*3.1. Qualitative and Semiquantitative Analysis of the PET/CT Images*

The biodistribution was as anticipated with minimal tracer seen in the blood pool with excretion through the urinary system. A few patients demonstrated liver and gallbladder uptake. There was minimal background activity noted in the fat and muscles. The mean SUVmax and SUVmean of the primary tumor on 18F-FDG PET/CT imaging was 22.1 ± 21.4 and 8.9 ± 7.6, respectively; while the mean SUVmax and SUVmean of the primary lesion on 68Ga-Nitroimidazole PET/CT imaging was 3.54 ± 1.5 and 1.8 ± 0.8 for the early time point and 3.43 ± 1.5 and 1.4 ± 0.4 for the delayed imaging. There was no statistical difference noted in the mean SUVmax and SUVmean of the primary lesion between the two time points on the 68Ga-Nitroimidazole scan (*p* = 0.461; *p* = 0.510, respectively). Qualitatively, 12 patients had discernable focal uptake above background, three patients had uptake similar to background and the remaining five patients had no uptake on 68Ga-Nitroimidazole scan. The mean TMR on 18F-FDG PET/CT and at both time points on

the 68Ga-Nitroimidazole scans were 38.3, 10.1 and 11.7, respectively. The mean TBRs on 18F-FDG PET/CT was 13.6 ± 11.3 and 13.5 ± 12.6 when using the aorta and ventricle as the blood pool surrogate. The median TBR on the delayed images on 68Ga-Nitroimidazole when using the aorta and ventricle were 2, while the mean TBR was 2.5 ± 1.6 and 3.3 ± 3.2. Figure 1 below demonstrates the semi-quantitative parameters on both 18F-FDG and 68Ga-Nitroimidazole PET/CT imaging.

**Figure 1.** The mean SUVmax, SUVmean, tumor to muscle ratio (TMR) and tumor to blood ratio (TBR) for 18F-FDG and 68Ga-Nitroimidazole at 30 and 60 min. \* PET: Positron Emission Tomography

The median MTV on the 18F-FDG PET/CT was 150.5 (18.41–548.93). The median hypoxic tumor volumes on the early and delayed images were 50 and 60.85, respectively. The hypoxic subvolume as a percentage of the MTV ranged from 2 to 98% (median 37%).

### *3.2. Immunohistochemistry Findings*

We successfully retrieved histological specimen for 15 patients. Five patients had negative staining for HIF-1α on immunohistochemistry. The remainder of the patients had varying intensities of staining for HIF-1α with three (3) patients having weak staining, two (2) had moderate intensity and five (5) patients had strong intensity. When considering the hypoxia score, seven patients were considered positive for hypoxia. Table 2 represents the intensity and percentage distribution of the HIF-1α expression within the tumor specimen. In 70% (n = 7) of the patients with any level of HIF-1α expression on immunohistochemistry, we demonstrated uptake on the 68Ga-Nitroimidazole hypoxia PET scan (Figures 2 and 3). Statistically there was a very weak, almost negligible correlation between these findings (r = 0.058; *p* = 0.837).


**Table 2.** Immunohistochemical analysis and correlation to 68Ga-Nitroimidazole imaging qualitative assessment.

Intensity of staining: 0 = no staining, 1 = weak intensity, 2 = moderate intensity, 3 = strong intensity. Sum of intensity and percentage distribution out of a total of 6 (see methods section for detailed description). Values above 3 were considered positive for hypoxia. Qualitative 68Ga-Nitroimidazole PET findings: 0 = no uptake, 1 = uptake similar to background, 2 = focal uptake above background, 3 = focal uptake markedly above background.

**Figure 2.** A 36-year-old female patient with FIGO stage II, squamous cell carcinoma of the cervix. (**A**). 18F-FDG PET transaxial image through the pelvis demonstrating uptake in the primary tumor (red arrow) with minimal uptake in the urinary bladder (blue arrow) (**B**). 68Ga-Nitroimidazole PET transaxial image through the same plane exhibiting spatially incongruent uptake with inhomogeneous uptake in parts of the tumor (red arrow) and intense activity in the urinary bladder (blue arrow). (**C**) is the corresponding CT only image in the same plane with target organs marked with arrows. The fractional hypoxic volume was 36%.
