Tumor Hypoxia as a Barrier in Cancer Therapy: Why Levels Matter

Abstract

Hypoxia arises in tumor regions with insufficient oxygen supply and is a major barrier in cancer treatment. The distribution of hypoxia levels is highly heterogeneous, ranging from mild, almost non-hypoxic, to severe and anoxic levels. The individual hypoxia levels induce a variety of biological responses that impair the treatment effect. A stronger focus on hypoxia levels rather than the absence or presence of hypoxia in our investigations will help development of improved strategies to treat patients with hypoxic tumors. Current knowledge on how hypoxia levels are sensed by cancer cells and mediate cellular responses that promote treatment resistance is comprehensive. Recently, it has become evident that hypoxia also has an important, more unexplored role in the interaction between cancer cells, stroma and immune cells, influencing the composition and structure of the tumor microenvironment. Establishment of how such processes depend on the hypoxia level requires more advanced tumor models and methodology. In this review, we describe promising model systems and tools for investigations of hypoxia levels in tumors. We further present current knowledge and emerging research on cellular responses to individual levels, and discuss their impact in novel therapeutic approaches to overcome the hypoxia barrier.

Keywords: cellular response; hypoxia level; imaging; model system; oxygen sensing; radiotherapy resistance; tumor microenvironment.

Figure 1

Distribution of hypoxia levels in tumors. (A) Illustration showing distribution of hypoxia levels from mild to severe in a section through a solid tumor. Non-hypoxic levels are seen close to capillaries (in red). (B) Close-up of a region in (A), showing the gradient in hypoxia levels from a capillary towards severe hypoxia. Symbols for different cell types and extracellular matrix (ECM) are indicated. (C) Illustration of how oxygen enhancement ratio (OER) increases with increasing oxygen concentration or tension in tumors, where OER is defined as the ratio between the dose needed to cause the same harmful effect to cells under anoxia and when oxygen is present. Data relative to maximum OER are shown and the curve is based on the study by Koch and coworkers described in [3]. The ranges indicated refer to the scale of the x-axis and are median pO₂ values reported across tumor types, median pO₂ cutoff for hypoxic fractions associated with poor radiotherapy outcome across tumor types, and median oxygen concentration reported for activation of hypoxia inducible factors (HIF). The pO₂ data are collected from [5] and the HIF data are based on [6,7]. (A–C) Hypoxia levels are indicated by the color code, with approximate oxygen concentrations (% O₂) and tensions (mmHg) provided by the x-axis in (C).

Figure 2

Model systems for studying hypoxia levels. (A) 2-dimensional (2D) monolayer cell cultures. Exposed to specific oxygen concentrations in a gas chamber. (B) 2D monolayer cell culture with an oxygen gradient created by a plate inserted into the petri dish at one end of the culture to limit oxygen and nutrient exchange with media. (C) Spheroid with a gradient in hypoxia levels from the periphery (non-hypoxic) towards the center (necrosis). (D) Animal model, showing a mouse with tumor grown on the back. The distribution of hypoxia levels in a section through the tumor is indicated. (A–D) hypoxia levels are indicated by the color code, with approximate oxygen concentrations (% O₂) and tensions (mmHg) provided by the x-axis in Figure 1C.

Figure 3

Invasive methods for quantification of hypoxia levels. (A) Polarographic needle electrodes for recording of pO₂ along tracks in a tumor. Frequency distribution of recorded pO₂ values is indicated. (B) Pimonidazole staining intensity in a histologic section from a xenograft tumor vs. distance from necrosis (below). The histologic section is shown above. (From “MRI Distinguishes Tumor Hypoxia Levels of Different Prognostic and Biological Significance in Cervical Cancer”. by Hillestad, T.; Hompland, T.; Fjeldbo, C.S.; Skingen, V.E.; Salberg, U.B.; Aarnes, E.-K.; Nilsen, A.; Lund, K.V.; Evensen, T.S.; Kristensen, G.B.; et al. 2020, Cancer Res., 80, 3993–4003, Copyright 2020 by American Association for Cancer Research [27]). (C) Spheroids indicating proteins upregulated at different hypoxia levels (above), and the combined expression data (below). (D) Gene expression signatures associated with defined hypoxia levels (left), and pie chart showing fractions of tumor with the defined level (right). (A,C,D) Hypoxia levels are indicated by the color code, with approximate oxygen concentrations (% O₂) and tensions (mmHg) provided by the x-axis in Figure 1C.

Figure 4

Imaging of hypoxia levels. (A) Optical pO₂ imaging of tumor in a dorsal skinfold chamber. The image with hypoxia level distribution is indicated. (B) Electron paramagnetic resonance (EPR) imaging of a tumor grown on the mouse back. Images with hypoxia level distribution are indicated. (C) Illustration of a positron emission tomography (PET) image showing uptake of hypoxia specific tracer (left), and pO₂ image based on the converted PET signal (right). The relationship between uptake (PET signal) and pO₂ is indicated. (D) Diffusion weighted (DW) magnetic resonance (MR) images showing apparent diffusion coefficient (ADC), reflecting oxygen consumption, (left, upper) and fractional blood volume (fBV), reflecting oxygen supply (left, lower), and the combined hypoxia level image (right). The images were collected from a patient with prostate cancer, and the DW-MR images are overlaid on T₂ weighted images of the pelvis. (A–D) Hypoxia levels are indicated by the color code, with approximate oxygen concentrations (% O₂) and tensions (mmHg) provided by the x-axis in Figure 1C.

Figure 5

Hypoxia sensing and response. Illustration of hypoxia sensing mechanisms (upper), biological responses (middle) in relation to hypoxia level from mild to severe, and subsequent treatment resistance (lower). HIF, Hypoxia-inducible factor; UPR, unfolded protein response; MTOR, mammalian target of rapamycin; ROS, reactive oxygen species; EMT, epithelial-mesenchymal transition. Hypoxia levels are indicated by the color code, with approximate oxygen concentrations (% O₂) and tensions (mmHg) provided by the x-axis in Figure 1C. Unknown hypoxia level is indicated in grey.

Figure 6

Glucose uptake at different hypoxia levels by combination of multimodality images. (A) Magnetic resonance (MR)-based hypoxia level image of a cervix tumor. (B) Glucose uptake in the same tumor by ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG)-positron emission tomography (PET) imaging. (C) Glucose uptake as a function of hypoxia level based on a voxel-by-voxel analysis of the co-registered images in (A,B). The hypoxia levels were divided into 20 sublevels, and the mean PET signal of each sublevel is plotted. (A,B) The hypoxia level and glucose uptake images are overlaid on T₂ weighted images of the pelvis. (A–C) Hypoxia levels are indicated by the color code, with approximate oxygen concentrations (% O₂) and tensions (mmHg) provided by the x-axis in Figure 1C.