BoneKEy-Osteovision | Commentary
Criteria for Evaluating Success of a Therapeutic Intervention in Osteogenesis Imperfecta: Application to Cell Transplantation and Bisphosphonates Therapy
DOI:10.1138/2001021
Developing an effective therapeutic approach for osteogenesis imperfecta (OI) can be justified purely on the physical and psychological toll the disease places on affected individuals and their families as well as its financial burden on the health-care system. However, OI is of interest to the basic scientist because of the pivotal role the disease has played in revealing the basic cellular and physiological consequences of a mutation in a structural component of the extracellular matrix. From molecular studies of OI, an appreciation of mutations affecting the triple helical domain of the type 1 collagen molecule has been gained that can be generalized to all collagen molecules. This knowledge was paramount to defining many of the heritable diseases of connective tissue. More recently identified types of OI that do not arise from mutations of type I collagen will lead to other matrix genes that underlie bone disease (). The strong dominant negative nature of an OI mutation affects the quality of the matrix that is produced and the synthetic and proliferative capacity of the cells that make the mutant molecules. Accumulation of a weakened matrix creates a state of unremitting high bone turnover in which defective matrix is continuously removed and replaced. In bone, the high bone turnover rate is a chronic stimulus for the mutant osteoblastic cell to proliferate and maximize its output of extracellular matrix proteins.
Somatic gene therapy (SGT) for OI represents an arduous challenge to the development of a therapy for a strong dominant negative disease. Here again, OI will be the experimental paradigm for working out the complex problems before success is realized. Studies of individuals who are somatic mosaics for an OI mutation suggested that SGT might be successful for a dominant negative genetic disease. For reasons that are still unclear, the admixing of normal and mutant cells in bone ameliorates bone fragility. Presumably normal cells out populate and out produce mutant cells such that the normal cells become the predominant functional osteoblast within bone. Thus one major hurdle to a successful SGT is the delivery of a sufficient number of cells to bone to have a meaningful biological effect. This is the importance of the recently reported follow-up study by Horowitz et al. () and this publication must be critically evaluated to determine if the goal of biologically significant engraftment has been attained.
The paper reports results on three of five OI children who underwent a successful whole bone marrow transplantation. Two of the treated individuals were reported in the original Nature Medicine paper (). Two other OI children of the same age and disease class who did not have transplantation served as controls. The clinical evaluation of the two children who underwent bone marrow transplantation but did not engraft are not discussed. They would have been valuable controls for the effect of the conditioning regimen on growth and bone mineral accretion. The report suggests that during the 18 months after transplantation there was a transitory increase in growth rate and total bone mineral accretion and a measurable decrease in fracture frequency. Whereas these responses are encouraging, these measurements are highly variable in young children which in part accounts for the dramatic swings in growth velocity and bone mineral density. Moreover, the natural history of OI, particularly in rapidly growing children, is highly variable. Other measurements which might make interpretation of the outcome more clear are not reported. For example, radiographs of the bone would have been particularly helpful to determine if the increase in bone mineral was associative with remodeling of the skeleton or an increase in cortical thickness. Total bone mineral density, as utilized in this report, is highly correlated with weight gain in normal children and its relationship to growth in OI children with varying disease severity has not been established. It would have been helpful to have provided BMD information on the vertebral site, for which data is available in a larger number of OI children at a similar range of age and disease severity (). Also lacking is any comment on the clinical behavior of these individuals other than fracture rate. Persistent bone pain, diaphoresis, muscle weakness and delayed motor milestones are constant symptoms of severe OI, all of which improve in OI children treated with bisphosphonates. Because a follow up bone biopsy was not performed, evidence of new bone matrix formation or the degree of bone engraftment is unavailable to assess outcome.
Despite the limited information, the authors interpret their findings as an optimistic report that standard or stromal cell enriched bone marrow transplantation may be beneficial in the treatment of OI. Because this report was published in a blood rather than a bone journal, it may influences other bone marrow transplantation teams to initiate this form of treatment for OI. This conclusion cannot be supported because (a) the report understates the impact that bisphosphonates have had in improving quality of life, bone density, bone remodeling and somatic growth in a large number of children () and (b) the study lacks the prospective and objective measures including bone biopsy () needed to assess the impact of the procedure on bone health.
The success of bisphosphonate treatment appears to be due to the high rate of bone resorption that is an intrinsic component of the pathophysiology of OI. By inhibiting bone resorption, total bone matrix increases. While the accumulated matrix is still abnormal, it would appear that more abnormal matrix results in better bone function than extremely deficient abnormal matrix. Particularly in young children in whom the bone formation rate is high, a moderate reduction in bone resorption can have a dramatic effect on symptoms of OI. Thus a bone marrow conditioning regimen which is likely to have a dramatic effect on osteoclast development that would persist for some time after the treatment was terminated could produce a transient increase in bone mass and growth by a mechanism similar to bisphosphonates. This possibility is dismissed by the authors using evidence that is unrelated to the underlying pathophysiology of OI and without pre and post conditioning measurements of bone derived urinary crosslinks. Given the relative risks of the pretreatment conditioning regimens () plus complications that can arise after the transplant (), dramatic differences in disease symptoms and objective measures have to be demonstrated before this can be recommended as an experimentally justified therapy, not to mention to be considered as state-of-the-art.
Transplantation studies in animals using whole bone marrow alone or bone marrow enriched with marrow stromal cells are given as evidence that bone engraftment can be readily achieved. Numerous studies have shown engraftment of cells marked either with a transgene or a unique endogenous gene widely distributed in many tissues including lung, brain and bone (). Whereas many studies imply that these cells attain the same differentiation status as their endogenous neighbors, there is only one study demonstrating that a bone specific gene is expressed by the transplanted cells (). In many cases, the marker gene does not discriminate whether this cell arises from a mesenchymal or macrophage lineage. Even when a relatively pure populations of stromal cells are used for transplantation, small contamination from the myeloid lineage could belie stromal cell engraftment.
The underlying genetic abnormality also determines the success of stromal cell transplantation. When the disease is non-cell autonomous, (i.e. the transplanted cells are engineered to secrete a deficient soluble factor), an improvement in the disease phenotype occurs, irrespective of where the cells established residence. This is most obvious when the transplanted cells express a cytokine () or clotting factor (). This mechanism probably explains the recent success of bone marrow transplantation for vitamin D resistant rickets (). Clinical improvement of a disease that is cell autonomous, such as OI, requires that the cells populate the affected bone, proliferate, differentiate and participate in bone turnover. This is a standard that has not yet been met in any murine studies. The one exception is when the transplanted cells have undergone prolonged expansion ex vivo (). For mouse this type of treatment rapidly leads to cell immortalization making interpretation of a transplantation experiment difficult. Perhaps these studies suggest that bone engraftment is possible, but the proliferative capacity of primary stromal cells is insufficient to establish adequate engraftment. On the other hand, immortalization may remove impediments to engraftment that are present in normal cells. In this case the cells may be assuming a more autonomous nature, resembling cancerous cells. Thus, progress to date in achieving stromal cell engraftment of bone is disappointing at best and certainly not sufficient to justify its implementation in humans.
Given the remarkable success of bisphosphonates for symptomatic improvement of OI, it is incumbent on researchers in the field of SGT to develop criteria that (a) demonstrate successful engraftment and (b) document improvements in bone health that are equivalent to or exceed medical therapy (table 1, column A). Whether the transplanted individual is mouse or man, it must be demonstrated that the transplanted cells populate the bone, expand in number over time, participate in the new bone formation and provide a continuous source of new bone cells over the life of the transplanted subject. These studies are easier to perform in the mouse when the transplanted cells are engineered to contain visual transgenes driven by bone specific promoters. Not only does the transgenic marker identify the source of the cell, it can reflect its level of participation in new bone formation. Moreover, this approach can be used to determine whether donor stem cells removed from the transplanted bone are still able to generate differentiated osteoblastic cells in vitro. If the transplanted cells are participating in new bone formation, the quality and quantity of the bone should improve over time. Radiographs should show remodeled bones with improved architecture and bone density measurements at well-defined sites should be increased in humans. μCT can be used to assess bone mass and architecture in mice. Bone histomorphometry and dynamic labeling studies should confirm these clinical measures of bone health. Particularly in the mouse, it should be possible to demonstrate improve mechanical properties of the bone after the intervention. Biochemical studies can also contribute to the impression of success by demonstrating that markers of new bone formation remained elevated while the level of markers reflecting bone degradation will gradually subside.
The clinical evaluation of transplanted individuals is equally important to the assessment of success. Bone pain and diaphoresis are major symptoms that are daily facts of life for many patients with OI. Muscle weakness can be profound and in growing children greatly delay acquisition of motor milestones. Successful interruption of the pathophysiological cycle of OI appears to greatly reduce these symptoms. Bone pain diminishes, diaphoresis is greatly reduced and muscle tone improves. Children begin to ambulate and adults find that their activities of daily living are easier to perform. Because the symptoms are so profound for individuals with OI, they need to be recorded in a prospective manner as clinical outcome measure. These types of measurements are difficult to record in the mouse, although cage activity can be quantitated and probably would be informative. Certainly fracture frequency, linear growth and weight gain are important to record although they can be influenced by many other factors unrelated to the intervention.
Table 1 utilizes the criteria discussed above to assess the relative success of bisphosphonate (column B) and cell therapy (column C) in the treatment of OI in either species. The table illustrates the success of bisphosphonate for improving measurements of bone health and quality of life. Ongoing studies with randomized controls should firmly establish the place of this form of therapy for OI. The one concern is whether the structural properties of the bone have improved despite the fact that functional and clinical measures of bone health have improved. Camacho et al. () have reported on preliminary studies of bisphosphonate treated oim/oim mice that show a significant reduction in the number of fractures, increased femoral metaphyseal density, increased femoral diameter, and reduced tibial bowing. Potential negative effects include a persistence of calcified cartilage in the treated oim/oim metaphyses and significantly shorter femora compared to non-treated oim/oim mice. No significant differences in weight gain or mechanical properties of cortical bone were found in alendronate-treated oim/oim mice. In either species, bisphosphonate treatment is the standard for intervention against which any new therapy will have to be judged.
Scoring the success of cell therapy by the same standard as bisphosphonate shows how relatively ineffective cell therapy is at this time. Investigators in this field need to recognize this and work towards developing the reagents and model systems to measure the functionality of the transplant and the response of the experimental subjects to the transplant. Given the experimental difficulties of performing these studies in a controlled and quantitative manner in humans, greater emphasis must be placed on studies in experimental animals in which the problems inherent to the procedure can be identified and solved. We need to be critical of the work that we perform, so as not to provide unrealistic expectations to OI patients and their families.
This work is licensed under a
Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.