Introduction and aims: During last decades better perinatal care and higher survival rates for very low birth weight (<1500 gr, VLBW) infants went hand by hand with an important reduction of incidence of severe brain lesions, as periventricular leukomalacia (PVL). At the same time more widespread use of brain MRI allowed to identify milder lesions, with unknown prognosis, as punctate white matter lesions (PWML). Recently two types of PWML were described: the hemorrhagic ones (visible on SWI sequence) and the non-hemorrhagic ones (negative on SWI and visible only on T1 and T2 sequences). The first aim of this study was to describe clinical risk factors and prognostic significance of both types of PWML. The second aim was to define the type and frequency of white matter lesions when compared to other prematurity-related brain lesions, describing their influence on neurological follow-up at 3 years of age. Methods: In the first part of the work we included all VLBW patients admitted to G. Gaslini Hospital NICU from January 2012 to August 2016 that have performed brain MRI at term-equivalent age; in the second part of the work we added to the population the patients admitted up to September 2017. For each patient we registered detailed clinical data and Griffiths Scale results at 2 and 3 years of age. The PWML, defined as small alterations of signal (high T1 and/or low T2) in the periventricular white matter were divided into SWI+ PWML and SWI- PWML based on their visibility on SWI sequence. The lesions were further classified based on their number, distance from the lateral ventricle, their position and morphological characteristics (linear, cluster or single). Multivariate analysis of risk factors for different types of PWML was carried out; SWI+ PWML e SWI- PWML were compared based on their anatomical characteristics and prognostic meaning at 2 years of age. In the second part of the work we have divided prematurity-related brain lesions - white matter lesions, germinal matrix hemorrhage – intraventricular hemorrhage (GMH-IVH) and cerebellar hemorrhage (CBH) – into three grades, with scores from 1 to 3. We have then analyzed the frequency of different types of lesions in our population. We have carried out linear measurements characterizing brain size, such as biparietal width (BPW) and trans-cerebellar diameter (TCD). Values of Developmental Coefficient on Griffiths Scale at 3 years of age were compared between patients with different types and grades of lesions and patients without brain lesions; we have also explored possible correlations between linear brain measurements and outcome. Results: In the first part of the study, PWML were registered in 19% (61/321) of VLBW infants, while only 25% of them (15/61) had lesions visible on SWI (SWI+). PWML were more frequent in patients with gestational age at birth below 28 weeks. Significant risk factors for PWML were absent or incomplete antenatal steroid treatment (p<0,05) and higher birth weight (p<0,05). Patients more at risk of developing multiple (more than 6) PWML were the ones intubated (p<0,05) and with higher Apgar score at birth (p<0,05). SWI+ PWML were associated with the presence of GMH-IVH (p=0,001) and lower gestational age at birth; these lesions were also closer to the lateral ventricle when compared to non-hemorrhagic punctate lesions (p<0,001). At 2 years of corrected age, patients with isolated SWI+ PWML had lower Developmental Coefficient on scale D (eye-hand coordination and fine motricity) when compared to the patients with normal imaging. The second part of the study showed that 187 out of 408 patients (46% of all VLBW) had at least one prematurity-related brain lesion on MRI, while 37 out of 408 (9%) had severe lesions (grade 3). The most frequent lesion was GMH-IVH (28% of the population), followed by white matter lesions (PWML e PVL, 21%) and CBH (17%). The presence of brain lesions, even if of grade 1, was related to a significant reduction of Developmental Coefficient on Griffiths Scale at 3 years of age when compared to infants without lesions. We have further observed significant differences in motor outcome (scale A of Griffiths) distributed progressively among patients with three grades of white matter lesions, from PWML of grade 1 to cystic PVL. We have observed positive correlation between values of Developmental Coefficient on Griffiths Scale at 3 years of age and BPW (r=0.18, p=0.0057), as well as TCD (r=0.27, p<0.001), even if excluding the patients with the brain lesions the correlations were not any more significant. Conclusions: We have observed that hemorrhagic punctate white matter lesions (SWI+ PWML) have risk factors, anatomical position and prognostic significance that differs with those of non-hemorrhagic ones, and that their presence is connected with lower gestational ages, as is known for other hemorrhagic lesions (GMH-IVH and CBH). It could be though useful to use this distinction in clinical practice. We have also observed that even low-grade lesions of all types have a negative influence on neurological outcome at 3 year of age, underlining the importance of precise diagnostics and elaboration of preventive strategies.
Introduzione e Scopo dello Studio. Negli ultimi decenni il miglioramento delle cure perinatali e l’aumento della sopravvivenza dei nati pretermine di peso molto basso alla nascita (< 1500g, Very Low Birth Weight, VLBW) è stato accompagnato da una drastica riduzione delle lesioni cerebrali gravi, come la leucomalacia periventricolare (PVL). Nel contempo un più diffuso uso della risonanza magnetica ha permesso di identificare lesioni sempre più lievi, con significato prognostico incerto, come lesioni puntate della sostanza bianca (Punctate White Matter Lesions, PWML). Recentemente sono stati evidenziati due sottotipi di PWML: di tipo emorragico (visibili alla sequenza SWI) e non-emorragico (negative alla SWI e visibili nelle sole sequenze T1 e T2). Un primo scopo dello studio è stato quello di evidenziare i fattori di rischio clinici e il significato prognostico in termini di follow-up psicomotorio di entrambi i sottotipi di PWML. Secondariamente, si è voluto definire la tipologia e la frequenza delle lesioni di sostanza bianca in confronto alle altre lesioni cerebrali tipiche della prematurità evidenziando eventuali differenze nell’outcome neurologico a tre anni di vita. Metodi. Per la prima parte dello studio sono stati inclusi tutti i pazienti VLBW ricoverati presso la U.O. Terapia Intensiva e Patologia Neonatale dell’Istituto G. Gaslini di Genova da gennaio 2012 ad agosto 2016 sottoposti ad RM cerebrale all’età corretta del termine; nella seconda parte dello studio la popolazione era ampliata aggiungendo pazienti ricoverati fino a settembre 2017. Per ogni paziente sono stati raccolti dati clinici dettagliati e valori di scala Griffiths a 2 e 3 anni di età corretta. Le PWML, definite come lesioni puntiformi della sostanza bianca periventricolare iperintense in sequenza T1 e/o ipointense in sequenza T2, sono stati suddivise in SWI+ PWML e SWI- PWML in base alla visibilità in SWI. Le lesioni sono state inoltre classificate in base al numero di lesioni per singolo paziente, alla distanza di lesioni dal ventricolo laterale, alla loro posizione ed alle caratteristiche morfologiche (lineari, a grappolo o singole). È stata eseguita una analisi multivariata dei fattori di rischio clinici per diverse tipologie delle PWML, inoltre le SWI+ PWML e SWI- PWML sono state confrontate in base alla sede anatomica e al significato prognostico a 2 anni. Nella seconda parte dello studio abbiamo caratterizzato le lesioni cerebrali tipiche della prematurità - lesioni di sostanza bianca, emorragia intraventricolare da sanguinamento della matrice germinativa (GMH-IVH), ed emorragia cerebellare (CBH) - in funzione di gravità, con gli score da 1 a 3. Abbiamo successivamente confrontato l’incidenza di diverse tipologie di lesioni. Sono stati eseguite le misurazioni di parametri lineari caratterizzanti dimensioni cerebrali, come diametro biparietale (BPW) e diametro trans-cerebellare (TCD). È stato eseguito confronto di valori di Griffiths a 3 anni di età tra pazienti con diversi tipologie e gradi di lesioni e pazienti senza lesioni cerebrali; sono state inoltre esplorate possibili correlazioni tra le misure cerebrali e l’outcome. Risultati. Nella prima parte dello studio, le PWML si sono registrate nel 19% (61/321) dei neonati VLBW, dei quali solo il 25% (15/61) presentavano lesioni visibili nella sequenza SWI (SWI+). Le lesioni puntate sono risultate più frequenti nei nati con età gestazionale > 28 settimane. Sono risultati fattori di rischio significativi per lo sviluppo delle PWML la profilassi steroidea incompleta (p<0,05) e un peso maggiore alla nascita (p<0,05). Erano più a rischio di sviluppare PWML multiple (più di 6) pazienti che hanno subito l’intubazione (p<0,05) e che avevano un punteggio di Apgar più alto alla nascita (p<0,05). Le SWI+ PWML sono stati associate con la presenza di GMH-IVH (p=0,001) e con età gestazionale alla nascita più bassa, inoltre queste lesioni sono risultate più vicino al ventricolo laterale rispetto alle puntate non emorragiche (p<0,001). Al follow-up a 2 anni di età, pazienti con SWI+ PWML isolate presentavano la riduzione di quoziente di sviluppo su scala D (coordinazione occhio-mano e motricità fine) rispetto ai pazienti con imaging normale. Dalla analisi dei dati nella seconda parte dello studio è emerso che 187 su 408 pazienti (46% di tutti i VLBW) presentavano almeno una lesione cerebrale alla risonanza magnetica, mentre 37 su 408 (9%) presentavano lesioni gravi (grado 3). La lesione più frequente era la emorragia intraventricolare da sanguinamento della matrice germinativa (28% della popolazione), seguita da lesioni della sostanza bianca (PWML e PVL, 21%) ed infine dall’emorragia cerebellare (17%). Presenza di lesioni cerebrali, anche di grado 1, comportava diminuzione significativa dei valori medi della scala di Griffith rispetto ai pazienti che non hanno presentato alcuna lesione cerebrale. Inoltre, abbiamo osservato differenze significative nell’outcome locomotorio (scala A di Griffiths) in modo progressivo tra i pazienti con tre livelli di lesioni della sostanza bianca, da PWML di grado 1 a PVL cistica. Per quanto riguarda le misure lineari, è stata osservata la correlazione positiva tra valori di Griffiths totale a 3 anni e BPW (r=0.18, p=0.0057), così come TCD (r=0.27, p<0.001), anche se le correlazioni non erano più statisticamente significativi se si escludevano i pazienti con lesioni. Conclusioni. Abbiamo osservato che le lesioni puntate di natura emorragica (SWI+ PWML) hanno fattori di rischio, posizione anatomica e significato pronostico diversi da quelli non-emorragiche, e che la loro presenza è collegata alle età gestazionali più basse, come succede anche per altre emorragie (GMH-IVH e CBH). Potrebbe quindi essere utile utilizzare questa distinzione anche in ambito clinico. Abbiamo inoltre osservato che anche per lesioni di basso grado (sia per le emorragie che per le lesioni della sostanza bianca) esiste una correlazione negativa con outcome a 3 anni di vita, il fatto che sottolinea l’importanza di diagnostica specifica e valorizza la ricerca in ambito di prevenzione di tali lesioni.
Risonanza magnetica cerebrale e outcome neurologico nei pazienti con peso molto basso alla nascita: ruolo delle alterazioni di sostanza bianca e delle altre lesioni cerebrali connesse con la prematurità
MALOVA, MARIYA
2021-05-26
Abstract
Introduction and aims: During last decades better perinatal care and higher survival rates for very low birth weight (<1500 gr, VLBW) infants went hand by hand with an important reduction of incidence of severe brain lesions, as periventricular leukomalacia (PVL). At the same time more widespread use of brain MRI allowed to identify milder lesions, with unknown prognosis, as punctate white matter lesions (PWML). Recently two types of PWML were described: the hemorrhagic ones (visible on SWI sequence) and the non-hemorrhagic ones (negative on SWI and visible only on T1 and T2 sequences). The first aim of this study was to describe clinical risk factors and prognostic significance of both types of PWML. The second aim was to define the type and frequency of white matter lesions when compared to other prematurity-related brain lesions, describing their influence on neurological follow-up at 3 years of age. Methods: In the first part of the work we included all VLBW patients admitted to G. Gaslini Hospital NICU from January 2012 to August 2016 that have performed brain MRI at term-equivalent age; in the second part of the work we added to the population the patients admitted up to September 2017. For each patient we registered detailed clinical data and Griffiths Scale results at 2 and 3 years of age. The PWML, defined as small alterations of signal (high T1 and/or low T2) in the periventricular white matter were divided into SWI+ PWML and SWI- PWML based on their visibility on SWI sequence. The lesions were further classified based on their number, distance from the lateral ventricle, their position and morphological characteristics (linear, cluster or single). Multivariate analysis of risk factors for different types of PWML was carried out; SWI+ PWML e SWI- PWML were compared based on their anatomical characteristics and prognostic meaning at 2 years of age. In the second part of the work we have divided prematurity-related brain lesions - white matter lesions, germinal matrix hemorrhage – intraventricular hemorrhage (GMH-IVH) and cerebellar hemorrhage (CBH) – into three grades, with scores from 1 to 3. We have then analyzed the frequency of different types of lesions in our population. We have carried out linear measurements characterizing brain size, such as biparietal width (BPW) and trans-cerebellar diameter (TCD). Values of Developmental Coefficient on Griffiths Scale at 3 years of age were compared between patients with different types and grades of lesions and patients without brain lesions; we have also explored possible correlations between linear brain measurements and outcome. Results: In the first part of the study, PWML were registered in 19% (61/321) of VLBW infants, while only 25% of them (15/61) had lesions visible on SWI (SWI+). PWML were more frequent in patients with gestational age at birth below 28 weeks. Significant risk factors for PWML were absent or incomplete antenatal steroid treatment (p<0,05) and higher birth weight (p<0,05). Patients more at risk of developing multiple (more than 6) PWML were the ones intubated (p<0,05) and with higher Apgar score at birth (p<0,05). SWI+ PWML were associated with the presence of GMH-IVH (p=0,001) and lower gestational age at birth; these lesions were also closer to the lateral ventricle when compared to non-hemorrhagic punctate lesions (p<0,001). At 2 years of corrected age, patients with isolated SWI+ PWML had lower Developmental Coefficient on scale D (eye-hand coordination and fine motricity) when compared to the patients with normal imaging. The second part of the study showed that 187 out of 408 patients (46% of all VLBW) had at least one prematurity-related brain lesion on MRI, while 37 out of 408 (9%) had severe lesions (grade 3). The most frequent lesion was GMH-IVH (28% of the population), followed by white matter lesions (PWML e PVL, 21%) and CBH (17%). The presence of brain lesions, even if of grade 1, was related to a significant reduction of Developmental Coefficient on Griffiths Scale at 3 years of age when compared to infants without lesions. We have further observed significant differences in motor outcome (scale A of Griffiths) distributed progressively among patients with three grades of white matter lesions, from PWML of grade 1 to cystic PVL. We have observed positive correlation between values of Developmental Coefficient on Griffiths Scale at 3 years of age and BPW (r=0.18, p=0.0057), as well as TCD (r=0.27, p<0.001), even if excluding the patients with the brain lesions the correlations were not any more significant. Conclusions: We have observed that hemorrhagic punctate white matter lesions (SWI+ PWML) have risk factors, anatomical position and prognostic significance that differs with those of non-hemorrhagic ones, and that their presence is connected with lower gestational ages, as is known for other hemorrhagic lesions (GMH-IVH and CBH). It could be though useful to use this distinction in clinical practice. We have also observed that even low-grade lesions of all types have a negative influence on neurological outcome at 3 year of age, underlining the importance of precise diagnostics and elaboration of preventive strategies.File | Dimensione | Formato | |
---|---|---|---|
phdunige_3907940.pdf
accesso aperto
Descrizione: phd thesis
Tipologia:
Tesi di dottorato
Dimensione
970.62 kB
Formato
Adobe PDF
|
970.62 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.