Machine-learning based approach to predict anti-CGRP response in patients with migraine: multicenter Spanish study

Evaluation of the concomitant use of prophylactic treatments in patients with migraine under anti-calcitonin gene-related peptide therapies: The PREVENAC study

BACKGROUND AND PURPOSE

Anti-calcitonin gene-related peptide (CGRP) therapies are recent preventive therapies approved for both episodic and chronic migraine. One of the measures of effectiveness is the withdrawal of other preventive treatments. The objective of this study is to quantify the impact of anti-CGRP drugs in concomitant preventive treatment in patients with migraine.

METHODS

This was an observational, retrospective, multicenter cohort study with patients from nine national headache units. Patients with migraine undergoing treatment for at least 6 months with anti-CGRP antibodies, who were initially associated with some preventive treatment (oral and/or onabotulinumtoxinA) were included. Demographic and clinical variables were collected, as well as variables related to headache. Differences according to withdrawal or nonwithdrawal were evaluated.

RESULTS

A total of 408 patients were included, 86.52% women, 48.79 (SD = 1.46) years old. Preventive treatment was withdrawn in 43.87% (179/408), 20.83% partially and 23.04% totally. In 27.45% (112/408), it was maintained exclusively due to comorbidity and in 28.6% (117/408) due to partial efficacy. The most frequent time of withdrawal was between 3 and 5 months after the start of treatment. The baseline characteristics associated with nonwithdrawal were comorbidities: insomnia, hypertension and obesity, chronic migraine, and medication overuse. In the multivariate analysis, the absence of high blood pressure, a greater number of preventive treatments at the start, and a lower number of migraine days/month after anti-CGRP treatment were independently associated with withdrawal of the treatment (p < 0.05).

CONCLUSIONS

Anti-CGRP antibodies allow the withdrawal of associated preventive treatment in a significant percentage of patients, which supports its effectiveness in real-life conditions.

A large language model-based generative natural language processing framework fine-tuned on clinical notes accurately extracts headache frequency from electronic health records

OBJECTIVE

To develop a natural language processing (NLP) algorithm that can accurately extract headache frequency from free-text clinical notes.

BACKGROUND

Headache frequency, defined as the number of days with any headache in a month (or 4 weeks), remains a key parameter in the evaluation of treatment response to migraine preventive medications. However, due to the variations and inconsistencies in documentation by clinicians, significant challenges exist to accurately extract headache frequency from the electronic health record (EHR) by traditional NLP algorithms.

METHODS

This was a retrospective cross-sectional study with patients identified from two tertiary headache referral centers, Mayo Clinic Arizona and Mayo Clinic Rochester. All neurology consultation notes written by 15 specialized clinicians (11 headache specialists and 4 nurse practitioners) between 2012 and 2022 were extracted and 1915 notes were used for model fine-tuning (90%) and testing (10%). We employed four different NLP frameworks: (1) ClinicalBERT (Bidirectional Encoder Representations from Transformers) regression model, (2) Generative Pre-Trained Transformer-2 (GPT-2) Question Answering (QA) model zero-shot, (3) GPT-2 QA model few-shot training fine-tuned on clinical notes, and (4) GPT-2 generative model few-shot training fine-tuned on clinical notes to generate the answer by considering the context of included text.

RESULTS

The mean (standard deviation) headache frequency of our training and testing datasets were 13.4 (10.9) and 14.4 (11.2), respectively. The GPT-2 generative model was the best-performing model with an accuracy of 0.92 (0.91, 0.93, 95% confidence interval [CI]) and R2 score of 0.89 (0.87, 0.90, 95% CI), and all GPT-2-based models outperformed the ClinicalBERT model in terms of exact matching accuracy. Although the ClinicalBERT regression model had the lowest accuracy of 0.27 (0.26, 0.28), it demonstrated a high R2 score of 0.88 (0.85, 0.89), suggesting the ClinicalBERT model can reasonably predict the headache frequency within a range of ≤ ± 3 days, and the R2 score was higher than the GPT-2 QA zero-shot model or GPT-2 QA model few-shot training fine-tuned model.

CONCLUSION

We developed a robust information extraction model based on a state-of-the-art large language model, a GPT-2 generative model that can extract headache frequency from EHR free-text clinical notes with high accuracy and R2 score. It overcame several challenges related to different ways clinicians document headache frequency that were not easily achieved by traditional NLP models. We also showed that GPT-2-based frameworks outperformed ClinicalBERT in terms of accuracy in extracting headache frequency from clinical notes. To facilitate research in the field, we released the GPT-2 generative model and inference code with open-source license of community use in GitHub. Additional fine-tuning of the algorithm might be required when applied to different health-care systems for various clinical use cases.

Effectiveness, tolerability, and response predictors of preventive anti-CGRP mAbs for migraine in patients over 65 years old: a multicenter real-world case-control study

OBJECTIVE

To evaluate clinical characteristics, effectiveness, and tolerability of preventive anti- calcitonin gene-related peptide (CGRP) monoclonal antibodies (mAbs) in the elderly. Anti-CGRP mAbs have demonstrated efficacy and safety in patients with migraine although there is limited information regarding the elderly.

DESIGN

We performed a multicenter case-control study of cases (patients over 65 years old) and controls (sex-matched patients under 55 years old) with migraine receiving anti-CGRP mAbs.

METHODS

We included the demographic characteristics, effectiveness-reduction in the number of monthly headache days (MHD) and monthly migraine days (MMD), 30%, 50%, and 75% responder rates-and treatment emergent adverse events (TEAEs). The primary endpoint was the 50% response rate regarding MHD at weeks 20-24; exploratory 50% response predictors in the elderly were evaluated.

RESULTS

In total, 228 patients were included: 114 cases , 114 controls-. Among cases 84.2% (96/114) were women, 79.8% (91/114) CM; mean age of cases 70.1 years old (range: 66-86); mean age of controls was 42.9 years old(range: 38-49). Cases had a higher percentage of vascular risk factors (P < .05),older age of onset (P < .001) and more reported prior preventive treatments (P < .001). Regarding effectiveness in cases, 50% response rate was achieved by 57.5% (42/73) at 20-24 weeks, with lower reduction in the MHD at 8-12 weeks (5 [7.2], 8 [9.1]; P = .001) and a higher reduction in MMD at 20-24 weeks (10.7 [9.1], 9.2 [7.7]; P = .04) compared to the control group. The percentage of TEAEs was similar in the 2 groups. Diagnosis of episodic migraine (EM) (P = .03) and lower number of MHD at baseline (P = .001) were associated with a 50% response in the elderly in univariate analysis.

CONCLUSIONS

Our study provides real world evidence of effectiveness and safety of anti-CGRP mAbs for migraine in patients without upper age-limit and possible predictors of anti-CGRP response in the elderly.

The Clinical Relevance of Artificial Intelligence in Migraine

Migraine is a burdensome neurological disorder that still lacks clear and easily accessible diagnostic biomarkers. Furthermore, a straightforward pathway is hard to find for migraineurs' management, so the search for response predictors has become urgent. Nowadays, artificial intelligence (AI) has pervaded almost every aspect of our lives, and medicine has not been missed. Its applications are nearly limitless, and the ability to use machine learning approaches has given researchers a chance to give huge amounts of data new insights. When it comes to migraine, AI may play a fundamental role, helping clinicians and patients in many ways. For example, AI-based models can increase diagnostic accuracy, especially for non-headache specialists, and may help in correctly classifying the different groups of patients. Moreover, AI models analysing brain imaging studies reveal promising results in identifying disease biomarkers. Regarding migraine management, AI applications showed value in identifying outcome measures, the best treatment choices, and therapy response prediction. In the present review, the authors introduce the various and most recent clinical applications of AI regarding migraine.

New management strategies for primary headache disorders: Insights from P4 medicine

Primary headache disorder is the main cause of headache attacks, leading to significant disability and impaired quality of life. This disorder is increasingly recognized as a heterogeneous condition with a complex network of genetic, environmental, and lifestyle factors. However, the timely diagnosis and effective treatment of these headaches remain challenging. Precision medicine is a potential strategy based on P4 (predictive, preventive, personalized, and participatory) medicine that may bring new insights for headache care. Recent machine learning advances and widely available molecular biology and imaging data have increased the usefulness of this medical strategy. Precision medicine emphasizes classifying headaches according to their risk factors, clinical presentation, and therapy responsiveness to provide individualized headache management. Furthermore, early preventive strategies, mainly utilizing predictive tools, are critical in reducing headache attacks and improving the quality of life of individuals with headaches. The current review comprehensively discusses the potential application value of P4 medicine in headache management.

Predictors of galcanezumab response in a real-world study of Korean patients with migraine

To assess factors associated with galcanezumab response in a real-world study of Korean patients with migraine. Predictors of the efficacy of monoclonal antibodies targeting calcitonin gene-related peptide (CGRP) or its receptor (anti-CGRP(-R) mAb) have been rarely investigated in Asians. We prospectively recruited and followed up patients with migraine who received monthly galcanezumab treatment in a single university hospital from June 2020 to October 2021. We defined the treatment response with ≥ 50% reduction in moderate/severe headache days in the 3rd month of treatment compared to baseline. Responders and non-responders were compared in terms of demographics, disease characteristics and severity, and previous response to migraine prophylactic treatments. Potential predictors of anti-CGRP(-R) mAb response were tested by using the univariable and multivariable logistic regression analyses. Among 104 patients (81.7% female; mean age 42.0 ± 13.02; 76.9% chronic migraine; and 45.5% medication overuse headache) included, 58 (55.7%) were responders. Non-responders had more chronic migraine, medication overuse headache, monthly headache days, days with acute medication, and daily headaches (i.e. chronic migraine persisting everyday without remission). The multivariable logistic analysis showed chronic migraine (OR 0.05 [95% CI 0.00-0.82], p = 0.036) and the number of previously failed preventive medication classes (OR 0.55 [95% CI 0.33-0.92], p = 0.024] were independently associated with treatment response. Chronic migraine and multiple failures from preventive medication are associated with poor galcanezumab response. Further studies are needed to investigate if earlier treatment before disease chronification or multiple failures may lead to a greater therapeutic gain from anti-CGRP(-R) mAb treatment.

Methodological considerations on real-world evidence studies of monoclonal antibodies against the CGRP-pathway for migraine: a systematic review

BACKGROUND

Real-world data are accumulating on the effectiveness, tolerability and safety of anti-calcitonin gene-related peptide pathway monoclonal antibodies for the preventive treatment of migraine. We performed a systematic review of the methodology of prospective, observational, clinic-based real-world evidence studies with these drugs in both episodic and chronic migraine.

METHODS

The objectives were to evaluate the definitions and reported outcomes used, and to perform a risk of bias assessment for each of the different studies. PubMed and EMBASE were systematically queried for relevant scientific articles. Study quality assessment of the included studies was conducted using the "National Heart, Lung and Blood Institute (NHLBI) Study Quality Assessment Tool for Before-After (Pre-Post) Studies with No Control Group".

RESULTS

Forty-six studies fitted the criteria for the systematic review and were included in the analysis. Ten studies (21.7%) defined a migraine day for the study, while only 5 studies defined a headache day for the study (10.9%). The most common primary endpoint/objective of the studies was change in monthly migraine days (n = 16, 34.8%), followed by responder rate (n = 15, 32.6%) and change in monthly headache days (n = 5, 10.9%). Eight studies (17.4%) did not define the primary endpoint/objective. Thirty-three studies were graded as "good" quality and 13 studies were graded as "fair".

CONCLUSION

Our analysis shows rather significant heterogeneity and/or lack of predefined primary outcomes/objectives, definitions of outcomes measures and the use of longitudinal monitoring (e.g. headache diaries). Standardization of terminology, definitions and protocol procedures for real-world evidence studies of preventive treatments for migraine are recommended.

TRIAL REGISTRATION

This study was registered with PROSPERO with ID CRD42022369366.

Searching for the Predictors of Response to BoNT-A in Migraine Using Machine Learning Approaches

OnabotulinumtoxinA (BonT-A) reduces migraine frequency in a considerable portion of patients with migraine. So far, predictive characteristics of response are lacking. Here, we applied machine learning (ML) algorithms to identify clinical characteristics able to predict treatment response. We collected demographic and clinical data of patients with chronic migraine (CM) or high-frequency episodic migraine (HFEM) treated with BoNT-A at our clinic in the last 5 years. Patients received BoNT-A according to the PREEMPT (Phase III Research Evaluating Migraine Prophylaxis Therapy) paradigm and were classified according to the monthly migraine days reduction in the 12 weeks after the fourth BoNT-A cycle, as compared to baseline. Data were used as input features to run ML algorithms. Of the 212 patients enrolled, 35 qualified as excellent responders to BoNT-A administration and 38 as nonresponders. None of the anamnestic characteristics were able to discriminate responders from nonresponders in the CM group. Nevertheless, a pattern of four features (age at onset of migraine, opioid use, anxiety subscore at the hospital anxiety and depression scale (HADS-a) and Migraine Disability Assessment (MIDAS) score correctly predicted response in HFEM. Our findings suggest that routine anamnestic features acquired in real-life settings cannot accurately predict BoNT-A response in migraine and call for a more complex modality of patient profiling.

Cluster and network analysis of non-headache symptoms in migraine patients reveals distinct subgroups based on onset age and vestibular-cochlear symptom interconnection

Objective

The present study endeavors to identify natural subgroups of migraine patients based on the patterns of non-headache symptoms, utilizing cluster analysis. Subsequently, network analysis was performed to estimate the structure of symptoms and explore the potential pathophysiology of these findings.

Method

A total of 475 patients who met the diagnostic criteria for migraine were surveyed face-to-face during the period of 2019 to 2022. The survey included collecting demographic and symptom data. Four different solutions were generated by the K-means for mixed large data (KAMILA) clustering algorithm, from which the final cluster solutions were selected based on a series of cluster metrics. Subsequently, we performed network analysis using Bayesian Gaussian graphical models (BGGM) to estimate the symptom structure across subgroups and conducted global and pairwise comparisons between structures.

Result

Cluster analysis identified two distinct patient groups, and the onset age of migraine proved to be an effective characteristic differentiating the two patient groups. Participants assigned to late-onset group showed a longer course of migraine, higher frequency of monthly headache attacks, and greater tendency toward medication overuse. In contrast, patients in early-onset group exhibited a higher frequency of nausea, vomiting, and phonophobia compared to their counterparts in the other group. The network analysis revealed a different symptom structure between the two groups globally, while the pairwise differences indicated an increasing connection between tinnitus and dizziness, and a decreasing connection between tinnitus and hearing loss in the early-onset group.

Conclusion

Utilizing clustering and network analysis, we have identified two distinct non-headache symptom structures of migraine patients with early-onset age and late-onset age. Our findings suggest that the vestibular-cochlear symptoms may differ in the context of different onset ages of migraine patients, which may contribute to a better understanding of the pathology of vestibular-cochlear symptoms in migraine.