Symptoms, investigations and management of patients with cancer of the oesophagus and gastro-oesophageal junction in Australia

In Australia, oesophageal cancer represents 1.2% of all cancers and is responsible for 2.1% of cancer deaths. Recently, there has been a striking increase in oesophageal adenocarcinoma (OAC) incidence, estimated at 4.2% per year in New South Wales,1 whereas the incidence of oesophageal squamous cell carcinoma (OSCC) has declined. Almost all of the increase in OAC incidence has occurred in males, contributing to a male–female ratio approaching 8 to 1.1 Gastro-oesophageal junction adenocarcinoma (GOJAC) has also increased in incidence. The incidence patterns for OAC and GOJAC contrast with those for OSCC. This parallels trends observed in other Western countries2,3 and is not due to changes in diagnostic criteria.4 The principal causes of the increase in OAC (and probably GOJAC) are thought to be increased prevalence of gastro-oesophageal acid reflux and obesity in Western populations.5-11 Changing patterns of obesity appear to be driving the rising incidence of OAC, with particular attention focusing on “male pattern” central adiposity, which is postulated to increase the production of mitogenic, obesity-related hormones.12,13 Falling rates of Helicobacter pylori infection may also play a role, as chronic infection causes hypochlorhydria and thus protects against reflux-mediated carcinogenesis.14

The prognosis for patients diagnosed with these cancers is poor; 1-year survival for patients with OAC in a NSW study was 49% for localised cancer, 43% for cancers with regional spread and 12% for disseminated cancers.1 Yet despite the rapid increases in incidence and the poor survival from oesophageal cancers, relatively little is known about the patterns of care for patients with these diseases. Here we report the findings of an investigation into the presentation and clinical management of a cohort of patients with carcinomas of the oesophagus or gastro-oesophageal junction.

Our study was based on a cohort of patients previously enrolled in the Australian Cancer Study (ACS), a population-based, case–control study undertaken to investigate risk factors for oesophageal cancer.6 For our study (the ACS Clinical Follow-up Study), we collected clinical information and outcome data on ACS patients. We collected data on the presenting symptoms of patients, investigations and treatment pathways.

We identified 3273 potentially eligible patients with oesophageal cancer, of whom 1618 were excluded for various reasons and 385 declined to participate (Box 1). Of the remaining 1270, 1140 patients had histologically confirmed oesophageal cancer and gave consent for access to their medical records. Completed case report forms were available for 1100 patients (370 OAC, 425 GOJAC and 305 OSCC) and linked questionnaire data were available for 1056 patients.

Patient demographics are shown in Box 2. Notably, age distributions were similar for the three patient groups, whereas sex distributions were markedly different for OSCC patients compared with OAC and GOJAC patients. Eighty-seven per cent of patients resided in cities and towns, 11% lived in moderately accessible regional locations, and 2% were from remote or very remote locations.

Our study provides the first comprehensive description of the presentation and management of Australian patients with oesophageal cancer. The cohort is not entirely representative, as enrolment into the study required a number of time-limiting steps, which meant that records for some potentially eligible patients could not be reviewed. For example, patients with late-stage disease (AJCC stages III or IV) were less likely to be enrolled, and we could not access files for patients who died before written consent could be obtained. This means that we may have underestimated the prevalence of such tumours, and thus probably overestimated the proportions of patients offered curative therapies.

Nonetheless, our study provides the first overall “snapshot” of these cancers and their management in Australia. The sample was also large, comprising an estimated 35% of all people in mainland Australia diagnosed with oesophageal cancer during the study period.

In a country defined by large distances and a dispersed population, the issue of access is important. Reassuringly, we found the geographical distribution of this cohort was similar to that of the 2001 census.

The most common presenting symptom was dysphagia, which on direct questioning was found to be present in over 70% of patients, a proportion similar to that found in other studies.16,17 Dysphagia occurs when the oesophageal circumference has been reduced by two-thirds,18 which is sufficient to compromise the lumen. The United Kingdom guidelines for managing oesophageal cancer outline a number of “alarm symptoms”, of which dysphagia is the first, and for which referral for endoscopy is recommended within 2 weeks of presentation.19 Our data suggest most patients do not recognise the importance of dysphagia as an alarm symptom. Only 7%–9% of patients reported reflux as their primary symptom, a similar proportion to that reported in a previous study.16 The underlying precipitant for these patients may have been a change, likely worsening, of previous reflux symptoms or the development of new symptoms.

A key finding was the infrequent recording of AJCC pretreatment tumour stage (7%), and although stage could be imputed for a further 42%, for about half of the patients it was impossible to determine the extent of their cancer. Of the staging investigations employed, CT scans were the most common, being used in more than 90% of patients in our study. The major benefit of CT is the ability to rapidly identify patients with distant metastases. Staging information is improved by also performing an EUS to assess local infiltration and local nodal status. Relatively few Australian centres were performing EUS at the time of our study, hence the low usage we observed. Compared with conventional staging modalities, FDG-PET scanning has been shown to detect distant metastases in 4%–28% of oesophageal cancer patients and to change management in 3%–40% of patients.20 Although fewer centres were performing FDG-PET than EUS in Australia at the time of our study, FDG-PET scans were reported for 42%–51% of the patients. Thus, despite limited availability, both FDG-PET scanning and EUS were performed on sizeable numbers of patients, suggesting rapid uptake for these modalities.

Putting aside the extent of incomplete reporting, our study differs from others in having relatively fewer patients with stage IV cancers. For example, a United States study reported stage IV disease in 48% of OAC patients and 52% of GOJAC patients21 — considerably higher than the proportions we observed. Reporting of cancer stage is not mandatory in Australia, hence there are no reliable population-based data for comparison. Instead, estimates of the distribution of cancer stage must be derived from chart reviews. As the use of chart reviews requires patient consent, and because consent is less likely among patients with late-stage disease, it is likely that all studies based on chart review will underestimate the incidence of late-stage disease.

We identified apparent differences in the surgical management of OAC and GOJAC. Specifically, our data suggest that patients with GOJAC are more likely to undergo surgery alone than OAC or OSCC patients. A similar finding was reported in an Irish study.22 Information bias may partly explain these differences, as the location of a tumour can be identified more precisely from surgical resection specimens than from endoscopy. Thus, patients who have undergone surgery are more likely to have the anatomical location of their tumour classed as gastro-oesophageal junction than patients who have not received surgery.

The proportion of surgical patients undergoing preoperative (neoadjuvant) therapy in our study was lower than the proportion in the US study21 but higher than that found in the Irish study.22 At the time that the patients in our study were being treated, there had been one international report of benefit from neoadjuvant CRT in patients with OAC,23 and one account of benefit from preoperative chemotherapy in both OAC and OSCC patients.24 An Australian phase II study assessing the role of neoadjuvant CRT and CRT alone for cure and palliation of OAC and OSCC had also been reported before our assessment.25 It is likely that these publications, along with an active Australian trial of preoperative CRT,26 raised awareness among Australian clinicians who treat such patients. This could explain the high prevalence of neoadjuvant therapy for OAC and OSCC, and may also explain why 10% of patients in our study were also enrolled in trials. While there is no evidence from trials that routine postoperative therapy improves survival, we observed reasonably high proportions of patients undergoing postoperative therapy. The reasons for this were not clear from the records.

We found that use of definitive CRT in our study was markedly higher for patients with OSCC (37%) than for patients with OAC (13%) or GOJAC (5%). In comparison, the Irish study observed proportions of patients undergoing definitive CRT as 12% for OAC, 6% for GOJAC and 12% for OSCC.22 A possible explanation for the difference may be a widespread perception among clinicians that adenocarcinomas are less sensitive to radiation than squamous cell carcinomas. There are limited data to determine the validity of this perception; however, there are more publications reporting benefits for CRT as an alternative to resection for OSCC27,28 than as a treatment for OAC.

From our observation, most patients present with late-stage disease, although precise staging information is infrequently recorded. We encourage efforts to increase the reporting of tumour stage for these cancers. Although dysphagia is common, it is not apparently recognised as an alarm symptom; education to rectify this deficiency may be warranted. Many patients are managed by teams comprising surgeons and medical and radiation oncologists. With the increasing use of combined modality treatment, and the need to use technologies such as EUS and FDG-PET scanning, it would seem that optimal patient management will be through specialist centres with a suitable caseload and interest in the disease. The incidence of OAC and GOJAC is rising rapidly in Australia, a factor which must be considered when planning for future health service needs. Finally, given the high proportion of patients with late-stage disease, and the acknowledged poor survival rates for these cancers, we need to continue to explore ways to reduce the disease burden through primary prevention and early detection.

1 Recruitment of patients into the ACS Clinical Follow-up Study

ACS = Australian Cancer Study. CRF = case report form. GOJAC = gastro-oesophageal junction adenocarcinoma. OAC = oesophageal adenocarcinoma. OSCC = oesophageal squamous cell carcinoma.