These have not assessed the differential patterns in plasma cytokine levels for patients receiving RT alone versus chemoRT. In the present study, the overall concentration of cytokines were different dependent upon treatment group for Eotaxin, IL-33, IL-6, MDC, MIP-1a and VEGF. In both groups, the peak elevation in plasma cytokine concentration for MIP-1a occurred at 4 weeks into treatment, whilst peak elevation in Eotaxin and VEGF occurred at 12 weeks after treatment completion. Changes in plasma concentrations of cytokines varied considerably between treatment groups at other time points. Our findings suggest that future studies investigating the kinetics of these plasma levels should not uniformly group patients receiving RT alone and chemoRT BAY-60-7550 together. In several previous clinical studies of patients treated for NSCLC there have been contradictory reports of associations between RT induced blood cytokine levels and clinical toxicity. A study by Arpin et al. also found changes in IL-6 to be prognostic for radiation pneumonitis, along with combined covariations of IL-6 and IL-10. Similar to our study, TNF-a was not correlated to toxicity. In a study by Crohns et al., VEGF, TNF-a, IL-1b, IL-6 and IL-8 levels in the serum were analysed in patients receiving various regimes of RT with mean dose of 46.9 Gy. These investigators were not able to demonstrate any significant changes from the baseline levels of these cytokines at two weeks or 3 months after the commencement of RT. Similarly, a study by Ru˝be et al. measured levels of TGF-b and IL-6 weekly during RT and could not find correlation with symptomatic pneumonitis and plasma level cytokines levels. In their study, patients received a range of treatments including definitive RT alone, definitive chemoRT, low dose twice weekly palliative accelerated RT. This is in contrast to several reports from Anscher’s group and a study from Zhao et al. indicating that elevation of TGF-b late during RT is associated with risk of pulmonary toxicity. In a study by Chen et al., levels of IL-1a and IL-6 but not TNFa were consistently elevated prior to and throughout treatment in patients whom developed radiation pneumonitis. In this same patient cohort, levels of e-selectin, l-selectin, TGF-b1 and bFGF varied but were not correlated with radiation pneumonitis. Again this cohort had considerable treatment heterogeneity, with 4 of the total 24 patients not having NSCLC, with an average delivered radiation dose of 60–64 Gy, and 3 of 15 patients having had their chemotherapy delivered neoadjuvantly prior to the radiation course. In the context of the previous literature, the strength of our study lies in the standardised treatment characteristics in our patient cohort and the large panel of cytokines tested. We discovered that in addition to IL-6, early changes in plasma levels of four previously unreported cytokines were associated with the risk of pulmonary toxicity. It is important to recognise the limitations of this study. In this study we were not able to control for potential confounding effects of patient stage and volume of irradiated amongst the RT and chemoRT cohorts.