Investigating In Vivo Tumor Biomolecular Changes Following Radiation Therapy Using Raman Spectroscopy.

Treatment resistance is a major bottleneck in the success of cancer therapy. Early identification of the treatment response or lack thereof in patients can enable an earlier switch to alternative treatment strategies that can enhance response rates. Here, Raman spectroscopy was applied to monitor early tumor biomolecular changes in sensitive (UM-SCC-22B) and resistant (UM-SCC-47) head and neck tumor xenografts for the first time in in vivo murine tumor models in response to radiation therapy. We used a validated multivariate curve resolution-alternating least-squares (MCR-ALS) model to resolve complex multicomponent Raman spectra into individual pure spectra and their respective contributions. We observed a significant radiation-induced increase in the contributions of lipid-like species (p = 0.0291) in the radiation-sensitive UM-SCC-22B tumors at 48 h following radiation compared to the nonradiated baseline (prior to commencing treatment). We also observed an increase in the contribution of collagen-like species in the radiation-resistant UM-SCC-47 tumors at 24 h following radiation compared to the nonradiated baseline (p = 0.0125). In addition to the in vivo analysis, we performed ex vivo confocal Raman microscopic imaging of frozen sections derived from the same tumors. A comparison of all control and treated tumors revealed similar trends in the contributions of lipid-like and collagen-like species in both in vivo and ex vivo measurements; however, when evaluated as a function of time, longitudinal trends in the scores of collagen-like and lipid-like components were not consistent between the two data sets, likely due to sample numbers and differences in sampling depth at which information is obtained. Nevertheless, this study demonstrates the potential of fiber-based Raman spectroscopy for identifying early tumor microenvironmental changes in response to clinical doses of radiation therapy.

Evaluating differences in optical properties of indolent and aggressive murine breast tumors using quantitative diffuse reflectance spectroscopy.

We used diffuse reflectance spectroscopy to quantify tissue absorption and scattering-based parameters in similarly sized tumors derived from a panel of four isogenic murine breast cancer cell lines (4T1, 4T07, 168FARN, 67NR) that are each capable of accomplishing different steps of the invasion-metastasis cascade. We found lower tissue scattering, increased hemoglobin concentration, and lower vascular oxygenation in indolent 67NR tumors incapable of metastasis compared with aggressive 4T1 tumors capable of metastasis. Supervised learning statistical approaches were able to accurately differentiate between tumor groups and classify tumors according to their ability to accomplish each step of the invasion-metastasis cascade. We investigated whether the inhibition of metastasis-promoting genes in the highly metastatic 4T1 tumors resulted in measurable optical changes that made these tumors similar to the indolent 67NR tumors. These results demonstrate the potential of diffuse reflectance spectroscopy to noninvasively evaluate tumor biology and discriminate between indolent and aggressive tumors.

Raman spectroscopy reveals phenotype switches in breast cancer metastasis.

The accurate analytical characterization of metastatic phenotype at primary tumor diagnosis and its evolution with time are critical for controlling metastatic progression of cancer. Here, we report a label-free optical strategy using Raman spectroscopy and machine learning to identify distinct metastatic phenotypes observed in tumors formed by isogenic murine breast cancer cell lines of progressively increasing metastatic propensities. Methods: We employed the 4T1 isogenic panel of murine breast cancer cells to grow tumors of varying metastatic potential and acquired label-free spectra using a fiber probe-based portable Raman spectroscopy system. We used MCR-ALS and random forests classifiers to identify putative spectral markers and predict metastatic phenotype of tumors based on their optical spectra. We also used tumors derived from 4T1 cells silenced for the expression of TWIST, FOXC2 and CXCR3 genes to assess their metastatic phenotype based on their Raman spectra. Results: The MCR-ALS spectral decomposition showed consistent differences in the contribution of components that resembled collagen and lipids between the non-metastatic 67NR tumors and the metastatic tumors formed by FARN, 4T07, and 4T1 cells. Our Raman spectra-based random forest analysis provided evidence that machine learning models built on spectral data can allow the accurate identification of metastatic phenotype of independent test tumors. By silencing genes critical for metastasis in highly metastatic cell lines, we showed that the random forest classifiers provided predictions consistent with the observed phenotypic switch of the resultant tumors towards lower metastatic potential. Furthermore, the spectral assessment of lipid and collagen content of these tumors was consistent with the observed phenotypic switch. Conclusion: Overall, our findings indicate that Raman spectroscopy may offer a novel strategy to evaluate metastatic risk during primary tumor biopsies in clinical patients.

Raman Spectroscopy and Machine Learning Reveals Early Tumor Microenvironmental Changes Induced by Immunotherapy.

Cancer immunotherapy provides durable clinical benefit in only a small fraction of patients, and identifying these patients is difficult due to a lack of reliable biomarkers for prediction and evaluation of treatment response. Here, we demonstrate the first application of label-free Raman spectroscopy for elucidating biomolecular changes induced by anti-CTLA4 and anti-PD-L1 immune checkpoint inhibitors (ICI) in the tumor microenvironment (TME) of colorectal tumor xenografts. Multivariate curve resolution-alternating least squares (MCR-ALS) decomposition of Raman spectral datasets revealed early changes in lipid, nucleic acid, and collagen content following therapy. Support vector machine classifiers and random forests analysis provided excellent prediction accuracies for response to both ICIs and delineated spectral markers specific to each therapy, consistent with their differential mechanisms of action. Corroborated by proteomics analysis, our observation of biomolecular changes in the TME should catalyze detailed investigations for translating such markers and label-free Raman spectroscopy for clinical monitoring of immunotherapy response in cancer patients. SIGNIFICANCE: This study provides first-in-class evidence that optical spectroscopy allows sensitive detection of early changes in the biomolecular composition of tumors that predict response to immunotherapy with immune checkpoint inhibitors.