PET Use for NSCLC Management Has Altered Resource Utilization, Increased Cost

Charles Bankhead

November 2013, Vol 4, No 9 - Economics of Cancer Care

Atlanta, GA—The adoption of positron emission tomography (PET) imaging for the management of patients with non–small-cell lung cancer (NSCLC) has had a mixed impact on clinical and economic outcomes, according to findings from a large retrospective cohort study.

The use of PET was associated with upstaging of early-stage disease and a reduction in “futile” attempts to control occult metastatic disease. As the adoption of PET increased, the use of radiotherapy and surgery declined, as did short-term inpatient costs. The use of chemotherapy increased, which was associated with higher overall costs.

“The ability of PET to affect patient management, healthcare resource use, and costs remains an important area of ongoing research,” Michaela A. Dinan, PhD, Medical Instructor, Department of Medicine, Duke University School of Medicine, Durham, NC, said at the 2013 American Society for Radiation Oncology annual meeting.

The study represented an extension of previous work by Dr Dinan and colleagues to ascertain changes in the use of imaging in Medicare patients with cancer. That analysis showed that from 1999 to 2006, the use of PET in Medicare patients with cancer increased substantially (Dinan MA, et al. JAMA. 2010;303:1625-1631). The overall cost of cancer care increased by 2% to 5% annually during the study period, but the cost of imaging increased by 5% to 10% annually.

The greatest increase in PET utilization occurred in patients with NSCLC. At the start of the study, few, if any, patients underwent PET imaging. By the end, more than 50% of patients had PET imaging at some point. “We found that no matter which subgroup of patients we looked at, all patients experienced large increases in PET use during the study period,” said Dr Dinan.

The team also found that increasing the use of PET was associated with the identification of more patients with advanced disease, which led to a closer look at resource utilization. “The presumption that appropriate detection of occult metastatic disease should reduce associated healthcare costs remains an open question at the population level,” said Dr Dinan.

The Study Results
To address the question, the team used the National Cancer Institute’s Medicare-linked Surveillance, Epide­miology, and End Results database to identify patients with NSCLC between 1996 and 2005, examining treatment utilization and the associated costs accrued within the first year after diagnosis.

Dr Dinan and colleagues hypothesized that the increasing use of PET would be associated with a decreased use of radiotherapy and surgery and lower short-term and total healthcare costs in the management of NSCLC.

Patients were grouped by years according to the evolution of PET use, resulting in cohorts of 9638 patients in the pre-PET era (1996-1997), 9551 patients in the initial-PET era (2000-2001), and 11,814 patients in the post-PET era (2004-2005).

The proportion of patients with NSCLC who underwent PET increased from 0.2% in the pre-PET period to 17.5% during initial PET use to 53.5% in the post-PET period (P <.001 for trend). This increased use was not associated with significant improvements in 2-year survival (28.7% in 1996-1997 and 29.9% in 2004-2005).

As hypothesized, the use of radio­therapy declined from 49.4% of patients in the pre-PET era to 42.9% in the post-PET period; the use of any type of surgical resection decreased from 29.1% to 24.7%; and the proportion of patients who received no treatment increased from 22.0% to 24.7% (P <.001 for all).

Inpatient costs averaged $28,924 per patient in the pre-PET period and $26,944 in the post-PET period. Outpatient costs increased from $18,411 to $25,266, and costs associated with PET use increased from $0 to $741, resulting in a net increase in total costs from $47,335 to $52,209 (P <.001 for all).

Using 1996 to 1997 as the reference point, the rate of surgical resection was significantly lower each year from 2001 (odds ratio [OR], 0.87) to 2004 (OR, 0.72). The proportion of patients receiving radiation therapy declined significantly starting in 2002 (OR, 0.85), and was 19% lower in
2004 (OR, 0.81).

In contrast, the growth in use of chemotherapy for NSCLC began in the first year after the reference period (OR, 1.30), increasing by more than 2-fold by 2003 (OR, 2.08) and 2004 (OR, 2.52).

“Consistent with previous research, this study suggests that PET has played a role in the upstaging of early-stage NSCLC and a corresponding reduction in futile attempts at local control of occult metastatic disease,” Dr Dinan concluded.