Does Prior Bariatric Surgery Predispose to Acetaminophen-Related Acute Liver Failure?

Abstract

INTRODUCTION:

Prior small studies have suggested that patients with prior metabolic and bariatric surgery (MBS) may have increased susceptibility to acetaminophen (APAP)-related acute liver failure/acute liver injury (ALF/ALI). The aim of this study was to compare the presentation, etiology, and outcome of adult ALI/ALF patients who were enrolled in a prospective registry study with prior bariatric surgery to those without.

METHODS:

Among 3,364 ALF/ALI patients in the Acute Liver Failure Study Group registry enrolled between January 1998 and August 2019, 85 (2.3%) reported prior MBS. On review, 6 cases were deemed not ALF and excluded from the analysis.

RESULTS:

Among the 79 MBS ALF/ALI patients, 95% were female and 86% were White. Etiology included APAP-related ALF/ALI in 78.5% of the MBS group vs 49.3% in the non-MBS group (P < 0.001). The proportion of patients with MBS increased from 1.6% (1998–2008) to 3.4% in the later years (2009–2019) (P < 0.001). There were proportionately more unintentional overdoses in the MBS cohort (71% vs 52%, P < 0.001) with smaller median total APAP dose ingestion (6,500 mg vs 12,000 mg, P = 0.009). Although MBS patients had more severe encephalopathy at presentation, overall, 21-day and transplant-free survival between the groups were similar.

DISCUSSION:

ALF/ALI is more frequent in MBS patients than in the general population. APAP toxicity is the most common cause in MBS patients but not the only diagnosis observed. MBS may predispose to severe unintentional APAP liver injury at lower doses. Both MBS patients and providers should be aware regarding this potential risk of increased susceptibility to APAP hepatotoxicity.

abstract-type="graphical">

INTRODUCTION

Bariatric surgery for morbid obesity is increasingly being performed, with over 289,000 Americans undergoing metabolic and bariatric surgery (MBS) in 2023 alone. Approximately 5.4 million individuals in the United States have had MBS at some point in the past (∼1.6% of the current US population) (1). The type of MBS procedure has changed over time with more patients currently undergoing laparoscopic sleeve gastrectomy compared with Roux-en-Y gastric bypass or adjustable gastric banding. MBS is one of the most effective treatments for obesity and leads to remission of the metabolic syndrome and improvements in metabolic dysfunction-associated steatotic liver disease (MASLD) in over 50% of people (2). However, rare cases of presumed acute liver failure (ALF) with massive steatosis have been reported in patients after MBS (3–5), especially after jejunoileal bypass (JIB) (6,7) or biliopancreatic diversion (7–9).

Acetaminophen (APAP) is the most common cause of ALF in adult Americans, accounting for 47% of ALF cases in the United States and approximately 16% of liver transplants for ALF in North America (10–12). A prior single center study demonstrated that MBS patients were over-represented among APAP ALF compared with other etiologies (13).

The Acute Liver Failure Study Group (ALFSG) has amassed detailed clinical and laboratory data on more than 3,300 ALF or acute liver injury (ALI) patients over the past 2 decades, providing an opportunity to review the experience with MBS and ALF/ALI in a North American population. The aims of this study were to (i) describe the presentation, etiology, and outcome of ALF/ALI patients with a history of MBS, to determine whether MBS may predispose to APAP and/or other forms of ALF including fatty liver disease; (ii) compare MBS patients with APAP-related ALF/ALI with the remaining non-MBS APAP ALF patients, and (iii) to determine risk factors of poor outcomes with ALF in MBS patients and whether cases of ALF in the setting of MBS are increasing over time.

METHODS

Patient characteristics and data collection

The ALFSG Registry was initiated in 1998 and has enrolled 3,364 adult patients among the 28 participating clinical centers (11). The database comprises patients admitted with a clinical diagnosis of ALF, meeting the following standard criteria: duration of symptoms of jaundice or illness less than 26 weeks before admission, altered mental status, and international normalized ratio (INR) ≥1.5. A separate category of severe ALI was also used after 2010, with the following criteria: an INR ≥2, absence of hepatic encephalopathy, and hepatic illness of ≤26 weeks' duration in patients without known chronic liver disease (14). Patients with previously recognized cirrhosis and superimposed acute-on-chronic-liver failure were excluded. Exceptions to the requirement of exclusion of cirrhosis were made for patients with acute presentation of Wilson disease with liver failure and those with first presentation of autoimmune-related ALF. Written informed consent was obtained from the patients' legal next of kin to collect information on medical history and management data during the first 7 days of enrollment, which was then documented on a standardized case report form. All sites complied with local IRB requirements and adhered to the Declarations of Helsinki and Istanbul.

Extensive demographic, laboratory, and clinical data, including imaging, blood, and urine samples, were collected for the first 7 days of admission or until transplantation, discharge, or death within 21 days of admission. APAP etiology was determined to be the primary cause of ALF/ALI if there was a history of potentially toxic APAP ingestion within a week of presentation, detection of any level of APAP in the serum, or alanine aminotransferase more than 1,000 U/L with any history of APAP ingestion, irrespective of APAP level. Other etiologies were diagnosed using standard algorithms (15). Etiologic diagnoses were made initially by the site investigator but were reviewed by a causality committee composed of senior hepatologists when more detailed information such as APAP protein adduct biomarkers was available. Six MBS patients were excluded from the study following committee review because of the subsequent finding of acute-on-chronic liver failure (ACLF) related to MASLD, not ALF—see further discussion in results.

Statistical methods

The results are expressed as percentages or mean ± SD unless otherwise stated. Comparison between groups of patients with spontaneous survival and adverse outcomes was performed using the 2-tailed Student t test and χ² test where appropriate. P < 0.05 was considered statistically significant. For survival analysis, time of entry was defined as the date of study enrollment. The survival time was calculated from the date of enrollment to the date of death or censoring (i.e., 3 weeks from enrollment in the study). The univariate Cox proportional hazards model was used to estimate the effect of each factor, including demographic variables, comorbidities, and presenting laboratory and clinical features. The variables with P < 0.10 on the univariate proportional hazards model were used in the Cox multivariate model to identify predictors of adverse outcome at 3 weeks. Backward conditional Cox regression modeling was used. All statistical analysis and graph formation was completed using SPSS version 15.0 (SPSS, Chicago, IL).

RESULTS

Between January 1998 and August 2019, 3,364 patients were enrolled in the ALFSG registry with ALF or ALI, of whom 79 (2.3%) were found to have a history of MBS. Among the 79, 65 presented with ALF, 14 with ALI and one of the 14 converted to ALF in hospital. The specific type of MBS was available in only 14 patients: 9 patients had undergone Roux-en-Y gastric bypass, 1 duodenal switch procedure, and 4 with sleeve gastrectomy. Seventy-five MBS patients (95%) were female, and 68 (86%) were White (Table 1). Coma grade on admission was frequently more advanced (grades 3 or 4), in 61% of the MBS patients compared with 47% of the remaining non-MBS ALF group (P < 0.001). Etiology of ALF in the overall group included 1,619 of 3,285 because of APAP (49.3%), as compared with 62 of 79 (78.5%) APAP in the MBS group, P < 0.001.

Table 1.

Demographic and clinical data for the bariatric surgery patients presenting with ALI or ALF

	N	Number (%) or median (range)
Age (yr)	79	38 (19–70)
Sex (female)	79	75 (94.9)
Race	79
White		68 (86.1)
African American		8 (10.1)
Other		3 (3.8)
Injury type	79
Enrolled as ALI		13 (16.5)
Enrolled as ALF		65 (82.3)
Converted from ALI to ALF		1 (1.3)
Primary etiology	79
Acetaminophen		62 (78.5)
Any ALI/ALF in pregnancy		1 (1.3)
Autoimmune hepatitis		1 (1.3)
DILI		5 (6.3)
Hepatitis B		2 (2.5)
Hepatitis C		1 (1.3)
Shock/ischemia		5 (6.3)
Indeterminate		2 (2.5)
BMI at study admission	70	30 (19–77)
Admission biochemistry
Hemoglobin (g/dL)	78	9.4 (6.8–13.6)
White blood count (×109/L)	78	8.3 (3.0–30.0)
Platelet count (×109/L)	77	150 (13–890)
INR	79	2.5 (1.1–12.0)
AST (IU/L)	78	2,446 (13–32,870)
ALT (IU/L)	77	1,949 (9–10,841)
Bilirubin (mg/dL)	79	4.2 (0.5–49.9)
Albumin (g/dL)	73	2.8 (1.4–3.8)
Ammonia (venous; μmol/L)	39	102 (25–690)
Creatinine (mg/dL)	79	1.2 (0.4–12.3)
Phosphate (mg/dL)	67	2.8 (0.6–13.2)
Usual intake of ETOH during the past 6 mo	70
None (less than 1 drink per week)		37 (52.9)
1 drink per week		1 (1.4)
2 drinks per week		6 (8.6)
3–6 drinks per week		7 (10.0)
7–14 drinks per week		4 (5.7)
>14 drinks per week		15 (21.4)
Number of weeks using ETOH during the past 6 mo	39	17 (0–26)
% NAC administered	79	72 (91.1)
Maximum coma grade (worst 7 d)	77
No coma		15 (19.5)
I		8 (10.4)
II		7 (9.1)
III		20 (26.0)
IV		27 (35.1)
Outcomes at 21 d
Waitlisted for transplant	78	11 (14.1)
Transplanted	79	10 (12.7)
Survival (overall)	79	61 (77.2)
Spontaneous survival	79	50 (63.3)

ALF, acute liver failure; ALI, acute liver injury; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; ETOH, ethanol; INR, international normalized ratio.

The demographics, presentation, and outcomes were compared between the 62 APAP MBS patients and the 1,619 APAP-ALF patients without MBS (Table 2). The median age of the 2 groups were similar (36 and 38 years, P < 0.08) whereas the median BMI was higher in the MBS group (28 vs 25, P < 0.001). There were no significant differences in race or ethnicity. MBS APAP patients were much more likely to be female (93.5% vs 72%, P < 0.001) and to report a history of diabetes (22.6% vs 10.8%, P = 0.012) and neurological disease (25.8% vs 14%, P = 0.015) compared with the non-MBS APAP patients. MBS APAP patients were also more likely to be anemic (hemoglobin 9.4 vs 11.3 g/dL; P < 0.001) and to have lower serum aminotransferase levels (median alanine aminotransferase 2,105 vs 4,100 IU/L, P < 0.001) on presentation compared with the non-MBS APAP patients. There were no significant differences in presenting INR, serum total bilirubin, or plasma ammonia levels between the 2 groups. Of note, there was a significantly smaller total dosage of APAP reported by the MBS patients (6,500 mg vs 12,000 mg, P = 0.009) and a much larger percentage of unintentional overdoses among the MBS cohort (71.2% vs 52.4%, P < 0.001). Reported alcohol use was available in 87% of the MBS APAP patients with most (48%) reporting abstinence but 29% reporting moderate-to-heavy use (Table 3). There were no significant differences observed in alcohol use between the 2 groups.

Table 2.

Comparison of the APAP MBS patients with APAP patients without MBS

	APAP MBS patients		Other APAP patients		P value
	N	Number (%) or median (range)	N	Number (%) or median (range)
Age (yr)	62	38 (19–70)	1,619	36 (17–80)	0.088
Sex (female)	62	58 (93.5)	1,619	1,166 (72.0)	<0.001
Race	62		1,619		0.601
White		57 (91.9)		1,341 (82.8)
African American		4 (6.5)		167 (10.3)
Other		1 (1.8)		111 (6.9)
Previous health and illnesses	62		1,619
Collagen/vascular diseases		5 (8.1)		74 (4.6)	0.210
Chronic liver disease		4 (6.5)		55 (3.4)	0.274
Endocrine/diabetes		14 (22.6)		175 (10.8)	0.012
Psychiatric disease		36 (58.1)		871 (53.8)	0.520
Neuro/seizure		16 (25.8)		227 (14.0)	0.015
Hypertension		11 (17.7)		212 (13.1)	0.337
Heart disease		7 (11.3)		92 (5.7)	0.099
Renal disease		4 (6.5)		80 (4.9)	0.548
Pulmonary disease		7 (11.3)		155 (9.6)	0.659
Substance abuse		29 (46.8)		605 (37.4)	0.143
GI disease		27 (43.5)		286 (17.7)	<0.001
HIV/AIDS		0		9 (0.6)	1.000
IDU at any time in past		3 (4.8)		95 (5.9)	1.000
BMI at admission	55	28 (19–66)	1,364	25 (14–64)	<0.001
Admission biochemistry
Hemoglobin (g/dL)	61	9.4 (6.8–12.0)	1,602	11.3 (4.3–24.5)	<0.001
White blood count (×109/L)	61	7.7 (0.3–27.7)	1,602	9.1 (0.3–83.0)	0.088
Platelet count (×109/L)	60	134 (13–890)	1,593	130 (4–613)	0.184
INR	62	2.4 (1.1–12.0)	1,617	2.9 (0.9–27.1)	0.035
AST (IU/L)	61	3,048 (41–17,871)	1,599	4,010 (25–40,850)	0.069
ALT (IU/L)	60	2,206 (73–10,841)	1,596	4,100 (14–29,060)	<0.001
Bilirubin (mg/dL)	62	4.1 (0.5–15.4)	1,618	4.0 (0.2–58.0)	0.484
Ammonia (venous; μmol/L)	29	102 (25–690)	626	96 (1.0–1,079)	0.631
Creatinine (mg/dL)	62	1.1 (0.4–12.3)	1,617	1.4 (0.3–11.9)	0.072
Phosphate (mg/dL)	52	2.6 (0.6–6.9)	1,378	2.4 (0.1–13.9)	0.632
APAP parent compound	56	29.0 (1.0–257.0)	1,407	33.1 (0–2,000.0)	0.328
APAP-cysteine adducts (nmol/mL)	13	1.517 (0.050–8.864)	562	6.077 (0.000–50.070)	0.002
APAP medication dosage (mg)	31	6,500 (325–86,050)	896	12,000 (325–75,000)	0.009
Number of APAP products reported taken	60		1,478		0.192
Intentionality	59		1,534		<0.001
Suicide attempt		7 (11.9)		612 (39.9)
Unintentional		42 (71.2)		804 (52.4)
Unknown		10 (16.9)		118 (7.7)
Therapeutic intent, primary indication	31		1,227		0.688
Acute pain syndrome (i.e., <1 wk of symptoms)		8 (25.8)		109 (27.8)
Infectious process		1 (3.2)		17 (4.3)
Subacute pain (i.e., 1–4 wk of symptoms)		6 (19.4)		71 (18.1)
Chronic pain (>4 wk of symptoms)		13 (41.9)		124 (31.6)
Other		3 (9.7)		71 (18.1)
% N-acetylcysteine given	79	72 (91.1)	1,617	1,482 (91.5)	0.873
Outcomes at 21 d
Waitlisted for transplant	62	7 (11.3)	1,613	299 (18.5)	0.180
Transplanted	62	6 (9.7)	1,619	101 (6.2)	0.282
Survival (overall)	62	51 (82.3)	1,619	1,331 (82.2)	1.000
Spontaneous survival	62	43 (69.4)	1,619	1,101 (85.2)	0.928

ALF, acute liver failure; ALI, acute liver injury; ALT, alanine aminotransferase; APAP, acetaminophen; AST, aspartate aminotransferase; BMI, body mass index; ETOH, ethanol; GI, gastrointestinal; IDU, intrauterine device; INR, international normalized ratio; MBS, metabolic and bariatric surgery.

Table 3.

Ethanol use in the MBS APAP vs all other APAP ALF/ALI patients

	N	MBS APAP	N	Other APAP	P value
Usual intake of ETOH during the past 6 mo	54		813		0.220
None (less than 1 drink per week)		26 (48.1)		333 (41.0)
1 drink per week		1 (1.9)		46 (5.7)
2 drinks per week		6 (11.1)		43 (5.3)
3–6 drinks per week		5 (9.3)		94 (11.6)
7–14 drinks per week		2 (3.7)		95 (11.7)
>14 drinks per week		14 (25.9)		202 (24.8)
Number of weeks using ETOH during the past 6 mo	35	12 (0–26)	556	24 (0–26)	0.082

ALF, acute liver failure; ALI, acute liver injury; APAP, acetaminophen; ETOH, ethanol; MBS, metabolic and bariatric surgery.

There were 6 patients in the cohort who were initially considered by the site investigator to have ALF, but on review by our committee were determined to have ACLF, the result of chronic MASLD, plus, in nearly all cases, septicemia. Following review, these 6 were not included because they were deemed not to have ALF. Table 4 describes the features of these 6 cases. None had very high aminotransferases, whereas all had steatohepatitis with many a high percentage of steatosis and most advanced fibrosis on histology. It is unlikely that they had acute hepatocellular dysfunction or injury to warrant the diagnosis of ALF, despite evidence of hepatic encephalopathy and coagulopathy. We did not find a single example of high aminotransferases or evidence of microvesicular fat leading to liver dysfunction in these 6 cases.

Table 4.

Demographic and clinical data for the 6 MBS patients with ACLF initially thought to be ALF but corrected to other, not ALF

	Case 1	Case 2	Case 3	Case 4	Case 5	Case 6
Age (yr)	33	33	25	37	39	36
Sex	F	F	F	F	F	F
Race/ethnicity	White	Hispanic	Hispanic	White	African American	White
PI etiology	Indeterminate	Indeterminate	Other: NASH	Other: NASH	Other: NAFLD	Other: acute fatty liver
Adjudicated etiology	Other: NASH	Other: NASH	Other: NASH	Other: not ALF	Other: fatty liver	Other: not ALF
Adjudicated diagnosis	MASH-ACLF	MASH-ACLF	MASH-ACLF	MASH-ACLF	MASLD/MASH-SALF	MASH-ACLF
Days from hospital to study admission	2	10	6	7	19	11
Weight loss surgery type	Roux-en-Y	Roux-en-Y	Roux-en-Y	“Gastric bypass”	“Gastric bypass”	Roux-en-Y
Time since surgery	6 yr	5 mo	4 mo	3 yr	19 yr	14 yr
History of rapid weight loss	?	Y	Y	Y	N	Y
History of nutrition deficit	Y	?	Y	Y	N	FSV?
Liver biopsy features
Steatohepatitis	Y	Y	Y	Y	Y	Y
Fibrosis stage	3	2	0	3 to 4 (early cirrhosis on explant)	0	3 at least
Infection present?	WBC+	Sepsis	UTI/sepsis	UTI/sepsis	Sepsis	PNA
Discharge status	Alive	Dead	Alive (21 d outcome unknown)	Transplanted	Dead	Dead
Study admission biochemical results
White blood count (×109/L)	12.6	19.9	14.4	7.7	9.9	12.5
Hemoglobin (g/dL)	7.9	10.3	10.1	8.5	8.9	6.5
Platelet count (×109/L)	153	169	80	134	81	120
INR	3.2	1.6	1.3	2.5	2.4	2.6
AST (IU/L)	238a	93	56	126	119	52
ALT (IU/L)	71a	22	50	27	38	17
Bilirubin (mg/dL)	27.9a	13.9	8.7	29.2	28.7	26.2
Ammonia (venous; μmol/L)	91b	95.0	32.0		57.0
Creatinine (mg/dL)	1.1a	1.3	0.6	1.9	2.3	2.7
Phosphate (mg/dL)		1.9	2.0		2.2	3.2

ACLF, acute-on-chronic liver failure; ALF, acute liver failure; ALI, acute liver injury; ALT, alanine aminotransferase; APAP, acetaminophen; AST, aspartate aminotransferase; INR, international normalized ratio; MASH, metabolic dysfunction-associated steatohepatitis; MASLD, metabolic dysfunction-associated steatotic liver disease; MBS, metabolic and bariatric surgery; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; PI, site principal investigator; SAHA, time from hospital admission to study admission.

a

Prestudy laboratory values.

b

Day 2 laboratory values.

The number of MBS ALF patients enrolled in the registry overall (79, 2.3%) seems to be higher than the percent of the US population having had MBS procedures: 1.62% at last estimate. Between 1998 and 2022, the percent of the US population undergoing MBS annually had increased from 0.006% to 0.07% (by 2009), and to 0.08% by 2022. This apparent plateau effect may be related to the impact of the COVID-19 epidemic. In our study, the number enrolled in the first 10 years was considerably less than the number enrolled in the next 10: 27 cases between 1998 and 2009 and 52 cases between 2010 and 2019 (Figure 1), yielding 1.6% during the first and 3.4% during the second period. Thus, the prevalence of ALF in the MBS population seems to be on the rise and would appear to have outpaced the ever-increasing number of subjects who have had bariatric surgery. The increase in ALF cases during the second time period may simply reflect, in part, an increase in the number of patients in the population having undergone MBS by that time (16).

[Figure omitted. See PDF]

DISCUSSION

The current large multicenter prospective cohort study confirmed a prior single center study suggesting that bariatric surgery patients are over-represented among patients with ALF/ALI, representing 2.3% of ALF patients vs 1.6% of the overall US population. There are multiple etiologies of ALF in patients who have undergone MBS, possibly including effects from the specific type of MBS itself. We identified 6 patients of the 85 total where the etiology was suspected to be directly related to MASLD in the setting of MBS. However, on careful review by a committee of senior hepatologists, we concluded that these patients did not represent ALF but rather ACLF accompanied by sepsis, and they were deleted from further consideration, several but not all having significant fibrosis. In our study, ALF secondary to steatohepatitis alone was not found, despite the apparent acute presentation in a few instances.

APAP hepatotoxicity, in addition to being the most frequently implicated ALF/ALI etiology, was even more prevalent in the MBS-ALF group at 78.5% compared with the 49.3% APAP toxicity prevalence observed in those with ALF/ALI who had not had MBS. The reasons for these findings remain unclear. It is possible that APAP toxicity evolves more rapidly in the setting of prior MBS either due to relative malnutrition of the MBS patient with depletion of intrahepatic and total body glutathione stores or perhaps solely due to rapid transit of medication to the jejunum or to both mechanisms. In support of the latter, 1 recent study demonstrated more rapid absorption and higher peak APAP concentrations after Roux-en-y gastric bypass compared with preoperative pharmacokinetics (17). In addition, data from a large multicenter study have shown an increase in drug-related and alcohol-related mortality after bariatric surgery, which are factors commonly associated with APAP overdose (18). Drug and alcohol use and abuse are common within the overall APAP overdose population outside of MBS as well (19). We could not discern a difference in reported alcohol use between the MBS and non-MBS groups in this study (Table 3).

Although obesity predisposes to MASLD and bariatric surgery can improve many metabolic factors, severe liver dysfunction may also occasionally occur after specific MBS procedures (5). For example, 1 study found the rate of ALF as high as 7% out of nearly 500 jejunoileal bypass (JIB) patients (6). Another review article of 14 studies found that a few patients who had undergone biliopancreatic diversion developed ALF a median of 20 months after MBS, which was a significantly shorter time period than those who developed ALF after JIB (7). A Belgian study followed 10 patients listed for liver transplantation because of liver failure after MBS: 9 had undergone biliopancreatic diversion and 1 had the JIB (8). We were unable to identify a single case of ALF related directly to steatohepatitis alone despite the initial diagnosis of ALF though this may be reflective of the type of bariatric surgery in our cohort. All 6 of our patients, the majority with Roux-en-Y gastric bypass MBS (Table 4), initially considered to have ALF without a qualifying other etiology such as APAP, were found to have septicemia and/or simply liver failure in the setting of chronic liver disease. Patients with MASLD have been shown to be unusually susceptible to septicemia, and this was present or suspected in all 6 cases (20).

There are limited published data regarding the potential for altered metabolism of APAP after MBS. In 1 pharmacologic study, patients' serum APAP concentrations after administration of a 1,500 mg liquid dose were measured at time points before and 3 and 12 months after Roux-en-Y surgery, showing a 2-fold increase in peak APAP concentrations postsurgery and higher steady-state APAP concentrations after standard 650 mg every 4 hour dosage compared with healthy controls, supporting the possibility that altered drug metabolism occurs in the setting of MBS. However, an increase in APAP-cysteine protein adducts was not seen in a further study, with proposed reason for this being down regulation of cytochrome P450 2E1 enzymes (17).

In favor of the hypothesis that severe malnutrition, micronutrient deficiencies, and a higher incidence of alcohol use disorder (AUD) may play a role is our observation that MBS patients were significantly more anemic than the non-MBS group. A cohort study from 2021 followed 17 patients with decompensated liver disease after MBS and found that all had malnutrition; 10 also had ongoing alcohol ingestion that could compound their worsening liver dysfunction (21). In another systematic review of nearly 50,000 patients who had undergone bariatric surgery, more than 80% of whom were female, there was an increase in the prevalence of AUD after MBS as compared with prior, with the highest prevalence of AUD after Roux-en-Y gastric bypass procedures (22). We did not observe a difference in either alcohol consumption or prevalence of excessive drinking in the MBS group compared with those without MBS; however, detailed data were somewhat limited (Table 3).

It is noteworthy that the APAP MBS patients in this study reported lower doses of APAP ingested and were less likely to have taken APAP as a suicide attempt (Table 2), again suggesting that relatively low-dose ingestions over several days with therapeutic intent can lead to more serious liver injury in this setting than would otherwise be expected (23). These findings from this study and the limited data available regarding altered APAP metabolism post-MBS do suggest that MBS predisposes to a higher risk of APAP toxicity when controlling for the dose and possibly an increased sensitivity at lower daily doses. Further research is needed to elucidate this possibility, especially since over 5 million Americans have undergone MBS and APAP is one of the most widely used over the counter medications in the country. Our data suggest that once MBS patients develop ALF/ALI they appear to do reasonably well and similarly to their non-MBS counterparts despite presenting with more advanced grades of encephalopathy. We could not demonstrate a difference in overall or transplant-free survival between the 2 groups.

It is unclear whether there are more cases of ALF in patients who have undergone MBS over time, that is, a higher incidence of ALF in MBS patients in the past decade than the prior one. Although MBS becomes increasingly popular providing more candidates for potential APAP overdose, the relationship of the MBS to APAP toxicity remains unclear. One might suppose that newer techniques (such as laparoscopic sleeve gastrectomy) might be less prone to toxicity than the earlier Roux-en-Y approaches, as a larger surface area of gastric and duodenal mucosa is available to help metabolize APAP, but further studies are needed to explore this hypothesis.

There are several important limitations to our study. First, while we were certain that a MBS had been performed, information regarding the type of procedure performed and when it was performed was not often reported. Moreover, the study relied on recollection of history directly from patients or from family members/friends in cases where patients were not aware and could not communicate. In addition, formal assessments of patient nutritional status were not undertaken, neither body composition analysis or nor presence of occult micronutrient deficiencies in either group. Future studies will need to account also for use of newer weight loss medications such as the glucagon-like receptor agonist agents (GLP-1, GIP-1 RA) in altering APAP metabolism in the presence of more likely marginal nutritional states.

In summary, more MBS patients than expected experience ALF/ALI and most of these cases represented APAP hepatotoxicity. These data, coupled with the lower cumulative dose of APAP ingested suggest that MBS patients may be at increased risk of inadvertent APAP hepatotoxicity. The MBS ALF/ALI APAP patient group was significantly more likely to be female, White, and experienced unintentional or lower dose-type injuries. Our registry demonstrates that the frequency of ALF/ALI in this population seems to be increasing over the past 2 decades. Future studies should aim to elucidate the reasons for why APAP ALF risk is increased in MBS patients.

CONFLICTS OF INTEREST

Guarantor of the article: Shannan R. Tujios, MD.

Specific author contributions: S.R.T., J.A.R., and W.M.L. were responsible for planning and conducting the study, interpreting data and drafting the manuscript. M.G. and V.D. responsible for interpreting data. I.W., A.A., J.A., and R.J.F. were responsible for interpreting data and drafting the manuscript. All authors approved final draft.

Financial support: The Acute Liver Failure Study Group was supported by R-01 DK 58369 and U-01 DK 58369, to UT Southwestern Medical Center, as well as the Jeanne Roberts Fund of the Southwestern Medical Foundation, Dallas, TX.

Potential competing interests: W.M.L. reports receiving research funding from Gilead, Novo Nordisk, Eiger, Alexion, Vivet, Camurus and Lipocine; and consulting: SeaGen, GSK, Veristat, Genentech, InhibRx. R.J.F. has received research funding from Takeda and Kezar Pharmaceuticals and served as consultant for Moderna.

Trial registration: ALFSG registry NCT number: NCT00518440.

Study Highlights

BACKGROUND

• ✓ Metabolic/bariatric surgery are increasingly utilized to treat obesity and metabolic dysfunction associated complications with over 250,000 Americans undergoing surgery annually. Acetaminophen is commonly recommended in this population.

FINDINGS

• ✓ A multicenter 20-year study of acute liver failure/injury determined acute liver failure especially from acetaminophen is more frequent in patients with history of metabolic/bariatric surgery than the general population. The majority were unintentional overdoses with a significantly lower median total acetaminophen dose reported.

IMPLICATIONS FOR PATIENT CARE

• ✓ Metabolic/bariatric surgery appears to predispose patients to unintentional acetaminophen hepatotoxicity at lower doses. Patients with metabolic/bariatric surgery should be counseled to avoid exceeding recommended acetaminophen dosages.

Author Notes

Correspondence: Shannan R. Tujios, MD. E-mail: [email protected].

REFERENCES

1.ASMBS. Estimate of bariatric surgery numbers, 2011–2023. (https://asmbs.org/resources/estimate-of-bariatric-surgery-numbers). Accessed January 8, 2024.

2.Verrastro, O, Panunzi, S, Castagneto-Gissey, L, et al. Bariatric–metabolic surgery versus lifestyle intervention plus best medical care in non-alcoholic steatohepatitis (BRAVES): A multicentre, open-label, randomised trial. Lancet 2023;401(10390):1786–97.

3.D'Albuquerque, LA, Gonzalez, AM, Wahle, RC, et al. Liver transplantation for subacute hepatocellular failure due to massive steatohepatitis after bariatric surgery. Liver Transpl 2008;14(6):881–5.

4.Eilenberg, M, Langer, FB, Beer, A, et al. Significant liver-related morbidity after bariatric surgery and its reversal: A case series. Obes Surg 2018;28(3):812–9.

5.Vespasiani-Gentilucci, U, Vorini, F, Carotti, S, et al. Hepatic complications of bariatric surgery: The reverse side of the coin. Acta Gastroenterol Belg 2017;80(4):505–13.

6.Aktas, A, Gokler, C, Sansal, M, et al. Acute liver failure following sleeve gastrectomy with jejuno-ileal bypass. Obes Res Clin Pract 2021;15(3):297–9.

7.Requarth, JA, Burchard, KW, Colacchio, TA, et al. Long-term morbidity following jejunoileal bypass. The continuing potential need for surgical reversal. Arch Surg 1995;130(3):318–25.

8.Addeo, P, Cesaretti, M, Anty, R, et al. Liver transplantation for bariatric surgery-related liver failure: A systematic review of a rare condition. Surg Obes Relat Dis 2019;15(8):1394–401.

9.Geerts, A, Darius, T, Chapelle, T, et al. The multicenter Belgian survey on liver transplantation for hepatocellular failure after bariatric surgery. Transplant Proc 2010;42(10):4395–8.

10.Yoon, E, Babar, A, Choudhary, M, et al. Acetaminophen-induced hepatotoxicity: A comprehensive update. J Clin Transl Hepatol 2016;4(2):131–42.

11.Reuben, A, Fontana, RJ, Davern, T, et al. Outcomes in adults with acute liver failure (ALF) from 1998–2013: An observational cohort study. Ann Intern Med 2016;164:724–32.

12.Karvellas, CJ, Leventhal, TM, Rakela, JL, et al. Outcomes of patients with acute liver failure listed for liver transplantation: A multicenter prospective cohort analysis. Liver Transpl 2023;29(3):318–30.

13.Holt, EW, DeMartini, S, Davern, TJ. Acute liver failure due to acetaminophen poisoningin patients with prior weight loss surgery: A case series. J Clin Gastroenterol 2015;49(9):790–3.

14.Koch, DG, Speiser, JL, Durkalski, V, et al. The natural history of severe acute liver injury. Am J Gastroenterol 2017;112(9):1389–96.

15.Ganger, DR, Rule, J, Rakela, J, et al. Acute liver failure of indeterminate etiology: A comprehensive systematic approach by an expert committee to establish causality. Am J Gastroenterol 2018;113(9):1319–28.

16.Alalwan, AA, Friedman, J, Park, H, et al. US national trends in bariatric surgery: A decade of study. Surgery 2021;170(1):13–7.

17.Chen, KF, Chan, LN, Senn, TD, et al. The impact of proximal roux-en-Y gastric bypass surgery on acetaminophen absorption and metabolism. Pharmacotherapy 2020;40(3):191–203.

18.White, GE, Courcoulas, AP, King, WC. Drug- and alcohol-related mortality risk after bariatric surgery: Evidence from a 7-year prospective multicenter cohort study. Surg Obes Relat Dis 2019;15(7):1160–9.

19.Randesi, M, Levran, O, Correa da Rosa, J, et al. Association of variants of arginine vasopressin and arginine vasopressin receptor 1A, with severe acetaminophen liver injury. Cell Mol Gastroenterol Hepatol 2017;3:500–5.

20.Ebrahimi, F, Simon, TG, Hagström, H, et al. Risk of severe infection in patients with biopsy-proven nonalcoholic fatty liver disease: A population-based cohort study. Clin Gastroenterol Hepatol 2023;21(13):3346–55.e19.

21.Mendoza, YP, Becchetti, C, Wan, T, et al. Malnutrition and alcohol in patients presenting with severe complications of cirrhosis after laparoscopic bariatric surgery. Obes Surg 2021;31(6):2817–22.

22.Cerón-Solano, G, Zepeda, RC, Romero Lozano, JG, et al. Bariatric surgery and alcohol and substance abuse disorder: A systematic review. Cir Esp (Engl Ed) 2021;99(9):635–47.

23.Craig, DG, Bates, CM, Davidson, JS, et al. Overdose pattern and outcome in paracetamol-induced acute severe hepatotoxicity. Br J Clin Pharmacol 2011;71(2):273–82.