Imaging of Tumor Hypoxia With Radionuclide-Labeled Tracers for PET

The hypoxic state in a solid tumor refers to the internal hypoxic environment that appears as the tumor volume increases (the maximum radius exceeds 180-200 microns). This state can promote angiogenesis, destroy the balance of the cell’s internal environment, and lead to resistance to radiotherapy and chemotherapy, as well as poor prognostic factors such as metastasis and recurrence. Therefore, accurate quantification, mapping, and monitoring of hypoxia, targeted therapy, and improvement of tumor hypoxia are of great significance for tumor treatment and improving patient survival. Despite many years of development, PET-based hypoxia imaging is still the most widely used evaluation method. This article provides a comprehensive overview of tumor hypoxia imaging using radionuclide-labeled PET tracers. We introduced the mechanism of tumor hypoxia and the reasons leading to the poor prognosis, and more comprehensively included the past, recent and ongoing studies of PET radiotracers for tumor hypoxia imaging. At the same time, the advantages and disadvantages of mainstream methods for detecting tumor hypoxia are summarized.

information between blood and tissue fluids. However, the vasculature of tumors differs from that of normal tissues. This structural difference increases the diffusion distance between the supplying vessels and tumor cells, forming a diffusion gradient and resulting in insufficient oxygen supply to the tumor cells that are away from the vessels, eventually leading to tissue hypoxia (1). As the volume of solid tumors continues to increase, especially when larger than 180-200 microns, coupled with the uncontrolled growth of tumor cells that consume high levels of oxygen and nutrients, tumor tissues will have insufficient local tissue perfusion or oxygen supply and lack of nutrients, resulting in the production of a series of acidic wastes and leading to extracellular acidosis. Collectively, these events ultimately result in tumor cells having a hypoxic microenvironment (2).
Hypoxia-inducible factor (HIF)-1 is a nuclear transcription factor that is widely present in mammalian cells. HIF-1 activates selected genes by transcription. These genes increase cell survival and proliferation to help cancer cells adapt to hypoxia by increasing angiogenesis, altering metabolism, degrading the extracellular matrix, inducing tumor cell dedifferentiation, and arresting the cell cycle (3)(4)(5). When the oxygen concentration is normal, HIF-1a is ubiquitinated and can be rapidly hydrolyzed by the ubiquitin-proteasome system. Therefore, HIF expression in normoxic cells is low (6). However, hypoxia prevents the ubiquitination of HIF-1a and induces an increase in the expression of HIF-1a (7). The transcriptional activity of HIF-1 also increases with the degree of hypoxia, resulting in activation of the p53 gene and expression of the cyclin-dependent kinase inhibitor (CKI) (8). Simultaneously, HIF-1a can also directly bind to the tumor suppressor gene p53 to promote p53dependent cell apoptosis (9). In hypoxic conditions, the inhibition of prolyl hydroxylase stabilizes HIF-1a, followed by its binding to HIF-1b and transportation to the nucleus. The transcriptional activator protein p300/CBP is recruited through a transcriptional activation region of subunit a, and then interacts with the hypoxia response element (HRE), composed of the core sequence 5′-RCGTG-3′ of the promoter region of the target gene, such as carbonic anhydrase-IX (CA-IX). The downstream factors or enzymes are regulated and can then play a regulatory role (10) (Figure 1).
HIF can regulate vascular endothelial growth factor (VEGF), transforming growth factor (TGF), tumor necrosis factor (TNF), and angiopoietin-2 (ANG-2) receptor, among others (11), to induce the proliferation of vascular endothelial cells. While inducing new blood vessels, HIF also regulates the vascular system of the host for the metastasis of tumor cells (12). Thus, the tumor can not only obtain nutrients and oxygen from the host through the new tumor blood vessels, but can also continuously deliver metastatic cells to the host through these blood vessels, while continuing to grow and inducing bloodvessel formation in other parts of the body, ultimately leading to tumor invasion and migration (13,14). It is noteworthy that the formation of new blood vessels may lead to reoxygenation of hypoxic areas. Using current standard radiotherapy, hypoxic cells can also gradually become reoxygenated (REOX) during the treatment process due to the killing of the radiosensitive oxygen-consuming cells near the capillaries, as the remaining cells can obtain more nutrients and oxygen (15). Activated HIF can also induce the formation of tumor cell stemness and screen out more aggressive cells.
Additionally, increasing experimental evidence suggests that CA-IX plays a direct role in many tumor phenotypic characteristics caused by hypoxia and acidosis, including increased local adhesion during cell proliferation, unstable cell contact, tumor interstitial crosstalk, and maintaining the stem cell phenotype, signal transduction, and other cancer-related phenomena (16). Importantly, CA-IX is not expressed in most normal tissues except in the stomach and gallbladder epithelium (17). However, it is usually overexpressed in tumors and supports tumor cell migration and invasion. Inhibiting the catalytic activity of CA-IX can significantly improve chemotherapeutic sensitivity or radiosensitivity. Similarly, inhibiting CA-IX can enhance the anti-angiogenic effect of anti-VEGF antibodies (18).

Significance
Under hypoxic conditions, up-regulated HIF can regulate a variety of cancer phenotypes, causing various adverse changes in tumors and tissues ( Figure 2).
Hypoxia can cause genetic damage. The rapid change in oxygen concentration produces free radicals, which act on tissue cells and cause DNA damage (19). Moreover, hypoxia may lead to the overexpression of tumor suppressor genes, proapoptotic factors, and anti-apoptotic factors to salvage damaged cells, but may make these cells polyploid. A study by Nelson et al. (20) shows that these cells have certain defects in their inherent apoptotic pathways, which are important signs of cancer and precancer.
The most important biological effect caused by hypoxia is resistance to radiotherapy and chemotherapy, which is directly related to disease prognosis. Molecular oxygen is very important for the sensitivity of radiotherapy. In conventional photon-based radiotherapy or light ion therapy, such as those using protons, oxygen has a high electron affinity and plays a central role in the fixation process (21,22). Hypoxia reduces the ability of the drug to produce free oxygen radicals, resulting in hypoxic cells that are less sensitive to chemotherapy.
Another interesting phenomenon is that when the hypoxic state causes cell death, the migration of tumor-associated macrophages (TAM) to the hypoxic area is inhibited, thereby preventing TAM from removing dead and damaged cells. Other effects of TAM include inhibition of the immune response of the body to tumor cells and inhibition of the activation of adjacent macrophages. These phenomena promote angiogenesis and lymphangiogenesis, thereby promoting tumor growth, migration, and transformation (23,24). Hypoxia-related TAM can upregulate HIF-1 and HIF-2 expression both in vivo and in vitro (25).
E-cadherin is an important intercellular adhesion molecule that mainly maintains the integrity and polarity of epithelial cells.
Loss of E-cadherin is a sign of tumor epithelial-to-mesenchymal transition (EMT) in tumor cells and is related to hypoxiainduced HIF activation.

Definition
There are considerable differences in the degree of hypoxia among different tumor types; thus, it is impossible to propose a definite value to define hypoxia. The currently recognized "gold standard" is to measure the distribution of the partial pressure of oxygen (pO2) using polarographic electrodes, but this method is invasive and greatly limits its use in tumor determination. As the oxygen requirement of tissues is met by blood, a method of responding to hypoxia by measuring the vascular supply was proposed. Cryophotometry or magnetic resonance imaging (MRI) to measure blood oxygen saturation, or MRI, computed tomography (CT), or positron emission tomography (PET) to measure tumor perfusion have been widely used ( Table 1). However, vascular supply is not the only factor that causes hypoxia in cells and tissues; thus, monitoring of the vascular supply alone cannot completely determine hypoxia (50). Later, it was discovered that the expression of certain genes and proteins changes during hypoxia. Therefore, it was proposed that the state of hypoxia could be determined by measuring the levels of the endogenous markers. These principal markers include HIF-1, CA-IX, glucose transporter (GLUT)-1 and -3, and osteopontin (OPN). However, some reactions that are not related to hypoxia can also lead to upregulation of the endogenous markers. Some studies have reported that reactive oxygen species, cancerspecific mutations, and changes in signal transduction pathways can cause these changes even in normoxic conditions (51).
However, the most widely used method for detecting hypoxia is based on PET/SPECT imaging. For example, foreign substances (imidazole drugs, cytarabine, or other substances) are labeled with radionuclides or fluorescent probes and used as tracers with specific chemical properties to target hypoxic areas within the tumor.
To summarize, pre-monitoring of hypoxic information before tumor diagnosis or treatment is extremely important in selecting accurate and effective individualized treatments and monitoring the efficacy and prognosis. As PET is more commonly used in a clinical setting for tumor diagnosis, in this review, we have focused and conducted a thorough evaluation on several methods that use markers in PET imaging to detect hypoxia. This study sheds light on the timely intensive treatment methodologies to overcome the resistance caused by hypoxia and prevent local recurrence.

MAIN METHOD
Since 1981, misonidazole (Miso) was introduced as the first marker for the identification of tumor hypoxia based on molecular imaging. Miso was labeled with 14 C for autoradiography to detect tumor hypoxia (52). Subsequently, two important PET tracer categories, 18F-labelled nitroimidazoles and Cu-labelled diacetyl-bis(N4-methylthiosemicarbazone) analogs, have been developed to specifically study the areas of tumor hypoxia (53). 18 F Janet Rasey and her colleagues at the University of Washington first proposed the noninvasive imaging of hypoxia by PET (54). They synthesized 18F-fluoromisonidazole and then verified its feasibility as a PET hypoxia tracer in several models of human tumors. Since then, researchers have extensively explored the possibility of using 18 F-labeled compounds as hypoxia tracers ( Figure 3).   (55). The reduction of nitroimidazole in cells is achieved by nitroreductases. The first one-electron reduction to a nitro group is reversible and the reduced substance is easily reoxidized by molecular oxygen (56). Therefore, substances reduced under aerobic conditions can be quickly reoxidized and can diffuse outside the cell without being retained in normal cells. However, under hypoxic conditions, the free nitro radicals are further reduced in cells, undergo protonation, combine with intracellular macromolecules, and are irreversibly retained in hypoxic cells (Figure 4). These radicals are not encountered in apoptotic or necrotic cells (57)(58)(59).However, the accumulation of nitroimidazole compounds in tumor cells is not only affected by hypoxia, but also by many other aspects, such as the expression level of multidrug resistance-associated protein 1 (MRP1) (60,61). It is reported that tumor cells pretreated with MRP1 inhibitors have significantly higher radioactivity than cells not treated with inhibitors. At the same time, the SUV mean ratio in tumors of mice in the treatment group is significantly higher than that of control mice in vivo PET studies. Therefore, it was found that MRP1 inhibitors can increase the accumulation of 18 F-FMISO in hypoxic cells (62). In other words, during this type of chemotherapy, it may lead to an overestimation of tumor hypoxia. 18

F-FMISO
Over the years, several fluorinated nitroimidazole-based labels have been developed for PET imaging. Currently, the most commonly used PET hypoxia tracer used clinically and in research is 18 F-fluoroimidazole ( 18 F-FMISO) (63,64). It is also the first drug to be clinically tested (65) and has been successfully used in glioma (66), breast cancer (67), head and neck tumors (68,69), lung cancer (70), and other models to reflect hypoxia, and also used to stratify patients for radiotherapy. However, after intravenous injection, only medium contrast images are obtained owing to its high relative lipophilicity and slow penetration from the blood and passive diffusion, which results in a relatively low uptake (71). Compared with conventional imaging methods such as CT and MRI, 18   Its low lipophilicity leads to faster clearance of unbound drugs by the blood, which can improve hypoxia-normoxia contrast at an early point in time. In preclinical animal studies, the head-tohead comparison of 18 F-FAZA, 124 I-IAZA, and 18 F-MISO showed that 18 F-FAZA had faster vascular clearance than that of 18F-FMISO after 3 hours of injection. Recently, clinical studies have successfully evaluated the feasibility of 18 F-FAZA in hypoxic imaging of brain glioma (73) and lymphoma (73); lung (74,75), head and neck (76,77), cervical spine (78), and cervical tumors (79); rhabdomyosarcoma (80); and rectal tumors (81). Generally speaking, 18-FAZA has better application values compared with 18 F-FMISO. 18 F-FETNIM 18 F-Fluoroerythronitromidazole (FETNIM) also has better hydrophilicity than 18 F-FMISO, allowing for rapid renal clearance and low liver absorption, which can also explain the negative positive relationship between tumor blood flow and initial tumor 18 F-FETNIM absorption (82). Clinical studies in esophageal cancer (83), head and neck tumors (84)(85)(86), and cervical cancer (87) have shown that the tumor-to-blood ratio calculated by imaging at 2 h pi is 1.4-2.48 within the range, and that high tissue uptake of FETNIM indicates reduced progression-free and overall survival. Although 18 F-FETNIM can also be used in imaging studies to determine hypoxia in lung cancer (88,89), some studies indicate that the uptake of 18 F-FMISO in the tumor/non-tumor ratio is significantly higher than that of 18 F-FETNIM. Therefore, whether 18 F-FETNIM has a better application value than 18 F-FMISO is still debatable (90).

F-RP-170
Another 18 F-labeled 2-nitroimidazole compound, 18 F-RP-170, has been used to study patients with glioma, and the average tpO2 in the high uptake area has been reported to be significantly lower than that in the low uptake area. In the high uptake area, a significant negative correlation between the standardized uptake value (SUV) and tpO2 has been noted, whereas the HIF-1a index in the high uptake area is significantly higher than that in the low uptake area. These findings are suggestive of the low oxygen selectivity of 18 F-FRP-170 (91,92). Studies on brain tumors (93) and lung cancer (94) show that the SUV in hypoxic tissue is higher than that in normal tissue. Compared with 18 F-FMISO, the shorter time interval before scanning and improved hypoxic contrast may make it more suitable for research and use in clinical imaging.

F-EF5
18 F-EF5 was studied as a PET tracer in 2001. Compared with many other hypoxia tracers, 18 F-EF5 has a higher octanol-water partition coefficient, which enables higher cell membrane permeability and longer plasma half-life (95), which can improve the uniformity of tumor uptake and tracer distribution. In the study of patients with head and neck squamous cell carcinoma (HNSCC), the median tumor-tomuscle 18 F-EF5 uptake ratio (T/M) was found to increase over time and reported as 1.38 (range, 1.1-3.2) after 3 h of tracer injection (96). The ability of 18 F-EF5 to detect hypoxia is encouraging. It was also found in preclinical studies that 18 F-EF5 PET could predict the response to graded radiotherapy in tumor models (97); however, its marking chemistry is more complicated by comparison. 18

F-HX4
A third-generation nitroimidazole tracer ( 18 F-HX4) (98) has been developed in recent years. A 1,2,3-triazole moiety is introduced using simple click chemistry to make the compound more hydrophilic, which, at the same time, also increases its renal clearance (99). These characteristics of 18 F-HX4 help in reducing the background signal faster, thereby improving the signal-to-noise ratio (98). Preliminary studies have shown that compared with 18 F-FMISO and 18 F-FAZA, 18 F-HX4 has a higher maximum tumor-to-blood ratio with a half-life of about 3h (100). In addition, a piece of evidence that can prove that 18 F-HX4 will accumulate in hypoxic areas is the observation of a strong and significant spatial relationship between the distribution of 18 F-HX4 and pyridimidazole positivity and CA-IX positivity (98,99,101). In non-small cell lung cancer (NSCLC) (102), HNSCC (103,104), and esophageal and pancreatic cancers (105), a clear correlation with 18 FDG is observed, which is important however, better repeatability (106). Therefore, 18 F-HX4 is a progressive next-generation tracer that can be used as a tool for monitoring treatment responses and radiotherapy planning.
Unfortunately, the exact mechanism of Cu-ATSM hypoxia selectivity is unclear and controversial and has conflicting findings (114,115). In conclusion, based on initial research, it is believed that Cu(II)-ATSM will diffuse rapidly into cells owing to its high membrane permeability and low redox potential, and will be metabolized by NADH/NADPH in the mitochondria that are dysfunctional due to hypoxia. Under normoxic conditions, Cu(I)-ATSM is reversibly oxidized to Cu(II)-ATSM and diffuses out of the cell. In contrast, under hypoxic conditions, the Cu(I)-ATSM complex is irreversibly retained in the cell. Since Cu(I)-ATSM is far less stable than Cu(II)-ATSM, under the mediation of pH and other conditions, copper dissociates from the complex to form ATSMH 2 , causing radioactive copper to remain in the cell (116,117) (Figure 6).
Compared with 18 F-FDG and fluorine-based hypoxic PET tracer, radiolabeled Cu-ATSM can specifically accumulate in the hypoxic areas of the tumor and form a certain difference. Moreover, Cu-ATSM exhibits better pharmacokinetics and signal-to-noise ratio, is not taken up by the bladder, and does not show any interference (118-122) ( Table 2).
Although its retention mechanism in hypoxic cells is still controversial, Cu-ATSM has been used as a PET tracer to study tumor hypoxia for head-and-neck (153,154) lung (155,156), cervical (157,158), and rectal tumors (159), and gliomas (160). Studies have shown that Cu-ATSM has other benefits when used for hypoxia imaging, such as in the staging and detection of recurrent prostate cancer (120,121); as an effective indicator to predict neoadjuvant chemoradiotherapy and survival rate of patients with rectal cancer (159); as an indicator of treatment response in head and neck cancer (161), and in brain tumors (162). In addition, due to its special radioactivity, many reports have highlighted its use as an effective radiotherapy agent. Since treatment is not within the scope of this review, it has not been discussed here (163,164).
In summary, the above-mentioned studies show that the use of radiolabeled Cu-ATSM complex in PET examination is feasible and that other benefits to the treatment of patients are also possible ( Table 3). Although the mechanism of action is unclear, it is still worth pursuing in-depth research and development on this compound. Ga (t 1/2 = 68 min , 89% b+, Eb+max= 1.92 MeV, 11% EC) is a generator-produced radionuclide that offers excellent coordination chemistry with several bifunctional chelating agents and provides rapid radiolabeling over a range of pH (191)(192)(193). Therefore, 68 Ga has been preferentially studied in hypoxic PET. For example, Sudhakara et al. developed 68 Galabeled agents based on mono-, bis-, and tris-nitroimidazole conjugates with the chelating agent 1,4,7-triazacyclononane-1,4,7-tris[methyl(2-carboxyethyl)phosphinic acid] (TRAP), to obtain a series of 68 Ga-labeled compounds. The radiochemical yields were extremely high and the uptake of tumor models in in vivo experiments revealed its remarkable ability to target hypoxic areas (194). Recently, Yoichi Shimizu et al. also confirmed that the 68 Ga marker could help in the visualization of tumor tissues and hypoxic areas within 2 h of intravenous injection (195). In short, studies on lung (196) and colon cancers (197,198) show that on the basis of the ease in production, 68 Ga-labeled hypoxia tracers are comparable to traditional 18 F-labeled tracers.

124/125/131
I Radioactive iodine is a commonly used radiotherapy agent for the diagnosis and treatment of thyroid diseases. The characteristics of its radiation make it well-sought after for radioimmunotherapy (199). Radioactive iodine isotopes play a role in the field of hypoxia imaging as they can be used to label various substrates that can target hypoxic regions. For example,   123 I-IAZA is the most common radiotracer with a chemical structure similar to that of 18 F-FAZA. 124 I is a popular choice owing to its longer half-life. The distribution of 131 I-IAZGP (200) is similar to that of pimonidazole, and has a similar CA-IX expression profile in colorectal cancer models. 124 I-IAZGP (201) can be used to image liver tumors within 6 h (202); however, there are insufficient tumor-absorption studies that support the clinical role of 124 I-IAZGP PET in patients with colorectal cancer and head and neck cancer. 124 I-FIAU (203) can be used to evaluate the expression of HSV1-tkeGFP fusion gene that is associated with hypoxia. The area where 125 I-IPOS accumulates is positively correlated with the HIF-1a-positive area (204). 125 I-M75 specifically accumulates in colorectal cancer xenografts within 48 h of administration and delineates hypoxia by targeting CA-IX (205). However, several radioactive iodine isotopes have long half-lives. Although these isotopes play a unique role in radioimmunotherapy, their use as imaging agents should be carefully assessed.

99m Tc
The 99m Tc-labeled complex is one of the earliest nonnitroimidazole contrast agents used to determine hypoxia in cells. 99m Tc-labeled 2-nitroimidazole was also developed, with BMS181321 being the first. However, these agents are unstable and lipophilic, which limits their application (206). Subsequently, other tracers were developed to overcome these shortcomings (207). By choosing different cores and ligands, the tracer was designed to exhibit different characteristics to optimize the selection. The [ 99m Tc(CO) 3 (H 2 O)] + nucleus is well studied for the labeling of nitroimidazole compounds (208). Recently, more focus has been placed on the introduction of tumor-targeting moieties (e.g., RGD). The 99m Tc nitroimidazole complex is retained longer in hypoxic tumor cells.

Others
In addition to the radionuclides discussed above, several others have been used for hypoxic PET. For example, 111 In is a promising candidate for imaging hypoxic areas to study head and neck tumors (209), prostate cancer (210), and colorectal cancer (211). The nitroimidazole and gastrin-releasing peptide receptor (BB2R) conjugate was labeled with 177 Lu (212); the CA-IX antibody G250-F(ab′)2 was labeled with 89 Zr (213) and some promising results were obtained. Several nuclides are used in radiotherapy although their imaging effects are unsatisfactory or there are many factors worthy of improvement; however, these studies could help develop new ideas and research methodologies.

DISCUSSION AND CONCLUSIONS
Assessing the relationship between tumor tissue and hypoxia has been a fascinating topic in cancer research for many years mainly due to its complex correlation with disease progression and treatment response. In clinical research, in addition to the use of oxygen electrodes for research, the use of exogenous probes such as FMISO/FAZA-PET, HX4, Cu-ATSM, and pimonidazole, as well as methods that target endogenous markers, such as CA-IX and HIF-1, are the mainstay to determine hypoxia (214). This review summarizes the progress of radionuclide-labeled PET tracers in hypoxia imaging. Each agent has its advantages and disadvantages; thus, it can only be used in a relatively limited range. Although some tracers have demonstrated the feasibility of their use in PET imaging hypoxia in several tumor entities in a clinical setting, so far, a perfect tracer with all desirable characteristics of an ideal PET hypoxia tracer has not been obtained.
One of the most important challenges in a clinical setting is the reproducibility of hypoxia imaging. The definition of the hypoxic area during measurement, selection of hypoxianormoxia threshold, time variability of PO 2 levels in tumors during continuous measurement, heterogeneity within and between tumors of the same patient, and factors such as receiving treatment affect the reproducibility of scanning (215).
Blood perfusion is one of the obstacles that limits the use of traditional hypoxia tracers in the tumor microenvironment. Insufficient perfusion limits the effective delivery of the tracer to tissues, affects tracer accumulation in normal or tumor tissues, and leads to poor imaging effects or contradictory findings. Therefore, the special relationship between perfusion and hypoxia requires further clinical research for the evaluation of hypoxia-perfusion patterns in different tumor types as well as to elucidate their relationship with clinical outcomes. With the recent advent of PET-MRI scanners, multiple imaging models and multifunctional tools that provide perfusion information and increase the accuracy of hypoxia measurement and provide complementary information with higher predictive values will soon become possible.
Many problems need to be solved urgently to obtain the ideal and perfect tracer. Defining a relatively standardized hypoxia threshold or observation index, comparing various tracers to study different tumor types, and screening a more optimized imaging plan may be suitable approaches to achieve this goal.
In recent years, the research and application of nanotechnology in tumors has resulted in a very valuable tool, which aims to provide various possible solutions to address the shortcomings of current imaging agents. Nanocarriers have better permeability and retention in solid tumor tissues; thus, they can promote molecular delivery to overcome the poor vascular system in tumor tissues. Nanocarriers can also be used to carry materials with different targeting properties, or can be used in imaging systems with multiple modes to improve their biocompatibility (by reducing cytotoxicity) and add other functions to improve the accuracy and utility of the tracer (216).
Owing to the increasing clinical demand for the assessment of hypoxia, imaging for tumor hypoxia is of great significance for tumor diagnosis, prognosis assessment, and planning treatment strategies. Due to hypoxia-induced malignant progression of tumor cells and their resistance to radiotherapy and chemotherapy, several ongoing clinical trials are focusing on adjusting the radiotherapy regimen according to the degree of oxygenation of the tumor and assessing the benefits.
Simultaneously, radiosensitizers and hypoxia activation studies in medicine are also underway. Thus, the goal of researchers who are studying hypoxia is to change clinical outcomes rather than just provide prognostic information.