Risk factors and associated diseases among preterm infants with isolated lenticulostriate vasculopathy

Abstract

Objective:

To determine the prevalence, risk factors and diseases associated with isolated lenticulostriate vasculopathy (LSV) among preterm infants.

Study Design:

Medical records of 84 preterm infants (gestational age (GA) 25 to 34 weeks) with isolated LSV in a case–control retrospective study over a period of 6.5 years were reviewed and compared with matched control infants. LSV was defined as ‘early’ if it was documented in head ultrasound (HUS) before or on the fifth day of life and ‘late’ if it was not present in the first HUS and recorded later during neonatal hospitalization.

Results:

A 3.9% prevalence of LSV was recorded among preterm infants (GA⩽34 weeks). Study and control groups were similar for all maternal parameters, neonatal outcomes and length of hospitalization. Infants with late LSV had more neonatal complications than control infants and were born with younger GA and lower birth weight in comparison to infants with early LSV. More infants with late LSV needed mechanical ventilation, were diagnosed with bronchopulmonary dysplasia and were hospitalized longer in comparison to infants with early LSV. Urine cytomegalovirus was negative in the entire study group.

Conclusions:

No risk factors or specific associated morbidities were identified among preterm infants with early isolated LSV. Infants with late isolated LSV were younger and had overall increased associated morbidities. Long-term outcome studies are needed to determine LSV impact.

Introduction

Lenticulostriate vasculopathy (LSV), presenting as hyperechogenic vessels in the thalamus and the basal ganglia, correlates with various neonatal diseases, among them fetal and neonatal infections, congenital malformations, chromosomal abnormalities, hypoxic/ischemic conditions, maternal diabetes, maternal drug abuse and prematurity.^{1, 2}

LSV has been described in 2.45% of preterm infants and in up to 20% of very preterm infants, although other brain abnormalities, mostly intraventricular hemorrhage (IVH), were recorded in some of these reports.^{1, 3, 4} Isolated LSV, defined as LSV with no other documented brain abnormality, is a condition often observed among preterm infants undergoing routine head ultrasound (HUS) studies during the neonatal period. Although isolated LSV is considered to be benign, its significance has not been fully investigated. Studies reporting LSV predictors and related morbidities have involved heterogeneous study groups, including term neonates and infants with other brain pathologies, that may influence the interpretation of LSV presence.^{4, 5} LSV was defined as ‘early’ if it was documented in HUS before or on the seventh or the tenth day of life in different studies and ‘late’ if it was recorded later during neonatal hospitalization.^{3, 6} All previous studies comprised small study groups (up to 22 cases of preterm patients). A study involving 21 preterm infants with isolated LSV⁶ did not find any differences in risk factors or associated perinatal complications between study and control infants, except for less use of antenatal steroids and more infants with low Apgar score at 1 min among infants with late LSV (after the tenth day of life). Long-term neurodevelopment was similar for both the study and control groups. Nevertheless, data are missing regarding the clinical impact of isolated LSV among preterm infants, namely, is LSV a result of brain pathologies emerging from perinatal conditions, such as maternal diseases and medications, labor factors and neonatal complications? Finding various risk factors for LSV may indicate the clinical impact of this diagnosis and potentially reflect on long-term outcomes.

The aims of the current study are to determine the prevalence, risk factors and diseases associated with isolated LSV among preterm infants.

Methods

The study included preterm infants of gestational age (GA) 25 to 34 weeks who were born at the Sheba Medical Center during the six-and-a-half-year period between January 2008 and January 2015 and who survived to discharge. Inclusion criteria were infants for whom isolated LSV was documented in at least one HUS exam during the neonatal hospitalization period. HUS screening is routinely conducted in our department for every preterm infant of GA⩽34 weeks. This screening includes bedside US evaluation during the first week of life (days 3 to 5), the second week of life (days 10 to –14) and at 1 month of age for those infants who are still hospitalized. Until 11 November 2013, HUS was performed using a HD11 Philips System (Philips Medical Systems, Bothell, WA, USA) using a 5- to 8-MHz curved-array transducer at the highest possible frequency and XRES adaptive image processing. Thereafter a Z.One US system (Zonare, Mountainview, CA, USA) with a 3 to 10 curved-array transducer at the highest possible frequency was employed. Each examination included both anterior and mastoid fontanelle scans. Images were documented and stored in the PACS for both the coronal and the sagittal planes as previously described.⁷ Pathological findings were documented as separate images and stored in the PACS. All exams were performed by a pediatric radiologist with at least 10 years’ experience in neonatal US or by a sonographer in the presence of a pediatric radiologist with the above mentioned experience. Studies were reported by the same radiologist present at or performing the US examination. Therefore, the images were reviewed twice: once during the examination and later on the same day while reporting. LSV was defined as the presence of curvilinear branched or single hyperechogenicity in the basal ganglia or thalamus. For each finding of LSV, the protocol in our department requires urine cytomegalovirus (CMV) screening.

LSV was defined as ‘early’ if it was documented in HUS before or on the fifth day of life and ‘late’ if it was not present in the first HUS and recorded later during neonatal hospitalization. The characteristics of infants with early and late LSV were compared.

Exclusion criteria were infants with abnormal ultrasonographic findings (except for LSV), including late (at age of 1 month) periventricular echodensity, IVH of all degrees, cystic periventricular leukomalacia, white matter hemorrhage or infarction and other pathological US documentation. Infants with documented choroid plexus cysts were considered to have normal HUS. For the purpose of the study and in order to reconfirm LSV diagnosis, a pediatric radiologist blindly re-evaluated all HUS exams of the study and control groups without knowing whether the child was affiliated with the study or control group. Only infants with reconfirmed isolated LSV findings were included in the study.

For each infant in the study group, we selected a control infant born at the same GA (completed weeks of gestation) closest (before or after) to the matched infant in the study group with documented and reconfirmed normal ultrasonography findings during the neonatal hospitalization period. When LSV occurred in one or both infants in a pair of twin neonates, they were matched to control twin neonates that conformed to the same order of twinning (first matched with first, second with second).

In order to find risk factors for LSV presence and to evaluate whether such a finding relates to various medical conditions, we collected retrospective data from the medical files of the mothers and infants. This data included: (1) Infants' perinatal conditions, such as GA; birth weight; weight for GA (appropriate—⩽10th percentile, small or large—⩾90th percentile for GA); gender; Apgar score; and delivery mode. (2) Maternal medical conditions and pregnancy complications such as maternal parity and gravida; twin chorionicity and amnionicity; evidence of twin-to-twin transfusion syndrome (TTTS); maternal gestational diseases (diabetes, hypertension, premature rupture of membranes of ⩾18 h and medications). (3) Postnatal complications such as respiratory distress syndrome and use and duration of respiratory support; neonatal hypotension and treatment given; patent ductus arteriosus as defined by clinical signs, including typical systolic murmur, bounding pulses and wide pulse pressure, confirmed by echocardiogram; bronchopulmonary dysplasia (BPD) as defined by the need for oxygen at 36 weeks postconception-corrected age; sepsis defined as positive blood culture; necrotizing enterocolitis defined by modified Bell's criteria of grade 2 to 3; retinopathy of prematurity of grade ⩾2; all documented HUS results during the neonatal period; and day of hospital discharge. All infants included in the study were Caucasians.

The institutional research ethics board at the Chaim Sheba Medical Center approved the study.

Statistical analysis

Medical parameter data for the study and control groups were compared. The continuous variables were compared using analysis of variance. The categorical variables were compared using Pearson’s chi-square test or Fisher's exact test. A P-value of <0.05 was considered significant. A logistic regression analysis was conducted to identify those variables that were most significantly associated with isolated LSV US findings.

Results

During the study time period, 2087 preterm infants (GA 25 to 34 weeks) were born at our medical center. Among them, 99 (4.7%) were primarily diagnosed with LSV in at least one HUS performed during postnatal hospitalization. However, our results show that during the earlier time period when the HD11 Philips System device was in use the prevalence of LSV diagnosis was 2.6%, whereas during the later study period when the Z.One US system was in use the prevalence of LSV diagnosis increased to 8.2%. Of those diagnosed, five infants were excluded owing to other brain pathologies (two with IVH, one with periventricular leukomalacia and one with ischemic lesions and late periventricular echodensity), four had major congenital malformations (ileal atresia, myxoma, heart defect and intestinal duplication), and six were excluded following the blind HUS recheck that did not reconfirm LSV. The remaining 84 infants comprised the study group. All cases had negative urine CMV.

During the study time period, 65 preterm infants did not survive to discharge. Of these, 24 died during the first 3 days of life and HUS was not performed. HUS did not reveal LSV among any of the other 41 infants who died.

Table 1 compares the study group and the matched control infants.

Table 1 Study characteristics for preterm infants with isolated LSV and control groups

The groups were similar for all maternal parameters. More infants in the study group were born by CS and overall neonatal complications were greater among the study group compared with controls.

Early LSV (found on first HUS less than fifth day of life) was determined in 50 infants, while late LSV (first found later than tenth day of life) was determined in 34 infants. When the LSV group was divided into early and late subgroups, and these are, in turn, compared with the control group, a significant difference with control infants is found only in the late LSV group, in the higher rate of overall complications (including infection, pneumothorax, bloody stool, renal failure, cardiac arrhythmia—supraventricular tachycardia) among infants with late LSV in comparison to control infants (17.% vs 0%, respectively, P=0.025).

Table 2 compares the characteristics of infants with early and late LSV. Mothers of infants with late LSV were more commonly treated with magnesium during pregnancy.

Table 2 Characteristics of infants with isolated early vs late LSV

Infants with late LSV were born with younger GA and lower birth weight compared with infants with early LSV. More infants with late LSV were diagnosed with BPD and required longer hospitalization.

Using logistic regression to predict preterm LSV (early and late) that included GA, birth weight, delivery mode, pregnancy diseases and neonatal outcomes, we found no perinatal and postnatal predictors.

Discussion

The current study included the largest reported study group (84 infants) of preterm infants with isolated LSV. The aim of the study was to evaluate whether diagnosis of isolated LSV among preterm infants is a marker of prenatal, perinatal or postnatal clinical conditions and morbidities. The study found no specific risk factors associated with early isolated LSV. Infants with late isolated LSV were younger and generally had increased associated morbidities.

Previous studies showed an LSV rate of 2.45% to 4.6% among preterm infants,^{4, 6, 8} although some studies reported rates as high as 20%.³ We speculate that this variability may be due to the use of different US systems and settings. With the advancement of technology, US reveals more findings. This is evident in the present study where LSV prevalence increased almost fourfold with the use of a different US system.

Most previous studies that evaluated risk factors for LSV among newborn infants used heterogenic study groups that included term infants and/or infants with other brain pathologies such as IVH.^{1, 2, 4} These studies reported several risk factors for LSV, including maternal magnesium treatment, hypoxic/ischemic conditions, respiratory distress syndrome and heart diseases. Two other studies that included small study groups showed no risk factors or associated morbidities, as was found in the current study.

Nevertheless, the attributes of infants with LSV diagnosed after birth (early) seem to differ from those with late appearance of LSV (late). Infants in the late group had smaller GA and birth weight and were sicker than those with early diagnosed LSV, as found in the current study and in previous reports.^{3, 6}

Among infants with LSV in the current study, 40% were defined as having late LSV, a lower prevalence than the 67% and 77% in previous studies.^{3, 6} This difference can be explained by the differences in GA in the study populations. Although the median GA in the current study population was 33 weeks, the study groups in the previous two studies included preterms with an average GA of 28.7 weeks and a median GA of 29 weeks,^{3, 6} correlating with late LSV as outlined in our own data.

Previous studies found that LSV was clinically correlated to congenital malformations, infections and hypoxic conditions,^{1, 2, 9, 10} suggesting a hypoxic/ischemic insult mechanism to the brain of the infant. Intrauterine insult will lead to early appearance (early LSV) and perinatal insult to late appearance (late LSV).³ The results of the current study are in line with this theory with respect to late LSV, which was found to correlate with younger GA, BPD and overall increased complications among preterm infants with LSV. With respect to early LSV, the current study did not find evidence of intrauterine insults among the studied parameters.

Among newborn infants, LSV is believed to be strongly correlated to intrauterine infection, mostly CMV. Indeed, CMV-positive symptomatic infants have a high rate (>50%) of LSV, which indicates central nervous system involvement.⁸ Our study results, which found no positive urine CMV among the entire study group with isolated LSV, confirm the results of previous studies,^{5, 8} indicating that LSV can commonly be found among preterm infants without obvious risk factors, including CMV infection. de Jong et al.⁸ concluded that TORCH (TOxoplasmosis, Rubella, Cytomegalovirus, Herpes simplex virus) evaluation is unnecessary when LSV is documented among Neonatal Intensive Care Unit infants.

Based on the assumption that LSV (mostly late onset) may be a marker for neonatal insult (hypoxic/ischemic), long-term outcomes are of great importance to evaluate its impact. The current study focused on risk factors and short-term diseases associated with isolated LSV and did not evaluate long-term outcomes. Two studies that evaluated long-term outcomes among infants with isolated LSV showed no neurodevelopmental abnormalities^{6, 11} while another study described idiopathic LSV as a predictor of neuropsychiatric disturbances.¹²

In summary, the results of the current study reinforce the assumption that there is no obvious evidence of hypoxic/ischemic insult among preterm infants with early isolated LSV. The differences between the early and late LSV subgroups may suggest different pathogenesis or timing between the groups. CMV is not likely to be identified among infants with isolated LSV.

In conclusion, no risk factors or specific associated morbidities were identified among preterm infants with early isolated LSV. Infants with late isolated LSV were younger and generally had increased associated morbidities. Long-term outcome studies are needed to determine the impact of LSV.