Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>



									                                                         616.11-002: 616.12-008.1

    ARMSTRONG. From the Department of Pathology, Cambridge
                      (Received for publication 11th June 1940.)
FOR some time previous to the war I was engaged in an attempt to
produce an experimental constrictive pericarditis in rabbits. The
method employed followed closely that of Beck [1929, 1930], who
irrigated the pericardium of dogs with Dakin's fluid. Owing to the
smallness of the animal, however, and the delicacy of the pericardium,
irrigation was not found feasible, and a simple injection of a small
quantity of the solution was made into the pericardium. The results
were surprising, for the animals, with very few exceptions if a strong
enough solution was used, died of massive pleural effusions within
fourteen days of operation. It soon became evident that the condition
produced in these animals, although not constrictive pericarditis, bore
great similarity to the results reported by Dible and Lynch [1938].
They observed the same phenomenon after the injection of small
quantities of tincture of iodine into the pericardium of rabbits. They
attributed the effusions to two processes: firstly, to an increase in the
permeability of the pleural capillaries as a result of the aseptic inflamma-
tion of the mediastinum; and secondly, to an obstruction to the egress
of fluid from the pleura along the mediastinal lymphatics. They were
clearly of the opinion that failure of the heart played no part in the
production of the hydrothorax, and deliberately rejected it as a cause,
after mature consideration. As my own work was primarily concerned
with the heart, my attention was focussed on this organ and its dysfunc-
tion. Working on these lines, my investigation of the causes of the
hydrothorax yielded results very different from those of Dible and
                          METHOD EMPLOYED.
    The approach to the pericardium was exactly similar to that em-
ployed by Dible and Lynch. A mid-line incision was made over the
sternum, and the pectoral muscle was separated from the sternum and
costal cartilages and reflected laterally. A thin sheet of muscle-the
pectoralis minor-was then separated and retracted laterally, leaving
the ribs and costal cartilages entirely bare. Fine-pointed scissors were
used to cut the 3rd, 4th, and 5th costal cartilages at their extreme

          Downloaded from Exp Physiol ( by guest on December 8, 2011
264                                  Armstrong
 anterior ends at their attachment to the sternum. It is of great in-
 portance that they should be separated at this site, for if the cut is
 made too far laterally the internal mammary artery may be divided,
 or a pneumothorax may ensue. When the costal cartilages had been
 freed, a vertical cut w,as made with the scissors through the fibres of
 the transversus thoracis muscle and the fascia posterior to it. Con-
 siderable attention must be paid to this step, for the fascia behind the
 transversus thoracis muscle is liable to be mistaken for the pericardium,
 in which case the injection is made into the anterior mediastinal tissues
 instead of into the pericardium itself, and no result will follow the
operation. After division of this fascia the heart and pericardium will
 be seen overlaid by the thymus and a small quantity of fatty tissue.
These tissues should then be retracted headwards after separation from
the pericardium with a blunt dissector. When this is done, a clear
and wide exposure of the pericardium and heart is obtained. The
injection is best made with a small needle, the last -3 of an inch of which
has been bent to an angle of about 130°.
     The substances used for injection were three in number:
       1. Tincture iodi mitis -0 2 c.c.
       2. Dakin's solution. Full strength. 0 2 c.c. to 0-5 c.c. Owing
            to the rapid deterioration of this solution it is important that
            only freshly prepared Dakin's fluid should be used.
     * 3. Gelatine, calcium chloride, and acriflavine solution, 0-5 c.c.
            to 1 0 c.c.
    The effects of these three solutions injected into the pericardium
were practically the same. No further mention will be made of the
gelatine-calcium chloride solution: it is quoted only to confirm Dible
and Lynch's finding that a variety of irritants produce the same result.
    Immediately after the injection is made into the pericardium, the
heart is seen to dilate, and the rate slows until only a few feeble con-
tractions are taking place. Very soon the rate and force of the contrac-
tions return, but the size of the heart remains larger than normal, at
least for the short time until the chest is closed. This acute dilatation
of the heart is most marked when Dakin's fluid is injected, but is well
seen, though to a less extent, after the injection of tincture of iodine.
    Mention must also be made here of certain other technical procedures
which were used before and after operation in these animals. In a
small series of cases the size of the heart was investigated by radio-
graphy before and after operation. The method adopted was to tie
the animal down on a long plywood board, reinforced by battens at the
sides and ends. The fore-paws were held firm by a string which passed
    * Composition gelatine 5
                               per cent., calcium chloride 5 per cent., in 1/2000 acri-
flavine in normal saline. This is recommended by Hunter for producing calcification
and hyaline fibrosis in the lung.

   Downloaded from Exp Physiol ( by guest on December 8, 2011
                 Experimental Heart-failure in Rabbits                              265
over two cleats anteriorly, and the hind legs were secured by a similar
process behind. The animal was prevented from rotating or moving
laterally by two padded flanges at right angles to the main plane of
the apparatus. These flanges formed a gutter about four inches deep
into which the thorax of the animal fitted. In this way standard
conditions were obtained, and the animal was always in the same
position for the series of radiographs which were taken. The film was
placed between the animal's chest and the plywood. The animals
were all radiographed in the vertical position, as it was found that a
clearer view of the heart was obtained in this way than when the
horizontal position was used.
    In a further small series of cases arterial blood-pressures were taken
before and after operation. The method employed was that of com-
pressing the arteries of the ear by a rubber capsule connected to a
mercury manometer. The systolic pressure was taken as the lowest
point at which the column of blood was broken. Readings were always
taken after the animals had been confined in a warm box until the
ear vessels showed a maximal dilatation. In the final acute experiment
the systolic and diastolic blood-pressures were taken by a direct method,
using a cannula and a Wigger's capsule.
    Venous pressures were taken in all cases by the direct method. A
long cannula was introduced into the jugular vein and passed down to
the base of the heart. The cannula was attached to a manometer
containing citrate solution. Readings were taken when it was seen that
there was a free movement of the column of fluid in the manometer.
In most cases the pressures were taken at the base of the heart, but
in a small series pressures were taken, in addition, from the jugular
veins themselves and from other veins, including the portal. In all
cases the zero reading was taken from the level of the base of the heart,
the position of which was directly measured in each animal.

                         AND TINCTURE OF IODINE.
   In the first instance, it was the post-mortem appearances which
suggested that the condition produced after the injection of Dakin's
fluid was identical with that produced by iodine. As the morbid
anatomical findings were essentially the same, a general description of
both will be given, and two short protocols of representative experiments
will be appended.
    Immediately after recovery from the anwsthetic the animals appeared
well, and by the following day the appetite was normal. The first sign
of trouble was usually a failure of appetite, but this seldom ensued
before the 6th or 7th day after operation. Even so, the animals
appeared well until they suddenly became dyspnceic at some time

         Downloaded from Exp Physiol ( by guest on December 8, 2011
266                                 Armstrong
between the seventh and fourteenth day. The onset of marked
dyspncea was seldom survived for longer than 24 hours. Dyspncea was
obvious in degree, and the animals would be seen with their heads
thrown back and the ale nasi distending with each laboured breath.
Even at this stage it was seldom possible to demonstrate effusions with
the stethoscope or by percussion. Often, but not invariably, cedema
of the scrotum and abdominal wall and of the subcutaneous tissue of
the chest might be demonstrated. Death ensued rapidly and took
place quite suddenly. If the effusions were tapped they rapidly
recurred, and it was impossible to stave off a fatal termination by this
    Post-mortem showed frequently, but not always, cedema of the
dependent parts, particularly of the scrotum and abdominal wall;
in bucks, cedema always appeared first in the scrotum.
    On opening the chest, 20 to 50 c.c. of fluid, sometimes blood-stained,
were usually found in each pleural cavity. Often there was a dense
network of fibrin covering the pleural surfaces, particularly that of the
mediastinum. When the fibrin was stripped off the pleura, the under-
lying membrane was shiny and showed no injection of the vessels.
This was true of the whole pleura excluding that immediately covering
the pericardium, where it was usually inflamed and somewhat shaggy.
There were, in fact, no signs of pleural inflammation, except in that
part which lay directly over the pericardium. The lungs were collapsed,
and except in the very rarest cases showed no trace of cedema.
    The pericardium was usually a little thickened, but in no case was
there any evidence that it was either dense or hard enough to constrict
the heart or interfere with its action. Pericardial effusion was never
seen. In no instance was there evidence of obstruction to the great
veins, either systemic or pulmonary, as they pierced the pericardium
to enter the heart. The heart itself was sometimes normal and some-
times abnormally flabby. In the cases in which gelatine and calcium
chloride were injected the flabbiness was extreme and, after removal,
the mushroom-like collapsing of the heart left no doubt that a toxic
myocarditis was present. Microscopy of the myocardium in cases
treated with Dakin's fluid and tincture of iodine showed no definite
changes; there was certainly no fatty degeneration. In many cases,
however, a change was seen in the nuclei, which had become fragmented
and appeared to lie in vacuoles in the cytoplasm. But this change
may well have been an artefact, and it was felt that little weight could
be attached to it.
    The mediastinal tissues were cedematous, and much serum oozed
from them when they were opened. A characteristic finding was that
the mediastinal glands were greatly enlarged, turgid with fluid, and of
a pale translucent plum-colour. Microscopy showed the sinuses widely
distended with coagulated lymph containing large numbers of red cells

  Downloaded from Exp Physiol ( by guest on December 8, 2011
                Experimental Heart-failure in Rabbits                          267
and lymphocytes. Innumerable dilated tortuous lymphatics were
seen leading to and from these glands. The lymphatics were readily
visible with the naked eye and often attained the size of the normal
thoracic duct.
    Frequently, but again not always, there was an effusion into the
peritoneal cavity. The liver was always congested and nutmeg in
appearance, and sometimes local areas of yellow necrosis were seen.
Under the microscope venous congestion was not always seen, but a
wide distension of the sinusoids with cedema fluid was an invariable
finding. The intestines were always empty and distended with quanti-
ties of gas. The mesenteric and systemic veins were congested.
    This post-mortem picture was built up on the appearances found in
some 80 rabbits in which injection of one or other of the substances
mentioned above had been made into the pericardium.
    In order to emphasise the identity of the post-mortem appearances
a summary of two protocols is appended.
Rabbit 0635. Weight 2020 g.
    17/5/39. Operation: Pericardium approached by the usual route.
0 3 c.c. tinct. iodi mitis injected into pericardium. No immediate
effect on the heart was apparent.
    20/5/39. Well. Radiographed. No effusions.
    24/5/39. Well. Radiographed. No effusions.
    26/5/39. Dyspnceic. CEdematous. Radiographed. Bilateral effu-
sions. Chest tapped. 25 c.c. removed from right chest; 20 c.c. from
left chest. Given 8 c.c. urethane and killed after acute experiment.
    Post-mortem.-Gross cedema of neck, chest wall, scrotum, and
abdominal wall. Bilateral pleural effusions. Patchy collapse of lungs.
Pleura normal except over pericardium. Heart flabby and small.
Trace of oily blood in pericardium. Pericardium slightly thickened.
Liver congested and showed a red reticular pattern on the surface.
Large, congested, plum-coloured mediastinal glands. Many dilated
lymphatics in the thorax, especially a large plexus anterior to the
aorta. No ascites. Gut congested and full of gas. Mesenteric and
systemic veins engorged.
Rabbit 0691. Weight 2080 g.
    21/6/39. Operation: Pericardium exposed by usual route. 0-5 c.c.
full strength Dakin's fluid injected into pericardium. Immediate
dilatation and slowing of the heart was seen.
    29/6/39. Radiographed. Heart shadow obscured by large right
pleural effusion. 5 c.c. of fine suspension of Hydrokollag 300 injected
into the effusion.
    30/6/39. Killed after acute experiment.
    Post-mortem. Scrotum cedematous. Bilateral pleural effusions.
The Hydrokollag was loculated in a fibrinous mass just above the

         Downloaded from Exp Physiol ( by guest on December 8, 2011
268                                 Armstrong
diaphragm. Much fibrinous cobweb in both pleural cavities. Peri-
cardium lightly adherent to the heart but not thickened. Lungs
collapsed. Enormous pale mediastinal glands distended with colourless
fluid. Ascites 50 c.c. Nutmeg liver. Congested bowel full of gas.
    These two experiments are representative of the fullv developed
condition found in rabbits treated with (a) tincture of iodine, and (b)
Dakin's fluid, and they serve to show the identity of the post-mortem
appearances in each case.

     I. To test the Hypothesis of Increased Capillary Permeability
                       and Lymphatic Obstruction.
    Dible and Lynch postulated an increase in the permeability of the
pleural capillaries. They reached this conclusion because they found
that the effusions were stained blue after the intravenous injection of
Trypan Blue. They stated that the dye must have escaped from the
capillaries into the effusion, and alleged that this was evidence of
abnormal capillary permeability. Although I have never used Trypan
Blue, several experiments were made for other purposes in which
Pontamine Blue 6 BX was injected intravenously into normal rabbits
a short while before they were killed. In all cases where peritoneal
fluid was present in the abdomen there are often a few c.c. of fluid
in the normal rabbit's abdomen this was well stained by the dye.
It would appear, therefore, that the rabbit's capillaries are normally
permeable to dyes, and that the normal body fluids will become stained
after intravenous injection. The staining of the effusions cannot be
accepted as evidence of unusual capillary permeability.
    Dible and Lynch also postulated an obstruction to the egress of
fluid along the lymphatics between the pleura and the mediastinal
glands. No mechanical obstruction was ever demonstrated, but they
considered that the high protein content and viscosity of the exudate
rendered it unable to traverse the lymphatics of the inflammatory zone.
Their evidence was based upon the finding that Indian ink injected into
the pleural effusions failed to reach the mediastinal glands in four days.
Indian ink injected into the pleura of the normal animal was always
found to have reached the mediastinal glands after this interval.
    Their experiments are open to criticism owing to the extreme liability
of Indian ink to flocculate and become unabsorbable when in contact
with fluids with a high protein content. The experiment already
quoted, Rabbit 0691, is a clear indication of this. Even the Hydro-
kollag 300, which is much less liable to flocculate than Indian ink, had
become inspissated, and after 24 hours was loculated in a position
which rendered its absorption impossible. Moreover, the distension of
 the lymphatics and the size of the mediastinal glands strongly suggest

 Downloaded from Exp Physiol ( by guest on December 8, 2011
                Experimental Heart-failure in Rabbits                           269
that the lymphatics were doing more, rather than less, than their normal
share of work. The following experiment appears conclusive:
Rabbit 0570.
    22/3/39. Pericardium approached by usual route. 0-25 c.c. tincture
iodi mitis injected into pericardium.
    30/3/39. Animal killed after acute experiment under urethane.
    Post-mortem. Showed bilateral pleural effusions-20 c.c. each side.
Very slight cedema. Congested liver.
    Acute Experiment.-The thoracic duct in the neck was cannulatedc
indirectly through the jugular vein after tying off all the tributaries
and the superior vena cava. The cannula then lay in a closed venous
sac into which the only entry was the thoracic duct. Heparin was
injected intravenously to prevent clotting. A good flow of slightly
blood-stained lymph was obtained from the cannula. 10 c.c. of slightly
blood-stained fluid were then withdrawn from the left pleural cavity,
and 5 c.c. of 5 per cent. Pontamine Blue 6 BX in normal saline were
injected into the effusion in its place. Twenty minutes later the
lymph escaping from the thoracic duct was lightly but clearly stained
with blue. Blood was then withdrawn from a vein, and the animal
killed. After centrifuging this blood the serum showed no trace of
    This experiment shows clearly that dyes injected into the pleural
effusion are recoverable from the lymph in a short space of time. The
dye could not have reached the lymph by way of the blood-stream,
because at this time the serum itself was unstained. The only explana-
tion is that the effusion and its contained dye were being rapidly
absorbed through the lymphatics. It would appear, therefore, that the
path of lymphatic absorption is intact. In view of the dilatation of
the lymphatics and the distension of the glands, it is probable that the
lymphatics are carrying along their path a larger than normal volume
of fluid.

  II. Observations suggesting that Failure of the Heart is the Causal
    (a) Morbid Anatomy.-The morbid anatomical findings, when con-
sidered in their entirety, are highly suggestive of heart failure. The
constancy and predominant nature of the pleural effusions are apt to
lead one astray if due weight is not given to the other findings. This is
probably the reason why Dible and Lynch turned to other factors for
an explanation.
    In addition to pleural effusions, all the animals showed congestion
of the liver; many showed cutaneous cedema, and quite a large pro-
portion showed ascites as well. These findings in themselves should
suggest cardiac failure.
 VOL. XXX., NO. 3.-1940.                                                  19

           Downloaded from Exp Physiol ( by guest on December 8, 2011
270                                Armstrong
    At the same time as these experiments were being made, Dr. A. N.
Drury was working on the effects of arteriovenous aneurysms produced
experimentally in rabbits. The method used was to divide one internal
carotid artery and to unite the proximal end -ith the jugular vein by
side to side anastomosis. He found that a small percentage of animals
so treated developed anasarca and died suddenly. Post-mortem showed
large bilateral pleural effusions and a picture in the rest of the body
identical with that found in the animals into whose pericardia Dakin's
fluid or iodine had been injected. This correspondence extended even
to the dilatation of the thoracic lymph channels and to the enlargement
of the mediastinal lymphatic glands. There was great enlargement
of the heart, engorgement of the svstemic veins and liver, cutaneous
cedema, and sometimes ascites.
     (1) Anastomosis, 2nd February 1939. iMarch 1: Ascites suspected.
 Mlarch 8: Scrotal cedema obvious and gross ascites. March 9, 1939:
Animal killed, 500 c.c. of ascitic fluid withdrawn. Scrotal cedema
very definite. Pleural effusion, bilateral, small in amount. Thoracic
lymph channels slightly dilated. AMediastinal lymphatic gland slightly
enlarged. Liver engorged. Generalised cedema indefinite.
     (2) Anastomosis, 2nd February 1939. MAlarch 10: Slight generalised
 oedema. Killed 17th MIarch. Bilateral pleural effusion, blood-stained
and considerable in amount. Thoracic lymph channels dilated.
 Mediastinal lymphatic gland enlarged. Ascites moderate in amount.
 Liver ungorged. Gross generalised cedema.
     In these cases there had not been the slightest interference with
 the thoracic viscera, the pleural capillaries w-ere undamaged, and the
 mediastinal tissues w-ere not the seat of experimental inflammation.
 They show clearly that a condition w-hich could be none other than a
 failure of the heart, occasioned by a large arteriovenous shunt, could
 produce post-mortem appearances identical N-ith those found in the
 animals under discussion.
     (b) Venitous Pressures. While the post-mortem evidence is sug-
 gestive, the results of experiments on the circulatory dynamics of these
 animals ale quite convincing.
     In taking venous pressures the method used was to pass a long
 glass cannula down the jugular vein to the base of the heart. The
 cannula was filled with citrate solution and a-as connected to a
 manometer. In all cases the thoracic walls were intact. Readings from
 the manometer w~ere taken in every case in reference to a zero line
 passing horizontally through the base of the heart. This zero line was
 measured directly in each case after the death of the animal, the thoracic
 walls having been reinoved and the heart completely exposed.
      Readings of the venous pressure at the base of the heart were taken

Downloaded from Exp Physiol ( by guest on December 8, 2011
                  Experimental Heart-failure in Rabbits                            271
in 28 animals as normal controls. Their range, in centimetres of water,
will be seen from fig. 1. The readings were nearly always at or below
zero. While readings were occasionally as low as - 3 or even - 5 cm.
water, the highest normal reading was + 2-2; this reading occurred
only once and was never exceeded.
    In the normal animal it was noticeable that the venous pressure
fluctuated with each respiration, and a distinct movement of the column

      FIG. 1.-Venous pressures taken at the base of the heart. 1. Controls; 2. After
    injection of iodine into pericardium; 3. After injection of Dakin's fluid into

of blood could be seen in the cannula where it met the citrate solution.
In the pathological animal it was noticeable that this respiratory
fluctuation was either diminished or obliterated. This is graphically
represented in fig. 1, where two points joined by a line represent the
variations in pressure in each animal. It will be noticed that in the
pathological animals these pressure fluctuations have been damped out,
and in some entirely obliterated. In these animals fluttering pulsations
of the veins coinciding with the heart-beat were seen. Sometimes the
veins pulsated almost like arteries, and it seems likely that in these
animals an incompetence of the tricuspid valve was present, the result
of dilatation of the a - v ring.
    Ten observations on the venous pressure were made in the patho-

          Downloaded from Exp Physiol ( by guest on December 8, 2011
272                                 Armstrong
 logical animals during acute experiments after which the animals were
 killed. All had pleural effusions. It will be seen from fig. 1 that the
 venous pressure was raised to a gross extent in every case, and that the
 rise was as significant in those treated with iodine as in those treated
 with Dakin's fluid. Whereas the normal venous pressure in the rabbit
is in the neighbourhood of ± 1 cm. water, in those treated with iodine
it varied from + 4-8 to + 10 cm. water anid in those treated with Dakin's
fluid it varied from + 4 5 to + 10 8 cm. water.
     (c) Enlargement of the Heart. It has already been stated that direct
visual observation at the time of operation shows that the heart dilates
immediately the irritant is injected into the pericardium. This acute
dilatation is greater when Dakin's fluid is injected than after the use
of iodine.
     It was decided, therefore, to ascertain whether there was any per-
manent increase in the size of the heart. For this purpose the animal
was radiographed before operation in order to obtain a control tracing
for each individual. Following operation it was radiographed again,
and in a few animals a series of plates were taken on successive days.
The outline of the heart was then traced from the film on to a strip of
transparent cellophane, and finally transferred to paper. The results
are seen in fig. 2.
     In these figures the outline of the heart after operation has been
superimposed on its outline before operation in order to show the
difference in size. No attempt has been made at accurate super-
imposition, so that the tracings only give information about general
enlargement of the heart. They give no indication of the chamber of
the heart which was enlarged. It must be emphasised, however, that
the method of radiography employed ensured that the animal was
in the same position when each photograph was taken. It is certain,
therefore, that each tracing provides accurate information about the
absolute size of the heart at any given time.
     It wTill be seen from fig. 2 that in nearly every case a significant
increase in the size of the heart was demonstrable after operation-
indeed it was seen at the operation itself and it persisted until the
death of the animal. As pericardial effusions were absent at post-
mortem it is clear that the increase in the cardiac shadow was due to
enlargement of the heart itself.
     Increase in the size of the heart may be caused by two conditions,
namely, hypertrophy and dilatation. As it is clear that there was
insufficient time for hypertrophy to develop, there is no doubt that in
these animals the increase in the size of the heart was caused by
     It may perhaps be argued that the rise in venous pressure was
not the cause of the pleural effusions but rather resulted from them;
that the increase in intrathoracic pressure due to the effusion was the

 Downloaded from Exp Physiol ( by guest on December 8, 2011
                   Experimental Heart-failure in Rabbits                               273
immediate cause of the rise in venous pressure. Unfortunately, no
readings of the venous pressure at the base of the heart could be taken
                           'r,,-A- 1,r.5
                                ,Xz                        :Z'90(los F'Lut3

                            .M75. PS,.
                                S.   S                 7        FLU,-0

                          T2'ft7&. Vf. q.l
                                                       7   D'71 vP
                                                                 p    6q

                                       v.?. 'oa        V0P. +0.

                                                      7DX7,V. P.++.o ia

                            ,   -8DYs.vL P.

                                                      8Dy4S.   V.P. +I,S
      FIG. 2. Tracings of rabbit hearts taken before and after operation. 1. Tracing
    from the normal animal before operation; 2. Tracing taken after operation. In
    each case the number of days after operation is indicated, followed by venous
    pressure value in cm. of water. The venous pressure recorded at the time of
    the final acute experiment is given in each case and is expressed in centimetres
         of water. The same venous pressures are recorded graphically in fig. 1.

before the onset of the effusions without killing the animal, and there is,
therefore, no information as to the time of appearance of these two


            FIG. 3.-Rabbit 0623. 1. Tracing of the heart before operation.
16/5/39. 0 5 c.c. Dakin's fluid injected into pericardium.
20/5/39. Radiographed. Tracing No. 2 made from film. Radiography at this time
           showed no effusion.
24/5/39. Radiographed. Heart still enlarged. Small right-sided pleural effusion.

factors. On the other hand, there is no doubt that the dilatation of
the heart preceded the onset of the effusions. Fig. 3 shows that in the
case of Rabbit 0623 there was marked dilatation of the heart four days

            Downloaded from Exp Physiol ( by guest on December 8, 2011
274                                 Armstrong
after operation, but that signs of effusion did not appear until a further
four days had elapsed. It is contended, therefore, that the combination
of cardiac dilatation and a rise of venous pressure, when taken in con-
junction Nith the post-mortem findings, offers unquestionable evidence
that the condition of these animals is occasioned by failure of the heart.
The absence of significant histological changes in the heart muscle
wN-hich caused Dible and Lynch to discard the hypothesis of cardiac
failure is undisputed, but it is felt that the present evidence from the
living animal far outweighs this consideration and demonstrates clearly
that cardiac failure is indeed the cause of the syndrome.

               III. Observations made on the Preparation.
    The foregoing remarks will have made it clear that a condition
closely resembling congestive cardiac failure in man can be readily
induced by experimental procedures in the rabbit. The preparation
offers an excellent opportunity for the study of the dynamics of cardiac
failure, and a number of investigations had been planned and would
have been carried out had it not been for the outbreak of war. The
few observations made, however, are perhaps worthy of record.
     (1) Dimination of the Pressure Gradient in the Great Veins. The
pressure in the jugular vein of the rabbit may easily be taken by inserting
the cannula into a large branch which joins the vein some distance
below the angle of the jaw. If the point of the cannula lies in the
branch just above its junction with the main vessel, the pressure reading
taken will represent the lateral pressure in the main trunk itself. From
this position the cannula can be rapidly slipped down the lumen of
the jugular vein to the base of the heart, and a reading can then be
taken in this situation. In the normal animal there is a well-marked
pressure gradient between the jugular vein and the entry of the great
veins into the heart itself. In the few normal animals that were studied
the difference of pressure between these two situations was 538 cm.
 w7ater. Five observations were recorded on rabbits in which an experi-
 mental failure of the heart had been induced (Table I). In all cases
 there was a very marked diminution in the pressure gradient between
 the jugular vein and the base of the heart; in three animals the gradient
 had been so decreased that it w-as impossible to measure it. The
 pressures were high in both situations, but the difference bet-een them
 had been greatly diminished.
     This observation is of importance as it helps to explain the nature
 of the failure of the circulation on the venous side of the heart. It is
 obvious that the venous return depends not only upon the hydrostatic
 pressure in the veins, but also upon the pressure gradient which exists
 along the course of the venous tree. No matter how high the venous
 pressure may be, no blood will flow towards the heart unless the pressure

 Downloaded from Exp Physiol ( by guest on December 8, 2011
                  Experimental Heart-failure in Rabbits                                275
in the peripheral veins is higher than that in the great venous trunks.
The velocity of blood-flow in a vessel varies directly with the pressure
gradient which occurs along its course, and any diminution of this
gradient will occasion also a decrease in the velocity of blood-flow.
It follows, therefore, that in these animals suffering from heart failure
there must have been a greatly diminished rate of inflow of blood into
the heart in spite of the rise in venous pressure.

            Normal animals.                          Animals in heart failure.

 Jugular    Base of   heart.   Difference.   No.      Jugular    Base of   Difference.
         vein.                                    ven.vein,       heart.

   6-9       - 1-3 to - 0 9        8-0       0692      +12        + 10 0         2.0
   8-2       +±22 to +2-7          5-8       0691      + 9-6      + 9.6          0
                                             0663      + 4.3      + 4-3          0
   6-3       +0-3to +0-8           5-8       0665      + 7.0      + 6.5          0.5
   6-0       +0-3 to +0-8          5-5       0664      + 8-0      + 50           3-0

    (2) Maintenance of the Arterial Pressure while the Heart is Failing.
It is well known in clinical medicine that a failure of the heart
does not always occasion a fall in systemic arterial blood-pressure.
The preparation under discussion conforms to this finding, for in all
of six animals which were investigated from this standpoint no signi-
ficant fall in arterial pressure occurred during the period of cardiac
    The method adopted for measuring the arterial blood-pressure
without destroying the animal has already been briefly described. It
provides a figure for the systolic blood-pressure only. It is, of course,
not a method of great accuracy, but the experimental error is probably
not greater than ± 10 mm. Hg. In all cases a minimum of four readings
was made and the mean of these taken as the final reading. Readings
were usually taken on four successive days before operation in order to
establish a control for the animal under observation. Readings were
then taken at varying intervals after operation. At the final acute
experiment the systolic and diastolic pressures were taken by the
direct method after cannulating the carotid.
    Representative results of two experiments are plotted in fig. 4
and show that up to and including the day of the final acute experiment,
when the animals were proved to be in failure, no significant fall in
arterial pressure had taken place.
    It would seem, therefore, that in the heart failure manifested by
these animals there was a redistribution of blood to the venous side.

           Downloaded from Exp Physiol ( by guest on December 8, 2011
276                                     Armstrong
This was reflected in the rise of venous pressure. The velocity of the
blood in the veins was also slowed, and its inflow into the heart was
impeded. These deductions follow from the reduction in the venous-
pressure gradient. Yet the systemic arterial pressure remained at
its normal level, suggesting that the left ventricle was doing its work
with unimpaired efficiency. The absence of either oedema or con-
gestion of the lungs supports this hypothesis. It is suggested, therefore,
that in these animals there is a failure of the heart which is much more
marked on the right than on the left side. It is, in effect, a prepon-
derating right-sided heart failure.

                                          1 2   0                     j   WSYSTOLI|C


        80                                                            066D4

              PRE-OPERATIVE         1    2 5 A 5        6 7 8 9            10
             FONTROL   READING5         DAYS AFTER      OPERATION

                 Fic.. 4.-Arterial pressuires before andI after operation.

    It is possible that the irritant injected into the pericardinin produced
a toxic myocardial degeneration of the superficial fibres of the heart.
If this is true, the effect would be felt over a definite depth of cardiac
muscle all over the surface of t,he heart. As the right ventricle is much
thinner than the left, a relatively greater proportion of its thickness
would be damaged by the irritant, and a correspondingly greater
degree of failuire of the r-ight ventricle would be manifested.
    (3) Observ~ations on the Protein Content of the Effusions.-It w%Aas
thought at first that information might be obtained either for or against,
heart failure as a cause of the syndrome from a study of the protein
content of the effusions. If the effusions were inflammatory in nature,
as suggested by Dible and Lynch, it was thought likely that they would
show the characters of an exudate, with a high total protein content
and a relatively low albumen-globulin ratio. If, on the other hand,
heart failure was the cause, it was thought that they would prove to
be transudates with a low protein content. All estimations were done
by the micro-Kjeldahl method.
    As has been shoin elsewhere, there is no doubt that the cause of the

 Downloaded from Exp Physiol ( by guest on December 8, 2011
                   Experimental Heart-failure in Rabbits                            277
condition was a state of failure of the heart. There is, therefore, no
doubt that the effusions were in fact formed by a process of "transuda-
tion." As will be seen from a study of Table II., the protein content

                                         Total     Albumen.     Globulin.    A./G.

   0360   Dakin's fluid    {Pleural
                                                                   1F06       5 55

   0361   Dakin's fluid    fSerum
                           0Ascitic       5-61
                                          4*13        4.34
                                                      3432         1D27
                                                                  0 81        3.4
                                                                              4 09

   0373   Dakin's fluid     (Pleural
                          a s
                             Ascitic      3-10
                                          3-70        2-51
                                                      3 08        0.59
                                                                  0 62        4-25

   0570   Iodine           (Serum
                           ~Pleural      5 52
                                         3.77         424
                                                      3.12        1-28
                                                                  0-65        33
                         (Serum          3-67         2-60        1-07        2-4
   0571   Dakin's fluid - Pleural        2-45         1-75        0 70        2-5
                          Ascitic        2 87         2-11        0 76        2-8

  0623    Dakin's fluid    (Serum
                                                      3 04
                                                                  1109       2-7

    03 Iodin
  0635            Serum
          IoiePleural                    5-52
                                         3-01         4244
                                                      2-24          471
                                                                  1 08
  0379      calcium            Pleural   3 09         2-64        0 45       5-9
            chloride           Ascitic   3-89         3-27        0-62       5J2

of the effusions, both peritoneal and pleural, if taken alone, would have
led one to believe that they were inflammatory in nature and caused
by a process of "exudation." The total proteins were high, and in
those cases in which simultaneous observations were made on the
serum the effusion proteins and the albumen-globulin ratio often
approached the figures for the serum itself. Moreover, fibrinogen was
always present, and, to judge by the amount of clot, was often found in
quite large quantities; it was never directly estimated. This again
is no bar to the diagnosis of transudation, for fibrin is present even in
the fluid normally found in the rabbits' peritoneum. Several estima-
tions were made by the same method on this normal fluid, and the
results are tabled in Table III. It will be seen that, in comparison
with the pathological effusions, it is much less rich in protein. Never-
theless, its protein content is higher than might have been expected,

          Downloaded from Exp Physiol ( by guest on December 8, 2011
278                                 Armstrong
and at least reaches the figure which would normally be considered a
    PER 100 C.C.

           No.     Total Protein.    Albumen.      Globtulin.     A.CG.

          0284          1-8             1-31          0-49          2-6
          0298          1-38            1-41          0-23          4-9
          0302          0-96           0-66           0 30          2-2
          0281          1-65            1-32          0-33          3.9
          0294          1-68            1-23          0-45          2- 7

    The logical conclusion to draw from these figures is that no hard
and fast line, such as is usually accepted, can be drawn between exudates
and transudates. The value of protein estimations in differential
diagnosis is of less value than is generally believed.
    This view is not out of keeping with the general teaching of
physiology and pathology, for in reality both are produced by the
same physical principles. Assuming a constancy in the value of
the plasma proteins, the factors at work in producing effusion are the
capillary pressure and the capillary permeability. In heart failure
and in inflammation these are both increased. In inflammation the
capillary pressure rises because of vascular dilatation, and the per-
meability is increased by the action of bacterial toxins; in heart failure
the capillary pressure rises because of the rise in venous pressure, and
the permeability is increased by anoxaemia resulting from reduction
in the velocity of the blood. On theoretical grounds there would seem
no reason to doubt that a severe rise in the venous pressure and a
severe slowing of the circulation can produce the same mechanical
factors as inflammation itself. While it cannot be denied that some
effusions in their protein content arc characteristic of the process of
transudation, and others, of inflammatory nature, are typical of exuda-
tion, it must not be forgotten that the one type merges imperceptibly
into the other. The overlap is of such dimensions that in all these
animals differential diagnosis from the effusion itself was quite
                     SUMMARY AND CONCLUSIONS.
    1. The effect of injecting Dakin's fluid and iodine into the peri-
cardium of the rabbit is described.
    2. The resulting changes are shown to be caused by heart failure,
and not by inflammation and lymphatic obstruction as heretofore

 Downloaded from Exp Physiol ( by guest on December 8, 2011
                  Experimental Heart-failure in Rabbits                         279
    3. The preparation is a simple method of producing experimental
failure of the heart.
    4. In these animals, in addition to a rise of venous pressure, a
reduction of the pressure gradient in the veins was demonstrated.
    5. During the course of the heart failure the arterial pressure
remained constant.
    6. From investigation of the protein content of the effusions the
conclusion was drawn that too rigid an interpretation of the characters
of transudates and exudates cannot be accepted. In the type of heart
failure under discussion the effusions have the characteristics normally
ascribed to exudates.
    This work was carried out during the tenure of a Post-graduate
Fellowship of the Medical Research Council. I wish to express my
thanks to the College Council of Trinity College for their additional
financial assistance.
    For the chest radiographs I am indebted to Professor Harris for
permission to use the X-ray apparatus of the Cambridge Anatomy
    My thanks are also due in large measure to Dr. A. N. Drury for his
assistance and guidance and for much helpful criticism.

BECK, C. S. (1929). Arch. Surg. 18, 1659.
BECK, C. S., and GRISWOLD, R. H. (1930). Ibid. 21, 1064.
DIBLE, J. H., and LYNCH, G. A. C. (1938). J. Path. Bact. 46, 271.
HUNTER, R. A. (1931). Tubercle, 12, 204.

         Downloaded from Exp Physiol ( by guest on December 8, 2011

To top