Heart Drugs, Not the Heart Disease Itself, May Drive Pancreatic Enzyme Spikes in Dogs With Mitral Valve Disease

For decades, veterinarians have noticed something puzzling in dogs with myxomatous mitral valve disease, the most common acquired heart condition in small-breed dogs. As the disease progresses and the heart fails, blood tests often show elevated pancreatic lipase activity, an enzyme marker normally associated with pancreatitis. The prevailing explanation has been intuitive: a failing heart pumps less efficiently, the pancreas—richly supplied with blood for its exocrine and endocrine functions—becomes under-perfused, and ischemic injury to the pancreatic acinar cells follows. But a new retrospective cross-sectional study from Chungbuk National University Veterinary Teaching Hospital in South Korea, published in the Journal of Veterinary Internal Medicine, turns that narrative on its head. After carefully separating dogs that received cardiac medications from those that did not, the researchers found that the disease itself showed no association with pancreatic lipase activity. The drugs, it seems, may be the real culprit.

The study cohort comprised 82 client-owned dogs, all free of clinical signs of pancreatitis and gastrointestinal disease. Sixty-seven dogs had myxomatous mitral valve disease, classified according to the American College of Veterinary Internal Medicine consensus statement: 24 in stage B1, 22 in stage B2, 9 in stage C without cardiac medication, and 12 in stage C with cardiac medication. Fifteen healthy dogs served as controls. The breed distribution reflected the typical small-breed predisposition of the disease, with Maltese, Pomeranians, Toy Poodles, and Spitz dominating the MMVD group. Serum pancreatic lipase activity was measured using a validated colorimetric assay based on the DGGR substrate, a point-of-care test widely used in veterinary practice.

The central finding emerged from a comparison that previous studies had never made. Among dogs not receiving cardiac medications, pancreatic lipase activity did not differ significantly across disease stages (Kruskal–Wallis H = 0.488; df = 3; P = .921). This was a striking result, because it meant that disease severity—whether the dog had early-stage valve changes or full-blown congestive heart failure—had no measurable effect on the pancreatic enzyme marker. The result directly contradicts earlier reports that had linked pancreatic lipase immunoreactivity to congestive heart failure and to echocardiographic indices of cardiac remodeling, including left atrial-to-aortic root ratio and normalized left ventricular internal diameter in diastole.

When medicated stage C dogs entered the analysis, the picture changed dramatically. These dogs had significantly higher pancreatic lipase activity than every other group. The Hodges–Lehmann median differences were substantial: 67.5 U/L above healthy controls (95% CI, 19.0–184.0), 57.0 U/L above stage B1 (95% CI, 12.0–145.0), 65.8 U/L above stage B2 (95% CI, 18.0–156.0), and 58.0 U/L above non-medicated stage C dogs (95% CI, 9.0–268.0). Effect sizes, measured with Cliff’s δ, ranged from 0.56 to 0.67—large by conventional standards. More clinically alarming, 25% of medicated stage C dogs (3 of 12) exceeded 400 U/L, a diagnostic threshold suggestive of pancreatitis, whereas no non-medicated stage C dog reached this level. Notably, these three dogs showed no signs of gastrointestinal disease during long-term therapy, and prior abdominal imaging in two cases confirmed no ultrasonographic abnormalities of the pancreas.

The correlation analyses added a temporal dimension to the finding. Among the 12 medicated stage C dogs, pancreatic lipase activity was positively correlated with the duration of furosemide administration (Spearman ρ = 0.67; 95% CI, 0.10–0.91; P = .028) and with the duration of angiotensin-converting enzyme inhibitor administration (ρ = 0.62; 95% CI, 0.01–0.89; P = .048). Pimobendan, the positive inotrope that is standard therapy for advanced MMVD, showed no significant correlation (ρ = 0.50; P = .104). The longer the dogs had been on the diuretic and the ACE inhibitor, the higher their pancreatic enzyme activity climbed—a dose–duration relationship that points toward a pharmacological mechanism rather than a hemodynamic one.

The multivariable linear regression analysis sealed the argument. After adjusting for age, ACVIM stage, and symmetric dimethylarginine concentration—a renal marker that could theoretically confound lipase levels through reduced clearance—medication status was the only independent predictor of pancreatic lipase activity (P = .004). On the log10-transformed scale, the regression coefficient for medication was 0.414, corresponding to an approximately 2.59-fold higher pancreatic lipase activity in medicated dogs compared with non-medicated dogs. The overall model was statistically significant (adjusted R² = 0.14; P = .01), and neither age nor SDMA reached independent significance. Renal dysfunction, historically invoked as an explanation for hyperlipasemia in dogs with cardiac disease, was excluded by design: dogs with serum creatinine above 1.4 mg/dL or SDMA above 18 μg/dL were excluded according to International Renal Interest Society guidelines.

The pharmacological mechanisms behind the furosemide association are plausible and well documented. High-ceiling diuretics reduce pancreatic blood flow by causing extracellular volume loss, an effect demonstrated in experimental pharmacology studies. Long-term furosemide treatment in rats has been shown to alter the electrolyte composition of pancreatic acinar cells and to cause accumulation of secretory material within the pancreatic ductules, leading to direct cellular toxicity and obstruction of the pancreatic tissue. Furosemide has previously been identified as a potential risk factor for pancreatitis in dogs in a case-control analysis of 101 cases. Angiotensin-converting enzyme inhibitors, meanwhile, have been linked to pancreatitis in humans, possibly through local angioedema or direct pancreatic toxicity, and a population-based case-control study found an increased risk of acute pancreatitis among ACE inhibitor users.

Pimobendan’s apparent neutrality is equally informative. An experimental study using a canine tachycardia-induced heart failure model showed that pimobendan preserved pancreatic blood flow and prevented acinar cell injury by maintaining cardiac output. This finding is highly relevant clinically, because pimobendan is the standard treatment for most dogs with advanced MMVD. The drug may, in effect, offset the hypoperfusion that heart failure would otherwise impose on the pancreas—suggesting that the standard cardiac regimen is not uniformly harmful, but that specific components of it carry pancreatic risk.

The implications for clinical interpretation are immediate. Elevated pancreatic lipase activity in a dog with MMVD has traditionally been read as evidence of pancreatic injury driven by the failing heart. The new data suggest that this interpretation should be reevaluated. The elevations reported in earlier studies might have been confounded by concurrent cardiac medication use—particularly furosemide and ACE inhibitors—which are the standard treatment regimens for dogs with congestive heart failure. The correlation between serum lipase activity and echocardiographic indices of remodeling reported in previous work might represent confounding by long-term medication administration rather than a direct effect of cardiac remodeling. Veterinarians interpreting pancreatic lipase results in dogs with MMVD should now consider medication history alongside disease stage.

The study has limitations that warrant cautious interpretation. As a retrospective cross-sectional analysis, it can describe associations but cannot establish causal relationships. The sample size in certain subgroups was limited, and the correlation analyses evaluating medication duration should be considered exploratory, potentially subject to type I errors due to the lack of adjustment for multiple comparisons. The statistical significance in the medicated stage C group was largely driven by three dogs with markedly elevated values, and non-medicated stage C dogs generally presented with their first episode of congestive heart failure, whereas medicated dogs were under long-term management—so disease chronicity could have contributed to the difference. Dose–response relationships could not be evaluated retrospectively, and concurrent abdominal ultrasonography was not performed for the entire cohort. Stored serum samples were used, but DGGR lipase activity is known to remain stable at −80°C for up to 12 months, making storage effects unlikely. Prospective studies with larger populations, longitudinal follow-up, and standardized pancreatic imaging are needed to confirm the relationship. Withholding essential cardiac medications from dogs with heart failure would be ethically inappropriate, so the retrospective design, for all its constraints, offers insights that a randomized trial never could.

Subject of Research: veterinary cardiology

Article Title: Evaluation of serum pancreatic lipase activity in dogs with myxomatous mitral valve disease

Article References: Kim, C., Chae, Y., Yun, T., Kang, B.-T., & Kim, H. (2026). Evaluation of serum pancreatic lipase activity in dogs with myxomatous mitral valve disease. Journal of Veterinary Internal Medicine, 40(5), Article aalag218. https://doi.org/10.1093/jvimsj/aalag218

Image Credits: AI Generated

DOI: 10.1093/jvimsj/aalag218

Keywords: myxomatous mitral valve disease, pancreatic lipase, furosemide, ACE inhibitors, pimobendan, canine pancreatitis, veterinary medicine, cardiac medications, Evaluation, serum, pancreatic, lipase