Part 2 of 2. Click here for Part 1.
With interest blooming in blood NfL as an easily accessed marker of neurodegeneration, disease severity, and prognosis (see Part 1 of this story), researchers are looking for practical applications. One study presented at the Alzheimer’s Association International Conference, held July 22–26 in Chicago, suggests that NfL could be used to stage candidate participants for early stage treatment trials. Another study is establishing NfL as a treatment marker in multiple sclerosis trials. Importantly, a study of cardiac surgery indicates that the protein could guide surgeons in developing more brain-sparing operating room techniques.
- Blood NfL reflects neuronal damage in the central nervous system.
- Will this improve enrollment and tracking of patients in clinical trials?
- Spiking NfL levels after heart surgery suggest current procedures put brain at risk.
Holly Soares of AbbVie, Inc., North Chicago, Illinois, probed how plasma NfL tests might help sponsors screen for clinical trials in early AD as a way to shorten recruitment time and identify the people most likely to progress in the short term. In Soares’ experience, recruiting presymptomatic trial participants comes with a failure rate between 70 and 80 percent. For example, Bristol-Myer Squibb’s Phase 2 trial of avagacestat screened 1,358 people to enroll 287; A4 screened 4,486 people with amyloid PET to enroll 1,169 amyloid-positive cognitively normal participants (Aug 2018 conference news). Could blood T-tau or NfL more quickly or cheaply enrich for people likely to have brain amyloid deposition?
In a biomarker study, Soares recruited 25 people with AD, 45 with MCI, and 25 healthy controls who were all screened for amyloid positivity with CSF and PET. While CSF p-tau and T-tau were elevated in the amyloid-positive people, Soares detected no difference in plasma tau. Plasma NFL was elevated in AD and in people with MCI who were amyloid-positive, but not people with MCI who were amyloid-negative. However, the extensive overlap in absolute plasma NfL values between the groups suggested, to Soares’ mind, that NfL would not help screen for brain amyloid positivity.
On the other hand, plasma NfL was useful to enrich for fast progressors, a strategy to reduce the number of people needed to enroll in a clinical trial. Analyzing data from 574 ADNI participants, Soares compared baseline plasma NfL and total tau to cognitive decline on the CDR-Sum of Boxes over the following 90 months. People with MCI who were ApoE4 positive and had high plasma NfL and T-tau declined the fastest.
Those markers could reduce the necessary size of trials significantly, Soares showed. For example, a hypothetical study to detect a treatment effect in one or two years would require 159 people per arm based on existing recruitment protocols, but could run with 35 per arm by enrolling high NfL/high T-tau/ApoE4-positive participants with MCI. The numbers got even better when including Aβ information. “NfL is very promising,” Soares concluded. “It’s probably telling us who is in an active stage and likely to decline, while blood Aβ reflects underlying amyloid brain pathology.”
What does total tau signify? Soares said that’s “still an open question. The two are telling us different things, and we’ll need more than one blood biomarker to help enrich populations,” she said.
Another use of serum or plasma NfL might be to track treatment effects of neuroprotective drugs. Most of this work has been done in multiple sclerosis, and it offers promise for other diseases. In people with MS, serum NfL concentration correlates tightly with brain lesions and disease activity (Barro et al., 2018), and normalize after treatment (Disanto et al., 2017; Novakova et al., 2017). At AAIC, Charlotte Teunissen, VU University, Amsterdam, showed additional unpublished data from the Kuhle lab of a trial evaluating the MS drug fingolimod. In that trial, two years of treatment drastically decreased elevated plasma NfL, back down to the level of healthy subjects.
Hoping that this kind of success will translate to Alzheimer’s, researchers are now aiming to standardize blood NfL assays between labs. They are also trying to define both clinical reference values and what constitutes a meaningful change in patients. To do that, Teunissen described an ongoing, multisite analytical validation of the Simoa NfL assay in serum from MS patients. Scientists at 17 centers in North America and Europe will compare Simoa and ELISA according to a standardized protocol and common materials.
Teunissen stressed that variability in NfL between people, even in control groups, is an important issue scientists don’t yet understand. Age and gender contribute: “We see enormous variability by age, which leads to steady increase in NfL levels. We also see gender differences, which only emerge if you look at large cohorts,” she said.
What can AD researchers learn from MS? On the plus side, the magnitude of the rise in NfL is similar in MS and AD, Teunissen said. However, the MS field has highly effective treatments, and patients with the highest NfL levels show the strongest treatment response. “If NfL works best in patients with highest levels, we might be less optimistic that we can identify treatment effects in AD trials,” Teunissen said. Nonetheless, she noted, “Some AD patients have quite high NfL levels, so we may be able to see effects in that subset.”
In pediatric spinal muscular atrophy and neuropathy due to HIV infection, effective treatment also lowered NfL, according to data presented by Henrik Zetterberg, who works at both the University of Gothenburg and University College London. Seeing these dynamic changes gives Zetterberg hope for using NfL as a marker of treatment efficacy in AD. Zetterberg is so bullish on NfL that he told Alzforum, “I believe a drug against AD that does not reduce NfL has a low chance of being effective. Some would say I put too much faith in one biomarker, but time will tell.”
By conducting studies in boxers and other athletes, Kaj Blennow, University of Gothenburg, Sweden, has laid much of the groundwork for understanding how NfL rises and falls in response to acute brain injury. Now, he is turning his attention to probing the known link between heart surgery and postoperative cognitive decline. Surgical procedures are associated with delirium and dementia in some older patients, especially in people who start out with low CSF Aβ42, the marker for brain amyloid (Evered et al., 2016). Clinicians are not sure why. Could general anesthesia be damaging the brain?
Possibly so. Recently, Blennow and colleagues reported that surgical procedures in older patients cause an increase in plasma NfL and total tau (Evered et al., 2018). In the study, patients older than 60 who were undergoing major surgery saw their total tau peak at 2.5 times baseline six hours after surgery, and then start to decline; their NfL rose less, and more slowly, and was still at its maximum at 48 hours. Blennow said this is similar to the pattern seen in acute brain injury. The results suggest that general anesthesia during surgery may be associated with neuronal damage in the short term.
In the 2018 study, three out of four patients got hip or knee replacements. At AAIC, Blennow presented a follow-up study on 26 heart surgery patients, and reported that it had even more dire consequences. The surgery caused a spike in total tau that peaked at 20 times normal levels during surgery. NfL also spiked 20-fold, but did so a bit later, five days after surgery. Blennow detected no such a spike in 25 non-cardiac controls who underwent head or neck procedures.
NfL levels rose whether or not surgeons used a heart-lung machine to circulate blood, but they shot up significantly higher—20-fold versus eightfold—for patients on the pump than for those off the pump. These results indicate that unexpected variables may be putting the brain at particular risk during cardiac surgery. Above all, they suggest cardiac surgeons may be well advised to use blood biomarkers to optimize their procedures and to know how their patient’s brains are doing, Blennow said.
Do these surges in NfL and tau foreshadow postoperative cognitive decline? “That’s a key question,” Blennow said. “In our second study, we have acute cognitive assessment, and have preliminary results that there is a correlation. However, the real problems occur after three to six months, and we don’t have that data yet. We’ll collect it this year,” he said.
Attempts to publish this data in a journal read by surgeons have thus far been unsuccessful, Blennow added.—Pat McCaffrey
- Blood Test for Neurofilament Light Chain Kicks Up Biomarker Research
- Memory Slips as Soon as Amyloid Appears, Two Decades Before Dementia
- Barro C, Benkert P, Disanto G, Tsagkas C, Amann M, Naegelin Y, Leppert D, Gobbi C, Granziera C, Yaldizli Ö, Michalak Z, Wuerfel J, Kappos L, Parmar K, Kuhle J. Serum neurofilament as a predictor of disease worsening and brain and spinal cord atrophy in multiple sclerosis. Brain. 2018 May 30; PubMed.
- Disanto G, Barro C, Benkert P, Naegelin Y, Schädelin S, Giardiello A, Zecca C, Blennow K, Zetterberg H, Leppert D, Kappos L, Gobbi C, Kuhle J, Swiss Multiple Sclerosis Cohort Study Group. Serum Neurofilament light: A biomarker of neuronal damage in multiple sclerosis. Ann Neurol. 2017 Jun;81(6):857-870. PubMed.
- Novakova L, Zetterberg H, Sundström P, Axelsson M, Khademi M, Gunnarsson M, Malmeström C, Svenningsson A, Olsson T, Piehl F, Blennow K, Lycke J. Monitoring disease activity in multiple sclerosis using serum neurofilament light protein. Neurology. 2017 Nov 28;89(22):2230-2237. Epub 2017 Oct 27 PubMed.
- Evered L, Silbert B, Scott DA, Ames D, Maruff P, Blennow K. Cerebrospinal Fluid Biomarker for Alzheimer Disease Predicts Postoperative Cognitive Dysfunction. Anesthesiology. 2016 Feb;124(2):353-61. PubMed.
- Evered L, Silbert B, Scott DA, Zetterberg H, Blennow K. Association of Changes in Plasma Neurofilament Light and Tau Levels With Anesthesia and Surgery: Results From the CAPACITY and ARCADIAN Studies. JAMA Neurol. 2018 May 1;75(5):542-547. PubMed.
No Available Further Reading