While some molecular imaging techniques can visualize the intrinsic differences, such as Raman scattering and autofluorescence, between normal and diseased tissues [9, 10], an exogenous molecular imaging agent is generally required [6] and administrated systematically or topically. In general, molecular imaging agents should have high selectivity, minimal background from nonspecific binding, suitable pharmacokinetics, excellent in vivo stability, and good safety profile [6]. While for topical application, the imaging agents are delivered to target organ directly thus smaller amount of imaging agents are required, which together reduce potential toxicity and blood clearance issues from intravenous injection [10]. This is of special importance for endoscopy-based molecular imaging of urothelial malignancies where imaging agents are introduced intravesically. Translational and clinical examples of molecular imaging though topical administration can be found for imaging of gastrointestinal cancer, another epithelial malignancy [11].

An exogenous molecular imaging agent is typically comprised of a targeting molecule and a signaling conjugate, which should not interfere with the function of the targeting molecule. Targeting molecules can be categorized based on size, ranging from small molecules, peptides, engineered proteins (e.g., affibody), aptamers, and monoclonal antibodies (mAb)—each of which possesses different pharmacokinetic and binding properties (Fig. 20.1) [6].

Small molecules (usually <500 Da) have the smallest size among all molecular imaging agent, which allows them to access and image intracellular targets. Unlike other molecular imaging agents that enable imaging by direct binding to specific target, small molecules can also visualize alternations in protein or cellular functions including metabolism, hypoxia, enzyme activity, and protein synthesis [6]. One clinical example is fluorodeoxyglucose (FDG) that targets increased glucose metabolism found commonly in cancer cells. ¹⁸F-FDG has been the most commonly used PET molecular probe for penile, bladder, prostate, and kidney cancer [12]. On the other hand, due to the small size, there are very limited numbers of signaling conjugates, and small molecules can be attached without altering their pharmacokinetic and targeting properties [6].

Peptides (~15 amino acids) are intermediate in size between small molecules and mAb and engineered proteins. Thus, peptides provide increased flexibility in signaling conjugates and generally offer superior selectivity and specificity than small molecules. While peptides generally have lower affinity compared to mAbs, the relatively smaller size makes them less immunogenic and allows penetration to deep tissue where mAbs are not accessible. The biggest limitation of peptide as targeting molecule is the in vivo stability, and strategies such as adding prosthetic group is mostly required for radiolabeling of peptides. An example of peptide-based molecular imaging agent for SPECT is ¹¹¹In-DTPA-octreotide, a somatostatin analog originally developed for scintigraphy of neuroendocrine tumors. More peptide imaging agents are under clinical trial or preclinical investigations [13]. One preclinical example of peptide-based molecular imaging is bombesin (BBN) analogs targeting the gastrin-releasing peptide receptor (GRPR) as molecular probes for PET imaging of prostate cancer. In vivo imaging results on animal models are promising [14].

Monoclonal antibodies can recognize tumor-associated antigens including adhesion molecules and cell surface proteins with ultrahigh affinity and specificity [15]. However, their utilization as molecular imaging agents are limited by high immunogenicity from murine origin, restricted tissue penetration, and slow clearance from blood associated with big size (150 kDa) [6, 16]. The immunogenicity issue can be overcome by production of chimeric and humanized mAb [15]. mAb derivatives with smaller size have also been developed to improve the pharmacokinetics of mAb while maintaining the specificity and affinity [17]. While single-chain variable fragment (scFv, ~25 kDa), which contains one antigen-binding site of mAb, clears too quickly to generate adequate accumulation on target, mAb derivatives with intermediate size such as diabodies (scFc dimer, ~55 kDa) and minibodies (scFv fused to single Fc domain, ~80 kDa) have shown promising imaging results on animal models [17]. Currently, there are at least 8 mAbs approved for SPECT molecular imaging [16]. Among them, ¹¹¹In-capromabpendetide (ProstaScint) in particular is approved for molecular imaging of prostate cancer [18].

Affibodies are emerging molecular imaging agents with significant potential for clinical translation [6]. Affibodies are small non-immunoglobulin affinity ligands capable of binding to a wide range of protein targets. They are selected from combinatorial libraries based on a 58-amino acid, three-alpha-helical Z-domain scaffold [17]. Affibodies have fast clearance from the blood with adequate tumor uptake [16]. There are growing interests in using affibodies as alternative to mAb and mAb derivatives. One example is affibody molecules that target epidermal growth factor receptor 2 as molecular probe for SPECT. A clinical study on breast cancer patients with ¹¹¹In- or ⁶⁸Ga-labeled ABY-002 has shown great potential to localize metastatic lesions [19]. Variants of ABY-002 also have been investigated on prostate cancer xenografts [20].

Aptamers are single-stranded DNA or RNA oligonucleotides that are activated upon binding to their targets. While aptamers offer high affinity and specificity comparable to mAbs, their application as molecular imaging agent is limited by low in vivo stability and short half-life associated with their small size [6]. Molecular imaging with aptamers in general is still in its infancy and needs further investigation [6]. Recently, a novel nucleolin-targeted DNA aptamer (AS1411) has been evaluated as therapeutic drug for metastatic renal cell carcinoma in a phase II trial study [21]. Low toxicity was observed in patients, indicating AS1411 may potentially be translated for molecular imaging purpose.

Targeting molecules can be conjugated to radioisotopes, fluorescent dyes, or nanoparticles to enable visualization. Molecular imaging with targeted nanoparticles is emerging as an exciting diagnostic tool. Nanoparticles can vary largely in size, shape, and material they can be composed of, with unique surface properties and reactivates [6, 22]. The large variety offers nanoparticles great flexibility in terms of imaging techniques they are compatible with, thus enabling multimodality molecular imaging [6, 22]. Nanoparticles are designed to be intrinsically near-infrared fluorescent [e.g., single-wall carbon nanotubes (SWNTs), quantum dots (Qdots)], or have magnetic properties [e.g., superparamagnetic iron oxide nanoparticles (SPIOs)] [23]. By surface modification [24] or additional radiolabeling [25], dual-labeled nanoparticles targeting the same ligands can therefore be recognized simultaneously by two or even three imaging techniques. In general, different imaging techniques are complementary rather than competitive [23], and the multimodality capacity of nanoparticles will facilitate the cross-talk between the whole-body scan imaging techniques (e.g., PET, CT) and optical imaging techniques (e.g., photoacoustic imaging, Raman spectroscopy), opening up the opportunity of tumor staging and characterization with single-molecular imaging agent.

The flexibility of nanoparticles also enables the capacity for multiplexed imaging. Many nanoparticles [e.g., quantum dots (Qdots) and surface-enhanced Raman scattering (SERS) gold nanoparticles] are tunable for different emission wavelength or Raman spectrum by alternation of size or structure. Same type of nanoparticles with different characteristics can be functionalized for simultaneous imaging of different tumor targets. In vivo studies on small animals have demonstrated simultaneous detection and separation of ten different types of SERS particles [26]. The Raman spectroscopy feature for detecting SERS nanoparticles has been incorporated into endoscopy [27], and molecular imaging of bladder cancer with targeted SERS particles is currently under active investigation.

The large surface area-to-volume ratio of many nanoparticles allows substantial payloads of targeting molecules, contributing to their ultrahigh sensitivity [6, 28], thereby minimizing the quantity of the imaging agent required and enabling molecular imaging with insensitive conventional imaging techniques such as CT [29] and MRI [30]. Nanocrystals such as Qdots offer additional benefits including 20-fold greater brightness and 100-fold greater stability to photo-bleaching than organic fluorophores [31], which is ideal for optical imaging.

While promising, in vivo application of nanoparticles can be challenging due to general pharmacokinetic issues associated with their large size (10 ~ 200 nm). While toxicity can also be an issue for nanoparticles such as Qdots with a core made of cadmium, Qdot-based molecular imaging may still have a role in evaluation of urogenital malignancies through topical (e.g., intravesical) administration, which could potentially minimize the toxicity and other limitations from systematic use. On the other hand, gold nanoparticles (AuNP) could be a safer alternative and have been approved for therapeutic use in humans [32]. Efficacy of therapeutic gold nanoparticles has also been demonstrated for treatment of prostate cancer on mouse xenograft [33], accelerating the clinical translation of molecular imaging of urological malignancies with targeted nanoparticles.

Current investigations of molecular imaging platforms incorporating conventional CT and MRI have emerged in recent years, which promise to add molecular information such as with PET and SPECT. PET and SPECT are radionuclide imaging techniques that trace whole-body distribution of intravenously administrated radiolabeled imaging agents [6], delivering excellent sensitivity along with unlimited depth of penetration and quantitative capability. As biochemical changes often precede anatomical ones in disease, PET and SPECT are able to detect such changes before CT and MRI. A limitation of PET and SPECT, however, is the lack of an anatomical reference frame (addressed by combining with CT/MRI). On the other hand, US is efficient in detecting vascular structures with high resolution and sensitivity [34], with transrectal US being the current standard for real-time guidance of prostate biopsy [35]. While it is relatively cost-effective and has a good safety profile, it is limited in its ability to image bone or air-containing structures and by its limited depth of penetration [6].

Optical imaging utilizes visible, ultraviolet, and infrared light to interrogate tissues of interest, with its advantages including superior resolution and relatively facile integration into the operating room setting. By integrating features such as fluorescence imaging into standard white-light endoscopic and laparoscopic procedures, the visual contrast between normal/benign and tumor tissues is enhanced, enabling real-time, intraoperative visualization and characterization of tumor tissues [84]. The advent of fluorescence-guided surgery has been a novel paradigm for cancer diagnostics and treatment [9].

Several promising optical imaging modalities have entered the urologic clinical arena, ranging on a macroscopic [i.e., photodynamic diagnosis (PDD), near-infrared fluorescence (NIRF)] to microscopic [i.e., optical coherence tomography (OCT), confocal laser endomicroscopy (CLE)] scale. PDD, NIRF, and CLE all require an exogenous fluorescent imaging agent. The potential in-human application of fluorescent molecular imaging has been presented in ovarian cancer [85] and colon cancer [86], adding promise for future application to urologic malignancies.

Several emerging molecular imaging modalities [i.e., photoacoustic imaging (PAI), Raman spectroscopy (RS), and multiphoton microscopy (MPM)] hold great promise in preclinical studies. PAI detects ultrasound transmission resulting from optical absorption by tissues. Imaging agents such as SWNT and AuNP emit strong photoacoustic signals via light absorption and can be used to enhance tissue contrast or functionalized with signaling molecules to enable molecular imaging with PAI [110, 111]. The advantages of PAI include superior depth of penetration (up to 5 cm) compared to optical imaging, significantly lower doses of imaging agent required (picograms to micrograms), and minimal background noise compared to traditional US [6]. The urologic application of PAI has been explored, with preliminary results of an ex vivo study on human prostate tissues suggesting that PAI can differentiate between normal, benign, and malignant tissues by detecting the intrinsic differences in photoacoustic signals [112]. In vivo visualization of prostate tumors has been shown in mouse models [113, 114], with a transrectal probe combining US and PAI under development to facilitate image-guided prostate biopsy. In a pilot study on ex vivo porcine bladders, PAI successfully visualized simulated tumors of injected biomaterials containing varying degrees of pigment, suggesting PAI can potentially visualize tumor depth in the bladder and provide complimentary staging information to diagnostic cystoscopies [115]. Photoacoustic endoscopy has been developed recently [116, 117], and PAI platforms contrasted with non-targeted non-ionizing imaging agents have been investigated for in vivo visualization of the bladder in small animals [27, 118, 119] to facilitate clinical translation. While PAI molecular imaging agents for urological malignancies are still under investigation, the feasibility of molecular imaging with PAI has been demonstrated in tumor-bearing mice for identifying breast cancer lesions using SWNT functionalized with Arg-Gly-Asp (RGD) peptide targeting tumor overexpressed integrin [120].

RS detects alternation in vibrational state of molecules in biological tissues under near-infrared light illumination (785–845 nm), which is known as a Raman shift [121]. Detection of multiple Raman peaks from the target tissue is plotted to create a spectrum of peaks, achieving a molecular “fingerprint” of the tissue examined without the need for exogenous contrast agent [121, 122]. RS’s drawbacks include time to obtain a spectrum (1–5 s), weak signals, and limited field of view. SERS gold nanoparticles have been demonstrated to augment relatively weak signals from Raman scattering to improve sensitivity. These nanoparticles, when conjugated with tumor-specific ligands, have been used for in vivo tumor detection in small animals [123]. Targeted SERS particles also have the potential to enable multiplexed targeted imaging during endoscopy [124]. The feasibility of in vivo bladder cancer diagnosis was demonstrated using a fiberoptic Raman endoscopic probe that could differentiate cancerous tissue from normal urothelium in near real time during cystoscopy [122]. Bladder cancer-specific SERS particles are currently under investigation. On the other hand, RS has also been considered for implications in prostate cancer with pathologic diagnosis, margin status, and diagnostic biomarkers [125].

MPM uses simultaneous absorption of 2–3 near-infrared photon to cause nonlinear excitation equivalent to that created by a single photon of blue light and allows for both in vivo imaging and ex vivo imaging of fresh, unprocessed, and unstained tissue via distinct intrinsic tissue emissions (ITEs). Imaging capabilities include submicron resolution and up to 0.5 mm depth of penetration. Urologic investigations using MPM have been done with periprostatic nerves in a rat model [126], visualization and quantification of renal uptake [127, 128], differentiating normal and pathologic human bladder biopsies [129], and testicular biopsies distinguishing normal from abnormal spermatogenesis in patients being worked up for nonobstructive azoospermia [130].