Serum Lipase Elevations in COVID-19 Patients Reflect Critical Illness and not Acute Pancreatitis.

Dear Editor:

We read with interest the article “ACE2 Expression in Pancreas May Cause Pancreatic Damage After SARS-CoV-2 Infection” by Liu et al. The authors noted abnormalities of serum amylase and lipase values among nearly 20% of hospitalized patients with COVID-19 and concluded that these enzyme abnormalities represent pancreatic injury and might even lead to chronic pancreatitis. However, many conditions may cause elevations in serum lipase, and these should not be attributed to pancreatic injury in the absence of acute pancreatitis (AP). We sought to establish the prevalence of abnormal serum lipase values and the prevalence of AP among a large cohort of patients hospitalized with COVID-19.

Consecutive patients hospitalized with COVID-19 at 36 North American medical centers during the earliest phase of the pandemic were identified and their charts were reviewed for clinical characteristics and outcomes, as previously described. AP was diagnosed according to the Revised Atlanta Criteria (2012), requiring 2 out of 3 of the following: characteristic abdominal pain, imaging abnormalities, or serum pancreas enzyme level greater than 3 times the upper limit of normal (ULN). For the outcome of elevated lipase, an extensive (n = 112 variables) univariable analysis was performed that included patient characteristics, comorbidities, medications, COVID-19 severity parameters and treatment regimens, and gastrointestinal symptoms (Table 1 ). Subsequently, a multivariable logistic regression model was fit for elevated lipase using stepwise selection among all variables with a significant difference between subjects with and without an elevated lipase at a level of P = .2.

Table 1

Univariable Associations with Elevated Lipase

	Normal lipase (n = 300)		Elevated lipase (n = 100)		P valuea
Patient characteristics
Age, y	60.27	15.48	59.43	14.90	.637b
Body mass index at presentation, kg/m2	30.29	7.76	31.67	8.55	.171b,c
Male sex	190	63.33	67	67.00	.508
Race					.774d
American Indian or Alaska Native	2	0.67	1	1.00
Asian	29	9.67	8	8.00
Black or African American	97	32.33	40	40.00
Native Hawaiian or other Pacific Islander	111	37.00	32	32.00
White	4	1.33	1	1.00
Multiple	57	19.00	18	18.00
Unknown
Hispanic ethnicity	49	18.22	18	19.35	.807
Cigarette smoking status					.112
Current smoker	12	4.26	9	9.57
Ex-smoker	83	29.43	30	31.91
Nonsmoker	187	66.31	55	58.51
Vaping status					.043d
Current vaping	1	0.54	3	4.55
Does not vape	183	98.92	62	93.94
Prior vaping	1	0.54	1	1.52
Alcoholism					.189
Current	19	7.14	11	12.94
No	233	87.59	68	80.00
Prior	14	5.26	6	7.06
Cannabis use					.530d
Current user	9	3.88	2	2.82
No	216	93.10	65	91.55
Prior user	7	3.02	4	5.63
Illicit drug use					1.000d
Current user	5	1.98	1	1.20
No	237	93.68	78	93.98
Prior user	11	4.35	4	4.82
Comorbidities
Hypertension	190	63.33	64	64.00	.905
Coronary artery disease/prior or current myocardial infarction	46	15.33	17	17.00	.692
Congestive heart failure	32	10.67	12	12.00	.712
Pulmonary hypertension	1	0.33	2	2.00	.156d
Chronic pulmonary obstructive disease	31	10.33	9	9.00	.700
Asthma	37	12.33	11	11.00	.722
Obstructive sleep apnea	37	12.33	10	10.00	.530
Interstitial lung disease/pulmonary fibrosis	4	1.33	0	0.00	.576d
Peripheral vascular disease	17	5.67	5	5.00	.800
Prior or current cerebrovascular accident or transient ischemic attack	25	8.33	7	7.00	.670
Dementia	10	3.33	1	1.00	.305d
Collagen vascular/rheumatologic disease	18	6.00	7	7.00	.721
Chronic liver disease	13	4.33	6	6.00	.587d
Cirrhosis	5	1.67	2	2.00	1.000d
MELD score before COVID-19 illness	14.00	8.89	20.00	0.00
Diabetes mellitus, uncomplicated	93	31.00	30	30.00	.851d
Diabetes mellitus with end-organ damage	28	9.33	9	9.00	.921
Moderate to severe kidney disease (creatinine >3 mg/dL before admission, end-stage renal disease, dialysis)	35	11.67	14	14.00	.538
Active/current malignancy, excluding nonmelanoma skin cancer	13	4.33	5	5.00	.783d
Prior malignancy	26	8.67	8	8.00	.836
Human immunodeficiency virus/AIDS	5	1.67	0	0.00	.338d
Solid organ transplant recipient	14	4.67	0	0.00	.026d
Bone marrow transplant recipient	4	1.33	1	1.00	1.000d
Irritable bowel syndrome	5	1.67	1	1.00	1.000d
Chronic diarrhea	2	0.67	2	2.00	.261d
Chronic constipation	8	2.67	1	1.00	.461d
Celiac disease	0	0.00	0	0.00	n/a
Prior biliary disease, including cholelithiasis, cholecystitis, choledocholithiasis, or cholangitis	7	2.33	1	1.00	.685d
Prior pancreatitis	4	1.33	2	2.00	.643d
Etiology					.400d
Acute pancreatitis	3	100.0	1	50.00
Recurrent acute pancreatitis	0	0.00	1	50.00
Chronic pancreatitis	0	0.00	0	0.00
Inflammatory bowel disease	2	0.67	1	1.00	1.000d
Other, not fitting into an above category	143	47.67	49	49.00	.817
Medications (within 1 mo of admission or current)
Angiotensin-converting enzyme inhibitor use	52	17.87	13	13.68	.344
Angiotensin receptor blocker use	46	15.75	12	12.50	.438
Nonsteroidal anti-inflammatory drug use	86	39.09	29	36.25	.655
Antibiotic use	85	30.14	27	29.03	.839
Proton pump inhibitor use	82	29.18	17	18.68	.049
H2 blocker use	39	13.88	8	8.70	.194
COVID-19 parameters
Highest level of care					.002
Inpatient ward	155	51.67	34	34.00
ICU	145	48.33	66	66.00
Duration of symptoms before first seeking medical attention, d	5.10	4.26	5.11	4.47	.924b,c
Duration of symptoms before hospitalization, d	6.68	4.85	6.29	4.90	.709b,c
Duration of hospitalization, d	14.84	13.60	19.86	15.73	.082b,c
If admitted to ICU, duration of ICU stay, d	14.23	12.29	19.21	13.06	.004b,c
Required mechanical ventilation	112	37.33	61	61.00	< .001
Required extracorporeal membrane oxygenation	6	2.00	4	4.00	.276d
Required vasopressor support	97	32.33	53	53.00	< .001
Final discharge disposition					.038
Deceased	57	19.00	28	28.00
Discharged to rehabilitation or nursing facility	48	16.00	21	21.00
Recovered (or almost recovered)	189	63.00	47	47.00
Still hospitalized at end of study	6	2.00	4	4.00
COVID-19-specific treatments
Remdesivir	16	5.33	9	9.00	.190
Hydroxychloroquine	168	56.00	66	66.00	.079
Chloroquine	2	0.67	0	0.00	1.000d
Azithromycin	178	59.33	64	64.00	.408
Glucocorticoids	41	13.67	15	15.00	.739
Interferon-α	0	0.00	0	0.00	n/a
Intravenous immunoglobulin	0	0.00	1	1.00	.250d
Lopinavir/ritonavir	7	2.33	3	3.00	.716d
Oseltamivir	16	5.33	5	5.00	.897
Tocilizumab	20	6.67	16	16.00	.005
Convalescent plasma	4	1.33	8	8.00	.002d
Anakinra	0	0.00	1	1.00	.250d
Ribavirin	1	0.33	0	0.00	1.000d
Sarilumab	3	1.00	2	2.00	.603d
Other	47	15.67	19	19.00	.437
None	59	19.67	13	13.00	.133
Gastrointestinal symptoms or signs
Anorexia	84	28.00	22	22.00	.239
Nausea	122	40.67	35	35.00	.315
Vomiting	79	26.33	24	24.00	.644
Abdominal pain (including cramps)	73	24.33	29	29.00	.354
Type
Diffuse	22	7.33	6	6.00	.651
Periumbilical	2	0.67	1	1.00	1.000d
Epigastric	10	3.33	8	8.00	.089d
Right upper quadrant	12	4.00	2	2.00	.532d
Left upper quadrant	6	2.00	2	2.00	1.000d
Right lower quadrant	5	1.67	1	1.00	1.000d
Left lower quadrant	9	3.00	2	2.00	.738d
Not specified	19	6.33	10	10.00	.221
Diarrhea	113	37.67	42	42.00	.441
Bloody diarrhea	1	0.33	0	0.00	1.000d
Hematemesis	2	0.67	3	3.00	.102d
Melena	4	1.33	3	3.00	.373d
Hematochezia (including bright red blood per rectum)	5	1.67	5	5.00	.130d
Dysphagia	6	2.00	2	2.00	1.000d
Odynophagia	1	0.33	0	0.00	1.000d
Constipation	21	7.00	3	3.00	.145
Hiccups	1	0.33	1	1.00	.438d
Jaundice	0	0.00	1	1.00	.250d
Other	11	3.67	3	3.00	1.000d
None	68	22.67	30	30.00	.140
Timing of gastrointestinal symptoms (if any) relative to respiratory/systemic symptoms (if any)					.813
Cannot tell from medical records review	24	10.34	5	7.14
GI symptoms came on concurrently with other symptoms	77	33.19	26	37.14
GI symptoms followed other symptoms	87	37.50	25	35.71
GI symptoms preceded other symptoms	42	18.10	14	20.00
GI symptoms were the only manifestation	2	0.86	0	0.00
Duration of gastrointestinal symptoms					.824
Cannot tell from medical records review	36	15.52	12	17.14
GI symptoms were present for a short portion (<25% of the duration) of the entire COVID-19 illness	92	39.66	24	34.29
GI symptoms were present for a significant portion (25%–75% of the duration) of the entire COVID-19 illness	69	29.74	21	30.00
GI symptoms were present for the entire/almost entire COVID-19 illness	35	15.09	13	18.57
Did gastrointestinal symptoms remain after resolution of other COVID-19 symptoms?					1.000d
No	132	89.80	43	91.49
Yes	15	10.20	4	8.51
If yes, how long?, d	9.67	9.07	—	—	n/a
Prominence of gastrointestinal symptoms					.156
Cannot tell from medical records review	12	5.17	1	1.43
GI symptoms were equally prominent to other symptoms related to COVID-19	48	20.69	16	22.86
GI symptoms were less prominent than other symptoms related to COVID-19	153	65.95	42	60.00
GI symptoms were less prominent than other symptoms related to COVID-19	19	8.19	11	15.71
Gastrointestinal diagnoses established shortly before, during, or shortly after COVID-19 illness
Esophagitis/esophageal ulcers	4	1.33	1	1.00	1.000d
Gastritis	3	1.00	3	3.00	.168d
Peptic ulcer disease	0	0.00	0	0.00	n/a
New enteritis	0	0.00	2	2.00	.062d
New colitis	2	0.67	3	3.00	.102d
Biliary disease, including cholelithiasis, cholecystitis, choledocholithiasis, or cholangitis	3	1.00	6	6.00	.009d
Hepatitis	10	3.33	4	4.00	.756d
Pancreatitis	0	0.00	6	6.00	< .001d
Other	26	8.67	11	11.00	.486d
None	258	86.00	75	75.00	.011

GI, gastrointestinal; ICU, intensive care unit; MELD, Model for End-Stage Liver Disease; n/a, not available.

Chi-square test unless otherwise noted.

Student t test.

Variable log transformed for analysis.

Fisher exact test.

Among 1992 subjects with polymerase chain reaction–confirmed COVID-19, serum lipase values were measured on the day of admission in 316 (15.9%). Although 58 subjects had an elevated value, only 9 (2.8%) were 3 times the ULN. Only 2 patients had both lipase >3 times the ULN and abdominal pain. Neither subject underwent cross-sectional imaging during the hospitalization, but an abdominal ultrasound in both cases demonstrated cholecystitis providing an alternative and more likely cause for AP than COVID-19. Serum lipase values were recorded during hospitalization in 400 (20.1%) subjects. Although in 100 subjects the levels were elevated, only 26 patients (6.5%) had a value greater than 3 times the ULN. Among these, 6 subjects met 2 of 3 Revised Atlanta Criteria and did not have an obvious alternative explanation for AP other than COVID-19 infection.

Within our cohort of patients hospitalized with COVID-19, there was no association between preexisting diabetes and elevated lipase values during hospitalization (P = .81). Furthermore, patients with diabetes with increased lipase values experienced similar mortality rates to patients with diabetes with normal lipase values (38.5% vs 24.0%; P = .08). There was no significant association between elevated lipase value and the use of angiotensin-converting enzyme inhibitors (13.7% vs 17.9%; P = .344) or angiotensin receptor blockers (12.5% vs 15.8%; P = .44) before admission (Table 1). Multivariable analysis demonstrated that elevated lipase values were associated with mechanical ventilation (n = 173; odds ratio, 2.55; 95% confidence interval, 1.6–4.08) and biliary disease (n = 9; odds ratio, 5.49; 95% confidence interval, 1.3–23.15).

Our findings demonstrate that AP occurs rarely in hospitalized subjects with COVID-19, despite the presence of angiotensin converting enzyme-2 receptors in both exocrine and endocrine cells. Mild elevations in pancreas enzymes were seen in approximately 20% of patients hospitalized with COVID-19; these were associated with severity of illness and did not indicate AP.

Since the start of the COVID-19 pandemic, there have been several reports on the development of AP among subjects with COVID-19 and its relation to AP severity. However, a systematic review determined that the work-up to accurately determine the cause of pancreatitis was suboptimal. Subsequent studies have posited that impaired circulation in the pancreas might explain the elevated serum lipase values among critically ill patients with COVID-19. Among critically ill patients with acute respiratory distress syndrome before the pandemic, elevations in lipase were common, but were not specific to the pancreas. Data from our cohort also suggest that elevated lipase values are associated with critical illness, rather than being directly related to pancreatic injury qualifying as AP.

In conclusion, elevations of pancreatic enzymes are relatively common (20%) in COVID-19 hospitalized patients, but rarely signify a clinical diagnosis of AP (1.5%). When AP is diagnosed, further investigation into the cause is warranted. As in other examples of critical illness, elevated serum lipase in COVID-19 patients likely reflects the overall disease severity.