COVID-19 Digestive System Involvement and Clinical Outcomes in a Large Academic Hospital in Milan, Italy.

Since February 2020, the COVID-19 pandemic has spread to Italy affecting more than 100,000 people. Several studies have reported a high prevalence of gastrointestinal (GI) symptoms, and investigated their potential association with clinical outcomes.1 The timing, clinical significance, and possible impact on viral spread of GI symptoms presentation have not been fully elucidated. Elevation of liver function tests and other laboratory values has also been reported; however, their prognostic significance has not been clearly established.2.

Since February 2020, the COVID-19 pandemic has spread to Italy affecting more than 100,000 people. Several studies have reported a high prevalence of gastrointestinal (GI) symptoms, and investigated their potential association with clinical outcomes. The timing, clinical significance, and possible impact on viral spread of GI symptoms presentation have not been fully elucidated. Elevation of liver function tests and other laboratory values has also been reported; however, their prognostic significance has not been clearly established.

We analyzed a cohort of reverse-transcriptase polymerase chain reaction–confirmed COVID-19 patients consecutively admitted to Humanitas Hospital, Milan, Italy, to describe the prevalence of GI symptoms and GI/liver tests abnormalities, and their association with clinical outcomes.

Methods

Clinical and laboratory data were extracted from electronic medical records. Presence of vomit and diarrhea (defined as passing of 3 or more loose stools per day) as reported at admission or during the week preceding admission were recorded as per electronic medical records. To explore the associations between GI clinical and laboratory parameters with clinical deterioration we used survival model for censored observations. The composite study end point was clinical deterioration defined as intensive care unit (ICU) transfer or death within 20 days of hospital admission. Time to event was defined as the time from hospital admission until the date of event or censoring. We used log-rank tests and Cox regression analysis. Missing data were not imputed. We presented hazard ratios with 95% confidence intervals (CI).

Results

From February 22 to March 30, 2020, 325 reverse-transcriptase polymerase chain reaction–confirmed COVID-19 patients had been admitted to the Humanitas Research Hospital. The analysis was restricted to 292 patients, after excluding those who were transferred to the ICU, or died, within the first day. Patients were predominantly males (68.2%) with a mean age of 65.0 ± 14.1 years. Diarrhea (27.1%) was the most frequent GI symptom. Patients’ characteristics are summarized in Table 1 .

Table 1

Association Between Clinical and Laboratory Gastrointestinal Parameters of Hospitalized COVID-19 Patients (n = 292) and Clinical Deterioration Leading to ICU Transfer or Death Adjusted by Age and Gender

	Mean ± SD or n (%)	Cox proportional hazards analysis
		Adjusteda HR	95% CI	P value
Demographic characteristics
Age, y	65.0 ± 14.1	—	—	—
≥65	161/292 (55.1)
<65	131/292 (44.9)
Gender		—	—	—
Female	93/292 (31.8)
Male	199/292 (68.2)
Gastrointestinal symptoms
Diarrhea	69/255 (27.1)	0.79	0.42‒1.46	.45
Vomit	11/274 (4.0)	—	—	—
Any gastrointestinal symptom (ie, diarrhea or vomit)	69/245 (28.2)	0.47	0.23‒0.97	.041
Blood biochemistry
Alanine aminotransferase ≥50 U/L	54/292 (18.5)	1.11	0.60‒2.04	.74
Aspartate aminotransferase ≥50 U/L	78/292 (26.7)	1.30	0.81‒2.08	.28
γ-Glutamyl transpeptidase ≥55 U/L	102/282 (36.2)	1.45	0.91‒2.30	.12
Alkaline phosphatase ≥150 U/L	27/280 (9.6)	1.62	0.87‒3.00	.13
Total bilirubin ≥1.2 mg/dL	31/292 (10.6)	1.39	0.76‒2.56	.29
Direct bilirubin ≥0.3 mg/dL	70/283 (24.7)	1.52	0.94‒2.44	.084
Indirect bilirubin ≥1.1 mg/dL	16/269 (6.0)	1.28	0.51‒3.20	.60
Amylase ≥100 U/L	43/288 (14.9)	1.54	0.88‒2.72	.13
Lipase ≥68 U/L	28/249 (11.2)	2.02	1.08‒3.80	.028

NOTE. Boldface indicates statistically significant P values.

CI, confidence interval; HR, hazard ratio; ICU, intensive care unit; SD, standard deviation.

Adjusted for age and gender.

As of March 30, 129 patients (44.2%) had been discharged, and 107 (36.6%) were still hospitalized. Clinical deterioration occurred in 82 patients (28.1%), including 27 (9.2%) patients who were transferred to ICU, and 56 (19.2%) who died.

Among admission parameters, the presence of any GI symptom (ie, diarrhea or vomit), and alkaline phosphatase, total bilirubin, direct bilirubin, and lipase levels were significantly associated with ICU transfer or death in the univariable analyses (Supplementary Table 1). Of these, the occurrence of any GI symptom (adjusted hazard ratio [aHR], 0.47; 95% CI, 0.23‒0.97; P = .041), alkaline phosphatase levels (aHR, 1.14; 95% CI, 1.05‒1.23, per 100 U/L increase; P = .001) and high lipase levels (aHR, 2.02; 95% CI, 1.08‒3.80; P = .028) remained significant after adjustment for age and sex (Supplementary Table 1). Importantly, the presence of GI symptoms was inversely associated with the risk of clinical deterioration, whereas higher levels of alkaline phosphatase and lipase predicted a poor prognosis.

Supplementary Table 1

Association of Gastrointestinal Clinical and Laboratory Parameters With Clinical Deterioration Leading to ICU Transfer or Death in Hospitalized COVID-19 Patients

Gastrointestinal characteristics at admission	Log-rank test		Cox proportional hazards analysis
	Chi-square (d.f.)	P value	Crude HR	(95% CI)	P value	Adjusteda HR	(95% CI)	P value
Gastrointestinal symptoms
Diarrhea	2.88 (1)	.089	0.60	0.33‒1.10	.097	0.79	0.42‒1.46	.45
Vomit	3.62 (1)	.057	—	—	—	—	—	—
Any gastrointestinal symptom (ie, diarrhea or vomit)	8.53 (1)	.004	0.37	0.18‒0.75	.006	0.47	0.23‒0.97	.041
Blood biochemistry
Alanine aminotransferase (per 10 U/L increase)	—	—	1.01	0.95‒1.07	.87	1.03	0.98‒1.08	.19
Alanine aminotransferase (≥50 vs <50 U/L)	0.36 (1)	.55	0.84	0.46‒1.51	.56	1.11	0.60‒2.04	.74
Aspartate aminotransferase (per 10 U/L increase)	—	—	1.03	0.99‒1.08	.12	1.04	0.997‒1.09	.065
Aspartate aminotransferase (≥50 vs <50 U/L)	1.22 (1)	.27	1.30	0.81‒2.08	.28	1.30	0.81‒2.08	.28
γ-Glutamyl transpeptidase (per 100 U/L increase)	—	—	1.16	0.96‒1.41	.13	1.20	0.98‒1.46	.074
γ-Glutamyl transpeptidase (≥55 vs <55 U/L)	1.11 (1)	.29	1.27	0.81‒2.01	.30	1.45	0.91‒2.30	.12
Alkaline phosphatase (per 100 U/L increase)	—	—	1.14	1.06‒1.23	< .001	1.14	1.05‒1.23	.001
Alkaline phosphatase (≥150 vs <150 U/L)	3.90 (1)	.048	1.83	0.99‒3.39	.055	1.62	0.87‒3.00	.13
Total bilirubin (per 1 mg/dL increase)	—	—	1.11	1.00‒1.23	.048	1.12	0.998‒1.25	.054
Total bilirubin (≥1.2 vs <1.2 mg/dL)	3.26 (1)	.071	1.70	0.94‒3.08	.079	1.39	0.76‒2.56	.29
Direct bilirubin (per 1 mg/dL increase)	—	—	1.16	0.96‒1.40	.12	1.18	0.96‒1.44	.11
Direct bilirubin (≥0.3 vs <0.3 mg/dL)	9.04 (1)	.003	1.96	1.25‒3.09	.004	1.52	0.94‒2.44	.084
Indirect bilirubin (per 1 mg/dL increase)	—	—	1.04	0.63‒1.72	.87	1.08	0.65‒1.78	.76
Indirect bilirubin (≥1.1 vs <1.1 mg/dL)	0.02 (1)	.88	1.07	0.43‒2.65	.89	1.28	0.51‒3.20	.60
Amylase (per 10 U/L increase)	—	—	1.04	0.98‒1.10	.23	1.04	0.98‒1.11	.17
Amylase (≥100 vs <100 U/L)	2.52 (1)	.11	1.56	0.89‒2.74	.12	1.54	0.88‒2.72	.13
Lipase (per 10 U/L increase)	—	—	1.06	1.00‒1.13	.051	1.07	1.002‒1.15	.042
Lipase (≥68 vs <68 U/L)	4.89 (1)	.027	1.98	1.06‒3.71	.033	2.02	1.08‒3.80	.028

NOTE. Boldface indicates statistically significant P values.

CI, confidence interval; HR, hazard ratio; ICU, intensive care unit; SD, standard deviation.

Adjusted for age and gender.

Discussion

In a cohort of COVID-19 admitted patients, we found a high prevalence of GI symptoms. Interestingly, the presence of diarrhea or vomit was associated with a better prognosis, independently of patient age and sex. It is worth noting that we excluded patients admitted in critical conditions in whom a detailed medical history about GI symptoms was not adequately assessed. Our findings could be explained by a prevalent GI viral localization rather than respiratory. GI tropism of SARS-CoV2 has been demonstrated in a recent study that detected SARS-CoV2 more frequently in the stools of patients presenting with diarrhea. ,

We also found that elevated lipase and alkaline phosphatase levels were associated with poor prognosis; whether this reflects a greater systemic inflammatory response or an early sign of multiorgan failure needs to be ascertained. Because angiotensin-converting enzyme-2 receptors are highly expressed by pancreatic islets, a possible direct cytopathic injury seems likely. The action of Sars-CoV on the pancreas through these receptors can induce acute hyperglycemia and transient type-2-diabetes, possibly leading to further complications and poor prognosis. The correlation we observed with alkaline phosphatase is intriguing because angiotensin-converting enzyme-2 receptors are also abundantly expressed on endothelial liver cells, which makes liver a potential target for SARS-CoV. However, liver biopsies of patients with COVID-19 have not shown signs of biliary tract damage or cholestasis, thus we cannot exclude that high alkaline phosphatase reflects bone diseases and systemic frailty.

In conclusion, we observed that biochemical elevations of liver and GI tests and GI symptoms are common at presentation in hospitalized patients with COVID-19, with high lipase and alkaline phosphatase levels and the absence of vomit/diarrhea predicting poor clinical outcomes.

Supplementary Appendix

Humanitas Covid-19 Task Force

ACCORNERO	STEFANO
AGHEMO	ALESSIO
ALI	HUSSAM
ANGELINI	CLAUDIO
ARCARI	IVAN
AROSIO	PAOLA
AZZOLINI	ELENA
BACCARIN	ALESSANDRA
BADALAMENTI	SALVATORE
BAGGIO	SARA
BARBAGALLO	MICHELA
BARBERI	CATERINA
BARBIC	FRANCA
BARBIERI	VIVIANA
BARBONE	ALESSANDRO
BASCIU	ALESSIO
BOCCIOLONE	MONICA
BOREA	FEDERICA
BORRONI	MARIO
BRESCIANI	GIANLUIGI
BRUNETTA	ENRICO
BULLETTI	CINZIA
CADONATI	CRISTINA
CALABRO'	LORENZO
CALATRONI	MARTA
CALVETTA	ALBANIA ANTONIETTA
CANNATA	FRANCESCO
CANZIANI	LORENZO
CAPRETTI	GIOVANNI LUIGI
CARLANI	ELISA
CARRONE	FLAMINIA
CASANA	MADDALENA
CECCONI	MAURIZIO
CERIOTTI	CARLO
CICCARELLI	MICHELE
CIMINO	MATTEO
CIUFFINI	LEONARDO
COLAIZZI	CHIARA
COLAPIETRO	FRANCESCA
COSTA	GUIDO
COZZI	OTTAVIA
CRAVIOTTO	VINCENZO
CRESPI	CHIARA
CRIPPA	MASSIMO
DA RIO	LEONARDO
DAL FARRA	SARA
D'ANTONIO	FEDERICA
DE AMBROGGI	GUIDO
DE DONATO	MASSIMO
DE LUCIA	FRANCESCA
DE SANTIS	MARIA
DELLE ROSE	GIACOMO
DI PILLA	MARINA
DIPAOLA	FRANCA
DIPASQUALE	ANDREA
DIPASQUALE	ANGELO
DROANDI	GINEVRA
FAZIO	ROBERTA
FERRANTE	GIUSEPPE
FERRARA	ELISA CHIARA
FERRARI	MATTEO CARLO
FERRI	SEBASTIAN
FOLCI	MARCO
FORESTI	SARA
FRANCHI	ELOISA
FRAOLINI	ELIA
FUGAZZA	ALESSANDRO
FURFARO	FEDERICA
GALIMBERTI	PAOLA
GALTIERI	ALESSIA PIERA
GAVAZZI	FRANCESCA
GENERALI	ELENA
GOLETTI	BENEDETTA
GUIDELLI	GIACOMO
JACOBS	FLAVIA
KURIHARA	HAYATO
LAGIOIA	MICHELE
LIBRE'	LUCA
LLEO	ANA
LOIACONO	FERDINANDO
LUGHEZZANI	GIOVANNI
MACCALLINI	MARTA
MAIORINO	ALFONSO FRANCESCO
MALESCI	ALBERTO
MANTOVANI	RICCARDO
MARCHETTINI	DAVIDE
MARINELLO	ARIANNA
MARKOPOULOS	NIKOLAOS
MASETTI	CHIARA
MILANI	ANGELO
MIRANI	MARCO
MORELLI	PAOLA
MOTTA	FRANCESCA
MUNDULA	VALERIA
NIGRO	MATTIA
OMODEI	PAOLO
ORMAS	MONICA
PAGLIARO	ARIANNA
PALIOTTI	ROBERTA
PARIGI	TOMMASO LORENZO
PEDALE	ROSA
PEGORARO	FRANCESCO
PELLEGATTA	GAIA
PELLEGRINO	MARTA
PETRIELLO	GENNARO
PICCINI	SARA
POCATERRA	DARIA
POLIANI	LAURA
PREATONI	PAOLETTA
PROCOPIO	FABIO
PUGGIONI	FRANCESCA
PUGLIESE	LUCA
RACCA	FRANCESCA
RANDAZZO	MICHELE
REGAZZOLI LANCINI	DAMIANO
REGGIANI	FRANCESCO
RODOLFI	STEFANO
RUONGO	LIDIA
SACCO	CLARA
SANDRI	MARIA TERESA
SAVI	MARZIA
SCARFO'	ISIDE
SHIFFER	DANA
SICOLI	FEDERICO
SOLANO	SIMONE
SOLITANO	VIRGINIA
STAINER	ANNA
STELLA	MATTEO CARLO
STRANGIO	GIUSEPPE
TAORMINA	ANTONIO
TESTONI	LUCIA
TORDATO	FEDERICA
TRABUCCO	ANGELA
ULIAN	LUISA
VALENTINO	ROSSELLA
VALERIANO	CHIARA
VENA	WALTER
VERLINGIERI	SIMONA
VESPA	EDOARDO
VOZA	ANTONIO
ZANUSO	VALENTINA
ZILLI	ALESSANDRA