UK-PBC Risk Score

UK-PBC model explained

Primary biliary cholangitis (PBC) is a type of chronic liver disease in which there is autoimmune destruction of the intrahepatic bile ducts. This in turn leads to cholestasis and progressive fibrosis.

Biliary injuries of this kind may end up in cirrhosis and liver failure requiring transplant.

However, rate of progression is variable and whilst some patients reach end stage liver disease (ESLD) within years of diagnosis, there are others in which the condition remains compensated.

Models like the UK-PBC aim to predict chances of adverse outcome in patients who are under treatment with the only licensed pharmacotherapy for PBC (by FDA), which is ursodeoxycholic acid (UDCA) (recommended 13-15 mg/kg/day dose).

The UK-PBC Research Cohort has fed data from thousands of patients from specialist centers across the entire UK, another reason why the model is backed by a highly representative cohort.

The UK-PBC risk score may be used to identify high-risk patients and adapt current therapy with potential second-line therapies. The score is recognized by the EASL PBC Guideline and with the development of recent therapies, such as that with obeticholic acid, should be used in the 1-year evaluation of patients to discover if the patient would benefit from a different therapy line.

The variables used in the PBC-UK model are:

■ 12 month: bilirubin, AST (aspartate aminotransferase) or ALT (alanine aminotransferase) and alkaline phosphatase (considered with upper limit of normal);

■ Baseline: albumin and platelets (considered with lower limit of normal).

The UK-PBC score predicts risk of liver failure at 5, 10 respectively 15 years via:

UK-PBC = 1-baseline survival function ^ exp(0.0287854 * (alp12-xuln-1.722136304) - 0.0422873 * (((altast12xuln/10) ^-1) - 8.675729006) + 1.4199 * (ln (bil12xuln/10) + 2.709607778) - 1.960303 * (albxlln-1.17673001) - 0.4161954 * (pltxlln - 1.873564875))

Where:

Baseline survivor function is:

■ 0.982 (at 5 years);

■ 0.941 (at 10 years);

■ 0.893 (at 15 years).

And:

■ alp12-xuln is ratio of Alkalin phosphatase measurement at 12 month and its upper limit of normal;

■ altast12xuln is ratio of either ALT or AST measurement at 12 month and its upper limit of normal;

■ bil12xuln is ratio of bilirubin measurement at 12 month and its upper limit of normal;

■ albxlln is ratio of baseline albumin measurement and its lower limit of normal;

■ pltxlln is ratio of baseline platelet measurement and its lower limit of normal.

Thus, the model provides estimates of the risk of developing ESLD within defined time points in the future.

It was also found that this scoring system may be accurate in evaluating long-term risk also in patients who are not under UDCA treatment. However, more research is required into how the UK-PBC can be used to support evaluation of patients before and during treatment.

About the original study

In 2016, Carbone et al. used data from the UK-PBC Research Cohort to construct a model which could predict long term outcome in patients suffering from primary biliary cholangitis and under treatment with ursodeoxycholic acid.

The prognostic model aimed to bypass the limitations of previous treatment response models and to provide risk of adverse outcome at 5, 10 and 15 years. The study involved a derivation cohort of 1,916 UDCA-treated patients.

Amongst the variables considered for inclusion were: patient age at PBC diagnosis, patient gender, blood test results (albumin, sodium, creatinine, bilirubin), PT, INR, immunoglobulins, spleen size or liver biochemistry.

Patients with additional chronic liver diseases (aside form PBC) were excluded and so were participants with PBC/autoimmune hepatitis (AIH) overlap syndrome. Patients not treated with UDCA for 12 months continuously or who had their treatment discontinued were also excluded.

Nonautomatic backward selection was used to derive the best-fitting Cox model, from which a multivariable fractional polynomial model was derived.

The model showed superior AUC values in derivation: 0.96, 0.95, and 0.94 for the 5-, 10-, and 15-year risk scores, respectively.

The model has been validated in a cohort of 1,249 UDCA-treated participants. All three predictive values achieve high prediction accuracies, with areas under the curse greater than 0.90.

Original source

Carbone M, Sharp SJ, Flack S, Paximadas D, Spiess K, Adgey C, Griffiths L, Lim R, Trembling P, Williamson K, Wareham NJ, Aldersley M, Bathgate A, Burroughs AK, Heneghan MA, Neuberger JM, Thorburn D, Hirschfield GM, Cordell HJ, Alexander GJ, Jones DE, Sandford RN, Mells GF; UK-PBC Consortium. The UK-PBC risk scores: Derivation and validation of a scoring system for long-term prediction of end-stage liver disease in primary biliary cholangitis. Hepatology. 2016; 63(3):930-50.

Article reviewed by Dr. Antonio Olveira