Liver Stiffness by Transient Elastography Correlates With Degree of Portal Hypertension in Common Variable Immunodeficiency Patients With Nodular Regenerative Hyperplasia

Abstract

Nodular regenerative hyperplasia (NRH) is associated with high morbidity and mortality in patients with common variable immunodeficiency (CVID). While liver biopsy is the gold standard for NRH diagnosis, a non-invasive technique could facilitate early disease recognition, monitoring, and/or immune intervention. We performed a cross-sectional analysis of ultrasound-based transient elastography (TE) in patients with CVID to evaluate liver stiffness and compared this between patients with (N = 12) and without (N = 6) biopsy-proven NRH. Additionally, these data were compared to a cohort followed at our institution for non-alcoholic fatty liver disease (NAFLD) (N = 527), a disease for which TE has routine diagnostic use. Clinical and pathologic features of NRH were evaluated as correlates of liver stiffness, and receiver operating characteristic curves were used to define a liver stiffness cutoff with diagnostic utility for NRH among CVID patients. CVID patients with NRH had a more severe disease presentation compared to those without. This included increased autoinflammatory disease comorbidities, combined B-cell and T-cell dysfunction, and abnormal liver biochemistries (specifically an increased mean alkaline phosphatase level [proximal to TE, 250 vs. 100 U/L; p = 0.03; peak, 314 vs. 114 U/L; p = 0.02). Results of TE demonstrated a significantly elevated liver stiffness in CVID patients with NRH (mean 13.2 ± 6.2 kPa) as compared to both CVID patients without NRH (mean 4.6 ± 0.9 kPa) and non-CVID patients with NAFLD (mean 6.9 ± 5.5 kPa) (p < 0.01). No single or composite histopathologic feature of NRH correlated with liver stiffness including nodule size, nodule density, sinusoidal dilation, fibrosis, and/or lymphocytosis. In contrast, liver stiffness by TE was significantly correlated with clinical parameters of portal hypertension, including an elevated hepatic venous pressure gradient, an increased splenic longitudinal diameter, presence of varices, and presence of peripheral edema. A liver stiffness of greater than or equal to 6.2 kPa was a clinically significant cutoff for NRH in CVID patients. We propose that TE has diagnostic utility in CVID, particularly in the presence of immunophenotypic features such as combined B-cell and T-cell dysfunction, autoinflammatory comorbidities, and/or abnormal liver tests. Elevated liver stiffness by TE should raise suspicion for NRH in patients with CVID and prompt expedited evaluation by hepatology.

Keywords: common variable immunodeficiency (CVID); fibroscan©; liver biopsy; liver disease; nodular regenerative hyperplasia (NRH); transient elastography (TE).

Figure 1

Liver stiffness by transient elastography is significantly elevated in CVID patients with NRH. Liver stiffness measurements (kPa) by transient elastography shown as mean (± SD) in CVID patients with biopsy-proven NRH (CVID + NRH, N = 12), CVID patients without NRH (CVID, N = 6), and non-CVID patients with non-alcoholic fatty liver disease (NAFLD, N = 527). Significance by one-way ANOVA with Tukey’s post-hoc correction; *, p = 0.01; **, p < 0.01. Red dotted line indicates a diagnostic cutoff value for liver stiffness of 6.2 kPa (defined using the ROC curve in Figure 4 ). CVID, common variable immunodeficiency; NRH, nodular regenerative hyperplasia; ROC, receiver operating characteristic.

Figure 2

Liver stiffness by transient elastography does not correlate with specific histopathologic features of NRH in CVID patients. Liver stiffness measurements (kPa) by transient elastography compared across histopathologic features of NRH in CVID patients with available liver biopsy (N = 7), including size of largest nodule (A), nodule density (B), centrilobular fibrosis (C), and sinusoidal lymphocytosis (D). Significance by one-way ANOVA with Tukey’s post-hoc correction or Spearman’s correlation (r) with significance (p) shown. Line of best fit (A, B) and mean value (C, D) are shown. NRH, nodular regenerative hyperplasia; CVID, common variable immunodeficiency.

Figure 3

Liver stiffness by transient elastography correlates with clinical parameters of portal hypertension in CVID patients. Liver stiffness measurements (kPa) by transient elastography compared across clinical parameters of portal hypertension in CVID patients, including splenic longitudinal diameter (A, N = 13 scored), presence of varices (grade 1–3) (B, N = 18 scored), presence of peripheral edema (C, N = 18 scored), elevated (>10 mmHg) hepatic venous pressure gradient (HVPG) (D, N = 8 scored), and clinically diagnosed portal hypertension in the electronic medical record based on any combination of these data (E, N = 18 scored). Significance by one-way ANOVA with Tukey’s post-hoc correction or Spearman’s correlation (r) with significance (p) shown. Line of best fit (A) and mean value (B–E) are shown. CVID, common variable immunodeficiency.

Figure 4

Receiver operating characteristic (ROC) curves for the diagnosis of NRH in CVID patients. ROC curves for (A) liver stiffness by transient elastography, (B) alkaline phosphatase (ALP) level in peripheral blood most proximal to the time of transient elastography, and (C) peak ALP level in peripheral blood, excluding acute illness, to diagnose NRH in patients with CVID. AUC, area under the curve; NRH, nodular regenerative hyperplasia; CVID, common variable immunodeficiency.

Figure 5

Clinical algorithm for early detection of nodular regenerative hyperplasia in individuals with CVID. CVID, common variable immunodeficiency; ULN, upper limit normal.