Higher Pre-Operative Non-Invasive Fibrosis Scores are not Associated with Surgical Outcomes in Patients with MASLD Undergoing Metabolic and Bariatric Surgery

Abstract

Background:

Metabolic dysfunction-associated steatotic liver disease (MASLD) is a leading cause of cirrhosis in Canada. Metabolic and bariatric surgery (MBS) leads to histologic improvement in patients with MASLD and obesity; however, patients with advanced fibrosis may be at risk for decompensation in liver function following MBS. There are currently no Canadian data evaluating the association between pre-operative advanced fibrosis and post-operative complications among patients with obesity undergoing MBS.

Methods:

We conducted a single-centre retrospective cohort study of individuals undergoing primary MBS (Roux-en-Y gastric bypass and sleeve gastrectomy) between September 2016 and August 2021. The primary exposure was pre-operative NAFLD (non-alcoholic fatty liver disease) fibrosis score (NFS). Outcomes of interest included post-operative complications and health care utilization at 30 and 90 days, and weight loss at 30 days post-operatively.

Results:

After exclusions, 204 patients who received MBS and had the data required to calculate NFS were included. The median age was 46 years (interquartile range 38–55 years) and 181 (89%) were female. A total of 55 (27%) patients had NFS, suggesting advanced fibrosis. There was no significant difference in 30- or 90-day complications and health care utilization rates in patients with higher NFS.

Conclusions:

Patients undergoing MBS have a high prevalence of steatosis and advanced fibrosis based on the NFS. Our data suggest that higher NFS is not associated with higher short-term post-operative complications in patients undergoing MBS. Further work needs to be done to determine the optimal method of screening for advanced liver disease in patients living with severe obesity.

Lay Summary

Metabolic dysfunction-associated steatotic liver disease (MASLD) is the leading cause of cirrhosis in Canada. Metabolic and bariatric surgery (MBS) can help treat MASLD by reducing the amount of fat and scarring in the liver. However, some patients with MASLD can experience worsening of their liver function after surgery. While liver biopsies remain the gold standard for diagnosing MASLD and grading its severity, they are also invasive and have risks associated with them. Non-invasive tests like the NAFLD (non-alcoholic fatty liver disease) fibrosis score (NFS), which use a combination of bloodwork and patient demographics, can be used to screen for advanced liver scarring in patients with MASLD. Our study aimed to determine how common advanced scarring was in patients undergoing MBS, and whether a higher NFS was associated with worse outcomes after surgery. Using the NFS, we discovered that over 25% of patients with MASLD undergoing MBS have advanced scarring; however, patients with higher NFS did not have more complications after surgery. Future studies should continue exploring the best methods of screening for MASLD and advanced scarring in patients living with obesity to determine who is at higher risk for complications after MBS.

Introduction

Metabolic dysfunction-associated steatotic liver disease (MASLD) is one of the leading causes of cirrhosis in Canada (1). Metabolic and bariatric surgery (MBS) can lead to histologic improvement in both steatosis and fibrosis in patients with MASLD and obesity (2,3), in addition to improved cardiovascular outcomes (4). However, decompensation in liver function following MBS has been reported even in the absence of cirrhosis (5). While liver biopsy is the gold standard for diagnosing fibrosis in MASLD, non-invasive fibrosis scores such as the NAFLD (non-alcoholic fatty liver disease) fibrosis score (NFS) have been developed to estimate the degree of advanced fibrosis in patients with MASLD (6). There are currently no Canadian data describing the prevalence of MASLD and advanced fibrosis using non-invasive fibrosis scores in patients with severe obesity who are undergoing MBS. Furthermore, it is not known whether higher NFS are associated with worse post-operative outcomes after MBS.

Methods

We conducted a single-centre retrospective cohort study of individuals with MASLD undergoing primary MBS (Roux-en-Y gastric bypass and sleeve gastrectomy) at Kingston Health Sciences Center between September 2016 and August 2021. MASLD was defined as the presence of imaging showing hepatic steatosis and the presence of obesity. Patients who were eligible for and underwent primary MBS (i.e. at least 18 years of age with BMI ≥40 kg/m² or ≥35 kg/m² and at least one obesity-related comorbidity on past medical history) and had pre-operative imaging documenting steatosis were included in the study. Patients who were having a revision or conversion operation after a prior MBS were excluded. All data were abstracted from the electronic medical record by two investigators (L Huynh, E Kouzmina). The primary exposure was pre-operative NFS calculated using data within 1 year prior to surgery. Baseline demographic data, obesity-related comorbidities, and pre-operative consultations with internal medicine or hepatology were also abstracted (Supplemental Table 1). Outcomes of interest included post-operative surgical complications, medical complications, and health care utilization (emergency department visit, hospitalization, or unplanned outpatient visit to a specialist provider) at 30 and 90 days (Supplemental Table 2). Weight loss outcomes were also captured using total and excess percentage weight loss at 30 days post-operatively. Descriptive statistics were reported as frequencies and percentages for categorical variables and mean with standard deviation for continuous variables. Demographics and outcomes were stratified by pre-operative NFS (±0.675) using Student's t-tests for continuous variables and Chi-squared tests for categorical variables. Multivariate logistic regression analysis was done to evaluate the association between NFS and 30-day health care utilization adjusting for age, sex, and pre-operative BMI. Two-sided p values <0.05 were considered to be statistically significant. Statistical analyses were conducted using STATA, version 21.1 (College Station, TX, USA). This study was approved by the Queen's University Health Sciences Research Ethics Board (SURG-452-18).

Results

Baseline characteristics

A total of 599 patients underwent primary MBS, 265 of which met criteria for MASLD based on pre-operative imaging documenting steatosis and the presence of obesity. Among patients with MASLD, 204 patients had data required to calculate NFS and formed the study cohort (Figure 1). The median age was 46 years (interquartile range 38–55), 181 (89%) were female, median pre-operative BMI was 47 kg/m² (interquartile range 43–52), and 66 patients (32%) had diabetes. A total of 92 patients (45%) were assessed by internal medicine pre-operatively, while 3 patients (1%) were assessed by hepatology. In total, 28 (14%) had an NFS consistent with F0–F2 fibrosis (<–1.455), 121 (59%) had indeterminate NFS values (–1.455 to –0.675), and 55 (27%) had scores consistent with advanced (F3–F4) fibrosis (>0.675). Patients with scores predicting advanced fibrosis (i.e. F3–F4) were more likely to be older, male, have a higher pre-operative BMI, higher rates of comorbidities (diabetes, COPD, sleep apnea), and more likely to be seen by an internist pre-operatively (Table 1).

Figure 1: Study cohort flowchart.

Table 1:

Baseline characteristics by NFS

	NFS <0.675 (n = 149)	NFS >0.675 (n = 55)	p value
Age, mean (SD)	44.6 (11.4)	50.5 (10.7)	<0.001
Female sex, n (%)	138 (92.6)	43 (78.2)	0.004
Initial BMI, mean (SD)	46.7 (6.1)	52.4 (8.3)	<0.001
ASA score ≥3	125 (84.9)	49 (89.1)	0.352
MSK pain	94 (63.1)	40 (72.3)	0.198
Cardiovascular disease (angina, heart failure, MI)	9 (6.0)	6 (10.9)	0.237
Diabetes	34 (22.8)	32 (58.20)	<0.001
Dyslipidemia	42 (28.2)	22 (40.0)	0.107
Stroke (TIA/CVA)	2 (1.3)	1 (1.8)	0.802
DVT/PE	5 (3.4)	4 (7.3)	0.227
COPD	3 (2.0)	6 (10.9)	0.006
Asthma	41 (27.5)	13 (23.6)	0.577
Chronic kidney disease	14 (9.4)	10 (18.2)	0.084
OSA	123 (82.6)	53 (96.4)	0.011
Depression	104 (69.80)	33 (60.0)	0.186
Anxiety	83 (55.7)	27 (49.1)	0.4
Internist consult, n (%)	59 (39.6)	33 (60.0)	0.009
Hepatologist consult, n (%)	2 (1.3)	1 (1.8)	0.802

ASA = American Society of Anesthesiologists; MSK = musculoskeletal; MI = myocardial infarction; TIA = transient ischemic attack; CVA = cerebrovascular accident; DVT = deep vein thrombosis; PE = pulmonary embolism; COPD = chronic obstructive pulmonary disease; OSA = obstructive sleep apnea

Post-operative outcomes

At 30 days, 18 patients (9%) experienced a post-operative complication, while 39 (19%) utilized health care services. At 90 days, 6 patients (3%) experienced a surgical complication, while 2 (1%) experienced a late medical complication, and 32 (16%) had unplanned health care utilization. There was no significant difference in total percentage weight loss or excess weight loss between the two groups. At 30 and 90 days post-operatively, there was no significant difference in post-operative complications or health care utilization in patients with higher NFS (Table 2). Multivariate analysis did not demonstrate any association between NFS and 30-day health care utilization rates after controlling for age, sex, or pre-operative BMI (Table 3).

Table 2:

Post-operative complications in patients undergoing bariatric surgery stratified by NFS

	NFS <0.675 (n = 149)	NFS >0.675 (n = 55)	p value
30-day surgical complications, n (%)	14 (9.4)	4 (7.3)	0.564
30-day healthcare utilization, n (%)	33 (22.1)	6 (10.9)	0.07
3-month surgical complications, n (%)	5 (3.4)	1 (1.8)	0.564
3-month medical complications, n (%)	1 (0.7)	1 (1.8)	0.461
3-month healthcare utilization, n (%)	26 (17.4)	6 (10.9)	0.254
*Total percentage weight loss, % (SD)	11.5 (4.5)	11.2 (4.5)	0.7038
*Excess percentage weight loss, % (SD)	25.7 (11.5)	22.9 (10.1)	0.1402
*Change in BMI, kg/m2 (SD)	5.4 (2.2)	5.7 (2.5)	0.3706

^*N = 180; NFS = NAFLD fibrosis score; SD = standard deviation

Table 3:

Logistic regression evaluating the association between NFS and 30-day health care utilization in patients undergoing bariatric surgery

	Univariate			Multivariate
	Odds ratio	95% CI	p value	Odds ratio	95% CI	p value
NFS >0.675	0.43	(0.2–1.1)	0.076	0.63	(0.2–1.8)	0.39
Male sex	0.88	(0.3–2.8)	0.82	1.30	(0.4–4.2)	0.70
Age (per year)	0.96	(0.9–1.0)	0.02	0.96	(0.9–1.0)	0.02
BMI	0.98	(0.9–1.0)	0.38	0.97	(0.9–1.0)	0.28

CI = confidence interval; NFS = NAFLD fibrosis score

Discussion

In our single-centre cohort study of patients with MASLD undergoing metabolic and bariatric surgery MSB in one region of Ontario, Canada, 27% had NFS suggestive of advanced fibrosis. Our findings are in keeping with previous studies of biopsy-proven MASLD in MBS patients, with a prevalence of at least F3 fibrosis ranging from 18% to 33% (3,4,7). Despite almost a quarter of our patients having advanced fibrosis based on their NFS, only three patients were seen by a hepatologist in the peri-operative period where they would have had access to further fibrosis staging with transient elastography. Nonetheless, our cohort had similar post-operative complication rates as other patients undergoing MBS (8). Our findings suggest that patients with MASLD have similar post-operative outcomes without routine peri-operative optimization with hepatology.

Our study demonstrated that higher NFS was not associated with worse 30- and 90-day post-operative outcomes following MBS. Previously, a large of cohort study of almost 20,000 patients without liver disease undergoing general anesthesia (cardiac and non-cardiac surgeries) demonstrated that higher fibrosis scores (using the FIB-4 score) were associated with increased peri-operative mortality (9). The lack of association between higher fibrosis scores and post-operative complications in our cohort may be due to the fact that patients in our cohort underwent a multidisciplinary assessment and medical optimization for MBS. Notably, patients with clinical or radiographic evidence of portal hypertension (e.g. presences of ascites, features of decompensated cirrhosis) would have been excluded from MBS as it is an established risk factor for poor peri-operative outcomes in patients with advanced liver disease (10). Furthermore, post-operative complications related to decompensation of liver disease may have not have been captured in the short follow-up period of this study, with onset of liver decompensation reported in the literature to vary from 8 months to 17 years (5).

Several limitations of our study must be acknowledged. First, clinical outcomes related to liver-related events were not specifically collected. Second, the single-centre nature of the study limited the generalizability of our results. Third, almost 30% of patients in our cohort did not have the required variables to calculate NFS, resulting in potential selection bias. Furthermore, whether patients in our cohort had underlying liver disease prior to surgery was not captured in the baseline demographics, which could result in possible confounding of results. For example, the presence of hepatic steatosis on imaging may have been secondary to concurrent alcohol use, although generally patients with active alcohol use would not have been eligible for MBS. Finally, our cohort had a modest sample size, which may have been underpowered to interpret the results.

In conclusion, our study demonstrates that patients at a single-centre hospital in Ontario, Canada, who are seeking MBS have a high prevalence of advanced fibrosis based on the NFS, although a higher NFS is not associated with worse short-term post-operative outcomes. Further work needs to be done to determine the optimal method of screening for advanced liver disease in this patient population, identifying patients at risk for post-operative complications, and determine which patients may require hepatology consultation for peri-operative optimization and risk stratification in a resource-limited health care system.

Funding Statement

N/A

Contributions:

Conceptualization: L Huynh, E Kouzmina, B Zevin, JA Flemming; Methodology: L Huynh, E Kouzmina, B Zevin, JA Flemming; Formal Analysis: L Huynh; Investigation: L Huynh, E Kouzmina, B Zevin, JA Flemming; Data Curation: E Kouzmina; Writing - Original Draft: L Huynh; Writing - Review & Editing: L Huynh, E Kouzmina, B Zevin, JA Flemming; Supervision: B Zevin, JA Flemming.

Ethics Approval:

This study was approved by the Queen’s University Health Sciences Research Ethics Board.

Informed Consent:

N/A

Registry and the Registration no. of the Study/Trial:

N/A

Data Accessibility:

The data will not be made publicly available. Researchers interested in accessing the study data can contact the corresponding author for further information.

Funding:

N/A

Disclosures:

N/A

Peer Review:

This article has been peer reviewed.

Animal Studies:

N/A