AT ILC 2017

AMSTERDAM (FRONTLINE MEDICAL NEWS) – Determining the ratio of neutrophils to leukocytes in the blood could help identify patients with alcoholic hepatitis that would and would not benefit from steroid treatment.

Patients who had a neutrophil to lymphocyte ratio (NLR) of between 5 and 8 before being treated with the corticosteroid prednisolone appeared to obtain a benefit versus no-steroid treatment (P = .007) while those with higher and lower NLR values did not, in an analysis presented at the International Liver Congress.

If patients with only an NLR of 5-8 were treated with prednisolone, then this would give a survival of 84.6% at roughly 3 months, compared with 88.1% for patients with an NLR of less than 5 and 57.4% for patients with an NLR above 8 (P less than .0001).

This could potentially help clinicians avoid putting some patients through a futile trial of steroid therapy, study author Ewan H. Forrest, MD , explained in an interview at the meeting, which is sponsored by the European Association for the Study of the Liver (EASL).

“The traditional approach would be to give steroids to patients with severe alcoholic hepatitis, wait 7 days, see if they are getting better, and if so, keep them on the steroids,” Dr. Forrest of the liver unit at Glasgow Royal Infirmary observed. Conversely, if patients are not doing better then steroids should be stopped.

“What we are increasingly aware of is that not only do some people not do well with steroids but also they actually do considerably more badly,” Dr. Forrest cautioned.

Usually, the response to steroid treatment in alcoholic hepatitis is measured by changes in serum bilirubin after a week of treatment, but this, of course, exposes patients to a “futile course of treatment with a risk of complication such as sepsis,” Dr. Forrest and his coauthors noted in a a late-breaking poster.

Determining the NLR has already been shown to help predict the prognosis of patients with several diseases with an underlying inflammatory component, such as cardiovascular diseases and several types of cancer. It also has proven useful in patients with liver disease, although not specifically in alcoholic hepatitis before this study, Dr. Forrest observed.

Data on patients with alcoholic hepatitis who had participated in the multicenter, double-blind, randomized STOPAH trial were used to see if the baseline NLR could help stratify patients who would benefit from steroid therapy.

STOPAH had compared the use of prednisolone or pentoxifylline for the treatment of alcoholic hepatitis but found no benefit for the latter, although there was a possible benefit of steroids for improving overall survival, at least in the short term ( N Engl J Med. 2015 Apr;372:1619-28 ).

Dr. Forrest noted that measurement of the lymphocyte count was not part of the original study design, so data to calculate the NLR were obtained retrospectively. As there had been little or no response to pentoxifylline in the trial, patients who had taken this drug were regarded as having had no treatment in the analysis.

In all, baseline NLR values could be worked out for 630 patients from the STOPAH trial, but 113 were excluded from further analysis as they met the prespecified exclusion criteria of gastrointestinal bleeding or sepsis.

Overall, a NLR of 5 or less, indicating milder liver disease, was associated with significantly better survival at 3 months than if the NLR was 5 or more (85.5% vs. 67.3%; P less than .0001), study findings suggested.

Dr. Forest noted that 29% of patients fell into the “sweet spot” of the NLR of between 5 and 8, where patients did benefit from steroids, but that the 23% of study subjects who had an NLR ratio above 8 did not. These patients may have had disease too severe to benefit from the prednisolone, he suggested, and tended to have a worse prognosis regardless. A baseline NLR greater than 8 was associated with acute kidney infection but not sepsis, the team found.

There was also no great effect of the steroid in the 48% of patients who had an NLR ratio less than 5, suggesting that maybe they had disease that was too mild to warrant such treatment and did well regardless.

Of course, these findings still need further validation, but they are “not far off” from clinical application, Dr. Forrest offered. Calculating the ratio is simple, can be done during a routine whole-blood cell count, and is potentially cost saving because it reduces the standard practice in the United Kingdom of giving “all-comers” 7 days of corticosteroid therapy as a trial to see if they get better, he said.

Dr. Forrest had no conflicts of interest to disclose.