Hyperlipasemia in COVID-19: Statistical Significance vs Clinical Relevance

Abstract

Less than a third of the patients admitted to hospital were tested for lipase levels, and neither the timing nor the indication for testing lipase levels has been elucidated. (1) Single center, USA 83 14 Hyperlipasemia associated with poor prognosis; none had clinically significant pancreatic involvement The division of patients in tertiles of lipase was unusual and requires explanation. The highest tertile (81–701 IU/L) includes a wide range and likely represents a heterogeneous population.

In the July 2020 issue of the journal, Barlass et al. (1) bring out concerns about poor outcomes in patients with pancreatic involvement in COVID-19. Although a predominantly respiratory pathogen, the novel coronavirus often leads to a multisystem disease. The involvement of the gastrointestinal lumen and liver is not uncommon. However, clinically significant involvement of the pancreas has been extremely rare. This single-center retrospective study screened 1,003 patients with COVID-19, of whom 294 were admitted to the hospital, and 83 underwent a test for lipase levels. Among these, 17% patients had significant elevation of lipase. This was associated with higher frequency of gastrointestinal involvement, intensive care unit (ICU) admission, and mechanical ventilation. Based on these findings, the authors concluded that elevated lipase was associated with poor outcomes.

Certain caveats in this retrospective chart review need to be addressed. Less than a third of the patients admitted to hospital were tested for lipase levels, and neither the timing nor the indication for testing lipase levels has been elucidated. It is possible that lipase testing was a part of a more comprehensive assessment of patients who were sicker, admitted to ICU, or had multisystem involvement to begin with. This brings about a huge selection bias in the study population. It would therefore be interesting to compare the clinical profile of patients who did not undergo lipase testing with the study population. Interestingly, 2 other studies addressing pancreatic involvement in COVID-19, one from China and another from the United States, had a much lower prevalence of significant hyperamylasemia (2,3) (Table 1).

Table 1.

Summary of studies: pancreatic involvement in COVID-19

[Figure omitted. See PDF]

Study	Setting	N	Lipase >3x	Outcome
Wang et al. (2)	Single center, China	62	0	Potential mild pancreatic injury in COVID; none had clinically significant pancreatic involvement
McNabb-Baltar et al. (3)	6 centers, USA	71	2	No impact of hyperlipasemia on clinical parameters; none had clinically significant pancreatic involvement
Barlass et al. (1)	Single center, USA	83	14	Hyperlipasemia associated with poor prognosis; none had clinically significant pancreatic involvement

The division of patients in tertiles of lipase was unusual and requires explanation. The highest tertile (81–701 IU/L) includes a wide range and likely represents a heterogeneous population. This is further supported by the wide confidence intervals of the odds ratios for intubation and ICU admissions.

Retrospective chart reviews often help bring out aspects of pathophysiology of new diseases, as is the case with COVID-19. However, they are often prone to strong biases, which need to be addressed systematically. Significant pancreatic involvement in COVID is rare as the current evidence stands. We therefore feel that suggesting lipase as be a marker of severe disease may not be supported by the current evidence. Such suggestions may provoke unnecessary testing, without clear implications of diagnosis or prognosis.

CONFLICTS OF INTEREST

Guarantor of the article: Rohit Gupta, MD, DM.

Specific author contributions: S.R.: drafting manuscript. A.S. and I.P.: critical editing. R.G.: concept, critical edits, and senior author.

Financial support: None to report.

Potential competing interests: None to report.

Author Notes

Correspondence: Rohit Gupta, MD, DM. E-mail: [email protected].