Multiphoton microscopy in surgical oncology- a systematic review and guide for clinical translatability

Highlights

• MPM offers functional real time imaging on a microscopic level without tissue preparation.

• Imaging of fresh tissues within an hour provides the best signal, including functional imaging.

• Tumor cells can be detected with a high sensitivity and specificity.

• Use of MPM in a sterile field of open surgery is feasible but requires extensive preparation.

Abstract

Background

Multiphoton microscopy (MPM) facilitates three-dimensional, high-resolution functional imaging of unlabeled tissues in vivo and ex vivo. This systematic review discusses the diagnostic value, advantages and challenges in the practical use of MPM in surgical oncology.

Method and Findings

A Medline search was conducted in April 2019. Fifty-three original research papers investigating MPM compared to standard histology in human patients with solid tumors were identified. A qualitative synopsis and meta-analysis of 14 blinded studies was performed. Risk of bias and applicability were evaluated.

MPM can image fresh, frozen or fixed tissues up to a depth 1000 μm in the z-plane. Best results including functional imaging and virtual histochemistry are obtained by in vivo imaging or scanning fresh tissue immediately after excision. Two-photon excited fluorescence by natural fluorophores of the cytoplasm and second harmonic generation signals by fluorophores of the extracellular matrix can be scanned simultaneously, providing high resolution optical histochemistry comparable to standard histology. Functional parameters like fluorescence lifetime imaging or optical redox ratio provide additional objective information. A major concern is inability to visualize the nucleus. However, in a subpopulation analysis of 440 specimens, MPM yielded a sensitivity of 94%, specificity of 96% and accuracy of 95% for the detection of malignant tissue.

Conclusion

MPM is a promising emerging technique in surgical oncology. Ex vivo imaging has high sensitivity, specificity and accuracy for the detection of tumor cells. For broad clinical application in vivo, technical challenges need to be resolved.

Introduction

Multiphoton microscopy (MPM) was first described in 1990 by Denk et al. [[1], [2], [3]] (Fig. 1).The imaging technique is based mainly on the detection and combination of two signals: two-photon excited fluorescence (TPEF), and second harmonic generation (SHG), that are detected at specific emission wavelengths after laser excitation. MPM offers in vivo, real time, high resolution, functional visualization without need for contrast agents, tissue staining, or tissue processing [2,4] up to a tissue depth of several hundred microns [5]. Since 1990, a number of practical uses have been established in vitro and in vivo, in live animals, and even in humans [6]. Today, three-dimensional (multiphoton tomography, MPT), or even four-dimensional (multiphoton tomography with fluorescence lifetime imaging, MPT-FILM) signal detection is possible; either of unlabeled fresh [7], frozen [8] or formalin fixed [9] human tissues. In all cases, MPM has the potential of producing images of a comparable resolution to standard staining techniques, such as hematoxylin and eosin (HE), or others [8].

Natural fluorescence from endogenous fluorophores and second harmonic generation signals can be scanned simultaneously after excitation with only one wavelength [4]. Simultaneous absorption of two photons of lower energy leads to excitement of endogenous fluorophores without phototoxic tissue damage by high energy photons (Fig. 2). TPEF signals are mainly produced by excitation of cytoplasmatic and mitochondrial fluorophores, mainly reduced nicotinamide adenine dinucleotide (NADH) and oxidized flavin adenine dinucleotide (FAD) [4,7,8,10], but also by extracellular molecules like flavines [11] and collagen [12]. In contrast, SHG signal generally derives from collagen [12] of the extracellular matrix (Fig. 3, Fig. 4). Longer (near infrared) excitation wavelengths enable a deeper tissue penetration [4,8] up to several hundred microns [5], or even 1000 μm in skin specimen [13]. In order to further differentiate between fluorescent structures, specific spectral ranges can be selected to create “virtual histochemistry” in fresh unlabeled tissue [14]. The spectral window of excitation wavelengths is limited by the phototoxicity of high energy photons, which can cause tissue damage, impaired reproduction capacity and apoptosis-like cell death [15].

Due to the relatively large dimensions of current MPM microscopes, use of the technology has mostly been limited to ex vivo imaging of histological specimen, which could theoretically be performed directly in the operating room under sterile conditions, for example to determine resection margins during surgical procedures [6,16]. Other limiting factors of current MPM technology are small scanning fields [17], and maximum scanning depth of 200 μm up to 800 μm, in certain circumstances up to 1000 μm, depending on the absorption and excitability of the tissue along with targeted magnification [6,8]. To date, in vivo multiphoton tomography (MPT) on humans has been used on skin lesions [[18], [19], [20]] and intraoperatively in a single patient with glioblastoma multiforme [6]. In these applications, the quality of MPM images was found to be comparable to ex vivo scans [6]. In the future, in vivo imaging may facilitate optical ad hoc biopsy combined with endoscopy or laparoscopy, “optical needle biopsy” [4], intraoperative real-time evaluation of resection margins [6,21] and nerve sparing surgery [22]. Even automated histopathological image analysis by machine (“deep”) learning using artificial intelligence has been investigated [10]. Tissue evaluation using deep learning has been reported to yield sensitivity up to 95% and specificity as high as 97% for ovarian cancer cells [10] and accuracy over 90% for hepatocellular carcinoma [23]. This technology might give clinicians the opportunity of intraoperative MPM evaluation in absence of highly trained and specialized pathologists in the future.

The technology has some limitations. As mentioned, contemporary MPM hardware is quite spacious, and tissue penetration depth is limited. Current technical research focuses on development of smaller probes with longer wavelengths, that penetrate deeper into tissues and are less influenced by absorption and scattering of other molecules in the tissue (e.g. hemoglobin) [24]. Other general technical obstacles in the practical application of MPM are inhomogeneous signal intensity in varying depths, low signal-to-noise-ratio and blurring of cell boundaries, which increases in deeper tissue layers [2]. These challenges have been addressed by computer processing methods such as super pixel-based image segmentation and watershed [2]. Nevertheless, in regard of resolution, MPM offers unique high quality imaging on a submicron level, superior to other functional imaging techniques such as positron emission tomography, magnet resonance imaging or computed tomography [24].

In most cases, the established criteria for malignancy have traditionally been based on HE findings and include nuclear-cell ratio (NCR), cell and nuclear polymorphism, including chromatin condensation and mitosis, as well as apoptotic cells [25] (Fig. 3, Fig. 4d). Crossing of suspect cells through the basal membrane is a sign of invasiveness [26]. In contrast to standard HE staining, cell nuclei [11] are not readily visualized directly by MPM, but are detected as an area with absence of TPEF/SHG signal [27,28], which is referred to as the “nuclear area” (Fig. 3, Fig. 4). Therefore, to date, visualization of important cancer characteristics, like chromatin condensation, mitosis and apoptotic cell is still considered to be more accurate in HE stained specimen [25,28] (Fig. 3, Fig. 4).

Neoplasms can be detected and characterized by MPM depending on the ratio of cell organelles in the TPEF signal (e.g. NCR) [29]. Additional information can be obtained by analyzing collagen content and –distribution by its SHG signal, which is usually not easily detected in HE stained sections [11,27]. Basement membranes can be visualized by their SHG signal to determine cancer cell invasion [26]. Cell-to-cell communication mechanisms like tunneling nanotubes and their up-regulation in certain cancer types give another hint of tumor biology [30]. The optical redox ratio (ORR) as a measure of cell metabolism, can be defined by quantifying the tissue signal intensity of NADH and FAD at specific excitation wavelengths [11]. The ratio of free to protein bound NADH is an indirect measure of glycolysis and oxidative phosphorylation, thus reflecting the metabolic state of a cell [7]. Tumor cells show an elevated ORR compared to tumor tissues after chemotherapy or healthy tissues [11,31]. Fluorescence lifetime imaging (FLIM) is another additional MPM technique to discriminate further between different types of cells. Instead of detecting the fluorescence intensity, this technique quantifies the duration of fluorescence lifetime after laser excitation [17]. Further complementary imaging techniques are stimulated Raman scattering and coherent anti-Stokes Raman spectroscopy (CARS) [32]. These imaging techniques are based on vibrational microscopy and inherent specific scattering of tissue molecules (e.g. lipids) after laser excitation, facilitating intrinsic chemical mapping of tissues [32].

In basic research, three-photon excited fluorescence is a promising technique, providing improved optical sectioning and resolution compared to TPEF in a depth of up to 500 μm across the scull in a living mouse [33]. Excitation at longer wavelengths results in deeper tissue penetration, weaker scattering and therefore a better resolution in deeper tissue layers [33]. To our knowledge however, this technology has not been used in patients with solid tumors.

The objective of this systematic review is to determine the value of MPM imaging in surgical oncology, to identify specific tumor characteristics in MPM imaging, to compare to gold standard histopathology, and to discuss current clinical applications, latest advances and future possibilities, as well as the current limitations for MPM imaging for surgeons in the clinical setting.