NICHD/Oncofertility Ovarian Tissue Slide Image Databank

In 2007, the Oncofertility Consortium (OC) was established with the support of an NICHD grant to address fertility concerns for children and patients of reproductive age diagnosed with cancer. The OC National Physicians Cooperative allowed for the collection of a portion of the ovarian tissue to be transferred to Northwestern University for research. As non-pathologic ovarian tissue has never been evaluated in large numbers due to the invasive nature of obtaining it, the tissue collected by the OC will allow for deeper evaluation of normal ovarian histology and function in young populations. The PAG program has partnered with Northwestern University/Oncofertility Consortium to create a digital ovarian tissue slide image database of ovarian tissue. The database has expanded to include other sites since that time including Children’s National, University of Pittsburgh, NIH and Washington University. Histological slides were prepared from fresh tissue blocks or cryopreserved ovarian tissue. Each slide typically contained 4-6 sections of tissue obtained from a single patient. Tissue sections were stained using Hematoxylin and Eosin (H&E) stain to enhance tissue contrast and facilitate subsequent image analysis. A total of 1857 initial slides were collected and subsequently scanned using a digital scanner at a 40x magnification level, resulting in an image resolution of 0.251 micrometers per pixel. Annotations of the ovarian follicles within the HALO image bank can be conducted using Aperio ImageScope software by trained members of the lab.  The primary objective of this study is to create a digital slide database with the goal of fostering collaborative research by making the slides accessible to scientists intramurally and extramurally both nationally and internationally*. We recommend that when evaluating follicle density and classification the NICHD Ovarian Nomenclature Follicle Workgroup recommendations are followed.  In addition, through collaboration with the NCI Artificial Intelligence Core we are developing tools to digitally evaluate ovarian follicles and histology in this tissue. This histologic digital database will be a useful resource to determine normal histology in pediatric populations and serve as control tissue for histologic normative data.**

Investigators interested in access to the image bank for research should contact Dr. Gomez-Lobo at veronica.gomez-lobo@nih.gov.

*A one-page proposal should be submitted including Title of study, significance, aim/hypothesis, methods, patient population to be studied, methods including number of samples needed, funding Source and NIH Biosketch or CV

Please Note: Any scientific products created as a result must attribute the grant numbers: NIH/NICHD U54HD076188 and the Oncofertility Consortium and NICHD Z1A HD009006

**There are inherent limitations to evaluating ovarian histology from a small portion of said ovary:

  1. How the tissue was fixed can affect the appearance of follicles. Tissues in this image bank have been fixed by labs or pathology departments using different protocols and fixatives. The details regarding these are not available.
     
  2. Area for follicle density calculation has not been clearly defined. The Follicle Workgroup of the Ovarian Nomenclature Workshop suggests using a depth of 1 mm from the surface epithelium for density calculations.
     
  3. Often samples of ovary are not clearly oriented such that the epithelium cannot identified and thus density measurement is limited. 
     
  4. In this image bank only small portions are fixed and evaluated resulting in some instances in only 1 section or 1 slide available. Furthermore, follicles are not evenly distributed within the ovary and are found in clusters that can vary by 100-fold. Thus, extrapolation from these small samples may likely not represent the true follicle density in an individual. 
     
  5. There is limited demographic/health data associated with these samples.
top of pageBACK TO TOP