ADVERTISEMENT
Diffusion-weighted magnetic resonance imaging as an early predictive marker of chemoradiotherapy response in squamous cell carcinoma of the anus: An individual patient data meta-analysis
Background
Patients with squamous cell carcinoma of the anus (SCCA) are at risk of locoregional recurrence after chemoradiotherapy (CRT). An early prognostic marker of treatment response would enable timely individualisation of treatment for patients with SCCA. Diffusion-weighted (DW) magnetic resonance imaging (MRI) features, such as apparent diffusion coefficient (ADC), may be prognostic; however, few trials with small patient numbers exist for SCCA. Using an individual patient data meta-analysis, we aimed to investigate in a larger cohort whether ADC based histogram parameters from paired DW-MRI at baseline and during CRT correlate with local or any relapse in patients with SCCA.
Methods
We included 3 prospective trials, run between 2013 to 2017, in patients receiving radical CRT for SCCA who had paired DW-MRI at baseline and during week 2 of treatment in our individual patient data meta-analysis. Individual clinical patient and treatment data, volumes of the macroscopic tumour and involved nodes >2 cm, and ADC based histogram parameters (ADCmax, ADCmean, ADCmin, skewness, kurtosis, and standard deviation) were combined into one dataset (one-stage meta-analysis). The association between ADC parameters from scans at the two time points and the percentage change in parameters between scans with local or any relapse was analysed using logistic regression. Deriving the area under the curve (AUC), we assessed each parameters ability to predict relapse. Based on findings from other tumour sites, and results from one of the trials in our meta-analysis, we investigated the effect of a change < 20% in ADCmean between the two scans.
Results
There were 90 targets analysed from 82 patients, including primary tumours and lymph nodes >2cm, admitted in the meta-analysis. Among all patients, 13.4% (n=11) relapsed locally and 19.5% (n=16) had any relapse. No ADC based histogram parameter at baseline, during week two of treatment or as percentage change between scans was associated with local or any relapse (p>0.05 for all). Patients with a change in ADCmean < 20% between scans had a higher local relapse rate (18.0%) than patients with change in ADCmean ≥20% (9.3%).
Conclusions
Our individual patient data meta-analysis did not identify definite parameters from paired DW-MRI at baseline or during week 2 of CRT as a useful biomarker in SCCA. However, further research should employ different MRI sequences and parameters, explore combinations of other parameter and imaging modalities, and utilise novel analysis methods.
Clinical trial identification
Clinical trials included in our individual patient data meta-analysis: NCT02145416 ACTRN12614001219673 NCT01937780.
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosures
All authors have declared no conflicts of interest.
