Multiphoton Microscopic Study of the Renal Cell Carcinoma Pseudocapsule: Implications for Tumour Enucleation

Background To utilise Multiphoton Microscopy as a novel imaging technique to characterize and quantify collagen at the Renal Cell Carcinoma Pseudocapsule, to assess for both intra-tumoral and inter-tumoral variation of collagen characteristics. MPM combines Second Harmonic Generation and Two Photon Excitation Fluorescence to image extracellular matrix architecture. Methods 20 partial nephrectomy specimen tissues were retrieved, cut into 5 -micron sections, mounted on slides and deparaffinized. The pseudocapsules were imaged with 2X and 20X objective at selected Regions of Interest. Corresponding clinical information was retrieved. Pseudocapsule thickness was determined. Collagen parameters measured included quantification by the Collagen Area Ratio, and qualitative measurements by the Collagen Fiber Density and Collagen Reticulation Index. Results The boundaries between tumor, PC and normal renal parenchyma were distinguished by Multiphoton Microscopy without the need for staining. In the thickest areas of the pseudocapsule, collagen content and density were quantitatively higher compared to the thinnest areas. Median Collagen Area Ratio was higher in the thickest compared to the thinnest areas of the PC (p=0.01). Clear Cell RCC specimens had a consistently higher Collagen Fiber Density in both the thickest and thinnest areas compared to non-Clear Cell RCC specimens (p=0.02). Conclusions In this study, we demonstrated the ability of Multiphoton Microscopy to quantify collagen characteristics of pseudocapsules without fluorescent labelling. Tumor enucleation for Renal Cell Carcinoma along its Pseudocapsule remains debatable with regards to oncological safety. Even with a complete and intact pseudocapsule, the pseudocapsule is not a homogenous structure, and varies in its thickness and its collagen characteristics within, and between, tumours.


Introduction
to a proposed standardized PC invasion scoring system to classify PC integrity, as current standard pathological reporting do not routinely report PC integrity. 2 A major component of the fibrous PC is collagen. 3 Collagen, being an abundant Extra-Cellular Matrix (ECM) component, is the strongest and hardest to penetrate by tumor cells. 4 During tumor expansion, collagen remodeling and reorientation occurs, with increased collagen deposition and density for mechanical strength to resist tumor growth. Degradation of collagen has also been shown necessary for tumor invasion. 4 However, limited literature exists on the collagen characteristics of PCs, which are at the interface of tumor expansion, and probably important in resisting invasion.

MPM combines Second Harmonic Generation (SHG) and Two Photon Excitation (TPE) Fluorescence to
image extracellular matrix architecture in great detail. 5 Although MPM has been used to image kidney tumors, MPM's capabilities are in the detailed evaluation of collagen and is potentially a better tool to evaluate collagen characteristics at the tumor-parenchymal boundary.
Tumor Enucleation for RCC has gained popularity as a Nephron-Sparing Surgery technique to maximally preserve renal parenchyma and reduce ischemia time. 6 Tumor Enucleation relies on dissection along the fibrous Pseudocapsule at the tumor-parenchyma interface. Hence, a deeper understanding of the inherent characteristics of the PC as a tumor-parenchyma boundary is important to determine the safety of Tumor Enucleation as an oncological approach. 7 Therefore, the aim of this study was to utilize MPM as a novel imaging technique to characterize and quantify collagen at the PC, and assess for both intra-tumoral and inter tumoral variations in the characteristics of the PC.

Materials And Methods 2.1 Sample Acquisition
The study protocol was approved by the National Healthcare Group Domain Specific Review Board,  9 Tumor stage was based on the updated 8th Edition American Joint Committee on Cancer Staging Manual. 10 Non-identifiable corresponding clinical information from our partial nephrectomy database was retrieved.

Sample Preparation and Multi-Photon Microscopy
The unstained de-paraffinized Formalin-Fixed Paraffin-Embedded tissues of 5-micron thickness were imaged using the commercially available laser-based Genesis200™ Multi-Photon Microscopy system (Histoindex Pte. Ltd, Singapore). Image acquisition was performed at 2X and 20X objective with 512 × 512-pixel resolution (Fig. 1). MPM is an imaging modality comprising SHG and TPE microscopy. SHG signals are detected when a laser of femtosecond range interacts with tissues that have unique, noncentrosymmetric structure within the ECM such as collagen fibers and elastin. When this SHG signal is combined with auto-fluorescence signal from the cells using TPE, MPM is able to penetrate deep into tissues to generate a high-resolution image that reflects both the architecture of the ECM and surrounding cellular structure. 11 On SHG microscopy, the fibrous PC was readily identified, with fluorescent green signals on SHG analysis corresponding to collagen fibers (Fig. 2). In contrast, collagen-devoid tumor tissues and collagen-poor normal renal parenchyma demonstrated minimal SHG signals.
Embedded within the proprietary stain-free imaging technology and analysis software lies the capability to detect sensitively and quantify fine collagen dynamics that are otherwise unobserved with traditional staining techniques. 11 The analysis is able to identify and characterize previously described SHG collagen characteristics that corresponds with the collagen profiles of the tissues. [12][13][14] The area occupied by collagen relative to the total area of the sample was expressed as the Collagen Area Ratio (CAR). The density of collagen fibers within a total given area was expressed as the Collagen Fiber Density (CFD). Both CAR and CFD quantify the collagen content of the tissues. After the collagen area was defined, a skeletonization of the collagen was applied, providing a schematic representation of the collagen fibers and their intersection points. The Collagen Reticulation Index (CRI) was defined as the total number of intersection points in the given collagen area, representing the degree of reticulation and interconnection of collagen fibers.
Imaging and analysis were performed for the PC regions abutting the adjacent normal parenchymal tissue for each of the specimens. The thickest and thinnest PC areas for each tumour specimen were designated Regions of Interests (ROIs) for PC analysis of collagen parameters in each specimen. At each ROI, clinically validated algorithms quantify collagen content and structure. These algorithms were validated against conventional stained H&E images by pathologists at the National University Hospital, Singapore. Collagen content was measured quantitatively by CAR and CFD, and collagen structure was measured by CRI.

Statistical Analysis
Parametric continuous variables were reported as Mean +/-Standard Deviation. Non-parametric continuous variables were reported as Median (Inter-Quartile Range). There were no missing values from data collected, and there were no side effects to the patients. The Mann-Whitney U test was used to compare non-parametric continuous variables. Statistical significance in this study was set as p < 0.05. All reported P values were 2-sided, and analyses were performed with SPSS (version 17.0).

Pseudocapsule MPM Imaging of RCC Histological Subtypes
The boundaries between tumor, PC and normal renal parenchyma were distinguished without the need for staining under light microscopy ( Fig. 1). Figure 4 shows the imaging and tumor specimen of a pT1a Papillary RCC specimen with a RENAL Nephrometry score of 6a. Figure 5 shows the imaging and tumor specimen of a pT1b Clear Cell RCC specimen with a RENAL Nephrometry score of 9p.

Pseudocapsule Collagen Characteristics
The median thickness of the thickest areas of the PCs was 0.5 (0.3-0.7) mm, compared to 0.2 (0.1-0.3) mm in the thinnest areas. In the thickest areas of the PC, collagen content and density were quantitatively higher compared to the thinnest areas (Table 2). Median CAR was significantly higher in the thickest compared to the thinnest areas of the PC (p = 0.01). Mean CFD was also higher (p = 0.09) in the thickest compared to the thinnest areas of the PC (p = 0.09). The thickest areas of the PCs showed measurable higher degrees of collagen cross-linking and collagen network complexity. Mean CRI was higher in the thickest PC areas compared to the thinnest areas of the PC (p = 0.19).  Table 3 and Table 4 shows the comparison between PC collagen characteristics between the thickest and thinnest areas of the PC, stratified by histological subtype, RENAL Nephrometry score and tumor size. Clear Cell RCC specimens had a consistently higher collagen fiber density in both the thickest and thinnest areas compared to non-Clear Cell RCC specimens (p = 0.02). There were no other statistically significant differences when comparing the other subgroups.

Discussion
This initial study demonstrated that the use of commercially available MPM enabled detailed imaging Multi-Photon Microscopy is a highly accurate tool for imaging of collagen and analysing its organization from molecular scale up to tissue architectural level. It combines the advantages of a non-linear imaging approach with a coherent modality able to probe molecular organization. [12][13] Another key advantage is the ability to process fresh, unstained tissues without the need for staining. 5  subtypes based on their tissue architecture, with a diagnostic accuracy of 95%. 20 It was suggested that MPM may potentially be used for real time unstained kidney tissue biopsy diagnosis and surgical margin assessment in future. 20 The findings in our report support the other reports that the presence and integrity of RCC pseudocapsules are known to vary between tumours. Higher rates of PC presence and PC completeness have been found in clear cell RCC. 3 Higher risk of PC invasion have been demonstrated in larger tumors and tumors with incomplete PC. 21 PC invasion has also been associated with higher grade and stage of tumors, which independently predict the risk of tumor recurrence. 1  Indeed, there are few studies analysing the characteristics of PCs beyond just structural integrity.
Provenazano et al described the process of increased extra-cellular matrix collagen deposition, remodelling and crosslinking to resist tumour expansion. 23 Wang et al studied of PCs in small renal masses, and established collagen and reticulin to be abundant within the fibrous stroma of PCs.
Pseudocapsules were shown to be a rim of compressed fibrous tissue at the tumour-parenchymal margin. Through special staining with Tri-chrome, they demonstrated consistently high collagen content in PCs across all tumour histological subtypes. 24 The high collagen content is the reason why MPM is potentially a game-changing tool for providing insights into the characteristics and biological behaviour of PCs. For example, in addition to inter tumour variations, our study showed that the RCC PC has significant thickness and collagen area variation even within each tumour along the tumourparenchymal interface. Despite the differences in PC thickness, collagen fibre density and the complexity of linkages between the fibres (as measured by the reticulation index) are not significantly different between the thinnest and thickest areas of PC for each tumour. Intriguingly, this suggest that the site of any PC invasion by its tumour is likely a function of the thickness alone and not 'structural' weakness.
This is the first study to employ the stain-free MPM technology to analyse collagen characteristics of RCC pseudocapsules. Our data suggests that each RCC PC is not a consistent layer when fine collagen characteristics are examined using the stain-free MPM technology. Even with a truly complete and intact PC, we have shown that there are quantifiable and demonstrable intra-tumoral variation of collagen quantity and quality within each tumour PC. This has implications on the technique of Tumour Enucleation, which relies on the complete dissection along the fibrous PC. The PC, being the boundary between normal tissue and tumour cells, is a key factor during enucleation and surgeons should be aware of such intra-tumoral variation in PC thickness to achieve precise and complete tumour enucleation. Despite favourable published data, concerns over the oncological safety and positive margin rates of enucleation remain, mainly related to the integrity of the PC. 1,22 Indeed, intertumoral differences of collagen density between clear and non-clear cell cancers in our study suggest different pseudocapsular toughness when exposed to surgical dissection during enucleation.
This may be the reason why anecdotally, surgeons may have avoided enucleation for papillary cell cancers for fear of rupturing the weaker PCs, resulting in tumour spillage.
The main drawback of the study is the small numbers and the retrospective nature. A larger number over a longer study period would enable us to better correlate collagen characteristics with different tumour histological subtypes, grades and stages. Indeed, with increasing use of better imaging techniques, it is anticipated others will utilize MPM to corroborate and extend our initial findings.

Conclusions
In this study, we have demonstrated the ability of MPM to quantify collagen characteristics of pseudocapsules without fluorescent labelling. It is important for Urologists to be aware that even with a complete and intact pseudocapsule, it is not a homogenous structure, and varies in its thickness and its collagen characteristics within and between tumours. Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Competing interests: The authors declare that they have no competing interests.
Funding: National University Health System Innovation Seed Grant 2017 (intra-institution). There was no extra-institution or commercial funding.
Authors' contributions: YQ was involved in the data analysis and manuscript preparation. WK was involved in the planning of the study. LY and TT were involved the histological preparation of the samples. HY was involved in the planning of the study and manuscript preparation. All authors read and approved the final manuscript.