- Our Science
- Khan Website
- Home
- Publications
- Gallery
- Staff
- Links
- Faculties
- Curriculum Vitae
- POB Website
- Home
Our Science – Khan Website
Javed Khan, M.D.
 |
 |
|
||||||||||||||||||||||
Biography
Dr. Khan obtained his bachelor's degree in 1984 and his master's degrees in 1989 in immunology and parasitology at England's University of Cambridge. He subsequently obtained his M.D. there and the postgraduate degree of MRCP (Membership of the Royal College of Physicians), equivalent to board certification in the United States. After clinical training in internal medicine and pediatrics as well as other specialties, he received a Leukemia Research Fellowship. In May 2001, Dr. Khan joined the Pediatric Branch, NCI, as a tenure track investigator. Dr. Khan and colleagues have published a new model for diagnosis of cancer using artificial neural networks (ANN), a form of artificial intelligence, and microarray technology. In April 2001, Dr. Khan was recognized by the American Association for Cancer Research for his work in tumor profiling by receiving a Scholar in Training Award.
Research
Our chief interest is to apply genomic techniques in the investigation of pediatric malignancies and translate this to the clinic. We have demonstrated that cDNA microarray analysis can be used to identify genetic fingerprints of a specific type of muscle cancer, rhabdomyosarcoma (RMS), and is able to distinguish one type of cancer from another. Together with other colleagues we were the first to apply hierarchical clustering and visualization tools including multidimensional scaling to demonstrate the relationships between cancers based on gene expression profiling. We then went onto apply this technique to determine that the PAX3-FKHR fusion oncogene found in RMS activated a myogenic transcription program, which is a critical component to the oncogeneic process in these muscle cancers. Our team was also the first to apply artificial neural networks to diagnose pediatric cancers from DNA microarray gene expression profiles. Recently we have utilized the similar techniques to predict the outcome of patients with pediatric malignances including Wilm's tumors and high-risk neuroblastoma.
Molecular Taxonomy of Pediatric Xenografts and Cancers
Neuroblastomas are cancers of neural crest origin with variable prognoses depending on age at presentation, stage, histology, presence of MYCN amplification, chromosomal ploidy, and deletion status of 1p36. Very little is known of the molecular mechanisms that confer good or poor prognosis in this and other malignancies. We have recently demonstrated that cancers can be diagnosed on the basis of gene expression profiling using cDNA microarrays and sophisticated pattern recognition algorithms such as Artificial Neural Networks. The Oncogenomics Section has expanded this concept further by profiling a series on neuroblastomas of different stages and prognosis. With these methods we are identifying tumor-specific expression patterns, or 'fingerprints', that uniquely identify a poor prognostic group, as well as those associated with specific genetic aberrations including MYCN amplification. By these techniques, we hope to classify expression profiles that correlate with prognosis and hence identified the genes that confer these biological properties. Once we have narrowed down the list of genes that defines a particular cancer or diagnostic or prognostic group cluster to a minimum number, we will use this to make smaller microarrays or other multiplex PCR-based assays for diagnostic purposes in the clinic.
We also plan to develop a publicly available database of a wide range of pediatric malignancies from patient tumors, including archival material and xenografts, as well as prospective samples from patients on treatment protocols at the Pediatric Oncology Branch (POB). These studies will increase the knowledge of these malignancies by identifying genes that are significant to the biology of these cancers. In addition, it may identify secreted proteins that can be used for diagnosis (e.g., AFP for germ cell tumors), as well as following therapy including the monitoring of tumor regression or recurrence. We may also identify new targets for therapy, including immune therapy, or discover novel molecular targets such as death pathway genes, uniquely expressed in these cancers.
Proteomic analysis of high-risk neuroblastoma using isotope-coded affinity tags (ICAT) analysis
Isotope-coded affinity tags (ICAT), allows the quantitative measurement of protein expression levels in different cell types and tissues. In this method proteins from two samples can be compared by chemically labeling both samples with the light and heavy isotopic forms of a reagent respectively. With this method we plan to sequence and identify up to 3000-4000 differentially expressed proteins between tumors with poor (death) and good (event free survival > 3yrs) outcome. These proteins represent potential targets for therapy, diagnostic and prognostic markers for high-risk patients as well as provide important clues on the biology of these tumors that fail to respond to conventional therapy.
Comparative Genomic Hybridization of Neuroblastoma
Recurrent non-random genomic alterations are the hallmarks of cancer and the characterization of these imbalances is critical to our understanding of tumorigenesis and cancer progression. We are performing Array-comparative genomic hybridization (A-CGH) on cDNA microarrays containing 42,000 elements in neuroblastoma. We are developing a novel probabilistic algorithm, called topological statistics, to increase the sensitivity of cDNA A-CGH for detecting single-copy alterations. Our method not only overcomes the shortcoming of relative low sensitivity of cDNA A-CGH but also enables a direct visualization of the statistical confidence for the observed alterations. Furthermore using probabilistic approaches and machine learning algorithms (ANNs and Support Vector Machines) to estimate the frequency of genomic imbalances in tumors of different stage and MYCN amplification status, we are attempting to identify unique patterns of genomic imbalance in all three categories. Using these approaches we identifying models of the evolution and progression of neuroblastoma.
Gene Expression Profiling of Normal Tissues
The non-specific nature of current cancer therapeutics, the lack of response in high-risk disease, and the severe, sometimes fatal, organ toxicities have driven cancer therapeutics towards rationally designed and specific molecularly targeted therapy. It is hoped that targeting genes expressed only in cancer, and not in normal tissue will minimize many of the side effects of therapy. To identify genes that are expressed only in cancer, it is of paramount importance to profile normal pediatric organ tissues. These unique tumor antigens can be used in the development of vaccine-based immune therapies for childhood cancer (a primary interest of Drs Helman and Mackall, POB), and will reduce the risk of serious autoimmune manifestations. To achieve this aim, we will perform gene expression analysis on approximately 200 normal pediatric tissue samples obtained from the Maryland Brain and Tissue Bank (http://medschool.umaryland.edu/btbank/). These samples include tissues from cerebrum, cerebellum, heart, lung, liver, pancreas, stomach, small intestine, large intestine, kidney, spleen, skeletal muscle, ovary, testes, ureter, uterus, bladder, adrenal gland, and prostate. The median age of the subjects from which the samples were obtained is 19yrs (range 2-40yrs), with a median post mortem interval (PMI) of 11hrs (range 4-19hrs). Information from the Bank suggests that the RNAs obtained from tissues within these PMIs are of good quality as judged by integrity of the ribosomal bands. The data generated by gene expression profiling will be collated and compared with the data generated from the malignant tissues. After publication, we will make the data available to outside investigators for searching via a web-based format.
Molecular Mechanisms of Drugs Using cDNA Microarrays
Another area, which we will focus on, is the monitoring of gene expression changes impacted by drugs. The choice and design of many chemotherapeutic agents currently used in cancer treatment has been traditionally empirical in nature. The molecular mechanisms of their actions are not well understood, and their mode of action is often indiscriminate targeting, both of tumor and normal cells. As a model system, we have investigated the gene expression alterations during neural differentiation of the neuroblastoma (NB) cell line SMS-KCNR by retinoic acid (RA) as well as Fenretinide (4-HPR) using cDNA microarrays. Neuroblastoma is the most frequently occurring extracranial solid tumor of childhood and has the highest rate of spontaneous regression of any human cancer. RA is known to stimulate morphological neural differentiation of NB. It has been shown to enhance neurite extension, increase membrane excitability, induce neurotransmitter enzymes, and reduce tumorogenicity, as well as improve prognosis for high-stage disease. Neuroblastoma will be treated with all-trans retinoic acid (ATRA), or 4-HPR or the solvent ethanol (control) and gene expression profiles will be performed at different time points. By this method we will identify pathways of critical genes involved in neuronal differentiation and apoptosis.
The combined approaches outlined in this proposed program will allow a comprehensive analysis of pediatric tumor genomes.
Collaborating with us are Malcolm Smith, Chand Khanna, Glenn Merlino, Carol Thiele, Jon Wigginton, NIH, Tim Veenstra, SAIC, Kathy Pritchard-Jones, Institute of Cancer Research/Royal Marsden NHS Trust, United Kingdom and the United Kingdom Childhood Cancer Study Group (UKCCSG).
This page was last updated on 9/11/2008.
