1a. The Administrative and Quality Control Office (AQCO)

The Administrative and Quality Control Office (AQCO) is designed to minimize redundancies and allow streamlined workflow that makes using the NCDMD an efficient, cost-effective, convenient and error-free experience for investigators. To make the most efficient use of resources and personnel, the AQCO is administered through the Gene Therapy Center of the UNC School of Medicine. The AQCO provides administrative oversight and support and ensures high quality results and documentation in accordance with Good Laboratory Practices (GLP) to facilitate translation to clinical care. This bilateral approach allows workflow tracking from the point of initial project submission to finished data and provides a communication rich environment, facilitating growth and development of the NCDMD.

The NCDMD is designed to provide an ideal evaluation package and reflects the current perception of “best practices” in the field. It is anticipated that the definition of “best practices” will continue to be refined and developed. Anticipating this progression, the AQCO will provide the complex coordination, quality measures, and consistent leadership necessary to adapt to increased demand and the evolving key measures and predictive values in the field.

1a(1). Administrative Plan.
Oversight: Directed by Dr. Jeff Beecham, the AQCO is a consistent, efficient service facility that provides workflow tracking; facilitates communication between investigators and service facilities; monitors cost effectiveness and the fee-for-service model; education and training; and general administrative oversight.

Project Coordination: The AQCO Project Coordinator provides accounting/ budget management (fee-for service invoicing, supplies, procurement of capital equipment, travel, etc.) to ensure accurate expense reporting and proper maintenance of the fee-for-service model (see below). The Coordinator provides specific administrative support to assure effective communication and coordination at the inter- and intra-institutional levels. Intra-institutional communication is supported via secure server access. The coordinator serves as the communication liaison – tracking and storing all NCDMD communication and acts as Secretary for the Steering Committee.

Program Management: The Program Manager is charged with supporting the growth and development of the NCDMD through: organization of Steering Committee meetings (described below); regular assessment of investigator satisfaction, providing a summary of recommendations to the Steering Committee; progress reporting; maintenance of the Data Sharing Plan; organization of an educational seminar series; and liaison with the External Quality Assurance consultant.

Fee-for-Service Model for NCDMD Projects: An overarching goal of the NCDMD is to develop and make available DMD dog models for studies including high-quality, standardized cGLP testing to support translation of new interventions to the clinic. Services are provided to investigators at a subsidized rate. The U-24 grant provides support for key infrastructure of the NCDMD (“fixed costs”), including personnel and equipment of the core facilities and the animal breeding colony. Fees paid by investigators are expected to cover costs associated with the projects themselves (“variable costs”), to include dogs entered onto projects and materials used in the core facilities. The fee-for-service model is monitored and directed to qualifying projects by the Steering Committee through the program solicitation process.

Data/Materials Transfer Information for NCDMD Projects: Information generated through NIH-funded research activities at the NCDMD will be disseminated through timely peer-reviewed publications, seminars and presentations at national scientific meetings. Improved resources generated at the Canine Muscular Dystrophy Facility (CMDF) and other facilities will themselves be made available at the time of public disclosure to NCDMD collaborators as well as qualified investigators within the scientific community. These materials will be used for internal research purpose only and will be disseminated in accordance with the NIH Grants Policy statement and the Principles and Guidelines for Recipients of NIH Research Grants and Contracts on obtaining and disseminating biomedical research resources. Final Notice: December 1999.

NCDMD Performance Evaluation: Is the Center performing well? In addition to regular review by the Operations Committee and the Steering Committee (described below), an evaluation tool will be provided to assess investigator satisfaction and allow open field text entry to identify problem areas in unique projects. Results from the surveys and recommendations will be presented to the Steering Committee to assure that the needs of investigators are being met.

1a(2). Quality Plan.
Oversight: The AQCO Director, Dr. Beecham, will ensure success of the NCDMD quality control program by providing GLP training, updates and supervision of service facility personnel on a Quality Assurance level. This level of oversight will avoid conflicting priorities and allow a clear process for quality reporting. Dr. Beecham reports directly to Drs. Kornegay and Samulski.

Monitoring: The Quality Assurance Coordinator ensures that standardized GLP reporting and documentation are recorded across service facilities. The QA Officer maintains control over master records and files and controls the security and maintenance of an electronic database of quality control/ quality assurance records. This includes controlled documents such as SOPs and data log templates, training records, raw material and equipment records, Drug Product Files and other controlled and/or archived documents. All facilities, operations and product files are reviewed by the QA Officer for completeness and accuracy. The QA Officer also performs internal quality assurance audits and liaisons with companies to perform audits of vendors. The QA Officer has independent authority and reports to the AQCO Director (Dr. Beecham) and Steering Committee.

The External Quality Assurance Consultant will provide independent, external QA oversight through periodic audits for the NCDMD and will be responsible for assuring that testing and facility functions comply with regulatory requirements of the FDA and similar agencies. The consultant will also provide regulatory advice, review drug master files for the NCDMD and products and facilitate and coordinate interactions with external agencies such as the FDA. The consultant will also liaison with the AQCO through the program manager to provide regulatory guidance and assistance involving interactions with investigators and the service facilities. This may include consulting with the Investigators to advise them in developing pre-IND applications. The QA consultant will also be active in Steering Committee meetings as a non-voting member.

1b. The Steering Committee
The NCDMD Steering Committee oversees management of the NCDMD, providing expert attention and an objective and multidisciplinary perspective while engaging in a number of activities, including proposal solicitation. The Steering Committee will establish stringent selection criteria to identify researchers from around the country for collaborative research projects that involve testing of promising candidates in dog models of DMD. We have been contacted by a number of investigators who wish to utilize GRMD dogs or one of the other canine models in preclinical studies. The NCDMD Steering Committee will determine which therapeutic strategies are pursued. As discussed above, it is anticipated that the definition of “best practices” will continue to be refined and developed. Anticipating this progression, the Steering Committee, in close coordination with the AQCO, will play a key role in providing the complex coordination, quality measures, and consistent leadership necessary to adapt to increased demand as new treatment strategies and biomarkers are developed.

1b(1). Constituency of the Steering Committee. Voting members (8) of the NCDMD Steering Committee will include the Committee Chair (Director of the NCDMD – Dr. Joe Kornegay); the Director of the UNC-CH Gene Therapy Center (Dr. Jude Samulski); the Director of the UNC Francis Owen Blood Research Laboratory (Dr. Tim Nichols); representatives from the joint NIH sponsors of the NCDMD, NINDS (Dr. John Porter) and NIAMS (Dr. Glen Nuckolls) who together have one vote; Dr. Richard (Dick) Moxley, a clinical neurologist from the University of Rochester, who cares for DMD patients; Mr. Charles (Chuck) Riesebeck, a parent of a boy with DMD; Dr. Mark Haskins from the University of Pennsylvania who has worked extensively with large animal models of human genetic diseases; and Dr. Paul Muhlrad, Research Program Coordinator, the Muscular Dystrophy Association. The QC consultant described in the Administrative and Quality Coordination Office (AQCO) will also be included as a non-voting member to provide expertise with regulatory or compliance issues.
1b(2). Steering Committee Operations. The Steering Committee will meet twice yearly at UNC-CH to conduct the above business. A major function will relate to whether interventions advance to the next stage of therapeutic development (see Selection of Projects for the NCDMD below). Given the importance of product advancement, a clear consensus will be sought. Five of the voting members must agree that the intervention should advance to the next stage or it will be terminated. The collaborating investigator will have the opportunity to appeal an adverse decision. Reversal would require approval by five members. In some cases, to expedite these or other time-sensitive matters, Steering Committee meetings may be held via conference call or video conference.

1c. Operations Committee
Day-to-day operations of the NCDMD will be managed through an Operations Committee composed of the NCDMD Director (Dr. Joe Kornegay), the CMDF (below) administrator (Ms. Janet Bogan), the Physiology, Histology/Molecular, and Imaging Services Facilities (see cores below) Directors or their designees, and the Quality Assurance Coordinator from the Administrative and Quality Coordination Office (AQCO). The Operations Committee will work in tandem with the Steering Committee to set policies of the NCDMD and its cores, to include development of needed procedures and fees to be charged. Most importantly, the Operations Committee will meet monthly to review progress of ongoing projects that are being conducted through the NCDMD.

1d. The Canine Muscular Dystrophy Facility (CMDF).
The CMDF is a key component of a new UNC-CH off-campus large animal resource and research facility (RRF) underdevelopment in Orange County 14 miles from the main campus.

1d(1). CMDF Specific Aims.
A). Monitor estrus and breed carriers so as to perpetuate these canine models.
B). Work with the Department of Laboratory Animal Medicine (below) to provide specialized care for dystrophic dogs.
C). Provide support to the core facilities (below) to include anesthesia for surgical biopsies and imaging.

1d(2). CMDF History and Qualifications. The CMDF is administered by Ms. Janet Bogan. Ms. Bogan has a bachelor’s degree in biochemistry and is a Registered Lab Animal Technologist and a Certified Manager of Animal Resources. She has worked extensively with dogs in this colony since 1989. Ms. Bogan has particular expertise in breeding management and neonatal and critical care. Her duties also include assisting with oversight of all proposed studies and supervision of personnel. Ms. Bogan will assist with training other personnel in this project on how to conduct experiments utilizing the canine model.

1d(3). CMDF Explanation of Services. As described below, in addition to managing the overall colony and breeding program, CMDF personnel support individual NCDMD projects and cores. Veterinary technicians anesthetize dogs for surgical biopsies and imaging techniques and also assist with these and other specialized procedures. The fees charged to NCMD investigators for procedures performed through the CMDF or with the assistance of CMDF personnel will be substantially subsidized because personnel and equipment costs will be covered through the U-24 funding mechanism. For sake of comparison, dog per diem and fees charged for representative procedures are presented in Table 1.

1d(4). CMDF Animal and Veterinary Care. The dogs are housed at the Canine Muscular Dystrophy Facility (CMDF). Care for animals is provided by the UNC-CH Department of Laboratory Animal Medicine (DLAM). Activities of the CMDF are supervised by appropriate university (IACUC), professional (AAALAC), and governmental (USDA) regulatory agencies. Separate Institutional (IACUC) protocols are maintained for the Standard Operating Procedures of the colony and for individual projects in which dogs are involved. The overall colony SOP defines policies that govern breeding (see separate summary below), pregnancy determination and parturition, and animal care, including the critical postnatal period when weight gain is an important predictor of pup survival (Table 2). Discomfort, distress, pain, and injury to the dogs are avoided or minimized, consistent with sound research design. No more dogs are used than is necessary to complete the specific aims of proposals.

1d(4a). Breeding and Neonatal Care. Conception rate and gestation in GRMD and other canine models are not impaired. Breeding affected males to obligate carriers produces an expected 1:1 ratio of affected to unaffected dogs (25% affected males, 25% affected homozygous females, 25% obligate female carriers, and 25% normal males). Phenotypic data from homozygous females and affected males do not differ so these dogs are used interchangeably on studies.¹ The mean litter size is 7 pups, with a range of 1-13. This compares favorably with the average litter size of 7.6 seen in golden retrievers.²

GRMD carriers cycle in estrus on an average 10-month interval. They are checked for signs of proestrus Monday, Wednesday, and Friday. At the first sign of proestrus, we begin measuring serum luteinizing hormone (LH) and progesterone levels. When LH is positive, ovulation has occurred. We initiate breeding every other day by AI when progesterone levels are > 3 ng/ml and vaginal cytology shows > 60% cornified epithelial cells. Breeding is continued until progesterone levels are > 10 ng/ml and white blood cells are seen on vaginal cytology, indicating the onset of diestrus. We evaluate ultrasound and measure relaxin levels 30 days from the onset of diestrus. If confirmed pregnant, the whelping date is calculated to be 27 days later. Body temperature is measured beginning a few days before this date to more precisely predict whelping. With this breeding protocol, conception rates are approximately 85%.

Carriers are not removed from the colony at a set age. Rather, we remove them based on the regularity of their estrus cycle, rate of pregnancy, litter size, and pup survival. With these factors in mind, most carriers continue to be bred until they are 5 to 7 years of age.

All bitches are observed at the anticipated time of whelping to ensure that appropriate neonatal care can be provided. Affected pups are weak at birth and require nutritional supplementation for up to 4 weeks. We closely monitor pups during the first 24 hours of life to ensure that colostrum is ingested. Colostrum from other dogs is harvested and frozen so that it can be used as a supplement for weaker pups that are pushed aside by more vigorous littermates and/or do not show satisfactory weight gain. Pups are weighed soon after birth and, on average, four times daily for the first 2 weeks (Table 2). Each pup's physical response to outside stimuli and hydration level are observed at weighing intervals. Newborn pups are evaluated on their overall responsiveness and strength. Righting response, whereby a pup is placed on its back and allowed to right itself, in addition to rooting, latching on to teats and suckling ability of the pup when placed near the dam are observed. Any evidence of weakness of these responses or failure to thrive is noted. This would include the aforementioned physical responses and any evidence of respiratory distress, abdominal distension and/or significant weight loss (i.e. sustained lack of weight gain or loss of 20 g. over a 24-hour period).

In our hands, survival of GRMD dogs is typically 70-80 percent. Adequate numbers of dogs can be obtained for long term comparative studies.^3-9 Currently, an average of 2.25 affected pups from each litter survive beyond 10 days of age. To provide an approximate number of affected dogs that can be produced each year by each carrier, the following formula can be used: Affected pups/carrier each year = 2.25 (Surviving Pups/Litter) X 0.83 (10 mos/12 mos; estrus cycles/yr) X 0.85 (conception rate). Using this formula, each carrier produces only about 1.58 pups/annually. Thus, to produce 15 affected dogs each year, one must maintain 10 carriers. With a combined per diem for routine and specialized care of ~ $9.00, the cost to maintain each carrier is ~ $3,285/year. Without the NCDMD subsidy, this cost must be borne by investigators, over and above the cost of the study dogs themselves. With funding from the U-24, the cost of the breeding colony is borne by the NCDMD.

To determine the effect of inbreeding and postnatal care on pup mortality and, thus, guide our overall colony management, we have systematically analyzed data from 66 litters with a total of 459 pups born over a ten-year period (Table 3). All dogs in the colony originated from a single founder male. Nine additional founding dogs (3 males, 7 females) were added to the colony over the course of the ten years reviewed. These dogs were assumed to be unrelated to the existing population and to each other since the exact relationships of these individuals were not known. The inbreeding ranged from mating of cousins to continual line breeding (sire x daughters/granddaughters). All litters were produced from three different types of matings:

1. normal females x affected males,
2. carrier females x normal males, and
3. carrier females x affected males.

Demographic data collected from each litter included inbreeding values for parents and offspring, initial number of offspring, fraction of offspring surviving until 14 days of age, sex and muscular dystrophy genotype, age of dam, parity (primiparous or multiparous dams), pup weights, pup condition and postnatal care. The muscular dystrophy genotype of each individual was determined by elevated serum creatine kinase values or by RFLP analysis. Pup morality was defined as any death within the first 14 days of life. Inbreeding coefficients were calculated using the computer program CompuPed (Man’s Best Friend Software, Franklin, IL) based on Wright's formula. Models of the relationship of mortality and percent inbreeding were created using logistic regression. Hosmer-Lemeshow goodness of fit tests for the logistical models were done. There was evidence of a positive relationship between inbreeding and mortality rate (p = 0.6950, where p > 0.05 indicates a good fit). Mortality increased with increased inbreeding (Table 3) for the overall colony, including normal, carrier, and affected dogs. A total of 375 dogs (81.7 %) had some calculated degree of inbreeding, and 81 (21.6%) of these died within the 14-day neonatal period. The coefficient for the indicator variable “affected” was 2.6458. This value must be exponentiated for interpretation. We found e2.6458 = 14.094 ('Odds Ratio'), implying that the odds of dying, at any level of percent inbreeding, are about 14 times greater for affected pups than for normal/carrier pups. Based on these data, we introduce new breeding stock unrelated to the colony, on average, at five-year intervals. We continue to evaluate this strategy and plan to begin introducing new breeding stock at three year intervals.

We have also compared pup mortality according to the level of postnatal care, classified as none, partial or complete. No intervention was defined as monitoring weights only. Partial intervention included weighing pups and supplementing those experiencing difficulty with bitch's milk, milk replacer, or both using gastric tubes or bottles. Complete intervention involved removing pups experiencing difficulty from the dam entirely and feeding as with the partial group. Litters were grouped according to their level of postnatal care, and mortality data were compared for evidence of increased pup survival in any given group. Pups with partial postnatal care tended to do better than those with either no or complete care.

In addition, we have assessed whether the GRMD trait contributes to fetal mortality by comparing the expected genotypic ratios with the actual ratios of pups born in each litter. An overall evaluation of these data was made in relation to percent inbreeding and mortality rate, as well as an individual comparison by genotype. The Chi-square statistic for each litter was plotted against the percent inbreeding and the plot seemed random with no obvious pattern. Consistent with the plot, the correlation coefficient (specifically Spearman's Rho) was not significantly different from 0. This suggests that inbreeding does not affect prenatal mortality in our colony.

Parity (primiparous or multiparous) did not affect pup survival.

Numerous pups were submitted for necropsy to determine cause of death. No consistent pathological findings were identified.

1d(5). CMDF Organizational Structure. The CMDF is a key component of the NCDMD, providing animal housing and personnel to meet investigator needs. The CMDF Administrator, Ms. Janet Bogan, serves on the NCDMD Operations Committee (see above), thus ensuring that activities of the CMDF are coordinated with the cores. Production goals of the CMDF and project milestones are reviewed at the monthly Operations meetings.

1e. The Service Facilities.
Core facilities, including the Physiology Testing Facility (PTF), Histology & Molecular Services Facility (H&MSF) and Large Animal Imaging Facility (LAIF), provide specialized support services for research projects using the dogs bred/maintained at the CMDF. These facilities are analogous to funded cores that support researchers who utilize the mdx mouse:

1. University of Pennsylvania Wellstone Center, Lee Sweeney, Director, U54-AR052646, Core B: Physiological Assessment. This research core performs ex vivo, in situ, and whole-animal assessments of muscle integrity and function; and
2. Children’s National Medical Center, Washington DC, Kanneboyina Nagaraju and Eric Hoffman, Directors, Department of Defense, Mouse Functional Testing Core Facility. Histology, histochemistry, blood chemistry, imaging, and various physiologic measures are performed.

These existing funded mouse cores provide evidence that infrastructure centers help standardize and improve the quality of data obtained during therapeutic development and provide a nice parallel for our Center. Indeed, the NCDMD cores serve a particularly important function, as relatively few investigators have expertise using large animal models.

1e(1) HISTOLOGY AND MOLECULAR SERVICES FACILITY (H&MSF)
1e(1a) H&MSF Specific Aims.

1. Serve as a centralized, on-site resource to facilitate performance of morphologic, immunohistochemical, ultrastructural, and biodistribution studies needed by NCDMD investigators.
2. Provide services of highly-qualified staff and access to state-of-the-art instrumentation, technical assistance, instruction and consultation for NCDMD investigators at low cost.
3. Provide standardized protocols using written Standard Operating Procedures according to Good Laboratory Practice (GLP) protocols required for pre-clinical studies to support applications for research permits from the FDA.

1e(1b) H&MSF History and Qualifications. Muscle samples will be processed through the Microscopy Services Laboratory (MSL) (http://www.med.unc.edu/microscopy/) (C. Robert Bagnell, PhD, Director) and Muscle Pathology Laboratory (Leigh B. Thorne, MD, Director) of the Department of Pathology and Laboratory Medicine at UNC-CH . Dr. Kornegay is also trained in general pathology, with special expertise in neuropathology and muscle pathology. He has previously conducted systematic pathologic studies to determine the natural history of GRMD (Figure 2)^8-10 and response of affected dogs to therapy (Figure 3).³ The molecular component of the H&MSF is provided through the Joint Vector Laboratories of the UNC-CH Gene Therapy Center (http://genetherapy.unc.edu/jvl.htm), which has facilitated an ongoing Phase I trial for AAV gene delivery in DMD patients. Ability to track biodistribution of genetic therapies administered to animal models is a requirement of the FDA for IND submissions.

 

 

Figure 2. Quantitative Pathology Studies. Histopathologic sections (H&E) (top) and histograph (bottom) demonstrating true hypertrophy of the cranial sartorius at 6 months (top left) of age and replacement of muscle with connective tissue and fat (top right) at 23 months in GRMD. Histograph from reference 9.

 

 

 

 

 

Figure 3. Vastus lateralis muscle biopsies taken from 6-month-old dogs with GRMD showing changes in response to chronic 2 mg/kg daily oral prednisone treatment. Compared to untreated GRMD controls (A, C, and E), cross sections from prednisone-treated GRMD dogs show differences in myofiber calcification (B & F) and fetal myosin expression (D).  A & B, H&E stain; C & D, antibody to fetal myosin; E & F, alizarin red stain. Bar = 100 µm). From reference 3.

 

 

 

1e(1c) Standard Protocols and Endpoints. To provide consistency for projects that may ultimately lead to an IND application, it is important to standardize sample selection and processing and conduct procedures using GLP. We feel our experience with GLP standards will be particularly important as preclinical studies move forward in the IND process. Thus, investigators utilizing the NCDMD are encouraged to take advantage of expertise within the H&MSF for not just histology but also molecular techniques. Services available through the H&MSF include a wide array of procedures detailed in the next section. The breath of these capabilities helps ensure that essentially all histologic and molecular endpoints needed for projects can be met. Results from histologic and molecular studies will be correlated with those from the PTF and LAIF (see below).

For most projects, quantitative histologic studies will be done using immunohistochemistry. Commonly used techniques will include H&E, dystrophin with multiple antibodies, Gomori trichrome, oil red O, fetal myosin, and Alizarin red. We anticipate that Evans blue dye uptake will routinely be used to identify pre-necrotic and necrotic fibers.⁸ The percentage area of biopsy or necropsy samples occupied by fat, fibrous connective tissue, and muscle will be done with conventional morphometic techniques available through the H&MSF (see below). We have experience with these procedures and techniques in both natural history and preclinical studies.^3,8-10 Morphometric assessment of muscle necrosis and/or regeneration (together with dystrophin expression depending on the therapeutic approach) will be the primary histologic endpoint.

1e(1d). H&MSF Services:
Prepare high-quality histologic sections for NCDMD investigators. A number of routine processing procedures and services are performed by the Core technical staff, including:

Histochemical stains of muscle. Specific stains vary with the particular therapeutic intervention. Stains routinely used to evaluate muscle include H&E, modified Gomori trichrome, NADH, Congo red, PAS, ATPase, alkaline phosphatase, acid phosphatase, non-specific esterase, oil red O, myophosphorylase, myoadenylate deaminase, and alizarin red.

Specific histologic assessment following administration of vectors in different organs at high dose (1x1012vg/kg) can be performed in skeletal muscle at the site of injection and the contralateral muscle, as well as other organs.

Processing and sectioning of frozen tissues and paraffin, epon, or glycol methacrylate embedded tissues for light microscopy.
Processing and ultrathin sectioning of tissues embedded in epoxy resin and/or immunohistochemical embedding media for transmission electron microscopy, including examination by TEM.

Processing tissues and examination by scanning electron microscopy.

Perform biodistribution studies of vectors administered into canine models through different routes of administration in accordance with FDA guidelines for submission of INDs. Genomic DNA (gDNA) from frozen tissue/blood will be extracted at the specified time-points.

1e(1e). H&MSF Organizational Structure and Explanation of Services. The central service function of this core facility is to provide high-quality histologic preparations and molecular biodistribution of vector therapeutic for NCDMD investigators. These procedures are performed by a highly-qualified technical staff according to standardized protocols to insure uniform quality. We will also provide technical assistance in histologic preparation and light/electron microscopy and, where appropriate, will assist with fixation and cryopreservation of specimens to optimize preservation of ultrastructure and epitopes to be probed by immunohistochemistry. For immunohistochemical analyses specific to an individual project, the H&MSF technical staff will assist in the design and optimization of immunohistochemical staining protocols and will insure that appropriate positive and negative controls are included. The ability to carry out routine GLP histology and biodistribution analysis for IND submission is a primary milestone objective of the H&MSF. Capability of providing GLP quality service to pre-clinical studies allows investigators to better design and stream line FDA formal toxicologic and biodistribution studies. This service will provide standardized analysis and a database for all investigators, shortening time and cost to clinical trials.

1e (2) PHYSIOLOGY TESTING FACILITY (PTF)
1e(2a). PTF Specific aims:

1. Develop noninvasive functional tests that can be used serially over time to define the natural history of GRMD and other canine models.
2. Consult with investigators using the NCDMD so as to develop protocols to effectively assess function of GRMD dogs and other canine models in response to a proposed intervention.
3. Conduct functional tests on GRMD and other dystrophic dogs and normal littermate controls to demonstrate benefit (where present) of a proposed intervention.
4. Work with investigators to interpret and present PTF data for sake of publications and IND applications.

1e(2b). PTF History and Qualifications. Dr. Joe Kornegay, the PTF Director, was instrumental in defining the GRMD model and has worked with affected dogs for 25 years. He has collaborated with others to develop a series of physiologic measurements of canine muscle force generation. These tests have been utilized to define both the natural history of GRMD^51,52 and the response of GRMD dogs to therapy.⁴⁹ They will be utilized through the PTF to achieve the Aims detailed above.

1e(2c). Standard Protocols and Endpoints. Functional assays available through the PTF are described in the next section. Tibiotarsal joint force decrement subsequent to eccentric contraction will be the principal functional endpoint.⁷ Secondary functional endpoints could include isometric distal limb force (tibiotarsal joint force)⁶ and tibiotarsal joint contractures.⁴ The quadriceps force assay detailed below may also be useful.

1e(2d). PTF Explanation of Services. All tests will be done utilizing GLP. For all tests, described below, dogs are premedicated with acepromazine maleate (0.02 mg/kg), butorphanol (0.4 mg/kg), and atropine sulfate (0.04 mg/kg), masked, intubated and subsequently maintained with isoflurane. Anesthetized dogs are positioned in dorsal recumbency. The pelvic limbs are alternately immobilized in a custom-made stereotactic frame that aligns the tibia parallel to the table at a 90^o angle to the femur.

---

Figure 4. Tibiotarsal force measurements from normal (open bars) and GRMD (closed bars) at 3, 4.5, 6, and 12 months of age. GRMD force is less than normal at all ages. Differential involvement is seen according to age, with flexion affected early and extension late. From reference 6.

---

Isometric distal limb (tibiotarsal joint) force assay.⁶ Baseline forces are measured by stimulating either the common peroneal (tibiotarsal joint flexion) or tibial (tibiotarsal joint extension) nerve using paired stimulating and reference 27-gauge monopolar needle electrodes placed just distal to the fibular head (common peroneal nerve) or within the gastrocnemious muscles (tibial nerve), respectively. As a result, the distal pelvic limb pulls (flexion) or pushes against (extension) a lever interfaced with a force transducer (Aurora Scientific, Ontario, Canada), providing a measure of isometric force. Supramaximal 150 V, 100 sec pulses are applied (Model S48 Solid State Square Wave Stimulator) in a tetanic run of 250 pulses (50/sec). Passive force is subtracted from total force produced; only active force generated by muscles is measured. Because dogs vary in weight, the absolute tetanic force (Newtons) is divided by the body weight (kg) to obtain weight-corrected isometric force (Figure 4). Dogs are repositioned after each force measurement, and the higher of two weight-corrected forces is recorded.

---

Figure 5. A. Left pelvic limb of a 6-month-old GRMD dog immobilized in a stereotactic frame, with needles positioned to stimulate the peroneal nerve. B. Histogram showing mean torque (N-m) generated by tetanic tibiotarsal joint flexion during a series of 10 tetanic contractions. At the conclusion of each tetanic contraction, these maximally-stimulated cranial tibial compartment muscles were forcibly stretched by the servomotor as the tibiotarsal joint was extended a further 30 degrees at a rate of 0.7 muscle length/sec) (see D). The mean value of the initial contractions (blue bar) was ~ 1.4 N-m, while the tenth contraction (red bar) was ~ 0.2 N-m, representing an 85% decrement. C. The characteristic tetanic mechanical potential produced by tibiotarsal joint flexion is illustrated. D. A single tetanic mechanical potential analogous to that seen in C is followed by a sharp further deflection reflecting the eccentric contraction induced by the servomotor, with an immediate return to baseline. The baseline deflection following the potential is due to vibration artifact. Data shown here were collected using a modification of the method described. The tetanic run was 700 ms, with the first 500 isometric followed by 200 ms of stretch.

---

Repeated eccentric contractions and force measurements. Dystrophin serves to buttress the muscle cell membrane. In the absence of dystrophin, the membrane is prone to tearing during minimal exercise. Dystrophic muscles are particularly prone to injury subsequent to eccentric (lengthening) contractions.¹¹ As an example, mdx mice have near-normal absolute force measurements but demonstrate greater-than-normal force decrement subsequent to eccentric muscle contractions.¹² We have previously shown that GRMD dogs also have a greater-than-normal force decrement with eccentric contractions.³ In an initial study, eccentric contractions were induced in flexor muscles of the cranial tibial compartment of the pelvic limb by stimulating the sciatic trunk in the mid-thigh area.⁷ This caused contraction of both the tibiotarsal joint flexors and extensors. Because the extensors are more powerful, eccentric (lengthening) contractions were induced in the flexors. More recently, we have developed a technique whereby the common peroneal nerve is stimulated while simultaneously extending the tibiotarsal joint with a servomotor (Aurora Scientific) coupled to a lever arm (Figure 6). Movement of the lever arm is controlled by use of a computer and customized LabView software (Aurora Scientific).

Before beginning the experimental contractions, isometric forces from single muscle twitches are measured. The sciatic and common peroneal nerves are alternately stimulated using square wave pulses of 100µsec duration. The percutaneous stimulating electrode is optimally positioned and voltage adjusted until twitch force (Pt) reaches a maximum. To maximally activate the muscles of interest, stimulation voltage is 50% greater than necessary to achieve Pt. Eccentric contractions are induced using square wave pulses of 100 µsec duration in a tetanic run for 1 sec at a frequency of 50 Hz. The contraction is held isometric at the optimal fiber length (L0) for the first 900 ms. In the final 100 ms, the muscles of the cranial tibial compartment are stretched by the servomotor at a rate of 0.7 muscle lengths/sec., such that the muscles are stretched to 107% of L0. Stretches are performed 10 times, once every 5 sec. Following a 4 minute rest, the series of 10 stretches is repeated. Thus, the muscles of the craniotibial compartment are repeatedly stretched to induce mechanical damage. Contraction-induced injury is quantified by the force deficit (Fd) using the following equation: Fd = (Po before stretch - Po after stretch / Po before stretch) X 100.

---

Figure 6. Quadriceps Force Assay. Top. The dog is positioned in dorsal recumbency with the pelvic limb immobilized. A (yellow) band is positioned at the distal tibia & secured to a force transducer. The femoral nerve is stimulated percutaneously at the inguinal area. Bottom. Histogram of mean ± SD force in the right and left limbs from five GRMD dogs at 2 months of age. There is excellent side-to-side consistency.

---

Quadriceps force assay. Motivated by the observation by Vignos that quadriceps strength is the most important determinant of continued ambulation in DMD patients,¹³ we have developed an additional technique to measure quadriceps muscle force (Figure 6). A simple load cell is used to capture isometric tension generated by quadriceps contraction after percutaneous femoral nerve stimulation. Data is captured by the same custom software described above. The extending distal limb pulls against a lever interfaced with a force transducer. We have evaluated a total of five GRMD dogs in the context of an ongoing regional limb gene therapy study. Pretreatment values were ~ 3.0 N/kg, with good consistency from side to side.

1e(2e). PTF Organizational Structure. The central service function of this core facility is to provide high-quality functional testing for NCDMD investigators. These procedures are performed by a highly-trained and qualified technical staff according to standardized protocols to insure uniform quality on a day-to-day basis. Quality control is monitored on a routine basis. The centralized location of this core withiin the animal facilitates ensures direct communication between the investigators and the PTF staff about their specific needs. The ability to carry out routine SOP analysis from the PTF facilitates the IND by providing efficacy data related to the chosen therapeutic. This service will standardize analysis for all investigators and also provide a database that will be available to all investigators. The direct outcome of this service should significantly shorten time and cost to clinical trials. Specific milestones dictating progression of projects to and through the PTF will be determined based on discussions with investigators and through oversight of the NCDMD Steering and Operations Committees. Typically, dogs will not be evaluated in the PTF unless certain milestones have been achieved through the Histology & Molecular Services Facility (H&MSF). This service is not offered by any commercial vendor and, therefore, can only be provided by the PTF.

1e(3). LARGE ANIMAL IMAGING FACILITY (LAIF)
The LAIF is a part of the UNC-CH Animal Imaging Center (AIC), whose primary goal is to serve as a centralized, on-site resource to facilitate performance of high quality animal image acquisition and analysis.

1e(3a). LAIF Specific Aims.

1. Establish user-defined imaging protocols providing imaging services and image analysis tools.
2. Develop novel MR imaging methods for proposed projects.
3. Establish required imaging resources for large animal studies.

1e(3b). AIC History and Qualifications: The UNC-CH AIC was established in 2003 and is directed by Dr. Weili Lin. The 2500 sq ft facility houses 3T and 9.4 MR scanners (both funded through High End Shared Instrument grants, PI: Lin), a SPECT scanner, a PET/CT scanner, an in vivo optical imaging system, a small animal surgical suite, and a short-term animal housing area. Additional funds have been committed to purchase an animal PET/CT scanner, a high resolution in vivo ultrasound imaging system, and an in vivo optical imaging system capable of providing tomographic images. Both PET/CT and high resolution ultrasound systems will be available for dog imaging. Currently, we provide services to faculty from several UNC-CH departments. Therefore, the required infrastructure to provide both image data acquisition and analysis is well established and can be easily adapted for investigators using the NCDMD. The LAIF will complement analyses conducted through the H&MSF and PTF by providing measurable image analysis. Our long-term goal is to establish routine non-invasive image analysis that will be adapted and utilized in clinical studies.

1e(3c). Standard Protocols and Endpoints. Imaging modalities available through the LAIF are described in the next section. Based on our preliminary results, fat percentage within selected pelvic limb muscles assessed with MRI will be the principal endpoint. Muscle volume and T1 and T2 values are potential secondary endpoints. We plan to develop a MRI image database of the natural disease progression in GRMD by the beginning of year 3 of the NCDMD. MRI will be used for projects in year 4 and beyond only if endpoints are useful in demonstrating treatment efficacy and/or correlate with results from functional and histologic studies. We anticipate each dog entered on projects will have MRIs completed three times within a 6 month to one year timeframe. Results from MRI studies will be correlated with those from the H&MSF and PTF (above).

1e(3d). AIC (LAIF) Imaging Modalities and Services. We have completed preliminary MRI studies on a 3T scanner available through the UNC-CH Animal Imaging Center. The animal imaging protocol (Table 4) was based on one used previously in DMD patients at UNC-CH (Fan J, Howard J, and Lin W, unpublished observations). Our animal protocol provides excellent resolution, thus allowing region-of-interest measurements of MRI parameters (Figure 7; Table 5). With the recent relocation of our colony, we did not have access to normal dogs for comparative studies. For these preliminary studies, MRI findings from GRMD dogs were compared to those in carriers. Quantitative studies have been completed in two GRMD dogs (2 months and 8 years old) and two carriers (2 months and 8 years old) and two carriers (2 months and 5 years old). We have focused on seven muscles of the proximal pelvic limb (cranial sartorius, quadriceps femoris [rectus femoris and vastus heads evaluated separately], biceps femoris, adductor, gracilis, semimembranosus, and semitendinosus). Signal-intense lesions presumably corresponding to edema associated with necrosis were seen on T2 weighted images in the 2-month-old GRMD dog, while the 8-year-old dog had increased fat deposition. The severity of these changes varied among muscles both visually (Figure 7) and quantitatively (Table 5). Signal-intense lesions in the 2-month-old dog were particularly pronounced in the rectus femoris, adductor, biceps femoris, and vastus lateralis and medialis muscles (Figure 7D). Fatty changes in the older GRMD dog were more prominent in the semimembranosus and semitendinosus muscles (Figure 7J). Volumetric changes did not vary dramatically between GRMD and carrier dogs.

---

---

---

Figure 7. GRMD MRI Studies. The four panels from left to right are MRI images from a 2-month-old GRMD carrier (A,C,E,G) and affected littermate (B,D,F,H) and 5-year-old GRMD carrier (I,K,M) and affected dog (J,L,N). A, B, I, and J are TSE-fat percentage and C, D, K, and L are TSE-fat saturation (see protocol in Table 4). Transverse sections of muscle have been segmented in E, F, M, and N for region-of-interest measurements (see Table 5) and are shown in three dimensions in G and H (2-month-old dogs only). Note, particularly, the signal-intense lesions in several muscles in D and J, representing fluid accumulation, acutely, and fatty change, chronically, respectively. Signal-intense lesions seen in J reverse with fat saturation in L.

---

Image Analysis. Our team led by Dr. Martin Styner, Co-director of the UNC Neuro Image Analysis Laboratory, is highly skilled at conducting image analysis studies and can offer a wide array of image analysis tools that are routinely applied in human, primate and rodent imaging studies. Many of our tools have been developed in-house and are freely available as cross-platform (Windows, Linux, Solaris, Mac OS X) binary distributions or as open-source¹⁴ (available at http://www.ia.unc.edu/dev/download/index.htm).

Structural Analysis. In regard to the methodology that will be applied to dogs evaluated through the NCDMD, probabilistic tissue segmentations as well as selected region-of-interest segmentation tools will be employed. Our atlas-based tissue classification^15,16 is a versatile tool that segments tissue classes using a probabilistic tissue distribution atlas. It is regularly applied for human and small animal brain tissue segmentation. The tissue segmentation procedure further provides intensity calibration and intensity inhomogeneity correction;¹⁷ the latter is especially important in the proposed image analysis due to the large field-of-view and the associated clearly visible intensity inhomogeneity. For the segmentation of the individual muscles, our region-of-interest (ROI) segmentation tool utilizes a prior atlas, derived as an unbiased average image in a separate atlas building step, that is deformed to match each intensity corrected and calibrated MRI image via non-linear high-dimensional fluid deformation.¹⁸ The atlas’s structural probabilistic ROI definitions are propagated to each subject’s image using the computed deformation field.¹⁹ Propagated ROIs are reviewed for accuracy and edited²⁰ only if they deviate significantly from critical, and easily defined, boundaries. The corresponding volumetric measurements for ROIs and tissue segmentations are automatically computed. Both whole ROI and parcellation computation methodology has been validated and evaluated for stability in brain imaging studies. Repeatability studies show coefficients of variance less than 1.5% for all brain measurements.

Atlas Building. Based on the images obtained from individual muscles, we intend to develop atlases of the pelvic limb muscles, separate for the lower and upper limb. These atlases will build via an unbiased average image computation scheme.^18,21 In combination with our tools, these atlases will allow us and others to investigate subtle anatomical muscular changes during disease progression or in response to treatment.

Tracking of Image Measurements Over Time. Our registration tools^18,21,22 allow the tracking of longitudinal structural changes in imaging studies.^23-26 These tools are particularly important for longitudinal studies to be conducted through the NCDMD.

Image Processing. The images are uploaded via a password-protected procedure to Dr. Styner at the UNC Neuro Image Analysis Laboratory (NiAL). Investigators will have remote access to the datasets via password-protected, encrypted secure shell login directly to the NIAL file server. All datasets, as well as derived data computed as part of the processing procedure, are stored on RAID 5 disk storage to protect against single disc failures. Data backup is performed daily with bi-weekly offsite storage of the backup tapes.

1e(3e). AIC (LAIF) Organizational Structure. As mentioned previously, the UNC AIF is currently providing services to many investigators across UNC-CH. We will utilize a modification of this model to provide services to NCDMD investigators. As currently structured, studies will be restricted to MRI. Investigators are invited to contact us for discussions of specific protocols in advance of developing the initial grant. Studies will most likely be based on the imaging protocol outline in Table 4. A specific milestone of this core is to accumulate an image database of the natural disease progression in GRMD correlated with histologic and functional studies from the H&MSF and PTF. This database will allow direct comparison with treated dogs and provide supportive data to therapy success and help support pre-clinical IND submission. The long-term goal of this facility is to expedite non-invasive analysis of various treatment paradigms used in the DMD canine model that will eventually support similar analysis in clinical studies. Support of this core will substantially reduce the cost of MRI studies (Table 1)

Aim 2. Conduct Translational Research Focused on Therapeutic Strategies for DMD

This research will lead to identification of promising candidates for advancement to IND/clinical trials. Projects to be conducted through the NCDMD will be chosen by the Steering Committee based on stringent quantifiable criteria. Continued support of these projects will be contingent on completion of milestones and validation of the need for extending services offered by the NCDMD.

2a. Selection of projects for the NCDMD. The NCDMD encourages submission of subsidy applications covering all treatment modalities from a variety of investigators with diverse academic and corporate backgrounds. The overall process for routing projects through the NCDMD from original submission to the investigator-initiated investigational new drug (IND) application is summarized below.

Sequential Steps for NCDMD Projects.

1. Investigators should contact NCDMD Director, Dr. Joe Kornegay, early in the process. Ultimate submission of a subsidy application is viewed as an iterative process.
2. Project application is reviewed and scored by the NCDMD Steering Committee. Key criteria to be evaluated include results from the mdx mouse, availability of preliminary data from normal dogs (pharmacokinetics of the proposed compound, etc), and the likelihood that the compound can be moved expeditiously to human trials. U-series projects funded through NINDS will already have well-defined milestones. For other projects, milestones may have to be developed or modified based on input from the Steering Committee. Approval of projects, either at the time of original review or at appeal, will require agreement by five of the eight Steering Committee members.
3. Approved projects will be initiated through the NCDMD Canine Muscular Dystrophy Facility (CMDF) and the NCDMD cores (histology/molecular, physiology, and imaging) with oversight by the NCDMD AQCO. Projects must complete sequential milestones to progress through the NCDMD. These milestones will vary with individual projects but will generally involve demonstrating initial success at the molecular or histological level before moving to physiological testing and imaging. Progress will be monitored through weekly meetings of the NCDMD Operations Committee.
4. The NCMD will communicate regularly with the Project Investigator and Steering Committee. Projects that do not achieve milestones will be terminated by the Steering Committee, with the opportunity for appeal.
5. Results will be reviewed by the NCDMD AQCO and reported to the Project Investigator for initiation of the IND application, if indicated.

Proposals are sought from researchers who wish to conduct collaborative research projects that involve testing of promising candidates in dog models of DMD. The NCDMD Steering Committee will determine which therapeutic strategies will be pursued. Criteria used to evaluate the scientific merit of applications will vary with the proposed method. In general, merit will be based on three major criteria scored 1-5 points:

1. Documentation of therapeutic benefit in the mdx mouse, as evidenced by systematic, published studies. All therapeutic approaches will be considered. Treatments will generally fall into one of three major categories:
   1. molecular therapy wherein dystrophin expression will be the major outcome;
   2. cellular therapy wherein sustained implantation of transplanted cells with associated dystrophin expression will be the major outcome; and
   3. pharmacologic management wherein the outcome parameter(s) will vary with the particular intervention but should include an intermediate measurement (biomarker) demonstrating that the compound is active at the proposed mechanistic site. With any approach, multiple outcome parameters could apply.  For assessment of skeletal muscle function in mdx mice, preference will be given to investigators who have shown improvement in force generation using an eccentric muscle contraction protocol. These studies will not be done through the NCDMD.
2. Documentation of safety in normal dogs, to include the Pharmacokinetics/Biodistribution and Toxicological profile of the compound/agent proposed. These studies could be done through the NCDMD. However, priority would be given to projects in which these data were already available.
3. Likelihood that the proposed therapeutic approach can be moved expeditiously into human clinical trials, assuming that promising results are identified in a canine dystrophic model.

U-series grants will already have well-developed milestones. Indeed, continued funding of these projects by the sponsoring Institute will be contingent on completing these milestones. It will be important for the NCDMD Steering and Operations Committee to work closely with these investigators to ensure that their milestones can be met. Other projects in which funding will ultimately come from other mechanisms must also have well-developed milestones. Sequential progression of projects through the NCDMD is contingent on completion of these milestones, as judged by the NCDMD Steering Committee.

2b(2). Canine Muscular Dystrophy Facility (CMDF) Go/No Go GSHPMD Model Milestone. The CMDF will produce and maintain a breeding colony of well-characterized dogs with genetically-determined muscular dystrophy. The GRMD model has been best characterized by our group and will continue to be used in most studies. However, German shorthaired pointers (GSHPMD) have a large DNA deletion, essentially amounting to a “dystrophin knock out”.²⁷ This “dystrophin knockout” model could be advantageous, as revertant fibers that would otherwise confound results of gene therapy studies should be absent. Moreover, the absence of revertant fibers provides a “cleaner” background in which to conduct immunologic studies of the relative roles that viral vectors and transgenes play in the immunologic response with viral-based gene therapy. So, as to be able to utilize this model in preclinical studies, we must characterize the phenotype. Thus, we have established a Go/No Go milestone for the GSHPMD model.

We will characterize the natural disease progression of GSHPMD in 3 to 12-month-old dogs using histologic, functional and imaging endpoints by April, 2012 (anticipated beginning of year 4 of the U24). We have obtained separate funding for these studies.

2b(3). Large Animal Imaging Facility (LAIF) Go/No Go Milestones. As described above, the Histology and Molecular Services Facility (H&MSF), Physiology Testing Facility (PTF), and Large Animal Imaging Facility (LAIF) are key components of the NCDMD, providing assays that will establish endpoints under GLP, dictating whether projects move forward in the preclinical IND process. We have extensive experience utilizing both histologic (H&MSF) and functional (PTF) assays but minimal experience with the MRI modalities. Thus, we have established two Go/No Go milestones for the LAIF.

1. We will develop an image database of the natural disease progression in GRMD dogs from 3 to 12 months of age and correlate these findings with histologic (H&MSF) and functional (PTF) studies from the same dogs by April, 2011 (anticipated beginning of year 3 of the U24).
2. MRI endpoints will have been useful in demonstrating treatment efficacy and/or correlate with results from functional and histologic studies by April, 2012 (anticipated beginning of year 4 of the U24) based on results from dogs evaluated in projects entered in years 1 - 3. We have obtained separate funding for these studies.

2b(4). Individual Project Milestones. Milestones for individual projects must be developed at three levels:

1. initiating canine projects;
2. conducting sequential studies in canine models; and
3. moving from canine studies to the IND application process.

2b(4a). Individual Project Go/No Go Milestones for Initiating GRMD Studies.

1. Demonstrated efficacy in the mdx mouse.
2. Ability to show a satisfactory pharmacokinetic/biodistribution or toxicological profile in normal dogs.
3. Likelihood that the proposed therapeutic approach can be moved expeditiously into human clinical trials, assuming that promising results are identified in a canine dystrophic model.

2b(4b). Individual Project Go/No Go Milestones for Completing Sequential Studies in GRMD Dogs. The sequence of canine model studies and associated Go/No Go milestones will vary with the therapeutic intervention. In general, milestones will have to be achieved before moving forward with complete functional, imaging, and/or further histological studies. Examples of milestones for the three general categories of DMD treatment include:

• Molecular: Demonstrate transduction of at least 25% of the myofibers within the transplanted field (could be an individual muscle or limb).
• Cellular: Demonstrate surviving transplanted cells with associated dystrophin expression within the treated field (could be an individual muscle of limb) for at least 3 months following the intervention.
• Pharmacologic: Develop an intermediate measurement (biomarker) demonstrating that the compound is active at the proposed mechanistic site.

For molecular and cellular treatments, it would also be essential to complete the proposed sequential progression of the intervention from, as an example, localized to regional to systemic treatments.

2b(4c). Individual Project Go/No Go Milestones for Moving From GRMD Studies to the IND Process and Human Trials. Milestones to be achieved before moving forward with the IND process and human trials would vary with the proposed intervention but would include those for fully completing the GRMD experiments (above) and additional functional, imaging, and/or further histological studies.

1. Reduce myofiber necrosis by 25% as evidenced by quantitative histopathological studies (counting of necrotic fibers, Evans blue dye uptake, etc).
2. Increase individual or grouped muscle force by at least 25%.

Literature Cited

1. Kornegay JN, Bogan DJ, Bogan JR: Effect of gender on phenotype in golden retriever muscular dystrophy.  Mol Ther 16 (S1):S200-201, 2008.
2. Kelley RL: Factors influencing canine reproduction and nutritional management of the pregnant bitch.  Canine Reproduction and Nutritional Health – Tufts Animal Expo, pp 9-14, 2001.
3. Liu JMK, Okamura CS, Bogan DJ, et al:  Effects of prednisone in canine muscular dystrophy.  Muscle Nerve 30:767-773, 2004.
4. Kornegay JN, Sharp NJH, Schueler RO, Betts CW:  Tarsal joint contracture in dogs with golden retriever muscular dystrophy.  Lab Anim Sci 44:331-333, 1994.
5. Kornegay JN, Sharp NJH, Bogan DJ, Van Camp SD, Metcalf JR, Schueler RO: Contraction tension and kinetics of the peroneus longus muscle in golden retriever muscular dystrophy.  J Neurol Sci 123:100-107, 1994.
6. Kornegay JN, Bogan DJ, Bogan JR, Childers MK, Cundiff DD, Petroski GF, Schueler RO: Contraction force generated by tarsal joint flexion and extension in dogs with golden  retriever muscular dystrophy. J Neurol Sci 166:115-121, 1999.
7. Childers MK, Okamura CS, Bogan DJ, Bogan JR, Petroski GF, McDonald K, Kornegay JN:  Eccentric contraction injury in dystrophic canine muscle.  Arch Phys Med Rehabil 83:1572-1578, 2002.
8. Childers MK,  Okamura CS,  Bogan DJ,  Bogan JR,  Sullivan MJ,  Kornegay JN:  Myofiber injury and regeneration in a canine homologue of Duchenne muscular dystrophy.  Am J Phys Med Rehab 80:175-181, 2001.
9. Kornegay JN, Cundiff DD, Bogan DJ, Bogan JR, Okamura CS:  The cranial sartorius muscle undergoes true hypertrophy in dogs with golden retriever muscular dystrophy.  Neuromuscul Disord 13:493-500, 2003.
10. Kornegay JN, Tuler SM, Miller DM, Levesque DC: Muscular dystrophy in a litter of golden retriever dogs.  Muscle Nerve 11:1056-1064, 1988.
11. Edwards RHT, Jones DA, Newham DJ, Chapman SJ: Role of mechanical damage in the pathogenesis of proximal myopathy in man.  Lancet 1:548-552, 1984.
12. Moens P, Baatsen PHWW, Marechal G: Increased susceptibility of EDL muscles from mdx mice to damage induced by contractions with stretch. J Mus Res Cell Motil 14:446-451, 1993.
13. Vignos PJ Jr, Spencer GE Jr, Archibald JC: Management of muscular dystrophy of childhood. JAMA 184:89-96, 1963.
14. Styner M, Jomier M, Gerig G: Closed and open source neuroimage analysis tools and libraries at UNC: IEEE Symposium on Biomedical Imaging ISBI 2006, pp. 702-705.
15. Styner M, Charles HC, Park J, Lieberman J, Gerig G: Multi-site Validation of Image Analysis Methods - Assessing Intra and Inter-site Variability, Proc. SPIE Vol. 4684, Medical Imaging 2002, p. 278-286.
16. Prastawa M, Gilmore JH, Lin W, Gerig G: Automatic Segmentation of MR Images of the Developing Newborn Brain.  Medical Image Analysis 9:457-466, 2005.
17. Styner M, Gerig G, Brechbühler C, Szekely G: Parametric estimate of intensity inhomogeneities applied to MRI, IEEE Transactions on Medical Imaging, 19:153-165, 2000.
18. Joshi S, Davis B, Jomier M, Gerig G: Unbiased diffeomorphic atlas construction for computational anatomy.  NeuroImage 23:S151-S160, 2004.
19. Gouttard S, Styner M, Joshi S, Gerig G: Subcortical structure segmentation using a probabilistic atlas prior. Proc SPIE Medical Imaging Conference, in print, 2007.
20. Yushkevich PA, Piven J, Cody Hazlett H., Gimpel Smith, R, Ho S, Gee JJ, Gerig G: User-guided 3D active contour segmentation of anatomical structures: Significantly improved efficiency and reliability.  NeuroImage 31:1116–1128, 2006.
21. Lorenzen P, Prastawa M, Davis B, Gerig G, Bullitt E, Joshi S: Multi-modal image set registration and atlas formation.  Medical Image Analysis MEDIA 10: 440-451, 2006.
22. Rueckert D, Sonoda LI, Hayes C, Hill DLG, Leach MO, Hawkes DJ: Non-rigid registration using free-form deformations: Application to breast MR images. IEEE Transactions on Medical Imaging18:712-721, 1999.
23. Gerig G, Davis B, Lorenzen P, Xu S, Jomier M, Piven J, Joshi S: Computational anatomy to assess longitudinal trajectory of brain growth, Proc. 3DPVT, 2006.
24. Cevidanes LH, Styner M, Proffit WR: Image analysis and superimposition of 3-dimensional cone beam computed tomography models. Am J Orthodon DentoFacial Orthoped 129:611-618, 2006.
25. Pappas I, Styner M, Malik P, Remonda L, Caversaccio M: Automatic method to assess local CT-MR imaging registration accuracy on images of the head. Am J Neuroradiol 26:137-144, 2005.
26. Cevidanes LHS, Franco AA, Gerig G, Proffit WR, Slice, DE, Enlow, DH, et al. Assessment of mandibular growth and response to orthopedic therapy with 3-dimensional magnetic resonance images. Am J Orthod Dentofacial Orthop, 128:16-26, 2005.
27. Schatzberg SJ, Olby NJ, Breen M, Anderson LVB, Langford CF, Dickens HF, Wilton SD, Zeiss CJ, Binns MM, Kornegay JN, Morris GE,  Sharp NJH:  Molecular analysis of a spontaneous dystrophin “knockout” dog. Neuromuscul Disord 9:289-295, 1999.