A method of predicting fibrosis risk in women with non-alcoholic fatty liver disease and postmenopausal obesity

Full Text

Introduction

Women in the postmenopausal period, especially those not receiving hormone replacement therapy, are at risk of developing non-alcoholic fatty liver disease (NAFLD). In 70–90 % of cases, the clinical and morphological form of hepatic steatosis is registered in NAFLD, which is characterised by a mild symptomatic course. Steatohepatitis develops in 10–30 % of patients, and liver fibrosis subsequently forms in 25–40 % of them, gradually leading to cirrhosis of the organ in 20–30 % of cases [1]. Obesity, which is recorded in more than half of postmenopausal women, increases the likelihood of developing hepatic steatosis by more than three times [2–5].

The prevalence of NAFLD among postmenopausal women is up to 37 %, according to some data [6], and in women aged 40–50 years with metabolic disorders, hepatic steatosis is detected in 60–75 % of cases. There is evidence that postmenopausal women have a higher risk of developing liver fibrosis and a worse life prognosis compared to men and women of reproductive age and in the premenopausal period [7].

The prognosis for patients is determined by the presence of liver fibrosis (LF), and the risk of adverse outcomes increases according to the stage of fibrosis [8]. Therefore, the development of accessible methods for assessing the risk of LF in patients with NAFLD is relevant, primarily in at-risk groups, which include women in the postmenopausal period.

The objective of the study is to develop a method for predicting fibrosis risk in patients with non-alcoholic fatty liver disease (NAFLD) and postmenopausal obesity.

Materials and Methods

The study involved 70 patients (average age 51.0 (49.0–52.0) years) with NAFLD and overweight or obesity in the postmenopausal period. Signs of hepatic steatosis were verified by ultrasound after excluding other causes of its development [8]. The patients' platelet count, glucose concentration, and activity of aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were determined (Table 1).

 

Table 1  Laboratory methods of examination in a group of patients with NAFLD and postmenopausal obesity

Parameter, unit of measurement	Reference values	Equipment	Reagents
ALT, U/L	0–55	Architect-4000 biochemical analyser (USA)	Abbott (USA)
AST, U/L	5–34
Glucose, mmol/L	3.9–6.1
Platelets, 10⁹/L	180–380	Medonic M20  (BOULE MEDICAL АВ, Sweden)	BOULE MEDICAL АВ (Sweden)

 

The body mass index (BMI), steatosis index HSI [9], and fibrosis indices APRI [10; 11] and FIB4 [12–14] were also calculated. The formulas for calculating the indices and their interpretation are presented in Table 2.

 

Table 2 Formulae for calculating steatosis and fibrosis indices

Indexes	Formula	Interpretation
HIS (Hepatic Steatosis index)	HSI = 8 • (ALT / AST) + BMI (+2, if female; +2, if diabetic)	Less than 30.0, steatosis is excluded; above 36.0, the risk of steatosis is high with a sensitivity of 93.1 % and specificity of 92.4 %
APRI (AST to Platelet Ratio Index)|	APRI = AST (U/L) / (upper limit of normal AST in U/L) • 100 / platelets (10⁹/L)	No more than 0.5, the probability of fibrosis is low (negative predictive value 83 %); more than 1.5, the probability of fibrosis is high (positive predictive value 68.4 %), "grey zone" from 0.5 to 1.5. According to updated data, the threshold value for predicting significant fibrosis is 0.88
FIB-4	Online calculator: https://www.hepatitisc.uw.edu/page/clinical-calculators/fib-4	FIB-4 < 1.30, no significant fibrosis (reliability 90 %); FIB-4 > 2.67, presence of severe fibrosis (reliability 80 %), "grey zone" from 1.3 to 2.67

 

In patients with elevated fibrosis index values, liver fibroelastometry was performed to confirm and establish the approximate stage of liver fibrosis according to the Metavir scale.

To describe the signs, the median (Me) and the boundaries of its confidence interval (xJ–xK) were used, with the significance of differences assessed using the Mann–Whitney U-test. Differences were considered statistically significant at p < 0.05. The constant and coefficients for the formula for assessing the risk of fibrosis formation were calculated using multiple regression analysis.

Results and Discussion

According to BMI data, in the group of patients with NAFLD, 24 (34 %) women had excess body weight, 20 (29 %) patients had obesity of the 1st degree, and 26 (37 %) had obesity of the 2nd–3rd degree. Transaminase activity and platelet counts in the overall group almost did not exceed reference values and amounted to: AST — 19 (17–22) U/L; ALT — 20 (17–27.5) U/L; platelet count — 256 (227.5–305) · 10⁹/L. However, 6 % of patients had platelet counts at the lower limit of the norm, and 40% had fasting hyperglycaemia. The HSI index in all patients exceeded 36, indicating a high risk of steatosis, and averaged 45.0 (41.5–47.3).

According to fibrosis indices, most patients had a low probability of fibrosis. The APRI index averaged 0.2 (0.2–0.3), which was less than the recommended value of 0.5, and the FIB4 index averaged 0.8 (0.7–1.0), respectively, was less than 1.3, indicating the absence of significant fibrosis [8]. However, in 12 patients with NAFLD, fibrosis index values were above the threshold values and fell into the «grey zone», which did not exclude the presence of the initial stages of liver fibrosis. Therefore, we formed two groups of women with NAFLD and obesity in postmenopause with low and high risk of liver fibrosis and, depending on this sign, analysed their biometric and laboratory parameters.

Patients with a high risk of liver fibrosis were significantly older (p = 0.044); they had a lower platelet count (p < 0.001), higher glucose concentration (p = 0.046), and higher transaminase activity. BMI and HSI index values did not differ significantly, although they were slightly higher in the group with a high risk of liver fibrosis (Table 3). Fibroelastometry performed on 6 patients from the high-risk group showed stage I fibrosis in 4 cases and stage II fibrosis in 2 cases.

 

Table 3 The parameters studied in groups of patients with NAFLD and obesity in postmenopause, depending on the risk of fibrosis, Ме (x_J–x_K)

Parameter	Group of women with low risk of liver fibrosis, n = 58	Group of women with high risk of liver fibrosis, n = 12	р
Age, years	50.5 (48.0–52.0)	53.0 (49.0–56.0)	0.044
BMI, kg/m²	30.0 (29.1–32.9)	33.3 (31.2–39.0)	0.094
Platelets, 10⁹/L	266.0 (253.0–292.0)	204.0 (152.0–220.0)	< 0.001
Glucose, mmol/L	5.2 (5.0–5.5)	5.8 (5.0–6.7)	0.046
ALT, U/L	20.0 (19.0–22.0)	35.0 (19.0–51.0)	0.011
AST, U/L	19.0 (17.6–19.3)	31.0 (29.0–38.0)	< 0.001
HSI	43.6 (40.0–46.3)	45.7 (39.9–51.4)	0.361
APRI	0.2 (0.2–0.2)	0.5 (0.5–0.6)	< 0.001
FIB4	0.7 (0.7–0.8)	1.5 (1.3–1.7)	< 0.001

Note: p is the significance level for the Mann–Whitney test.

 

Four factors that significantly affect the probability of fibrosis were selected: platelet count, glucose level, AST, and ALT. The predicted risk of liver fibrosis formation was calculated using the formula:

Y = 0.4258 – 0.0038 · X₁ + 0.0567 · X₂ +
+ 0.0350 · X₃ – 0.0110 · X₄ ,

where Y (rounded to the nearest whole number) is the predicted risk of liver fibrosis in patients with NAFLD and obesity in postmenopause, X₁ is the platelet count 10⁹/L, X₂ is the glucose level, mmol/L, X₃ is the AST level, U/L, X₄ – is the ALT level, U/L. If Y is equal to or greater than one, the presence of a risk of liver fibrosis formation in women with NAFLD and obesity in postmenopause is predicted. If Y is less than one, the absence of a risk of fibrosis formation in women with NAFLD and obesity in postmenopause is predicted.

The model is effective (multiple correlation coefficient R = 0.8221; the proportion of the influence of the factors included in the model is R² · 100 = 67.7 %) and statistically significant (F-test = 21.5159; p < 0.0001). The diagnostic sensitivity of the proposed method for predicting the risk of fibrosis is 100 %, the diagnostic specificity is 97.1 %, with a reproducibility rate of 91.7 % and an agreement rate of 97.8 %.

Previously, other authors proposed a number of models for assessing the risk of liver fibrosis progression in patients with NAFLD. For example, T. S. Krolevets et al. (2018) proposed a regression mathematical model that includes parameters of waist circumference, levels of soluble leptin receptors, and leptin levels in the blood serum with an expected fibrosis risk indicator: 1 for low and 2 for high [15]. Other authors also proposed an equation for assessing the risk of fibrosis in patients with NAFLD and obesity by determining HOMA-IR index and resistin levels, with a sensitivity of 100% and specificity of 78.9 % [1]. The use of the APRI index for determining severe fibrosis and liver cirrhosis class was also proposed [16].

Our method for predicting the risk of fibrosis in NAFLD includes accessible laboratory tests, has good diagnostic characteristics, no “grey zone”, and is designed for a risk group (postmenopausal women).

Clinical case 1

Patient S., 56 years old. Menopause for 5 years. Has obesity of the 1st degree. On examination: height 155 cm, body weight 83 kg; body mass index 34.5 kg/m². The diagnosis of NAFLD was confirmed by ultrasound. Laboratory examination: platelets — 152 × 10⁹/L, glucose — 7.4 mmol/L, AST — 31 U/L, ALT — 56 U/L, APRI index — 0.5, FIB4 index — 1.53. Liver fibrosis was confirmed by elastography — F1 = 6.9 kPa (on the Metavir scale). Next, the predicted risk of liver fibrosis formation was calculated: Y = 0.4258 – – 0.0038 · 152 + 0.0567 · 7.4 + 0.0350 · 31 – – 0.0110 · 56 = 0.7, rounded to the nearest whole number, – 1. The result obtained showed that the patient is at risk of further progression of liver fibrosis.

Clinical case 2

Patient K., 52 years old. Menopause for 2 years. Has obesity of the 1st degree. On examination: height 165 cm, body weight 83 kg; body mass index 30.5 kg/m². The diagnosis of NAFLD was confirmed by ultrasound. Laboratory examination: platelets — 368 · 10⁹/L, glucose — 5.42 mmol/L, AST — 19 U/L, ALT — 20 U/L, APRI index — 0.12, FIB4 index — 0.47. Next, the predicted risk of liver fibrosis formation was calculated: Y = 0.4258 – – 0.0038 · 368 + 0.0567 · 5.42 + 0.0350 · 19 – – 0.0110 · 20 = – 0.22, rounded to the nearest whole number, – 0. The result obtained showed that the patient has a low risk of liver fibrosis.

Conclusions

1. The risk of developing liver fibrosis in women with NAFLD and obesity in postmenopause increases with age.
2. A method has been developed for predicting the risk of liver fibrosis in women with NAFLD and obesity in postmenopause using publicly available laboratory markers: platelets, glucose, AST, and ALT. The method has good diagnostic characteristics with no “grey zone”. If the index value is more than one, the presence of a risk of liver fibrosis in postmenopausal women with steatosis is predicted (specificity — 97.1 %); if the value is less than one, a low risk of fibrosis is predicted (sensitivity — 100 %).
3. The use of the proposed method will expand the possibilities of minimally invasive diagnostics, ensure earlier detection of the risk of developing liver fibrosis in this risk group, which will determine further management tactics for these patients.

About the authors

I. A. Bulatova

Ye.A. Vagner Perm State Medical University

Author for correspondence.
Email: bula.1977@mail.ru
ORCID iD: 0000-0002-7802-4796

DSc (Medicine), Head of the Department of Normal Physiology, Professor of the Department of Faculty Therapy № 2, Occupational Pathology and Clinical Laboratory Diagnostics

Russian Federation, Perm

T. P. Shevlyukovа

Tyumen State Medical University

Email: bula.1977@mail.ru
ORCID iD: 0000-0002-7019-6630

DSc (Medicine), Professor of the Department of Obstetrics and Gynecology of the Institute of Maternity and Childhood

Russian Federation, Tyumen

A. A. Sobol

Ye.A. Vagner Perm State Medical University; Women's Health Clinic

Email: bula.1977@mail.ru

Laboratory Assistant of the Department of Pathologic Physiology, Therapist, Gastroenterologist

Russian Federation, Perm; Perm

I. L. Gulyaeva

Ye.A. Vagner Perm State Medical University

Email: bula.1977@mail.ru
ORCID iD: 0000-0001-7521-1732

DSc (Medicine), Head of the Department of Pathologic Physiology

Russian Federation, Perm

V. S. Sheludko

Ye.A. Vagner Perm State Medical University

Email: bula.1977@mail.ru

PhD (Medicine), Leading Researcher

Russian Federation, Perm

References

Supplementary files