In vivo analysis techniques for antibody drug: Recent advances and methodological insights

Abstract

Antibody drugs, such as monoclonal antibodies and antibody-drug conjugates, have shown significant potential in treating diseases due to their high specificity and affinity. In vivo analysis of antibody drugs with non-invasive and real-time techniques is of importance to understand dynamic behavior of drugs within living organisms, and help evaluate their pharmacokinetics and efficacies. This review summarizes the advances and in vivo analysis methods of antibody drugs, including the techniques of radiolabeled imaging, near-infrared fluorescence imaging and surface-enhanced Raman spectroscopy. The principles, applications, and challenges of each technique are discussed, which provides insights for the development of antibody drugs and in vivo analytical methods.

Graphical abstract

Highlights

• •Progress of antibody drugs and the importance of in vivo analysis are introduced.
• •Non-invasive and real-time methods for in vivo analysis of antibody drugs are summarized.
• •The application, limitations, and possible solutions of in vivo analysis methods are discussed.

1. Introduction

Antibody drugs exploit the inherent specificity and target recognition properties to achieve precise therapeutic engagement, characterized by high specificity, low immunogenicity, minimal adverse reactions, and prolonged plasma residence times. Advances in genetic engineering and chemical conjugation technologies have enabled the development of diverse antibody formats, including monoclonal antibodies (mAbs) and antibody-drug conjugates (ADCs) [[1], [2], [3]]. Since the first US Food and Drug Adiministration (FDA) approval of muromonab-CD3 (OKT3) targeting cluster of differentiation 3 (CD3) antigen in 1986 [4,5], over 120 antibody drugs have been licensed for different indications like oncology [6,7], neurological [8], and autoimmune disorders [9] by 2024.

Understanding the dynamic behavior of antibody drugs in vivo is critical for pharmacokinetic and pharmacodynamic studies, including analysis of absorption, biodistribution, metabolism, and excretion. Such insights further enable the evaluation of drug safety and efficacy, identification of potential side effects, and optimization of dosing regimens. Various analysis techniques are involved in this process. Immunoassay methods, such as enzyme-linked immunosorbent assay (ELISA), are widely used for quantitative analysis of antibody drugs due to the high specificity and sensitivity [[10], [11], [12]]. Liquid chromatography-mass spectrometry (LC-MS) method enables structural characterization of antibody drugs and identification of their catabolites through high-resolution mass detection [[13], [14], [15]]. Surface plasmon resonance and flow cytometry have also been reported for the analysis of antibody drugs [[16], [17], [18], [19]]. However, these techniques are mainly suitable for in vitro analysis and are limited by the disability of real-time monitoring. The pharmacokinetic and pharmacodynamic studies rely on discrete blood sampling at predefined intervals. Tissue-level biodistribution and efficacy evaluation necessitate terminal sacrifice of animal models for sample procurement.

Real-time in vivo analysis of antibody drugs enables continuous monitoring of biodistribution and dynamic changes in living systems, thereby minimizing animal sacrifice. However, there are a series of challenges needed to be addressed for in vivo analysis. The structural and functional complexity of antibody drugs results in slow distribution kinetics and diverse metabolic pathways, necessitating analytical methods with prolonged signal stability. Furthermore, complex biological environments of as blood and tissue poses significant challenges for in vivo tracing of antibody drugs, which requires high signal to noise ratio and strong penetration detection capability of analysis methods. Notwithstanding these challenges, promising progress has been made in real-time analysis of antibody drugs in living organisms [20]. Imaging techniques like radiolabeled imaging and fluorescence imaging have been successfully applied across species from murine models to humans for biodistribution and metabolism studies of antibody drugs [21,22]. Additionally, emerging techniques like surface-enhanced Raman scattering (SERS), have exhibited promising application potential for in vivo tracing [23]. These innovative techniques significantly bolster the advancement and clinical utilization of antibody drugs.

Currently, comprehensive reviews have systematically evaluated analytical methodologies for antibody drugs [20,24]. These reviews gave a comprehensive summary of analytical methods for one specific drug, such as mAbs and ADCs. By contrast, this mini-review aims to introduce briefly the advances of different types of antibody drugs, and summarize emphatically the principle and representative examples of in vivo analysis techniques, including radiolabeled imaging, near-infrared fluorescence imaging (NIRFI), and SERS. Emerging techniques like fluorescence lifetime imaging (FLIM) that have received limited coverage in previous reviews are specifically introduced. In addition, the advantages, limitations, and future perspectives of these techniques are also discussed, aiming to shed light on their development potential and application scope.

2. Antibody drugs

2.1. Characteristics

In contrast to traditional small molecule therapeutics, antibody drugs exhibit larger molecular masses and more intricate structures, therefore performing distinct characteristics in administration modes, therapeutic mechanisms, distribution and metabolism [25]. Due to their poor membrane permeability and gastrointestinal instability, antibody drugs are typically administered via intravenous, intramuscular, or subcutaneous injection [26]. Additionally, antibody drugs exhibit low renal clearance, prolonged plasma half-lives (e.g. ADCs: about 1 week [27]), and multifaceted therapeutic mechanisms, including cell growth inhibition, immune effector activation, and cytotoxic payload delivery [28]. Finally, antibody drugs also exhibit unique distribution kinetics and metabolic pathways. Targeted accumulation at disease sites represents a hallmark feature, while metabolic elimination occurs primarily through macromolecular catabolic processes such as proteolytic hydrolysis [29], immune system clearance [30], and nonspecific clearance [31]. These processes render their metabolic behavior exceedingly complicated. A prime example is target-mediated drug disposition in immune clearance, where antibody-target binding initiates pharmacological responses, altering metabolic kinetics in a dose-dependent manner.

2.2. Classification

The Y-shaped monomer of an antibody generally consists of two variable antigen binding domains (Fab) and a constant domain (Fc). Various antibody drugs have been developed, leveraging the basic antibody structure, which include mAbs, bispecific antibodies (BsAbs), ADCs, Fc fusion proteins, antibody fragment drugs, and antibody prodrugs (Fig. 1).

Fig. 1.

Common types of antibody drugs.

Therapeutic mAbs are currently the most common and commercially valuable antibody drugs. The evolution of mAbs undergoes a meticulous engineering journey, from murine antibodies to fully humanized versions, aiming to reduce immunogenicity [25,32,33]. Advanced methodologies, such as phage display and single B cell production technology, have been adopted to produce mAbs for optimizing clinical performance [34,35]. The first US FDA-approved monoclonal antibody (mAb), OKT3, was used for the treatment of organ transplant rejection and autoimmune diseases [5]. Up to now, a variety of mAbs have been developed and approved by US FDA including blinatumomab [36], moxetumomab pasudotox-tdfk [37], brentuximab vedotin [38], cetuximab [39], rituximab [40], and necitumumab [41], which are applied in malignant tumors [42], malaria [43], and organ transplant rejection [44].

BsAbs are a family of antibodies designed to recognize two antigens or two different epitopes of the same antigen [45]. The original concept of BsAbs were first proposed by Nisonoff et al. [46] in 1960. BsAbs possess the potential of bridging functional receptors or cells. T-cell redirection bridged by BsAbs to specifically eliminate tumor cells is the most common therapeutic mechanism. For example, ertumaxomab links T cells and breast cancer cells to induce cell elimination through HER2-binding domain and T cell-binding domain [47]. Several BsAbs have been approved by US FDA and more than 100 BsAbs are in the development stage and clinical research, which are mainly focused on cancer therapy.

ADCs consist of three key components: (1) an antibody for target recognition; (2) a chemically stable linker; and (3) a cytotoxic payload. The linker covalently connects the antibody's Fc region to the payload, which is released either through intracellular antibody degradation or site-specific linker cleavage upon target engagement. The drug-to-antibody ratio (DAR), typically ranging from 2 to 4, is a critical quality attribute that directly influences both therapeutic efficacy and systemic toxicity [48]. In addition to cytotoxic small molecule drugs, a variety of payloads are alternative, such as radionuclides, peptides, and oligonucleotides. Mylotarg was the first ADC approved by the US FDA for acute myeloid leukemia therapy. Whereafter, the second-generation ADC like brentuximabvedotin and the third-generation ADC like polatuzumab vedotin were successively approved with the improvement of solubility and site-specific conjugation [49,50].

Fc fusion proteins are biologics created by the fusion of Fc fragments of immunoglobulin and functional protein. Fc fusion proteins show merits of extended half-life in plasma and reduced immunogenicity [51]. Etanercept is the first Fc-fusion protein drug for the treatment of autoimmune diseases approved by US FDA in 1998, which is formed by the extracellular part of human tumor necrosis factor receptor 2 and the Fc segment of immunoglobulin G1 [52]. In recent years, various Fc fusion protein drugs have been developed for oncology, immunology, inflammatory diseases, and neurological diseases. At least 13 Fc-fusion protein drugs are approved worldwide and about 40 Fc-fusion protein drugs are in clinical development by August 2023 [[53], [54], [55]].

With the pursuit of higher efficacy and safety, other antibody drugs like antibody fragment drugs and antibody prodrugs have also been developed. Antibody fragment is part of a full-length antibody that retains the specificity of the complete antibody and have better penetration ability in terms of pharmacokinetic properties. Several antibody fragment drugs have already been approved for sale [[56], [57], [58]]. Antibody prodrugs refer to a class of drugs that need to be metabolized in the body before exercising their biological functions to avoid the toxicity of the active antibody. Prodrug strategies are engineered to undergo site-specific activation within pathological tissues, thereby enabling precise therapeutic targeting to enhance efficacy while minimizing systemic side effects [59].

3. Overview of in vivo analytical methods for antibody drugs

Antibody drugs possess unique physicochemical properties and in vivo actions, which place demands on in vivo analysis methods. Firstly, it is a slow process of the distribution from blood to surrounding tissues and clearance of antibody drugs. Long-term signal stability is a critical requirement for in vivo analytical methodologies to ensure accurate characterization of antibody drugs in preclinical and clinical settings. Furthermore, analytical techniques should exhibit high anti-interference capacity and excellent penetration capability to address the complex physiological environments encountered in biological systems. Moreover, non-invasive and real-time monitoring are also imperious demands for analysis techniques to reduce the damage to living organisms. Considering these factors, imaging technology is a feasible method for in vivo analysis. At present, various imaging techniques, such as NIRFI and radiolabeled imaging, have been applied and developed continuously to accurately analyze the dynamic distribution of antibody drugs. In addition, other rapid and non-invasive techniques like SERS also shows potential for in vivo analysis (Table 1 [[60], [61], [62], [63], [64], [65], [66], [67], [68], [69], [70], [71], [72], [73]]).

Table 1.

Summary of detection methods.

Method	Signal unit	Advantage	Challenge	Antibody drugs	Type of drugs	Sample source	Refs.
SPECT	131I	High sensitivity, high penetration capability, and one-time full-body imaging	Relatively high cost and low spatial resolution, potential radioactive hazard	Meprazumab	mAb	Human volunteers	[60]
	99mTc			scFv	Antibody fragment	Mouse model	[61]
PET	89Zr	High sensitivity, high tissue resolution, high penetration capability, and one-time full-body imaging	High cost, potential radioactive hazard	Natalizumab	mAb	Mouse model	[62]
	134Ce			DOTA-trastuzumab	ADC	Mouse model	[63]
	64Cu			64Cu-hJF5-DyLight650	mAb	Mouse model	[64]
NIRFI	IRDye800CW	Low cost, simplicity of operator	Photobleaching, poor water solubility and tissue diffusion capacity	scFv-Fc IgG fragment Fab fragment	Antibody fragment	Mouse model	[65]
	Cy5			Cy5-Ab–SS–SN38	ADC	Mouse model	[66]
	Ag2Se QDs	High fluorescence stability, high signal-to-noise ratio	Potential biocompatibility and toxicity problems	Kadcyla	ADC	Mouse model	[67]
	CdSeTe/CdS QDs			Cetuximab	mAb	Mouse model	[68]
	NaYbF4			Atezolizumab	mAb	Mouse model	[69]
FLIM	IRDye 800CW	High sensitivity, high signal-to-noise ratio	Low imaging speed	Monoclonal anti-EGFR antibody	mAb	Mouse model	[70]
	AF700 AF750			Trastuzumab	mAb	Mouse model	[71]
SERS	Cy7 CyNAM Cy7.5	High sensitivity, cost-effectiveness, rapid detection	Shallow detection penetration, low spatial resolution	Tetraspanin-8 antibody	Antibody	Mouse model	[72]
	DTTC			Cetuximab	mAb	Mouse model	[73]

PET: Positron emission tomography; SPECT: Single-photon-emission computed tomography; NIRFI: Near-infrared fluorescence imaging; scFv: Single chain variable fragment; QDs: Quantum dots; FLIM: Fluorescence lifetime imaging; SERS: Surface-enhanced Raman spectroscopy; mAb: monoclonal antibody; ADC: antibody-drug conjugate.

3.1. Radiolabeled imaging

Radiolabeled imaging techniques utilize radioactive element-labeled antigens or antibodies to recognize antibodies or antigen and further detect the radioactive signal. Radioimmunoassay is one of the most accurate and sensitive methods for diagnosis of various diseases with the advantages of high sensitivity, high penetration capability, real-time monitoring, and in-situ analysis, which has been a powerful tool in clinical applications. Multiple subdivision techniques have been used to study the dynamic distribution and evaluate the therapeutic effect of antibody drugs, such as single-photon-emission computed tomography (SPECT) and positron emission tomography (PET) [74,75].

SPECT imaging is a radiolabeled imaging technology by detecting the photons generated by radionuclide decay of radioisotope (Fig. 2A [76]). SPECT imaging is a useful tool for studying the biological distribution and tumor targeting dynamics of antibody drugs. For example, meplazumab, a humanized mAb targeting cluster of differentiation 147 (CD147), has been investigated as a corona virus disease 2019 (COVID-19) therapeutic candidate. Ye et al. [60] labeled meplazumab with ¹³¹I to study the dynamic biodistribution of ¹³¹I-meplazumab in mice and human volunteers by SPECT imaging. The half-life of ¹³¹I (t_1/2 = 8.02 day) is similar to that of meplazumab. After injection with ¹³¹I-meplazumab, SPECT scans were performed at different time points to monitor the dynamic distribution. The clearance half-life of ¹³¹I-meplazumab in humans was 223.5 h, longer than that in mice (117.4 h). Meplazumab exhibited high uptake in human liver and spleen, aligning with the distribution of CD147 in human body (Fig. 2B [60]). The pharmacokinetic characteristics of ¹³¹I-meplazumab in humans was revealed to fit well with the two-compartment model. This study confirmed the safety, tolerability, and targeting ability of ¹³¹I-meplazumab in human body, providing an important experimental basis for the clinical transformation of meplazumab. In addition to mAbs, SPECT imaging is also reported to analyze antibody fragments. Vascular cell adhesion molecule 1 (VCAM1) is a suitable target for molecular imaging of atherosclerosis due to its role in leukocyte recruitment and adhesion to plaque. Single chain variable fragment (scFv) of VCAM1 can be used to detect vulnerable plaques in animal models of atherosclerosis. Deigner et al. [61] used ^99mTc-labeled scFv to successfully image the unstable plaques in atherosclerosis-induced mice and rabbits by SPECT imaging, which is of great significance for the early diagnosis and risk assessment of atherosclerosis.

Fig. 2.

Schematic diagram of imaging principle and examples of single-photon-emission computed tomography (SPECT) and positron emission tomography (PET). (A) Schematic illustration of principle of SPECT [76]. (B) SPECT imaging of ¹³¹I-meplazumab in healthy volunteers [60]. (C) Schematic illustration of principle of PET [80]. (D) The structure of ⁸⁹Zr-natalizumab [62]. (E) PET/computed tomography (CT) images of ⁸⁹Zr-natalizumab in mice after injection for 7 days [62]. Reprinted with permission from Refs. [60, 62].

SPECT has some inherent limitations to its applications. Compared with PET, the spatial resolution and sensitivity of SPECT are relatively low. In addition, the traditional single-machine SPECT imaging system is difficult to correct attenuation, limiting its use to non-quantitative imaging [77]. Some improvements in hardware and software are applied to solve these limitations. Yang et al. [78] developed an ultrahigh energy-resolution spectral SPECT imaging system. The resolution and sensitivity were significantly improved after usage of 24 CdTe imaging spectrometers and a multi-channel readout circuitry. Shcherbinin et al. [79] developed a quantitative reconstruction algorithm to correct attenuation, scatter, collimator blurring, and collimator septal penetration in SPECT imaging system, which significantly improved the accuracy of quantification with error levels of 3%–5% for various isotopes.

In PET, the positrons emitted by radioisotopes annihilate with nearby electrons to generate photons for detection (Fig. 2C [80]). Compared with SPECT, PET exhibits more accurate quantitative ability, higher spatial resolution, and higher sensitivity. PET provides an important approach for analyzing the efficacy of drugs and implementing precise treatment for cancer patients [20]. Radiolabeling antibody for in vivo characterization is known as immuno-PET, which is currently a fastest-growing tool for biomarker discovery, disease diagnosis, and drug tracing [81].

Fluorine-18 (¹⁸F) and carbon-11 (¹¹C) are currently widely used as radioactive elements in diagnostic applications. While the short half-life of ¹⁸F and ¹¹C is insufficient to monitor the antibody drugs with long pharmacokinetic half-life. Therefore, long half-life isotopes, such as ⁶⁴Cu [82,83], ¹²⁴I [84], ⁹⁰Y [85], ⁶⁸Ga [86], ⁸⁹Zr [22,62], are employed for visualization of antibody drugs in vivo. Among them, ⁸⁹Zr is the most widely used because of the long half-life (t_1/2 = 78.4 h), which is matched to the long-circulating accumulation time of antibody drugs. ⁸⁹Zr-based PET imaging has been used to analyze the biological distribution and study efficacy of various antibody drugs, including mAbs [61], ADCs [62], and antibody fragment drugs [87]. Kim et al. [62] labeled the sulfhydryl part of natalizumab with ⁸⁹Zr to monitor the targeted distribution and provide in vivo pharmacokinetic information of natalizumab (Fig. 2D [62]). The highest tumor uptake of ⁸⁹Zr-natalizumab (2.22% ± 0.41% ID/g) was monitored by PET imaging at 7 days postinjection (Fig. 2E [62]). This long-term therapeutic drug monitoring and accurate quantification are crucial for optimizing antibody-based treatment of hematological malignancies. Sijbrandi et al. [22] investigated the efficacy of ADCs with different linkers with PET imaging. Maleimide (Mal) and ethylenediamine platinum (II) moiety (Lx) are used as the connector payload auristatin F (AF) respectively. The dynamic distribution of ⁸⁹Zr-trastuzumab, ⁸⁹Zr-trastuzumab-Mal-AF and ⁸⁹Zr-trastuzumab-Lx-AF was monitored in mice. The findings demonstrated that ⁸⁹Zr-trastuzumab-Lx-AF exhibited superior tumor targeting capability and the most favorable survival rate. It suggested that the Lx linker is a promising candidate for enhancing the curative effect of ADCs.

Although the abovementioned isotopes have shown excellent performance in PET imaging, the relatively strong coordination properties make them chemically unstable. Bailey et al. [63] used ¹³⁴Ce as PET radionuclide to monitored the uptake of 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA)-based ADCs in mice. ¹³⁴Ce possesses a long half-time (t_1/2 = 75.8 h) and high chemical stability due to the inability of coordination. Micro-PET imaging revealed time-dependent accumulation of [¹³⁴Ce]Ce-DOTA-Trastuzumab in tumors. Elevated tumor uptake was maintained throughout the 214 h observation, with minimal bone and liver uptake, indicating the in vivo stability of ¹³⁴Ce and long-term tumor targeting of [¹³⁴Ce]Ce-DOTA-Trastuzumab. Beyond cancer, antibody drugs are also harnessed for the diagnosis of lung diseases. Henneberg et al. [64] developed a non-invasive imaging method for early diagnosis and therapeutic effect monitoring of invasive pulmonary aspergillosis based on double-labeling humanized mAb (HJF5). HJF5 antibody was labeled with radionuclide ⁶⁴Cu and fluorophore Dylight650. Accurate co-locate and co-quantify aspergillus fumigatus infection and antibody tracer in lung was monitored by immuno-PET/magnetic resonance imaging (MRI) , which was further verified by 3D fluorescence microscope imaging in vitro.

Besides the isotope choice and the protein structure design, PET imaging also faces some technical challenges. For example, the Compton effect of high-energy photons in living organisms leads to the attenuation or scattering effect, thus decreasing the resolution and sensitivity. Some technical improvements have been investigated to solve the limitations. A series of attenuation correction algorithms, such as combining zero echo time sequence with Dixon sequence, were developed to improve the accuracy of attenuation correction [88]. In another study, the position of annihilation events was determined more accurately by measuring the time difference of arrival of photon pairs, thus reducing the uncertainty in the reconstruction process and improving the spatial resolution of images [89].

3.2. Near-infrared fluorescence imaging

Radiolabeled imaging has become an indispensable tool in clinical diagnosis due to its strong penetration ability and high sensitivity. However, there are still some limitations that cannot be ignored, such as potential radiological hazards and high equipment costs. Fluorescence imaging is expected to become an alternative method with the advantages of less damage to organisms, low cost, and convenient operation [90]. Fluorescence imaging enables real-time and non-invasive monitoring to obtain dynamic information inside cells and tissues, revealing key information to understand biological processes and disease mechanisms, making it a powerful tool in disease diagnosis and biomedical research [91]. However, the low penetration depth of visible fluorescence restricts its application in living organism, due to the interference of light absorption and scattering by biological tissues. To address this challenge, near-infrared fluorescence (NIRF) technology is developed. Longer emission wavelengths of near-infrared (NIR) probes can effectively reduce light absorption, light scattering, and background fluorescence in biological tissues, thus performing high penetration depth and signal to noise ratio (Fig. 3 [92]). Owing to these advantages, NIRFI technology plays an increasingly important role in bioimaging of antibody drugs at the living level [93]. Different fluorescent probes, such as organic dyes and quantum dots (QDs), are used as labels to track the biodistribution of antibody drugs and evaluate the drug efficacy, which provides a promising way to understand the in vivo action mode of antibody drugs [[93], [94], [95]].

Fig. 3.

Schematic diagram of near-infrared fluorescence imaging (NIRFI) Reprinted with permission from Ref. [92]. UV: ultraviolet.

Organic NIRF dyes have been widely used in fluorescence imaging with merits of simple functionalization and diverse luminous spectra [96]. Commonly used NIRF dyes include indole heptamethine cyanine dyes [97], boron-dipyrromethene [98], and rhodamine derivatives [99]. A variety of NIRF molecules have been reported for in vivo imaging of antibody drugs. El-Sayed et al. [65] labeled a series of antibody fragments with IRDye800CW to analyze the distribution and clearance characteristics in vivo, which provides important information for image-guided surgery. In particular, they found that larger antibody fragments containing Fc domain (such as IgG and scFv-Fc) exhibited longer fluorescence signal duration. Kobzev et al. [66] synthesized a Cy5-labeled therapeutic ADC (Cy5-Ab–SS–SN38) for NIRFI and evaluated the targeted ability to HER2+ breast cancer. The ADC is composed of HER2-specific antibody trastuzumab and payload SN38 (the active metabolite of irinotecan), and then labeled with NIRF dye Cy5 (Fig. 4A [66]). The real-time accumulation of Cy5-Ab–SS–SN38 in tumors was observed through fluorescence imaging. The efficacy and tolerance of ADCs depend on a great extent on the chemical linker connecting antibodies and small molecules, which affects the stability of ADCs and the release rate of payloads [100,101]. NIRFI can be used to evaluate the cutting efficiency of linker in ADCs. Usama et al. [102] developed a cleavage-activated fluorescent probe of cyanine carbamates with the maximum emission wavelength of 750 nm (denoted as Pan-CyLBam). This probe exhibited excellent cellular permeability and lysosomal accumulation due to the tertiary amine moiety. In particular, the fluorescence is turned off when conjugated to the antibody and turned on after cleavage, providing insight into the cleavage site and efficiency of ADC in complex tissue settings. Common ADC linkers including reduction-cleavable disulfides and cathepsin-cleavable linkers were quantitatively compared in vivo with the fluorogenic turn-on Pan-CyLBam (Fig. 4B [102]). Stronger fluorescence signal in tumor was observed from the mice treated with cathepsin-sensitive probes (Fig. 4C [102]), indicating the higher tumor activation ability of cathepsin-cleavable linkers than that of disulfides linkers. In addition, fluorescence signal also provided significant insight into off-target cleavage pathways. The significant signals in liver at early time points indicated the presence of liver uptake/cleavage and probe clearance through hepatobiliary pathways (Fig. 4D [102]).

Fig. 4.

Near-infrared fluorescence imaging (NIRFI) of fluorescent dye near-infrared (NIR) probe. (A) Cy5-labeled therapeutic ADC (Cy5-Ab–SS–SN38) for targeted NIRFI of HER2+ breast cancer [66]. (B) Structures of antibody-drug conjugates (ADCs) with different linkers labeled by cleavage-activated fluorescent probe. Reproduced with permission from Ref. [102]. (C) Fluorescent imaging after the injection of probes in mice with MD Anderson-Metastatic Breast-468 (MDA-MB-468) tumors for 48 h [102]. (D) Quantification of the fluorescent signal from the liver and the tumor at different time points [102]. Reprinted with permission from Refs. [66, 102].

NIRFI provides a convenient method for in vivo analysis of antibody drugs. However, there are still some concerns needing to be addressed. The poor water solubility and tissue diffusion capacity of organic dyes limit the further application, which expected to be solved by reasonable structure design. For example, introducing shielding units to avoid aggregation or charged moiety to increase hydrophilicity can effectively to improve solubility and tissue penetration capacity [103,104]. It is also notable that the conjugated fluorescent molecules may change the pharmacokinetics of antibody drugs. Glaudemans et al. [77] investigated the biodistribution of Alexa750-labeled and ⁶⁴Cu-labeled mAbs and antibody fragments. It was found that Alexa750-labeled antibody exhibited shorter blood half-life and higher liver uptake than those of ⁶⁴Cu-labeled counterparts.

Besides organic fluorescent dyes, nanoparticles are also feasible choice to construct NIR probes such as QDs and rare earth doped nanoparticles. These nano-materials have the advantages of good light stability, high excitation intensity, and long fluorescence life [105]. At present, different kinds of nanoparticles have been used for NIRFI of antibody drugs. Tsuboi et al. [67] developed a QDs probe to evaluate efficacy of ADCs. CdSeTe/CdS QDs conjugated with annexin V-enhanced green fluorescent protein (EGFP) showed NIRF (λ_max = 850 nm) and specific recognition of tumor apoptosis induced by ADCs. Through NIRFI, the tumor shrinkage was observed after injection with ADCs in mice for a long time (24 days). Cetuximab is a mAb against epidermal growth factor receptor (EGFR), which is used to treat EGFR-positive metastatic colorectal cancer and squamous cell carcinoma of the head and neck. Zhu et al. [68] combined Ag₂Se QDs with NIRF and cetuximab to prepare Ag₂Se-cetuximab nanoprobe through biological conjugation strategy. In vivo NIRFI results of nude mice carrying Ag₂Se-cetuximab nanoprobes at different time points clearly showed the accumulation and retention of nanoprobes in the tumor site of tongue cancer, especially after 13 h. This probe further exhibited remarkable effects in inhibiting tumor growth and improving survival rate.

Rare earth doped nanoparticles are a common class of NIR-II probes. Zhong et al. [69] developed Zn-doped and erbium-based rare-earth nanoparticles NaYbF₄ (2% Er, 2% Ce, 10% Zn@NaYF₄) with a core-shell structure, which possessed NIR-IIb emission window at about 1600 nm and therefore deep-tissue optical imaging ability (Fig. 5A [69]). Then hydrophilic polymer layers and atezolizumab, an anti-programmed death-ligand 1 (PD-L1) mAb, were conjugated to the nanoparticle (denoted as ErNPs-aPD-L1) for molecular imaging of PD-L1 in mice with colon cancer (Fig. 5B [69]). The tumor-to-normal tissue signal ratios was up to 40, which was difficult to realize in previous fluorescent molecular imaging (Fig. 5C [69]). Furthermore, researchers combined ErNPs-aPD-L1 and PbS QDs-labeled anti-cluster of differentiation 8α (CD8α) mAb to target different proteins for two-plex molecular imaging (Fig. 5D [69]), which has potential clinical value for evaluating the therapeutic effect and predicting the patient's response to immunotherapy.

Fig. 5.

Near-infrared fluorescence imaging (NIRFI) of rare earth doped nanoparticles near-infrared (NIR) probe. (A) Schematic design of Zn-doped erbium-based rare-earth nanoparticles (ErNPs) and corresponding transmission electron microscopy characterization [69]. (B) Schematic illustration of the hydrophilic ErNPs with cross-linking polymeric layers and amine groups for antibody conjugation [69]. (C) NIRFI of CT-26 tumor mice treated with ErNPs-programmed death-ligand 1 (aPD-L1) and free ErNPs at different time points [69]. (D) Two-plex fluorescence imaging of a colon tumor 26 (CT-26) tumor mouse at 24 h post intravenous injection of mixed ErNPs-aPD-L1 and PbS-anti-cluster of differentiation 8α (aCD8α) [69]. Reprinted with permission from Ref. [69].

The potential toxicity of QDs and rare earth nano-material caused by heavy metal components remains a concern for in vivo imaging [100]. Surface modification of biocompatible molecules like hydrophilic polymer is proved an effective strategy to enhance the biocompatibility [106,107]. In the example of erbium-based rare-earth nanoparticles NaYbF₄ [69], the hydrophilic and cross-linked coating layers endowed the nanoparticles biocompatibility and safety. 90% of nanoparticles were excreted form mice within 2 weeks and no detectable toxicity was observed.

In addition to traditional fluorescence intensity imaging, FLIM is also an effective means to analyze antibody drugs. FLIM obtains images by measuring the fluorescence lifetime of fluorescent molecules [108]. Compared with fluorescence intensity imaging, FLIM is largely impervious to influences such as the intensity of stimulated luminescence, probe concentration, and photobleaching, but uniquely related to microenvironment [109]. Moreover, it is capable of distinguishing different fluorophores with similar fluorescence spectra [110], thereby offering enhanced signal to noise ratio, specificity, and sensitivity. FLIM has been proved to have significant advantages in improving the specificity and sensitivity in the analysis of antibody drugs.

The slow clearance of antibodies in vivo causes the nonspecific antibody accumulation in clearance organs like the liver, which interferes with the accurate imaging of tumor tissue. Pal et al. [70] labeled antibodies against EGFR with a NIRF dye IRDye 800CW (called anti-EGFR-800) to image human breast cancer in vivo with FLIM. The fluorescence lifetime of anti-EGFR-800 in tumor was longer (0.7 ns) than that of non-specific probe in liver (0.63 ns), which significantly improved the contrast between tumor and background, thus enhancing the sensitivity and specificity of imaging. Verma et al. [71] combined FLIM and Förster resonance energy transfer (FRET) to quantify targeted drug delivery of trastuzumab, a mAb targeting HER2. Trastuzumab was conjugated to NIR probes donor fluorophore (Alexa Fluor 700, AF700) or acceptor fluorophore (Alexa Fluor 750, AF750). The FLIM FRET signal was observed when different NIR fluorophore-labeled trastuzumab bound to HER2 in intact in HER2-positive breast cancer cells (Fig. 6 [71]), which could delineate the tumor margins with high signal-to-noise ratio and quantify the bound and unbound fraction of antibody. Trastuzumab labeled with AF700 or AF750 was injected into the mice with different types of HER2-positive tumor xenografts. Higher FLIM FRET levels were observed in HER2+ breast tumors with high vascularity and low collagen than that in HER2+ ovarian tumors with low vascularity and high collagen, indicating the higher fraction of bound and internalized trastuzumab-HER2 complexes in HER2+ breast tumors.

Fig. 6.

Schematic illustration of fluorescence lifetime imaging (FLIM) and förster resonance energy transfer (FRET) methods revealing the effect of tumor microenvironmental factors on delivery of near-infrared (NIR) labeled trastuzumab. Reprinted with permission from Ref. [71].

Compared with fluorescence intensity imaging, the speed of FLIM is generally lower due to the repeated scanning to collect enough photons for each pixel to fit the fluorescence lifetime. Some improvements in hardware and software were put forward for fast imaging. Koenig et al. [111] employed the updated hardware, time-correlated single photon counting card with ultra-short dead time detector, to build a FLIM imaging systems with increased count rates, which could speed up imaging acquisition by more than 100 times. In another research, Marois et al. [112] developed enhanced fitting algorithm to improve the imaging speed.

NIRFI has made great progress in the analysis of antibody drugs in vivo. Though there still exist some challenges like limited penetration depth, spatial resolution and permeability of tissue, it is anticipated that limitations will be improve greatly in the future, with the swift advancement of fluorescence probe synthesis, laser technology, detector technology, super-resolution imaging, and algorithm.

3.3. Surface-enhanced Raman spectroscopy

Besides widely used radiolabeled imaging and fluorescence imaging, other cutting-edge methodologies have also exhibited promising application potential for in vivo analysis of antibody drugs. SERS detects the scattering spectra of molecules under laser excitation, providing fingerprint-like information about molecular vibrations and rotations. SERS exhibits distinguishable narrow bandwidth (<1 nm) and efficient laser excitation in NIR region, making it an excellent tool for in vivo protein analysis [113]. SERS probes typically contain Raman reinforced nanoparticles, Raman label compounds and antibodies. The antibody trace is reflected by detecting the signal of the Raman label compounds, which can be used to study the biodistribution, metabolic process, and affinity. Kang et al. [72] constructed NIR-SERS probes for in vivo validation of antibody candidates. Au/Ag hollow-shells were first formed on the surface of silica core as substrates. Different NIR dyes were conjugated to the surface of Au/Ag hollow-shells by chemisorption to prepare NIR-SERS dots with absorbance bands in the range of 740–790 nm (denoted as SERS dot [Cy7], SERS dot [CyNAM], and SERS dot [Cy7.5]) (Fig. 7A [72]). Then three tetraspanin-8 (TSPAN8) antibody candidates against human colon carcinoma (HCT8) were conjugated severally to the surface of NIR-SERS dots (denoted as C2-SERS dot [Cy7], C4-SERS dot [CyNAM], and C5-SERS dot [Cy7.5]). After injecting the mixture of three probes into human colon cancer xenograft mice, the dynamic biodistribution in organs and tumor tissue of different antibody candidates could be observed with a fiber-coupled portable Raman system (Fig. 7B [72]). SERS spectra and further normalized quantitative analysis revealed that C2 antibody showed the strongest affinity for tumor tissue (Fig. 7C [72]), which was in good agreement with that of SPECT. This method showed great potential as a cost-effective and accurate multiplexing tool for in vivo validation of antibody drugs. In another example, Conde et al. [73] conjugated the mAb cetuximab to SERS dots to construct probe for targeting EGFR on xenograft tumor mice models. The Raman signal is observed simultaneously with extensive tumor growth inhibition in mice, indicating this probe ideal for tumor detection and drug efficacy evaluation.

Fig. 7.

Schematic diagram of principle and examples of near-infrared (NIR)-surface-enhanced Raman spectroscopy (SERS) probe. (A) Schematic diagram of NIR-SERS probes, composed of Au/Ag hollow-shell assembly, NIR dye, silica shell and antibodies. (B) Photograph of SERS measurement on a human colon cancer xenograft mouse with a fiber-coupled portable Raman system after injection for 1 h. (C) SERS spectra of the skin, liver, and tumor sites of the mouse. Reprinted with permission from Ref. [72].

Although SERS shows merits of high detection sensitivity and cost-effectiveness, several problems hamper the application for in vivo tracing of antibody drugs. Current SERS methods rely on labeled detection mode. Label-free protein fingerprinting is highly valuable, but challenging in complex in vivo environment. In addition, shallow detection penetration and low spatial resolution limits the in vivo tracing in large animals and deep-tissue imaging. Improvements in method or instrument have been developed to solve these problems. He et al. [114] reported a biomimetic recognition-driven plasmonic nanogap-enhanced Raman scattering for direct protein fingerprinting and quantitation. Silver nanoparticles were first coated with two kinds of molecularly imprinted polymers that recognized the N-and C-termini of target proteins, respectively. The gap between the two particles specifically trapped target protein into well-defined plasmonic nanogaps, therefore achieving protein fingerprinting with ultrahigh sensitivity and quantitation robustness. Zhang et al. [115] integrated ultra-bright SERS nanotags and transmission Raman spectroscopy to construct a deep-seated detection/imaging system. The detection penetration was up to 14 cm in ex vivo porcine tissues. Combining SERS and other imaging methods is also an effective strategy to provide complementary bioimaging features. Pan et al. [116] synthesized gold nanorod and ultrasmall iron oxide nanoparticle composites as dual-mode probes for SERS-magnetic resonance imaging of PD-L1 in triple-negative breast cancer, which showed the merits of high sensitivity and deep penetration. These advances pave the way for further in vivo application of SERS.

4. Conclusion

Antibody drugs have been experiencing exponential growth due to the high specificity, low immunogenicity, and long half-life. Numerous and various antibody drugs are approved to the clinic treatment for multiple diseases, such as cancer, neurological diseases and autoimmune diseases. In vivo analysis of antibody drugs provides direct information for revealing pharmacokinetic profiles and therapeutic mechanisms within living organisms. Multiple techniques have been applied for the in vivo assessment of targeting and efficacy in preclinical and clinical settings. Among these, radiolabeled imaging techniques including SPECT and PET have been powerful tools for non-invasive quantification at the human scale, leveraging their high sensitivity and deep tissue penetration. However, the associated risks of radiation exposure and prohibitive costs should be considered in the development of analysis methods. NIRFI shows the merits of high signal-to-noise ratio, low biological damage, and easy operation. Various fluorescent labels like organic dyes and inorganic nanoparticles are developed to conjugate to antibody drugs. Different imaging modes based on intensity and lifetime provide variety of options and convenience for the construction of analytical methods. In addition, SERS has also been reported for in vivo targeting evaluation of antibody drugs, which can achieve rapid and cost-effective analysis. There is no doubt that the abovementioned techniques will keep continuous improvement and playing essential roles with the development of antibody drugs.

Despite significant advances in antibody drugs analysis methods in vivo, there are still several challenges. First, unimodal approach has the inherent limitations like low resolution and high background interference. Exploring multi-modal analytical strategies can combine the advantages of different methods for comprehensive and accurate analysis, such as PET-fluorescence imaging and PET-MRI imaging. In addition, current approaches locate drugs mainly by detecting the signal of labeled units. Changes in structure and composition of antibody drugs are important information, but difficult to identify. Incorporation of structure resolution techniques into in vivo analysis is a possible approach. For example, MS methods are widely used in the structural resolution of antibody drugs, but mainly focus on ex vivo analysis. Recently, some in vivo MS methods have been reported for real-time analysis of small molecules, such as open-flow microperfusion combined with MS and ultrasonic sputter desorption MS, which show potential for structure resolution of antibody drugs in vivo. Finally, most current studies are conducted in mouse models. The in vivo analysis of antibody drugs in human being is certainly of greater significance but faces difficulties in feasibility and safety. The detection methods need to be comprehensively improved in terms of penetration depth, signal stability, and detection security to pave the way for experimenting on human being. In summary, the in vivo analysis of antibody drugs is a developing field, and future research will introduce new technologies and methods to better understand and evaluate the dynamic behavior and therapeutic effects of antibody drugs in vivo.

CRediT authorship contribution statement

Xiaolu Miao: Writing – review & editing, Writing – original draft. Beilei Sun: Writing – review & editing, Writing – original draft. Jian Zhang: Supervision. Jinge Zhao: Supervision. Bing Ma: Writing – review & editing, Supervision. Yongming Li: Writing – review & editing, Supervision. Weizhi Wang: Writing – review & editing, Supervision, Resources, Funding acquisition.

Declaration of competing interest

The authors declare that there are no conflicts of interest.