Peter A Humphrey, MD, PhD

I am a urologic pathologist with extensive experience and a large number of publications in prostate pathology and prostate cancer research. The focus of my research has mainly been on prostate cancer, with emphasis on diagnostic surgical pathology of the prostate, clinico-pathological investigations, and molecular biomarkers of clinical relevance.

The early phase of my research career, at Duke University, reflected my interest in molecular pathology and membrane-bound polypeptide growth factor receptors and was devoted to study of the epidermal growth factor (EGF) receptor in a brain tumor model system. We characterized the protein products of amplified and rearranged epidermal growth factor genes in gliomas, and defined a series of mutant proteins. As a junior faculty member at Duke I obtained extramural funding from several sources and concentrated on one specific deletion mutant epidermal growth factor receptor, which we named variant III (vIII). I designed and had synthesized a peptide that spanned the fusion junction of the mutant vIII, and demonstrated, as we reported in a PNAS paper, that antibodies reactive with this peptide bound specifically to the mutant and not wild-type EGF receptor.

My main clinical and research interests are in prostate cancer. After moving to Washington University, I explored the expression and functional significance of the polypeptide growth factor hepatocyte growth factor and its receptor, c-met in prostate cancer. We documented, in a series of papers, how overexpression of these two proteins is linked to aggressive prostate cancer. We specifically showed, for the first time, that c-met expression is associated with the high-stage prostate cancer and the development of androgen-independence.

My research efforts in surgical pathology of the prostate gland have aimed to refine the diagnosis of prostate cancer through definition of the diagnostic and clinico-pathologic features of structural variants, such as deceptively benign-appearing pseudohyperplastic, atrophic, microcystic, PIN-like, and foamy gland variants. Diagnostic awareness of these variants is critical in establishing a diagnosis of malignancy for these patients. The papers on the pseudohyperplastic, microcystic, and PIN-like (stratified epithelial) variants were the first in the literature. I have also sought quantitative prognostic attributes for prostate cancer, such as percentage of prostate chips with cancer, prostate cancer size in needle core and radical prostatectomy tissues, and length of prostate cancer surgical margin involvement. Additional clinico-pathologic investigations on prognostic factors have generated publications on quantification of the frequency and admixture of high histological Gleason grade patterns, that represent the most aggressive form of prostate cancer, and the prognostic significance of grade at the surgical margin. Another major research initiative has been the development and validation of tissue biomarkers in prostate cancer. These functionally significant biomarkers include hepsin, NKX3.1, prolactin receptor, cytoglobin, alpha-methylacyl-CoA racemase (AMACR), and ERG. Several of these biomarkers (hepsin, prolactin receptor, and cytoglobin) were discovered by our gene expression profiling and publications on them highlight potentially novel functional roles for them in prostate cancer. The published work on AMACR and ERG immunohistochemistry has defined their comparative diagnostic utility in foamy gland and minimal adenocarcinomas of the prostate. This is of significance since AMACR is one of the most widely used tissue biomarkers in diagnosis of prostate cancer.

Translational research efforts have also focused on testis and kidney cancer. A biomarker study that has had great impact on diagnostic urologic surgical pathology is the development of SALL4, a zinc finger transcription factor involved in stem cell character maintenance, as an excellent diagnostic marker for testicular germ cell tumors. It is now widely used as a diagnostic tool in surgical pathology. In kidney cancer research we have characterized the clinico-pathologic and genetic features of rhabdoid renal cell carcinoma, a particularly aggressive form of kidncy cancer.

Currently, my research is focused on definition and utilization of new technologies in the tissue diagnosis of prostate cancer, including multiphoton microscopy (in collaboration with Dr. Rick Torres) and artificial intelligence/machine learning.