Pregnancy—related diabetes mellitus in two dogs
B
H)fn;N$W j r0Introduction4x3{+k/m&G+K,^3n0
中国宠物医师网/] Zzq&A ^
Diabetes mellitus can arise as a consequence of in—— sulin resistance associated with other disease states， or drug administration．This association is rec09— nized in dogs with hyperadrenocorticism，in dioestrus，with acromegaly and glucagonoma，or following administration of drugs such as gluc0—— corticoids and progestagens．The hormonal changes during late pregnancy in dogs also cause profound insufin resistance，and uptake of glucose by peripheral tissues in response tO insulin was re—— duced by 30％of values during anoestrus．Yet，al— though its occurrence is mentioned in authorita—— rive texts，reports of diabetes mellitus during pre9— nancy in dogs are scarce． Transient diabetes mellitus is very uncommon in dogs and when it occurs it is usually associated with insulin antagonism from disease or drug therapy．Transient diabetes meUitus associated with ketoacidosis was reported in a dog with stressful concurrent illness，and treated with glucocorti— coids．Dioestrous bithes with diabetic ketoacidosis may also sometimes have resolution of insulin——re—— quiring diabetes mellitus following successful ther——^4v@2ba0
中国宠物医师网_k7Mnp@7Rbn$F
apy and ovariohy’sterectomy． With the rare exception of functional pituitary tu- mors，excessive production of growth hormone in canine acromegaly is derived from the mammary gland，where it is induced by either progesterone or synthetic progestagens．As a consequence，signs of acromegaly are noted in a consistent relation—— ship with dioestrus or administration of progesta- gens，although they may wax and wane through several heat or treatment cycles before diagnosis． Typically，signs begin 3—5 weeks after oestrus in non—pregnant bitches，and include respiratory stri— dor，increased soft tissue mass，formation of skin folds，fatigue，enlargement of inter-dental spaces， polyuria and polydipsia，and hyperglycaemia．Di- abetes meUitus occurs commonly in acromegaly because growth hormone causes profound glucose intolerance and insulin resistance．The fact that concentrations of progesterone during pregnancy in the bitch are similar to those during non—preg- nant dioestrus suggests that acromegaly should be equally likely in pregnant bitches，yet reports of its occurrence in pregnancy could not be found． Two cases of diabetes mellitus in bitches occurring in association with pregnancy with contrasting presentations and maternal outcomes are reported here．D he1xNi0q9hJ7@0
c0u h {%_6Moq0Case lvk UV L5JS5k0
kd-\H {5Ph0A 6——year——0ld pregnant Labrador Retriver was presented for investigation of facial swellin9 1 4 days before her expected due date(EDD)．The bitch had delivered nine viable pups 3 years pre- viously，but despite regular cycling since then had failed to conceive when mated during four sepa- rate oestrus periods，until this pregnancy．The bitch had thickened skin，with increased folds af- fecting the face，neck，trunk and tail—base area， and also had upper respiratory stridor．Mild hy— perglycaemia of 8．6(reference range 3．9—6．1) mmol／L was identified，there was no glucosuria， the urine was dilute(specific gravity l．006)and the urinary sediment was inactive．Glucosuria was subsequently demonstrated l o days prior to the EDD．Significant fasting hyperglycaemia(21．05 mmol／L)and concurrent glucosuria were present 7 days prior to the EDD．At this time，the con— centration of fructosamine in serum was slightly elevated at 383(reference range l85—341)umol／ L，and a urinary tract infection was identified．Ap— propriate antibiotics were administered based on sensitivity testin9．Acromegaly with secondary di- abetes mellitus was suspected．Treatment was ini— tiated with 0．375 IU／kg intermediate——acting porcine insulin l 2一hourly with meals．This dose was increased incrementally t0 0．6 IU／kg during the subsequent 5 days，without achieving eugly— caernja． An elective Caesarean section and ovariohysterec—— tomy were performed 61——63 days after predicted ovulation，as estimated from progesterone concen- tratlons at pr0—oestrus．Two live puppies were delivered，both with meconium-stained amniotic fluid．The pups were both underweight，viz 345 and 335 9 at birth，whereas the mean birth weight of 5 1 0ther Labrador Reriever puppies from six litters in the same kennels was 452(range 370- 564)9．Both puppies appeared alert and active initially，but their condition deteriorated and thev died within 36 h of birth despite supportive care with warmth，oral dextrose and colostrums．Blood glucose concentrations were measured during the terminal management of both pups，but in neither中国宠物医师网a K.o EA,El3f
中国宠物医师网1bkY{P1H6L X~
case was a value ascertained before dextrose had been administered．The blood ducose concentra— tion was 3．5 mmol／L in one pup，and ranged be— tween 2．2 and 4．2 mmol／L in the other pup． Post—operatively，pre—insulin blood glucose values were measured twice daily for 3 weeks，and l 2一h glucose curves were performed on Day 2，4，7， 14，21，28 and 35．During this period，doses of insulin were increased by 0．1—0．2 IU／kg in re— sponse to persistent hyperglycaemia during a glu—— cose curve，and decreased whenever a blood glu— cose of<5．5 mmol／L was demonstrated(on three occasions)．Doses of insulin ranged between 0．3 and 0．7 IU／kg during this period．Despite these adjustments，insufin was unable to be withindrawn，and l 8 months later the bitch re- mains well—controlled on a dose of 0．4 IU／k9． By 2 months post—operatively，the facial swellin9， skin thickenin9，excess folds and respiratory stridor had resolved． L y`ssF3e0
-Y9n\0{;f&L0Case 2中国宠物医师网,pE7}
ozW9|
3oC&c|6`cy0{0A 6-year——0ld female Siberian Husky that had de—— livered five viable pups l year previously was presented with a primary complaint of dystocia． Following the birth of a single，live pup no fur— ther abdominal contractions were observed．The bitch was pale and mildly tachy’cardic(heart rate 1 50 beats per minute)．There was no response to two injections of l 0 IU oxytocin administered in—— tramuscularly l h apart．A Caesarean section was performed and a further three live pups were de—— livered 5 h after the birth of the first pup．Due to continued haemorrhage at the sites of placental at—— tachment and because there were no future plans to breed the bitch，and ovariohysterectomy was performed at the same time．Four of the pups weighed approximately 450 9，similar to the five pups of the bitch’s previous litter which ranged from 450 t0 538(mean 491)9，and one pup was small，at 255 9．All pups appeared strong and were vocalizing at birth．The bitch recovered from anaesthesia well and displayed an interest in the pups，and was sent home in the care of the own— er． Two days later，the bitch was presented severely ill with diabetic ketosis．The dog had lost weight since surgery and was weak and unable to stand， and there was vomitin9，melaena，tachypnoea， tachycardia，pyrexia and severe dehydration．Clin— ical pathology tests were performed in—house． There was severe hyperglycaemia(出ucose 36．75， reference range 4．28—6．94 mmol／L)．a high urea (29．2，reference range 2．5—9．64 mmol／L)，and a creatinine of l 60(reference range 44—1 59)umol／ L．There were mild elevations in hepatic enzyme activities (alanine aminotransferase 242，reference range 5—5 1 U／L；alkaline phosphatase l 79．refer— ence range l4—134 U／L)，and a normal concen— tration of bihrubin(<2，reference range 0一1 5 u— mol／L)．The concentration of cholesterol was ele— vated at l 0．72(reference range 2．84—8．27)mmol／ L．The total concentration of calcium was 2．06 (reference range l．98—3)mmol／L，and concentra— tion of phosphorus was 4．13 (reference range 0．81—2．191 mmol／L．Amylase activity was normal at 504 (reference range 500—1 500)U／L．The concentration of total protein was 54(reference range 52—82)9／L，and that of albumin was 26 (reference range 22—39)9／L．A severe leucocyt0— sis(total white cells 39．1，reference range 6—16．9x 109／L)and anaemia(haematocrit l5．3，reference range 37—55％；haemoglobin 54，reference range中国宠物医师网 k'~ A7P P
7d {.G/C,Ixx0{01 20——l 80 9／L)were present．A reticulocytosis was not evident(reticulocytes 0．3，reference range 0— 1．5％1．A platelet count of 375 (reference range 200—5001 x 1 09／L was measured．A blood smear was not examined．The urine contained moderate concentrations of ketones(+++，80—1 60 mmol／ L)and high concentrations of glucose (++++，> 1 1 1 mmol／L)．It was recognized that，in retr0— spect，the bitch had been polydipsic for approxi— mately 2 weeks，but the owner had considered this normal．Radiographic and ultrasonographic examinations were not performed．Abdominal pain exceeding that expected 2 days after lapar0—— tomy and ovariohyrsterectomy was difficult to de—— tect．Possible differential diagnoses included an underlying acute primary small intestinal disease or acute pancreatitis，but further investigations were not performed． The bitch recovered following intensive therapy 、)1，ith intravenous fluids and insulin．The former was initiated using whole blood and followed by 0．9％saline．Fluids were supplemented with glu— cose，as necessary，to maintain blood glucose con— centreations above l 4 mmol／L，until ketosis had resolved．Potassium could not be measured during this initial period and therefore potassium supple— mentation of fluids was based on the minimum expected potassium requirement during treatment of ketoacidosis．The clinical setting precluded hourly monitoring and so low——dose soluble in—— sulin was administered intramuscularly approxi—— mately every 3 h，to reduce hyperglycaemia and correct the ketoacidotic state．An initial dose of 0．25 IU／kg was used，then adjusted according to blood glucose concentrations for subsequent treatments．This protocol represented a modifica— tion of that recommended by Feldman and Nel一 3k)11． Additional therapy comprised broad——spectrum an—— tibiotics，antiemetics and gastrointestinal protec— tants for symptomatic control of gastrointestinal signs．After 36 h of intensive therapy，intermedi— ate——acting porcine insulin was commenced at an initial dose of 0．5 IU／k9．The insulin requirement of the bitch gradually declined and insulin was able to be discontinued 7 days later．Intravenous fluids were continued for 7 days，at which point the bitch was well enough to be discharged． Haematological and biochemical screening per—— f0咖ed at a commercial laboratory l l days after admission with ketosis showed only mild abnor— malities and resolution of leucocytosis，resolution of elevated concentrations of urea，improvements in the activtities of liver enzymes，a mild non—re— generative anaemia (haematocrit 32，reference range 37-55％)and thrombocytosis(928，refer— ence range 200-500 x 109／Ll．Intermittent vomit— ing and inappetence gradually resolved over 3 weeks．The bitch remains euglycaemic l 8 months later and had gained weight，and there had been no glucosuria or other clinical signs of illness． All four puppies were managed at home by an exDerienced breeder but died within 5 days of birth．中国宠物医师网i||3} KC2\"[
#qns5CI%}(K0Discussion中国宠物医师网+?;w,lO'N.R zM
中国宠物医师网9?!]H2Q8tIr:z3Ob \~+M
Although not confirmed by assessment of growth hormone concentrations，Case l presented with signs that could be consistent with acromegaly，of increased skin thickness and prominent skin folds around the face，neck，and shoulder area，respira— tory stridor，and development of diabetes mellitus． Because mammary production of growth hor——中国宠物医师网#c#i+P q Y bz
中国宠物医师网j~r`"hN{r6t
mone is known to be stimulated by progesterone， and progesterone concentrations are similar during pregnancy and non—pregnant dioestrus in dogs，it is surprising that acromegaly occurring during pregnancy is not more frequently reported． Growth hormone concentrations were elevated in only l／45 pregnant dogs test，however the ages of those pregnant bitches were not reported．Bitches are generally used for breeding at a younger age than the mean age at which those authours re—— ported acromegaly was seen fmean 8．5±standard error 0．5 years)，and therefore acromegaly might be less likely to be seen during pregnancy．The occurrence of considerable inter——dog variation in the magnitude of increase in growth hormone may also mean that certain individuals are more predisposed to developing acromegaly than others． In case 2，diabetes mellitus was not diagnosed un— til after delivery，but the history of polydipsia su9— gests its presentce for some time during late pregnancy．Most dogs with diabetes mellitus have had consistent signs for a period of at least 2 weeks．Just as with acromegaly，diabetes mellitus might be expected to occur with equal frequency during pregnant and non——pregnant dioestrus in the dog since both progesterone and growth hor—— mone cause insulin resistance．However，the scarcity of published cases of．diabetes occurring during pregnancy suggests its occurrence is un—— common．Indeed，peripheral insulin resistance during pregnancy may be adaptive，giving priority for酉ucose to fetal and placental tissues during a period in which the demand for glucose is more than doubled，and reducing the likelihood of hy— poglycaemia which may develop during fastin9，or as a result of the insulin response during a low— carbohydrate meal．In the non—pregnant dioe— strous d09，despite lower insulin resistance this fe— tal drain of glucose is not present，perhaps predis— posing non——pregnant dogs to develop diabetes more frequently than pregnant ones． The presence of ketonuria at the same time as se—— vere hyperglycaemia in Case 2 supports a diagn0—— sis of diabetic ketosis rather than pregnancy ket0— sis．in which ketonuria without glycosuria is found，and blood glucose concentrations are low． Although the clinical presentation of Case 2 was more typical of diabetic ketoacidosis than diabetic ketosis，acidosis was not confirmed and the clinical presentation could also be consistent with concur—— rent diabetic ketosis and severe acute primary gas—— trointestinal or pancreatic disease．Signs of severe anaemia，pyrexia and leucocytosis suggestd such complicating disease．The anaemia and dispropor— tionate elevation of urea compared to creatinine were thought to be the result of gastrointestinal bleedin9．The haematocrit of healthy pregnant bitches at term can range from as low as 26(mean 30．6％)，and the amount of blood lost in the gas— trointestinal tract may not have been as substantial as the haematocrit of l5．3％may otherwise have indicated．Surgical losses during oVariohysterect0— my are likely also to have contributed to the low haematocrit at this time．Although concentrations of amylase were not elevated，the possibility of pancreatitis remains，since amylase is frequently not elevated in pancreatitis．Whatever the cause， such a severe concurrent illness could contribute to insulin resistance and ketogenesis by increasing secretion of counter——regulatory hormones such as catecholamines and cortisol，already likely to have been promoted by the stress of dystocia and surgery．Pancreatitis could additionally cause tran— sient or permanent damage to islets during the-vse*J7I.B v1?0
N7yZw[3jEy3foY0pathological process，reducing their capacity to se— crete insulin．These factors would have added to ketogenic factors normally present in late pregnan—— cy such as high endogenous concentrations of growth hormone．Antagonism of the actions of insulin，promotion of fatty acid mobilization and stimulation of the release of glucagons by such counter——regulatory hormones facilitates ketogen—— esis．When not balanced by tissue uptake and uti— lization of ketones，as is normally enabled by in— sulin，severe ketogenesis results． Persistence of insulin——requiring diabetes mellitus after resolution of secondary diabetogenic factors in Case l suggests that，whether or not underlying pancreatic pathology pre—exsited，there had been further irreversible destruction of islets．The metabolic state of diabetes mellitus can contribute to destruction of beta——cells through the effects of 出ucose and lipid toxicity．These effects are de— pendent on the duration and severity of hyper—— glycaemia．Decreasing concentrations of blood glucose to below l 4 mmol／L will therefore mini—— mize the risks of progression of destruction of pan—— creatic islets；however．this is di伍cult to achieve at the best of times in dogs let alone in the face of the severe insulin resistance of pregnancy．Very Kgh doses of insulin are likely to be required and intensive monitoring of blood舀ucose concentra— tions needed．Immediate termination of the pre9— nancy by ovariohy’sterectomy may be a more pragrnatic approach to achieving good glycaemic control in the bitch．Regardless，given the risks of future dioestrous—related diabetes meUitus，ovari0— hysterectomy should be performed before the next oestrus． The possibility that the predisposition for diabetes is inherited must also be considered．Although the bitch in Case 2 remained euglycaemic without in—— sulin therapy some l 8 months after the episode of ketosis，diabetes mellitus may redevelop，especially if other factors causing insulin resistance are not well controlled． A high risk of adverse fetal events，including abor— tion，small unthrifty pups and overly large pups (macrosomia)，has been described in diabetic bitches．Vascular effects of diabetes may reduce | placental blood supply，contributing to abortion or I poorly—grown pups，and macrosomia may result J from the anabolic effects of fetal hyperinsulinemia， l a response to excessive delivery of glucose，amino I acids and厶tcy acids t。the fetus在。m the matemal J circulation．In a study involving bitches with ex— I perimentally—induced diabetes mellitus，higher l numbers of stillborn pups，with no gross abnor— I malities，were b。m t。six diabetic bitches than t。 I six control bitches．The birthweights of the pups l was also reported in a colony of diabetic Golden l Retrievers．Excessive feteal growth is the most I frequent complication of gestational diabetes melli— l tus in humans． I However，in the two cases presented here high I birthweight was not present even though poor l maternal glycaemic control was demonstrated in I Case l and suspected in Case 2．The risks and l types of adverse fetal outcomes may depend on the degree and duration of hyperglycaemia in the dam，and whether the onset of diabetes meUitus predates organogenesis and placentation． The pups of diabetic bitches may also be predis— posed to hypoglycaemia after birth，when cessa— tion of maternal defivery of fuels is not immedi— ately matched by a reduction in insulin secretion by the fetus．However，limited experimental evi— dence suggests that a predisposition to hypogly——#t7u5m6wQF|E t(r
I0
中国宠物医师网t P`fX4@
caemia is not likely in the pups of diabetic bitches in the absence of other predisposing factors．Al一 though elevated concentrations of insufin were present in pups of diabetic dogs at birth，hypogly— caemia did not occur even when such pups were fasted；hepatic glycogen stores were elevated，and glycogenolysis occurred at increased rates immedi—— ately after birth，suggesting a degree of hepatic in— sensitivity to the effects of hyperinsulinism．Al一 though gluconeogenesis is initially attenuated，nor— mal rates are attained by 24 h of age．A blood glucose concentration as low as 2．2 mmol／L was measured in one pup from Case l，despite glucose supplementation，a value below the reported range of 2．8-7 mmol／L for canine neonates of between 1 and 3 days of age．Several other factors，how— ever，may have precipitated hypoglycaemia in that pup from Case l，including hypothermia，hypoxi— a，or rejection by the bitch；small or underweight pups，as this was，may be especially susceptible． In the cases presented here，the lOSS of l 00％of the pups in both litters is unusual and suggests that the metabolic state of the dams may have been a contributing factor．This can not be assumed， however，since several other risk factors were also present，including low birthweight，dystocia，e— mergency Caesarean section，and matemat illness． It is not possible to identify the factors involved in the neonatal deaths in these cases because of a lack of detailed investigation． These two cases demonstrate that diabetes meUitus can occur in association with pregnancy in dogs and that diabetic ketosis can occur during transient diabetes mellitus in dogs．The clinical findings in Case l also suggest that acromegaly can occur during pregnancy——related dioestrus but it may be uncommon during pregnancy because of the dif_ ference between the usual age of onset of this condition and the age that dogs are usuaHy bred． Diabetes meUitus may be less common during pregnancy than non——pregnant dioestrus because of the nett glucose drain by the fetus，which balances the profound insulin resistance of pregnancy． Transient diabetes is uncommon in dogs，but the insulin—antagonistic effects of stressful illness， surgery，dystocia and pregnancy were sufficient to result in diabetic ketosis in the absence of absolute insulin deficiency in Case 2． Diabetes meUitus should be included as a difieren- tial diagnosis in all pregnant dogs with polyuria and polydipsia．Even mild hyperglycaemia should not be ignored during pregnancy as，despite their insulin—resistant state，normal bitches maintain euglycaemia during pregnancy．Persistence of dia— betes after resolution of insulin antagonism in Case l suggests destruction of pancreatic islets， which could have been the result of poor metabolic contr01．Either control of blood glucose concentration to below l 4 mmol／L or immediate termination of the pregnancy via ovariohysterec—— tomy may have prevented this．Maternal hyper— glycaemia may contribute to adverse fetal out—— comes in dogs but further study is required to de—— terminF the nature of th e risk中国宠物医师网0OXv m$L
