Do We Have Adequate Surveillance in Cancer Survivor Care?

Published on: 

A poster discussion session at the 2017 Annual Meeting of the American Society of Clinical Oncology examined retrospective surveillance data in 3 different cancers: non-small cell lung cancer, head and neck cancer, and colorectal cancer.

Survivor care is important among patients who have undergone treatment for cancer. Survivorship care and follow-up can help patients discuss issues that may be associated with their treatment, share their financial concerns, and, most importantly, ensure disease-free survival. A poster discussion session at the 2017 Annual Meeting of the American Society of Clinical Oncology examined retrospective surveillance data in 3 different cancers: non-small cell lung cancer (NSCLC), head and neck cancer (HNC), and colorectal cancer (CRC). Katherine Van Loon, MD, MPH, from the University of California, San Francisco, was the discussant.

The first study, conducted by researchers affiliated with multiple healthcare systems in Philadelphia, Pennsylvania, evaluated the impact of radiation therapy and latency period on the risk of developing new primary lung cancers after a head and neck cancer after HNC.

The population-based study of 85,154 patients with HNC in the Surveillance, Epidemiology, and End Results (SEER) database found a total of 4209 patients with new primary lung cancers. Compared with the no radiation group, those who received radiation therapy had a higher incidence of the primary lung cancers across all latency periods, from less than 1 year (standardized incidence ratio, 3.45 vs 2.18, respectively) to 10 to 15-year follow-up (standardized incidence ratio, 3.19 vs 1.88, respectively). The highest incidence for the radiation-treated group was observed in the 1 to 3-years latency period (4.57 vs 2.41).

The authors concluded that in patients with HNC, the risk of developing a new primary lung cancer is associated with radiation treatment, with the greatest risk observed within 10 years of the initial HNC diagnosis. They also recommend that screening for patients who smoke should be considered, especially within 10 years of the primary HNC diagnosis.

Van Loon said that while the large historic sample size from the SEER data was a significant strength of the study, retrospective analysis place limitations on the observations. Further, the predominance of squamous cell carcinoma raises questions on de-novo vs metastatic nature of the observed lung cancer. She also pointed out the lack of data on patient exposure to risk factors as a study limitation.


The next study evaluated the impact of post-treatment surveillance in CRC with the purpose of this study was to determine if the intensity of post-treatment surveillance is associated with time to recurrence detection, treatment, or overall survival (OS). The authors examined the primary records of 10,636 stage I to III patients with CRC from Commission on Cancer accredited hospitals who were diagnosed between 2006 and 2007, and data was merged with records in the National Cancer Database. A predicted and observed number of imaging and carcinoembryonic antigen (CEA) tests per patient were determined and clustered by hospital; patients were then categorized into high or low intensity categories.

Of the 6279 patients, those who underwent high-intensity imaging (50.6%) or CEA surveillance (51.2%) in the 3 years after CRC treatment, had a mean of 2.9 imaging studies and 4.7 CEA tests. Patients with low-intensity imaging underwent a mean of 1.4 imaging studies and 1.6 CEA tests.

Patients’ 5-year recurrence rates were no different based on the intensity of surveillance: stage 2 and 3 patients who underwent high-intensity imaging and CEA testing had a slightly higher resection rate, without any improvement in 5-year OS.

The authors concluded that higher intensity surveillance was not associated with earlier detection of recurrent disease or improved OS. It did, however, result in a slightly higher resection rate. They went on to recommend less frequent testing for surveillance in patients with CRC.

Van Loon praised the highly annotated design and large real-world sample of patients used by the trial. “The findings add to results of FACS and GILDA trials, so intensive follow-up may not improve OS for CRC survivors,” she added.

Questions remain about whether a subset of patients with recurrence may benefit from more intensive surveillance, Van Loon concluded.

The final abstract discussed was the receipt of recommended surveillance with imaging in survivors of early stage NSCLC.

Lifelong imaging surveillance for early cancer detection is recommended in lung cancer survivors who have a high risk for recurrence and second cancers and a 5-year survival of only 50%. The authors of the study examined the rates and determinants of regular surveillance imaging in NSCLC survivors.

computerized tomography

Examination of the SEER-Medicare linked database identified 10,680 patients with stage I and II NSCLC, diagnosed over the 10 years between 2001 and 2011 and treated with surgery. Patients were censored at the time of recurrence/second cancer, loss of insurance or 3 months before death. In this population, receipt of and/or positron emission tomography imaging during the surveillance periods of 7 to 18, 19 to 30, 31 to 42, and 43 to 60 months from the date of surgery was assessed.

The study found that 79% and 40% survivors had follow-up information until the end of 30- and 60-month surveillance periods, respectively. With a median follow-up of 7.6 years, 71% of the survivors received imaging in the first 18 months after surgery, but only 56% and 44% of survivors continued to receive regular imaging by 30- and 60-month of follow-up periods, respectively.

Survivors, the analysis found, were less likely to receive imaging if they were older (≥80 years), black, not married, had rural residence, did not receive adjuvant therapy, had stage I disease (compared to stage II) and were diagnosed in 2006 or earlier. In adjusted analysis, survivors receiving recommended imaging up to 18 months from surgery had improved survival compared to survivors who did not (Hazard ratio [HR], 0.86; 95% CI, 0.81 to 0.92). Survival benefit was also observed in survivors who had regular imaging up to 5 years from surgery (HR, 0.68; 95% CI, 0.60 to 0.76).

The fact that more than 50% of the lung cancer survivors did not receive recommended long-term surveillance imaging had the authors conclude that adherence to regular surveillance even at 5 years from initial surgery is associated with improved survival.

Van Loon noted that the large cohort size was a definite plus for the study, as was the use of the SEER-Medicare linked database. The limitations of the study were that it was retrospective, was missing the population younger than 66 years, and did not use non-Medicare sources. Additionally, 60% of patients were missing follow-up at 60 months.