Frederick M. (Frederick Madison) Allen.

Total dietary regulation in the treatment of diabetes online

. (page 66 of 76)
Online LibraryFrederick M. (Frederick Madison) AllenTotal dietary regulation in the treatment of diabetes → online text (page 66 of 76)
Font size
QR-code for this ebook


changed or increased food tolerance. But after any interval, early or
late, the characteristic change can be produced by adding to the diet
any kind of food beyond the assimilation, or by anything else that
brings on active diabetic symptoms. Its independence of digestion is
shown by its occurrence in fasting animals after removal of sufficient
pancreatic tissue to cause active diabetes during fasting. It is like-
wise independent of other pathologic processes, such as give rise to
diabetes. For example, hydropic degeneration does not give rise to
fibrosis, neither does it ever result from simple fibrosis of any island,
no matter how extreme. But when diabetes occurs, the presence of
fibrous or other changes does not prevent the typical hydropic proc-
ess, which in fact is generally found side by side with the fibrous
and hyaline changes in the t3^e of cases in which the latter have been
most emphasized by Opie and others. A point of particular practical
importance is that simple prolonged hyperglycemia, without glyco-
suria, regularly and definitely produces this change in diabetic ani-
mals; it is merely slower than with the more intense condition of
hyperglycemia plus glycosuria. A slight degree of vacuolation may
also be found in transitory diabetes, not only with glycosuria, but
even when the animals for some days following operation show merely
a considerable hyperglycemia on carbohydrate diet. The change
may be demonstrated in a tiny bit of pancreas removed at this time
without affecting the condition. Subsequently it may be impossible
to bring on diabetes by any amount of starch and sugar feeding, and
though temporary hyperglycemia may result, the island cells are not
vacuolated. It must be concluded that the condition following
operation was a genuine mild diabetes, even if evidenced only by
hyperglycemia; and that subsidence of inflammation and functional
and structural restoration and perhaps hypertrophy in the pancreas
remnant resulted in a final cure.

4. Mechanism. — Experiments have been conducted to show the
possible r61e of humoral and nervous agencies. Of the former, it
might be supposed that the blood sugar is predominant, and that
hyperglycemia per se is the stimulus to the islands. The evidence is
against such an assumption. Vacuolation of the islands does not
follow even the most prolonged hyperglycemia in normal animals,



628 CHAPTER vm

one example being the experiment described in the former mono-
graph by one of the authors, in which glycosuria was maintained by
glucose injections in a cat during the greater part of 17 months, with-
out producing vacuolation in the pancreatic cells or any other accom-
paniments of diabetes. Adrenalin and other forms of glycosuria like-
wise do not cause this change. Phloridzin has given useful results.
Phloridzin poisoning, continued even for many months, does not
cause hydropic degeneration; the unchanged sugar tolerance thereafter
would be sufficient proof, even without the microscopic examina-
tions. But when dogs are made severely diabetic by removal of all
but a small fragment of pancreas, and, beginning even before opera-
tion, receive phloridzin in small doses so as not to produce fatal
acidosis or intoxication and yet keep the blood sugar continuously
at an actually subnormal level, hydropic degeneration occurs in fully
t3rpical fashion. Of nervous influences, it was shown in the former
monograph that the Bernard puncture ordinarily causes no vacuo-
lation in the islands, but in the single predisposed dog in which true
diabetes followed the puncture, the characteristic hydropic change
occurred. Romans,^* producing hyperglycemia and glycosuria by 7 to
10 hours faradic stimulation of the splanchnic nerves in animals
possessing an entire pancreas or only a half or a fifth of it, was un-
able to demonstrate positive alterations in the islands. Even poten-
tially diabetic animals, however, fail to show vacuolation from hyper-
glycemia due to glucose injections within such an interval. The
questionable swelling and diminished granulation of the cells men-
tioned by Romans may possibly indicate a positive effect, which was
prevented from becoming inanifest only by the time limitations. A
paralytic injury seems to be excluded by both denervation and graft
experiments. The best procedure for this purpose is to leave a small
remnant about the lesser duct to aid digestion, and to transplant the
button of duodeniun bearing the main duct along with the entire
uncinate process beneath the skin of the abdomen, preserving the
blood supply of the graft through the inferior pancreaticoduodenal
vessels, and removing all the rest of the pancreas. Division of this
pedicle some weeks later isolates the graft from all intra-abdominal
connections. Removal of successive portions from both the duodenal
remnant and the subcutaneous graft then shows that both diabetes



ETIOLOGY AND PATHOLOGY 629

and island changes are absent until the total remaining pancreatic
tissue is reduced to the degree required to produce diabetes in an
ordinary operation; food then shows the usual influence, and the
usual hydropic changes occur in both the duodenal remnant and the
subcutaneous graft. The graft of course acquires a nerve supply,
presumably vasomotor, along with the blood supply from the sub-
cutaneous tissue. Also the intrapancreatic ganglia survive, and no
experiments have succeeded in eliminating the possibility of their
regulating action*. What is proved is that both the normal internal
secretion and the hydropic process go on apparently unchanged in the
absence of any possible specific stimulus or control from centers out-
side the pancreas. It may be conjectured that the regulation of the
function of the islands, as probably of endocrine organs in general, is
both nervous and humoral. Both experimental and clinical evidence
suggests the possible influence of nervous stimuh, at least for harm.
The humoral agency is proved; but it does not lie in hyperglycemia,
which is merely a symptom of active diabetes. A special hormone
is conceivable. But as mass action is an explanation of growing im-
portance for physiological regulations, it may be that variations in the
concentration of the internal secretion (whether this be combined
with food substances such as sugar, or free) in the blood serve to govern
the formation and discharge of this substance by the cells.

5. Consequences. — The hydropic degeneration results in numerical
atrophy and finally almost complete disappearance of the islands of
Langerhans. Weichselbaum and Stangl did not distinguish between
this clear-cut process and the more recondite "atrophy" in which the
island cells resemble lymphocytes. It seems probable that in the
earlier stages compensatory regeneration and hyperplasia also occur.
This view is supported clinically by the marked recovery of tolerance ■
often possible at such a stage, developmentally by the formation
of islands from ducts, which Laguesse and Bensley proved to con-
tinue in postembryonic life, and pathologically by the rather fre-
quent finding of enlarged, seemingly hyperplastic islands, and of
others that appear as if newly formed. Duct-Uke strands of cells,
and small, compact, generally oval islands, unUke those usually seen
normally and sometimes looking definitely like proliferations from
ducts, have been mentioned by Homans and are a familiar observa-



630 CHAPTER vrtr

tion in diabetic animals. In man such islands seem to be as a rule
spherical, since their cut section is approximately round. It has
seemed fitting descriptively to refer to them as the "morula" type of
islands. They are not merely contracted islands with empty capil-
laries, but conform to Bensley's description of secondarily developed
islands, and in their most typical form they are pushed in between
the acini as morula-hke masses of cells without capsule or the ordi-
nary capillary and trabecular framework. They are strikingly fre-
quent in some cases of diabetes, being sometimes ahnost the only
type of island present. The finding of a considerable proportion of
such islands may create a suspicion of diabetes; but, even granting
the interpretation of them as "young" islands, it is not surprising
that they may be found in the absence of diabetes, since regenera-
tion need not be Kmited to the diabetic pancreas, and may iti fact
sometimes be the means of preventing diabetes. The islands of this
type may suffer hydropic change and be lost like the others. Whether
they are functionally equal to the ordinary islands, whether the cells
are commonly paler and more subject to exhaustion and degeneration,
are only speculative suggestions at present. Regeneration, even if
a reality, fails in the later stages. Apart from the uncertainties of this
question, two important facts are fully established. On the anatomi-
cal side, the deficit in number and mass of islands demonstrated in
diabetes by Opie, Weichselbaum, Heiberg, and others, as far as it is
not due to destructive processes causing the diabetes, is the result of
the diabetes itself through hydropic degeneration; and in many cases,
especially some of the worst type in young persons, this latter process
is responsible for the loss of most of the islands. On the physiological
side, this phenomenon explains clearly the progressive loss of tolerance
caused by improper diet, and enforces the lesson of the difficulty or
impossibility of repairing the damage.

6. Significance. — ^Active diabetes must first be present in order for
the hydropic change to occur. In the partially depancreatized ani-
mal, a small fragment of pancreas evidently attempts unsuccessfully
to carry the whole metabolic burden. In human patients, the islands
present are apparently likewise stimulated to meet a demand beyond
their capacity. Presumably the cells respond with an actual or at-
tempted increase of secretory activity for a longer or shorter time



ETIOLOGY AND PATHOLOGY 631

while appearing morphologically normal. At length secretion is dis-
charged more rapidly that it can be formed, so that the normal fine
granulation becomes more sparse and is replaced by vacuolation.
Finally nuclear degeneration and complete disintegration of the cell
result. This interpretation is so clearly suggested by the anatomic
appearances, and so plainly confirmed by the feeding experiments,
that there has been full agreement on the point among recent investi-
gators. For this reason, the term "exhaustion" has come to be used
as synonymous with vacuolation.

Physiologically, this phenomenon may be compared to the dis-
charge of the acinar cells of the pancreas by secretin. There is a
difference, in that secretin or other hormones have never been known
to cause destruction of cells and actual disappearance of the tissue
upon which they act. Whether the difference Hes in intensity or
duration of stimulation or in pecuharities of the cells is unknown.
A further distinction is that an internal secretory function is here
concerned, and the process is a clear-cut example of anatomic break-
down of an endocrine organ by functional overstrain. A strikingly
clear relation between function and structure is thus shown, and the
conception of diabetes as the overstrain of a metaboUc function, long
probable on clinical grounds, is strongly confirmed. The experi-
ments with partially depancreatized animals are so simple and definite
that this biologically important phenomenon is made easy to study.

III. Clinical Application.

This includes practical deductions concerning the etiology, anatomic
diagnosis, and treatment of diabetes.

A. Clinical Etiology.

The etiology of diabetes will probably become fairly clear as soon
as the pathology of the pancreas is adequately studied. Notwith-
standing all the past work, the pancreas remains a neglected and
little known organ from the simple anatomic-pathologic standpoint, as
may easily be seen from the cursory remarks on it in text-books.
The normal variations must first be better known. Occasional atj^pi-
cal appearances were noted by Opie and others, and Oertel and



632 CHAPTER vin

Anderson^' are the most recent authors to interpret these as indicating
a natural sequence of degeneration and regeneration, though a ques-
tion of their true normality is raised by the statement that they
occur chiefly after middle Ufe. A comprehensive investigation must
be comparative. The present war, for example, may afford unusual
opportunity for extensive observations on the pancreas of supposedly
healthy individuals; but even if a high proportion show pecuharities,
the question whether these are normal phenomena or reactions to
morbid influences can be answered only by comparison, with a series
of animal species. The same mihtary opportunity might be utilized
to obtaiQ a series of specimens from the more primitive races of man.
It might then be possible to judge whether the erect posture, civilized
habits, or other factors render man specially lia,ble to pancreatic
disorders and hence to diabetes.

An elementary fact, which seemingly should have attracted notice
before, is that the pancreas is one of the most frequently diseased
organs in the body. There is even ground to question whether a
strictly normal pancreas may not appear as the exception rather than
the rule in miscellaneous autopsies. Doubtless the percentage of
abnormaUties is lower in the young and rises with age, just as the
incidence of diabetes is known to increase with age. The clinical
diagnosis is missed in the great majority of cases, because of the
deep situation of the organ hindering physical examination, the ab-
sence or slightness of digestive disturbance, and the fact that only
the most violent forms of pancreatitis give the symptoms recognized
as characteristic, while in milder or more chronic cases the local or
general signs are indefinite or imperceptible, or if present are ob-
scured by other morbid conditions. With due regard for nervous,
hereditary, and other contributing influences, the central problem of
the practical etiology of diabetes is the cause of the pancreatitis which
may be predicated in every case, as a rule with no exceptions yet
demonstrated.

Three forms of injury are distinguishable. One is the well known
inflammation produced by bile or obstructed pancreatic secretion, the
immediate harmful agency being chemical but the cause back of it

^'Oertel, H., and Anderson, C. M., Royal Victoria Hospital (Montreal), Scien-
tific Reports, 1916, Series B, 163-173.



ETIOLOGY AND PATHOLOGY 633

generally bacterial.'" Another is damage by bacterial products, lo-
cally or from a distance. There can be no doubt of the occasional
r61e of organisms, as those of syphilis, lying in the parenchyma, or of
colon or other bacteria in the ducts. Also any pathogens circulating
in the blood might perhaps attack either the capillaries or the epi-
thehum. Injury by soluble toxins from remote foci seems more
vague and dubious. Such a possibiHty is supported by threefold
evidence: (a) familiar lesions in viscera, including the pancreas,
from extensive burns and other severe intoxications; (&) the fre-
quent hyaline and rare amyloid changes, often affecting particu-
larly the islands, and best explained as of toxic origin; (c) the easily
demonstrable impairment of assimilation which so often accompanies
even sHght infections. The injury seems to be direct, since it is
difl&cult to imagine any important metaboUc alteration from a trivial
cold, for example. It also appears to be functional, since rigid con-
trol of symptoms by dietary restriction during the attack ordinarily
permits recovery of the full tolerance thereafter, while any permanent
lowering resulting when symptoms are not thus controlled is readUy
explainable by hydropic degeneration. It is possible, however, to
conceive of intoxication causing first functional deterioration and
then structural decay, in the form of pyknosis of nuclei, atrophy of
cells, and fibrosis of capillaries and islands..

Miscellaneous localized infections found in a certain proportion of
diabetics may be variously interpreted as pure accidents such as
befall nimierous non-diabetics, or as a result of the lowered resist-
ance characteristic of diabetes, or as a source of infection or intoxi-
cation directly causing the diabetes. It is probable that each of these
three possibilities is sometimes correct. The demonstrable fact of
damage to the islands of Langerhans by infectious disturbances, and

'" Concerning pancreatitis from bile, see Halsted, W. S., Bull. Johns Hopkins
Hosp., 1901, xii, 179-182; Opie, E. L., Ibid., 182-188; Am. J. Med. Sc, 1901, cxxi,
27-43; Flexner, S., J. Exp. Med., 1906, viii, 167-177.

Concerning pancreatitis from injection of bacteria, acids, alkalies, and other
irritants, see Flexner, Johns Hopkins Hosp. Rep., 1900, ix, 743-771; Flexner and
Eearce, Univ. Penn. Med. Bull., 1901-02, xiv, 193-202. Glycosuria is frequent
with acute pancreatitis, according to Flexner's experiments and the clinical ex-
psrience of Emerson, Bull. Johns Hopkins Hosp., 1908, xix, 95-96.



634 CHAPTER vni

the need and benefit of clearing up discoverable foci in any case, do
not prove, however, that the focus or organism in question is the
original cause of the diabetes. Removal of infectious foci responsible
for continuous or recurrent intoxication naturally prevents the in-
jury due to such intoxication. The clearing up of slighter foci, as
dental caries or pyorrhea, in the absence of perceptible systemic
symptoms, though advisable on general principles, seldom has much
influence upon the assimilation. The rule with all foci is that their
treatment gives relief from a distinct aggravating influence, but
never a cure of the diabetes or a subsequent course different from
cases without known foci. It is reasonably certain that a diabetes
due exclusively to gall stones from typhoid or colon bacillus infection,
or to pancreatitis from S)rphiHs, pneumonia, or mumps, would show
precisely the usual aggravation from a staphylococcus middle ear
abscess or a streptococcus tonsUUtis. Therefore conclusions should
not too lightly be drawn that such a focus is the primary cause of the
diabetes, either as a source of circulating toxin or as a portal of entry
of organisms.

This problem can only be solved by more careful investigation of
the incidence and causes of pancreatitis. '^ Whipple's study, already
mentioned," consisted in examinations of the pancreas in 230 un-
selected autopsies. Of these, 105 appeared normal; the others showed
more or less pancreatitis, classified in five groups. The 6 cases of
acute diffuse pancreatitis, occurring with pneimionia or other infec-
tions without recognizable symptoms, were mentioned above in con-
nection with the similar inflammation which in animals is proved to
be productive of diabetes. The three other classes of acute changes
were 41 cases of focal necrosis, apparently not serious in degree, 5
cases of fat necrosis, and 7 cases of acute hemorrhagic pancreatitis.
The remaining group of chronic pancreatitis embraced the majority of
cases. FeiUng ^^ expresses the opinion that mmnps is the commonest
cause of pancreatitis in childhood. He cites authors who have diag-

'* Cf. Egdahl, A., A Review of One Hundred and Five Reported Cases of Acute
Pancreatitis, with Special Reference to Etiology; with Report of Two Cases,
Bull. Johns Hopkms Hasp., 1907, xviii, 130-136: McCrae, T., Acute Pancreatitis
in Tjrphoid Fever, Assn. Am. Phys., May 7, 1918.

'2 Feiling, A., Quart. J. Med., 1914r-lS, viii, 263-264.



ETIOLOGY AND PATHOLOGY 635

nosed pancreatitis in respectively 4 out of 60 and 5 out of 33 cases
of mumps. Autopsies on cases of mumps are rare, and it is reason-
able to suppose that pancreatitis escapes notice more often than it
is recognized. There have, however, been only a few reports of
glycosuria or diabetes with mumps. The hyperglycemia or glycosuria
frequently reported in connection with other infections is sometimes
diabetic, more often doubtful in character.'^

The occurrence of acute pancreatitis without diagnostic symptoms is
highly suggestive in regard to the etiology of diabetes. It is conceiv-
able that the condition occurs either with general infections or in-
dependently, and that diabetes may result immediately or after
various intervals. In delayed cases, the acute inflammation may
render the organ susceptible to later injuries, or time may be re-
quired for the damaged islands to break down under metabolic strain.
Thus, in children particularly, it is imaginable that an infection occa-
sioning slight or imperceptible S3anptoms may leave the patient ap-
parently as well as before, yet really with pancreatic deficiency which
months or years later, under the burden of growth, becomes manifest
as diabetes, seemingly of acute onset. The possibiKty was mentioned
that with diabetic heredity such inflaimnations may be either more
frequent, or, by reason of an inherently subnormal functional power
of the organ, more serious in their consequences; and accordingly there
will be special interest in examinations of the pancreas of both the
non-diabetic and the -diabetic members of diabetic families. The
hardness of the pancreas so often noted by surgeons in operations is
hkewise of interest, though internists and pathologists dispute the
rehability of such observations. While the commonest origin of
chronic pancreatitis may be gall stones and biliary infections, there is
room for inquiry concerning other prolonged causes, and also to
what extent acute inflammations are responsible for progressive later
changes, through obstruction of circulation or secretion by fibrous
tissue.

However uncertain the symptoms, there is reason to believe that
pancreatitis of any marked degree is not a matter of indifference for
the general health; and the pathologic facts warrant the prediction

^ Allen, Glycosuria and Diabetes, pp. 564 and 798.



636 CHAPTER vni

that it will some day receive more attention as a cause of acute and
chronic illness. When an adequate diagnostic method is developed,
the new domain of pancreatitis will largely be carved out from the
diagnoses which today are essentially expressions of ignorance, e.g.
"dyspepsia," "nervous indigestion," "neurasthenia," "gastroenter-
itis," "autointoxication," "biliousness," "catarrhal jaundice," as well
as other indefinite abdominal troubles or vague general impairment
of health. Particularly with regard to "catarrhal jaundice," it is
worth while to question how often the stasis of bile is due to swelling
of the head of the pancreas, or, if due to primary bile duct infection,
how often this spreads to the pancreas, and whether in all such cases
thought should not be directed to the pancreas as the organ chiefly
in danger. Clinical observations (cf . table. Chapter III) support the
view that diabetes and indigestion generally do not occur together,
partly because, as Opie showed, the type of pancreatitis is often dif-
ferent, and also doubtless because the malnutrition of pancreatic
indigestion opposes the development of active diabetes.

The frequency of pancreatic inflammation may give reason for
surprise that the damaged organ is so often able to prevent diabetes,
but removes the principal element of mystery from the cases that
occur.

B. Anatomic Diagnosis.

Three chief difi&culties have hindered the study of pancreatic
pathology.

First, the organ is subject to incidental changes which may efface
its ordinary characteristics. Thus, OerteP* describes cellular atrophy
and collapse of structure to such an extent that islands and acini
are not distinguishable. These changes, which have nothing to do
with diabetes, but are purely the expression of pancreatic involution
from cachexia or other cause, are necessarily an obstacle in diagnosis,
especially important because terminal cachexia is so common in dia-
betes. Here the islands may show appearances like Weichselbaum's
atrophy, with shrunken cells and pyknotic nuclei, without proving
any specific disease in them. It is necessary also to be critical of

'^Oertel, H., Royal Victoria Hospital {Montreal), Scientific Reports, 1916,
Series B, 155-162.



ETIOLOGY AND PATHOLOGY 637

reports of the gross weight of the pancreas, since this may be found
low without indicating anything more than emptiness of the organ
and general malnutrition.

Second, cytological observations have not been considered in the
purpose or in the. preparation of material of most researches in the
past. Autopsy specimens are seldom ideal, often for reasons outside



Online LibraryFrederick M. (Frederick Madison) AllenTotal dietary regulation in the treatment of diabetes → online text (page 66 of 76)