Cerebral Sinovenous Thrombosis in the Asphyxiated Cooled Infants: A Prospective Observational Study

Accepted Manuscript Cerebral sinovenous thrombosis in the asphyxiated cooled infants: a prospective observational study M. Radicioni, V. Bini, BSc, P. Chiarini, A. Fantauzzi, F. Leone, R. Scattoni, P.G. Camerini PII:

S0887-8994(16)30443-X

DOI:

10.1016/j.pediatrneurol.2016.09.006

Reference:

PNU 8990

To appear in:

Pediatric Neurology

Received Date: 21 June 2016 Revised Date:

7 September 2016

Accepted Date: 7 September 2016

Please cite this article as: Radicioni M, Bini V, Chiarini P, Fantauzzi A, Leone F, Scattoni R, Camerini P, Cerebral sinovenous thrombosis in the asphyxiated cooled infants: a prospective observational study, Pediatric Neurology (2016), doi: 10.1016/j.pediatrneurol.2016.09.006. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

ACCEPTED MANUSCRIPT

Title

RI PT

Cerebral sinovenous thrombosis in the asphyxiated cooled infants: a prospective observational study.

SC

Running Title

M AN U

Cerebral sinovenous thrombosis and whole-body cooling.

Authors

M. Radicionia, V. Binib, BSc, P. Chiarinic, A. Fantauzzia, F. Leonec, R. Scattonia, PG.

a

TE D

Camerinia.

Neonatal Intensive Care Unit, S.M. della Misericordia Hospital of Perugia, Perugia,

Italy; bDepartment of Medicine, University of Perugia, Perugia, Italy; cNeuroradiology,

AC C

EP

S.M. della Misericordia Hospital of Perugia, Perugia, Italy

Corresponding Author: Maurizio Radicioni, Neonatal Intensive Care - S. Maria della Misericordia Hospital of Perugia, S. Andrea delle Fratte - 06156 Perugia, Italy. Tel: +39 075 5786464. Fax: +39 075 5786467. e-mail: [email protected].

Manuscript's word count: 2258 Conflict of Interest: Authors declare no conflicts of interest to disclose.

2

ACCEPTED MANUSCRIPT ABSTRACT OBJECTIVE: Cerebral sinovenous thrombosis is considered unusual in the asphyxiated cooled infants, but reliable data regarding the true incidence of this comorbidity is lacking. This study aimed to assess its incidence in a population of

RI PT

asphyxiated cooled infants by performing routine brain magnetic resonance venography.

METHODS: All the asphyxiated infants cooled at our institution were subject to brain

SC

magnetic resonance venography after rewarming. Assessing the incidence of cerebral sinovenous thrombosis was the primary goal. Secondary analyses included group

M AN U

comparisons for laboratory tests and monitored parameters, relationship between variables, logistic regression models and receiver operating characteristic curve for cerebral sinovenous thrombosis prediction.

RESULTS: Cerebral sinovenous thrombosis were detected in 10/37 (27%) infants,

TE D

mainly at the superior sagittal sinus (8/10). These infants showed higher blanket (p< 0.001) and lower esophageal temperatures (p= 0.006), lower platelet count (p= 0.045) and received more red blood cell transfusions (p= 0.038) than the cooled

EP

infants without thrombosis. Blanket temperature was independently associated with cerebral sinovenous thrombosis (p= 0.049), and 32°C /hr. was the optimal cutoff value

AC C

to predict the event (sensitivity: 90%; specificity: 88.5%). CONCLUSION: High incidence or cerebral sinovenous thrombosis in neonates treated with therapeutic hypothermia, suggests that MRV is likely indicated in many of these children. High blanket temperature may be one variable that helps identify patients at higher risk.

Keywords: cerebral sinovenous thrombosis; magnetic resonance venography; neonatal hypoxic ischemic encephalopathy; stroke; therapeutic hypothermia

3

ACCEPTED MANUSCRIPT INTRODUCTION Cerebral sinovenous thrombosis (CSVT) is a rare finding with potential high neurological morbidity, characterized by partial/complete occlusion of one/multiple cranial venous sinuses and hindering of blood flow in the tributary venous system.

RI PT

CSVT affects at least 0.67 cases per 100.000 children per year, with higher rates among newborns (1-3). As in the adult and children, CSVT etiology in neonates is often multifactorial, but male sex, perinatal complications due to complicated/assisted

SC

deliveries and systemic illness are more frequently observed (2,4,5). Early presentation of symptoms, even at birth, may suggest the possibility for a fetal origin and a

M AN U

relationship with birth asphyxia (6,7). However, this condition is considered an unusual finding in the asphyxiated cooled infants, although accurate data are still lacking (8-11). The true extent of the problem is likely underestimated, because clinical signs and symptoms are similar to asphyxia and diagnosis can be missed if proper imaging of cerebral venous circulation is not performed. Our goal was to evaluate CSVT incidence

TE D

in a population of asphyxiated cooled infants through brain magnetic resonance

EP

venography (MRV) routinely performed after rewarming.

METHODS

AC C

As of July 2010, we began to treat the asphyxiated infants admitted at our third level neonatal intensive care unit with whole-body cooling to esophageal temperature of 33.5 °C maintained for 72 hours by using the servo-contr olled Blanketrol® III hyperhypothermia system [Cincinnati Sub-Zero OH 45241-1528 USA] in the gradient mode (10C - Smart Mode). Eligibility and treatment of the asphyxiated infants were conducted according to established criteria and methods (11). Video electroencephalography recordings were performed at admission, as an entry criterion, during the rewarming phase and in occasion of clinical seizures. Correct placement of the esophageal probe (third lower, esophageal) was checked with chest X-ray and

4

ACCEPTED MANUSCRIPT subsequently verified if abnormal swings of the esophageal and/or blanket temperatures were noted. All infants received phenobarbital and fentanyl infusion for sedation and pain control, as by internal protocol. Esophageal, skin, rectal, and blanket temperatures, vital signs, intravenous fluids, urine output and inotropic support were

RI PT

hourly recorded on a predesigned data form. Vasoactive-inotropic score was calculated according to what previously published by Gaies et al (12). Type and number of transfused blood components were also documented. Laboratory assessment,

including platelet count (PLT), prothrombin time, activated partial thromboplastin time,

SC

fibrinogen and D-Dimer was performed on admission and every 24 hours during the same period; D-Dimer values were dropped from the analysis because the method of

M AN U

measurement changed during the study period. Brain magnetic resonance imaging (MRI) was planned within the first week of life for each patient, or when clinically feasible. All brain MRIs were executed using institutional protocols for neonatal brain imaging on 1.5-Tesla scanner [Optima MR360 Advance - General Electric Medical

TE D

Systems] with a contiguous 2D time-of-flight magnetic resonance venography (MRV). No contrast agent was administered. All imaging results were independently reviewed by two authors experienced in evaluating neonatal brain MRI (P.C.; F.L.), blinded to the

EP

clinical conditions. Clearly visible complete or partial absence of flow in ≥ 1 cranial venous sinus on brain MRV was recognized as a sign of thrombosis, with or without

AC C

associated parenchymal lesions. All the CSVT positive infants were subject to follow-up brain MRI/MRV with timing decided at the discretion of the neuroradiologist, and cranial venous sinus patency was assessed as normal, improved or persistently occluded. Anticoagulation treatment protocol consisted of subcutaneous low molecular weight heparin (LMWH) at the starting dose of 1 mg/kg twice a day, titrated to anti-factor Xa activity between 0.1-0.3 units/mL for at least six weeks. A thorough screening for thrombophilia including antithrombin III deficiency, protein C deficiency, protein S deficiency, activated protein C resistance, factor V Leiden mutation, prothrombin gene mutation G20210A and thermolabile form of the methylene tetrahydrofolate reductase

5

ACCEPTED MANUSCRIPT gene was executed in the CSVT positive infants. Authors have complied with the World Medical Association Declaration of Helsinki regarding ethical conduct of research involving human subjects and had an approval for this study by the local Ethics Committee (CEAS N.:1356/09).

RI PT

Statistical analysis

The Shapiro-Wilk test was applied to assess the normal distribution of variables. Due to their asymmetric distribution, the non parametric Mann–Whitney's U-test was used for

SC

comparisons between CSVT positive and CSVT negative patients, and Chi-square test for comparisons of categorical variables. Multivariate logistic regression models were

M AN U

fitted for the prediction of the CSVT as binary variable (1=CSVT; 0=no CSVT), incorporating as explanatory variables all the variables that showed a p-value ≤ 0.25 in the bivariate analysis (14). To avoid multicollinearity problems, predictors that were in strong correlation with other explanatory variables were dropped from the models.

TE D

Multicollinearity in logistic regression models is a result of strong correlations between independent variables. The existence of multicollinearity inflates the variances of the parameter estimates. That may result, particularly for small and moderate sample sizes, in lack of statistical significance of individual independent variables while the

EP

overall model may be strongly significant. Multicollinearity may also result in wrong

AC C

signs and magnitudes of regression coefficient estimates, and consequently in incorrect conclusions about relationships between independent and dependent variables. Goodness-of-fit of logistic regression models was checked using Hosmer

and Lemeshow test and Odds Ratios (ORs) with 95% confidence intervals (CIs) were also calculated. The predictive accuracy of variables strongly related to the CSVT development was also quantified as the area under the receiver operating characteristics (ROC) curve. Finally, the relationships between variables were tested using the Spearman’s rho correlation coefficient analysis. All statistical analysis was

6

ACCEPTED MANUSCRIPT performed using IBM-SPSS® version 22.0 (IBM Corp., Armonk, NY, USA, 2013) and a two-sided p-value <0.05 was considered significant.

RI PT

RESULTS Thirty-seven asphyxiated infants underwent whole-body cooling at our institution from July 2010 through November 2015. Two brain MRI/MRV were performed in each infant

SC

at a median of 7 (IQR= 4.5 days) and 29.5 (IQR= 15.5 days) days of life, respectively. In the first brain scan, hypoxic-ischemic injuries were detected in 17/37 (46%) and

M AN U

CSVT in 10/37 (27%) infants, with male predominance (8/10) and main involvement of the superior sagittal sinus (SSS) (8/10). An overview of the affected sinuses, associated brain lesions and thrombosis evolution is provided in Table 1. Incidence of clinical seizures, severity of encephalopathy and anticonvulsant therapy was similar in both CSVT positive and negative infants. CSVT positive infants showed higher blanket

TE D

(33.3 [1.8] vs. 29.9 [2] °C; p< 0.001), lower esoph ageal hourly temperatures (33.4 [0.1] vs. 33.5 [0.1] °C; p= 0.006), lower PLT (121 x 10 3 [54] vs. 164 x 103 [48] /µL; p= 0.045) and received more red blood cell transfusions (RBC) (0.5 [0-2] vs. 0 [0-2]; p= 0.038)

EP

during cooling (Table 2). Bivariate logistic regression models showed significant relationships between CSVT development and blanket temperatures (OR: 3.289 [95%

AC C

CI: 1.416-7.641]; p= 0.006), esophageal temperatures (OR: 0.888 [95% CI: 0.7910.998]; p= 0.046), and PLT (OR: 0.982 [95% CI: 0.966-0.999]; p= 0.039) (Table 3). In separate multivariate logistic regression models, only blanket temperatures (adjusted for esophageal temperatures, PLT and RBC) were independently associated with CSVT event (T 0-24 hr., OR 2.839; 95%CI: 1.085-7.430; p= 0.034. T 24-48 hr., OR=3.324; 95%CI: 1.165-9.479; p= 0.025. T48-72 hr., OR=2.774; 95%CI: 1.093-7.044; p= 0.032. T 0-72 hr., OR=10.5; 95%CI: 1.009-110.8; p= 0.049, respectively). In the predictive accuracy analysis (ROC curve) the variables with the strongest relationship with the CSVT occurrence (blanket temperatures), showed the areas under the curve

7

ACCEPTED MANUSCRIPT (AUC) ranging from 83.8% to 89.8% (Figure 1). In Table 4 are reported sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) at the best cutoff for each hourly blanket temperature. Partial or complete sinus recanalization was observed at follow-up brain MRV in all

RI PT

patients except one, who suffered from an extensive thrombosis since the beginning. Two infants in the CSVT negative group died early in infancy because of complications due to hypoxic-ischemic encephalopathy. None of the patients treated with

SC

anticoagulation had life-threatening complications; however, it must be noticed that a small temporal hemorrhage was observed in a patient with superior sagittal sinus

M AN U

thrombosis, whose development could be related to a short-lived deep venous system thrombosis not detected by MRV as also to anticoagulation therapy. No death occurred in the CSVT positive infants. A mutation in homozygous of methylene tetrahydrofolate reductase gene was found in two infants of the CSVT positive group, but plasmatic levels of homocysteine were within the normal range. Preliminary data on motor and

TE D

cognitive development at 12 and 24 months of life using the Griffith's scale showed

EP

appropriate neurological development in both infant groups.

DISCUSSION

AC C

Neonatal CSVT more often occurs in the superficial venous system of the brain, probably because of its vulnerability to the mechanical forces during delivery. The normal moulding and overlapping of cranial sutures during complicated/assisted deliveries may result in the damage of the immediately underlying venous sinuses, thus promoting CSVT (2). Prolonged supine position and passive drainage of the cerebral venous system may also contribute by impairing venous flow within the cranial sinuses (15). Therefore, it is plausible that the asphyxiated cooled infants may be particularly prone to develop CSVT. Despite the male predominance among the CSVT positive infants, CSVT frequency was not influenced by sex or by the mode of delivery in our

8

ACCEPTED MANUSCRIPT case series. Clinical seizures and/or abnormal electrical activity were not as frequent as expected in the affected infants, likely because of anticonvulsant therapy and mild hypothermia. CSVT diagnosis relied heavily on neuroimaging, thus suggesting the necessity for the routine assessment of the cranial venous system patency in these

RI PT

patients. Brain MRV can image blood flow and detect its cessation within the cranial venous sinuses, so that CSVT can be readily identified (16,17). One potential pitfall when using this technique in newborns is the occurrence of flow gaps within the cranial venous sinuses due to the age-related smaller caliber, slower venous flow, and skull

SC

molding (18), particularly at the posterior portion of SSS. However, it is conceivable that the high incidence of CSVT observed at this level in our patients might be related

M AN U

to mechanical pressure effects of the occipital bone in supine position (15). Theoretically, slowing of blood flow in the cranial venous sinuses may be also due to concurrent pathological conditions (e.g.: polycythemia; poor cardiac output) in the asphyxiated infants, but significant differences were not observed between CSVT

TE D

positive and negative infants. According to our experience, isolated low platelet counts and high blanket temperatures during cooling should raise the suspicion of underlying CSVT in these babies. While platelet consumption has been associated with neonatal

EP

CSVT, with or without venous infarction (19), this is the first report concerning the possible association between CSVT development and high blanket temperatures

AC C

during cooling. In fact, CSVT positive infants needed higher blanket temperatures to maintain their core target temperature, as if they were more prone to overcooling. Such peculiar behavior suggests a reduced thermogenic response, but the incidence of known causes for altered endogenous heat production (e.g.: anticonvulsant therapy; severity of encephalopathy) was similar in the CSVT positive and negative infants. Evidence suggests that hypothalamus is the most important region for autonomic temperature control and its damage elicits inhibition of brown adipose tissue and shivering thermogenesis in cool environments (20). Although thalamic and hypothalamic injuries were not detected by brain MR, concomitant deep venous system

9

ACCEPTED MANUSCRIPT thrombosis resulting in vasogenic edema of this area cannot be completely excluded, given that thromboses may have been short-lived and these structures are much more commonly seen at contrast-enhanced MRV than at 2D time-of-flight MRV (21). Wide fluctuations in temperature above and below a target temperature may limit the

RI PT

effectiveness of therapeutic hypothermia. Both manual and servo-control methods are recognized to be equally effective in maintaining the target temperature during cooling, although the core temperature shows less variability using the servo-controlled

equipment (22). The Blanketrol III system has additional automatic control modes

SC

designed to limit the magnitude of the difference between the esophageal and

circulating water temperature (gradient modes) and in turn decreases fluctuations in

M AN U

temperature above and below the target temperature (23). Low birth weight (24) and malposition of the esophageal probe (25) can induce swings in the esophageal temperature during whole-body cooling by a servo-controlled system, but these factors were not relevant in our study. Based on our results, blanket temperatures may be

TE D

used as a predictor variable for CSVT development in the asphyxiated cooled infants. The diagnostic capability expressed as the area under the curve (AUC) was 85.2% with good sensitivity (80%) and specificity (85%) at the cutoff value of 32 °C hourly mean

EP

from the first day of treatment. The CSVT risk was 2.84, 3.32, and 2.77 times greater for each degree of temperature increase, in the first, second and third day of treatment,

AC C

respectively. Therefore, high blanket temperatures during cooling should direct towards early and appropriate neuroimaging assessment of the cranial sinus venous patency. Low-molecular-weight heparins are often used to treat thrombosis in infants, but evidence on anticoagulant therapy in the context of hypoxic ischemic encephalopathy is lacking, and diffuse brain injury may be a relative contraindication to anticoagulation for small, nonprogressive CVST (26). In our patients, enoxaparin therapy titrated to anti-Xa activity in the prophylactic range was associated with partial and/or complete recanalization of the sinuses in almost all cases at the follow-up MRV, without hemorrhagic complications due to anticoagulation. However, our study was not

10

ACCEPTED MANUSCRIPT designed to define the effect of the enoxaparin therapy on CSVT evolution, and definitive indications cannot be drawn.

RI PT

CONCLUSION Our results suggest that routine assessment of the cranial venous system patency should be recommended in the asphyxiated cooled infants, because they are

SC

particularly prone to develop CSVT. Further studies with larger sample sizes are needed to determine the impact of the detection and treatment of CSVT in infants

Acknowledgments

M AN U

undergoing therapeutic hypothermia.

Authors would like to thank Tiziana Becchetti, Francesca Di Genova, Cristiana Germini,

TE D

Carla Lupi, Daniele Mezzetti, Stefania Troiani, Aurelia Zeryngite and the nursing staff of the Neonatal Intensive Care Unit, for their valuable collaboration in the clinical care of

EP

the patients.

AC C

Funding Source: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Abbreviations

CSVT, cerebral sinovenous thrombosis LMWH, low molecular weight heparin MRI, magnetic resonance imaging MRV, magnetic resonance venography PLT, platelet count

11

ACCEPTED MANUSCRIPT PLTc, platelet concentrates RBC, red blood cell

RI PT

SSS, cerebral superior sagittal venous sinus

REFERENCES

1. deVeber G, Andrew M, Adams C, Bjornson B, Booth F, Buckley DJ, et al. Canadian Pediatric Ischemic Stroke Study Group. Cerebral sinovenous

SC

thrombosis in children. N Eng J Med 2001; 345:417-23.

2. Berfelo FJ, Kersbergen KJ, van Ommen CH, Govaert P, van Straaten HL, Poll-

M AN U

The BT, et al. Neonatal cerebral sinovenous thrombosis from symptom to outcome. Stroke 2010; 41:1382-88.

3. Schmidt B, Andrew M. Neonatal thrombosis: report of a prospective Canadian and international registry. Pediatrics 1995; 96:939-43.

TE D

4. Golomb MR, Dick PT, MacGregor DL, Curtis R, Sofronas M, deVeber GA. Neonatal arterial ischemic stroke and cerebral sinovenous thrombosis are more commonly diagnosed in boys. J Child Neurol 2004; 19:493-97.

EP

5. Dlamini N, Billinghurst L, Kirkham FJ. Cerebral venous sinus (sinovenous) thrombosis in children. Neurosurg Clin N Am 2010;21:511-27.

AC C

6. Barron TF, Gusnard DA, Zimmerman RA, Clancy RR. Cerebral venous thrombosis in neonates and children. Pediatr Neurol 1992; 8:112-16.

7. Ramenghi LA, Govaert P, Fumagalli M, Bassi L, Mosca F. Neonatal cerebral sinovenous thrombosis. Semin Fetal Neonatal Med 2009; 14:278-83.

8. Roach ES, Golomb MR, Adams R, Biller J, Daniels S, Deveber G, et al. Management of stroke in infants and children: a scientific statement from a Special Writing Group of the American Heart Association Stroke Council and the Council on Cardiovascular Disease in the Young. Stroke 2008; 39:2644-91.

12

ACCEPTED MANUSCRIPT 9. Al Yazidi G, Boudes E, Tan X, Saint-Martin C, Shevell M, Wintermark P. Intraventricular hemorrhage in asphyxiated newborns treated with hypothermia: a look into incidence, timing and risk factors. BMC Pediatr 2015; 15:106. 10. Azzopardi D, Robertson NJ, Cowan FM, Rutherford MA, Rampling M, Edwards

encephalopathy. Pediatrics 2000; 106:684-94.

RI PT

AD. Pilot study of treatment with whole body hypothermia for neonatal

11. Battin MR, Teele RL. Abnormal sagittal sinus blood flow in term infants following a perinatal hypoxic ischaemic insult. Pediatr Radiol 2003; 33:559-62.

SC

12. Shankaran S, Laptook AR, Ehrenkranz RA, Tyson JE, McDonald SA, Donovan EF, et al. Whole-body hypothermia for neonates with hypoxic-ischemic

M AN U

encephalopathy. N Engl J Med 2005; 353:1574-84.

13. Gaies MG, Gurney JG, Yen AH, Napoli ML, Gajarski RJ, Ohye RG, et al. Vasoactive-inotropic score as a predictor of morbidity and mortality in infants after cardiopulmonary bypass. Pediatr Crit Care Med 2010; 11:234-38.

TE D

14. Hosmer DW, Lemeshow S. Applied Logistic Regression 2nd ed. New York, N.Y.: Wiley; 2000. ISBN 0-471-35632-8. 15. Tan M, Deveber G, Shroff M, Moharir M, Pontigon AM, Widjaja E, et al. Sagittal

EP

sinus compression is associated with neonatal cerebral sinovenous thrombosis. Pediatrics 2011; 128:429-35.

AC C

16. Eichler F, Krishnamoorthy K, Grant PE. Magnetic resonance imaging evaluation of possible neonatal sinovenous thrombosis. Pediatr Neurol 2007; 37:317-23.

17. Connor SE, Jarosz JM. Magnetic resonance imaging of cerebral venous sinus thrombosis. Clin Radiol 2002; 57:449-61.

18. Widjaja E, Shroff M, Blaser S, Laughlin S, Raybaud C. 2D time-of-flight MR venography in neonates: anatomy and pitfalls. AJNR 2006; 27:1913-18. 19. Govaert P, Smith L, Dudink J. Diagnostic management of neonatal stroke. Semin Fetal Neonatal Med 2009; 14:323-28.

13

ACCEPTED MANUSCRIPT 20. Tansey EA, Johnson CD. Recent advances in thermoregulation. Adv Physiol Educ 2015;39:139-48. 21. Leach JL, Fortuna RB, Jones BV, Gaskill-Shipley MF. Imaging of cerebral venous thrombosis: current techniques, spectrum of findings, and diagnostic

RI PT

pitfalls. Radiographics 2006;26(S1):S19-41. 22. Strohm B, Azzopardi D, Brocklehurst P, Currie A, Edwards D, Halliday H, et al. Temperature control during therapeutic moderate whole-body hypothermia for neonatal encephalopathy. Arch Dis Child Fetal Neonatal Ed 2010; 95:F373-75.

SC

23. Laptook AR, Kilbride H, Shepherd E, McDonald SA, Shankaran S, Truog W, et al. Temperature control during therapeutic hypothermia for newborn

M AN U

encephalopathy using different Blanketrol devices. Ther Hypothermia Temp Manag 2014; 4:193-200.

24. Shankaran S, Laptook AR, McDonald SA, Higgins RD, Tyson JE, Ehrenkranz RA, et al. Temperature profile and outcomes of neonates undergoing whole

TE D

body hypothermia for neonatal hypoxic-ischemic encephalopathy. Pediatr Crit Care Med 2012; 13:53-9.

25. Lavinio A, Sanfilippo F, Gupta AK. Therapeutic hypothermia: heat transfer from

EP

warmed tracheal tubes to oesophageal temperature probes poses risk of lifethreatening overcooling. Br J Anaesth 2012; 108:328-29.

AC C

26. Dang LT, Shavit JA, Singh RK, Joshi SM, Leber SM, Barks JD, et al. Subdural hemorrhages associated with antithrombotic therapy in infants with cerebral atrophy. Pediatrics 2014;134(3):889-93.

14

ACCEPTED MANUSCRIPT FIGURE 1 (legend) ROC curve analyses performed on the 0-24 hr., 24-48 hr., 48-72 hr. and 0-72 hr. cumulative blanket temperature (expressed as mean/hour) to set a cutoff value for achieving the best diagnostic accuracy (sensitivity plus specificity) in predicting CSVT

RI PT

during and at the end of treatment. The overall diagnostic accuracy (AUC) for each

SC

curve was also reported.

•

M AN U

Key Notes

Cerebral sinovenous thrombosis is reported to be an unusual finding in the asphyxiated cooled infants, but reliable data are lacking.

•

Systematic assessment of the cranial sinus venous patency with brain magnetic

TE D

resonance venography may increase cerebral sinovenous thrombosis detection in the asphyxiated cooled infants. •

High blanket temperatures during whole-body cooling may raise the suspicion

AC C

EP

of underlying cerebral sinovenous thrombosis in these patients.

RI PT

ACCEPTED MANUSCRIPT

TABLE 1. Overview of the affected sinuses, associated brain lesions and thrombosis evolution

Days of life at first brain MRV Follow-up MRV

SC

BRAIN HEMORRHAGES Right Temporal No Left Parasagittal and Peritrigonal No No No No No Right Frontal and Peritrigonal Right Intraventricular and Peritrigonal

M AN U

TE D

b

CEREBRAL VENOUS SINUS Superior Sagittal (Posterior) Superior Sagittal (Anterior) Superior Sagittal (Posterior) Superior Sagittal (Middle) + Left Straight Superior Sagittal (Posterior) Superior Sagittal (Posterior) Left Transverse + Left Sigmoid Superior Sagittal (Anterior) Superior Sagittal (Posterior) + Right Straight Right Straight + Right Jugular Vein

EP

a

MRVa 7 6 10 5 12 6 7 5 7 10

AC C

SEX M F M M F M M M M M

SINUS PATENCYb Normal Normal Normal Normal Normal Improved Improved Improved Improved Obstructed

ACCEPTED MANUSCRIPT

M AN U

TE D EP AC C

Gestational age,1 (wks) Birth weight,2 (gr.) Male gender3 Cesarean section3 Arterial blood pH2,a Arterial base deficit,2,a (mmol/L) Sarnat score1,a Anticonvulsant drugs,1,d Clinical seizures,3,d Vasoactive inotropic score1,b Mean arterial pressure,2,b (mmHg) Intravenous fluids,1,b (mL/Kg/die) Urine output,2,b (mL/Kg/hr.) Blanket temperatures,2,b (°C) Esophageal temperature,2,b (°C) Rectal temperature,2,b (°C) Blood glucose,2,c (mg/dL) Serum creatinine,1,c (mg/dL) Haematocrit,2,c (%) Prothrombin time,1,c (s) Activated partial thromboplastin time, 2,c (s) Platelet count,2,c (x 103/µL) PFC transfusions,1,d PLTc transfusions,1,d RBC transfusions,1,d 1 median (min-max) 2 mean (SD) 3 n (%) a Within 60’ from birth b Hourly values during cooling c Determinations during cooling d Events during cooling

CSVT Positive Infants (N = 10) 40 (38-41) 3144 (540) 8 (80) 8 (80) 6.96 (0.15) 17.9 (4.9) 1,5 (1-2) 1 (1-3) 2 (20) 487.2 (0-2121413) 52.4 (5.3) 78.6 (57-106) 2.8 (0.9) 33.3 (1.8) 33.4 (0.1) 33.0 (0.7) 96 (47) 0.7 (0.2-1.9) 40.5 (4) 21.9 (16.3-39.4) 47.5 (12.3) 121 (54) 1.5 (0-8) 0 (0-3) 0.5 (0-2)

SC

VARIABLES

RI PT

TABLE 2. Comparison of Baseline Characteristics, Laboratory Tests and Monitored Parameters between CSVT Positive and Negative Patients CSVT Negative Infants (N = 27) 40 (34-42) 3350 (547) 17 (63) 11 (41) 6.96 (0.15) 19.5 (5.7) 2 (0-3) 1 (1-2) 6 (22) 113.5 (0-1295198) 52.2 (5.3) 64.8 (45-136) 2.8 (0.8) 29.9 (2.0) 33.5 (0.1) 33.1 (0.4) 100 (29) 0.5 (0.3-2.0) 40.4 (6) 18.1 (14-33.6) 48.9 (9.2) 164 (48) 0.5 (0-9) 0 (0-4) 0 (0-2)

P .639 .338 1.00 .085 .910 .496 .617 .411 1.00 .342 .973 .116 .644 <.001 .006 .573 .230 .156 .798 .560 .711 .045 .420 .062 .038

RI PT

ACCEPTED MANUSCRIPT

TABLE 3. Bivariate Logistic Regression Analysis Models (dependent variable: CSVT yes/no)

EP

AC C

n (%) mean (SD) 3 median (min-max) a Hourly values during cooling b Mean value of determinations during cooling 2

TE D

1

CSVT Negative Infants (N = 27) 17 (63) 11 (41) 29.9 (2.0) 33.5 (0.1) 164 (48) 0 (0-2)

SC

Male gender1 Cesarean section1 Blanket temperatures,2,a (°C) Esophageal temperature,2,a (°C) Platelet count,2,b (x 103/µL) RBC transfusions,3,a (n)

CSVT Positive Infants (N = 10) 8 (80) 8 (80) 33.3 (1.8) 33.4 (0.1) 121 (54) 0.5 (0-2)

M AN U

Explanatory Variables

P

.691 .928 .006 .046 .039 .075

OR (95% CI) 1.373 (0.288-6.544) 1.052 (0.353-3.135) 3.289 (1.416-7.641) 0.888 (0.791-0.998) 0.982 (0.966-0.999) 3.190 (0.888-11.456)

ACCEPTED MANUSCRIPT

TABLE 4 - Area under the ROC curve (AUC), sensitivity, specificity, PPV and NPV of predictive variables. 95%C.I.

Criterion °C/hr.

Sensitivity (%)

95%C.I.

Specificity (%)

95%C.I.

PPV

BT 0-24 hr.

85.2

69.4-94.8

32.0

80.0

44.4-97.5

88.5

69.8-97.6

72.7 46.4 - 99.0

92.0 81.4 - 102.6

BT 24-48 hr.

87.3

72.0-96.0

31.4

80.0

44.4-97.5

84.6

65.1-95.6

66.7 40.0 - 93.3

91.7 80.6 - 102.7

BT 48-72 hr.

83.8

67.8-94.0

31.7

90.0

55.5-99.7

69.2

48.2-85.7

52.9 29.2 - 76.7

94.7 84.7 - 104.8

BT 0-72 hr.

89.8

75.1-97.4

32.2

90.0

55.5-99.7

88.5

69.8-97.6

75.0 50.5 - 99.5

95.8 87.8 - 103.8

M AN U

SC

RI PT

AUC (%)

AC C

EP

TE D

BT= blanket temperatures

95%C.I.

NPV

95%C.I.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT