Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Robert G. Cooks

Committee Chair

Robert G. Cooks

Committee Member 1

Peter T. Kissinger

Committee Member 2

Philip S. Low

Committee Member 3

Mary J. Wirth


My dissertation research has focused on the development of ambient ionization – mass spectrometry (MS) for clinical measurements, specifically intrasurgical cancer detection. The molecular differences between normal and cancerous tissue were detected via direct tissue analysis in vitro by touch spray ionization (TS) or by analyzing sectioned or smeared tissue using desorption electrospray ionization (DESI). The physical form of the tissue, e.g. in vitro sampling, sectioned, or smeared, was inconsequential in differentiating normal from cancerous tissue; however, the spectra acquired by TS and DESI differed due to differences in ionization processes. We envision that TS-MS and DESI-MS could impact diagnostic medicine, for example in providing surgeons with rapid, near real-time information as to tissue disease state, i.e. normal or tumor. Disease state information provided to surgeons about discrete pathologically ambiguous areas, ideally intrasurgically via TS or DESI-MS smear analysis, could improve the completeness of tumor resection while minimizing unintended damage to adjacent tissue.

Touch spray ionization was developed for intrasurgical detection of cancer; TS greatly simplifies MS analysis by using the same device for in vivo sampling and subsequent ionization. Frozen tissue sections were sampled and analyzed by TS-MS providing the ability to differentiate normal from human prostate cancer, via lipid profiles, using multivariate statistics. The next proof-of-concept step for TS-MS was the analysis human kidney cancer specimens in vitro, immediately following resection. TS-MS analysis of untreated kidney tissue emulated intrasurgical use, e.g. the presence and co-sampling of biofluids such as blood. Regardless, normal renal tissue and kidney cancer was differentiated using lipid profiles and multivariate statistics.

Desorption electrospray ionization (DESI) - MS imaging of tissue sections differentiated normal from tumor in all cancers studied. DESI-MS imaging of human prostate and human kidney tissue sections were performed to corroborate TS-MS results. Human brain cancer, a major focus of my dissertation research, was studied by imaging tissue sections using DESI-MS to establish the characteristic chemical features, e.g. lipid and metabolite profiles, that distinguish normal brain parenchyma from gliomas and different brain tumors. It was found that information in the negative ion mode lipid profile, positive ion mode lipid profile, and negative ion mode metabolite profile is able to discriminate brain parenchyma (grey and white matter) and gliomas, the most common form of malignant brain tumor. Further, the negative mode lipid and metabolite profiles also proved capable of discriminating different types of brain tumors (gliomas, meningiomas, and pituitary tumors) which account for ~80% of all central nervous system tumors. DESI-MS imaging of effaced or otherwise pathologically ambiguous frozen tissue sections offered the ability to determine the underlying brain parenchyma in cancerous samples – something that traditional morphologic evaluation was not able to determine. Further, DESI-MS was able to detect molecular changes resulting from varying amounts of glioma tumor cells present within infiltrated tissues. The tumor cell percentage of these samples was predicted using N-acetyl-aspartic acid, a neurometabolite which was found to decrease in cancerous tissue, and matched well with histopathologic evaluation.

The transition from DESI-MS imaging of sectioned tissue to DESI-MS analysis of tissue smears was driven by the time restriction of intrasurgical application. The potential of DESI-MS analysis of smears was first demonstrated upon canine non-Hodgkin’s lymphoma fine-needle aspirate smears which provided similar sensitivity and specificity values to that of tissue section imaging but is technically less demanding and decreased analysis time. DESI-MS imaging of tissue sections established that MS profiles contained the sufficient information for diagnosis; whereas DESI-MS analysis of tissue smears made the intrasurgical analysis of human brain tumors feasible. The observed lipid or metabolite profiles were not significantly altered by the physical act of smearing and their signal intensities were comparable to those of tissue sections. Further, the chemical information obtained from tissue smears was equivalent to those of tissue sections as determined by canonical component analysis. The culmination of my dissertation research was the creation and implementation of an intrasurgical DESI-MS tissue smear analysis method for human gliomas. Preliminary results from the initial intrasurgical cases analyzed using the developed DESI-MS method are discussed.