Date of Award
Doctor of Philosophy (PhD)
Cancer and Developmental Biology
Michael A. Dyer, Ph.D.
R. Kiplin Guy, Ph.D. Suzanne J. Baker, Ph.D. Dianna A. Johnson, Ph.D. Clinton F. Stewart, Ph.D.
Rodent models play an essential role in the development of new chemotherapeutics and dosing regimes. It is often difficult to carryout a clinical study for pediatric cancers due to the small patient population. Retinoblastoma, a pediatric cancer of the eye, is one example of a pediatric cancer that can benefit from preclinical studies. Over the years various retinoblastoma rodent models have been developed used to test various combination of broad-spectrum systemic chemotherapy. It was found form these studies that the combination of topotecan and carboplatin was effective. However both drugs cause myelosuppression and therefore administrating both of these drugs systemically is not possible. An alternative effective therapy in the clinic was the use of a subconjunctival administration. We thought if we could administer both drugs, one by systemic and one by a subconjunctival injection, perhaps we could decrease the systemic exposure with good tumor response. Detailed pharmacokinetic studies were conducted to understand the subconjunctival injections of topotecan and carboplatin. It was found that both drugs could successfully penetrate the eye and increase drug exposure. In addition, in the presence of a tumor, drug exposure to the vitreous was greater.
Additionally comparative pharmacodynamic studies combining topotecan subconjunctival injection with carboplatin intraperitoneal or carboplatin subconjunctival injection with topotecan intraperitoneal were conducted. The tumor response, systemic toxicity and local toxicity were studied. There was tumor response in both combinations and no ocular toxicity was seen with a single eye subconjunctival injection for either drug. However, rats that received the combination with topotecan subconjunctival injection and carboplatin intraperitoneal experienced great toxicity and morbidity. The data and observations suggest the death is due to dehydration. Therefore it was concluded that the alternative combination was better.
The above data suggested an appropriate drug combination and schedule for a preclinical study. However, the noninvasive methods to follow tumor progression and choosing the correct genetic model needed to be determined. This was essential to ensure the preclinical study could be easily translated for future clinical studies. A characterization study of five modalities, retina camera, optomotry, tonometer, ultrasound and MRI, was done with retinoblastoma mice. We determined the feasibility of each technique. It was found that the retina camera could detect the tumor the earliest in a high throughput manner. Additionally, the tonometer and optomotry machines could assess ocular health. While the ultrasound and MRI could image the eye and tumor in one field of view, MRI could capture the posterior chamber in more detail along with the extraocular space. With different software programs, the tumor to eye ratio volume measurement were determined and compared to the gold standard of enucleation, embedding, serial sectioning and hand tracing. It was found that there was a better correlation between the ultrasound and hand tracing histological sections.
Concurrently, the tumor progression of six different genotypes was assessed. The tumor progression depended on the number and different genes deleted. Additionally, based on genotypes, it was determined there was not a strong genotypic trend in the increase in IOP or the loss of vision. From the studies of tumor progression we have learn more about the influence of genes on tumor progression, which will benefit additional genetic studies in mouse model systems and human tumors.
Nemeth, Katie Marie , "The Use of Preclinical Models to Improve the Treatment of Retinoblastoma" (2010). Theses and Dissertations (ETD). Paper 175. http://dx.doi.org/10.21007/etd.cghs.2010.0225.