We thank Dr. Price and colleagues for their interest and the opportunity to
further clarify our approach to the internal rotation contractures that
develop in so many of these children. It was our frustration with methods
similar to the ones that they describe that led us to explore and develop the
arthroscopic approach. Approximately 70% of children who present with internal
rotation contractures will have glenohumeral
deformity2,3.
The children with centered glenohumeral joints to whom the correspondents
allude represent a minority. For the majority of patients with advanced
contractures and deformity, there is increasing evidence that the described
surgical method does not consistently relocate the glenohumeral joint and may
impede optimal skeletal development.
Van der Sluijs and
colleagues4 and
Birch and
colleagues5
independently reported on yet another method of releasing the subscapularis,
one that releases it from its insertion, aiming for a step lengthening of the
tendon while preserving the anterior capsule (when possible). Both groups of
authors reported that in patients with severe contractures it is not possible
to achieve reduction of the glenohumeral joint without releasing the anterior
capsule. In fifteen of the nineteen patients in the series of van der Sluijs
et al., release of the contracture could not be achieved without release of
the anterior capsule. In addition, Birch and colleagues postulated that in
many instances, excessive retroversion results in an external rotation
contracture once the glenohumeral joint is reduced. For retroversion of
>40°, they recommended an internal rotational osteotomy as part of the
same procedure5. In
that study, seventy of 183 patients were managed with combined soft-tissue
release and rotational osteotomy. Another recent study, by Waters and Bae,
demonstrated that procedures that avoid the anterior capsule fail to result in
glenohumeral
remodeling6, further
establishing that extra-articular procedures are incompletely effective in
such cases. So the question becomes, for children with severe contractures
and/or those with advanced glenohumeral deformity, do surgeons who do not
address the subscapularis tendon and the underlying joint capsule consistently
achieve a complete release that will allow glenohumeral remodeling? The
foregoing studies and our own suggest not.
It is not clear why the correspondents compare attempted arthroscopic
release of an extra-articular structure such as the Achilles tendon with our
procedure, but, contrary to their intention, this comparison does highlight
the appeal of minimizing surgical trauma with percutaneous and arthroscopic
approaches. The comparison also brings to light the reality that most
caretakers of these children have considerable expertise in areas other than
shoulder surgery (neurosurgery, plastic surgery, hand surgery, and pediatric
orthopaedics). Only surgeons who are experienced in shoulder arthroscopy
should consider this form of management. Our program, and others that have
adopted the arthroscopic approach, combine the efforts of a hand/peripheral
nerve surgeon and a shoulder specialist to address the complexity of many of
these problems.
As discussed in our report, the existing literature is woefully inadequate
in describing the loss of internal rotation that comes from any method of
treatment, and better methods are needed to quantify and document this
concern. This is certainly true of the clinical results reported by the
correspondents and in all other clinical series that employ similar
methodology. It is somewhat ironic that our attempt to deal with this issue
candidly has become a point of vulnerability in a field that has, until
recently, skirted the issue. It is a mistake to think that children who
undergo a release of the subscapularis from its origin have normal
subscapularis function. The truncated and atrophied subscapularis muscle can
be seen on magnetic resonance imaging and corresponds to a limitation of
active internal rotation on clinical examination. In fact, loss of internal
rotation in some of our earlier open cases exceeded that seen in many of our
arthroscopic cases. Yes, it is true that we would opt for improved internal
rotation in nearly all of our patients. However, at the present time, the
state of the art does not offer these children a perfect solution that
provides for a complete range of motion in all directions with normal
glenohumeral development. We concur with Dr. Birch et al. that a contemporary
surgical approach must achieve glenohumeral reduction for children with
skeletal remodeling potential and then must restore the functional orientation
of the arm if needed. This can be done with open or arthroscopic means, in one
or more operations.