Icon Legend

This session is not in your schedule.

This session is in your schedule. Click again to remove it.

Biomedical Imaging and Instrumentation

Biomedical Imaging and Instrumentation - Poster Session D

Poster G3 - Quantifying collagen changes in immunotherapy-treated mouse colon tumors using in vivo second harmonic generation imaging

Friday, October 25, 2024

3:30 PM - 4:30 PM EST

Location: Exhibit Hall E, F & G

Presenting Author(s)

Samantha K. Burkard (she/her/hers)

Undergraduate Research Assistant
University of Wisconsin - Madison and Morgridge Institute for Research
Madison, Wisconsin, United States

Co-Author(s)

Alexa R. Heaton, PHD (she/her/hers)

Assistant Scientist
Morgridge Institute for Research and University of Wisconsin - Madison, United States

Primary Investigator(s)

MS

Melissa C. Skala, PhD

Professor
Morgridge Institute for Research and University of Wisconsin - Madison, United States

Introduction: Collagen is an important structural protein, a major component of the extracellular matrix (ECM), and a primary interest of cancer research. Previous work in breast cancer and melanoma has identified changes in collagen morphology, referred to as Tumor-Associated Collagen Signatures (TACS), which can provide prognostic value. Namely, during tumorigenesis, collagen fibers often straighten and become more aligned, alongside an increase in collagen density. Other work has also found a correlation between cancer treatment and accelerated fiber shortening. This remodeling of tumor collagen fibers is thought to promote cancer cell invasion and restrict immune cell infiltration into the tumor. One method to track these changes in tumor collagen fibers is second harmonic generation (SHG) microscopy, an imaging technique that does not require labels or dyes. The resulting collagen images can be segmented and analyzed to quantify collagen morphology changes. Here, we used open-source CurveAlign and its package CT-FIRE to capture collagen morphology metrics at an image level and single fiber level. We compare collagen morphology from in vivo SHG images of immunotherapy or vehicle treated mouse colon cancer tumors and ask whether collagen changes are correlated with treatment.

Materials and

Methods: MC38 tumor-bearing mice from vehicle (PBS) or treated (radiation therapy, IL-2, and αCTLA-4) groups were imaged in vivo on days zero (prior to treatment), six, and nine using SHG microscopy. Pretreatment (n = 31), vehicle (n = 50), and treated (n = 50) images from three independent experiments were analyzed using CurveAlign and CT-FIRE software. CurveAlign was used to quantify the coefficient of alignment (0 to 1, where 1 is most aligned, calculated by doubling each angle, then dividing the length of the vector sum by the sum of individual vector lengths) and density (measured by finding the proportion of pixels over a brightness threshold of 35, chosen via ROI analysis) of each field of view (FOV). CT-FIRE was used to quantify single fiber length and width (number of pixels) as well as straightness (score from 0 to 1 where 1 is straightest, calculated as the distance between end points divided by fiber length,) and angle (measured in degrees relative to the horizontal). Lastly, each FOV was also qualitatively scored using a curliness ranking system: 1 – straight, 2 – wavy, 3 – curly, similar to the process that pathologists use to score cancer histology images. Ordinary one-way ANOVA was performed to calculate statistical significance across groups.

Results, Conclusions, and Discussions: We successfully acquired in vivo MC38 tumor images at multiple time points during therapy using SHG imaging (Fig. 1A). We found that on day 6, vehicle and treated tumors exhibited similar collagen morphology across all metrics except width and density. Strikingly, treated tumors exhibited significantly wider collagen fibers on day 6 and day 9 (Fig. 1B). Treated tumor collagen was also found to be denser on day 6 only. Our qualitative curliness scoring, integrated across days 6 and 9 for vehicle and treated groups, showed that vehicle tumors had more straight fibers (52% score 1) while treated tumors had more wavy and curly (88% score 2 and 3). Both groups exhibited heterogeneity in their collagen phenotype (Fig. 1C). Surprisingly, we did not find treated tumor collagen fibers to be significantly longer or less straight.

The increased width observed in treated mouse tumors is consistent with the expectation that less taut, more curly collagen fibers will be wider – as collagen reorganizes towards a healthier phenotype. As treated collagen fibers became wider, it was surprising that length and straightness did not significantly change as these metrics may also be affected as collagen fibers become less taut. Perhaps these nonsignificant changes in collagen length and straightness are due to heterogeneity across these MC38 tumors, possibly attributed to their fast growth dynamics which may allow less time for collagen reorganization. The increased density observed on day 6 of treatment is consistent with our previous melanoma studies, attributed to the day 0 radiation therapy which may promote an influx of fibroblasts, increasing collagen deposition. The qualitative curliness scoring illustrated tumor heterogeneity across vehicle and treated mouse tumors, with a clear increase in curliness in the treated group only.

CurveAlign and CT-FIRE are valuable tools to quantify collagen morphology changes across treatment groups. Here we have shown that radio-immunotherapy treated colon tumors experience collagen reorganization, with width and density being the most correlated to treatment. With continued investigation, collagen morphology metrics may have clinical significance as a biomarker in the evaluation of patient treatment decisions.

Acknowledgements (Optional):