Symptomatic seizures in preterm newborns: a review on clinical features and prognosis

Spag­no­li et al, tal­ian Jour­nal of Pediatrics2018

Neona­tal seizures are the most com­mon neu­ro­log­i­cal event in new­borns, show­ing high­er preva­lence in preterm than in full-term infants. In the major­i­ty of cas­es they rep­re­sent acute symp­to­matic phe­nom­e­na, the main eti­olo­gies being intra­ven­tric­u­lar haem­or­rhage, hypox­ic-ischemic encephalopa­thy, cen­tral ner­vous sys­tem infec­tions and tran­sient meta­bol­ic derange­ments.

Cur­rent def­i­n­i­tion of neona­tal seizures requires detec­tion of parox­ys­mal EEG-changes, and in preterm new­borns the inci­dence of elec­tro­graph­ic-only seizures seems to be par­tic­u­lar­ly high, fur­ther stress­ing the cru­cial role of elec­troen­cephalo­gram mon­i­tor­ing in this pop­u­la­tion. Imag­ing work-up includes an inte­gra­tion of ser­i­al cra­nial ultra­sound and brain mag­net­ic res­o­nance at term-equiv­a­lent age. Unfavourable out­comes fol­low­ing seizures in preterm infants include death, neu­rode­vel­op­men­tal impair­ment, epilep­sy, cere­bral pal­sy, hear­ing and visu­al impair­ment. As exper­i­men­tal evi­dence sug­gests a detri­men­tal role of seizures per se in deter­min­ing sub­se­quent out­come, they should be prompt­ly treat­ed with the aim to reduce seizure bur­den and long-term dis­abil­i­ties. How­ev­er, neona­tal seizures show low response to con­ven­tion­al anti­con­vul­sant drugs, and this is even more evi­dent in preterm new­borns, due to intrin­sic devel­op­men­tal fac­tors. As a con­se­quence, as lit­er­a­ture does not pro­vide any spe­cif­ic guide­lines, due to the lack of robust evi­dence, off-label med­ica­tions are often admin­is­tered in clin­i­cal prac­tice.

Neona­tal seizures in new­borns are clin­i­cal­ly defined as abnor­mal, stereo­typed and parox­ys­mal dys­func­tions in the cen­tral ner­vous sys­tem, occur­ring with­in 44 weeks of ges­ta­tion­al age.

From a clin­i­cal point of view, neona­tal seizures are divid­ed into clonic, myoclonic, ton­ic and sub­tle seizures. Clonic seizures con­sist of repet­i­tive, rhyth­mic (1–4/s) con­trac­tions of mus­cle groups of the limbs, face or trunk. They can be focal or mul­ti­fo­cal. They are usu­al­ly elec­tri­cal­ly-con­firmed. Myoclonic seizures are iso­lat­ed or repet­i­tive con­trac­tions of prox­i­mal or dis­tal mus­cles, less reg­u­lar and per­sis­tent than clonic jerks. They can be gen­er­al­ized, focal or mul­ti­fo­cal. Ton­ic seizures are clas­si­fied as gen­er­al­ized or focal. Focal ton­ic seizures con­sist of sus­tained, but tran­sient, asym­met­ri­cal pos­tur­ing of the trunk or extrem­i­ties or ton­ic eye devi­a­tion, while gen­er­al­ized ton­ic seizures are bilat­er­al sym­met­ri­cal ton­ic pos­tur­ing (with flexor/extensor pre­dom­i­nance or mixed). Elec­tro­graph­ic cor­re­late is incon­sis­tent for both seizure types. Sub­tle seizures include motor automa­tisms (such as chew­ing, swal­low­ing, suck­ing, repet­i­tive tongue move­ments, “cycling”, “box­ing”, “ped­al­ing”, “swim­ming”) and auto­nom­ic signs (changes in heart rate or breath­ing pat­tern, flush­ing, sali­va­tion, pupil dilata­tion).

Neu­ro­phys­i­o­log­i­cal­ly a seizure is defined as the pres­ence of a parox­ys­mal EEG change with a clear onset and ter­mi­na­tion, last­ing at least 10 s, and an evo­lu­tion in fre­quen­cy, mor­phol­o­gy and ampli­tude.

As above said, the preva­lence of neona­tal seizures is high­er in preterm (22.2%) than in full-term new­borns (0.5%) but stud­ies about clin­i­cal­ly-defined seizures show a larg­er per­cent­age of neona­tal seizures than EEG-based stud­ies, espe­cial­ly aEEG-based ones. For exam­ple, a prospec­tive study about infants born before the 32nd week of ges­ta­tion­al age found a 5% inci­dence of neona­tal seizures. In con­trast, 11.9% inci­dence has been report­ed in stud­ies on clin­i­cal­ly-diag­nosed seizures.

Clin­i­cal-only seizures (with­out any elec­tro­graph­ic evi­dence), are usu­al­ly con­sid­ered non-epilep­tic and relat­ed to the loss of cor­ti­cal inhi­bi­tion on sub­cor­ti­cal struc­tures.

Parox­ys­mal non-epilep­tic motor phe­nom­e­na should be tak­en into con­sid­er­a­tion in the dif­fer­en­tial diag­no­sis.

No gen­er­al con­sen­sus has been achieved on the def­i­n­i­tion of neona­tal sta­tus epilep­ti­cus. A com­mon­ly adopt­ed def­i­n­i­tion con­sid­ers it as con­tin­u­ous seizure activ­i­ty last­ing ≥30 min or the occur­rence of recur­rent seizures ≥50% of the record­ing time (1–3 h)