Biomarker Panel Predicts Slow or Fast ALS Course

A diagnosis of amyotrophic lateral sclerosis typically means a person will survive for only two or three years—but some people live a decade or more. Patients, and their physicians, could benefit from a method that would help them predict how much time the person has left to live. In the December JAMA Neurology, researchers from the Pennsylvania State College of Medicine in Hershey present a step forward in the form of a panel of inflammatory biomarkers in both blood and cerebrospinal fluid (CSF). By analyzing half a dozen biomolecules, the researchers correctly predicted disease duration in 94 percent of cases in a small cohort. “Nothing is 100 percent certain,” cautioned senior author James Connor, “but a test like this could help people understand if their disease is likely to move quickly or slowly.”

ALS researchers have been searching urgently for biomarkers for the diagnosis and prognosis. One hurdle, Connor suggested, has been the focus on single parameters. Researchers now realize that because ALS likely describes a syndrome with multiple etiologies, one biomarker may not fit all. The Hershey researchers, and others, have begun to compile lists of biomarkers that, together, could diagnose ALS, though no clinically validated test exits yet (Mitchell et al., 2009; Mitchell et al., 2010; Ranganathan et al., 2005). In the current study, they turn their attention to prognosis. Defining a panel of prognostic ALS markers, particularly by combining two different body fluids, is a relatively new approach, commented Robert Bowser of the Barrow Neurological Institute in Phoenix, who was not involved in the study.

First author Xiaowei “Bill” Su and colleagues examined plasma from 29 people and CSF from 33, all of whom had donated the fluids at the time of diagnosis and had since died. Some lived as little as nine months after diagnosis, others 16 years. The researchers genotyped the participants for the hemochromatosis gene (HFE), in which polymorphisms appear in about one-third of people with ALS, but only in one-tenth of the general population (Wang et al., 2004; Goodall et al., 2005). Too few subjects were mutation carriers to make any inference about the gene’s effect on survival. HFE encodes an iron sensor. None of the people in the study had a family history of ALS and the researchers checked no other known ALS mutations, which are quite rare.

Because the immune system is known to play a role in ALS (see Oct 2008 news story; Sep 2009 news story; Nov 2009 news story), Su and colleagues selected a commercial immunoassay kit to measure 27 cytokines and growth factors in the fluid samples. HFE variants are somewhat prevalent in ALS and Connor has a longstanding interest in this gene (see Alzforum Webinar), hence the scientists also quantified eight molecules involved in iron metabolism.

Then, Su used multivariant modeling to identify biomarker panels that would best predict disease duration. The model started with the biomarker that correlated most tightly with survival, then successively added the next-best biomarkers, re-evaluating the correlation at each turn until it identified the most predictive set. Considering the 29 plasma samples, the model identified a panel that predicted disease duration “right on” in 18 cases, Connor said. For the 33 CSF samples, the computer came up with a panel that hit the mark in 21 cases. Eighteen participants donated both plasma and CSF, and for those the model accurately predicted the survival of all but one person, who passed away sooner than predicted.

The panel included six markers; high levels of these five predicted longer survival:

  • Plasma interferon-g-inducible protein 10 (IP-10), an anti-inflammatory cytokine that suppresses macrophages and antigen-presenting cells.
  • Plasma interleukin-5, a cytokine that promotes differentiation of white blood cells.
  • Plasma L-ferritin, the light chain of the iron storage molecule ferritin.
  • CSF monocyte chemoattractant protein-1 (MCP-1), which attracts certain immune cell types and is also implicated in Alzheimer’s disease.
  • CSF:Plasma ratio of Interferon-g (IFN-g), a pro-inflammatory molecule that activates macrophages and IP-10 production.

In contrast, high CSF interleukin 8 (IL-8), a pro-inflammatory cytokine that activates neutrophils, predicted shorter lifespan.

Overall, the results indicate that an increase in pro-inflammatory molecules combined with a decrease in anti-inflammatory mediators predicts faster ALS progression, Connor said.

Plenty Left To Do
“It is an interesting paper that needs to be replicated in a much larger set of samples,” commented Merit Cudkowicz of Massachusetts General Hospital in Boston in an email to Alzforum. “It is too preliminary to be conclusive,” agreed Stanley Appel of The Methodist Hospital Research Institute in Houston. In addition, Connor and Bowser said it would be important to examine these biomarkers in a longitudinal study.

Connor and colleagues hope that a panel like this, if verified, could aid clinical trials. By comparing a volunteer’s predicted to actual survival, researchers might be better able to determine if a drug worked. Trialists might also find that patients on a slow or fast track respond differently to medicines, suggested Lucie Bruijn of the ALS Association. While this would not necessarily allow researchers to plan smaller or shorter trials, Connor said, it would enable them to analyze or reanalyze results with a clearer understanding of predicted disease duration. Notably, past data from several trials are now available to researchers (see Dec 12 news story). Appel cautioned that a biomarker based on inflammatory molecules likely would be a poor general outcome measure for all ALS clinical trials. It could be useful for medicines that specifically target that immune state (for example, see Mar 2010 news story; Feb 2012 news story; Jan 2013 news story).

In addition to traditional fluid biomarkers, scientists are testing novel measures of muscle physiology (see Oct 2011 news story) and brain imaging (see Jan 2009 news story).

Several roadblocks are slowing the development of robust biomarkers. Standardization is one challenge, noted Bowser. To identify and confirm biomarkers, investigators need large sets of samples that were all collected in a standardized manner. A biorepository managed by the Northeast ALS Consortium can help here, he noted (Sherman et al., 2011). In addition, commercial assays used in discovery, such as the kit Su worked with, typically vary between lots and are unsuitable for clinical applications, Bowser said (see Nov 2009 news series on biomarker quality control in Alzheimer’s).

Even if those difficulties are surmounted, the best proof of a biomarker would be to show it correlates with survival benefits in a successful drug trial, noted Bruijn. Catch 22: That requires a good treatment.


To view commentaries, primary articles and linked stories, go to the original posting on Alzforum.org here.

Copyright © 1996–2018 Biomedical Research Forum, LLC. All Rights Reserved.

Share this:
Facebooktwittergoogle_plusmailFacebooktwittergoogle_plusmail