Mathew et al. [5
] in 1982 were the first to call attention to the possibility that migraine may be transformed into daily headache over time.
Five years later, the same group of researchers [6
] reported a case series of 630 CDH patients and attempted to subdivide them into three different groups depending on their headache type: “Type I starts as daily or near-daily headache with no change in the severity and lacks migrainous features; Type II starts as daily or near-daily headaches with occasional more severe headache with some migrainous features; Type III (transformed or evolutive migraine) starts as a clear-cut occasional episodic migraine …. with increasing frequency over the next many years ….. evolving into chronic daily headaches”. As many as 489 of the 630 CDH patients of Mathew et al. (77.7%) belonged to Type III, while only 57 (9.0%) and 84 (13.3%) belonged to Type II and Type I, respectively. These authors in 1987 were not yet talking about chronic tension-type headache (CTTH), nor about new daily persistent headache (NDPH), a form of primary headache first described just 1 year earlier by Vanast [11
]. Both were still unknown definitions, even though the two first CDH types could seemingly be identified with these two headache forms. However, back then they already suggested the term “transformed or evolutive migraine” for the third type.
] himself in 1993 definitely chose the name “transformed migraine” (TM), which he included in the first true CDH classification, alongside CTTH and NDPH (Table ).
Classifications of chronic daily headache introduced in the mid-1990s
The following year, Silberstein, together with Lipton, Solomon and Mathew again [13
], proposed that hemicrania continua (HC) should be added to these three forms and that each of the four forms thus identified should be distinguished depending on the presence or absence of medication overuse (Table ). They also set precise diagnostic criteria for each form. TM criteria are reported in Table .
Table 2 Diagnostic criteria for transformed migraine by Silberstein et al. 
In 1995, following the description of a broad case series of CDH patients seen at the Parma and Pavia headache centres in Italy, Manzoni et al. [14
] first introduced the name “chronic migraine” (CM), which they included, alongside migraine with interparoxysmal headache (MIH), within the migraine forms that evolve unfavourably over time until they lose the typical symptom-free interval between an attack and the next (Table ). According to the Italian authors, CM and MIH differentiate from each other for the type of headache that sets in the originally free intervals between attacks: while retaining the clinical features of migraine in CM, in MIH this interval headache loses its similarities to migraine: in some cases it has the same features of tension-type headache, but in other cases it has undefined features and this prevents it from being definitely included in the migraine or the tension-type headache group.
The proposals introduced in the mid-1990s by the US authors [13
] and the Italian authors [14
] differ from each other in several respects. First, they use different approaches to identify CDH entities: while Silberstein et al. [13
] adopt a more detailed and exhaustive classification that includes also NDPH and HC, Manzoni et al. [14
] prefer a classification that takes into account only the forms most frequently encountered in clinical practice. Secondly, and more importantly, the differences between them concern: (a) the terminology used to describe migraine forms that evolve unfavourably over time (TM for Silberstein and CM for Manzoni); (b) the temporal requirements for these headache forms (presence of headache at least 15 days per month for at least 1 month for Silberstein and at least 6 days per week for at least 1 year for Manzoni); and (c) Manzoni et al.’s attempt to identify possible clinical subtypes in the group of migraine forms evolving unfavourably (CM and MIH with interval headache with or without tension-type headache features).
If we browse the scientific CDH literature produced from the mid-1990s to as late as 2006, we can see how predominant Silberstein et al.’s systematization became. Almost all studies conducted over this long period of time and aimed at defining the epidemiological, pathogenetic and therapeutic aspects of CDH basically followed Silberstein et al.’s classification [13
], both in terms of headache forms included in the classification, and in terms of their respective diagnostic criteria and the terminology proposed to give an official name to each form. In this connection, it is worth noting that there was a general and widespread acceptance of the TM name.
On the other hand, we are sorry to say that the 2004 edition of the International Classification of Headache Disorders classification (ICHD-2) [15
] only partially—and not always appropriately—integrated those considerations in its final changes over the 1988 first edition of the IHS classification [4
In the first place, the ICHD-2 [15
] editors did not deem it advisable to devote a separate chapter to CDH within their classification—which, all things considered, is actually a decision we can agree on.
Secondly, of the four different CDH forms identified by Silberstein et al. [13
] (Table ), only one, CTTH, was already included in the 1988 first edition of the IHS classification [4
] and then again in the 2004 ICHD-2 classification [15
]. Two other forms, HC and NDPH, did not appear in the IHS classification [4
] but were included in the ICHD-2 classification [15
], where they were coded to Group 4 “Other primary headaches”. The last of Silberstein et al.’s [13
] CDH forms, TM, was not recognized as such in the ICHD-2 [15
], which however included for the first time CM and its diagnostic criteria, coded to 1.5.1 as a complication of migraine (Table ). Thus, with respect to the most important and certainly most frequent of all CDH forms, i.e. migraine evolving unfavourably over time, the ICHD-2 [15
] eventually retained the CM name, originally proposed by Manzoni et al. [14
], but with diagnostic criteria (>15 days per month for at least 3 months) that are very different from those suggested by the Italian authors (at least 6 day per week for at least 1 year). In addition, the ICHD-2 [15
] ignored the MIH proposed by the same authors [14
]. The result is that only for some of the patients affected by this clinical entity—precisely those with an interval headache resembling tension-type headache—can the ICHD-2 [15
] provide a diagnosis, but even so it would be a dual diagnosis, of migraine without aura and of tension-type headache. MIH patients with an interval headache not resembling either migraine or tension-type headache remain unclassified.
Diagnostic criteria for chronic migraine by the ICHD-2 (2004) and by the ICHD-2R (2006)
With respect to Silberstein et al.’s classification [13
], the ICHD-2 classification [15
] neither recognizes the TM definition nor agrees with its diagnostic criteria (Tables , ). Additionally, the ICHD-2 [15
] does not envisage, either for CM or even for CTTH, HC and NDPH, a differentiation based on the presence or absence of medication overuse, because it prefers to retain Medication-overuse headache as an autonomous clinical entity (coded to 8.2 of the 2004 classification), much as the 1988 IHS classification [4
] had already done.
As defined by the diagnostic criteria of the ICHD-2 classification [15
], CM seems: (a) to resemble more a high-frequency migraine than a migraine evolving over time to a daily or near-daily form and eventually losing some, and in certain cases, many of its typical migraine features, such as unilateral and/or throbbing pain and/or nausea and vomiting as accompanying symptoms; (b) to be scarcely relevant to actual clinical practice, because a patient with more than 15 days of headache per month is highly unlikely to use symptomatic drugs for less than 10–15 days per month.
Since its inclusion in the ICHD-2 classification [15
], then, CM has always appeared as an ambiguous clinical entity and one that would not be of much use either for clinical practice or research.
Bigal et al. [16
] in the US tried to re-classify 638 CDH patients seen over a period of 20 years at the New England Medical Center for Headache in Stamford, Connecticut. By strictly applying the diagnostic criteria of the ICHD-2 [15
], they managed to establish a diagnosis of CM only in nine cases. In contrast, using the diagnostic criteria proposed by Silberstein et al. [13
], the number of patients that in the same case series could be classified as suffering from TM without medication overuse would be as high as 158. Therefore, the authors concluded that CM diagnosis, as defined and formally applied in the ICHD-2 [15
], does not offer any considerable benefit to the CDH diagnostic issue.
Another major shortcoming of the ICHD-2 classification [15
] is that, as was clearly demonstrated in the same analysis by Bigal et al. [16
], most CDH patients receive multiple diagnoses, in several cases as many as four or five, including some that are only probable. Since the very introduction of the ICHD-2 [15
] in 2004, then, it has been clear that the diagnostic criteria of CM needed to be changed.