Introduction The correlation between SUV on PET-CT and prognosis in NSCLC has been the subject of much debate. We were interested in whether SUV_max values could be used to determine which cases of early stage (1–2) NSCLC were likely to progress.

Methods We reviewed all 93 histologically proven early stage NSCLCs seen at our tertiary centre over a one year period. We defined those with SUV_max in the upper quartile as the high SUV group and compared these with the remainder. Historical data were considered to allow subsequent outcomes to be established.

Results Median follow up was 772 days during which time there was a 17% mortality rate [all-cause median time to death 338 days]. The median SUV_max for the cohort was 10.1, and those in the upper quartile all had results over 15.0.

The high SUV_max group (n = 27) and low SUV_max group (n = 68) had similar baseline characteristics, received similar treatment regimens and there were no significant differences in tumour size between the groups. Disease progression and mortality were both significantly higher in those with SUV_max in the upper quartile, despite this group tending to have earlier disease (see Table 1).

Retrospective analysis using Youdin’s index suggested that the optimal threshold for predicting disease progression was not significantly different when cases with nodal involvement were excluded [SUV_max 15.0 vs 15.5]

Conclusions Our results suggest that SUV_max may indeed help identify those patients with early stage NSCLC at higher risk of progression. In our large cohort those with an SUV_max of >15 were over 3 times more likely to develop progressive disease than those with lower results and this was independent of tumour size or nodal involvement. Whether individuals in the higher-risk group would benefit from increased surveillance or adjuvant therapy remains to be established.