Invited comment: A new theory of anorectal function
(D. Chatoor, A. Emmanuel - issue 4, 2008)
We have been interested to study Drs Chatoor and
Emanuel's analysis and comments on our work. We note
that they do not comment overall on the musculo-elastic
theory itself, or on the concepts underlying our series of
publications. As we explained in the preamble we sought to
test the musculo-elastic theory of pelvic function and continence
by challenging its predictions, a method proposed
as the basis of the scientific method by the late Sir Karl
Popper, and regarded as the most rigorous test procedure.
Thus the studies we reported were designed to test the musculo-
elastic theory by seeking direct tests that would refute
the theory.
For example, in Study No 1, if we failed to demonstrate
the three predicted directional movements, then the Theory
would be invalidated. Since the theory states that faecal
incontinence (FI) is caused by lax suspensory ligaments, in
this case, the pubourethral and uterosacral, and we found a
>80% cure of FI on repair of those ligaments, then the theory
remains valid, although not necessarily proven. Popper held
that nothing in science is ever finally proven, but that any
theory remains valid until it is refuted. The classic example
of this principle in Physics is Newton's Laws of Gravity,
which required modification in relation to Einstein's Theory
of Relativity, yet still remain accurate enough that they are
used to calculate satellite orbits.
We define idiopathic FI as a disorder of faecal continence
despite a normal external and internal anal sphincter, and in
the absence of any causative central neurological or other
disorder.
Chatoor and Emanuel state that "the aetiopathogenesis"
of urinary and faecal incontinence is rarely a singular process.
Insults to the pelvic floor are usually multiple (including
chronic straining, increased intra abdominal pressure
effects, parturition and the menopause)". This statement
reflects generally held opinion but is essentially phenomenological
rather than mechanistic. It does not lead to understanding
of the abnormal functional forces acting on the
pelvic floor leading to incontinence. Our musculo-elastic
theory, on the other hand leads to predictions of abnormalities
in pelvic floor function, and that their correction should
mitigate FI when present. It also offers functional explanation
of symptoms such as straining and abdominal pressure
effects as secondary manifestations of connective tissue
damage, especially to the suspensory ligaments, because
this abnormality will cause unloading of muscle forces, and
weakening of the vector acting across the pelvic floor during
its normal function.
The comments made by Chatoor and Emanuel regarding
collagen abnormalities as risk factors for the development of
FI, and probably also applicable to stress urinary incontinence,
regarding collagen abnormaities are entirely in accord with
our perspective of these problems, and are especially valuable
since they should be amenable to experimental testing.
We note that Chatoor and Emanuel have misinterpreted
some of our ideas, and comment as follows:
Experimental study No 1.
We did not set out to alter the puborectal angle, which
we agree is not in itself crucial to faecal continence. We set
out to restore the functional integrity of certain intrapelvic
ligaments, in order to restore muscle forces. Any resultant
change in the puborectal angle would therefore be secondary
to restoration of ligamentous function and muscle force
vectors acting across the pelvic floor.
We regard the outer longitudinal muscle of the rectum
(LMA) as contracting against, and angulating, the levator
plate, which is attached to the ligamentous structure via its
fascial coverings. We do not suggest the LMA merges with
the uterosacral ligaments.
We have not claimed to quantify muscle forces, but rather
to confirm the direction of these muscle forces. We agree
that the position of the ligaments is necessarily only an estimate,
but it is evident by studying the radiographic illustration
of the normal subject with the pelvic floor in the
"resting position", that there is a definite bend at midurethra,
which is consistent with the anchoring point of the three
muscle forces we have described. With regard to the uterosacral
ligaments, we compared the straining to the resting
data. The downward angulation of the tip of the levator plate
is the directly in line with the position of the cervix, the
insertion point of the uterosacral ligaments.
It is not possible to assess laxity in the living subject, even
with MRI.
Experimental study No 2.
We have indeed suggested a possible mechanism; that is,
restoration of the pubourethral anchoring point for levator
plate contraction and we provide a diagram to illustrate this
concept.
Experimental study No 3.
Chatoor and Emanuel offer an explanation of the phenomenon
described on the basis of increased sensory input
caused by digitation of the vagina, leading to an enhanced
reflex response. We prefer the notion, based on clinical
examination that this change in function was simply due to a
mechanical change associated with restitution of abnormalities
in function associated with lax connective tissue structures.
This concept is testable by more detailed experimental
work in affected patients.
Experimental study No 4.
The bladder was full of urine when the testing was done.
Although we accept that there is discussion regarding the
best techniques for measurement of abdominal pressure
changes, any criticisms regarding the methods we used
apply equally to both the squeezing and straining maneouvres,
and so would apply to both.
Experimental study No. 5.
Chatoor and Emanuel make a definitive statement "One
of the frequent causes of faecal incontinence in the elderly
is internal sphincter atrophy." This may or may not be so.
We must emphasize that we are suggesting a different
approach to understanding the functional basis of faecal
incontinence and stress urinary incontinence; i.e., that it may
be due primarily to ligamentous laxity. This concept does
not exclude a role for other factors, especially anal sphincter
tears, and even internal anal sphincter dysfunction, but we
do not think the latter is a likely cause of faecal incontinence
as a unique and solitary abnormality. We agree entirely that
we have taken an arbitrary 2 mm definition of internal anal
sphincter thinning on the basis of advice from our Radiologist.
What we did demonstrate was:
- Only a minority of patients with FI had IAS thinning.
- IAS thinning had no impact on whether such patients were cured following the tape insertion procedure.
One has to conclude therefore that IAS defect was not a
major cause of FI in our patients.
As regards the ultrasound probe, a 7 Mhz probe may be
less sensitive but its use was consistent, and this does not
therefore alter our conclusions.
Once again we point out that our observations and interpretations
are open to further experimental study, which will
verify or refute them.
Experimental study No 6.
We note the comments. However, the presenting symptoms
were relieved by the procedure, and we have presented
our observations and interpretation of the mechanism of
benefit.
Experimental study No 7.
We agree that the puborectalis is an important part of the
anorectal closure mechanism; indeed, this has been known
for more than 30 years. We regard the main function of the
puborectalis muscle as anchoring the anorectum, so that closure
of the anus can be effected by stretching the rectum
backwards and downwards. A relevant analogy might be
maintenance of urinary continence after excision of he distal
part of the urethra. In that case, we consider it is the
backward/downward stretching which narrows the urethral
tube, exponentially raising its resistance according to the 4th power law of Poiseuille.
Experimental study No 8.
The histological findings described were features that do
not occur in normal subjects, as compared with historical
data on the anatomy and histology of these perineal muscles.
Biopsies from control subjects are clearly unethical,
not least since normal subjects are not subject to surgical
procedures. We have made the point that histological features
in muscles in which muscle fibres could be detected
were consistent, in part, with changes found in limb muscles
after tenotomy. The essential feature of this abnormality is
that it is due to unloading of muscle by the tenotomy, or by
ligamentous laxity in the case of our pelvic floor biopsies,
and that these changes are reversible.
Even a weak muscle, caused by direct injury, denervation
or myopathy can function better if its ligamentous attachments
are reinforced. People with severe muscle atrophy,
as in inherited neuropathies, poliomyelitis or myopathy can
remain mobile as long as joints and ligaments are intact. It is
well-recognized that when there is stretching of ligaments,
mobility may be lost.
We agree however, that other than stating that the tape
provides a strong insertion point, we did not demonstrate
how the contractility improved.
Experimental study No 9.
A strength of this report is that it represents the experience
of a generalist gynaecologist who based his report on a preoperative
questionnaire. Symptoms are important. Gradation
of the severity of incontinence would not have contributed
to the aim of testing the hypothesis although we agree that
will be necessary in attempting to better define the indications
for the procedure. The important point to note is that
patients were cured of their FI symptoms. This report reflects
the remarkably successful results of one practitioner.
Experimental study No 10.
Chatoor and Emanuel take issue with our algorithm. This
is intended to reflect our experience. We agree that it differs
in some respects from other algorithms of pelvic floor
dysfunction and, of course, like all algorithms it certainly
oversimplifies the issues. We draw attention to the various
algorithms in the book "The Pelvic Floor", (Pemberton J,
Swash M, Henry MM, WB Saunders 2002) which also represent
summaries of ideas and practice. The unexpected
placement of certain symptoms, such as nocturia, unexpectedly
as a posterior defect represents our experience of its
resolution following a posteriorly-directed, reconstructive
tape-insertion procedure.
In reply to Chatoor and Emanuel we again point to the
results described - a large percentage of these patients were
cured or improved of their FI symptoms following operations
that only repaired suspensory ligaments. The symptoms
and their relation to the algorithm have been validated
in two ways.
- Use of 'simulated operations'; i.e., anchoring specific ligaments and observing the effect on symptoms such as urge, SI and pelvic pain.
- Tracking pre-operative symptom fate after site-specific repair to the three zones in large numbers of patients. The surgical techniques are fully described elsewhere in the literature.
We agree with Drs Chatoor and Emanuel, that the algorithm
may not have described the functional defects sufficiently
for optimal cure of all patients in this group. For example,
some patients required a second procedure directed to another
pelvic floor ligament. However, because we wished to assess
whether laxity in either the anterior or posterior suspensory
ligaments (or both) caused FI, we followed the protocol
described. The operations are minor in-patient procedures, so
that an argument can be made for repairing both the anterior
and posterior in all patients with FI.
We agree that it would have been helpful to repeat the tests
in failed cases. However, this re-examination was dependent
on patient consent.
Mean anal pressure and functional anal length were determined
by standard methods, using balloon manometry, in
the Dept of Colorectal Surgery at the Royal Perth Hospital.
Our results regarding the difficulty in securing reliability
in pudendal nerve terminal motor latencies speak for themselves.
The neurological co-author (M Swash), who introduced
this technique as an experimental method in the 1980s
never intended that it should be used in clinical practice in
individual patients, for the very reason that the length of
the terminal segment of the pudendal nerve over which the
measurement was made could not be verified and might vary
from test to test. At the time this method provided useful
verification of damage to the pudendal nerve, when treated
as a change in group data, but the standard deviation of the
grouped results was always too great for application in individuals.
Our results reported here reflected these comments.
With regard to "validated scoring sheets' we believe that
nothing could be more validating than a patient stating
she does not soil anymore. This is a yes/no response that
answers the question absolutely.
Poisseuille's Power Law may be confusing. However, we
set out to explain why some patients with no obvious anatomical
defects were incontinent, or had emptying problems.
By regarding the anorectum as a tube which is opened
or closed by muscle forces, this explanation becomes rationalized
in terms of physics. If we then understand that these
same muscle forces effectively contract against suspensory
ligaments, as do all muscles in the body, then the argument
becomes rationalized also in biomechanical terms.
Experimental study No 11.
Once again, Chatoor and Emanuel have ignored our
results - a large percentage of these patients were cured
or improved of their FI symptoms with simple out-patient,
daycase, operations which repaired suspensory ligaments.
Cystocoeles were present in a number of these patients
and required repair. This necessity afforded us the opportunity
to test whether this repair also improved the FI cure
rate; it did not.
We agree with Chatoor and Emanuel that it would have
been helpful to add subgroup data in the paper. However, the
numbers of patients involved were too small to allow this.
We agree that the question, which ligament causes which
symptom is of critical importance. Our approach as stated,
was to restore all the anatomical defects, based on the principle
of "restore the anatomy, and you will restore the function".
Further work may yet address this question.
Experimental study No 12.
Chatoor and Emanuel make no comment on the major
observation made in this report by Abendstein; that is, a
large percentage of these patients were cured or improved
of their rectal intussusception and obstructive symptoms by
repairing the posterior suspensory ligaments, thus avoiding
a major invasive abdominal procedure. From our perspective,
this observation validates another of the musculo-elastic
theory's predictions. All these patients were symptomatic
prior to surgery.
Again, we do agree that subgroup data in this study would
have been beneficial to our aims.
We believe overwhelmingly that psychological problems
in this disorder are secondary to the incontinence. It is
remarkable how many such problems disappear overnight
after successful surgery.
Final comments
In their criticisms Chatoor and Emanuel have unmasked
what is perhaps a major imperative for future research, close
collaboration with colleagues whose knowledge and advice
will provide more information, and better test the musculoelastic
theory.
We have presented information that supports a concept
that emphasizes the role of ligaments applied to
the muscle forces activating anorectal closure (continence)
and evacuation. Our aim has been to indicate a
new direction for treatment and research. We hope that
the many doubts and questions raised by Chatoor and
Emanuel will be taken up and used to further test the
musculo-elastic theory. What we have done can only
be considered a small beginning which, we trust, will
unfold into a rich new era of research in anorectal function
and dysfunction.
March 2009
MICHAEL SWASH
