Interim CT shrinkage and FDG-PET uptake predicted head and neck outcomes

KEY POINTS

• The systematic review identified 98 studies evaluating tumour response during curative head and neck radiotherapy using 16 different assessment modalities. FDG-PET was studied in 32 reports involving 1,159 patients, MRI in 31 involving 1,732 patients, and CT in 11 involving 1,195 patients.
• Timing of interim assessment was highly heterogeneous, ranging from the first to final treatment week and being reported variously by treatment week, fraction number, delivered dose, or days after radiotherapy initiation.
• Low tumour-volume reduction on interim CT showed the strongest prognostic association. Patients with low shrinkage had substantially poorer locoregional control, with a pooled hazard ratio of 4.6 (95% CI 2.4–8.8).
• High absolute SUVmax on interim FDG-PET was associated with poorer overall survival, with a pooled hazard ratio of 1.97 (95% CI 1.19–3.24), and poorer locoregional control, approximately HR 1.7 (95% CI 1.1–2.8).
• Change in mean apparent diffusion coefficient on MRI did not demonstrate convincing prognostic value for locoregional control, with a pooled hazard ratio close to 1.0 (95% CI approximately 0.9–1.0).
• Only 19.4% of the 98 studies contained sufficiently compatible data for quantitative synthesis; for MRI, the proportion was just 12.9%. Of 26 corresponding authors contacted for additional data, only eight replied and four ultimately supplied usable information.
• All studies included in the quantitative meta-analysis had moderate or high risk of bias. SUVmax and tumour-shrinkage cut-offs differed between studies, and the review specifically cautions that a prognostic association does not establish that response-guided dose escalation or de-escalation will improve outcomes.

CLINICAL TAKEAWAY

SOURCE

Clinical Oncology