Molecular imaging—its current role in cancer

Abstract

Medical imaging has undergone extensive growth over the last few decades and now plays a central role in clinical oncology. The future of imaging in the management of oncology patients is molecularly targeted imaging agents. Molecular imaging differs from conventional anatomical imaging in that imaging probes are utilized to visualize target molecules-of-interest. It is envisioned that molecular imaging will have a major impact on oncology and personalized medicine by allowing earlier diagnosis, assessing early response to treatment and by predicting treatment response. It will, hopefully, also have an impact on drug development by streamlining preclinical and clinical tests for new drug candidates.

Molecular imaging is the in vivo characterization and measurement of biological processes at the cellular and molecular level. It differs from traditional medical imaging where morphologically visible results of disease processes are visualized. Molecular imaging allows not only localization of a tumor in the body but also allows imaging of the expression and activity of specific molecules (such as protein kinase) as well as biological processes (such as angiogenesis, hypoxia, apoptosis), which influence tumor behavior and response to treatment. This exciting new field has the potential to transform the future of oncologic medicine.

Imaging of angiogenesis

Angiogenesis is a process whereby new blood vessels are formed to supply oxygen and nutrients to growing tumors, and is an important requirement for tumor growth, invasion and metastasis. Initial tumor growth is limited by the restricted access to oxygen, nutrients and growth factors. Therefore, tumors initially grow toward pre-existing nearby vessels giving rise to a perivascular tumor growth pattern. This pattern only supplies the periphery of the tumor and gradual tumor growth leads to central hypoxia, which up regulates the expression of different angiogenic factors.¹ The angiogenic process is driven and characterized by angiogenesis-related markers that are either only present in endothelial cells during angiogenesis but not in mature vessels, or are significantly upregulated. These markers include the family of vascular endothelial growth factors receptors (VEGFRs) and its ligands, ανβ₃ cell adhesion molecule integrins or matrix metalloproteinases (MMPs).²

The blood vessels in tumors have an abnormal structure, demonstrate chaotic organization and have a high number of dysfunctional vessels. There is uncontrolled development of new blood vessels secondary to the overproduction or proangiogenic factors, which results in increased microvessel density (MVD). These vessels have abnormal, poorly developed walls, which results in very leaky vessels. This increased permeability and the absence of functional lymphatic vessels within the tumors results in increased interstitial fluid pressure, which reduces the transvascular pressure gradient and fluid flow.³ The decreased blood flow (BF) as a result of increased interstitial fluid pressure and other factors such as hemoconcentration, leads to impairment of oxygen supply and tumor hypoxia. In an attempt to overcome the oxygen shortage the tumor cells express more proangiogenic factors, which then recruit even more abnormal vessels resulting in the switching on of invasive and metastatic programs in the tumor cells for escaping from the hostile hypoxic microenvironment.

Tumor vasculature can be differentiated from normal tissue based on structural features, the function of the blood vessels and the specific expression of angiogenesis-related markers. Pathophysiological changes within the tumor secondary to the structural abnormalities of the new vessels can be exploited on contrast-enhanced imaging. These changes include an increase in capillary permeability, volume of extravascular-extracellular space (EES), interstitial pressure and tumor perfusion. The effects of angiogenesis on tumor perfusion and permeability can be assessed. A high blood volume (BV) value is the expression of an increased microvascular network due to the formation of new vessels, whereas an increased BF is an expression of a large number of arteriovenous shunts, which, because of their low resistance, cause increased flow in the microvasculature. An increased BF causes a corresponding decrease in the mean transit time (MTT) of blood in the vascular network. The greater permeability of the newly formed vessels is responsible for a higher permeability value or permeability surface (PS).⁴

Contrast-enhanced techniques, including perfusion computed tomography (CT), dynamic contrast-enhanced magnetic resonance imaging (MRI) (DCE-MRI), dynamic susceptibility-enhanced MRI, dynamic microbubble-enhanced ultrasound (US) or positron emission tomography (PET) with oxygen-labeled water, have been used for the functional evaluation of tumor vasculature. DCE analysis may be based on a qualitative, semiquantitative or quantitative approach. Curveology is based on the qualitative analysis of the shape of the signal-intensity time curve, in which the speed of the filling phase, the peak intensity and the kinetics following the peak of the enhancement (decrease or washout, plateau or slow accumulation of contrast media) are evaluated. Semiquantitative analysis is based on the evaluation of parts of the curve, including time to peak enhancement, bolus arrival time, maximum upslope, initial area under the curve (IAUC) and maximum enhancement. These parameters have an incompletely defined, complex relationship with tumor physiology and represent a composite of physiological processes (Table 1). Finally, a quantitative analysis can be performed using mathematical models to obtain information on different biologically relevant physiological parameters related to perfusion, capillary permeability or both.

Table 1.

Common parameters derived from DCE imaging methods that evaluate the tumor microvasculature

Parameter	Physiological process
IAUC	Includes information of BF, BV, permeability, EES volume and MVD
Time to peak	Depends on tissue perfusion
Washout	Is the velocity of enhancement loss
BF	Is how much BFs per tissue mass per time, and reflects the delivery of oxygen and nutrients to the tumor
BV	Is the volume of capillary blood contained in a certain volume of tumor and correlates with the amount of functional blood vessels within a tumor
MTT	Is the mean time taken by blood to pass through the capillary network from the arterial inflow to the venous outflow
Transfer constant Ktrans	Volume transfer constant of a contrast agent between the blood plasma and the EES and represents a complex combination of tissue BF and permeability
ve	Distribution volume of the contrast agent per unit volume of tissue and reflects the fractional volume of the intravascular-extracellular space
Return constant kep	Rate constant of the contrast agent between the EES and the blood plasma and depends mainly on permeabilitykep = Ktrans/ ve
PS area product	Is the product of permeability and total surface area of capillary endothelium in a unit volume or mass of tissue and reflects the total diffusional flux across the capillaries and depends on the contrast agent, vessel surface area, vessel leakiness and the volume of the interstitial distribution space

Several markers related to angiogenesis including VEGF/VEGFR, ανβ₃ integrins, hypoxia-inducible factor-1 or MMPs can be targeted for single-photon emission computed tomography (SPECT)/PET imaging. There are differences between SPECT and PET imaging. SPECT uses radioisotopes that decay via electron capture, gamma emission or both, whereas PET radioisotopes decay via emission of positrons. PET also offers greater spatial information and permits more accurate attenuation correction, compared with SPECT.

Integrin ανβ_3, which is significantly upregulated on tumor vasculature but not on quiescent endothelium, binds to arginine-glycine-aspartic acid (RGD)-containing components of the extracellular matrix. Probes for SPECT or PET imaging of integrins’ expression have been developed by conjugating the RGD-containing peptides with glucose- or galactose-based sugar amino acids.² However, the usefulness of imaging integrins as a surrogate marker for angiogenesis depends on the level of ανβ₃ expression on the vasculature, or tumor, or both. Some tumors, such as melanoma, demonstrate intense ανβ₃ expression on tumor cells and therefore tracer uptake may not be an optimal surrogate markers for angiogenesis in these tumors.

Increased expression of VEGF and VEGFRs by tumor cells and by the tumor-associated vasculature, respectively, correlates with tumor growth rate, MVD, tumor metastatic potential and poorer prognosis in a variety of malignancies.² Therefore the ability to image the VEGF or VEGFR expression and occupancy of VEGFR in vivo could potentially provide important information, including assessing the efficacy of antiangiogenic cancer therapy. However, although several preclinical trials on probes currently used for imaging the VEGF and VEGFRs pathways, such as antibodies against VEGF and radiolabeled VEGF-A, have shown promising data, further validation and improvements need to be achieved before these enter clinical practice.⁵

A recent study which detailed a method for conjugating a therapeutic antibody to a molecular magnetic resonance imaging nanoparticle concluded that cet-PEG-dexSPION could be a promising theranostic nanomedicine for therapeutic targeting of EGFR-expressing tumor cells using the therapeutic antibody cetuximab and non-invasive monitoring of treatment efficacy.⁶

Imaging of hypoxia

Hypoxia is defined as diminished availability of oxygen required for cells to metabolize normally. It occurs when the tumor becomes large enough to disrupt the balance between the supply and consumption of oxygen. In solid tumors, hypoxia-induced proteome changes lead to cell cycle arrest, differentiation, apoptosis and necrosis. However at the same time, hypoxia-induced proteome and/or genome changes may promote tumor progression via mechanism enabling cells to overcome nutrient deprivation, to escape the hostile environment and to favor unrestricted growth. Tumor hypoxia has been demonstrated to be important in invasion, angiogenesis, apoptosis, metastasis, chemoresistance and radioresistance.⁷ Therefore being aware of the degree and extent of hypoxia prior to initiation of treatment is important in treatment planning and having this information early after the onset of treatment helps predict chemoresistance and radioresistance.

Blood oxygen level dependent (BOLD) MRI has the potential to diagnose tumor hypoxia. In oxygen-deficient states hemoglobin will present as deoxyhemoglobin, which is paramagnetic and results in increased transverse relaxation of the surrounding protons while oxyhemoglobin does not. Decreased oxygenation in blood results in decreased signal intensity on T2*-weighted images.⁷ The partial pressure of oxygen (pO₂) can be estimated by using the relationship between BOLD-MRI signal and vascular oxygenation.

Because of the development of nitroimidazoles for use as radiosensitizers, multiple nitroimidazole compounds have been developed for use as PET hypoxia imaging markers, of which ¹⁸F-Fluoromisonidazole (¹⁸F-FMISO) is the most widely used. FMISO is only sensitive to the presence of hypoxia in viable cells. Studies have demonstrated that significant ¹⁸F-FMISO uptake requires a pO₂ level of <10 mmHg.⁸ The limitations thus associated with the use of ¹⁸F-FMISO include the long wait time between injection and scanning to allow the tracer to be cleared from the plasma and the normoxic tissues, the modest signal-to-noise ratio and the reproducibility of intratumor distribution with successive scans. Second-generation nitroimidazoles, which are more water soluble and not significantly degraded by most of the oxidizing mechanics in the body, are being investigated.

Another PET agent being investigated to delineate areas of hypoxia is Cu-Diacetyl-bis (N4-Methylthiosemicarbarcone) (Cu-ATSM). Cu-ATSM undergoes reduction in living cells and is rapidly cleared from the aerobic cells but becomes trapped in hypoxic cells. Therefore the tracer could allow the differentiation between dead, hypoxic, non-functional and viable tissues.⁹

Imaging of apoptosis

Apoptosis, or programmed cell death, can be triggered by an array of factors, such as a lack of needed growth factors, anti-hormonal therapy, immune reactions, ischemic injury, ionizing radiation and chemotherapy. Early detection of tumor response to therapy is critical to shorten the period of uncertainty following initiation of therapy, and to rapidly identify the most effective treatment for individual cancer patients.

Apoptosis is characterized by the activation of a series of protesases, commencing with the release of cytochrome-c, which leads to self-assembly of the apoptotic protease-activating factor-1 (apaf-1) and capase-9 into the apoptosome, which goes on to activate caspase-3/7 and results in cell death by increasing DNA cleavage. Concurrently, blebbing of the cell wall in the dying cells exposes phosphatidylserine or phosphatidylethanolamine to the cell surface.¹⁰

One of the most extensively investigated probes is annexin-V, a ubiquitous intracellular human protein with a high affinity for membrane-bound phosphatidylserine. This target normally resides in the inner leaflet of the plasma membrane, but is flips to the outer leaflet after scramblase activation by apoptotic cells and is consequently exposed to binding by annexin –V.¹¹ Trials of patients receiving chemotherapy have demonstrated increased probe accumulation in patients that later achieved remission.^12,13

Recent proof-of-principle studies have found that an ¹⁸F-labeled capase-3 sensitive nano-aggregation PET tracer (¹⁸F-S-SNAT) is a promising new apoptosis-specific PET tracer for imaging tumor response to chemotherapy.^14,15 Another recent study discussed the preparation of the novel tracer ¹⁸F-ML-8 and examined its potential application in imaging of cyclophosphamide-induced apoptosis in tumor bearing mice.¹⁶

Tumor receptor imaging

There are many advantages associated with the ability to measure receptor expression by imaging rather than by histological inspection. These include its’ non-invasiveness and the ability to assess sites, which are difficult to sample. In addition it allows the assessment of entire disease burden and avoids sampling error from biopsies when receptor expression is heterogeneous, as well as the potential for serial monitoring over time or after drug treatment. Tumor receptor imaging can measure the therapeutic target expression and the presence or absence of the target, along with the level of target expression could be used to direct patient selection for targeted therapy.

Trials in patients with breast cancer found that the average ¹⁸F-fluoroestradiol (FES) standardized uptake value (SUV) was higher in patients who responded to endocrine therapy (targeted to estrogen receptors) versus non-responders and that if the FES SUV across disease sites was <1.5, patients were unlikely to respond to endocrine treatment.^17,18 A recent study in breast cancer patients found that treatment plans were changed in nearly 50% of patients based on the 18F-FES PET/CT results.¹⁹

Another molecular target which has been studies in breast cancer is HER2, which is overexpressed in 25–30% of breast cancers and is involved in angiogenesis and metastasis.²⁰ Several HER2 inhibitors have been developed and imaging HER2 expression and changes in expression at tumor sites may assist oncologists to determine if and which kind of HER2 inhibitor should be used during different phases of treatment. To date, HER2 monoclonal antibodies, such as trastuzumab, have been labeled with multiple different tracers to determine HER2 expression. One such technique, ⁶⁴Cu trastuzumab PET, could visualize primary tumor lesions >2 cm and metastatic brain lesions >1 cm.^21,22 Another HER2 imaging technique, ⁸⁰Zr-trastuzumab PET imaging, has be demonstrated to visualize HER2 positive tumors in humans, and due to its longer half-life it produces clearer images but induces higher radiation exposure.²³

Figure 1.

(A, B) Benign lymph node: A) Axial precontrast T2*-weighted image shows a hyperintense portocaval node (arrow). B) 48 h after administration of lymphotrophic nanoparticle, the node shows a dramatic reduction in signal intensity indicating benignity (arrow). (C, D) Malignant lymph node: C) Axial precontrast T2*-weighted image shows a hyperintense peripancreatic node (arrow). D) 48 h after the administration of a lymphotrophic nanoparticle, the node shows no signal change indicating malignant infiltration (arrow). (Reprinted with permission from McDermott et al.,³³ Copyright 2013).

Prostate specific membrane antigen (PSMA) is a rational target for imaging of prostate cancer due to its high expression in this disease and its association with prostate tumor aggressiveness. PSMA-targeted PET imaging, using a ⁶⁸Ga-labeled prostate membrane specific antigen (PMSA) ligand, has been shown to be able to detect primary prostate cancer, relapses and metastases with high contrast.^24–26 A different PSMA-targeted PET tracer, N-[N-[(S)-1,3-dicarboxypropyl]carbamoyl]-4-(18)F-fluorobenzyl-L-cysteine (DCFBC), was able to detect clinically significant high grade (Gleason 8 and 9) and larger sized primary prostate tumors reliably. This agent may play a role as a non-invasive imaging biomarker for differentiating indolent from aggressive disease and thereby improving risk-adaptive management.²⁷

Folate receptor (FR) is expressed in different epithelial cancers, such as ovarian, breast, endometrial and non-small cell lung cancers, with few non-malignant tissues demonstrating FR expression therefore making it an attractive therapeutic target. Recent phase 2 clinical trials have found that ^99mTc-etarfolatide imaging may be able to identify patients who are most likely to benefit from treatment with vintafolide.^{28,29 99m}Tc-etarfolatide imaging may potentially also be useful as a prognostic tool as FR expression has been found to be prognostic for ovarian and lung cancer.³⁰

Biomarker-targeted probes can also be linked with nano-particle based contrast agents and be imaged with MRI. As an advantage over other imaging modalities, such as PET, SPECT and CT, there is no ionizing radiation involved with MRI. A recent proof-of-concept study found that iron oxide magnetic nanoparticles conjugated with PMSA-targeting antibody, J591, enhanced MRI of prostate cancer in a preclinical model of orthotopic prostate cancer zenografts in mice.³¹

Lymphotrophic imaging

Accurate nodal staging is important in the management of any primary malignancy. Conventional cross-section imaging, such as CT or MRI, used for nodal evaluation is limited as they rely on size criteria for the detection of metastases. Lymphotrophic nanoparticles have a monocrystalline, inverse spinel, superparamagentic iron oxide core and contain a dense packing of dextrans to prolong their circulation time. After administration, these nanoparticles are slowly extravasated from the vasculature into the interstitial space, from where the particles drain into the lymph nodes via the lymphatic system. Within a normal lymph node, the nanoparticles are internalized by macrophages which can be detected as a drop in signal on T2- and T2*-weighted MR images due to the significant susceptibility these nanoparticles produce. In a node, which is completely or partially infiltrated by malignant cells, there is a lack of nanoparticle uptake because of the absence of functioning macrophages which results in these nodal areas retaining their signal intensity on both T2- and T2*-weighted images.³² Images are obtained 24 h after the administration of the nanoparticles, as time is required for the circulating nanoparticles to be internalized by the macrophages.

Conclusion

Although remarkable progress has been made in molecular imaging, further work is needed to bring this technique into routine clinical practice. Hopefully, the future of clinical molecular imaging will include (i) the ability to detect physiologic or molecular changes which indicate the presence of cancer when it is still at an early and curable stage, (ii) the ability to assess treatment response and adjust treatment protocols in real time and (iii) the ability to streamline cancer drug development process.