Comparison of hematologic variables between dogs with congenital intrahepatic and extrahepatic portosystemic shunts

Abstract

Background

Dogs with congenital intrahepatic portosystemic shunt (IHPSS) are predisposed to gastrointestinal inflammation, ulceration, and bleeding, unlike dogs with congenital extrahepatic portosystemic shunt (EHPSS). Limited information is available about hematologic differences between dogs with IHPSS and dogs with EHPSS.

Objective

Compare hemogram variables between dogs with IHPSS and EHPSS. We hypothesized that hematologic variables would differ between the 2 populations, with a higher frequency and severity of anemia and microcytosis in dogs with IHPSS.

Animals

Twenty‐six client‐owned dogs with IHPSS and 35 client‐owned dogs with EHPSS.

Methods

Retrospective cross‐sectional study. Dogs were included if a CBC was performed before shunt attenuation. Contingency analysis was performed to determine if the frequency of clinical signs and of hematologic variables below the reference range differed between groups. Hematologic and selected biochemical variables were compared between groups using an analysis of covariance with age as a covariate.

Results

Gastrointestinal clinical signs (IHPSS, 81% vs EHPSS, 34%; P = .01), anemia (31% vs 6%; P = .01), microcytosis (77% vs 29%; P = .002), and hypochromia (77% vs 49%; P = .03) were more common in dogs with IHPSS than in dogs with EHPSS. Dogs with IHPSS had lower packed cell volume (34% vs 41%, P = .04), hemoglobin concentration (11.5 g/dL vs 13.7 g/dL, P = .03), mean corpuscular volume (57 fL vs 65 fL; P = .001), and mean corpuscular hemoglobin concentration (32 g/dL vs 33 g/dL; P = .04) than dogs with EHPSS.

Conclusions and Clinical Importance

Dogs with IHPSS had a higher frequency of anemia, microcytosis, and hypochromia and exhibited more gastrointestinal clinical signs.

Abbreviations

ALP

alkaline phosphatase

ALT

alanine aminotransferase

BUN

blood urea nitrogen

EHPSS

extrahepatic portosystemic shunt

GGT

gamma‐glutamyl transferase

HCT

hematocrit

Hg

hemoglobin

IHPSS

intrahepatic portosystemic shunt

MCH

mean corpuscular hemoglobin

MCHC

mean corpuscular hemoglobin concentration

MCV

mean corpuscular volume

PCV

packed cell volume

PSS

portosystemic shunt

RBC

red blood cells

RDW

red cell distribution width

1. INTRODUCTION

Congenital portosystemic shunt (PSS) is a common vascular anomaly in dogs and broadly categorized as intrahepatic (IHPSS) or extrahepatic (EHPSS). One of the most common hemogram abnormalities in dogs with PSS is microcytosis, characterized by a low mean corpuscular volume (MCV), and present in 60%‐72% of PSS dogs with or without anemia. ¹ , ² The specific pathogenesis for PSS‐associated microcytosis is yet to be determined, but possibilities include a deficiency in the transport of iron because of decreased transferrin synthesis (as a negative acute phase protein or because of decreased hepatic synthetic function), concurrent iron deficiency secondary to gastrointestinal hemorrhage, inflammation, abnormal lipid metabolism leading to alterations in erythrocyte membrane lipid content, or a combination of these abnormalities. ³ , ⁴ The iron status in dogs with a PSS has been evaluated. Previous studies in dogs with congenital PSS identified that these dogs have changes in iron status similar to animals with functional iron deficiency secondary to inflammatory disease, characterized by decreased serum iron concentration, decreased total iron‐binding capacity, and adequate or increased hepatic iron stores in the bone marrow and Kupffer cells. ¹ , ² , ⁵ , ⁶ Microcytosis can resolve in affected congenital PSS dogs after shunt attenuation. ¹ , ² , ⁷ , ⁸

Clinical differences between dogs with IHPSS and dogs with EHPSS could lead to differences in their respective hemograms. Unlike dogs with EHPSS, dogs with IHPSS are reported to have upper gastrointestinal inflammation, ulceration, and bleeding. ³ Gastrointestinal ulceration was reported in 15% of dogs with IHPSS before endovascular treatment and in 21% after endovascular treatment. ⁸ The frequency of gastrointestinal ulceration in dogs with IHPSS raises concern for the development of more marked hematologic abnormalities associated with acute and chronic blood loss, leading to iron deficiency and possibly more severe anemia in these dogs. The veterinary literature on hematologic changes and iron status in dogs with congenital PSS focuses primarily on dogs with EHPSS or those with surgically‐induced EHPSS. ⁵ Studies that include dogs with IHPSS often do not distinguish between dogs with IHPSS and EHPSS. ¹ , ² , ⁶ To our knowledge, a single published study directly compared clinicopathologic variables between dogs with IHPSS and EHPSS, and the evaluation was limited to only packed cell volume and MCV. ⁹ Furthermore, we have had a clinical suspicion of gastrointestinal blood loss in some dogs with IHPSS at presentation based on clinical signs (eg, melena) or after successful intravascular attenuation based on unresolved microcytosis and regenerative anemia despite resolution of biochemical abnormalities.

Our aim was to document differences in hematologic abnormalities between dogs with IHPSS and EHPSS. Our primary objective was to characterize differences in hemogram variables between these 2 groups. We hypothesized that dogs with IHPSS would have differences in hematologic variables compared with dogs with EHPSS, including a higher frequency and severity of anemia and microcytosis.

2. MATERIALS AND METHODS

Medical records were retrospectively reviewed from the Oregon State University Veterinary Teaching Hospital between 2011 and 2022. Dogs were included in the study if (1) they had a confirmed single congenital PSS diagnosed by computed tomography (CT) or ultrasonography performed by a board‐certified radiologist, and (2) they had a CBC performed at the Oregon Veterinary Diagnostic Laboratory (ADVIA 2120 Hematology Analyzer, Siemens Healthineers) before shunt attenuation. Complete blood counts were performed immediately after blood collection or after refrigeration overnight to minimize effects of storage on the variables assessed. Data were collected from the most recent CBC performed in preparation for shunt attenuation if multiple CBCs were performed within this time frame. Age at diagnosis, sex, neuter status, body weight, current medications, and clinical signs at the time of the CBC were recorded in the electronic medical record system by the primary clinician (rotating intern or internal medicine resident) under the direct supervision of a board‐certified internal medicine specialist. These findings then were retrieved from the electronic medical record system and collated by the primary author (Y.C.). Clinical signs were grouped by major organ system, with gastrointestinal signs including change in appetite, vomiting, hematochezia, diarrhea, and ptyalism; neurologic signs including circling, head pressing, pacing, ataxia, dullness, seizures; and, urinary signs including polyuria, polydipsia and pollakiuria. Melena and hematemesis were categorized separately from gastrointestinal signs because they imply upper gastrointestinal ulceration and erosion as opposed to hematochezia localizing to the large intestine. Measurement of red blood cells (RBC, 10⁶/μL), hemoglobin (Hb, g/dL), spun packed cell volume (PCV, %), hematocrit (HCT, %), mean corpuscular volume (MCV, fL), mean corpuscular hemoglobin (MCH, pg), mean corpuscular hemoglobin concentration (MCHC, g/dL), red cell distribution width (RDW, %), absolute reticulocyte count (/μL), platelet count (10³/μL), and mean platelet volume (MPV, fL) were recorded. Microscopic blood smear review for red blood cellular morphology, evidence of platelet clumping with manual platelet count, and presence of hemoparasites were recorded. Because of the low prevalence of hemoparasites in dogs in our geographical region, PCR testing for hemoparasites was not performed. Serum biochemistry was performed at the Oregon Veterinary Diagnostic Laboratory (AU480 Chemistry Analyzer, Beckman Coulter) and was included for analysis if performed on the same date as the CBC. The following serum biochemistry results that represent liver pathology or function were included: glucose (mg/dL), blood urea nitrogen (BUN, mg/dL), albumin (g/dL), total protein (g/dL), cholesterol (mg/dL), and total bilirubin (mg/dL) concentrations, and alanine aminotransferase (ALT, U/L), alkaline phosphatase (ALP, U/L), and gamma‐glutamyl transferase (GGT, U/L) activities.

2.1. Statistical analysis

Dogs were categorized into IHPSS and EHPSS groups based on shunt morphology. Mean corpuscular volume was assessed for covariance with reticulocyte count using Spearman rank correlation. Contingency analysis (Pearson's chi‐squared or Fisher's exact test as appropriate) was performed to determine if the frequency of clinical signs and hematologic variables falling below the reference range differed between groups. A Mann‐Whitney U test or Student t test was used to compare age and body weight, respectively, between groups. Statistical analysis was performed using GraphPad Prism (version 9.5.1 for Windows, GraphPad Software, Boston, Massachusetts USA) or Microsoft Excel (Microsoft Office 365, 2023). Data were adjusted for multiple comparisons using a 2‐stage linear step‐up Bejamini, Krieger, and Yekutieli procedure with a false discovery rate of 5%.

Analysis of covariance (ANCOVA) was performed with age as a covariate to investigate for covariate influences on the differences in variables between dogs with IHPSS and dogs with EHPSS. The contingency analysis was restricted to those CBC variables that were found to be statistically different between groups. The ANCOVA was performed using SAS (version 9.4 for Windows, SAS Institute Inc, Cary, North Carolina USA). The assumption of normality was evaluated by inspection of QQ‐plots, histograms, and skewness. Data that were judged not normally distributed were log‐transformed and evaluated again and assessed as being normally distributed. P‐values were adjusted for multiplicity using the adaptive linear step‐up method of Benjamini and Hochberg. A significance threshold of P < .05 was used.

3. RESULTS

3.1. Population characteristics

Thirty‐three dogs with IHPSS and 56 dogs with EHPSS were identified on medical record review. Seven dogs with IHPSS and 21 dogs with EHPSS were excluded because they had no record of a CBC completed at the Oregon Veterinary Diagnostic Laboratory before shunt attenuation. As such, 26 dogs with IHPSS and 35 dogs with EHPSS were included in the final analysis, for a total of 61 dogs. The majority (58/61, 95%) of all dogs had a CBC performed within 2 months of diagnosis, although a minority (3/61, 5%) had a CBC performed >6 months after diagnosis in preparation for shunt attenuation.

The IHPSS population consisted of 8 female intact, 2 female spayed, 14 male intact, and 2 male neutered dogs. The most common breeds were primarily large breed dogs and included mixed breed dogs (n = 8), Labrador retriever (n = 4), and golden retriever (n = 3). The median age at the time of CBC collection was 8 months (range, 4‐24 months) and the median body weight was 16.1 kg (range, 3.3‐39.9 kg). All dogs had CT angiography imaging performed to diagnose IHPSS, with 12 categorized as left divisional, 12 right divisional, and 2 central divisional.

The EHPSS population consisted of 9 female intact, 12 female spayed, 6 male intact, and 8 male neutered dogs. The most common breeds were primarily small breed dogs and included Pug (n = 5), mixed breed (n = 4), and Yorkshire terrier, Maltese, Havanese, and shih tzu (n = 3 each). The median age at the time of CBC collection was 24 months (range, 3‐120 months), and the median weight was 4.4 kg (range, 1.2‐33 kg). Thirty‐three dogs had CT angiography imaging performed, with the 2 remaining dogs having abdominal ultrasonography. The most common shunt morphology was splenocaval (26%; 9/35) followed by splenophrenic (23%; 8/35), right gastrocaval (20%; 7/35), portoazygous (11%; 4/35), splenoazygous (9%; 3/35), portocaval (9%; 3/35), and left gastrocaval (3%; 1/35).

For dogs with IHPSS, the most common presenting clinical signs were of neurological origin, followed by gastrointestinal signs and nonspecific lethargy (Table 1). Melena or hematemesis was reported in 3 dogs (3/26, 12%), and 2 of these dogs had a history of at least 1 red blood cell transfusion for treatment of anemia before collection of blood for CBC. Of these 2 dogs, 1 dog had a transfusion performed 1 month before its CBC and no further transfusion information was available for the second dog. For dogs with EHPSS, the most common presenting clinical signs were of neurological origin, followed by urinary signs, nonspecific lethargy and gastrointestinal signs (Table 1). No dogs with EHPSS had a history of reported melena or hematemesis, although 1 dog had a history of packed RBC transfusion because of progressive regenerative anemia (PCV, 10%) 3 months before its CBC. This dog did not have overt clinical signs of bleeding or gastrointestinal ulceration documented in the medical record, and the cause for the anemia was not determined. Only 1 dog with IHPSS and none of the dogs with EHPSS had ascites. Only 1 dog with IHPSS and 1 dog with EHPSS had hematochezia.

TABLE 1.

The frequency of clinical signs recorded at the time of diagnosis for dogs diagnosed with an intrahepatic portosystemic shunt (IHPSS, n = 26) and dogs with an extrahepatic portosystemic shunt (EHPSS, n = 35).

	IHPSS		EHPSS
	n	%	n	%	P‐value	Significance level
Neurological signs (circling, head pressing, pacing, ataxia, dull, seizure)	22	85	20	57	.09
Lethargy	15	58	12	34	.11
Gastrointestinal (hypo/anorexia, vomiting, hematochezia, diarrhea, ptyalism)	21	81	12	34	.01	**
Melena/hematemesis	3	12	0	0	.11
Urinary (polyuria/polydipsia, pollakiuria)	12	46	15	43	.84
Failure to thrive	6	23	8	23	.89

Note: Adjusted P‐value with the associated significance level (*P ≤ .05; **P ≤ .01; ***P ≤ .001) for each clinical sign is provided.

Dogs with IHPSS had a lower median age (P < .001) and higher median body weight (P < .001) at diagnosis than dogs with EHPSS. In our study, 6 of 26 (23%) dogs with IHPSS and 4 of 35 (11%) dogs with EHPSS were ≤6 months of age at the time of their CBC. More dogs with IHPSS experienced gastrointestinal clinical signs compared with dogs with EHPSS (P = .01). The frequency of neurologic signs, lethargy, melena, hematemesis, urinary signs, and failure to thrive was not significantly different between groups (Table 1). Seven dogs with IHPSS (7/26; 27%) were receiving a proton‐pump inhibitor (omeprazole) at the time of their CBC, of which 6 were anemic (Hct <36%), whereas none of the dogs with EHPSS were receiving a proton‐pump inhibitor or other gastric acid suppressant.

3.2. CBC findings before shunt attenuation

Compared with dogs with EHPSS, dogs with IHPSS had significantly lower Hb, PCV, MCV, MCH, MCHC, BUN, and higher cholesterol (Table 2). One EHPSS dog without anemia (Hct >36%) had an extremely high MCV (90.2 fL). Reticulocyte count for this dog was not available in the medical record, but only a few polychromatophils were noted on blood smear review. This pet was a toy poodle, a breed known to exhibit macrocytosis. ¹⁰ When this outlier was removed, a significant difference remained between groups for MCV (P < .001). Significantly more dogs with IHPSS than dogs with EHPSS were below the reference range for Hb, PCV, MCV, MCH, and MCHC (Table 3). Fourteen dogs with IHPSS (14/26, 54%) and 4 dogs with EHPSS (4/35, 11%) had anemia characterized by Hct <36%, of which the majority of these dogs with IHPSS (11/14) and all of the dogs with EHPSS had reticulocyte counts performed. Of the anemic dogs with a reticulocyte count, only 1 of 11 of the dogs with IHPSS had a regenerative response (defined as reticulocyte count >126 × 10³/μL or increased polychromatophils noted on blood smear) and none of the dogs with EHPSS had a regenerative response. When analyzing MCV for covariance with reticulocyte count, there was no correlation in dogs with IHPSS (r = 0.17, P = .5) and dogs with EHPSS (r = −0.13, P = .6).

TABLE 2.

Hematologic and selected serum biochemistry variables between dogs with congenital intrahepatic portosystemic shunt (IHPSS) and dogs with congenital extrahepatic portosystemic shunt (EHPSS).

		IHPSS		EHPSS
	Reference interval	n	Median (range)	n	Median (range)	P‐value	Significance level
RBC (106/μL)	5.20‐8.50	26	6.3 (4.2‐8.7)	35	6.4 (4.5‐7.8)	.47
Hb (g/dL)	11.5‐20.5	26	11.5 (5.4‐15.8)	35	13.7 (8.8‐17.2)	.03	*
Hct (%)	36.0‐58.0	26	35.2 (20.2‐48.9)	35	42.5 (26.4‐51.2)	.06
PCV (%)	31‐57	26	34 (20‐46)	35	41 (26‐49)	.04	*
MCV (fL)	63.0‐77.0	26	57.0 (38.8‐73.0)	35	65.3 (53.9‐90.2)	.001	***
MCH (pg)	20.0‐27.0	26	18.5 (10.4‐23.4)	35	21.3 (17.1‐29.9)	.001	***
MCHC (g/dL)	33.0‐35.0	26	32.0 (22.8‐34.1)	35	33.0 (30.1‐35.9)	.04	*
RDW	12‐14	26	13.3 (11.4‐18.8)	35	13 (11‐17)	.22
Reticulocytes (/μL)	12 000‐126 000	19	60 300 (9000‐122 092)	19	58 900 (10 800‐167 445)	.39
PLT (103/μL)	150‐470	26	170 (66‐447)	35	221 (103‐421)	.07
BUN (mg/dL)	8‐26	26	5 (2‐20)	35	7 (3‐19)	.03	*
Glucose (mg/dL)	67‐124	26	92 (59‐119)	35	91 (49‐124)	.49
Cholesterol (mg/dL)	144‐367	26	173 (57‐418)	35	119 (32‐304)	.04	*
Total protein (g/dL)	5.2‐7.0	26	4.6 (2.7‐5.6)	34	5.1 (3.7‐7.3)	.27
Albumin (g/dL)	2.6‐3.8	25	2.4 (1.8‐3.0)	35	2.5 (1.5‐3.3)	.07
Total bilirubin (mg/dL)	0.1‐0.6	26	0.3 (0.1‐0.5)	35	0.2 (0.1‐0.6)	.45
ALP (U/L)	10‐115	26	154 (58‐263)	35	106 (17‐599)	.40
GGT (U/L)	1‐6	26	3 (0‐8)	35	3 (0‐26)	.64
ALT (U/L)	18‐97	26	79 (25‐351)	35	88 (11‐1388)	.12

Note: Data presented as counts and median with range (minimum‐maximum). Adjusted P‐value of ANCOVA with age as covariate and the associated significance level (*P ≤ .05; **P ≤ .01; ***P ≤ .001) for each variable is provided.

TABLE 3.

The frequency of selected hematologic variables falling below the reported reference range between dogs with an intrahepatic portosystemic shunt (IHPSS, n = 26) and extrahepatic portosystemic shunt (EHPSS, n = 35), displayed as the number and percentage of dogs below the reference range.

		IHPSS		EHPSS
	Reference interval	n Below	%Below	n Below	%Below	P‐value	Significance level
Hb (g/dL)	11.5‐20.5	13	50	3	9	.002	**
PCV (%)	31‐57	8	31	2	6	.01	*
MCV (fL)	63‐77	20	77	10	29	.002	**
MCH (pg)	20‐27	19	73	10	29	.002	**
MCHC (g/dL)	33‐35	20	77	17	49	.03	*

Note: Adjusted P‐value with associated significance level (*P ≤ .05; **P ≤ .01; ***P ≤ .001) is provided.

4. DISCUSSION

We found that dogs with IHPSS had a higher frequency of anemia, microcytosis, and hypochromia when compared to laboratory reference ranges. Even when not outside reported reference ranges, dogs with IHPSS had lower median Hb, PCV, MCV, MCH, and MCHC than dogs with EHPSS, emphasizing the need for careful evaluation of laboratory results and trends in dogs with portosystemic shunting. Complete evaluation of iron status and systemic inflammation was not performed on the dogs in our study, and thus the cause of more marked anemia and microcytosis in the dogs with IHPSS was undetermined. However, similar to certain dogs in our study, previous studies reported signs of upper gastrointestinal bleeding (ie, melena or hematemesis) in dogs with IHPSS before shunt attenuation, raising concern for both acute and chronic gastrointestinal hemorrhage. ⁸ In our study, the frequency of melena or hematemesis was low in dogs with IHPSS and no dog with EHPSS exhibited clinical signs of upper gastrointestinal bleeding. Estimations from studies in humans indicate that 50 to 100 mL of blood is needed to cause melena. ¹¹ , ¹² Because large volumes of blood are needed to identify melena, the absence of melena does not exclude subclinical gastrointestinal ulceration, erosion, and hemorrhage. Inconsistent owner observation of their dog's feces could also contribute to underreporting of melena. As such, upper gastrointestinal bleeding is likely an underestimated cause of anemia in this population. Additionally, we were unable to compare hemogram variables before and after shunt attenuation because only a few of these dogs had long‐term follow‐up (>6 months) at our institution, and as such persistence of anemia and microcytosis after attenuation could not be evaluated.

Upper gastrointestinal endoscopy in dogs with IHPSS indicates that mucosal petechiation and active or healing ulceration are common. ³ A study involving 100 dogs with IHPSS undergoing shunt attenuation found that 21% of dogs with long‐term follow‐up after transvenous coil embolization had signs consistent with gastrointestinal hemorrhage and ulceration, primarily melena, of which 15% had signs before the procedure. ⁸ Postoperative gastrointestinal bleeding was also associated with shorter survival time (929 vs 2535 days). ⁸ The frequency of portal hypertension after coil embolization is low (<1%), making it an unlikely cause of the gastrointestinal hemorrhage observed during follow‐up. ⁸ , ¹³ Thus, the authors of that study prescribed life‐long antacid treatment of gastrointestinal ulceration, which they report decreased the frequency of gastrointestinal hemorrhage and improved postoperative survival times. ⁸ Some dogs with IHPSS included in our study were receiving omeprazole at the time of their CBCs, which might have contributed to an underestimated frequency of anemia by decreasing the frequency of gastric ulceration.

Despite reports of gastrointestinal bleeding in dogs with IHPSS, limited information is available on iron status specifically in dogs with IHPSS to support absolute iron deficiency. Researchers have evaluated the iron status in dogs with congenital PSS to determine whether functional or absolute iron deficiency is a cause of microcytosis in this patient population. ¹⁴ Functional iron deficiency occurs when there is insufficient incorporation of iron into reticulocytes despite adequate iron stores in the body. Absolute iron deficiency occurs when total body iron stores are low, and in dogs this deficiency is most commonly caused by chronic gastrointestinal bleeding. ¹⁴ Previous studies support functional iron deficiency in dogs with congenital PSS characterized by decreased serum iron concentrations, decreased total iron‐binding capacity, and adequate iron stores in the bone marrow and liver. ¹ , ² , ⁵ , ⁶ However, the published literature primarily reports on dogs with EHPSS or does not compare findings between IHPSS and dogs with EHPSS. However, 1 study noted that dogs with IHPSS had lower hepatic stainable iron scores compared with dogs with EHPSS, further supporting the possibility that some dogs with IHPSS may have absolute iron deficiency. ¹ Further research is needed to determine if absolute iron deficiency occurs in this population. If so, it could be a therapeutic target to improve long‐term outcome for dogs with IHPSS, such as iron supplementation and additional peptic ulcer treatment.

Among the biochemistry analytes included in the statistical analysis, dogs with IHPSS had significantly lower BUN and higher serum cholesterol concentrations. Urea and cholesterol are synthesized by the liver, and as such serum concentrations may be lower in animals with portosystemic shunting. ⁴ It is possible that dogs with IHPSS in our population had a higher fraction of shunted blood than the dogs with EHPSS, contributing to more marked decreases in hepatic perfusion and synthesis of urea. However, many other factors influence BUN and serum cholesterol concentrations, such as diet, upper gastrointestinal bleeding, renal insufficiency, and gastrointestinal malabsorption, and therefore the relevance of these differences is unclear. Further investigation into the differences in serum biochemical variables between dogs with IHPSS and dogs with EHPSS should control for these additional factors.

In our study, dogs with IHPSS were significantly more likely to have gastrointestinal clinical signs. A previous report documented that >85% of dogs with IHPSS had evidence of moderate to severe lymphoplasmacytic (with or without eosinophilic) mucosal and submucosal inflammation present on endoscopic biopsy samples, which suggests that dogs with IHPSS have a concomitant gastrointestinal disorder. ³ , ⁸ The effect of diet on gastrointestinal clinical signs was not evaluated in this study given inconsistent reporting. Further investigation into the pathophysiology of gastrointestinal inflammation and clinical signs in dogs with IHPSS is needed to help guide pre‐ and postoperative monitoring and treatment recommendations.

The dogs with IHPSS in our study were younger than the dogs with EHPSS. The higher frequency of gastrointestinal signs in dogs with IHPSS might have contributed to more expedient veterinary evaluation and diagnosis at a younger age. Because hemogram results can differ by age in dogs, an analysis of covariance was performed to control for age in the statistical analysis. The neonatal RBC exhibits macrocytosis with MCV decreasing to adult values by 4 weeks of age. ¹⁵ Hematologic variables of dogs decrease markedly by 3 days of age and continue to decrease over the next several weeks, and then incremental increases in RBC indices can be detected by 2 months of age with adult reference ranges usually reached by 2 to 6 months of age. ¹⁵ None of the dogs in our study were <4 months of age. Ten dogs (16%) were between 4‐6 months of age. Furthermore, certain breeds are known to have nonpathological macrocytosis (sighthounds, toy and miniature poodles) ¹⁰ , ¹⁶ and others are known to have nonpathological microcytosis (Japanese breeds). ¹⁷ Because these breeds were not common in our population, these breed‐specific variations were not considered likely to have affected the statistical analysis.

Our study had some limitations. First, previous studies have evaluated the utility of reticulocyte indices, such as reticulocyte hemoglobin content, in dogs with anemia, and 1 study established certain cut‐offs to help distinguish absolute iron deficiency, anemia of inflammatory disease, PSS, and breed‐associated microcytosis based on reticulocyte hemoglobin concentration and content. ⁴ However, reticulocyte indices are not reported at our institution, and as such comparison of reticulocyte indices between dogs with IHPSS and dogs with EHPSS was not possible in our study. Second, in most cases CBCs are performed immediately upon blood collection and submission to the laboratory at our institution. However, it is possible that some samples were stored and refrigerated overnight. Mean corpuscular volume increases and MCHC decreases upon storage over 24 hours, and more rapidly in unrefrigerated samples. ¹⁸ , ¹⁹ Proper overnight refrigerated storage likely minimized this artifact in our population. Third, evidence of gastrointestinal hemorrhage was solely based on reported clinical signs in the medical records and no other tests for iron storage were performed to help elucidate differences in total body iron status in our population of dogs. Future studies investigating gastrointestinal hemorrhage in dogs with PSS could include fecal occult blood, gastrointestinal endoscopy, serum iron panels, liver iron quantification, and bone marrow iron evaluation. These additional tests could help alleviate the likely underestimation of gastrointestinal hemorrhage when relying on clinical history, as previously discussed. Because inflammation can lead to functional iron deficiency by upregulating hepcidin and thus causing sequestration and decreased absorption of iron, iron panels should be interpreted concurrently with markers of inflammation, such as C‐reactive protein, hepcidin, and leukograms. ⁷ , ²⁰ , ²¹ Fourth, 3 dogs (2 dogs with IHPSS and 1 dog with EHPSS) had received blood transfusions, 2 of which were within 3 months of their CBCs. After an initial decrease within the first 24 hours posttransfusion, at least 70% of transfused red blood cells are expected to survive and have a normal lifespan (approximately 120 days). ²² , ²³ , ²⁴ As such, recent transfusion in these dogs may have impacted analysis of CBC variables.

In conclusion, dogs with IHPSS had significantly lower Hb, PCV, MCV, MCH, and MCHC and a higher frequency of anemia, microcytosis, and hypochromia than dogs with EHPSS. Dogs with IHPSS also had a higher frequency of gastrointestinal signs (hyporexia, anorexia, vomiting, hematochezia, diarrhea, ptyalism). Further investigation is needed to evaluate iron status and gastrointestinal health in dogs with IHPSS.

CONFLICT OF INTEREST DECLARATION

Authors declare no conflicts of interest.

OFF‐LABEL ANTIMICROBIAL DECLARATION

Authors declare no off‐label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION

Authors declare no IACUC or other approval was needed.

HUMAN ETHICS APPROVAL DECLARATION

Authors declare human ethics approval was not needed for this study.

Couture Y, Keys D, Summers S. Comparison of hematologic variables between dogs with congenital intrahepatic and extrahepatic portosystemic shunts. J Vet Intern Med. 2024;38(3):1458‐1464. doi: 10.1111/jvim.17081