You, John J. MD, MSc*†; Wong, Rebecca K.S. MBChB, MSc‡; Darling, Gail MD§; Gulenchyn, Karen MD‖; Urbain, Jean-Luc MD, PhD¶; Evans, William K. MD#
Curative therapy for esophageal cancer can include surgical resection, with or without neoadjuvant chemoradiation therapy, or chemoradiation therapy alone. Each of these treatment modalities is associated with substantial morbidity and carries a risk of important complications, including death. The risks of aggressive therapies with curative intent would not be worth incurring for patients who harbor distant metastatic disease. Therefore, accurate staging of patients before curative therapy is important to ensure that these intensive therapies are directed to patients for whom the benefits of treatment outweigh the harms.
Several studies have evaluated the diagnostic performance of 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET) for the staging of patients with esophageal cancer. FDG-PET has been reported to have a pooled sensitivity of 67% and pooled specificity of 97% to detect distant lymph node and organ metastases (M stage), compared with pathology as the reference standard.1 A systematic review also found that FDG-PET has superior diagnostic performance compared with computed tomography (CT) for the detection of distant metastases in patients undergoing staging of esophageal cancer, owing to its similar specificity (93% versus 91%) but greater sensitivity (71% versus 52%).2 There are also data suggesting that FDG-PET may have an impact on clinical management when used for preoperative staging of esophageal cancer, but the magnitude of the impact is highly variable across studies. For instance, among patients with apparently resectable disease based on conventional staging (e.g., CT scanning of the chest and abdomen), the proportion of patients upstaged by PET varies fivefold across studies, from as few as 4.0% to as many as 23.1% of patients.3–11 Moreover, these studies recruited patients from university-based, tertiary or quaternary care centers, in which the patient population and patterns of care may not be broadly representative of more routine clinical practice. As a result, the clinical utility of PET/CT for staging of patients with esophageal cancer in a real-world setting remains unknown.
Beginning in 2000, the government of the province of Ontario, Canada, implemented an evidence-based approach to the introduction of PET technology in which it would only fund PET based on high-quality evidence of its value in clinical decision making. In this approach, for clinical scenarios where the evidentiary base suggested a potential role for PET but robust data were lacking, such as the use of PET for staging patients with potentially resectable esophageal cancer, further prospective data collection through a provincial field evaluation (i.e., registry) was recommended.12,13 In this article, we present data from the Ontario PET Cancer Registry (OPCR) about the frequency of clinically relevant findings on PET/CT (e.g., upstaging of disease) when added to conventional staging for patients with potentially resectable esophageal cancer, and report health care utilization and outcomes after PET/CT as measured through linkage of OPCR data to health administrative databases, to evaluate the clinical utility of PET/CT in a real-world setting.
PARTICIPANTS AND METHODS
We conducted a multicenter prospective cohort study (OPCR) of patients receiving PET/CT for staging of potentially resectable esophageal cancer. Using encrypted unique identifiers, we linked OPCR data to provincial administrative databases (all housed at the Institute for Clinical Evaluative Sciences, Toronto, Ontario, Canada) to determine health care utilization and outcomes after PET/CT. The study received full Research Ethics Board approval from all participating institutions.
The OPCR study was a multicenter (n = 6 centers) prospective field evaluation funded by the Ontario Ministry of Health and Long-Term Care to provide access to PET for clinical indications for which current evidence suggested potential benefit of PET, but for which it was deemed that further evidence was required to more rigorously establish its clinical utility. Furthermore, as a field evaluation, the intent of the OPCR was to evaluate the clinical utility of PET in a real-world setting. Therefore, referrals to the OPCR for PET scans were permitted from both academic and community-based clinicians.
Patients were eligible for inclusion in the OPCR for staging of esophageal cancer if they had histologically proven esophageal cancer that seemed to be resectable based on conventional imaging (which could include endoscopic ultrasound). Upon providing informed consent to participate in the OPCR study, patients were also asked for their consent to anonymous linkage of their personal health information to provincial administrative databases for the purpose of determining patterns of health care utilization before and after PET.
In this article, we report data regarding patients receiving a first PET/CT scan for esophageal cancer staging in the OPCR that was performed between July 2009 and January 2011. Records were excluded from the study cohort if they were for a repeat PET/CT scan for esophageal cancer.
PET/CT Imaging Procedure
The PET/CT scanners were located at six centers in Ontario. Machines were required to meet specified performance criteria and to undergo quality control evaluation on each day imaging was performed. The PET/CT scanners were the following with full-ring bismuth germanate detectors: a Discovery ST 64 (General Electric, Waukesha, Wisconsin) in London, a Biograph Duo (CTI/Siemens, Knoxville, TN) at Princess Margaret Hospital in Toronto, a Philips Gemini Dual machine (Philips Electronics NV, Eindhoven, The Netherlands) at the Ottawa Hospital and at Sunnybrook Health Sciences Centre in Toronto, a Biograph 16 (Siemens, Knoxville, TN) at St. Joseph’s Healthcare Hamilton, and a 64 slice Gemini TF with lutetium-yttrium orthosilicate detectors (Philips Electronics NV, Eindhoven, The Netherlands) in Thunder Bay. To ensure consistent exam quality across all sites, studies were performed using the NEMA NU2-2001 phantom to verify calibration inaccuracy, verify reconstructed image resolution less than 10 mm Full Width Half Maximum, and qualify reconstruction methods at each site. Acquisition protocols were developed at each site to meet a minimum patient noise equivalent counts of more than 30 Mcounts/meter (+10%). Compliance was assured by monthly monitoring and quarterly review by the Quality Assurance Subcommittee.
The PET/CT examination was performed after a fast of 6 hours; blood glucose was required to be less than 10 mmol per liter before intravenous administration of 18F-FDG (5 MBq/kg, not exceeding 550 MBq). PET acquisition was preceded by a low-dose CT, and a whole-body PET/CT scan in supine position was obtained from the base of the skull to the upper half of both femurs. The examination was interpreted by the nuclear medicine physician with knowledge of the clinical history and access to correlative imaging.
At study entry, data were collected regarding patients’ age, sex, and stage (tumor, node, metastasis [T, N, and M] status, according to the American Joint Committee on Cancer [AJCC] staging manual, 6th edition).14 Nuclear medicine physicians at participating centers interpreted the PET/CT images and, based on these findings, recorded data regarding the N stage (Nx, N0, or N1) and M stage (Mx, M0, M1a, or M1b) of the tumor, again using the 6th edition of the AJCC staging manual.
To determine health care utilization before and after PET/CT, we linked the OPCR data to provincial administrative databases by using an encrypted unique identifier. The province of Ontario, Canada, provides its residents with universal coverage for hospital-based and physician services. The Ontario Health Insurance Plan (OHIP) Database contains data on all billing claims for insured physician services, surgical procedures, and diagnostic procedures performed across the province. Therefore, linkage to the OHIP Database provides the ability to comprehensively track, on a province-wide basis, health services received by patients enrolled in the OPCR. To characterize the staging workup received by patients in the 3 months before PET/CT, we searched the OHIP Database for claims for the following procedures: CT chest (X406, X407, X125), CT abdomen/pelvis (X409, X410, X126, X231, X232, X233), CT brain (X400, X401, X188, X402,X405, X408), magnetic resonance imaging of brain (X421), nuclear medicine bone scan (J850, J650, J851, J651), endoscopic ultrasound (E800, E801, S236, S237), and neck ultrasound (J105, J405).
To track health services use after PET/CT, we searched the OHIP Database for claims for surgical resection of esophageal cancer (S089, S090, S123, S125, S128) and chemotherapy (G339, G345, G359, G381). Actual radiotherapy visits are not reimbursed through OHIP; therefore, as a surrogate for radiotherapy, we searched for OHIP fee codes related to radiotherapy planning sessions (X310, X311, X312, X313). We also searched for fee codes for the following palliative procedures: esophageal dilation (E696, E698, Z523, Z525, Z529), esophageal stenting (E629, S082, S083), or laser debulking of tumor (E692, E695). We tracked repeat PET/CT scanning for esophageal cancer directly in the OPCR database. Finally, we determined vital status through linkage to the Registered Persons Database.
Our primary outcome was the proportion of patients with any clinically important change in stage based on PET/CT findings (i.e., results that could be expected to change management, such as a change from apparently resectable M0 disease before PET/CT to incurable, distant metastatic M1b disease based on PET/CT imaging). A detailed description of M stage categories,14 and the definition of clinically important change in stage used for this study are listed in Table 1.
We also determined, through linkage to administrative data, the proportion of patients receiving surgical resection, chemotherapy, radiotherapy, or palliative procedures at 3 and 6 months after the index PET/CT scan, and overall survival through to a last follow-up date of September 1, 2011.
We used descriptive statistics (mean and SD for continuous variables and proportions for categorical variables) to summarize the baseline characteristics of the study cohort (e.g., age, sex, time since esophageal cancer diagnosis, previous surgical resection, recent chemotherapy, recent radiotherapy, and clinical stage before PET), the results of PET/CT scanning, and subsequent health care utilization.
We compared the proportion of patients receiving different types of treatment (surgery, chemoradiotherapy, etc.) according to post-PET/CT M-stage grouping by using multiple comparisons with Cochran–Armitage trend test and p values were adjusted by the Hommel method. We used a time-to-event analysis to characterize overall survival of our cohort, following patients from the time of PET/CT to the earliest of either death or September 1, 2011 at which time patients were censored. We constructed Kaplan–Meier survival curves for the overall cohort, as well as stratified by pre-PET/CT M stage and post-PET/CT M stage, and used the log-rank test to examine for differences in survival across these strata.
There were 504 records in the OPCR for a first PET/CT scan for esophageal cancer staging between July 2009 and January 2011. After excluding 13 patients who did not consent to data linkage, there were 491 patients in our study cohort, followed for a median of 336 days. The study cohort is described in Table 2. Patients had a mean age of 65.1 years (SD 10.6 years), were predominantly male (79.6%), and were typically staged using CT chest (91.6%) and CT abdomen/pelvis (91.0%). On the basis of pre-PET clinical assessment and conventional imaging, the substantial majority of patients (93.5%) had apparently resectable disease (i.e., M0 disease) at the time of enrollment.
Change in Stage after PET/CT
PET/CT imaging led to clinically important changes in stage for a total of 118 of 491 patients (24.0%): 107 patients (21.8%) were upstaged by PET/CT results (i.e., from M0 to M1a or M1b; or from M1a to M1b), and 11 patients (2.2%) were downstaged by PET/CT results (from M1b to M0; or from M1a to M0). There were 74 patients (15.1%) who were upstaged to M1b status based on results of PET/CT (Table 3).
Health Care Utilization and Outcomes after PET/CT
By 6 months after PET/CT, the majority of the study cohort had undergone some form of specific therapy for their esophageal cancer, most commonly surgical resection with or without adjunctive therapy (44.4%) or combination chemoradiotherapy (25.1%; Table 4). Patterns of care at 6 months after PET/CT differed depending on the imaging findings on PET/CT, with greater use of surgical resection with or without adjunctive therapy in patients with M0 disease (54.4%) than those with M1a (25.0%; p < 0.001) or M1b (7.3%; p < 0.001) disease, and greater use of combination chemoradiotherapy in patients with M1b disease (43.9%) than those with M1a (22.2%; p = 0.002) or M0 (21.2%; p < 0.001) disease. Use of palliative procedures (esophageal dilatation or esophageal stenting) was similar in patients with M0 (30.8%), M1a (22.2%), and M1b (29.3%) disease (p = 0.65). Health care utilization at 3 months is shown in Supplemental Table 1 (Supplemental Digital Content 1, http://links.lww.com/JTO/A502).
During follow-up, 10 patients (2.0%) and 22 patients (4.5%) had received a repeat PET/CT scan for esophageal cancer at 3 months and 6 months, respectively; all patients who received repeat PET/CT scans for esophageal cancer staging had M0 disease on their index PET/CT scan.
Overall survival for the entire study cohort (median survival, 603 days), is shown in Figure 1A. Survival decreased significantly with higher M stage based on conventional (pre-PET/CT) staging (median survival for M0, 628 days; M1a, 319 days; M1b, 217 days; p = 0.007; Fig. 1B) and these differences became more pronounced based on reclassified stage after PET/CT (median survival for M0, 701 days; M1b, 227 days; p < 0.001; Fig. 1C; note: median survival could not be calculated for patients with stage M1a disease on PET/CT because more than 50% of the patients were still alive at maximum follow-up).
This prospective, multicenter field evaluation of patients with potentially resectable esophageal cancer found that PET/CT led to clinically important changes in stage for one in every four patients scanned. In most cases this was because of upstaging of disease to an incurable (M1b) status and these patients had lower rates of esophageal resection during follow-up.
Strengths of our study include its prospective, multicenter design, and size. To our knowledge, it is the largest published experience with PET/CT for staging of patients with potentially resectable esophageal cancer. More importantly, previous studies have enrolled patients from highly specialized, university-based centers.3–11 Our study is, to our knowledge, the first to confirm the clinical utility of PET/CT when used in a real-world setting.
Our study also has limitations. First, our data are based on the 6th edition of the AJCC staging system, because this was the system in use at the time our study was conducted.14 It is likely that fewer patients in our cohort would have been upstaged by PET/CT had the current 7th edition been used instead. This is because nodes considered as nonregional lymph node metastases and classified as M1a under the 6th edition are now classified as M0 under the 7th edition.15 Therefore, patients in our study who were M0 based on conventional imaging and upstaged because of findings of M1a disease on PET/CT would not have been upstaged according to the 7th edition. Reclassification of data in Table 3 suggests that 15.1% of patients (74 of 491) in our study (approximately 1 in every 7 patients) would have been upstaged by PET/CT to M1 status under the 7th edition, which still represents a clinically important impact of PET/CT, given that patients found to have M1 disease would no longer be candidates for curative surgical resection. Second, we did not have access to the results of biopsies that may have been performed to further evaluate FDG-avid lesions detected on PET/CT, and recognize that this is an important limitation of our study. However, it is reassuring to note that prior studies have reported that FDG-PET has high specificity (i.e., >90%) for the detection of distant metastases,1,2 and a low false-positive rate of 3.7%.10 Nonetheless, it remains unknown whether some patients in our cohort may have had false-positive results on PET/CT and may have had potentially life-saving therapy incorrectly withheld as a result. Third, our study did not include a control group of otherwise similar patients who did not undergo PET/CT. Therefore, our findings do not permit a direct assessment of the magnitude of the impact of PET/CT on clinical decision making. However, our observation that the rates of surgical resection were much higher among patients with M0 status on PET/CT compared with those with M1b status on PET/CT does suggest that PET/CT had an important influence on subsequent management by avoiding futile, aggressive interventions in patients with incurable metastatic disease.
We found that PET/CT stage was associated with different patterns in subsequent management, whereby patients with no evidence of metastatic disease on PET/CT were more likely to undergo surgical resection. Although most patients (54%) in our cohort with M0 disease based on PET/CT staging underwent esophagectomy, some patients did not. However, the rate of surgical resection we observed is similar to rates of 46% and 30% seen in other population-based cohorts of localized esophageal cancer in the United States and Australia, respectively.16,17 Reasons for the nonsurgical management (e.g., with definitive chemoradiotherapy)18,19 of patients with M0 disease in our cohort may have included poor performance status, high perioperative risk, or patient preferences.
Our finding that 15.1% of patients were upstaged to M1b status after PET/CT is broadly consistent with other studies that have examined this issue (range, 4.0%–23.1%).3–11 In particular, the rate of upstaging in our study is consistent with data from the prospective, multicenter American College of Surgeons Oncology Group trial, in which the proportion of patients upstaged to M1b status was 4.8%, but as high as 14.3% when patients with a positive PET scan were classified as upstaged if no biopsy was done but confirmatory imaging was positive.10 However, the reclassification rate in our study is higher than the rate of upstaging (4.0%) observed in a recent prospective study conducted by van Westreenen et al.11 at three centers in The Netherlands. It is possible that some physicians occasionally underreported or underestimated patients’ actual stage at entry into our study (e.g., patients may have been labeled as having M0 disease, but with CT imaging showing lesions consistent with metastatic disease), which would have allowed patients to access PET through the provincial registry study; however, we have no evidence to support this assertion. A more plausible explanation for the lower rate of upstaging in the study by van Westreenen et al. likely relates to the more intensive, protocolized, multimodality nature (including endoscopic ultrasound) of the pre-PET diagnostic staging used at that academic center. In contrast, our field evaluation was conducted in a jurisdiction where endoscopic ultrasound was not widely available and consisted of a more general, population-based sample of patients who were thought to have potentially resectable esophageal cancer based on conventional real-world staging. Therefore, our data provide an assessment of the impact of PET/CT when used for the staging of patients with potentially resectable esophageal cancer in a general population setting.
In conclusion, we found that PET/CT led to clinically important changes in stage for an appreciable number of patients with potentially resectable esophageal cancer (i.e., a number needed to scan of 4). In particular, PET/CT can upstage patients with apparently resectable disease based on conventional staging, can prevent patients with incurable disease from receiving futile, aggressive interventions, and can contribute importantly to the clinical management of these patients.
The Ontario PET Cancer Registry (OPCR) was funded by the Ontario Ministry of Health and Long-Term Care. This study was supported by the Institute for Clinical Evaluative Sciences (ICES), which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (MOHLTC). The opinions, results and conclusions reported in this article are those of the authors and are independent from the funding sources. No endorsement by ICES or the Ontario MOHLTC is intended or should be inferred.
The authors thank the site investigators (Dr. Marc Freeman, University Health Network/Princess Margaret Hospital, Toronto; Dr. William Wong-Pack, St. Joseph’s Healthcare Hamilton, Hamilton; Dr. Lisa Ehrlich, Sunnybrook Health Sciences Centre, Toronto; Dr. Laurent Dinh, The Ottawa Hospital, Ottawa) and PET technologists at participating study centers for their diligence in collecting data for the OPCR; Ian Purdy, Jennifer Bennie, Say-Ry Pheng, Cindy Fong, Thi Ho, Eriola Asllani, and Limei Zhou, all at the Institute for Clinical Evaluative Sciences, Toronto, Ontario, for central project coordination and analytic support; and Angela Bonin and Beth Manganelli, Bay Area Health Trust, Hamilton, Ontario, for central administrative support.
1. van Westreenen HL, Westerterp M, Bossuyt PM, et al. Systematic review of the staging performance of 18F-fluorodeoxyglucose positron emission tomography in esophageal cancer. J Clin Oncol. 2004; 22:3805–3812
2. van Vliet EP, Heijenbrok-Kal MH, Hunink MG, Kuipers EJ, Siersema PD. Staging investigations for oesophageal cancer: a meta-analysis. Br J Cancer. 2008; 98:547–557
3. Block MI, Patterson GA, Sundaresan RS, et al. Improvement in staging of esophageal cancer with the addition of positron emission tomography. Ann Thorac Surg. 1997; 64:770–6
4. Kole AC, Plukker JT, Nieweg OE, Vaalburg W. Positron emission tomography for staging of oesophageal and gastroesophageal malignancy. Br J Cancer. 1998; 78:521–527
5. Meltzer CC, Luketich JD, Friedman D, et al. Whole-body FDG positron emission tomographic imaging for staging esophageal cancer comparison with computed tomography. Clin Nucl Med. 2000; 25:882–887
6. Flamen P, Lerut A, Van Cutsem E, et al. Utility of positron emission tomography for the staging of patients with potentially operable esophageal carcinoma. J Clin Oncol. 2000; 18:3202–3210
7. Buchmann I, Hansen T, Brochhausen C, et al. FDG-PET in the initial staging of squamous cell oesophageal carcinoma. Nuklearmedizin. 2006; 45:235–241
8. Katsoulis IE, Wong WL, Mattheou AK, Damani N, Chambers J, Livingstone JI. Fluorine-18 fluorodeoxyglucose positron emission tomography in the preoperative staging of thoracic oesophageal and gastro-oesophageal junction cancer: a prospective study. Int J Surg. 2007; 5:399–403
9. Duong CP, Demitriou H, Weih L, et al. Significant clinical impact and prognostic stratification provided by FDG-PET in the staging of oesophageal cancer. Eur J Nucl Med Mol Imaging. 2006; 33:759–769
10. Meyers BF, Downey RJ, Decker PA, et al. American College of Surgeons Oncology Group Z0060 The utility of positron emission tomography in staging of potentially operable carcinoma of the thoracic esophagus: results of the American College of Surgeons Oncology Group Z0060 trial. J Thorac Cardiovasc Surg. 2007; 133:738–745
11. van Westreenen HL, Westerterp M, Sloof GW, et al. Limited additional value of positron emission tomography in staging oesophageal cancer. Br J Surg. 2007; 94:1515–1520
12. Evans WK, Laupacis A, Gulenchyn KY, Levin L, Levine M. Evidence-based approach to the introduction of positron emission tomography in ontario, Canada. J Clin Oncol. 2009; 27:5607–5613
13. Wong R, Walker-Dilks C, Raifu A. Evidence-based guideline recommendations on the use of positron emission tomography imaging in oesophageal cancer. Clin Oncol (R Coll Radiol). 2012; 24:86–104
14. Greene FL, Page DL, Fleming ID, Fritz A. Esophagus American Joint Committee on Cancer: AJCC Cancer Staging Manual, 6th Ed. 2002; New York Springer 91–98
15. Edge SB, Byrd DR, Compton CC. American Joint Committee on Cancer Staging Manual. 2010; 7th Ed New York Springer 103
16. Cen P, Banki F, Cheng L, et al. Changes in age, stage distribution, and survival of patients with esophageal adenocarcinoma over three decades in the United States. Ann Surg Oncol. 2012; 19:1685–1691
17. Stavrou EP, Ward R, Pearson SA. Oesophagectomy rates and post-resection outcomes in patients with cancer of the oesophagus and gastro-oesophageal junction: a population-based study using linked health administrative linked data. BMC Health Serv Res. 2012; 12:384
18. Stahl M, Stuschke M, Lehmann N, et al. Chemoradiation with and without surgery in patients with locally advanced squamous cell carcinoma of the esophagus. J Clin Oncol. 2005; 23:2310–2317
19. Bedenne L, Michel P, Bouché O, et al. Chemoradiation followed by surgery compared with chemoradiation alone in squamous cancer of the esophagus: FFCD 9102. J Clin Oncol. 2007; 25:1160–1168