To The Editor:
I am writing to comment on the article "Spontaneous
Repair of Full-Thickness Defects of Articular Cartilage in a Goat
Model. A Preliminary Study" (83-A: 53-64, Jan. 2001), by
Jackson et al.
The authors studied the healing responses to defects measuring
6 ¥ 6 mm that were created in the femoral condyles of goats.
After a nice presentation of the radiographic and histological results,
they concluded that defects this large do not heal spontaneously. Jackson
et al. noted that they were not aware of any previous work using
this animal model.
I would like to raise a couple of points that may be useful.
First, 6 ¥ 6-mm defects in the femoral condyles of goats do
sometimes heal spontaneously. In a study published previously in
the British volume of The Journal1,
three of seven defects had healed at six months and three of eight
had healed at twelve months after surgery. Since the study by Jackson
et al. examined only two and four specimens at these respective
time-periods, it is possible that they were led to a slightly erroneous
conclusion by failing to have a sufficient number of data points.
I am surprised that The Journal has accepted a paper where this
is the case. Surely, this principle, that it is not safe to reach
a conclusion that a phenomenon never or always happens, on the basis
of a small number of observations, should apply to all papers?
My second point relates to the aim of articular cartilage repair—restoration
of a functional weight-bearing surface. Jackson et al. mentioned
the use of finite-element stress analysis to predict elevated contact
pressures around cartilage defects, but this can only provide a theoretical
approximation of a complex situation. It is a pity that the condyles were
sectioned for histological examination without measuring the contact pressure
distribution, with use of either pressure-sensitive paper (Prescale-Fujifilm)
or the computerized Tekscan system. This measure provides a sensitive evaluation
of the functioning of the reformed surfaces that correlates nicely with
the histological evaluation. In the earlier study cited above, for
example, we were surprised to find elevated contact pressures where
some of the drill-holes had been. The histological examination showed
that this was caused by a layer of thin fibrocartilage overlying
a reformed but irregular subchondral plate. Since pressure mapping
is illuminating, relevant, and nondestructive, I would like to suggest
that it be used more widely when evaluating the healing and restoration
of articular surfaces.
D.W. Jackson, P.A. Lalor, H.M.
Aberman, and T.M. Simon reply:
The specific article that Dr. Amis (a coauthor) refers to focused
primarily on assessing biologic implants used to repair surgically
created cartilage lesions1. The
left knees of nineteen goats were implanted with three different
grafts (glutaraldehyde-fixed bovine meniscal xenograft, glutaraldehyde-fixed
bovine costal cartilage xenograft, and viable osteochondral allograft).
In all animals the same knee received a minimum of three grafts.
In fifteen of the nineteen knees, one additional ungrafted defect was
made. The ungrafted defects were 6 mm in both diameter and depth
and were made in the anterolateral femoral condyle. In Table I of
their article, Shahgaldi et al.1 determined
that, of the ungrafted lateral femoral condyle lesions, three of seven
at six months and three of eight at twelve months had "healed." They reported
that these ungrafted defects remained grossly discernible at all review
periods. These authors also reported that the contact pressures
of the ungrafted defects were not normal at any time.
Our experience with articular cartilage lesions in the lateral
femoral trochlea has shown a similar tendency toward varying degrees
of healing. In our series, we made one lesion per knee, as opposed
to three or four lesions, we introduced no foreign material into
the joint, and we chose a location that shows the least spontaneous
healing on the basis of our previous experience with multiple sites
in this model. Gross examination after twelve months suggested closure
of the defect at the superficial layer in a few of the lesions in
the medial femoral condyle.
We did not feel that any of the lesions that we created showed
substantial healing. "Healing" is a term used
differently by different authors. In our study, we purposely chose
the most difficult anatomic site in the goat knee to evaluate for
any spontaneous healing that would occur after introduction of defects
of this size. If the natural history of these surgically created
defects represented a healing process, we feel that it would be
appropriate to evaluate its functional result. Our study did not involve
assessment of contact-pressure distribution in the zone of influence because
healing had not occurred. We are not aware of previously published reports
of similar work in this location—that is, looking at the
natural history of a critical-sized defect in the middle third of
the weight-bearing surface of the medial femoral condyle. We are
aware of similar, unpublished work by other investigators that supports
our observations of lesions in this location.
Our work was an assessment of the natural history of this surgically
created lesion in a specific and potentially clinically relevant
location. It becomes the basis now for the evaluation of treatments
that may prevent the secondary changes that we observed to be suggestive
of continued degeneration. We agree with the conclusion at the end
of the article by Shahgaldi et al. that "large osteochondral
defects do not heal."1