New daily persistent headache: a start with an uncertain end

As headache neurologists, it is a common experience to meet a new patient with headache that began on a particular day several months or years ago, and which has, unfortunately, been present ever since. This presentation may be encountered even more often in pediatric headache clinics, as new daily persistent headache (NDPH) may have a predilection for onset in adolescence¹. The International Classification of Headache Disorders, 3^rd edition (ICHD-3) diagnostic criteria for NDPH specify that the headache had a “distinct and clearly remembered onset”, became continuous within 24 hours of starting, has been present for a minimum of 3 months and is not better ascribed to another primary or secondary headache disorder². The headache phenotype may resemble chronic migraine or chronic tension-type headache. While prospective studies on NDPH are limited, it is typical for the headache to last for years, no matter what types of treatment are given³, at least among patients coming for care in a subspecialty headache clinic. To the lay person, and even to many medical professionals, the notion that a person could have one spontaneous, continuous headache for years may seem preposterous. As an invisible illness, with no demonstrable laboratory or imaging biomarker, NDPH is a set up for patients to experience social stigma and can be truly life shattering.

Despite its impact, we know shockingly little about NDPH. For example, what are its causes? NDPH is a syndrome with presumably multiple possible etiologies and inciting factors⁴, and there is no consensus on which secondary causes of headache must be ruled out before diagnosis. For those who have a pre-existing history of infrequent migraine, is it simply a migraine attack that started in some unusual way, or for some reason failed to “turn off”? For those in whom NDPH began in the setting of a systemic viral illness such as Epstein Barr Virus or COVID-19^5–7, is there an inflammatory or auto-immune trigger responsible? It is unclear whether focusing on the inciting event at NDPH onset, or on the NDPH symptoms (i.e., the “phenotype”) may provide the most useful clues towards clarifying the underlying pathophysiology and effective treatments for this condition. A common practice is for clinicians to “treat the headache phenotype”, i.e., to use migraine therapies if the NDPH phenotype is migrainous, and tension-type headache therapies (which practically speaking may overlap with migraine therapies) if it is featureless. However, with neither biomarkers, neuroimaging data, an understanding of underlying pathophysiology, nor any randomized clinical trials to inform treatment decisions, we are practicing in an evidence free zone.

Experts may not even agree as to what it means for headache to be “new”. For example, if prior to onset of the continuous headache, the patient experienced migraine attacks at a frequency of once per year with no attack acceleration pattern, can a diagnosis of NDPH be made? Chronic migraine would certainly have been the diagnosis if their headache frequency had built up to daily gradually rather than suddenly. What if their pre-existing migraine frequency was once a month, or once a week? Where is the line in the sand? Normally ICHD diagnoses start with expert consensus and then successively refine as our pathophysiologic understanding of a headache disorder advances. However, with NDPH our understanding is still so much in its infancy that even the basics of who is eligible to be diagnosed with NDPH is a topic of debate.

What it means for NDPH to improve is even less well agreed upon, and this is a major barrier to selecting outcome measures for clinical trials. In migraine treatment trials, migraine days/month or headache days/month are common primary outcome measures. With NDPH, these outcomes are impractical as, at least for studies of moderate duration (3–6 months being typical for migraine trials), chances are that in a population with NDPH the number of headache days will be “30 out of 30” at the start of the trial and still 30 at its end, as few patients remit in the short-to-medium term³. Clinically, simply lowering the intensity of baseline headache can be seen as a major therapeutic accomplishment in NDPH, given how hard it can be to make any progress with this intransigent disorder. Should we therefore make the number of moderate/severe headache days/month the outcome measure of choice for NDPH trials? A small survey study of adolescents who were experiencing continuous headache due to chronic migraine found this to be the outcome measure that they and their parents valued most⁸, but it is unknown if this preference generalizes to those with NDPH or to adults. Measures of headache-related disability or quality of life might be more highly valued. Another option would be to use the Patient Global Impression of Change (PGIC)⁹, which asks patients how much better or worse they feel on a treatment, and mirrors what clinicians often aim to assess subjectively at follow-up visits. However, perhaps as clinical researchers we tend to aim too low. Should long duration trials that aim to identify whether a treatment shortens the time to first headache free moment, or first headache free day, even if that takes years to attain, be the goal? If so, what would appropriate comparator treatments be, as having a placebo arm for such a long trial would likely be unethical and make enrollment more challenging? Clearly patients and their caregivers should be the ones to answer the question of what matters most in NDPH treatment. Qualitative and/or quantitative research that delves into which patient-reported outcomes are most important to this patient population is urgently needed. Complacency about poor outcomes in NDPH cannot be accepted, and we must strive to set measurable, practical, and meaningful outcomes to assess the benefits of current and future therapies. It’s time to get persistent.