William Osler.

Modern medicine : its theory and practice, in original contributions by American and foreign authors (Volume v. 5) online

. (page 78 of 126)
Online LibraryWilliam OslerModern medicine : its theory and practice, in original contributions by American and foreign authors (Volume v. 5) → online text (page 78 of 126)
Font size
QR-code for this ebook

over the liver and spleen, and this may be so noticeable that the condition
was graphically described by Curschmann^ as Zuckenjussleher (iced liver),
and is sometimes spoken of as universal chronic perihepatitis; these names
suggest a more intimate connection with the liver than really exists, since
more careful examination would show that, although more prominent over
the liver, the chronic peritoneal change is not confined to that region.

The pericardial pseudocirrhosis described by Pick^ is probably allied
to this form of chronic peritonitis; Pick, however, believed that a latent
adherent pericardium gives rise to circulatory disturbance in the liver,
hepatic fibrosis, ascites, and that, although as a result of the ascites and
repeated tappings, chronic peritonitis might be induced, it was more or
less accidental and a secondary event. Nicholls^ and Kelly^ have drawn
special attention to the association of chronic peritonitis with similar intra-
thoracic changes.

2. The association with arteriosclerosis and granular (arteriosclerotic)
kidneys was insisted on by Hale White,'^ who found it in 19 out of 22 cases
of universal perihepatitis and peritonitis, but some observers, especially
Nieholls, do not attach much importance to its etiological influence. The
relation between arteriosclerosis and the peritoneal change may be either
(a) direct; it has been suggested that arteritis in the small vessels of the
peritoneum sets up a chronic fibrotic change. Another hypothesis is that
the ursemic poisons, circulating in the blood of patients with arteriosclerosis
and the resulting renal disease, produce chronic inflammatory changes in
the serous membranes; or (6) indirect; the diminished bactericidal power
of the blood, shown by Flexner^ to be common in cases of chronic renal dis-
ease, may allow infection with organisms of no great virulence to take place.

3. Simple chronic peritonitis may be met with in combination with
various intra-abdominal lesions, or its starting point may be quite obscure.
The associated lesion, such as syphilitic disease of the liver, may reduce the
resistance of the peritoneum, or, as in the case of chronic duodenal ulcer
or cholecystitis, may provide a local focus for peritoneal infection. In such
cases a widespread chronic inflammatory process results instead of the local
reaction which would be expected; this may be compared with the formation
of large cheloids after minute injuries, and possibly the excessive reaction
may depend on a low resistance, inherent or acquired, of the peritoneum
in that individual.

Alcoholism has been regarded as a factor, and would reduce the resistance

1 Deut. med. Woch., 1884, x, 564.

2 Ztschr. f. klin. Med., Berlin, 1896, xxix, 385.

^ Studies from the Royal Victoria Hospital, Montreal, 1902, i, No. 3.

^ American Journal of the Medical Sciences, Philadelphia, 1903, cxxv, 116,

^ Transactions of the Clinical Society of London, 1888, xxxi, 219,

" Journal of Experimental Medicine, 1896, i, 559.


of the peritoneum and by setting up oastro-intestinal inflammation favor
microbic infection; of the infiuenee of otiier cheniieal j)()i.sons, such as lead,
there is no evidence. The chronic venous engoroenient due to heart and
other obstructive diseases would act in much the same manner. The rela-
tion of ascites to the chronic peritonitis is usually considered to be that the
chronic peritonitis causes the ascites, but it is conceivable that in some
circumstances the reverse may hold good; thus, in chronic ascites due to
backward j)ressure of heart disease, rej)eated tappings may favor a low
form of infection, which is not necessarily introduced from without, but may
be carried by the blood stream. That tapping may lead to chronic inflam-
mation is shown by examination of the cells found in ascitic effusions; an
effusion which at the first tapping shows a j)re})on(lerance of endothelial
cells, and is therefore mechanical in origin, may in subsequent tapjMugs
show an altered cell count due to superadded changes (Cade,^ Gilbert and
Villaret^). Chronic peritonitis set up in this way may be at first local and
in the neighborhood of the puncture, and gradually become more widespread
and possibly universal.

In some cases traumatism appears to be the exciting factor, and it is quite
conceivable that by damaging the viscera it would allow of a low form of
infection of the peritoneum. Lastly, no obvious cause may be forthcoming.

Pathogeny. — Since the process is evidently a chronic and progressive
form of inflammation, it has naturally been surmised that it is the result of
infection with microorganisms of a low grade of virulence. Nicholls sug-
gests the Bacillus tuberculosis, Bacillus coli, and Bacillus typhosus. Louis
(1825) regarded chronic peritonitis as always tuberculous; in modern times
Picchini (1891) argues that subacute and chronic polyorrhymenitis is prac-
tically always tuberculous, and F. Taylor, who quotes him, accepts this in
a modified way. The recognition in recent years of a special form of tuber-
culosis, the chronic hyperplastic, which is best known in connection with the
ileocsecal part of the intestine, but has also been recognized in the ureter
and female urethra (Hartmann^), in burste, and elsewhere, suggests that
many cases of chronic peritonitis are tuberculous; in other words, that
there may be a chronic hyperplastic tuberculous peritonitis, which bears
the same relation to ordinary tuberculous peritonitis that chronic hyper-
plastic tuberculosis of the intestine does to ordinary intestinal tuberculosis.
As bearing on this view, it may be pointed out that, by means of Jousset's
method of inoscopy, tubercle bacilli are now found in many cases of ascites
which do not otherwise suggest peritoneal tuberculosis, and that tuberculosis
with fibrotic changes is a not uncommon complication of arteriosclerosis.
The last point may help to explain the association of arteriosclerosis and
granular kidney with simple chronic peritonitis.

Morbid Anatomy. — The appearances in simple chronic peritonitis
vary considerably from the more marked changes of the "iced" organs to
a general slight opacity of the peritoneum. In a well-marked case the peri-
toneum is covered by a thick membrane which resembles a thin layer of
cartilage; it is dull white or glistening, and pearly in color, and may show
pigmentation. This dense layer, especially over solid organs, such as the

* Arch, de med. exper. et d'anat. path., Paris, 1906, xviii, 769.
2 Compt. rend. Soc. biol, Paris, 1906, Ix, 820.

^ Mem. et bull. Soc, chir., Paris, 1906, and Ann. gyn. et d'obstet., Paris, 1907,
biiv, 1.


liver and spleen, for which it forms a firm casing, mjiy show a few c](;pressions
like the pitting of sand by rain-drops. It is possibl(! that th(;se deficiencies
are the result of rupture of the dense fibrous tissue;, induced by its own con-
traction. On the other hand the thick membrane may be locally accentu-
ated in the form of fibrous elevations or nodules, perhaps more often seen
on the parietal peritoneum; to this condition the name "peritonitis fibrosa"
has been applied; these fibrotic elevations when small may closely resemble
miliary tubercles, or when large, nodules of disseminated malignant disease.

The thick membrane, which may be even one-quarter of an inch in depth,
can usually be peeled off the liver and spleen, leaving their capsules in a
fairly normal condition. From its marked tendency to cicatricial contraction
it squeezes the organs and leads to atrophy; in extreme instances these
organs may therefore become so atrophied as to be isolated in the abdominal
cavity and lose their normal relations. The omentum becomes shortened
and rolled up into a firm cord, often in irregular knots, running transversely
across the abdomen; in some instances hardly a trace of the omentum is left.
In like manner the mesentery becomes contracted, so that the small intestines
are anchored to the spine and cannot float in the ascitic fluid, "and, if a
hernia had existed, it will sometimes be found to have been completely
reduced" (Hodgkin'^). At the same time there may be such considerable
diminution in both the length and diameter of the alimentary canal that
the valvulae conniventes become very closely placed — Prof. W. H. Welch
has shown me a specimen in which the longitudinal section of the intestine
looked like a comb — and the colon may be no larger than the small intestine.
This gives rise to impaired absorption. The condition has been called
"peritonitis deformans." Subacute or acute peritonitis may supervene.

Adhesions are often present, but their extent varies greatly. They are
more often seen over the solid and relatively fixed organs, since the peristalsis
of the intestines interferes with their formation and may rupture them when
formed. The remains of adhesions may be seen as filamentous or villous
tags on the small intestine, and sometimes resemble miliary tubercles.
Adhesions are not uncommon between the liver or spleen and the diaphragm,
and may be dense or delicate. The presence of definite cords may give rise
to acute strangulation of the intestine, but this is rare in this form of chronic
peritonitis. When there are sufficient adhesions the ascitic fluid may become
encysted, and when the adhesions are so numerous as to obliterate a large
area of the peritoneal cavity fluid effusion may be slight in amount.

The liver and spleen are often described as "iced," or as showing universal
chronic capsulitis, and in the case of the liver as "Zuckergussleber." This
condition, also called universal chronic perihepatitis, was formerly regarded
as a distinct morbid entity, and many cases of chronic ascites, in reality
due to chronic peritonitis, were regarded as the result of chronic perihepatitis,
the existence of the more widespread peritoneal change being ignored or
regarded as a secondary result of the recurrent ascites and consequent tap-
pings. From the frequency of ascites in hepatic cirrhosis the liver was con-
sidered the important factor in the production of ascites, and disease of its
capsule was therefore accepted as a sufficient cause of ascites without
any reference to the peritoneal change. This was not unnatural, since in

* Lectures on the Morbid Anatomy of the Serous and Mucous Membranes, i, p. 152,
1836, London.


iiianv cases the change is iiuich more manifest over the Hver than elsewhere.
On the Hver and spleen the thickened membrane is confined to the serous
surfaces, and does not affect the areas between the reflections of peritoneum,
such as the coronary ligament. The thickness of the mcjnbrane varies in
dili'erent parts, it modifies the shape of the viscera, rounds off the edges, and
covers up natural depressions. The gall-bladder may be almost buried,
and is usually collapsed, but consA-iction of the bile duct or portal vein,
contrary to what might naturally lie expected, is very rare. In a case at
one time under my care prevertebral chronic peritonitis leading to obstruc-
tion of the lymphatics above the level of the rece[)taculum chyli and so to
elephantiasis was found at the autopsy (Bernstein and Price^). The organs
are compressed, and usually show chronic venous engorgement, but any
consideral)le extension (jf fibrosis from the capsule into the substance of the
organs is very rare.

MicroficopicaUii there is extremely dense fibrous tissue laid down in parallel
lamelljE, with a few nuclei between. The fibrous tissue is prone to undergo
hyaline change, hence the condition has been called hyaloserositis (Nicholls) ;
the structure of the fibrous tissue is the same as that of the so-called lamellar
fibromas of the spleen, which, indeed, are localized foci of the same process.
The deeper layers of the membrane show bloodvessels with collections of
leukocytes and mast cells. Nicholls was unable to find any evidence of
fibrin. As the wrinkled peritoneum can be seen underneath, the formation
of the membrane is due to organization of exudation, and not to hyperplasia
of the peritoneum.

Clinical Picture. — The condition most commonly occurs in middle
life, but it may be seen in children; the sex incidence is about equal.

The cases may present different manifestations, but the characteristic
result is chronic ascites, which requires numerous tappings and recurs
with remarkable persistence. The interval between successive tappings
may diminish in progressive cases so that paracentesis may become necessary
every week; more than 100 tappings may be required, and in Rumpf's
cases the abdomen was tapped 301 times in sixteen years. After tapping,
the rolled up omentum may be clearly felt running across the abdomen, and
is often more distinctly palpable toward the left hypochondrium. It must
not be regarded as the margin of an enlarged liver, or as the greater cur-
vature of the stomach infiltrated with carcinoma. The other signs of ascites
will be found. The abdomen when distended may be almost uniformly
dull, since from retraction of the mesentery the small intestines are tethered
to the spine and unable to reach the abdominal wall. In rare instances
friction is audible through a stethoscope applied to the abdominal walls.
In some cases the fluid is encysted and the abdomen is irregular, so that the
diagnosis may be very difficult. Signs of concomitant disease, such as
adherent pericardium, valvular disease of the heart, pleural effusion, or renal
disease, may be present. The urine is somewhat scanty and may present
the characteristics of granular kidney. Jaundice is very rare, and when
present is due to some complication.

Symptoms. — The onset is, as a rule, gradual; but symptoms may date
from a febrile attack, and are usually vague. The first thing noticed is
sweUing of the abdomen followed by abdominal uneasiness and a sense of

> British Medical Journal, 1907, i, 617.


fulness and weight, but pain is absent or slight and only brought on by
movement. There is naturally some loss of strength, and the muscles are
flabby, but emaciation is slow in appearing. Patients may feel well and
except for the abdominal embarrassment, able to be up. Respiration may
be somewhat hampered, constipation is usual and no doubt in part depends
on the weakened state of the abdominal muscles, and there may be some
flatulency and dyspepsia. Gastro-intestinal hemorrhage does not occur.
In the later stages considerable wasting and oedema of the feet supervene.
It is noteworthy that oedema of the feet occurs long after ascites has been
well established.

The fluid varies to some extent in its characters, usually it is like serum,
of a citron color, and of a specific gravity of 1015 or more. It may clot
and form threads of fibrin or large flocculi. In some instances, in which
subacute inflammation has supervened, it is turbid from the presence of
cells. It may be milky or chyliform, and after tapping may become blood-

Diagnosis. — This rests mainly on the presence of chronic and recurrent
ascites, for which no other cause is forthcoming. It must be distinguished
from other causes of ascites, such as tuberculous and malignant peritonitis
and hepatic cirrhosis. It is probable that in the future cytological examina-
tion of the ascitic fluid will render this comparatively easy.

Prognosis. — This should be guarded; recovery is rare, but the course
is slow, and as many as 100 tappings may be required. Some cases of the
recurrent ascites which are diagnosed as simple chronic peritonitis certainly
recover after a number of tappings. Nicholls found that the duration varied
between two and sixteen years. The average duration of Ramsbottom's
9 cases was 624 days.^ Death is generally due to some intercurrent infection,
such as pneumonia, but in cases associated with adherent pericardium and
chronic mediastinitis it may be due to cardiac failure.

Treatment. — Treatment should, in the first instance, be directed to
the cause, such as renal disease, heart affections, syphilis; otherwise it is
palliative and symptomatic. No permanent good can be anticipated from
laparotomy and the attempt to produce artificial adhesions on the lines of
the Talma-Morison operation for the relief of the ascites of portal cirrhosis
of the liver. Paracentesis should be performed when required; often this
gradually becomes more frequent and may be necessary every week. Injec-
tion of adrenalin chloride, one dram in an ounce of distilled water, into the
peritoneal cavity before the cannula is withdrawn is worth a trial.

Diuretics, such as caffeine, may be employed from time to time, and tonics
will usually be necessary. Good hygienic conditions and fresh air are desir-
able. It IS not necessary for the patient to remain in bed in the earlier
stages. Some relief to the feeling of weight in the distended abdomen may
be obtained from the use of a binder or belt.


The ordinary clinical heading of tuberculous peritonitis covers cases with
somewhat different morbid lesions; thus, although tuberculous inflamma-

^ Medical Chronicle, Manchester, 1906, xlv, 7.


tion confined to the peritoneum is, strictly speaking, nil that is implied, it is
customary to include cases which have in aildition tul)crcuk)us ulcers of the
intestine and tuberculosis of the intra-abdominal lymj)hatic fjlands, since
these conditions are commonly combined and often cannot be separated
durino; life.

The incidence of peritoneal infection in tuberculosis has been differently
estimated. In 531 cases of pulmonary tuberculosis examined after death
at the Brompton Hospital, there was tuberculous peritonitis in 22, or 4.1
per cent. (Fowler and Godlee^). In 197 cases of tuberculosis at the Boston
City Hospital the peritoneum was infected in 14, or 7.1 per cent. (Bottomley^).
in 1393 cases of tuberculosis at Breslau there was tuberculous peritonitis
in 226, or 16.2 per cent. (Borschke^). In 300 cases of fatal tuberculosis of
various parts of the body at St. George's Hospital the peritoneum was
afi'cctcd in 56, or 18.6 per cent.

Etiology. — Age. — The disease may occur at any period of life, but it
is rare in babies and in old age. Osier records a patient aged eighty-two years,
while among 59 cases of tuberculosis in children tabulated by Colman*
the youngest case of fatal peritoneal tuberculosis was one and a half years.
It is most frequent between twenty and forty years of age. The belief
sometimes expressed, that the disease is commoner in children, is probably
the result of confusion between chronic intestinal disturbance with a dis-
tended abdomen, formerly spoken of as "tabes mesenterica," and true peri-
toneal tuberculosis. In 306 cases of tuberculous peritonitis under the
age of fifteen, nearly half were between the ages of three and seven years,
and the sex incidence was ecjual (Faludi^).

Sex. — There is a marked divergence between statistics obtained from
postmortem records and those of operation as to the sex incidence. Post-
mortem records show that it is more frequent in males; in Konig's 107
cases, 89 were males. In operation cases females are greatly in excess;
according to Nothnagel, whose estimate is probably based on Konig's series
of operation cases, 90 per cent, are in women; of Osier's" 131 cases, 60, or
54 per cent., were women. The preponderance of female cases discovered
at operation appears to depend on the frequency with which tuberculosis
of the Fallopian tubes serves as the starting point for peritoneal tuberculosis.

Disposing Factors. — Cirrhosis of the liver shows a decided tendency to be
complicated by peritoneal tuberculosis ; thus in 584 cases of hepatic cirrhosis,
obtained by combining the statistics of Lancereaux, Kelynack, Yeld, Rol-
leston and Fenton, tuberculous peritonitis occurred in 53, or 9 per cent.
No doubt concomitant or antecedent alcoholism favors the incidence of
tuberculosis generally, but the catarrhal condition of the intestine and the
portal obstruction probably play a part in the localization of the infection.
Injury has, in some cases, preceded the onset of symptoms. Pregnancy,
according to H. Kelly, ^ has a definite causal relation to tuberculous peri-
tonitis; in 28 per cent, of his cases the illness was dated from childbirth or
a miscarriage. Heredity would naturally be expected to be of some impor-

' Diseases of the Lungs, p. 358, London, 1S98.

2 Medical and Surgical Report, Boston City Hospital, 1900, p. ILS.

^ Virchow's Archiv, 1892, cxxvii, 121.

* British Medical Journal, 1893, ii, 740. ^ Jahr. f. Kinderh., 1905, Ixii, 304.

^ Johns Hopkins Hasp. Rep., 1891, ii, G7. ' Operative Gynecology, 190G, ii, 237.


tance in the incidence of tuberculous peritonitis, hut in 25 cases investigated
by Bottomley there was a history of tuberculosis in 2 only.

Bacteriology. — Peritoneal tuberculosis has been regarded by Raw^
as due to infection with bacilli of bovine tuberculosis absorbed from the
alimentary canal and introduced with the food. He considers that pulmo-
nary tuberculosis and the secondary intestinal ulceration which occurs as
the result of swallowing sputum in at least 50 per cent, of fatal cases of chronic
pulmonary tuberculosis, are due to infection with the bacillus of human

Tubercle bacilli are comparatively seldom found in the peritoneal effusion
by ordinary microscopic examination. Jousset's method of inoscopy is,
however, more successful in leading to their detection. This comparative
scarcity of bacilli in the fluid is also seen in tuberculous pleural effusion, and
is thought to depend on adhesion of the bacilli to the serous membrane.
Injection of the ascitic fluid into guinea-pigs shows that bacilli are present
in the effusion.

Paths of Infection. — The miliary tubercles seen on the peritoneum
in some cases of generalized tuberculosis are due to hsemic infection, the
bacilli reaching the peritoneum by the bloodvessels, but in these cases the
peritoneum merely shares with the rest of the body in the universal outburst
of tubercles and seldom or never shows any accompanying inflammation.
Tuberculous infection usually reaches the peritoneum through the lym-
phatics. The bacilli may be conveyed from the intestine and vermiform
appendix, the lymphatic glands, the pleurae, the suprarenals, prostate, vesic-
ulse seminales, and the testes.

According to Nothnagel, the lungs are primarily affected, without any
intestinal lesions, in the majority of cases of peritoneal tuberculosis; but in
Borschke's 226 cases of tuberculous peritonitis there was primary pul-
monary tuberculosis with marked intestinal ulceration in 140 cases, or 62
per cent. It is, however, remarkable how often tuberculous ulceration of
the intestine secondary to pulmonary tuberculosis is seen without diffuse
tuberculous peritonitis. Thus among 382 fatal cases of pulmonary tuber-
culosis at the Brompton Hospital, London, tuberculous ulcers were present
in the intestine in 296, or 77.4 per cent., while tuberculous peritonitis occurred
in 15 only, or 4 per cent. (Fowler and Godlee). The occurrence of tuber-
culous peritonitis secondary to primary intestinal tuberci^osis is variously
estimated. German statistics are to the effect that primary intestinal
tuberculosis even in children is very rare, and it is therefore not surprising
that among Borschke's 226 cases of tuberculous peritonitis not a single
one was secondary to primary intestinal tuberculosis. British statistics
give a percentage of primary intestinal tuberculosis in children ranging
up to 28.1 per cent, of the cases of tuberculosis (Shennan^) and averaging
19 per cent, as against 4 per cent, in Germany and 3 per cent, in the United
States (Harbitz^). In women tuberculous disease of the Fallopian tubes
is a prolific source of tuberculous peritonitis. Primary tuberculosis of the
peritoneum, in which there is no tuberculous focus elsewhere in the body,
and in which the path by which bacilli have got into the peritoneum is not
obvious, is rare. Borschke's series of 226 cases contained only 2.

^ British Medical Journal, 1906, ii, 357.

^ Edinburgh Hospital Report, 1900, vi, 130.

^ Journal of Infectious Diseases, Chicago, 1905, ii, 189.


Morbid Anatomy. — Before describino' the morbid appearances of diffuse
tuberculous peritonitis a few words may l)e devoted to localized tuberculosis

Online LibraryWilliam OslerModern medicine : its theory and practice, in original contributions by American and foreign authors (Volume v. 5) → online text (page 78 of 126)