Bone morphogenetic proteins (BMPs) have a number of potential
applications in bone healing and in the treatment of other bone
disorders. Potential applications include the following:
accelerating normal bone healing
increasing the prevalence of the union of fractures and potentially increasing
the strength of the callus formed in initial fracture healing
increasing the effectiveness of autologous or allogeneic bone grafting
to improve fracture healing, heal bone defects, or treat nonunions
accelerating the rate of consolidation and increasing the prevalence
of union in distraction osteogenesis
replacing autogenous bone graft in treating deficiencies in bone union
to eliminate the cost and morbidity of harvesting autologous bone
increasing the rate of consolidation and strength of bone union in
reconstructive procedures such as spine arthrodesis and other arthrodeses
inducing bone formation in pathological bone defects
increasing bone mass disorders such as postmenopausal and disuse
osteoporosis
In clinical trials of BMPs, investigators thus far have chosen spine
arthrodesis, fracture nonunions, and high-risk acute fractures as
models for evaluating the BMPs. Spine arthrodeses have been chosen
because they are commonly performed; pseudarthrosis is fairly frequent
and can lead to failure of the procedure, and the need for harvesting large
amounts of autologous bone graft adds significantly to the morbidity
of these procedures. In addition, adequate autologous bone is usually
not available, especially in children. Fracture nonunions have been chosen
because they represent the failure of the normal biologic process
of healing, the previous standards of care procedures have failed,
and the morbidity and disability from nonunions is high. High-risk
fracture models such as open fractures of the tibia have been chosen
because the prevalence of nonunion is high compared with that of
other fractures, they are common, and again the morbidity and disability
from failure of union is significant.
In approaching these clinical scenarios, investigators can study a
number of potential outcomes. The BMPs can be used to augment the
standard of care to improve healing through acceleration of union
or increasing the incidence of union. The BMPs with their carrier
can be used to replace the need for autologous bone graft to reduce
costs and the morbidity of bone graft harvest. Or, the BMPs can
be used as an adjunct to or substitute for traditional standard
of care methods such as in the treatment of nonunions or bone defects
to provide a simpler lower morbidity rate and more efficacious method
of treatment.
The formulations of BMPs used thus far require a matrix into which
they are instilled, which is designed to distribute the BMP at the
treatment site and, in current formulations, requires open surgical
placement. Critical to success are the appropriate dose and the
maintenance of BMP in the site of treatment for a sufficiently long
period of time. Particularly attractive is the percutaneous installation
of BMPs through needles or canals, as this avoids the morbidity
associated with open surgical exposure. Even if BMP were found to be
equivalent to the current standard of care methods of treatment,
as opposed to superior, the ability to treat with a percutaneous
technique would greatly enhance the value of the BMP products used
because of the reduced morbidity of treatment as compared with that
of open methods.
Clinical trials such as those just outlined are exceedingly challenging
to the companies testing the product and to the academic and non-academic
clinical centers carrying out the clinical trial. As outlined by
Dr. Riedel in his comments, effective execution of a clinical trial
requires flexibility, compromise, dedication, scrupulous attention
to protocol to avoid bias, and long-term commitment to acquire sufficient patients
for the study and to follow them for a sufficiently long period
of time so that statistically proved differences between treatment
groups can be obtained.
Studies of various models of healing of acute fractures are challenging
because the standard of care methods produce high rates of success.
For example, in the treatment of open fractures of the tibia, delayed
unions are common but nonunion rates are only 2-5%. Since
the standard of care in these fractures is locked intramedullary
nailing, the majority of fractures are stable immediately following
treatment; this eliminates instability as a method of assessing
progression and completion of fracture union. The efficacy of a BMP
being compared with autologous bone graft in fractures treated with
intramedullary nails can thus be measured only by noninvasive means
such as radiographs, in attempts to assess the rate and amount of bone
formation and achievement of union as measured by bridging of at
least two cortices as viewed on anterior-posterior and lateral views.
The endpoint of success is usually measured as unrestricted pain-free
weight-bearing, assuming that late failure does not occur. Lastly,
even if differences can be shown between control and experimental
groups, is this difference sufficient to be of clinical significance
and will this difference in clinical outcome be worth the cost or
other collateral morbidities that might occur?
The effectiveness of an open tibial fracture study could be improved
by the use of external fixation rather than intramedullary nail
fixation; this would provide the opportunity to serially assess
the changing mechanical stability of the fracture site. The lack
of metallic implants at the fracture site would also improve the
ability to assess the rate, amount, and density of bone formation.
This could be made quantitative by the use of selected cuts on computed tomography
(CT) where an accurate volumetric assessment of bone formed would
be possible, as well as by measuring its density in Hounsfield units. Although
external fixation is a standard, accepted, available treatment for
open fractures of the tibia, recently it has lost favor as compared
with intramedullary nailing. If fracture union were found to be
superior with BMPs, then the more invasive methods such as intramedullary
nailing could fall by the wayside in favor of less invasive techniques
such as external fixation. Such a clinical research model deserves
re-exploration.
Clinical studies of treatment of nonunions are equally challenging
in that nonunions are not common and nonunions tend to be quite
variable in their presentation. In addition, current standard methods
of treatment utilizing internal fixation and bone grafting achieve
high rates of success after the index procedure for treatment of
the nonunion, varying from an initial success rate of 75% in
nonunion of the femur treated with intramedullary nailing and bone grafting
to a success rate of 98% in nonunions in upper extremities.
If a BMP formulation were available that could be used with closed
intramedullary exchange nailing or with an open procedure in the
treatment of diaphyseal nonunions of the femur, this would be a
good model considering that the initial union rates after the index
nailing procedure for the nonunion range from 50 to 74%.
The key to most of these studies, particularly those attempting
to evaluate the rate of union and amount of bone formation over time,
is the ability to assess the bone formed in the site being treated.
Interpretation of plain x-rays is quite subjective and variable,
particularly when the interpretation of radiologists is compared
with that of orthopaedic surgeons. Study designs that can use more
quantitative methods of assessments such as CT scans, which allow
for the actual measurement of volume of bone formed, its distribution as
it affects the polar moment of inertia, and the density, must be considered.
In summary, I congratulate the biotechnology corporations that have
pursued the development of BMPs and clinical studies of their efficacy,
as well as the hundreds of surgeons and dozens of institutions that
have participated in these challenging studies. Significant results
have been obtained in a number of studies that point the way to
the development of even better-designed studies and encourage us
that fairly soon we will be using BMPs in the treatment of bone-healing
problems.