Early Response Prediction of Multiparametric Functional MRI and 18F-FDG-PET in Patients with Head and Neck Squamous Cell Carcinoma Treated with (Chemo)Radiation

Abstract

Background: Patients with locally-advanced head and neck squamous cell carcinoma (HNSCC) have variable responses to (chemo)radiotherapy. A reliable prediction of outcomes allows for enhancing treatment efficacy and follow-up monitoring.

Methods: Fifty-seven histopathologically-proven HNSCC patients with curative (chemo)radiotherapy were prospectively included. All patients had an MRI (DW,-IVIM, DCE-MRI) and ¹⁸F-FDG-PET/CT before and 10 days after start-treatment (intratreatment). Primary tumor functional imaging parameters were extracted. Univariate and multivariate analysis were performed to construct prognostic models and risk stratification for 2 year locoregional recurrence-free survival (LRFFS), distant metastasis-free survival (DMFS) and overall survival (OS). Model performance was measured by the cross-validated area under the receiver operating characteristic curve (AUC).

Results: The best LRFFS model contained the pretreatment imaging parameters ADC_kurtosis, K_ep and SUV_peak, and intratreatment imaging parameters change (Δ) Δ-ADC_skewness, Δ-f, Δ-SUV_peak and Δ-total lesion glycolysis (TLG) (AUC = 0.81). Clinical parameters did not enhance LRFFS prediction. The best DMFS model contained pretreatment ADC_kurtosis and SUV_peak (AUC = 0.88). The best OS model contained gender, HPV-status, N-stage, pretreatment ADC_skewness, D, f, metabolic-active tumor volume (MATV), SUV_mean and SUV_peak (AUC = 0.82). Risk stratification in high/medium/low risk was significantly prognostic for LRFFS (p = 0.002), DMFS (p < 0.001) and OS (p = 0.003).

Conclusions: Intratreatment functional imaging parameters capture early tumoral changes that only provide prognostic information regarding LRFFS. The best LRFFS model consisted of pretreatment, intratreatment and Δ functional imaging parameters; the DMFS model consisted of only pretreatment functional imaging parameters, and the OS model consisted ofHPV-status, gender and only pretreatment functional imaging parameters. Accurate clinically applicable risk stratification calculators can enable personalized treatment (adaptation) management, early on during treatment, improve counseling and enhance patient-specific post-therapy monitoring.

Keywords: MR diffusion weighted imaging; MR dynamic contrast enhanced; PET/CT; functional imaging; head and neck; outcomes analysis; prognosis; radiation therapy/oncology; squamous cell carcinoma; tumor response.

Figure 1

Overview of ADC, IVIM, DCE and FDG-PET imaging acquisition in a patient with left tonsillar carcinoma before and 10 days into (chemo)radiotherapy, and with whom locoregional failure occurred. The upper row shows the ADC map on which the tumor is delineated, in order to extract DWI and IVIM parameters. The subtle spatial mismatch was due to a slightly different angulation of the neck. The ADC histogram shows high pretreatment positive ADC_skewness (blue histogram), and an increase towards a higher intratreatment ADC_skewness (orange histogram). Furthermore, a high pretreatment ADC_kurtosis was associated with LRF (orange line). The middle row shows the population-based arterial input function (AIF) and a tumor concentration time curve. The images are DCE images at the 75 temporal phase, on which a colored functional map of the parameter K^trans is superimposed in the delineated tumor. The color scale shows the range between 0 and 1 mMol/L. The ¹⁸F-FDG-PET image in the lowest row shows the tumor delineation (red ROI) on the attenuation-corrected ¹⁸F-FDG-PET image (black/white SUV scale ranges between 0 and 10), with a threshold of >50% SUV_peak and in anatomical correlation with a diagnostic CT scan.

Figure 2

Flowchart of patient inclusion.

Figure 3

The median of significant multivariate prognostic pretreatment, intratreatment and delta-parameters per single imaging modality for locoregional recurrence-free survival, distant metastasis-free survival and overall survival (see Tables S2 and S3 for the complete tables). HPV-negative tumors were scored with the number 0 and HPV tumors with the number 1.

Figure 4

Significant multivariate prognostic pretreatment, intratreatment and delta parameters of all imaging techniques combined for locoregional recurrence-free survival, distant metastasis-free survival and overall survival (See Tables S2 and S3 for the complete tables). Overall, for each patient outcome, the intercept and the slopes per median parameter was shown. The median slopes were found lower in patients with locoregional control (LRC) than locoregional failure (LRF), lower in no distant metastasis (no DM) than distant metastasis (DM), and lower in survival than death, which resulted in a lower risk for an adverse outcome. HPV-negative tumors were marked with a 0, and HPV-positive tumors with a 1. Gender was marked with a 0 for females and 1 for males.

Figure 5

Kaplan–Meier survival curves and the log-rank test for the most optimal prognostic models: (A) the combination of pretreatment, intratreatment and delta parameters for the cumulative incidence of locoregional recurrence, divided into high/medium/low risk groups; (B) prediction model with pretreatment parameters for the cumulative incidence for distant metastasis, divided into high/medium/low risk groups; and (C) prediction model with pretreatment parameters prognostic for overall survival, divided into high/medium/low risk groups.

Figure 5

Kaplan–Meier survival curves and the log-rank test for the most optimal prognostic models: (A) the combination of pretreatment, intratreatment and delta parameters for the cumulative incidence of locoregional recurrence, divided into high/medium/low risk groups; (B) prediction model with pretreatment parameters for the cumulative incidence for distant metastasis, divided into high/medium/low risk groups; and (C) prediction model with pretreatment parameters prognostic for overall survival, divided into high/medium/low risk groups.