P < 0.001).

Preoperative staging

All of the patients managed by the MDT underwent a preoperative CT scan compared with 29 (54%) of the control patients (χ² = 30.02, DF = 1, P < 0.001). All of the patients deemed potential surgical candidates by the MDT also underwent a preoperative EUS compared with five (9%) of the control patients (χ² = 91.12, DF = 1, P < 0.001).

Details of global treatment modalities utilized

Table 3 documents the different types of treatment modalities utilized for all of the patients in the study. Of the 77 patients who underwent surgery with curative intent, 54 patients (70%) underwent a potentially curative R0 esophagectomy, 16 (21%) underwent open and close laparotomy, six (8%) underwent open and closed thoracotomy, and four (5%) underwent palliative esophagectomy (three patients CRM+[cross-reacting material positive], one patient was subsequently proven to have a liver metastasis). Of the 67 MDT patients who underwent surgery with curative intent, 53 patients (79%) underwent a potentially curative R0 esophagectomy, nine (13%) underwent open and close laparotomy (χ² = 11.90, DF = 1, P = 0.001), no patients underwent open and close thoracotomy (χ² = 5.45, DF = 1, P = 0.02), and 5 (7%) underwent palliative esophagectomy (four patients CRM+, one patient was subsequently proven to have a liver metastasis).

Chemotherapy associated morbidity and mortality

One patient undergoing neoadjuvant chemoradiotherapy developed grade III toxicity in the form of life-threatening sepsis, and one patient undergoing neoadjuvant chemotherapy developed significant neutropenia.

Operative morbidity and mortality

The details of the major operative morbidity and mortality are shown in Table 4. Of the three MDT patients who suffered fatal complications, two had received neoadjuvant chemoradiotherapy and both died of acute respiratory distress syndrome (ARDS).

Details of pathological response to neoadjuvant therapy

Of the patients who received neoadjuvant therapy, based on TNM criteria, 14 patients had no response (eight chemotherapy, six chemoradiotherapy) and nine patients a partial response (five chemotherapy, four chemoradiotherapy). All of the three patients deemed to have had a complete pathological response had received neoadjuvant chemoradiotherapy.

Survival

Corrected cumulative survival by treatment, calculated by life table analysis, is shown in Fig. 2. The cumulative 5-year survival following R0 esophagectomy for the control patients was 10% compared with 52% in the MDT patients (χ² = 15.05, DF = 1, P = 0.0001). The median survival following R0 esophagectomy in control patients was 18 months (95% CI = 10–26) compared with 66 months (15–116) in MDT patients. The cumulative survival related to the type of neoadjuvant therapy is shown in Fig. 3. The factors found to be significantly associated with the duration of survival on univariate analysis are shown in Table 5. The hazard function related to MDT treatment or not, is shown in Fig. 4.

Outcomes related to individual consultant surgeon's teams

The details of the individual consultant's surgical workload and outcomes are shown in Table 6. There was no statistical correlation between consultant surgeon's individual workloads and patients' 1- and 5-year survival (Spearman's correlation coefficients 0.324 and 0.243, P = 0.478 and P = 0.599 respectively).

Multivariate analysis

The prognostic variables entered into the model were ASA grade, neoadjuvant chemoradiotherapy, extent of lymphadenectomy, neoadjuvant treatment, lymph node involvement, MDT management, anesthetist, type of operation and operating surgeon. Forward and backward stepwise regression was used. MDT management (HR = 0.337, 95% CI = 0.201–0.564, P < 0.001), lymph node metastases (HR = 1.728, 95% CI = 1.070–2.792, P = 0.025), and ASA grade (HR = 2.207, 95% CI = 1.412–3.450, P = 0.001) were found to be the most important predictors of survival as determined by Cox's proportional hazards model (global χ² for the model was 28.405, DF = 3, P < 0.001).

Discussion

The principal finding of this study was that the key recommendations of the Clinical Outcomes Group (COG) guidelines regarding specialist services and multiprofessional teams were implemented in a large UK district general hospital and were associated with improved outcomes after surgery for patients with esophageal carcinoma. The proportion of patients undergoing optimal staging investigations increased significantly following MDT introduction, and this was accompanied by a significant reduction in the proportion of patients undergoing open and closed laparotomy and thoracotomy. Operative mortality fell over fourfold from 26% to 5.7%, while survival at 5 years for patients undergoing potentially curative surgery improved fivefold, from 10% to 52%. Moreover, the referral rate by hospital clinicians to the MDT increased significantly from one-third of patients in the 1990s to over three-quarters in 2003.

There are a number of potential weaknesses of this study. Retrospective case note review depends on adequate documentation of appropriate data at the time of presentation. Statistical comparisons between various treatment groups may be invalidated by older patients with significant comorbidity being treated by non-surgical means. Analysis of subgroups within a study may lead to bias, while comparisons of groups may prove to be not statistically significant simply because the study has insufficient power to demonstrate real differences. Accurate preoperative staging of patients' tumors was not possible in the early 1990s. Identification of patients who had received potentially curative rather than palliative resection was difficult because of inadequate information in the case notes. Hence the results after strictly curative resections may have been better than the overall results reported here. In any retrospective review, definition of the case subjects and of the controls is critical in determining the outcome of the study and subsequent conclusions. In this study we have chosen MDT management, that is specific specialist consultant investigation and treatment in the fields of gastroenterology, pathology, radiology, nursing, oncology, anesthesia followed by consultant team surgery by two specialist surgeons. Arguably, we could just as easily have used time – before and after January 1998. However, in the mid-1990s the same two consultant surgeons were practicing independently, but without each other's technical surgical support in theatre, or the additional resource of the other key upper gastrointestinal specialist team clinicians – in particular, a specialist consultant oncologist and a specialist consultant anesthetist. The overall surgical outcomes with these particular surgeons clearly improved significantly (Table 6), following the introduction of a multidisciplinary team framework.

Clearly this is not a randomized controlled trial and the number of patients is relatively small, but the results nonetheless demonstrate what can be achieved by specialist care in a large district general hospital in Britain. The improved outcomes cannot be explained by poorer-than-average results in our historical control group, as our results in the early 1990s are comparable with those reported by Sagar et al. from a large UK teaching hospital.³ In this latter study, potentially curative resection was possible in only 25% of patients, compared with 16% at the Royal Gwent Hospital in the early 1990s. Operative mortality was 27% after potentially curative esophagectomy (c.f. 26% in the present study) and 5-year survival was 7% (c.f. 10%). The results of surgical treatment of patients with esophageal cancer at the Royal Gwent Hospital during the early 1990s were therefore probably in keeping with those of most other centers in the UK at that time.

Alterations in anesthetic and surgical technique that occurred in the period of the study were reflected in the operative procedures used. The MDT group included patients who had transhiatal esophagectomy¹⁰ and patients in whom somewhat more extensive lymphadenectomies were performed than had been the custom in the early 1990s. Critics of the transhiatal Orringer esophagectomy have suggested that it is an inadequate cancer operation, which should be reserved for patients who have impaired respiratory function. However, its proponents argue that operative mortality and long-term survival are at least as good as with standard resection techniques.¹¹ Advocates of en-bloc resection with extensive lymphadenectomy argue that improved survival can only be achieved by meticulous wide excision of the tumor and its lymphatic field¹² as has been reported in rectal cancer¹³ and stomach cancer.^14,15 We used transhiatal esophagectomy in patients who were deemed to have T1/2 N0 carcinomas of the lower third of the esophagus irrespective of cell type, and in patients classified as ASA III. This approach is largely substantiated by the 5-year survival of our patients, which was 37% after transhiatal esophagectomy alone, compared with 38% after transthoracic esophagectomy alone. Five-year survival related to the surgical approach after neoadjuvant treatment was also similar at 75% after transhiatal esophagectomy compared with 80% after transthoracic esophagectomy.

Adjuvant or neoadjuvant chemotherapy and radiotherapy have been associated with improvements in local control and survival in several other cancer sites, in particular breast and colorectal cancer.^16,17 However, similar therapeutic approaches in the management of esophageal cancer have produced conflicting results. The large randomized UK trial of neoadjuvant chemotherapy (OEO2) reported a 2-year survival advantage of 43% versus 34% for patients who received two cycles of cisplatin and 5FU prior to surgical resection when compared with surgery alone.¹⁸ This contrasts with the US intergroup study, which failed to show any overall benefit in survival after preoperative chemotherapy.¹⁹ Our results suggest a significant survival benefit of 58% after neoadjuvant chemotherapy with surgery, compared with 51% after surgery alone. Neoadjuvant chemoradiotherapy is even more controversial. Of the five randomized trials reported, only Walsh et al. demonstrated a statistically significant survival advantage,²⁰ while Urba et al. demonstrated a 3-year survival advantage of 30% after chemoradiotherapy and surgery compared with 16% after surgery alone.²¹ Our results are in keeping with the above, with a 3-year survival of 68% for patients receiving neoadjuvant chemoradiotherapy and surgery, compared with 37% after surgery alone. Absolute conclusions are difficult to draw from the above trials, because of the considerable heterogeneity that exists with respect to radiation dose/fractionation, scheduling of chemotherapy, histology and outcomes after surgery alone. Moreover, many of the trials were underpowered in statistical terms. It is possible that overall improvements in survival from a combined approach are offset by higher operative mortality rates as suggested by Bosset et al.;²² indeed two-thirds of our operative deaths occurred in patients who had received preoperative chemoradiation. Certainly good outcomes can be achieved with definitive chemoradiation alone²³ and the place of selective surgery after definitive chemoradiation is the subject of an ongoing European Organisation for Research and Treatment of Cancer trial. A potential further advantage of specialist multidisciplinary teams is the identification of a significant cohort of patients who, although not surgical candidates, may be suitable for treatment with potentially curative chemoradiotherapy.

The true value of specialist surgical treatment per se in improving outcomes for patients with cancer of the esophagus remains uncertain. As far as we are aware, there have been no studies of the effectiveness of the service model recommended by the National Health Service Executive, and any relationship between patient throughput and survival remains controversial.^5,24,25 Nonetheless, the results of this study are encouraging, and mirror our previous findings in the surgical treatment of gastric cancer, where surgical subspecialization was associated with significant improvements in outcome.^26,27 Furthermore, these findings support the view that multiprofessional management can harness the diverse talents of an array of specialist clinicians with a common interest in upper gastrointestinal oncology, so that management plans are stage-directed and patient-tailored, to optimize outcomes after surgery for patients diagnosed with cancer of the esophagus.

Acknowledgments

We grateful for the contribution of the following Consultant General Surgeons who treated patients with esophageal cancer in the early and mid-1990s at the Royal Gwent Hospital: Mr D. E. Sturdy MS FRCS, the late Mr G. Griffith OBE FRCS, the late Mr M. Price Thomas FRCS, Mr K. D. Vellacott DM FRCS.

References