Serial monitoring of pancreatic lipase immunoreactivity, C‐reactive protein, abdominal ultrasonography, and clinical severity in dogs with suspected pancreatitis

Abstract

Background

Diagnosis of pancreatitis is based on clinical signs, pancreatic lipase immunoreactivity (cPLI), and abdominal ultrasonography (AUS). Diagnostic discrepancies exist between test results which might be related to differences in the timeline for resolution of these abnormalities after pancreatic injury.

Hypothesis/Objectives

To evaluate disease severity, ultrasonographic findings, and serum biomarkers of pancreatitis in dogs over a period of 28‐days.

Animals

Sixteen client‐owned dogs with a clinical suspicion for acute pancreatitis based on history/physical examination, an abnormal SNAP cPLI, and ultrasonographic evidence of pancreatitis.

Methods

Prospective observational study. Clinical severity (modified clinical activity index [MCAI]), cPLI, C‐reactive protein (CRP), and AUS were evaluated at days 0, 2, 7, and 28. Owner assessed overall health (OH) was noted. Dogs were stratified into baseline cPLI ≥400 μg/L vs <400 μg/L groups for reporting.

Results

The median CRP, MCAI, and OH were 111.9 mg/L, 10, and 4/10 respectively in the cPLI ≥400 μg/L group. The median CRP, MCAI, and OH were 58.0 mg/L, 6, and 6/10 respectively in the cPLI <400 μg/L group. None of these variables were significantly different between groups. Most dogs (4/5) in the cPLI <400 μg/L group had a history of suspected pancreatitis (ie, suspect acute on chronic disease). cPLI and MCAI rapidly decreased in dogs with a baseline cPLI ≥400 μg/L, whereas sonographic evidence of pancreatitis persisted for a longer time period.

Conclusions and Clinical Importance

Ultrasonographic evidence of pancreatitis in the absence of overt clinical or biochemical abnormalities might represent a resolving injury rather than active disease.

Abbreviations

AP

acute pancreatitis

AUS

abdominal ultrasound

CP

chronic pancreatitis

CRP

C‐reactive protein

DGGR

1,2‐o‐dilauryl‐rac‐glycero‐3‐glutaric acid‐(6′‐methylresorufin) ester

IQR

interquartile range

MCAI

modified canine activity index

OH

overall health

PLI

pancreatic lipase immunoreactivity

QOL

quality of life

UPASS

ultrasonographic pancreatic assessment severity score

1. INTRODUCTION

Pancreatitis is the most common disorder of the exocrine pancreas in dogs. ¹ Despite its common occurrence, the diagnosis remains challenging. Histopathology was once considered the diagnostic gold standard for pancreatitis in dogs but the presence of localized inflammatory lesions, that can be missed by biopsy sampling, and the invasive nature of sample collection means that this approach is no longer routinely used. ² , ³ Most clinicians now utilize a clinical best practice standard comprised of an overall assessment of a dogs history, physical examination findings, serum biomarker concentrations, and abdominal ultrasonographic findings. Common serum biomarkers evaluated in these dogs include pancreatic lipase concentration (cPLI) or activity and C‐reactive protein (CRP) concentration. Serum cPLI concentrations are highly sensitive and specific for a diagnosis of acute pancreatitis (AP), but this biomarker has a reduced diagnostic sensitivity in dogs with chronic pancreatitis (CP). ⁴ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ C‐reactive protein is a biomarker of systemic inflammation and concentrations are increased in dogs with AP. ¹² While the clinical diagnosis approach to pancreatitis offers important advantages over histopathology, it is not without its limitations including the presence of discrepancies between test results and the low sensitivity of serum biomarkers in CP. ³ Serum cPLI can be abnormal in cases in which pancreatitis is not the major disease process, suggestive of a secondary pancreatopathy.

Diagnostic discrepancies are noted between ultrasonographic findings and both catalytic and immunologic lipase assays. One study noted a fair level of agreement between 1,2‐o‐dilauryl‐rac‐glycero‐3‐glutaric acid‐(6′‐methylresorufin) ester (DGGR) activity and pancreatic ultrasonographic findings in dogs (κ = 0.29‐0.35), while another study documented a weak correlation between cPLI concentrations and pancreatic ultrasonographic findings in dogs (r _s = .0178). ¹³ , ¹⁴ Both of these aforementioned studies were retrospective meaning that the timing of serum sample collection relative to ultrasonographic examination was unable to be well controlled and this could have contributed to the reported diagnostic discrepancies. This is of particular relevance when we consider the short half‐life of pancreatic lipase (2 hours) in healthy dogs (Dossin O, Rick M, Ridge T, et al. Pharmacokinetics of pancreatic lipase in healthy dogs [Abstract]. In: European College of Veterinary Internal Medicine Congress. Seville, Spain. 2011:238). These diagnostic discrepancies prompted additional studies in this area. ¹² , ¹⁵ , ¹⁶ These studies strengthened evidence that ultrasonographic evidence of pancreatitis and lipase activity do not follow the same clinical timeline for onset and resolution after pancreatic injury but are limited by the short‐term nature of follow up, missing or limited data, the presence of concurrent disease, exclusion of suspected CP cases, or their retrospective nature.

Our study had 2 aims. First, to prospectively evaluate disease severity, ultrasonographic findings, cPLI, and CRP concentrations in dogs with clinically acute presentations of suspected pancreatitis over 28‐days. Second, to compare findings in dogs with elevated (≥400 μg/L) vs gray‐zone (200‐399 μg/L) cPLI concentrations.

2. MATERIALS AND METHODS

2.1. Study overview

Prospective observational study. From June 2021 to March 2023, 24 client‐owned dogs presenting to a university teaching hospital with suspected clinically acute pancreatitis were enrolled in the study. Five dogs dropped‐out of the study, and 3 dogs had a cPLI <200 μg/L leaving a final study size of 16 dogs. Initial clinical suspicion for pancreatitis was based on 3 factors, (i) an abnormal cage‐side pancreatic lipase assay result (SNAP cPLI, IDEXX Laboratories Inc, Westbrook, ME), (ii) a minimum of 2 of the following clinical signs: hyporexia, vomiting, diarrhea, lethargy, and abdominal discomfort, and (iii) at least 2 sonographic features of acute pancreatitis, which included pancreatic enlargement, hypoechoic pancreatic parenchyma, and an abnormal peripancreatic mesenteric echogenicity. ¹⁷ Exclusion criteria included dogs with a PCV <15% (because of blood draw volume requirements) and dogs with major concurrent disease. Major concurrent disease included dogs with IRIS acute kidney injury (AKI) grade ≥3 and dogs with below reference interval serum cobalamin concentrations (when measured). Other excluded dogs were based on a consensus agreement between a board‐certified internal medicine specialist and an ACVIM small animal internal medicine candidate.

After case identification, the owners of each dog were contacted by study investigators for voluntary enrollment in the study. If permission was granted informed consent was documented. The study protocol was approved by the Michigan State University IACUC (PROTO202000315). Immediately after enrollment a serum sample was collected for quantification of cPLI (Spec cPL, Texas A&M Gastrointestinal Laboratory, College Station, TX) and CRP (Gentian canine‐specific immunoturbidimetric CRP assay, Cornell University Animal Health Diagnostic Center, Ithaca, NY; collected within 15 minutes of enrollment, and within 12 hours of abdominal ultrasonography). Samples were stored on‐site in a −80°F freezer until batch analysis at each laboratory. The cPLI and CRP assays utilized have been validated for use in dogs. ¹⁸ , ¹⁹ Owners were also asked to complete a survey outlining the clinical features of pancreatitis (Supplementary Material 1 and 2). The modified canine activity index (MCAI) was also calculated by study investigators as an indicator of clinical disease severity. ¹² To prevent confusion about what this clinical index describes we have elected to use the alternate term modified clinical activity index, while continuing to use same acronym (MCAI). Owners were also questioned regarding prior episodes of pancreatitis or clinical signs suggestive of pancreatitis. If historical episodes (<3 months previously) of pancreatitis or related clinical signs were noted, or if extensive questioning revealed a more chronic (but mild) history of GI signs this was noted in the study file. The AUS, cPLI, CRP, clinical signs questionnaire, and MCAI were repeated on days 2, 7, and 28 after enrollment (see Figure 1). On the rare occasion that 1 of these follow up examinations occurred on an unavailable date (eg, university holiday) the appointment was scheduled for the next available day.

FIGURE 1.

Study timeline. This figure visually outlines the screening criteria and subsequent analyses that were performed at days 0, 2, 7, and 28 post enrollment on each dog in the study.

2.2. Abdominal ultrasonographic protocol and review

A complete AUS, consistent with the criteria outlined in the ACVR and ECVDI consensus statement, was performed at each visit. ²⁰ All AUS procedures were performed on a Canon Aplio i700 ultrasound (Canon Medical Systems USA, Inc, Tustin, CA) by a board‐certified radiologist or a trained sonographer with extensive experience in veterinary ultrasonography. When the initial AUS was performed by the sonographer, a board‐certified internal medicine specialist or radiologist reviewed images to confirm study eligibility. In addition to the standard views outlined in the consensus statement, a color doppler image of the portal vein in long axis was also captured to evaluate for potential hepatopedal blood flow, in addition to a view of the caudal vena cava in left lateral recumbency at the right 10th‐13th intercostal space, to assist in ruling out portal hypertension as a cause of pancreatic edema. All stored images and cine clips were reviewed in a blinded fashion at study completion by JL or HC. Individual imaging features and a composite organ score (UPASS) were recorded, as utilized in a prior study. ¹⁴ The UPASS is an indicator of the number of sonographic abnormalities of the pancreas. The presence or absence of gastrointestinal (GI) wall abnormalities were also recorded at each visit. ²¹ , ²²

2.3. Grouping of enrolled dogs

Dogs were subsequently stratified into groups based on the pancreatic lipase immunoreactivity concentration at day 0 (enrollment), (i) cPLI ≥400 μg/L, and (ii) cPLI <400 μg/L. The frequency of prior signs associated with GI disease were also noted.

2.4. Data analysis

A prestudy sample size calculation was performed to determine potential differences in AUS findings between dogs with a cPLI ≥400 μg/L, and those with a cPLI <400 μg/L. A study size of 10 dogs in each group was recommended, to detect a difference of 2 in UPASS (a pancreatic organ score) between the populations with a power 0.8. ¹⁴ The type 1 error probability associated with this test of the null hypothesis is 0.05. Data sets were assessed for normality by Shapiro‐Wilk testing. Normally distributed data were reported as means ± SD, whereas nonnormally distributed data were reported as medians and interquartile range (IQR). The Mann‐Whitney test, with Bonferroni correction, was used to compare each parameter at baseline between the 2 subgroups. Wilcoxon signed ranks test, with Bonferroni correction, was employed for paired comparisons between 2 time points for each parameter. Bootstrapping technique with 500 replications was adopted to calculate 95% confidence intervals for medians at each time point for each parameter. Statistical analyses were performed using R Statistical Software (v4.3.0; R Core Team 2021). For all analyses, a P < .05 was considered statistically significant.

3. RESULTS

3.1. Animals

The median age of dogs was 11.5 years (IQR: 6.5 years). Of the study dogs, 9 were female spayed and 7 were male neutered. There were no intact dogs. The median weight of study dogs was 10.9 kg (IQR: 16.2 kg). The study group encompassed the following breeds: mixed breed dog (n = 8), Australian Cattle Dog (n = 2), Yorkshire Terrier (n = 2), Bichon Frise (n = 1), Maltese (n = 1), Miniature Pinscher (n = 1), and Welsh Terrier (n = 1). No dogs had hyporexia and lethargy alone. No dogs died during the study period.

3.2. Comparing dogs at day 0 with an initial cPLI ≥400 μg/L vs <400 μg/L

Eleven dogs had baseline cPLI concentrations ≥400 μg/L, while 5 dogs had an abnormal SNAP cPL but a cPLI concentration < 400 μg/L. Four of 5 dogs with a baseline cPLI <400 μg/L had a chronic history of clinical signs consistent with pancreatitis. ¹⁷ Whereas only 1 of the 11 dogs with a baseline cPLI ≥400 μg/L had a chronic history of clinical signs consistent with pancreatitis. ¹⁷

When comparing these groups there were no significant differences between CRP, MCAI, and OH. Median values are reported in Table 1.

TABLE 1.

Baseline comparison between dogs with and without a Spec cPL ≥400 μg/L.

DAY 0	CRP	MCAI	UPASS	OH
Spec cPL ≥400 μg/L (n = 11)	111.9 (166.7)	10.0 (2.3)	7 (1)	4.0 (3.3)
Spec cPL <400 μg/L (n = 5)	56.4 (95.5)	7.5 (3.8)	5.5 (3)	6.5 (4.3)
Statistical comparison	P = 1	P = .56	P = .84	P = 1

Abbreviations: CRP, C‐reactive protein; MCAI, modified clinical activity index; OH, owner assessed overall health; UPASS, ultrasonographic pancreatic assessment severity score; X (Y), median (IQR).

3.3. Comparing cPLI, CRP, UPASS, MCAI, and OH over time

3.3.1. Dogs with a baseline cPLI ≥400 μg/L

Median cPLI concentrations were not significantly different between day 0 and 2 (P = .055) or between day 7 and 28 (P = .55) but were different between days 0 and 7 (P = .003), between day 2 and 7 (P = .015), and between day 2 and 28 (P = .015). Similarly, median CRP concentrations were not significantly different between day 0 and 2 (P = .068) or between day 7 and 28 (P = .194) but were different between all other time points (P = .002‐.003). Median UPASS were not significantly different between day 0 and 2 (P = .107), day 0 and 7 (P = .061), or day 2 to 7 (P = .517) but were improved when day 0 or day 2 were compared directly to day 28 (P = .022‐.038). Owner assessed OH scores were not significantly different between day 0 and 2 (P = .083) but were significant at all other time comparisons (P = <.001‐.009). Median and 95% CI (bootstrap) values for each timepoint are provided (Table 2). Graphical representations of this data are also provided (Figure 2).

TABLE 2.

Spec cPL, CRP, UPASS, MCAI, and OH scores over time.

	Day 0		Day 2		Day 7		Day 28
Baseline group	cPLI ≥400 (n = 11)	cPLI <400 (n = 5)	cPLI ≥400 (n = 11)	cPLI <400 (n = 5)	cPLI ≥400 (n = 11)	cPLI <400 (n = 5)	cPLI ≥400 (n = 11)	cPLI <400 (n = 5)
Spec cPL (μg/L)	1990 (1184‐2000)a	245.5 (177‐382)	402.5 (207‐1349)a	189 (48‐374)	156.0 (101‐475)b	131 (48‐253)	144 (70‐202)b	122 (65‐295)
CRP (mg/L)	111.9 (38‐207)a	56.4 (5‐207)	45.2 (31‐82)a	49.1 (5‐144)	11.5 (6‐25)b	9.3 (6‐61)	5 (5‐7.8)b	6.5 (5‐10)
UPASS	7 (6‐7)a	5.5 (4‐7)	6 (5‐6)ab	4 (2‐6)	4 (4‐5)ab	3.5 (0‐4)	3 (2‐5)b	2 (2‐3)
MCAI	10 (8.5‐11)a	8 (4‐10)	5 (3.5‐6)b	5 (3‐9)	2 (1‐4)c	2 (0‐3)	0 (0‐2)c	1 (0‐1)
OH	4 (1.5‐5)a	6.5 (0‐8)	6.5 (5‐9)ab	6 (4‐6)	8.5 (7‐10)bc	8 (5‐10)	10 (9.5‐10)c	9.5 (8‐10)

Note: Values without shading represent the cPLI ≥400 μg/L group while the gray shading represents the cPLI <400 μg/L. Superscript is utilized to indicate statistical significance, a change in superscript letter over time points indicates an adjusted P value <.05. No comparisons in the cPLI <400 μg/L were significant and as such superscript were not added to these values.

Abbreviations: CRP, C‐reactive protein; MCAI, modified clinical activity index; OH, owner assessed overall health; UPASS, ultrasonographic pancreatic assessment severity score; X (Y), median (95% CI via bootstrap method).

FIGURE 2.

Graphical representation of changes in biochemical values and scoring systems over time. This figure represents changes in biochemical parameters and scoring systems over time in dogs with a high (Spec cPL ≥400 μg/L) and low baseline (Spec cPL < 400 μg/L) pancreatic lipase concentrations. The center point represents the median value, while the lines represent the 95% CI (bootstrap method). Superscript is utilized to indicate statistical significance, a change in superscript letter over time points indicates an adjusted P value <.05. No comparisons in the cPLI <400 μg/L were significant and as such superscript was not added to these values. CRP, C‐reactive protein; MCAI, modified canine activity index; OH, owner assessed overall health; Spec cPL, pancreatic lipase concentration; UPASS, ultrasonographic pancreatic assessment severity score.

3.3.2. Dogs with an Initial cPLI <400 μg/L

Median cPLI (P = 1.00) and CRP (P = .188‐1.00) concentrations were not significantly different between any time points (P = .188‐1.00). UPASS (P = .128‐1.00), MCAI (P = .134‐1.00) scores, and owner assessed OH (P = .585‐1.00) scores were not significantly different between any time points. Median and 95% CI (bootstrap) values for each timepoint are provided (Table 2). Graphical representations of this data are also provided (Figure 2).

4. DISCUSSION

In this study we prospectively describe the progression of clinical signs, ultrasonographic findings, pancreatic lipase, and C‐reactive protein concentrations across a 28‐day period in dogs with clinically acute presentations of suspected pancreatitis. ²³ This follow up period is longer than previously reported and allows for improved comparison between clinical, biochemical, and ultrasonographic indicators of pancreatitis. ¹⁶ Additionally we compare findings in dogs with baseline cPLI concentrations ≥400 μg/L with those <400 μg/L (with a concurrent abnormal SNAP cPL). This later group could be postulated to represent acute worsening of chronic disease, as enzyme leakage from a pancreas suffering from chronic disease is reduced. ²⁴ The majority (4/5) of dogs in the baseline cPLI concentrations <400 μg/L group also had a chronic history of clinical signs consistent with pancreatitis, compared to a minority (1/11) of dogs in the baseline cPLI concentrations ≥400 μg/L group, which further supports this hypothesis.

In dogs with baseline cPLI concentrations ≥400 μg/L, median cPLI concentrations fell between each measured time point, with significant decreases noted at all time points except in the acute phases of hospitalization (day 0‐2). Similar patterns were also noted in CRP, a biomarker of systemic inflammation. Our CRP findings mirror that of a recent study that noted minimal changes in CRP concentrations during a 48 hour (brief) period of monitoring. ²⁵ Lipase activity decreases to near or within RI within approximately 2 days of treatment this is similar to our group in which the median cPLI concentration at day 2 was 402.5 μg/L, just above the diagnostic threshold for pancreatitis. ¹⁶ When compared to cPLI, CRP and OH, the composite ultrasonographic score (UPASS) did not drop significantly during the early phases of the disease and significant decreases were only noted between days 0 and 28, and between days 2 and 28, suggesting a slower rate of improvement relative to inflammatory biomarkers and clinical signs. This mirrors what has been reported anecdotally in that ultrasonographic resolution of pancreatitis can lag behind resolution of clinical signs and immunologic pancreatic lipase assay results. A recent retrospective study analyzing a more brief time period also noted that a history of pancreatitis was correlated with ultrasonographic evidence of pancreatitis, but not lipase activity. ²⁶ The collective results interpretation of these studies mean that ulltrasonographic evidence of pancreatitis in the absence of other supportive evidence, should be interpreted with caution. ³ , ²⁶ The results of this study, in conjunction with 2 prior studies, clearly demonstrates that the timeline for detection of abnormal lipase and ultrasonographic findings differ both in onset of abnormalities and their resolution. ¹⁵ , ¹⁶ This helps to explain discrepancies seen between biochemical and sonographic findings in prior studies, but further adds to the complexity of a diagnosis of pancreatitis. ¹³ , ¹⁴ , ¹⁵ Given that median cPLI concentrations decreased to just above the upper end of the RI during early hospitalization (by day 2), it would be imperative to maintain strict sampling timelines in future studies into pancreatitis in dogs, as even a small change in timeline could significantly affect the results of the study. This is inherently more challenging in retrospective studies and strengthens the need for well‐designed prospective studies in this area. The results of preexisting retrospective studies should be interpreted with caution.

For dogs with baseline cPLI concentrations <400 μg/L there were no statistically significant changes in measured variables over time; however, this is likely influenced by the small group size, limited spread of data, and the use of adjusted P‐values to allow for consideration of multiple comparisons. When median values are compared graphically (Figure 2), there is a low rate of change in cPLI concentration over time, whereas the change in CRP, MCAI, UPASS, and owner assessed OH followed similar patterns to the group with initial cPLI ≥400 μg/L (Figure 2). These dogs, which might have acute on chronic disease, experience clinical signs of pancreatitis, despite minimal changes in cPLI concentrations. It is important to note that this cohort of dogs did not have pancreatic histopathology performed and it is not possible to definitively rule out other causes of the reported clinical signs. However, this is considered less likely given that no other major diseases were identified in the dogs history or in the 28‐day prospective follow up period. The lack of histopathology is consistent with common clinical practice, indeed given the invasive nature of pancreatic biopsies and the risk for morbidity, pancreatic biopsies have been described by some as ethically unjustifiable unless a laparotomy is being performed for other reasons. ²⁴ A recent retrospective study using a catalytic lipase assay and a more abbreviated follow up period, also noted that pancreatic lipase activity did not significantly change over time in dogs with suspected CP, and that the lipase activity did not differ between clinically severe and clinically mild disease. ²⁷ The combined interpretation of these studies suggests that pancreatic lipase concentration or activity alone cannot be used alone to exclude pancreatitis as a differential diagnosis, particularly when acute on chronic disease or CP is possible. ²⁷ This recommendation is consistent with prior studies that have documented a reduced diagnostic sensitivity of cPLI for the diagnosis of CP. ¹¹ Chronic or acute on CP is challenging to diagnose and an array of diagnostic tools including lipase concentrations, imaging, and potentially histopathology are required to obtain a diagnosis. ¹¹ , ²⁴ Additional studies such as the 1 reported here are needed to better define this group of dogs.

Our study had limitations predominantly related to the small sample size (risk of being underpowered), a lack of histopathologic evidence of pancreatitis, and the inability to definitively rule out other causes of the clinical signs in dogs with an initial cPLI <400 μg/L. Five of the original 24 dogs did not complete the study, which reduced our sample size, and 3 dogs were excluded because of cPLI concentrations <200 μg/L. It would have been ideal to have additional sampling time points between days 7 and 28 to further elucidate the timeline of clinical improvement in ultrasonographic parameters; however, this might have further reduced recruitment. The OH surveys utilized were not independently validated which is a limitation. However, the authors believe that owner perception of their pets health are important goals in the management of pancreatitis. Recent retrospective studies have also emphasized the importance of duration of clinical signs before presentation when evaluating pancreatic lipase concentrations in dogs. ¹⁶ , ²⁶ Because of temporal factors (study initiation 2021) this was not specifically addressed in our study questionnaire and data on this area is limited in our study. The prospective use of a longer follow up period and the use of an immunologic lipase assay in our study make the results of our study and the earlier studies complementary and further our understanding of biomarker and imaging findings over time in dogs with pancreatitis. ¹⁶ , ²⁶ Given this future study designs should be cognizant of both prehospitalization and in‐hospital timelines when taking serum samples for biomarker quantification. No dogs died during the study which might mean that more severe presentations of pancreatitis were not captured in this study group. A lack of severe cases could impact the results of the study.

CONFLICT OF INTEREST DECLARATION

Authors declare no conflict 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

Approved by Michigan State University IACUC, PROTO202000315.

HUMAN ETHICS APPROVAL DECLARATION

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

Supporting information

Supplementary Material 1. Supporting information.

Supplementary Material 2. Supporting information.

Mitchell L, Wang S, Lawver J, Cridge H. Serial monitoring of pancreatic lipase immunoreactivity, C‐reactive protein, abdominal ultrasonography, and clinical severity in dogs with suspected pancreatitis. J Vet Intern Med. 2024;38(2):987‐994. doi: 10.1111/jvim.17020