ESTRO 2024 - Abstract Book

S1440

Clinical - Head & neck

ESTRO 2024

There were 198 OS events at a median of 2.5 years (IQR 2.9 years). For surviving patients (n = 459), the median follow-up was 6.9 years (IQR 3.6 years). There were 213 PFS events, including 113 cancer events and 100 competing events. 384 patients (58.4%) received intensified treatment. Median and mean Ω scores were 0.8 (IQR 0.1) and 0.8 (standard deviation [SD] 0.1) respectively. 381 patients (58%) had a high Ω score (> 0.8) and 276 (42%) a low score (<= 0.8). 246 patients (37.4%) were low -risk according to HN005 eligibility criteria. On logistic regression, increasing Ω score was associated with a marked increase in the odds of receiving intensified treatment (odds ratio [OR]: 18.0, 95% CI: 10.9 - 29.6, p < 0.001), in contrast to low-risk patients (i.e., HN005-eligible; OR: 1.09, 95% CI: 1.01 - 1.18, p = 0.03). Figure 1 demonstrates cumulative incidences of cancer events (solid lines) and competing events (dashed lines) for patients with high and low Ω scores (red and blue lines respectively, Figure 1A) and who were high - and low-risk (red and blue lines respectively, Figure 1B) per HN005 eligibility criteria.

On gce regression, the Ω score as a continuous variable predicted relative risk of cancer events (Ω ratio 1.19, 95% CI 1.04- 1.36, p=0.01). An Ω cut - off of >= 0.81 appeared optimal (Ω ratio 2.14, 95% CI 1.18 -3.90, p=0.01); 337 patients [51.3%] had an Ω score >= 0.81. Conversely, HN005 trial eligibility did not predict relative risk (Ω ratio 1.32, 95% CI 0.62-2.83, p=0.47).

Conclusion: