The present diagnostic criteria of NDPH (of ICHD-II) exclude the prominent migrainous features in NDPH. There were very few case series of NDPH before the introduction of ICHD criteria of NDPH [1, 7, 12]. However, these all case series had patients with prominent migrainous features. Even after the development of ICHD criteria for NDPH, most authors bypassed it and used broader set of criteria which includes migrainous features [2, 3, 6, 8, 9, 11]. In this retrospective study, we examined the patients fulfilling the Kung et al.’s criteria for NDPH, which allows to include migrainous features. Our study is probably the third largest case series on NDPH (after Robin et al.’s and Peng et al.’s observations).
In our cohort, about two-thirds of patients (65 %) fulfilled the IHS criteria for NDPH. Twenty-two patients (35 %) had prominent migrainous features (NDPH-mf). In similar observation in other larger cohorts, 34–44 % patients had fulfilled the criteria for NDPH-ICHD [2, 3, 9]. Epidemiological and clinical features of these two groups were substantially similar, except the presence of migrainous features in NDPH-mf, which were the differentiating features between two. There were statistically significant differences for unilateral pain, nausea, vomiting, photophobia, and phonophobia.
Rozen  reviewed the literature for the presence of migrainous features in patients with NDPH. The prevalence for nausea in patients with NDPH was 33–68 %. Our 49 % patients reported nausea. Takese et al.  used the strict ICHD criteria to include NDPH patients and in that series photophobia was noted by only 3 % patients, while none of the patients had phonophobia. However, in other series both photophobia and phonophobia were common (photophobia 27–66 % and phonophobia 17–61 %). The prevalence of photophobia (33 %) and phonophobia (19 %) in our case series was toward the lower side of the existing range.
In original description of NDPH (Vanast case series) , patients did not have a prior history of headache. However, most other series on NDPH had patients with a prior history of headache. Robbins et al.  included only those patients who had headache frequency of <4 per month. In Peng et al. series , headache frequency was ≤1/month. In our observation, more than half (54 %) had a history of episodic headache (≤1/month). Comparison between patients with NDPH with past history of episodic tension-type headache and NDPH with past history of migraine was done (Table 2). Migrainous features and cranial autonomic features were more common in NDPH with the past history of migraine. Although, none of these features were statistically significant, such observations (high migrainous features and cranial autonomic features in patients with a past history of episodic migraine) have not been observed previously in any case series. This observation indicates that NDPH patients with a past history of episodic migraine may have more migrainous features. There may be several explanations for this association. As migrainous features may be the part of NDPH, this co-association is a normal phenomenon. This might also be because of superimposed attacks of migraine attacks in these patients. Although, no patients had escalation of frequency of migraine just prior to NDPH onset, we cannot rule out the possibility of abrupt transition of episodic migraine into chronic migraine in these patients, as about 20–30 % patients with CDH may have history of abrupt transition from episodic migraine into chronic migraine [12, 14]. There is no guideline to differentiate NDPH with migrainous features to chronic (daily) migraine with a history of abrupt transition from episodic headache to chronic headache.
We reviewed the literature to delineate the interrelation between the prior history of headache and migrainous features (nausea, photophobia, phonophobia). Most studies included migrainous features in their criteria. Prior history of headaches was noted between 25 and 38 % [2, 3, 7, 8] in these case series. Only one large series strictly followed ICHD criteria (excluding migrainous features) . Nausea was reported by 33 % patients. However, photophobia was noted by only 3 % patients. None of the patients had phonophobia. The prior history of headache was noted in only 7 % patients in that series. Therefore, a possibility exists that a patient with a past history of episodic headache might have more migrainous symptoms. Moreover, patients with a past history of episodic migraine might have attacks of migraine even on the background of NDPH. This suggests that the presence of prominent migrainous features should be judged cautiously in patients with a past history of episodic headache (especially migraine).
Therapeutic responses and prognosis
The first description of NDPH considered this entity as a “benign or self-limiting” form of headache ; however, most other observations considered it as the most refractory headache disorder. A review of the literature suggests that most studies were done on the patients who had headache duration of more than 6 months. It may be the reason for getting more refractory form of NDPH . It is observed that CDH with daily pain is more treatment refractory than CDH without daily pain . Patients with NDPH have ‘daily and continuous’ headache from the onset. Therefore, a possibility to become refractory to treatment is more with NDPH than any other CDH. A recent study done on patients with shorter duration (median 5 months) had demonstrated a relatively good prognosis . Mean duration of illness in our case series was 27.5 months. However, our patients showed favorable out come. 37 % patients showed “excellent” response (no or less than 1 headache per month). Another 30 % patients had “good” response (>50 % reduction in headache frequency or days per month). Only 14 % patients had poor response. We categorized the patient in three groups to assess the response to treatment: NDPH (3–6 months), NDPH (>6–24 months), and NDPH (>24 months). We compared the response to treatment between the groups. Although, none of the values reach the statistical significance, patients with shorter duration (at the time of first visit) had better outcome. Excellent response was more than two times higher in patients with a history of less than 6 months in comparison to patients with history of >2 years (58 vs. 26 %). Although, a possibility of self-limiting form of NDPH cannot be ruled out in patients with NDPH of shorter duration (<6 months), our observations indicate that intervention in early stage could prevent chronification. However, a number of other explanations are also plausible. A possibility of self-limiting form is more in patients with headache of shorter duration. In the same line, chance of being refractory to the treatment is more in patients with history of longer duration. These all may be reasons for the good responses in patients with shorter duration.
There is no specific treatment strategy for NDPH [5, 17]. On this issue Rozan  writes “Most headache specialists will treat NDPH with the same acute and preventive medications that they use to treat chronic migraine….” Unfortunately, there is no formal evidence-based recommendation for optimal therapy in patients with CDH. Most published guidelines are the personal experience of the authors or the summary of the available evidences. Probably it is the reason for heterogeneity in treatment of patients with NDPH in most case series (including ours).
A few authors suggest that aggressive intravenous therapy for intractable CDH is more cost- and time-effective mode of treatment . Krusz suggests that intravenous therapy may help in breaking a long, unremitting cycle of CDH. Intravenous sodium valproate is one of the safest intravenous drugs used for the various headache disorders and its effects have been demonstrated in both intractable acute attack and chronic daily headaches [19–21]. In one open-label study, intravenous valproate was effective in a number of primary headache disorders (including CDH) . Improvement in headache was noted by 80 % patients with CDH by intravenous valproate in another open-label study .
Various studies have demonstrated the beneficial effects of injectable steroids on acute intractable attack of migraine and other headache disorders. Anecdotal evidences suggest that intravenous steroid for the short term may effective even in CDH patients [17, 18]. A recent meta-analysis suggests that addition of steroid to the standard therapy for the management of acute migraine headache may decrease the incidence of the recurrence of headaches . The dose and duration of steroids are highly variable. Most authors used steroids for a few days to about 2 weeks. Bonuccelli et al.  gave injectable dexamethasone (with amitriptyline) for 2 weeks in patients with CDH. Trucco et al.  gave intravenous dexamethasone (8 mg daily) (with other drugs) for 7–15 days. The authors  suggest that a higher dose or intravenous route (of steroids) may be required for patients with CDH/medication overused headaches.
Intravenous MPS is generally considered as safe as majorities of the side effects are transient and self-limiting and do not require specific treatment .
Our observation showed a lower percentage of medication overuse (13 %). This observation was similar to that of Kung et al.’s case series  where medication overuse was noted in only in 8.7 % patients. However, a few recent observations reported higher proportion of medication overuse in patients with NDPH. Robbins et al.’s  reported medication overuse in about 45 % patients. In Peng et al.’s series , medication overuse was noted in about 35 % patients with NDPH. The factors predicting development of the medication overused headache have not been explored in the literature. However, the presence of such wide variability of prevalence of medication overuse headache in patients with NDPH suggests multifactorial involvement.
Chronic daily headache has been linked to various comorbid psychiatric conditions. Associated psychiatry disorders can lead to poor response to therapy. However, there are very few studies in patients with NDPH in which psychiatric evaluation was performed. Robbins et al.’s reported self-reported anxiety and self-reported depression in about one-third of the patients. In Peng et al. case series , psychiatry comorbidities were noted in more than half of the patients. However, in both case series, associated psychiatric comorbidities were not related with the outcomes. Detail psychiatric evaluations were not done in our patients. Self-reported depression and anxiety were noted by a few patients in our case series (19 and 16 % respectively) (Table 1). No association was noted between clinical outcomes and self-reported depression and anxiety. However, as comorbid psychiatry disorders are known to have a poor response to therapy, better outcome in our case series may be because of low rates of comorbid depression and anxiety.
The etiology of NDPH is poorly defined. About 40–60 % patients recognize triggers at the onset of the headache [2, 3, 5, 12]. Our 56 % patients had recognized triggers at the onset of NDPH. Besides infectious etiology (29 %), other triggers in our patients were injury, stress, surgery, and post partum state. Injury was reported as a trigger by seven (11 %) patients. Three patients had mild head trauma. A possibility of chronic post-traumatic headache attributed to mild head injury (ICHD-II code 5.2.2)  exists in these patients. However, it was less likely as none of the patients had symptoms and/or signs suggestive of concussion (a must to fulfill the criteria of 5.2.2) . Although most case series did not report head trauma as a trigger, it was the second most common inciting factor (23 %) in Mack’s series . Surgery and stress are two important triggers for the development of NDPH. We think injury or trauma may be analogous to surgery. Moreover, injury itself may produce stress (another trigger for NDPH). Most case series excludes the injury as triggers for NDPH. However, our case series and Mack’s observation suggest that injury may be a trigger for the development of NDPH, and mechanisms responsible for it may be entirely different from that of post-traumatic headache. Though statistically not significant, patients with recognized triggers showed better outcome. A study with larger sample of patients is required to confirm the observations.
Ethnic and geographical variation should also be taken into consideration for such type of observation, especially for post infectious variant, as infectious agents vary with the environmental factors. All patients received drugs before reporting to us. Therefore, we cannot rule out a possibility of delayed response of previously used drugs in a few patients. Therefore, our observations should be judged very cautiously.
Limitation of the study
It is a retrospective study and possibilities of unrecognized selection bias and recall bias exist. There was a low recall of the headache onset day compared to other NDPH series. Therefore, a possibility of recall bias about onset of headache may be high in our case series. In addition, headache management and treatment were not standardized. Besides these, we cannot rule out even the possibility of other cause of headache (secondary), as full evaluation for secondary headache was not done (contrast enhanced MRI and MRV were not done in each patient). The patients in our study were seen in an adult tertiary neurology clinic. Therefore, our observations cannot be generalized as our sample of patients may not truly represent patients with new onset headaches due to referral and other biases. Despite a number of limitations in the study, the positive outcome suggests that further research is necessary.