Ambulatory pulse oximetry as a clinical aid for the diagnosis and treatment response of radiation pneumonitis.

PURPOSE: Radiation pneumonitis (RP) is a clinical diagnosis, with no single best method of detection currently available. This study evaluated whether a decline between resting (rPO) and ambulatory (aPO) pulse oximetry (PO) levels after concurrent chemotherapy and radiation therapy (RT) can serve as a clinical aid in diagnosing and evaluating treatment response of grade 2-3 RP.

METHODS AND MATERIALS: Between March 2007 and November 2013, rPO and aPO values were obtained from 55 patients immediately after definitive thoracic RT and at each subsequent visit, for up to 4 visits. Median values of the decline from rPO to aPO were compared between those with and without subsequent RP. A logistic regression model was used to determine an association between a drop in PO and, independently, clinically defined RP.

RESULTS: RP was identified in 19 of 55 patients, with a median time to diagnosis of 56 days after RT. Twelve patients (22%) were diagnosed with grade 2 RP and 7 (13%) with grade 3 RP. According to a Wilcoxon rank sum test, the median calculated drop between rPO and aPO was greater in RP patients than in those without RP (median 4.21 and 1.01, respectively; P<.0001). After adjustment for total tumor dose and age, multivariate analyses revealed a 64.8% increase in the chance of RP development with every unit of decline in PO (P=.0014). After initiation of treatment with a corticosteroid, the mean difference in PO drop was compared with patients' baselines and demonstrated a statistically significant improvement, with peak PO value recovery after 2 weeks of corticosteroid therapy (P=.0001).

CONCLUSIONS: Patients diagnosed with RP demonstrated an early, measurable drop between rPO and aPO that was detected at or before diagnosis. Consequent recovery in PO followed treatment with corticosteroids. PO measurements are cost-effective and readily available, and they can be a valuable tool to aid in diagnosing RP and gauging treatment response.