Predictive markers, models and systems for brain damage in premature infants
By combining serum biomarkers and quantitative MRI technology with multiple clinical indicators, a predictive model for brain injury in preterm infants was constructed, which solved the problem of accuracy in early diagnosis of brain injury in preterm infants and achieved efficient prediction and early intervention.
Patent Information
- Application Number
- CN202210139827.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-02-15
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2042-02-15
AI Technical Summary
Existing technologies are insufficient for the early and accurate diagnosis of brain injury in premature infants, especially brain injury-related brain injury (BIPI), and conventional examination methods suffer from poor diagnostic specificity or are complex to perform.
A predictive model for brain injury in premature infants was constructed by combining serum biomarkers (such as B-FABP and S100β protein) with quantitative MRI technology and multiple clinical indicators. The predictive model was established by logistic regression analysis and the prediction was made using factors such as invasive mechanical ventilation time, acidosis, and frequent apnea.
This study improved the early diagnostic accuracy and predictive ability of BIPI in preterm infants. The area under the ROC curve of the model reached 0.998, with a sensitivity of 97.6% and a specificity of 98.6%, which is helpful for early intervention and improved prognosis.
Smart Images

Figure CN114842976B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to disease prediction model technology, and specifically relates to a prediction marker, prediction model and system for brain damage in premature infants. Background Art
[0002] Premature infants and extremely low birth weight infants still have a high mortality rate and disability risk, and the incidence of premature infant-related diseases such as brain injury in preterm infants (BIPI), bronchopulmonary dysplasia, neonatal necrotizing enterocolitis and retinopathy of prematurity is gradually increasing. BIPI is one of the main causes of neonatal mortality and is closely related to visual impairment, cognitive behavioral disorders, and long-term cerebral palsy. BIPI brings a heavy burden to families, society, and health systems, and seriously affects the future life and learning of premature infants. Early diagnosis of BIPI and early intervention and treatment are particularly important.
[0003] BIPI is a disorder of brain hemodynamics in premature infants caused by various infections, asphyxia, hypoxia, ischemia, etc., and there are different degrees of brain structural or functional damage. According to neuropathology, brain damage in premature infants can be divided into: ① non-hemorrhagic brain damage: including periventricular leukomalacia (PVL), periventricular hemorrhagic infarction (PVHI) and late ventriculomeglay; ② hemorrhagic brain damage: including periventricular-intraventricular hemorrhage (PVH-IVH), brain parenchymal hemorrhage, subarachnoid hemorrhage (SAH), etc.; ③ other parts of the body: such as cerebellum and brainstem damage. PVL is divided into cystic PVL, non-cystic PVL and diffuse white matter damage (white matter damage, WMD). PVH-IVH and PVL are common types of brain damage in premature infants. The incidence of PVH-IVH has gradually decreased over the past few decades. Studies in the early 1980s showed that the incidence of PVH-IVH in extremely low birth weight infants weighing less than 1500g was as high as 40% to 50%, while this figure dropped to 20% in the early 21st century.
[0004] At present, it is believed that the occurrence of BIPI is related to multiple factors. Foreign studies have reported that the levels of tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6) in the cord blood of children with BIPI are significantly higher than those in normal newborns, indicating that infection and inflammatory response are related to the occurrence of BIPI. Studies have found that early-onset sepsis is a common risk factor for hemorrhagic brain injury and non-hemorrhagic brain injury. Kidokoro et al. found that gestational age is a risk factor for BIPI and is negatively correlated with the occurrence of BIPI. Since BIPI often lacks typical clinical manifestations and signs, diagnosis mainly relies on imaging examinations such as cranial ultrasound and magnetic resonance imaging (MRI). Bedside cranial ultrasound has the advantages of simple operation, convenience, non-invasiveness and low cost, but it also has limitations such as unclear exploration and poor diagnostic specificity. MRI has the advantages of non-invasiveness, no radiation, strong diagnostic sensitivity and specificity, but it is complicated to operate, the examination time is late, and it is difficult for newborns to cooperate, which limits its clinical application. Therefore, a single examination method often inevitably has certain drawbacks, and the combined diagnosis of BIPI using multiple technologies is crucial.
[0005] Related studies have found that serum S100β protein levels are significantly increased after myocardial injury, suggesting that its specificity as a marker of brain injury is limited. In addition, due to its short duration in serum and the inconvenience of collecting umbilical cord blood samples in NICU, it is still necessary to make more rational use of its characteristics and explore other more specific and longer-lasting serum markers. At present, there are no reports on the predictive significance of B-FABP in BIPI in premature infants. Magnetic resonance imaging (MRI) is a non-ionizing imaging technology with high soft tissue resolution, which can be used to evaluate the brain structure of premature infants in detail. MRI technology can be used to evaluate the contrast of white matter and gray matter, the myelination process, ischemic pathological evolution originating from arterial abnormalities, vascular lesions of the brain Willis circle and its branches, hemorrhagic lesions, focal lesions >1mm, diffuse WMD, etc. Therefore, for children with WMD such as PVL and PVHI, MRI has the highest diagnostic value. At present, the MRI sequences routinely used in clinical evaluation of the neonatal brain are mainly T1WI, T2WI, diffusion weighted imaging (DWI) and inversion recovery sequences (such as FLAIR sequences). Brain injury is diagnosed by abnormal MRI signals, but subtle changes in brain structure are ignored and brain injury cannot be identified early before the signal changes. There is no clear conclusion on the application of quantitative detection technology of cranial MRI in predicting and diagnosing the occurrence and short-term neurological outcomes of premature infants. Summary of the invention
[0006] The present invention discloses for the first time the predictive role of serum markers combined with quantitative MRI technology in the occurrence of BIPI in premature low birth weight infants, providing a theoretical basis for early prevention and treatment of BIPI and improved prognosis. The present invention discloses for the first time the correlation between early peripheral blood B-FABP level changes and the occurrence of BIPI in premature infants.
[0007] Application of predictive factors in the prediction of brain injury in premature infants, wherein the predictive factors are one or more of the following: invasive mechanical ventilation time (days), acidosis, frequent apnea, platelet count > 0.27, serum S100β protein level (ng / ml) on the first day after birth, serum B-FABP level (ng / ml) on the first and seventh days after birth, LVD (mm), and IHD (mm).
[0008] The application of predictive factors in establishing a prediction model for brain injury in premature infants, wherein the predictive factors are one or more of the following: invasive mechanical ventilation time (days), acidosis, frequent apnea, platelet count > 0.27, serum S100β protein level (ng / ml) on the first day after birth, serum B-FABP level (ng / ml) on the first and seventh days after birth, LVD (mm), and IHD (mm).
[0009] A prediction model for brain damage in premature infants, the regression equation is: P =1 / [1+e-(-28.179)+0.458X1+1.028X2+1.932X3+1.401X4+1.489X5+0.595X6+0.410X7+0.429X8+1.740X9], where X1, X2, X3, X4, X5, X6, X7, X8, X9 are invasive mechanical ventilation time (days), acidosis, frequent apnea, platelet count>0.27, serum S100β protein level on the first day after birth (ng / ml), serum B-FABP level on the first and 7 days after birth (ng / ml), LVD (mm), IHD (mm), respectively. P is the probability of the Logistic model.
[0010] Application of the above-mentioned prediction model of brain damage in premature infants in the prediction of BPD in premature infants.
[0011] The method for establishing the prediction model of brain damage in premature infants mentioned above includes the following steps: collecting clinically relevant indicators of premature infants; then substituting the relevant indicators into the regression equation: P=1 / [1+e-(-35.032+0.431X1+0.245X2+8.363X3-0.321X4+0.707X5+1.236X6)], and obtaining the prediction model of BPD in premature infants; the clinically relevant indicators of premature infants include invasive mechanical ventilation time (days), acidosis, frequent apnea, platelet count>0.27, serum S100β protein level on the first day after birth (ng / ml), serum B-FABP level on the first and 7th days after birth (ng / ml), LVD (mm), and IHD (mm).
[0012] A prediction system for brain damage in premature infants includes a prediction module, a data input module and a prediction output module, wherein the prediction module includes the above-mentioned prediction model for brain damage in premature infants, the data input module is used to input relevant indicator data into the above-mentioned prediction model for brain damage in premature infants, and the prediction output module outputs the predicted probability of brain damage in premature infants according to the prediction model for brain damage in premature infants.
[0013] A computer equipped with the above-mentioned prediction system for premature infant brain damage.
[0014] The area under the ROC curve of the analysis of risk factors related to brain damage in premature infants and the risk assessment model given in the prior art is 0.76, and the Hosmer.Lemeshow goodness-of-fit test result is: (χ2 value = 4.016, P value = 0.675>0.05). It is believed that the model has good ability to identify risk factors and fitting effect. The AUC of the prediction model for BPD in premature infants constructed by combining 9 indicators with Logistic regression in the present invention is 0.998, the sensitivity is 97.6%, and the specificity is 98.6%, which is significantly higher than the existing prediction model.
[0015] The present invention uses premature infants who were admitted to the NICU of the Children's Hospital of Soochow University within 24 hours after birth from July 1, 2018 to December 31, 2020, with a gestational age of <37 weeks, a birth weight of <2500g, and a hospitalization time of ≥14 days as research objects. By analyzing the clinical data of premature infants and pregnant mothers, the serum S100β protein and B-FABP levels of premature infants on the 1st, 7th, and 14th days after birth were dynamically detected. A cranial MRI examination was performed at a corrected gestational age of 34 to 38 weeks and the brain structure was quantitatively measured. A prospective clinical study was conducted to explore the predictive role of serum markers combined with quantitative MRI technology in the occurrence of BIPI in premature low birth weight infants, and to provide a theoretical basis for early prevention and treatment of BIPI and improvement of prognosis. The present invention discloses for the first time the correlation between early peripheral blood B-FABP level changes and the occurrence of BIPI in premature infants. Data and methods. The BIPI prediction model was constructed by Logistic regression with the invasive mechanical ventilation time (days), acidosis, frequent apnea, platelet count>0.27, serum S100β protein level (ng / ml) on the first day after birth, serum B-FABP level (ng / ml) on the first and seventh days after birth, LVD (mm), and IHD (mm) as independent variables. The regression equation was: P =1 / [1+e-(-28.179)+0.458X1+1.028X2+1.932X3+1.401X4+1.489X5+0.595X6+0.410X7+0.429X8+1.740X9], where X1, X2, X3, X4, X5, X6, X7, X8, X9 are invasive mechanical ventilation time (days), acidosis, frequent apnea, platelet count>0.27, serum S100β protein level on the first day after birth (ng / ml), serum B-FABP level on the first and 7 days after birth (ng / ml), LVD (mm), IHD (mm), respectively. P is the probability of the Logistic model. The chi-square value of the Logistic regression prediction model is 2.689. P <0.001, indicating that the above indicators have explanatory power for the prediction of BIPI. Hosmer-Lemeshow was calculated using the classification interaction table, df=8, P >0.05, indicating that the prediction model has good goodness of fit. Taking the predicted variable as the test variable, the results showed that the area under the ROC curve of the above 9 indicators combined with the Logistic regression model was 0.998, which was greater than the prediction of each indicator alone. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 It is the measurement method of various indicators in magnetic resonance imaging.
[0017] Figure 2ROC curves for predicting the occurrence of BIPI by serum S100β protein and B-FABP.
[0018] Figure 3 ROC curve for predicting BIPI using cranial MRI brain structure measurements.
[0019] Figure 4 ROC curve for the Logistic regression model predicting the occurrence of BIPI. DETAILED DESCRIPTION
[0020] (1) Duration of invasive mechanical ventilation, acidosis, frequent apnea, and platelet count > 0.27 are independent risk factors for BIPI in premature low birth weight infants.
[0021] (2) The serum S100β protein level on the first day after birth and the serum B-FABP levels on the first and seventh days after birth of premature infants with BIPI were significantly higher than those of premature infants without BIPI, and their levels were positively correlated with the severity of BIPI. When the serum S100β protein level on the first day after birth was greater than 6.325 ng / ml, and the serum B-FABP level on the first and seventh days was greater than 12.14 ng / ml and 12.32 ng / ml, respectively, the risk of premature infants developing BIPI increased.
[0022] (3) B-FABP is a serum biomarker for predicting BIPI in preterm infants with a longer peak duration than S100β protein.
[0023] (4) The LVD and IHD of cranial MRI in premature infants with BIPI were significantly larger than those in premature infants without BIPI. When the measured values were greater than 13.75 mm and 1.65 mm, respectively, the risk of BIPI was predicted to increase.
[0024] (5) The duration of invasive mechanical ventilation (days), acidosis, frequent apnea, platelet count>0.27, serum S100β protein level (ng / ml) on the first day after birth, serum B-FABP level (ng / ml) on the first and seventh days after birth, LVD (mm), and IHD (mm) were combined to construct a regression model, which could better predict the occurrence of BIPI in premature infants.
[0025] 1. Inclusion criteria
[0026] Premature infants with gestational age <37 weeks, birth weight <2500 g, and hospitalization time ≥14 days who were admitted to the NICU of Children's Hospital of Soochow University within 24 hours after birth from July 1, 2018 to December 31, 2020 were selected as the research subjects. This study was approved by the Ethics Committee of the NICU of Children's Hospital of Soochow University (Ethics Number: 2021CS022) with the informed consent of the patients' families.
[0027] Exclusion criteria:
[0028] (1) Age at admission >24 hours and hospital stay <14 days;
[0029] (2) Patients with neurological malformations or severe complications;
[0030] (3) Patients with genetic metabolic diseases or chromosomal abnormalities;
[0031] (4) Patients with bilirubin encephalopathy and intrauterine TORCH infection;
[0032] (5) Those with surgical diseases or undergoing surgical operations during hospitalization;
[0033] (6) Those who died within a few hours after being admitted to the hospital and receiving active rescue treatment;
[0034] (7) Patients with incomplete clinical data.
[0035] 2. Diagnostic criteria and grouping
[0036] 2.1 Diagnostic criteria for brain damage in premature infants
[0037] According to the Expert Consensus on Diagnosis and Prevention of Brain Injury in Premature Infants formulated by the Chinese Medical Doctor Association in 2012
[39] And the "Expert Group Opinion on the Diagnosis, Prevention and Comprehensive Management of White Matter Damage in Premature Infants" formulated by the 2015 Neonatal Neurology Forum Expert Group
[40] , combined with my country's 20-item neonatal behavioral neurological assessment (NBNA)
[41] To make a diagnosis, premature infants must meet the following conditions: ① abnormal imaging examination (cranial ultrasound and / or MRI); ② abnormal NBNA score; including or excluding: ③ high-risk factors for brain injury; ④ clinical manifestations; they can be diagnosed as BIPI.
[0038] The specific diagnostic criteria are as follows:
[0039] (1) High-risk factors causing brain damage: hypoxia-ischemia, hemodynamic disorder, infection, abnormal coagulation function, abnormal delivery history, maternal and pregnant comorbidities or complications, etc.;
[0040] (2) Clinical manifestations: atypical, may be accompanied by central apnea, bradycardia, hypotension, hypertension, drowsiness, irritability, convulsions, abnormal muscle tone, abnormal primitive reflexes, etc., or may be without clinical symptoms;
[0041] (3) Ultrasound manifestations of the head: including subependymal hemorrhage, SAH, subdural hemorrhage, IVH, PVH, infarction, PVL, etc. IVH is graded according to the Papile classification system under ultrasound.
[42] It is divided into four grades: Grade I: unilateral or bilateral subependymal hemorrhage in the germinal matrix layer; Grade II: subependymal hemorrhage breaks into the ventricles, that is, intraventricular hemorrhage, without ventricular dilatation; Grade III: intraventricular hemorrhage with ventricular dilatation; Grade IV: IVH with periventricular hemorrhagic infarction. PVL is divided into four grades according to the de Vries method under ultrasound.
[42] : Grade I: bilateral symmetrical hyperechoic masses around the ventricles, lasting ≥7 days, and no cysts appear afterwards; Grade II: bilateral local hyperechoic masses around the ventricles, forming small cysts around the ventricles at least 2 weeks after birth; Grade III: bilateral extensive hyperechoic masses around the ventricles, which turn into extensive periventricular cystic changes after a few weeks (at the earliest 2 weeks), and the cysts can fuse into pieces; Grade IV: bilateral extensive hyperechoic masses around the ventricles, involving the white matter, which turn into diffuse cystic changes around the ventricles and subcortical superficial white matter after a few weeks (at the earliest 2 weeks).
[0042] (4) MRI examination: According to the 2012 Expert Consensus on Diagnosis and Prevention of Brain Injury in Premature Infants
[39] , MRI abnormalities in children with BIPI include: in the early stage of non-hemorrhagic white matter injury, T1WI shows high signal in the white matter area, and T2WI shows low signal or equal signal; in the later stage, T1WI signal disappears or shows low signal or white matter volume decreases, T2WI shows high signal, and severe cases have ventricular morphological changes. In the early stage of hemorrhagic injury, T1WI shows high signal and T2WI shows low signal, and in the later stage, both T1WI and T2WI show high signal.
[0043] (5) 20-item behavioral neurological assessment of newborns (NBNA score)
[41] : It is divided into 5 parts, namely 6 items of behavioral ability, 4 items each of passive muscle tension and active muscle tension, 3 items of primitive reflexes, and 3 items of general reactions. Each score has 3 grades, namely 0 points (failed to elicit and significantly abnormal), 1 point (slightly abnormal) and 2 points (completely normal), with a total score of 40 points, and <35 points is abnormal. See Appendix 1 for details.
[0044] 2.2 BIPI Grading Standard
[0045] Mild brain injury [43,44] : ① Bleeding involving only one cerebral sulcus or lobe; ② Bleeding in one or more ventricles but not accompanied by ventricular dilatation; ③ White matter lesions in MRI are ≤2 mm and range ≤3 sites.
[0046] Severe brain injury [43,44] : ① Bleeding involves two or more cerebral lobes; ② Intraventricular hemorrhage with ventricular dilatation; ③ White matter damage lesions in MRI are >2mm and / or range >3 sites; In severe cases, extensive subcortical white matter is involved.
[0047] 2.3 Diagnostic criteria for primary underlying diseases and common complications in premature infants
[0048] (1) Neonatal respiratory distress syndrome (NRDS)
[45] :It occurs in newborns within 12 hours after birth, with progressively worsening dyspnea accompanied by groaning, shortness of breath, irregular breathing, and inspiratory three-point retraction sign. X-rays can show decreased translucency of both lungs and bronchial inflation signs. In severe cases, white lungs may be present.
[0049] (2) Neonatal necrotizing enterocolitis (NEC)
[46] The main clinical manifestations are abdominal distension, bloody stools, and vomiting. In severe cases, there are concurrent shock and multiple organ failure. Abdominal X-rays are characterized by cystic gas accumulation in some intestinal walls, and pathology is characterized by necrosis of the distal ileum and proximal colon.
[0050] (3) Bronchopulmonary dysplasia (BPD) in premature infants
[47] Premature infants or extremely low birth weight infants who are still receiving mechanical ventilation, CPAP or FiO2>30% at 28 days after birth and 36 weeks of corrected gestational age are diagnosed with BPD. If FiO2≤30% and oxygen saturation is between 90% and 96%, they will undergo an oxygen weaning trial, and those who fail to wean are diagnosed with BPD.
[0051] (4) Respiratory failure
[48] : It is characterized by early accelerated respiratory rate and late slowed respiratory rate, labored breathing, flaring nasal wings, and the three-recess sign. Arterial blood gas analysis shows that the partial pressure of oxygen (PO2) is ≤60mmHg with or without the partial pressure of carbon dioxide (PCO2) >50mmHg. It is a common serious respiratory complication.
[0052] (5) Apnea of prematurity
[49] : Respiratory apnea lasts ≥ 20 seconds, or respiratory apnea lasts < 20 seconds with a slow heart rate < 100 beats / minute, with cyanosis of the skin, decreased muscle tension, poor response, and decreased oxygen saturation. Frequent apnea refers to repeated apnea ≥ 3 times / hour.
[0053] (6) Retinopathy of prematurity (ROP)
[50] : It is divided into 5 stages according to the severity: Stage I: a dividing line appears between the vascular and non-vascular areas in the temporal peripheral area of the retina; Stage II: the fundus dividing line is raised and ridge-like changes appear; Stage III: retinal blood vessels dilate and proliferate with fibrous tissue proliferation; Stage IV: tractional retinal detachment occurs from the periphery to the posterior pole; Stage V: complete retinal detachment.
[0054] (7) Neonatal septicemia [51,52] :The main symptoms are fever or no temperature rise, poor response, no appetite, jaundice, hepatosplenomegaly, shock, etc., and any one of the following conditions is met at the same time: ① ≥2 positive non-specific blood tests: increased or decreased white blood cell count, band nucleus cell / neutrophil number ≥ 0.16, platelet count <100×109 / L, increased C-reactive protein, increased procalcitonin; ② Cerebrospinal fluid examination shows purulent meningitis changes; ③ Pathogenic bacteria DNA is detected in the blood. Among them, those with onset time ≤ 3 days of age are early-onset sepsis (EOS), which is mostly vertical transmission of maternal pathogens; those with onset time > 3 days of age are late-onsetsepsis (LOS), which is mostly hospital-acquired or community-acquired.
[0055] (8) Pulmonary hemorrhage of the newborn (PHN)
[53] : Refers to massive bleeding in the lungs, involving at least two lobes of the lungs. Clinically, bloody fluid can be aspirated from the oral cavity, nasal cavity or trachea of the child, and bleeding caused by surgical injury or gastric bleeding is ruled out.
[0056] (9) Neonatal hyperbilirubinemia
[54] : After the birth of a newborn, hyperbilirubinemia is diagnosed when the serum bilirubin level exceeds the 95th percentile of the neonatal bilirubin curve.
[0057] (10) Parenteral nutrition associated cholestasis (PNAC)
[55] PNAC: It is a common and serious complication in premature infants during parenteral nutrition. Clinically, if continuous parenteral nutrition lasts for more than 14 days, accompanied by jaundice, lighter stool color, and increased serum direct bilirubin (>34μmol / L), and other causes of cholestasis are excluded, PNAC is considered to be diagnosed.
[0058] (11) Extrauterine growth retardation (EUGR)
[56] : Premature infants do not meet the expected intrauterine growth after birth. Referring to the 2013-Fenton curve, at the time of discharge or corrected gestational age of 36 weeks, their growth and development indicators are below the 10th percentile level of infants of the same age.
[0059] (12) Ventilator associated pneumonia (VAP)
[57] : Pulmonary infection in intubated children occurring after 48 hours of mechanical ventilation or within 48 hours after extubation, accompanied by changes in respiratory status, lung rales, and new radiological abnormalities.
[0060] (13) Neonatal anemia
[58] : Within 7 days after birth, the hemoglobin (Hb) concentration is less than 145g / L, among which 120g~145g is mild anemia, 90~119g is moderate anemia, and below 90g is severe anemia.
[0061] (14) Hypoglycemia [59,60] : Regardless of gestational age and age, a newborn baby can be diagnosed if the whole blood glucose is ≤2.2mmol / L. Hypoglycemia can be divided into mild (>1.7~2.2mmol / L), moderate (>1.0~1.7mmol / L), and severe (≤1.0mmol / L) according to the severity.
[0062] (15) Hyperglycemia
[60] : The domestic diagnostic standard is now mostly whole blood glucose > 7.0mmol / L.
[0063] (15) Score for neonatal acute physiology, perinatal extension, version Ⅱ (SNAPPE-Ⅱ)
[61] : The scoring standard based on SNAP-Ⅱ developed by Richardson in 2001 was used, including 9 variables: mean blood pressure, lowest body temperature, PO2 / FiO2, lowest pH value, urine output per hour, convulsions, birth weight, 5-min Apgar score, and small for gestational age. The scoring range is: 0-9 points for mild, 10-19 points for moderate, and >19 points for severe. The higher the score, the more serious the condition. See Appendix 2 for details.
[0064] 2.4 Classification of antimicrobial drugs
[0065] In 2015, the Ministry of Health issued the "Guidelines for the Clinical Application of Antimicrobial Drugs".
[62] According to the efficacy, safety and indications of antimicrobial drugs, they are divided into three categories: non-restricted, restricted and special use. At the same time, referring to the relevant provisions of the Classification Management Catalog of Clinical Application of Antimicrobial Drugs in Jiangsu Province and the Classification Management Catalog of Clinical Application of Antimicrobial Drugs in the Children's Hospital Affiliated to Soochow University, the classification standards for antimicrobial drugs are as follows:
[0066] Non-restricted antimicrobial drugs refer to antimicrobial drugs that have been proven safe, effective and relatively cheap through long-term clinical use, such as penicillin, amoxicillin, and flavopiridol.
[0067] Restricted antimicrobial drugs refer to antimicrobial drugs that have been proven safe and effective through long-term use, but have limitations in terms of safety, efficacy, impact on bacterial resistance, drug prices, etc., and need to be restricted in use based on patient conditions, populations, indications, or drug prescription volumes, such as ceftriaxone, latamoxef, amoxicillin / sulbactam, cefoperazone / sulbactam, azithromycin injection, etc.
[0068] Special-use antibiotics refer to newly marketed antibiotics that are expensive, have little clinical data on their efficacy or safety, or are not superior to currently used drugs, such as vancomycin, meropenem, imipenem, linezolid, etc.
[0069] 2.5 Study Groups
[0070] According to the BIPI diagnostic criteria, premature infants who developed BIPI were classified as the BIPI group, and among premature infants who did not develop BIPI, those whose general data matched those of the BIPI group were classified as the non-BIPI group. [43,44] , the BIPI group was divided into mild and severe BIPI subgroups; according to the abnormalities of cranial MRI
[39] , the BIPI group was divided into MRI abnormal group and MRI normal group.
[0071] 3. Research Methods
[0072] 3.1 Observational study on premature infants who met the inclusion criteria.
[0073] (1) Clinical data of premature infants and pregnant mothers were collected for comparison: ① General information of the two groups of premature infants: gender, gestational age, birth weight, head circumference, mode of delivery, 1-minute Apgar score, 5-minute Apgar score, twin or multiple pregnancy, IVF, small for gestational age; ② General information of the mothers of the two groups of premature infants: maternal age, parity, parity, amniotic fluid contamination, placental abruption, placenta previa, premature rupture of membranes ≥18h, cholestasis, gestational diabetes, gestational hypothyroidism, anemia during pregnancy ③ Primary diseases and complications of premature infants in the two groups: NRDS, NEC, BPD, PHN, respiratory failure, frequent apnea, ROP, EUGR, EOS, LOS, VAP, hyperbilirubinemia, infectious pneumonia, septic shock, central nervous system infection, hypoglycemia and its severity, duration of hypoglycemia, hyperglycemia, PNAC, anemia, patent ductus arteriosus with left-to-right shunt (patent ductusarteriosus, PDA), PDA ≥ 1.5 mm, atrial septal defect (ASD), pulmonary hypertension, acidosis, SNAPPE-Ⅱ score; ④ Treatment measures during hospitalization of the two groups of premature infants: hospitalization days, use of invasive mechanical ventilation, invasive mechanical ventilation time, oxygen therapy time, high-frequency ventilation, normal-frequency ventilation, oxygen concentration ≥ 30%, use of caffeine before weaning, rate of re-entry after failed weaning, PICC catheterization rate and duration, PICC complications (combined with blood-borne infection, tube detachment, tube blockage, rupture or tube failure, etc.), umbilical vein catheterization rate, intravenous nutrition time ≥ 30 days, number of blood transfusions (≥ 3 times), intravenous fluid glucose rate (1st, 3rd, 7th day), antibiotic use, porcine pulmonary surfactant use, and mouse nerve growth factor use; ⑤ Laboratory examinations of the two groups of premature infants at admission: white blood cell count (×10 9 / L), platelet count (×10 9 / L), platelet distribution width (%), platelet volume (%), PO2 (mmHg), PCO2 (mmHg), hematocrit (%), serum sodium (mmol / L), serum potassium (mmol / L), serum calcium (mmol / L), lactate (mmol / L), albumin (g / L), prealbumin (mg / L); ⑥ Nutritional supply of the two groups of premature infants during hospitalization: oral fluid intake, intravenous fluid intake, total fluid intake, enteral calorie intake, parenteral calorie intake, total calorie intake, time to reach full enteral nutrition, etc. on days 1, 3, 7, 14, and 21 after birth.
[0074] (2) After admission, the two groups of premature infants underwent peripheral venous puncture, and 0.5-1 ml of peripheral venous blood was collected on the 1st, 7th, and 14th days after birth. The blood was centrifuged at 3000 r / min for 5 minutes at room temperature of 20°C. The upper serum was placed in a dry EP tube and stored in a -80°C refrigerator within 1 hour for unified testing. In this experiment, the double antibody one-step sandwich enzyme-linked immunosorbent assay (ELISA) was used to measure the serum S100β protein and B-FABP levels, and the differences between the two groups were compared.
[0075] (3) All premature infants were given cranial ultrasound examinations on the 1st, 3rd, and 7th days after birth. If any abnormality was found, they were re-examined once a week. If no abnormality was found, they were re-examined once every 2 weeks until discharge. NBNA scoring was performed after the corrected gestational age reached 40 weeks. A cranial MRI examination was performed at the corrected gestational age of 34 to 38 weeks. The following measurements were made: lateral ventricular diameter (LVD), biparietal width (BPW), interhemispheric distance (IHD), third centricle of cerebrum diameter (V 3rd The differences in MRI brain structure measurements were compared between the BIPI group and the non-BIPI group, and between the BIPI group, the MRI normal group, the MRI abnormal group, and the non-BIPI group.
[0076] 3.2 Ultrasound examination of the head
[0077] Ultrasound doctors perform bedside cranial ultrasound examinations on premature infants using a Japanese TOSHIBA TA700 color Doppler ultrasound diagnostic instrument with a probe frequency of 3 MHZ. When the infant is asleep or quiet, the examiner performs continuous ultrasound scans on the infant's brain to observe the brain structure, blood flow, and abnormal echoes of brain tissue. After the examination, the ultrasound doctor judges the cranial ultrasound image.
[0078] 3.3 MRI of the head
[0079] The child was sedated with an intravenous injection of 10 mg / kg of sodium phenobarbital 15 to 30 minutes before the examination. After the child fell asleep, the imaging physician performed a cranial MRI examination on the child using a GE3.0T superconducting high magnetic resonance imaging device, selected the positive cross circle, and performed conventional MRI and DWI scans. The axial T1WI scanning parameters covering the entire head were: TE31ms, TR3316ms, FOV200mm, matrix 512×512, and slice thickness 5.0mm; the axial T2WI scanning parameters covering the entire head were TE103ms, TR6443ms, FOV200mm, matrix 512×512, slice thickness 5.0mm, and gap 5.0mm; the DWI scanning parameters were TE78ms, TR3954ms, FOV200mm, matrix 512×512, slice thickness 5.0mm, and gap 5.0mm. Measurement method [37,63] ① The left / right LVD (L-LVD / R-LVD) was measured on the level of the lateral ventricle in the coronal section of the brain, and the two were added together; ② The BPW and IHD were measured on the coronal section with the line connecting the bilateral cochlea and the basal osteotomy as the horizontal line; ③ V 3rd It is measured on the axial T1WI sequence at the anterior-posterior combined plane; ④ The maximum TCD is measured on the coronal section at the level of the lateral ventricle; ⑤ The cortical thickness is obtained by measuring the distance between the inner and outer surfaces of the cortex (the inner surface of the cortex is the boundary between white matter and gray matter, and the outer surface is the boundary between gray matter and cerebrospinal fluid); ⑥ The thickness of the corpus callosum is measured at the knee, middle and splenium of the midsagittal section of the brain, and the average value is calculated; see for details. Figure 1 After the examination, the radiologist reads the film and makes a diagnosis without knowing the child.
[0080] 3.4 NBNA Rating
[0081] After the corrected gestational age of premature infants reaches 40 weeks, the rehabilitation physician will perform a 20-item NBNA score. The examination is performed between two feedings of the infant, and the environment must be quiet, the light is semi-dark, and the room temperature is set between 22℃ and 27℃. The examiner should complete the examination within 10 minutes. After the examination, the rehabilitation physician will score the infant.
[0082] 3.5 Specimen testing
[0083] (1) Take out the sample to be tested from the -80℃ freezer and thaw it at room temperature for 20 minutes. At the same time, dilute 20× washing buffer with distilled water at a ratio of 1:20.
[0084] (2) Set up standard wells and sample wells on the strips. Add 50 μL of standard of different concentrations to the standard wells, add 50 μL of the specimen to be tested to the sample wells, and do not add anything to the blank wells.
[0085] (3) Add 100 μL of horseradish peroxidase-labeled detection antibody to each well except the blank well, seal the wells with a sealing film, and incubate in a 37°C incubator for 60 min.
[0086] (4) Discard the liquid and pat dry on absorbent paper. Fill each well with 350 μL of washing solution. After 1 minute, shake off the washing solution and pat dry. Repeat this process 5 times.
[0087] (5) Add 50 μL of substrate A and B to each well and incubate at 37°C in the dark for 15 min.
[0088] (6) Add 50 μL of stop solution to each well and measure the OD value of each well at a wavelength of 450 nm using an ELISA reader within 15 minutes.
[0089] (7) Use ELISA Calc software to draw a standard curve with the OD value of the standard as the horizontal axis and the concentration value of the standard as the vertical axis. Obtain a linear regression equation. Substitute the OD value of the sample into the equation to calculate the concentration of the sample (ng / ml).
[0090] 4. Main experimental reagents and equipment
[0091]
[0092] 5. Statistical methods
[0093] SPSS 25.0 statistical software was used to analyze all data. The measured data were normal distribution and presented as mean ± standard deviation ( ±s), independent sample t test was used for comparison between two groups, and Levene method was used for comparison among three groups to test whether variance homogeneity was satisfied. If it was satisfied, variance analysis was used, and if it was not satisfied, Krusal-Wallis H test was used. If it did not conform to normal distribution, the median (25th and 75th percentiles) [M(P 25 , P 75 )], Mann Whitney U test was used for comparison between two groups, and Krusal-Wallis H test was used for comparison among three groups; count data were expressed as percentage (%), and chi-square test or continuity-corrected chi-square test was used for comparison between two groups (n≥40, 1≤T min ≤5) analysis; Logistic regression analysis was performed on the factors with statistical differences in univariate analysis to obtain independent risk factors for BIPI. P<0.05 was considered statistically significant; the ROC curve was drawn to evaluate the predictive efficiency of the indicators and the area under the curve was calculated; Youden index = sensitivity + specificity - 1, and the maximum Youden index was used to calculate the critical value; Pearson correlation analysis was used for continuous variables, and Spearman correlation analysis was used for categorical variables to obtain the correlation coefficient r, P <0.05 is considered to be a correlation between the two variables.
[0094] result
[0095] 1. Basic clinical information and data
[0096] 1.1 Comparison of general conditions between the two groups of premature infants
[0097] During the study, a total of 238 premature infants with gestational age less than 37 weeks were observed. Nine cases died within a few hours after active rescue or were discharged automatically, seven cases had genetic abnormalities, two cases had chromosomal abnormalities, eight cases were transferred to surgery during hospitalization, one case was congenital small intestinal atresia, and one case was esophageal atresia. Among them, 43 cases were in the BIPI group. Among the 167 premature infants without BIPI, 77 premature infants whose general data matched those of the BIPI group were included as the non-BIPI group. There were 120 cases in the two groups, including 54 males and 66 females, with an average gestational age of (32.03±2.18) weeks and an average birth weight of (1870.3±548.83) g. There were 32 cases with mild brain damage and 11 cases with severe brain damage in the BIPI group; 14 cases were in the abnormal MRI group and 29 cases were in the normal MRI group. There were no significant differences between the BIPI group and the non-BIPI group in terms of gender, gestational age, birth weight, head circumference, mode of delivery, whether it was a test-tube baby, whether it was multiple or twin, whether it was small for gestational age, and whether the 1-min and 5-min Apgar scores were ≤7 points ( P >0.05), see Table 1 for details.
[0098]
[0099] Note: a is the chi-square test result 2 Value, b is the result of t test t value.
[0100] 1.2 Comparison of high-risk factors between the two groups of premature infants
[0101] 1.2.1 Comparison of general conditions of mothers of premature infants in the two groups during pregnancy
[0102] There were no significant differences in maternal age >35 years or <20 years, parity, parity, amniotic fluid contamination, placental abruption, placenta previa, premature rupture of membranes ≥18h, cholestasis, gestational diabetes, gestational hypothyroidism, gestational anemia, severe preeclampsia, antenatal use of glucocorticoids, antenatal use of magnesium sulfate, and antenatal infection between the two groups of premature infants ( P >0.05), see Table 2 for details.
[0103]
[0104] Note: a is the chi-square test result 2 value, c is the chi-square test result of continuity correction 2 Value, —— cannot be estimated.
[0105] 1.2.2 Comparison of primary diseases and complications between the two groups of premature infants
[0106] Compared with the non-BIPI group, the BIPI group had significantly higher incidence of NRDS, BPD, PHN, respiratory failure, frequent apnea, EUGR, EOS, VAP, septic shock, central nervous system infection, hypoglycemia, severe hypoglycemia, duration of hypoglycemia, anemia, acidosis, and SNAPPE-Ⅱ scores than the non-BIPI group ( P <0.05); however, there were no significant differences between the two groups in NEC, ROP, LOS, mild hypoglycemia, moderate hypoglycemia, hyperglycemia, hyperbilirubinemia, infectious pneumonia, PNAC, left-to-right shunt PDA, PDA ≥ 1.5 mm, ASD, and pulmonary hypertension ( P >0.05), see Table 3 for details.
[0107] Table 3 Comparison of primary diseases and complications between the two groups of premature infants
[0108]
[0109] Note: a is the chi-square test result 2 value, c is the chi-square test result of continuity correction 2 Value, d is the result of Mann-Whitney U test Z Value, —— cannot be estimated.
[0110] 1.2.3 Comparison of treatment between the two groups of premature infants during hospitalization
[0111] The BIPI group had significantly more days of hospitalization, use of invasive mechanical ventilation, duration of invasive mechanical ventilation, oxygen therapy duration, high-frequency ventilation, oxygen concentration ≥30%, use of caffeine before weaning, intravenous nutrition time ≥30 days, sugar rate on the first day, blood transfusion ≥3 times, use of porcine lung surfactant, and use of special antimicrobial drugs than the non-BIPI group, and the differences were statistically significant ( P <0.05); there were no significant differences between the two groups in conventional ventilation, re-ventilation rate after failed weaning, PICC placement, PICC time, PICC complications, umbilical vein placement, glucose rate on the 3rd and 7th days, use of non-restricted antibiotics, and use of restricted antibiotics ( P >0.05), see Table 4 for details.
[0112]
[0113] Note: a is the chi-square test result 2 Value, b is the result of t test t value, c is the chi-square test result of continuity correction 2 Value, d is the result of Mann-Whitney U test Z value.
[0114] 1.2.4 Comparison of laboratory test results between the two groups of premature infants
[0115] In the laboratory test results on the day of admission, the two groups had white blood cell counts>20×10 9 / L ratio and its median level, platelet> 300×10 9 There were significant differences in the proportion and median level of platelet distribution width>16.5%, the proportion and median level of platelet hematocrit>0.27, hematocrit, lactate, albumin, PO2≤60mmHg, and PCO2≤24mmHg ( P <0.05); the two groups had white blood cells <4×10 9 / L ratio, platelets <100×10 9 There was no significant difference in the proportion of / L, the proportion of platelet distribution width <12% and its median level, the proportion of platelet hematocrit <0.1, serum sodium, serum potassium, serum calcium, prealbumin, and PCO2>50mmHg ( P >0.05), see Table 5 for details.
[0116]
[0117] Note: a is the chi-square test result 2 Value, b is the result of t test t Value, d is the result of Mann-Whitney U test Z Value, —— cannot be estimated.
[0118] 1.3 Comparison of nutritional status between the two groups of premature infants
[0119] 1.3.1 Comparison of enteral nutrition between the two groups of premature infants
[0120] The start time of enteral feeding and the time of complete enteral feeding in the BIPI group were significantly later than those in the non-BIPI group, and the difference was statistically significant ( P <0.05), see Table 6 for details.
[0121]
[0122] Note: d is the result of Mann-Whitney U test Z value.
[0123] 1.3.2 Comparison of fluid intake between the two groups of premature infants
[0124] The oral fluid volume and total fluid volume of the BIPI group on the third day after birth were significantly less than those of the non-BIPI group, and the intravenous fluid intake on the 7th and 14th days was significantly higher than that of the non-BIPI group, and the difference was statistically significant ( P <0.05); there was no significant difference in the amount of intravenous fluid on the 3rd day, the amount of oral fluid on the 7th and 14th days, and the total fluid volume between the two groups ( P >0.05), see Table 7 for details.
[0125]
[0126] Note: b is the result of t test t Value, d is the result of Mann-Whitney U test Z value.
[0127] 1.3.3 Comparison of calorie intake between the two groups of premature infants
[0128] The enteral calories and total calories in the BIPI group were significantly lower than those in the non-BIPI group on the third day after birth, and the parenteral calories on the 7th and 14th days after birth were significantly higher than those in the non-BIPI group. The differences were statistically significant ( P <0.05); there was no significant difference in parenteral calories on the third day after birth, enteral calories on the 7th and 14th days, and total calories between the two groups ( P >0.05). See Table 8 for details.
[0129]
[0130] Note: b is the result of t test t Value, d is the result of Mann-Whitney U test Z value.
[0131] 1.4 Logistic multivariate regression analysis of BIPI in the two groups of premature infants
[0132] The occurrence of BIPI was taken as the dependent variable, and the factors with statistically significant differences after univariate comparison in the above clinical data were taken as independent variables for logistic multivariate regression analysis. The results showed that the duration of invasive mechanical ventilation (OR=2.135, 95%CI: 1.253-3.639), acidosis (OR=9.943, 95%CI: 1.782-55.481), frequent apnea (OR=7.950, 95%CI: 1.697-37.241), and platelet count>0.27 (OR=28.207, 95%CI: 2.286-296.228) were independent risk factors for BIPI in premature infants after birth, as shown in Table 9 for details.
[0133]
[0134] 2. Comparison of S100β protein and B-FABP in the two groups of premature infants
[0135] 2.1 Comparison of serum S100β protein between the two groups of premature infants
[0136] ① Comparison of results between groups: The serum S10β protein level of premature infants in the BIPI group was significantly higher than that in the non-BIPI group on the first day after birth, and the difference was statistically significant ( P <0.05); however, there was no significant difference in serum S100β protein levels between the two groups on days 7 and 14 after birth ( P >0.05); see Table 10 for details.
[0137] ② Comparison of results within the group: There was no significant difference in serum S100β protein levels on days 1, 7, and 14 in the non-BIPI group ( P >0.05). The serum S100β protein level in the BIPI group on the first day after birth was significantly higher than that on the 7th and 14th days, and the difference was statistically significant ( P <0.05), there was no significant difference in serum S100β protein levels on days 7 and 14 ( P >0.05); see Table 10 for details.
[0138]
[0139] Note: *: Compared with 14 days after birth, P <0.05; b is the result of t test t Value, m is obtained from variance analysis F value.
[0140] 2.2 Prediction of BIPI by serum S100β protein
[0141] The ROC curve was drawn with the serum S100β protein level (ng / ml) on the first day after birth as the variable and non-BIPI as the reference to evaluate the predictive efficacy of serum S100β protein level for BIPI. The AUC was 0.931 (95%CI: 0.888-0.973, P <0.05), when the critical value is 6.325ng / ml, the sensitivity is 95% and the specificity is 81%, see Figure 2 .
[0142] 2.3 Comparison of serum B-FABP levels between the two groups of premature infants
[0143] ① Comparison of results between groups: The serum B-FABP levels of premature infants in the BIPI group were significantly higher than those in the non-BIPI group on the first and seventh days after birth, and the difference was statistically significant ( P <0.05); however, there was no significant difference in serum B-FABP levels between the two groups on day 14 after birth ( P >0.05); see Table 11 for details.
[0144] ② Comparison of results within the group: There was no significant difference in serum B-FABP levels of premature infants in the non-BIPI group on days 1, 7, and 14 after birth ( P >0.05). The serum B-FABP levels of premature infants in the BIPI group on the first and seventh days after birth were significantly higher than those on the 14th day, and the difference was statistically significant ( P <0.05); however, there was no significant difference in serum B-FABP levels between the first day and the seventh day (P>0.05); see Table 11 for details.
[0145]
[0146] Note: *: Compared with 14 days after birth, P <0.05. b is the result of t test t Value, m is obtained from variance analysis F value.
[0147] 2.4 Prediction of BIPI by serum B-FABP
[0148] The ROC curves were drawn with serum B-FABP levels on the first and seventh days after birth as variables and non-BIPI as reference to evaluate the predictive efficacy of serum B-FABP levels on the first and seventh days for the occurrence of BIPI. The AUCs were 0.876 (95%CI: 0.808-0.945, P <0.05), 0.902 (95%CI: 0.845~0.909, P<0.05), when the critical values were 12.14ng / ml and 12.32ng / ml, the sensitivity was 79% and 79%, and the specificity was 84% and 90%, respectively. Figure 2 .
[0149] 3. Comparison of MRI brain structure measurement results between the two groups of premature infants
[0150] In this study, there were 14 premature infants with abnormal MRI in the BIPI group, accounting for 32.6% (14 / 43), including 2 cases of cystic PVL, 3 cases of diffuse WMD, 1 case of PVL+PVH, 2 cases of PVL+ventriculomegaly, 1 case of PVHI, 3 cases of severe HIE, 1 case of IVH+ventriculomegaly+hydrocephalus, and 1 case of PVH-IVH+SAH. There were 29 premature infants with normal MRI in the BIPI group, accounting for 67.4% (29 / 43).
[0151] 3.1 Comparison of MRI brain structure measurements in different groups of premature infants
[0152] By measuring the brain structure of MRI images, it was found that the LVD, IHD, and V 3rd The corpus callosum thickness was significantly smaller than that of the non-BIPI group, and the difference was statistically significant ( P <0.05); there was no significant difference in BPW, TCD, and cortical thickness between the BIPI group and the non-BIPI group ( P >0.05), see Table 12 for details.
[0153]
[0154] Note: b is the result of t test t Value, d is the result of Mann-Whitney U test Z value.
[0155] The BIPI group was further divided into the BIPI group premature infants with abnormal MRI and the BIPI group premature infants with normal MRI. The results showed that the LVD, IHD, and V 3rd The corpus callosum thickness was significantly smaller than that of the non-BIPI group ( P <0.05), while the differences in LVD, IHD, V 3rd There was no significant difference in the thickness of the corpus callosum ( P >0.05); in the BIPI group, there were no significant differences in BPW, TCD, and cortical thickness among the abnormal MRI group, the normal group, and the non-BIPI group ( P >0.05), see Table 13 for details.
[0156]
[0157] Note: #: Compared with the non-BIPI group, P <0.05; e is obtained by Kruskal-Wallis H test H Value, m is obtained from variance analysis F value.
[0158] 3.2 Prediction of BIPI by MRI brain structure measurements
[0159] ① Using LVD as a variable and non-BIPI as a reference, the ROC curve was drawn to evaluate the predictive efficacy of LVD for BIPI. The results were as follows: AUC was 0.625 (95% CI: 0.519-0.731, P <0.05), when the critical value was 13.75 mm, the sensitivity was 60.5% and the specificity was 66.2%, see Figure 3 .
[0160] ② Using IHD as a variable and non-BIPI as a reference, the ROC curve was drawn to evaluate the predictive efficacy of IHD for BIPI. The results were as follows: AUC was 0.677 (95% CI: 0.571-0.782, P <0.05), when the critical value was 1.65mm, the sensitivity was 65.9% and the specificity was 67.6%, see Figure 3 .
[0161] ③ With V 3rd As a variable, the ROC curve was drawn with non-BIPI as the reference to evaluate V 3rd The results of the prediction of BIPI were as follows: AUC was 0.642 (95% CI: 0.536-0.749, P <0.05), when the critical value was 2.35mm, the sensitivity was 32.6% and the specificity was 91.9%. Figure 3 .
[0162] ④ The ROC curve was drawn with corpus callosum thickness as the variable and non-BIPI as the reference to evaluate the predictive efficacy of corpus callosum thickness on BIPI. The results were as follows: AUC was 0.606 (95% CI: 0.498-0.715, P <0.05), when the critical value was 2.31 mm, the sensitivity was 73.0% and the specificity was 48.8%, see Figure 3 .
[0163] 4. Indicators combined with Logistic regression to build a BIPI prediction model
[0164] The BIPI prediction model was constructed by Logistic regression with the invasive mechanical ventilation time (days), acidosis, frequent apnea, platelet count>0.27, serum S100β protein level (ng / ml) on the first day after birth, serum B-FABP level (ng / ml) on the first and seventh days after birth, LVD (mm), and IHD (mm) as independent variables. The regression equation was: P =1 / [1+e-(-28.179)+0.458X1+1.028X2+1.932X3+1.401X4+1.489X5+0.595X6+0.410X7+0.429X8+1.740X9], where X1, X2, X3, X4, X5, X6, X7, X8, X9 are invasive mechanical ventilation time (days), acidosis, frequent apnea, platelet count>0.27, serum S100β protein level on the first day after birth (ng / ml), serum B-FABP level on the first and 7 days after birth (ng / ml), LVD (mm), IHD (mm), respectively. P is the probability of the Logistic model. The chi-square value of the Logistic regression prediction model is 2.689. P <0.001, indicating that the above indicators have explanatory power for the prediction of BIPI. Hosmer-Lemeshow was calculated using the classification interaction table, df=8, P >0.05, indicating that the prediction model has good goodness of fit. Taking the predicted variable as the test variable, the results show that the area under the ROC curve of the above 9 indicators combined with the Logistic regression model is 0.998, which is greater than the prediction of each indicator alone. See Table 14, Table 15, Figure 4 .
[0165]
[0166]
[0167] Neurodevelopmental disorders caused by BIPI are still the main problem left over by surviving premature infants. At present, there is no unified conclusion on the causes of BIPI at home and abroad. Foreign studies have shown that gestational age and birth weight are independent risk factors for BIPI and are negatively correlated with the incidence rate. Other risk factors include male, twin or multiple pregnancy, hypoglycemia, chorioamnionitis, low Apgar score, etc. It is currently believed that the occurrence of BIPI is related to postnatal hypoxia, ischemia, infection, malnutrition, glucose metabolism and other factors in premature infants.
[0168] Frequent apnea and acidosis are independent risk factors for BIPI proposed in the present invention. The reasons are as follows: frequent apnea causes alveolar collapse at the end of expiration, reducing gas exchange, and long-term apnea causes a decrease in PO2 and an increase in PCO2, which reflexively causes bradycardia and hypoxemia. Hypoxia can cause abnormal discharges in the cerebral cortex, inducing or aggravating brain damage. At the same time, hypoxia causes an increase in PCO2 and a decrease in pH, and energy metabolism disorders cause acidic substances to accumulate, resulting in metabolic acidosis. Metabolic acidosis can increase capillary permeability and vascular fragility, further affecting the autonomic regulation of cerebral blood flow, causing hemodynamic disorders in the brain, and thus causing brain damage.
[0169] The concept that antenatal application of magnesium sulfate may reduce the risk of cerebral palsy was first proposed by Nelson et al. in the 1990s. Subsequently, Doyle LW et al. used magnesium sulfate to evaluate its neuroprotective effect on pregnant women at risk of premature birth. The results showed that the risk of cerebral palsy in premature infants who used magnesium sulfate antenatally was reduced by nearly 1 / 3 compared with the control group. However, the mechanism by which magnesium sulfate exerts its neuroprotective effect is unclear. The reason may be that magnesium sulfate can block glutamate receptors, and the vasodilator effect of magnesium sulfate can increase uterine blood perfusion, thereby reducing the incidence of perinatal hypoxic-ischemic brain damage. However, there was no significant difference in the antenatal application of magnesium sulfate between the two groups of premature infants in the present invention.
[0170] A large number of research experiments have shown that infection plays an increasingly important role in the occurrence of BIPI. Hassanein et al. showed that IL-6 ≥ 1.077x10 -7 It is an independent risk factor for PVL. The reason is that the release of inflammatory factors increases after infection, affecting the formation of cerebral blood vessels, thereby causing hemodynamic disorders. Studies at home and abroad have found that EOS can increase the risk of BIPI. The present invention proposes that EOS and septic shock are both related to the occurrence of BIPI, but EOS and white blood cell counts > 20×10 9 / L is not an independent risk factor for BIPI.
[0171] The present invention discloses for the first time the relationship between serum S100β protein and BIPI. The results show that the serum S100β protein level of premature infants in the BIPI group was significantly higher than that in the non-BIPI group on the first day after birth. There was no significant difference in the serum S100β protein level between the two groups on the 7th day after birth. At the same time, the change in the S100β protein level was positively correlated with the severity of BIPI. S100β protein is mainly present in the glial cells of the central and peripheral nervous systems. Under normal circumstances, S100β protein cannot penetrate the blood-brain barrier and be released into the peripheral blood. When brain damage occurs in premature infants, glial cells are necrotic and the blood-brain barrier is damaged. S100β protein penetrates the blood-brain barrier and is released into the peripheral blood and excreted through the kidneys with a half-life of about 30 to 100 minutes. The ROC curve in this study showed that the area under the curve of serum S100β protein for predicting BIPI was 0.931. When the S100β protein level was 6.325 ng / ml, the sensitivity of predicting the occurrence of BIPI was 95.3% and the specificity was 81.8%, indicating that serum S100β protein has a good predictive effect on the occurrence of BIPI. The present invention shows that the serum B-FABP level of premature infants in the BIPI group gradually increased and reached a peak on the first day after birth, and remained at a high level on the seventh day after birth, and then showed a downward trend, and the change in serum B-FABP level was positively correlated with the severity of BIPI. The ROC curve showed that the AUC of serum B-FABP level on the first day and the seventh day for predicting the occurrence of BIPI were 0.876 and 0.902, respectively. When the critical values were 12.14 ng / ml and 12.32 ng / ml, respectively, the sensitivity was 79% and 79%, and the specificity was 84% and 90%, respectively, indicating that B-FABP is a serum biomarker with a longer peak duration than S100β protein, but its ROC predictive efficiency is not as good as that of S100β protein, and large-sample, multi-center studies are still needed to verify its value in predicting BIPI.
[0172] Kidokoro performed MRI scans on 97 premature infants with a gestational age of less than 30 weeks and found that the IHD and LVD of PVL children were significantly increased compared with the control group, and the TCD was reduced. Studies have found that enlargement of the lateral ventricles and reduction of cerebellar volume are associated with neurodevelopmental disorders, among which cerebellar damage can cause cognitive and behavioral defects in children. In addition to being related to hypoxia, ischemia and other causes, supratentorial hemorrhage can cause blood decomposition products (heme and iron) to deposit in the cerebrospinal fluid. Iron deposition and catalysis of reactive oxygen to cause free radical attacks can also cause cerebellar atrophy and TCD reduction. There was no significant difference in TCD between the two groups of premature infants in the present invention.
[0173] The present invention discloses a new joint prediction model for BIPI in premature infants. ROC curve analysis shows that with invasive mechanical ventilation time (days), acidosis, frequent apnea, platelet count>0.27, serum S100β protein level (ng / ml) on the first day after birth, serum B-FABP level (ng / ml) on the first day and 7 days after birth, LVD (mm), and IHD (mm) as independent variables, the areas under the curve for predicting the occurrence of BIPI are 0.755, 0.611, 0.696, 0.657, 0.931, 0.876, 0.902, 0.646, and 0.677, respectively; Logistic regression analysis was further performed in combination with multiple indicators, and the results showed that the area under the ROC curve of the joint prediction model was 0.998, the sensitivity was 97.6%, and the specificity was 98.6%, all of which were greater than the prediction of the 9 indicators alone. This indicates that the model can better predict the occurrence of BIPI, and can help to identify and diagnose BIPI early in clinical practice, improve the prognosis of children with BIPI, and reduce the possibility of adverse neurodevelopmental outcomes caused by BIPI.
[0174] In summary, the occurrence of BIPI is related to multiple factors. Infection, hypoxia, ischemia, malnutrition, abnormal glucose metabolism, and changes in blood viscosity can all lead to the occurrence of BIPI. Changes in the levels of serum markers S100β protein and B-FABP are closely related to the occurrence of BIPI and are positively correlated with its severity. MRI brain LVD, IHD, V 3rd The measurement of corpus callosum thickness can provide a possibility for MRI to predict the occurrence of BIPI from a quantitative perspective, and the combined construction of a regression model with multiple indicators can better predict the occurrence of BIPI.
Claims
1. A prediction model for brain damage in premature infants, characterized in that: The regression equation was: P=1 / [1+e-(-28.179)+0.458X1+1.028X2+1.932X3+1.401X4+1.489X5+0.595X6+0.410X7+0.429X8+1.740X9], where X1 was the duration of invasive mechanical ventilation, in days; X2, X3, and X4 were acidosis, frequent apnea, and platelet count>0.27, respectively; X5 was the serum S100β protein level on the first day after birth, in ng / ml; X6 and X7 were the serum B-FABP levels on the first and seventh days after birth, respectively, in ng / ml; X8 was the diameter of the lateral ventricle, in mm; X9 was the interhemispheric distance, in mm; and P was the logistic model probability.
2. A prediction system for premature infant brain damage, comprising a prediction module, a data input module and a prediction output module, characterized in that: The prediction module includes the prediction model for premature brain damage according to claim 1.
3. A computer equipped with the system for predicting brain damage in premature infants as claimed in claim 2.
4. The computer according to claim 3, characterized in that Including monitor, keyboard, and hard disk.
Citation Information
Patent Citations
Application of intestinal microorganism as BIPI (Brain Injury in Premature Infants) marker
CN112760368A
