Distribution and Timing of Distant Metastasis after Local Therapy in a Large Cohort of Patients with Esophageal and Esophagogastric Junction Cancer

Patients with esophageal and/or esophagogastric junction cancer (EAC) often have poor prognosis even when the cancer is localized. The incidence of EAC continues to rise in the USA; in 2013, an estimated 17,790 people were diagnosed with the disease and there were 15,210 deaths [1]. When EAC is localized, patients are treated with preoperative chemoradiation (trimodality therapy, TMT) or definitive chemoradiation (bimodality therapy, BMT) [2,3,4]. We reported that distant metastases (DMs) are fairly frequent after TMT and BMT [5,6,7]. Detailed information on the timing and exact frequency of DMs in this patient population has not been fully reported. We report here on the frequency, sites and timing of DMs in a large cohort of patients. Our ultimate goal is to develop a clinical model that will associate clinical parameters with a high likelihood of DM, prior to local therapy or surgery. If such a clinical model is established, it may provide an opportunity to properly select therapies for the high-risk subset. Currently, no such model exists.

The study cohort was identified from a prospectively maintained database at the Department of Gastrointestinal Medical Oncology, University of Texas M.D. Anderson Cancer Center, between 2002 and 2013. We included all patients with localized EAC who received TMT or BMT with a curative intent. All patients were staged by imaging study and upper endoscopic ultrasound. The therapy decision (i.e. TMT or BMT) was made at the multidisciplinary conference. The American Joint Committee on Cancer, 6th edition, was used for clinical staging. No other selection criteria were applied. The institutional review board approved this analysis.

Chemotherapy with radiation included a fluoropyrimidine (intravenous or oral) and either a platinum compound or a taxane. Radiation with a total of 45-50.4 Gy was delivered in 1.8 Gy per fraction by one of the conformal techniques. Platinum or taxane was given 5× weekly and fluoropyrimidine was given 5× on 5 days/week. In TMT-eligible patients, an esophagectomy was performed after 6-8 weeks after the completion of chemoradiation. The primary surgeon selected the surgical technique [transthoracic (Ivor-Lewis), transhiatal, total (three-field technique) or minimally invasive esophagectomy with lymph node dissection].

Imaging studies and endoscopy were performed upon completion of local therapy. Additional follow-up data were obtained from our institution's tumor registry and the hospital records or the Social Security database.

Time to DM was defined as the time from the end of local therapy. Overall survival (OS) for patients who developed DM was defined from the diagnosis of DM to death or the last follow-up. The rate of distribution and timing of relapse were tabulated by frequency and percentage. OS values were obtained by the Kaplan-Meier method. Statistical analyses were performed using SPSS software (IBM SPSS statistics 21.Ink).

We studied a total of 629 consecutive patients (356 who had TMT and 273 who had BMT) between 2002 and 2013. The median age was 63 years (range 20-91). Most of the patients were men (88.1%) and white (89.3%). The main primary site was the esophagogastric junction (84.9%) and the most frequent histology was adenocarcinoma (87.4%). The clinical characteristics of these patients are summarized in table 1. The median follow-up time was 37.2 months (interquartile range 17.8-65) for patients who were still alive.

The patterns of failure are summarized in figure 1. Forty percent (144 of 356 patients) who had TMT and 64% (175 of 273 patients) who had BMT had a relapse. Of 144 relapses after TMT and 175 after BMT, 83 and 62%, respectively, were DM.

The 7 most common anatomical sites of DM, as the first exclusive event, were the lung (for TMT 20% and for BMT 17%), distant lymph nodes (13 and 20%), liver (14 and 11%), peritoneum (10 and 13%), bone (6 and 11%), brain (8% in both groups) and pleura (7 and 5%). The cumulative rates of DM were 90.8 and 96.3% after TMT and BMT within 2 years after the completion of treatment and 97.5 and 99.1% after TMT and BMT within 3 years, respectively (table 2).

The median OS time of the total of DM was 10.2 months (95% CI 7.8-12.7) for TMT patients and 7.8 months (95% CI 5.7-9.9) for BMT patients. Those with pleural DM had the shortest OS: 1.5 months (95% CI 1.0-2.1) for TMT and 3.5 months (95% CI 3.3-3.6) for BMT.

Our analysis, representing the largest cohort of patients to date, demonstrates that DM as a first sign of relapse after local therapy of EAC patients is common. The TMT patients, with relatively less-advanced local EAC had a slightly lower frequency of DM compared to the BMT patients who more often had advanced local EAC. It is, of course, disappointing if a TMT patient develops DM (especially within 12 months of surgery). If we can identify patients who are destined to develop DM, we may be able to change the clinical algorithm (e.g. delay surgery for a certain period of time if the risk for DM is high); however, it is unclear what could be done for BMT patients destined to develop DM (prolonged induction systemic therapy?). Currently, there is no clinical or biomarker model associated with a high likelihood of the development of DM. Such a model would be desirable to explore various clinical approaches.

This study is retrospective and has all the limitations of a retrospective analysis. A strength of the study is that it has the largest number of patients ever reported for characterizing DM.

In conclusion, our data show that DM is a common occurrence after local therapy of patients with EAC. A clinical variables' model to predict DM would be desirable.

We acknowledge the U.T.M.D. Anderson Cancer Center Multidisciplinary Research Grants, CA138671, CA172741, CA129926 from NCI (JAA) and Generous Donors.

The authors have no financial or any other conflict of interest.