Multiparametric Imaging of Tumor Hypoxia and Perfusion with 18F-Fluoromisonidazole Dynamic PET in Head and Neck Cancer

Abstract

Tumor hypoxia and perfusion are independent prognostic indicators of patient outcome. We developed the methodology for and investigated the utility of multiparametric imaging of tumor hypoxia and perfusion with ¹⁸F-fluoromisonidazole (¹⁸F-FMISO) dynamic PET (dPET) in head and neck cancer. Methods: One hundred twenty head and neck cancer patients underwent 0- to 30-min ¹⁸F-FMISO dPET in a customized immobilization mask, followed by 10-min static acquisitions starting at 93 ± 6 and 160 ± 13 min after injection. A total of 248 lesions (≥2 cm³) were analyzed. Voxelwise pharmacokinetic modeling was conducted using an irreversible 1-plasma 2-tissue-compartment model to calculate surrogate biomarkers of tumor hypoxia (k₃), perfusion (K₁), and ¹⁸F-FMISO distribution volume. The analysis was repeated with truncated dPET datasets. Results: Substantial inter- and intratumor heterogeneity was observed for all investigated metrics. Equilibration between the blood and unbound ¹⁸F-FMISO was rapid in all tumors. ¹⁸F-FMISO distribution volume deviated from the expected value of unity, causing discrepancy between k₃ maps and total ¹⁸F-FMISO uptake and reducing the dynamic range of total ¹⁸F-FMISO uptake for quantifying the degree of hypoxia. Both positive and negative trends between hypoxia and perfusion were observed in individual lesions. All investigated metrics were reproducible when calculated from a truncated 20-min dataset. Conclusion:¹⁸F-FMISO dPET provides the data necessary to generate parametric maps of tumor hypoxia, perfusion, and radiotracer distribution volume. These data clarify the ambiguity in interpreting ¹⁸F-FMISO uptake and improve the characterization of lesions. We show total acquisition times can be reduced to 20 min, facilitating the translation of ¹⁸F-FMISO dPET into the clinic.

Keywords: 18F-fluoromisonidazole; FMISO; dynamic PET; head and neck cancer; hypoxia; perfusion.

FIGURE 1.

Sagittal images of 4 representative patients. (Left to right) Late 10-min ¹⁸F-FMISO PET/CT scan; TBR; k₃ map representing hypoxia-mediated entrapment of ¹⁸F-FMISO; K₁ map, a surrogate measure of perfusion; and ¹⁸F-FMISO DV, representing overall concentration of unbound ¹⁸F-FMISO relative to blood. Window levels are set according to scale bars and are adjusted for maximum visibility.

FIGURE 2.

(A) Line profile through TBR, k₃, and DV parametric maps for patient 1 from Figure 1 with HPV-negative, p16-positive T1N2c tumor originating in base of tongue. Minimum and maximum values of each parameter are normalized between 0% and 100% corresponding to actual parameter values provided in brackets in top of figure. (B) k₃–TBR voxelwise scatterplot for patient 1, color-coded according to DV. (C) Modeled time–activity curves (solid line) superimposed on measured time–activity curves (squares) from 2 voxels in hyperperfused and relatively normoxic (area 1 on Figure 1; blue) and hypoperfused and hypoxic tumor subregions (area 2 on Figure 1; red) for patient 2 with HPV-positive, p16-positive T1N2a tumor originating in tonsil. Parameter values for each voxel are given in top of figure. Also displayed is time–activity curve for IF (black). (D) TBR as function of time after injection is shown for patient 3 with HPV-positive, p16-positive T2N2b tumor originating in tonsil, whose lesion exhibited uniformly high DV (1.17). Also included are corresponding data for normoxic submandibular (SMG) and parotid glands. (E) Scatterplot of perfusion (K₁) and hypoxia (k₃) parameters for patient 3 (crosses; HPV-positive, p16-positive T2N2b tumor originating in tonsil) and patient 4 (dots; HPV-positive, p16-positive T3N1 tumor originating in tonsil), exhibiting a positive (Pearson r = 0.71) and negative (Pearson r = −0.61) trend between K₁–k₃, respectively.

FIGURE 3.

Correlation matrix of Pearson correlation between all investigated parameters. Upper right and lower left triangle regions with respect to the main diagonal of a matrix show coefficients derived on the voxel and regional level, respectively.

FIGURE 4.

(A) Scatterplot of mean intratumor k₃ and TBR for all analyzed lesions, color-coded according to mean intratumor DV. (B) Corresponding k₃–TBR intensity histogram for voxelwise values from pooled data. Included are nominal hypoxia thresholds based on TBR = 1.4 and regression-derived k₃ = 0.0031 min⁻¹.

FIGURE 5.

Frequency histogram of ¹⁸F-FMISO equilibration time T* (A) and ¹⁸F-FMISO DV (B) for pooled voxelwise data.

FIGURE 6.

(A) (Left) Scatterplot of average intratumor k₃ (top) and K₁ (bottom) for FD (abscissa) versus truncated 20-min dataset (TD; ordinate), color-coded according to root-mean-square deviation (RMSD) of residuals. (Right) Corresponding intensity histograms of voxelwise k₃ (top) and K₁ (bottom) values from pooled data. Identity (dashed) and regression (solid) lines are superimposed in all foregoing cases. (B) (Left to right) k₃ parametric maps as calculated with FD and TD, k₃ difference map (i.e., k₃(FD) − k₃(TD)), and equivalent maps for K₁, for all patient example tumors shown in Figure 1.