Frederick M. (Frederick Madison) Allen.

Total dietary regulation in the treatment of diabetes online

. (page 64 of 76)
Online LibraryFrederick M. (Frederick Madison) AllenTotal dietary regulation in the treatment of diabetes → online text (page 64 of 76)
Font size
QR-code for this ebook


mous with the preceding. The strict criterion of traumatic diabetes
heretofore has been absence of glycosuria before the injury and its
presence within a reasonably short time thereafter. This is incom-
plete cKnical evidence, for doubtless in many such cases hyperglycemia
already existed. Scientifically conclusive proof could be furnished
only by pathologic examination. It is a justified prediction that the
pancreas in almost any case of tramnatic diabetes is the seat of abnor-
malities antedating the trauma. Experiments with direct and indirect
injuries of the pancreas in animals, and the clinical knowledge con-
cerning acute pancreatitis, practically exclude the possibiUty of any
immediate production of diabetes by direct pancreatic traimia; either
the damage will faU short of this, or it wiU cause death rather promptly.
It is conceivable, however, that acute injury of a previously normal
pancreas may inaugurate a process which will later result in diabetes.
Such cases have never yet been demonstrated and must necessarily
be rare. No clear-cut example of traumatic diabetes has been seen
at this hospital. The recent war has given the death blow to
trauma as an important cause of diabetes. For practical purposes,
the best foimded view-point is that trauma, shock, nervous injuries,
etc., bring on diabetes only when the pancreas is already diseased.
Autopsies upon all possible cases of traumatic diabetes, especially if
death occurred shortly after injury, are highly important in this
connection. Medicolegally and otherwise, the status of tramna is
like that of other exciting causes, in that it seemingly gives rise to
some cases which otherwise might not have become active. Whether
diabetes might have been anticipated without such an exciting cause
could be judged to some extent by a qualified pathologist, but not in-
fallibly. Compensation to the injured individual is not thus invali-
dated, since there is no doubt that any exciting cause may at least
hasten the onset of an impending diabetes or aggravate an. eixifeting
diabetes. i ' *



608 CHAPTER vm

8. Infection and Inflammation. — The most advanced views of the
infectious etiology of diabetes were expressed recently by Woodyatt"
as follows:

"Diabetes mellitus, a name now given by clinicians to any case in which as
the chief incident — or as one of several incidents — there occurs from natural
causes a diminution of the endocrine function of the pancireas sufficient to cause
the symptom diabetes. The factors which determine such diminutions are
infections of the pancreas, the term implying an interaction between a suscep-
tible tissue and a suitable pathogenic germ. So called diabetic predispositions con-
sist in inherited susceptibilities to certain types of infection. Well known varia-
tions in the clinical course of diabetes in the same case at different times and in
different cases are not wholly ascribable to variations in diet, etc., but often to
infectious moments. Concurrence of other diseases as Basedow's, dyspituitarism,
myocarditis, arteriosclerosis, etc., with diabetes, are not due to 'correlation of
ductless glands,' nor to metabolic disturbances secondary to diabetes, but to
simultaneous infections in different tissues." Prophylaxis is held to consist in
preventing diabetogenous infections (tonsillitis, sinusitis, pyorrhea, parotitis,
cholecystitis, ulcer, prostatitis) rather than in dietary restrictions. "These
views are based on the literature, an analysis of 100 chnical cases with aid of
D. E. Abbott, the character and frequency of pancreas lesions in 538 animal
inoculations with different strains of streptococcus group by E. C. Rosenow,
clinical improvements following treatment of foci of infection in certain cases,
previous views of Rumpf, Lepine and others, general development of knowledge
concerning relation of foci of infection to visceral diseases (Billings, Rosenow,
Irons, and others), knowledge of the selective affinity of certain strains for cer-
tain tissues (Rosenow)." Woody att has also placed strong emphasis upon the
occasional cases of acute diabetes accompanying acute infectious disease, in
which apparently normal carbohydrate assimilation is regained after recovery
from the infection.

The above may stand as the clearest expression of the Rosenow
school concerning diabetes, related to their corresponding doctrines
concerning other diseases. Notwithstanding an active following,
this contention in general is not at present regarded as estabhshed by
the majority of the medical profession. For diabetes in particular,
importance is assigned to focal and other infections by J. R. Wil-
liams, while Joslin in his wide experience, and Greeley^^ in an analysis

1* Woodyatt, R. T., Abstract of Proceedings of Seventh Annual Meeting of the
Americam Society for the Advancement of Clinical Investigation, 1915, 25-28.
^ Greeley, H. P., Wisconsin Med. J., 1915-16, xiv, 464-468.



ETIOLOGY AND PATHOLOGY 609

of the large material of Hodgson's sanitarium, were unable to find
evidence that such infections are more prevalent among diabetics
than among other hospital patients, or that they are a determining
factor in the chnical course. The interpretation of transitory dia-
betes accompanying an acute infection is necessarily ambiguous,
since the latter might be a primary or merely an exciting cause, and
the assimilation might subsequently appear normal even though the
pancreas were reduced considerably in mass or function. Proof
of the frequency 9f pancreatic lesions in experimental septicemia wiU
be a valuable contribution, but such experiments cannot in themselves
be decisive unless they produce either diabetes or a marked and per-
manent lowering of sugar tolerance. The above writers have cour-
ageously recognized that the hypothesis of an infectious etiology of
all cases of diabetes must be extended to various other metabolic
disorders. The particularly close relation of diabetes and obesity
must hold also here. While it is not impossible that obesity may
yet be numbered among the infectious diseases, such a bold concep-
tion must appear tpday as prophecy rather than proof. The un-
certainties in the whole subject of diabetes are such that tangible
evidence is the chief need.

The one tangible and outstanding fact is the frequency with which
pancreatitis has been described by all authors of all shades of opinion
who have made accurate microscopic observations in diabetes. The
present pathologic study has given an unexpectedly sweeping cor-
roboration of this lesson from the earUer work. It is possible to set
up a general dictum, "without pancreatitis, no diabetes," and chal-
lenge pathologists to produce exceptions. A diabetic pancreas in
which careful search fails to reveal more or less evidence of present
or past inflammation is at least a great rarity, and, if ever found,
will merely illustrate the remarkable completeness of organic
recovery sometimes possible after acute inflammation even of a degree
sufficient to produce diabetes. The principal stumbling-block has been
the apparent discrepancy between structure and function, as found
in the presence of diabetes with seemingly insufficient anatomic altera-
tions, and the absence of diabetes with more extreme visible change;
and this disagreement persists even with distinctions between inter-
lobular and interacinar forms of pancreatitis, and with comparisons



610 CHAPTER vin

limited to destruction of islands rather than of total parenchyma.
Discussion must partly be deferred to the subsequent section on pa-
thology; but much of the uncertainty necessarily connected with
cUnical conditions is cleared away by consideration of animal experi-
ments, divisible into the following four groups.

(a) When the duct and perhaps part of the blood supply of a pan-
creatic remnant are occluded, gradual atrophy brings on the well
known Sandmeyer diabetes, as late as 13 months after operation in
one of that author's cases. Such diabetes has never been produced
with the whole pancreas m situ, even when the ducts have been in-
jected with paraflSn by Claude Bernard and others, or recently with
alcohol by Auer and Kleiner." The extreme degeneration and
sclerosis finally resulting were shown in the pathologic examinations
of Sandmeyer, J. H. Pratt, and Auer and Kleiner. Nevertheless,
diabetes has not been observed unless two-thirds or more of the
gland has first been removed (or, rarely, destroyed by gangrene).
The reason evidently is that indigestion, emaciation, and cachexia
are in progress at the same time, and prevent the famihar diabetic
sjTnptoms even when the destruction of islands has reached a point
at which the animal must be classed as potentially diabetic. With
the entire pancreas present, thtese other disturbances will cause death
before frank diabetes develops; but removal of most of the gland
causes the internal secretory deficiency to become manifest earlier in
proportion as the pancreatic remnant is smaller.

(b) After simple partial pancreatectomy leaving the remnant with.
natural duct drainage, the tissue generally remains soft, lobulated,
and normal, except for a narrow zone of fibrosis adjoining the area
of amputation. More or less hypertrophy occurs sometimes; other-
wise conditions are found unchanged even after several years. The
nutrition and assimilative power are correspondingly maintained
unless the remnant is so small that the animal is potentially diabetic,
in which case dietary and metabolic influences may cause functional
and structural decay of the islands, without inflammation or fibrosis
and without alterations in the acinar tissue.

1^ Auer, J., and Kleiner, I. S., Proc. Soc. Exp. Biol, and Med., 1916-17, xiv,
151-153.



ETIOLOGY AND PATHOLOGY 611

(c) Sometimes more or less fibrosis takes place in a pancreas
remnant, presumably as the result of undue trauma in the operation.
Upon tliis hint, it was found possible to produce diabetes with con-
siderably larger remnants than usual, by setting up inflammation by
crushing between the fingers. The course of the fibrosis follow-
ing acute injury varies. Sometimes the process apparently halts,
so that no further impairment of structure or function is evident in
rather extended observations. In other cases the sclerosis continues,
so that diabetes develops after several weeks or months, and the pan-
creas remnant is found atrophic, sometimes almost as extremely as
after occlusion of the duct.

(d) As a better means of inducing aseptic inflammation, rubber-
covered clamps were applied to cut off the blood supply for 20 min-
utes to 2 hours continuously. Diabetes results more easily in propor-'
tion as the pancreas remnant is smaller, but has been produced with
fully a third of the pancreas present. Part or the whole of the

. imcinate process is the most convenient remnant for this purpose.
Shorter periods of clamping are used at first; if diabetes does not re-
sult, gradually longer stasis can be applied in later operations without
death from fat necrosis. Apparently only technical obstacles pre-
vent producing diabetes with the entire pancreas present. This would
have much theoretical interest, but the success actually achieved is
sufficient to estabhsh the principle. Incidentally, the pancreatic
gangHa withstand anemia for as long as 2 hours, so that the hope of
observing the effects of the loss of the intrinsic nerve supply has been
disappointed. The after-etfect of this acute experimental inflamma-
tion reproduces strikingly the pancreatic pictures seen with clinical
diabetes. Various grades of chronic fibrosis are found; but the
most interesting result is a pancreas which is soft, lobulated, and
normal appearing in gross, and which microscopically is charac-
terized by scarcity of island tissue, without corresponding destruction
•of acini, and with visible evidences of pancreatitis apparently alto-
gether too slight to account for the condition. Either the islands
are injured more easily and profoundly than the acini, or their power
of recovery is less. The trivial cHnical disturbance in the dogs even
during the early period of most intense pancreatitis is also remark-
able, but it could not be safely inferred that human patients would



612 CHAPTER VIII

be equally little affected. It is therefore of interest that Whipple"
found acute inflammation of the pancreas precisely similar to the ex-
perimental form in 6 out of 230 unselected autopsies, most often with
pneumonia and a smaller proportion with other infections, and stated:
"These cases were all under careful observation in the wards of the
Johns Hopkins Hospital, and gave no s5Tnptoms of pancreatic disease."
Though the two columns of infections in the General Summary table
(beginniag of Chapter HI) may at first glance seem imposing, a classi-
fication of the 76 cases of this series will show something like the fol-
lowing four groups:

(a) 23 cases characterized by more or less numerous infections,
the causal relationship of which to the diabetes must be purely specu-
lative, without any definite indications. These are cases Nos. 1, 4,
6, 8, 9, 11, 13, 14, 15, 19, 20, 21, 29, 31, 32, 38, 42, 44, 48, 49, 53,
63, 72.

(b) 29 cases not only lacking any suggestive infectious etiology of
the diabetes, but rather exceptionally free from infections in general.
These are cases Nos. 2, 7, 10, 12, 18, 22, 25, 34, 35, 39, 45, 46, 47,
50, 51, 52, 54, 55, 57, 58, 64, 65, 66, 68, 70, 71, 73, 74, 75.

(c) 9 cases in which an infection stands in suggestive relation with
the outbreak of diabetes. Patient No. 3 not only had "colitis" 5
years before admission, but, just before the first diabetic symptoms,
suffered appendicitis and appendectomy, followed by phlebitis ap-
parently indicating septicemia. Patient No. 5 at the beginning of
his diabetes had symptoms which may have represented acute pan-
creatitis. No. 24 gave a clear-cut history of obstructive jaundice,
without evidence of gall stones oo: biliary infection, followed within a
year by the gradual onset of diabetes. The child No. 26 had some
unknown disturbance which caused vomiting, followed immediately
by diabetes. The suspicious condition in another child, No. 28, was
merely fever of unknown origin; the subsequent slight choreiform
movements do not prove that it was poliomyelitis, and it might
have been pancreatitis. In No. 36, the beginning of diabetes is
significantly related with general sepsis. No. 37 had nothing but an
alveolar abscess, then an ordinary cold; diabetes promptly followed.

"Whipple, G. H., Bull. Johns Hopkins Bosp., 1907, xviii, 391-396.



ETIOLOGY AND PATHOLOGY 613

No. 40 was admitted for pneumonia; he had apparently been a well
man before, and his apparently complete recovery from the diabetes
suggests pneumococcus pancreatitis as the probable cause. The
young boy No. 76 had only otitis media, indicated by earache and
fever. Diabetes quickly followed, and the only further evidence of
pancreatic involvement is that digestive upsets have since occurred
from sKght causes.

(d) IS cases in which at least a possibility of connection exists be-
tween diabetes and pancreatic injury produced by some infection
acquired many years previously. Three of these (Nos. 16, 41, 67)
are cases of syphilis. No. 17 had a serious combination of pneu-
monia and empyema 20 years before admission, and a later history of
indigestion. No. 23 had suffered from "bloody dysentery" 14 years
previously. No. 27 had had mumps with orchitis at the age of 18, a
pancreatic involvement being imaginable; there was also "jaundice"
at the age of 20. No. 30 had passed through some ordinary infections,
and gave a significant history of "nervous indigestion" and pale feces
for years past. No. 33 twice suffered from acute nephritis following
colds. The pancreas is certainly as susceptible to infection as the
kidney, and there was a history of indigestion. In No. 43, this com-
bination of sepsis and nephritis, with subsequent diabetes, is especially
striking. No. 56 had inflamed cervical glands in childhood, and was
also subject to gastrointestinal attacks with fever of unknown origin.
No. 59 gave a history of biliousness, nausea, and vomiting through-
out childhood. No. 60 had had "gastric fever" in 1896. No. 61 in
childhood had suffered from intractable diarrhea; subsequently he
had typical rheumatism; and as the infectious origin of his cardio-
renal disease is not doubted, the same may be assimied as the cause
of pancreatitis. No. 62 gave an, unintelligible description of some
childhood trouble possibly related to her diabetes. In No. 69,
sepsis may be thought of as a possible cause.

The elements of personal judgment and clinical uncertainty are
illustrated here as necessarily in all such inquiries. It is to be recog-
nized that groups (a) and (6) comprise 52 cases, or over 68 per cent
of the series. That is, this large majority shows no perceptible rela-
tion between known infections 'and diabetes. The relations sug-
gested in groups (c) and (d) are purely a matter of interpretation and



614 CHAPTER vni

carry no demonstration in any instance. The connection supposed in
group {d) between events so many years apart might easily and
sometimes doubtless rightly be considered imaginary.

The most definite group, (c), is open to quite different interpre-
tations, by reason of the well known facts that diabetics are specially
liable to infections and that existing or latent diabetes is aggravated
by infection. Thus, patient No. 40 might have had some preexisting
diabetic tendency; the latent diabetes may have been awakened by
the pnevunonic infection and subsided again with the subsidence of
the latter. Zealous advocates of infection might choose to change
cases Nos. 6 and 38 from group {a) to (c). But there is here no evi-
dence of anything but pneimionia in persons already mildly diabetic;
thus, patient No. 6 considered herself well before the pneumonia,
when there was no knowledge of diabetes; she continues to consider
herself well now, when it is known that glycosuria is constantly
present. Patient No. 57 was included in group (J), because measles
at the age of 8 was the only illness known before his diabetes. The
accidental fact that he studied medicine afforded the sole informa- .
tion that glycosuria was present then, 8 years before the supposed
onset of diabetes. Except for this, it would have appeared that
repeated tonsillitis and then furunculosis were followed by diabetes,
and the case might have been placed in group (c) on suspicion of
staphylococcus pancreatitis. In patient No. 41, diabetes was first
discovered in connection with a cold and sore throat, as in some of '
the cases in group (c) ; but in him the more probable cause of diabetes
is syphilis. It is probable that minor abnormalities, as of the teeth
and lymph glands (see again General Summary, Chapter III) are no
more common in diabetics than in 'others, unless as the result of the
diabetes; and an etiologic position has never been demonstrated for
them in any case. In the absence of other infections, some might
lay stress upon caries and pyorrhea in suCh a case as No. 67 ; but here *
syphilis is known to exist. AU the cKnical conditions are therefore
confusing, and the only trustworthy guide is the pathology. This
compels recognition of the fact that, unless these diabetics are dif-
ferent from the large series ' covered by the present microscopic
study and from the many others in the Mterature, the basis of the
disease in all of them is pancreatitis. The existence of pancreatic



ETIOLOGY AND PATHOLOGY 615

inflammation, and the search for its cause, is thus extended to all
four groups alike. The cause ordinarily to be supposed is bacterial;
but a connection with any other general or focal infection is clinically
discoverable at best in only a minority of cases, and even in them is
subject to many doubts and mistakes.

The basis of belief in the inflammatory origin of diabetes is there-
fore essentially a generalization from the classical studies of diabetic
pathology, with experiments and deductions to clear away some
apparent inconsistencies and confusion. The view of Woodyatt and
others of the Rosenow school, concerning the status of focal infec-
tions, specific relations of bacteria to organs, etc., is a distinctly new
suggestion, concerning which the above observations do not decide.
This contention in diabetes is on the same basis as in a variety of
other diseases, and must stand or fall as the general evidence may
determine.

II. Pathology.

Since the former review of this subject,^' papers have appeared by
Koch," supporting the archaic notion that the islands are merely
degenerate and functionless portions of the parenchjmaa, and by
Major,'"', dealing with a comparative study of the pancreas in 35
non-diabetic and 13 diabetic necropsies.

The pathologic changes of the pancreas in diabetes are divisible
into those causing the diabetes and those resulting from the diabetes.

A. Changes Causing Diabetes.

For general purposes, details of fibrous, fatty, hyaline, and other
alterations may be ignored, and all cases grouped roughly in three
classes, as follows.

1. Cases of extensive loss of parenchyma, involving islands and
acini alike, and in the most extreme instances comparable to the
Sandmeyer diabetes of dogs. The earliest pancreatic lesions described
by Cawley and others were naturally of this class.

^' Allen, Glycosuria and Diabetes, Chapter XXI.

"Koch, K., Virchows Arch. path. Anat., 1913, ccxi, 321-330.

2" Major, R. H., J. Med. Research, 1914, xxxi, 313-330.



616 CHAPTER vin

2. Cases of selective injury of islands. There is never much
destruction of islands without some involvement of the acinar tissue,
and all gradations between this and the first group are met, but the
best instances, notably of fibrous and hyaline change, well justify
Opie's use of them as the basis of the insular hypothesis.

Under this same heading may be mentioned also certain peculiar
pictures, described early by Weichselbaum and StangP^ and recently
by Wilhams and Dresbach,^^ but not heretofore correlated as stages of
a special process. They may in fact not represent a distinct and pro-
gressive sequence, but comparison between different cases and be-
tween different portions of the same pancreas frequently reveals the
following graded examples. In the seemingly incipient form, the
island appears sometimes congested, frequently hypertrophic, and
perhaps irregular in structure, but the essential feature is the striking
pyknosis of certain nuclei, with the cytoplasm about them often only
a narrow band. Even at this stage, sHght fibrosis or at least a few
round cells are present. Through successive degrees, this condition
seems to pass over into what Weichselbaum and Stangl called atrophy
of the islands, in which condition they are small, more or less fibrous,
and characterized predominantly by the shrivelled cells with pyknotic
nuclei and scanty cytoplasm. These pictures, though common, are
doubtful in interpretation. The change is distinguishable from typi-
cal hydropic degeneration; for though pyknosis of nuclei is a feature
of the latter, it is plainly subsequent to the vacuolation of the cyto-
plasm. Shrunken cells may be present in the same island with
hydropic cells, but apparently never become hydropic themselves,
perhaps because they are no longer functional. Weichselbaum and
Stangl considered the atrophic islands not diagnostic of diabetes, be-
cause found in some non-diabetic cases.' In 1911, Weichselbaum
again mentioned this atrophy, in which the island cells resemble
l)nmphocytes, and this time considered it probably a result of hydropic
degeneration, therefore diabetic. This picture is very rare in experi-
mental diabetes, but what is probably a true reproduction of it has
finally been observed in a few diabetic animals. Here also the

^'^ Weichselbaum and Stangl, BibHograpKy of Chapter 1.

22 Williams, J. R.. and Dresbach, M., Am. J. Med. Sc, 1917, cliii, 65-78.



ETIOLOGY AND PATHOLOGY 617

shrinkage of island cells is accompanied by round-cell infiltration,
and the change is probably inflammatory in origin.

Its presence raises a suspicion of diabetes; and even if the latter be
absent, the reason might be only that the change is not yet sufficiently
advanced or general. Widespread " atrophy" of islands is diagnostic
of diabetes, generally of the severest form. Islands showing this
change in advanced degree are certainly functionless, because they
are often present in large numbers in cases with scarcely any food



Online LibraryFrederick M. (Frederick Madison) AllenTotal dietary regulation in the treatment of diabetes → online text (page 64 of 76)