Amyotrophic lateral sclerosis displays substantial clinical heterogeneity that continuously challenges neurologists and clinical trialists worldwide. Consequently, investigators search for robust predictive tools to stratify patients accurately. The TRICALS risk score provides a phenotype-based survival prediction algorithm that integrates key clinical features. Meanwhile, fluid biomarkers reflect ongoing axonal degradation and disease activity. Investigators recently evaluated whether clinical phenotypes can replace invasive cerebrospinal fluid biomarkers during therapeutic trials. Specifically, researchers examined data from the MIROCALS trial to test whether clinical scores can substitute for neurofilament measurements. Their rigorous analysis confirmed that clinical models cannot replace biological quantification. However, combining both modalities yields valuable complementary prognostic information.

Understanding the TRICALS Risk Score in Clinical Trials

Amyotrophic lateral sclerosis presents variable progression rates, sites of onset, and survival trajectories. Therefore, therapeutic clinical trials require rigorous stratification to prevent baseline imbalances. The TRICALS risk score addresses this challenge by generating an individualized survival prediction for each patient. Specifically, this validated algorithm incorporates diagnostic delay, age at symptom onset, site of onset, baseline functional decline, and respiratory capacity. Clinicians value this framework because it relies entirely on non-invasive clinical evaluations from routine visits. Consequently, trial protocols frequently adopt this composite metric to harmonize inclusion criteria and balance treatment cohorts. However, clinical phenotypes represent macroscopic outcomes of cellular damage rather than direct measures of molecular injury. While functional disability unfolds over months, axonal destruction occurs continuously at microscopic levels. Thus, phenotyping models provide an indirect approximation of the disease process. Although clinical scores establish useful prognostic expectations, researchers needed to verify whether they could replace fluid biomarkers. Recent evidence clearly indicates that phenotypic assessment captures only an incomplete fraction of the active neurodegenerative cascade. Therefore, clinical scores require biological supplementation during therapeutic evaluation.

The Biological Significance of Neurofilament Biomarkers

Neurofilaments provide structural scaffolding within motor neurons and their long axons. Consequently, when motor axons degenerate, neurofilament fragments leak into the cerebrospinal fluid and circulation. Cerebrospinal fluid concentrations of phosphorylated neurofilament heavy chain directly reflect the biological tempo of axonal destruction. Furthermore, higher neurofilament levels strongly correlate with accelerated functional decline and shortened survival. In the MIROCALS trial, baseline cerebrospinal fluid neurofilaments served as a pivotal prognostic biomarker. In addition, the molecular marker demonstrated a profound statistical interaction with interleukin-2 treatment response. Specifically, participants with elevated baseline neurofilaments experienced substantial mortality reduction under the experimental regimen. Because neurofilaments directly quantify axonal damage, they capture therapeutic responses targeting neuroinflammation. In contrast, clinical assessment scores often lag behind rapid cellular changes. Consequently, molecular quantification provides immediate biological insight that outward physical signs cannot convey. Direct fluid measurements remain irreplaceable when tracking active motor neurodegeneration across clinical studies. Furthermore, these structural proteins offer objective indices that avoid subjectivity inherent in functional questionnaires. Thus, biological markers provide unprecedented mechanistic clarity during early trial phases. Therefore, trialists rely on fluid biomarkers to uncover true therapeutic efficacy.

Model Substitution and Statistical Performance Discrepancies

To test whether clinical scores can eliminate invasive lumbar punctures, researchers analyzed parameter substitution. Specifically, investigators substituted the TRICALS risk score for cerebrospinal fluid phosphorylated neurofilament heavy chain in survival models. However, replacing the fluid biomarker caused model performance to deteriorate sharply. The negative two log likelihood increased dramatically from 793.36 in the biomarker model to 928.25 in the substituted model. In addition, investigators evaluated whether the clinical score reproduced the treatment interaction identified by neurofilaments. Although the interaction between clinical risk and treatment reached statistical significance, its contribution to model fit was modest. The clinical interaction produced a chi-squared change of only 4.76, whereas the neurofilament interaction yielded 14.24. Furthermore, the original biomarker model revealed a pronounced survival benefit with a hazard ratio of 0.28. These statistical outcomes indicate that phenotypic metrics lack the sensitivity needed to duplicate biomarker stratification. Consequently, relying exclusively on phenotypic scores blunts critical signals of therapeutic efficacy. Moreover, researchers risk missing genuine drug effects if they discard sensitive molecular indicators. Thus, direct fluid quantification preserves statistical power in heterogeneous patient populations. Therefore, statistical substitution fails to capture essential biological responses during interventional drug trials.

Shared Variance and the Unseen Biology of ALS

Correlation analyses provide clear mathematical reasons for the performance gap between clinical models and fluid biomarkers. Specifically, the TRICALS risk score and cerebrospinal fluid neurofilament levels exhibited only a moderate correlation of 0.357. Consequently, the clinical score accounted for merely 12.7 percent of the shared variance in neurofilament concentrations. This finding confirms that over 87 percent of neurofilament variability remains independent of routine clinical parameters. Therefore, two patients presenting with identical clinical characteristics can harbor markedly different rates of active axonal damage. For example, individuals with matching age, onset site, and functional scores frequently experience divergent cellular injury rates. Consequently, clinical assessment alone cannot determine whether motor neurons are degenerating rapidly or slowly. Because neurofilaments reflect real-time axonal destruction, they expose biological processes that clinical examinations cannot detect. Furthermore, this hidden variance explains why clinical trials relying solely on clinical staging often show mixed outcomes. Thus, omitting molecular biomarkers leaves crucial pathophysiological events completely unmeasured. Additionally, direct biochemical assays expose occult disease dynamics long before outward physical changes manifest. Hence, fluid biomarkers provide indispensable diagnostic resolution. In conclusion, clinicians cannot infer molecular disease activity from functional scores alone.

Synergistic Integration of Phenotypes and Molecular Markers

Rather than viewing clinical scores and molecular biomarkers as competing methods, researchers should recognize their synergistic potential. The study demonstrated that adding the clinical score to the biomarker model enhances overall prognostic performance. Specifically, baseline clinical variables reflect general functional reserve, systemic resilience, and disease duration. Meanwhile, fluid neurofilaments index the immediate biological aggressiveness of motor neuron injury. Consequently, integrating both tools yields comprehensive prognostic accuracy for clinical trial stratification. When trialists evaluate experimental interventions, clinical scores assist in balancing baseline functional severity across study cohorts. Concurrently, biomarker stratification identifies biological subgroups most likely to demonstrate significant therapeutic benefit. This dual approach establishes a pragmatic blueprint for precision neurology in neurodegenerative diseases. Furthermore, combined survival models improve statistical efficiency and reduce required participant numbers. Therefore, clinical research teams can design more informative trials while minimizing unnecessary patient burden. Moreover, this cooperative framework helps neurologists optimize therapeutic selection for individual patients. In addition, future investigation will likely extend this multimodal strategy to other neurodegenerative disorders. Consequently, integrated profiling represents the optimal path forward for complex clinical trials. Thus, uniting phenotype and biology advances clinical research toward personalized medicine.

Can the TRICALS risk score replace lumbar puncture for ALS trial stratification?

No, the clinical risk score cannot replace neurofilament testing. While the clinical score predicts survival based on clinical phenotypes, it explains only 12.7 percent of neurofilament variance. Cerebrospinal fluid biomarkers directly capture ongoing axonal destruction, providing critical mechanistic insight and stratification power that clinical scoring systems alone cannot deliver.

Why did neurofilament biomarkers demonstrate a superior treatment interaction?

Neurofilaments directly quantify active motor axonal destruction and cellular response to disease mechanisms. In the MIROCALS trial, low-dose interleukin-2 modulated immune-mediated neuroinflammation, which directly influenced axonal injury rates. Because the fluid biomarker reflects this precise biological process, it demonstrated a far stronger statistical interaction with therapeutic efficacy than outward clinical phenotype.

How should clinicians integrate phenotypic scoring with fluid biomarker testing?

Clinicians and researchers should combine phenotypic scores and fluid biomarkers rather than selecting one over the other. The clinical score captures overall functional reserve and disease progression rate. Meanwhile, neurofilaments measure real-time axonal damage. Combining both tools enhances prognostic accuracy, balances clinical trial arms, and enables targeted patient stratification.

Disclaimer: This content is for informational and educational purposes only... Refer to the latest local and national guidelines for clinical practice.

References

1. Al Khleifat A et al. TRICALS risk score augments but cannot replace neurofilament as a prognostic biomarker in ALS. J Neurol. 2026 Oct 06. doi: 10.1007/s00415-026-14040-4. PMID: 42836993.
2. van Eijk RPA, Westeneng HJ, Nikolakopoulos S, et al. Refining clinical trial design in amyotrophic lateral sclerosis: the TRICALS risk score. Lancet Neurol. 2020;19(9):788-796.
3. Benatar M, Wuu J, Andersen PM, et al. Neurofilaments as biomarkers in amyotrophic lateral sclerosis. Curr Opin Neurol. 2021;34(5):715-722.