METHOD OF PROGNOSIS OF BRONCHOPULMONARY DYSPLASIA IN PREMATURE INFANTS
20230040391 · 2023-02-09
Inventors
Cpc classification
A61B8/5223
HUMAN NECESSITIES
A61K38/30
HUMAN NECESSITIES
G16H50/20
PHYSICS
A61B8/5207
HUMAN NECESSITIES
A61K31/58
HUMAN NECESSITIES
G16H50/30
PHYSICS
International classification
A61K31/58
HUMAN NECESSITIES
G16H50/20
PHYSICS
Abstract
The present invention relates to a method for prognosis the risk of bronchopulmonary dysplasia in premature infant by adjusting a score obtained from lung ultrasound with the gestational age of the infant.
Claims
1. An ex vivo method for determining whether a premature infant is having or is susceptible to have bronchopulmonary dysplasia (BPD) comprising: (a) obtaining a score from a lung ultrasound image of the infant, wherein images from the lung ultrasound are graded, wherein the lung ultrasound images contain images from the upper-anterior, lower-anterior and lateral areas, and wherein the score is obtained by combining the grades from the ultrasound images in a function; (b) obtaining gestational age of the infant; and (c) obtaining an end value by dividing the score of (a) by the gestational age of (b), wherein the infant is classified as at risk of having bronchopulmonary dysplasia when the end value is higher than a predetermined threshold value.
2. The method of claim 1, wherein the infant is 7 days old.
3. The method of claim 1, wherein the infant is 14 days old.
4. The method of claim 1, wherein the score of (a) is obtained by providing a grade from 0 to 3 to each of the 3 areas (upper anterior, lower anterior, and lateral areas) of each lung of the infant, wherein 0 indicates A-pattern, corresponding to the presence of the only A-lines in the lung area, 1 corresponds to B-pattern, defined as the presence of at least 3 well-spaced B-lines), 2 corresponds to a severe B pattern, defined as the presence of crowded and coalescent B lines with or without consolidations limited to the subpleural space, and 3 corresponds to a lung with extended consolidations, and summing all obtained grades to obtain a total score ranging from 0 to 18.
5. The method of claim 1, wherein the score of (a) is obtained by providing a grade from 0 to 3 to each of the 5 areas (upper anterior, lower anterior, upper posterior and lower posterior and lateral areas) of each lung of the infant, wherein 0 indicates A-pattern, corresponding to the presence of the only A-lines in the lung area, 1 corresponds to B-pattern, defined as the presence of at least 3 well-spaced B-lines), 2 corresponds to a severe B pattern, defined as the presence of crowded and coalescent B lines with or without consolidations limited to the subpleural space, and 3 corresponds to a lung with extended consolidations, and summing all obtained grades to obtain a total score ranging from 0 to 30.
6. The method of claim 1, wherein the infant is 7 days old, wherein the score of (a) is obtained by providing a grade from 0 to 3 to each of the 3 areas (upper anterior, lower anterior, and lateral areas) of each lung of the infant, wherein 0 indicates A-pattern, corresponding to the presence of the only A-lines in the lung area, 1 corresponds to B-pattern, defined as the presence of at least 3 well-spaced B-lines), 2 corresponds to a severe B pattern, defined as the presence of crowded and coalescent B lines with or without consolidations limited to the subpleural space, and 3 corresponds to a lung with extended consolidations, and summing all obtained grades to obtain a total score ranging from 0 to 18 and wherein the threshold value is 0.23.
7. The method of claim 1, wherein the infant is 7 days old, wherein the score of (a) is obtained by providing a grade from 0 to 3 to each of the 5 areas (upper anterior, lower anterior, upper posterior and lower posterior and lateral areas) of each lung of the infant, wherein 0 indicates A-pattern, corresponding to the presence of the only A-lines in the lung area, 1 corresponds to B-pattern, defined as the presence of at least 3 well-spaced B-lines), 2 corresponds to a severe B pattern, defined as the presence of crowded and coalescent B lines with or without consolidations limited to the subpleural space, and 3 corresponds to a lung with extended consolidations, and summing all obtained grades to obtain a total score ranging from 0 to 30 and wherein the threshold value is 0.43.
8. The method of claim 1, wherein the infant is 14 days old, wherein the score of (a) is obtained by providing a grade from 0 to 3 to each of the 3 areas (upper anterior, lower anterior, and lateral areas) of each lung of the infant, wherein 0 indicates A-pattern, corresponding to the presence of the only A-lines in the lung area, 1 corresponds to B-pattern, defined as the presence of at least 3 well-spaced B-lines), 2 corresponds to a severe B pattern, defined as the presence of crowded and coalescent B lines with or without consolidations limited to the subpleural space, and 3 corresponds to a lung with extended consolidations, and summing all obtained grades to obtain a total score ranging from 0 to 18 and wherein the threshold value is 0.31.
9. The method of claim 1, wherein the infant is 14 days old, wherein the score of (a) is obtained by providing a grade from 0 to 3 to each of the 5 areas (upper anterior, lower anterior, upper posterior and lower posterior and lateral areas) of each lung of the infant, wherein 0 indicates A-pattern, corresponding to the presence of the only A-lines in the lung area, 1 corresponds to B-pattern, defined as the presence of at least 3 well-spaced B-lines), 2 corresponds to a severe B pattern, defined as the presence of crowded and coalescent B lines with or without consolidations limited to the subpleural space, and 3 corresponds to a lung with extended consolidations, and summing all obtained grades to obtain a total score ranging from 0 to 30 and wherein the threshold value is 0.59.
10. A method for treating a premature infant that is having or is susceptible to have bronchopulmonary dysplasia (BPD), comprising performing the method of claim 1, and providing appropriate care when the end value obtained from the method is above the predetermined cutoff value.
11. The method of claim 10, wherein the appropriate care comprises administering Vitamin A to the infant.
12. The method of claim 10, wherein the appropriate care comprises administering Insulin-like Growth Factor-to the infant.
13. The method of claim 10, wherein the appropriate care comprises administering surfactant protein D to the infant.
14. The method of claim 10, wherein the appropriate care comprises administering a steroid to the infant.
15. The method of claim 14, wherein steroid is administered by intra-tracheal instillation.
16. The method of claim 14, wherein the steroid is budesonide.
Description
FIGURES
[0068]
[0069]
[0070]
Abbreviations: BPD: bronchopulmonary dysplasia; D7: seventh day of life; D14: fourteenth days of life; eLUS: extended lung ultrasound score; LUS: lung ultrasound score.
[0071]
EXAMPLES
Material and Methods
[0072] Study Design
[0073] A multicenter, pragmatic, international, observational, non-invasive, prospective, diagnostic accuracy study was conducted in five academic tertiary referral neonatal intensive care units (NICU) in France and Italy. The NICU at Paris Saclay University Hospital served as coordinating center. The study was approved by local ethical boards and parental/guardian consent was obtained following local regulations. The study was registered in the ISRCTN Registry and details are available there. The study was pragmatic as the participation did not change the clinical management, which was provided according to local NICU protocols, essentially based on optimal prenatal care and international guidelines for neonatal resuscitation and respiratory management of preterm neonates. Participating NICU teams are proficient in lung ultrasound and routinely use the technique in their clinical care, according to clinicians' evaluation.
[0074] Patients
[0075] All extremely preterm inborn neonates with gestational age 30 weeks whose parents agreed to participate were considered eligible for the study, if they do not have any of the following a priori exclusion criteria: 1) complex congenital malformations; 2) chromosomal abnormalities; 3) pulmonary hypoplasia; 4) congenital anomalies of surfactant proteins or any other suspected congenital lung disorders. At the end of recruitment, all data were sent to the coordinating centre, merged and reviewed to check for completeness and accuracy. Local investigators were contacted if clarification or more data were needed and infants were excluded post hoc in case of: 1) death or transfer to other hospitals before 36 weeks post-menstrual age; 2) missing data needed for the BPD diagnosis.
[0076] Data Collection
[0077] Data were prospectively collected into customized, secured, electronic spreadsheets by local investigators in each center. Data were completely anonymous in accordance with local and European privacy regulations, with local investigators maintaining an identification log. Basic demographics and common clinical data obtained during routine care were collected. Lung ultrasound, ventilatory and gas exchange data at definite timepoints were also registered (see below). A detailed list of collected data and standardised definitions used in the study is available in the ISRCTN Registry.
[0078] Lung Ultrasound Protocol
[0079] Lung ultrasound was performed upon NICU admission (day 0 (D0)) and at seven (D7), fourteen (D14) and twenty-eight (D28) days of postnatal age. D0 lung ultrasound was performed upon NICU admission and always before surfactant administration, if any. A final ultrasound was performed at 36 weeks post-menstrual age (36W). Lung ultrasound was performed in some centers with “hockey stick” micro-linear (15 MHz) and in others with a broadband linear (10 MHz) probe, according to the availability. At each time-point, a LUS specifically created and validated for newborn infants was calculated in real-time: this score is based on classical lung ultrasound semiology and is calculated over 6 chest areas (3 per each side, ranging from 0 to 18), as we previously described (Brat et al, op. cit). Additionally, investigators using a micro-linear probe also calculated an extended score (eLUS) over 10 chest areas (5 per each side, ranging from 0 to 30) including the scan of the upper posterior and lower posterior chest areas. Lung ultrasound patterns used for score calculation are described in the study definitions and illustrative examples are provided in Brat et al (op. cit.). Scans were performed by automatically adjusting the gain; depth and focus were set according to patients' size and the sign of interest. Lung ultrasound was performed in incubators, when the neonate was quiet and lying supine or slightly tilted to scan the posterior zones, during routine clinical care to minimize discomfort. Within 1 hour from lung ultrasound, if the patient had normal temperature and peripheral perfusion, blood gases (transcutaneous partial pressure of oxygen [PtcO.sub.2] and carbon dioxide [PtcCO.sub.2]) were measured with adequately calibrated transcutaneous devices (TCM4®, Radiometer Medical, Copenhagen, Denmark), used according to the
[0080] American Association for Respiratory Care guidelines and manufacturer's recommendations. Probes were applied until the achievement of a stable measurement and anyway for a maximum of 15 minutes, and, at the same time, mean airway pressure (P.sub.aw) and vital parameters were recorded. During transcutaneous measurements, ventilatory parameters were not changed and, for neonates receiving non-invasive respiratory support, pressure leaks have been minimized, by using appropriately sized interfaces and closing the mouth with gentle pressure on the jaw.
[0081] Outcomes
[0082] The primary outcomes were: (1) to efficaciously monitor lung aeration in neonates with CPIP by describing the relationship between lung ultrasound scores and gas exchange at different time points; 2) to demonstrate accuracy of LUS to predict BPD at 36 weeks post-menstrual age. Secondary outcome was to compare the performance of the classical and extended LUS to monitor lung aeration and predict BPD. BPD was diagnosed according to Jobe and Bancalari's criteria (Jobe and Bancalari. Am J Respir Crit Care Med. 2001; 163(7):1723-1729) by a clinician blinded to LUS data.
[0083] Calculations and Statistics Interim Analysis
[0084] The following indices were calculated to describe oxygenation (Brat et al (op. cit.)): (1) PtcO.sub.2 to FiO.sub.2 (P/F) ratio; (2) Alveolar-arterial gradient=PA—PtcO.sub.2, where PA indicates alveolar oxygen pressure and is given by (FiO2×[760−47])—(PtcCO.sub.2/0.8); (3) arterial to Alveolar (a/A) ratio =PtcO.sub.2/PA; and (4) oxygenation index (OI)=P.sub.aw×FiO.sub.2×100/PtcO.sub.2.
[0085] Sample size was calculated for the two primary outcomes as follows. To monitor lung aeration and function we targeted a correlation coefficient between LUS and OI of at least 0.6, based on previous data obtained in a similar population of extremely preterm neonates (De Martino et al. Pediatrics September 2018, 142 (3) e20180463). To predict BPD occurrence, we targeted an area under curve (AUC) of at least 0.7, and considering as null hypothesis the prediction by chance (AUC=0.5) and a positive/negative (i.e.: BPD/no BPD) case ratio of 1. For both calculations α and β were set at 0.05 and sample size resulted 98 and 100 for the two outcomes respectively. Given the easiness to recruit and the time needed to diagnose BPD we enlarged the recruitment to ensure having an equal number of positive and negative cases. An interim analysis was performed at 50% of the enrollment and no changes to the study protocol were made.
[0086] Data were expressed with mean (standard deviation) or median [interquartile range], as appropriate. Basic population data were compared with χ.sup.2 or Fisher and Student or Mann-Whitney test, as appropriate. Lung ultrasound scores calculated at the various timepoints were compared with repeated measures-ANOVA, using the BPD diagnosis as between subjects' factor and followed by Bonferroni post hoc test. Correlation analyses with lung ultrasound scores were performed using Spearman coefficients, followed by multivariate linear regressions with backward- stepwise method, adjusting for gestational age and the diagnosis of BPD. Covariates were removed from the model if p-value was >0.10. Gestational age was chosen as covariate because of its association with BPD; birth weight was not included because it is correlated with gestational age and creates significant multicollinearity. Results will be graphically shown in scatter plots with trendline generated by local regression smoothing procedure (Epanechnikov's kernel with 85% span).
[0087] Receiver operator characteristics (ROC) procedure analyzed the accuracy of lung ultrasound scores on different timepoints to predict BPD: curves were compared with DeLong's method and results are reported as area under the curve (AUC and 95% confidence interval). Then, lung ultrasound scores with highest AUC were entered in multivariate, logistic, backward-stepwise models together with gestational age and their interaction term. Covariate treatment was as above; goodness-of-fit was evaluated with Hosmer-Lemeshow test. Results were used to create gestational age-adjusted lung ultrasound scores. They were subjected to ROC analyses and post-test probability was estimated using the Fagan nomogram. Analyses were performed with SPSS 25.0 (SPSS Inc, Chicago, Ill.—USA), MedCalc 13.3 (MedCalc bvba, Ostend, Belgium), and GPower 3.1 (HHU, Dusseldorf, Germany). p-values<0.05 was considered significant.
RESULTS
[0088] One-hundred and seventy-nine neonates were eligible and 32 were excluded (for death (7) or transfer (4) before 36 weeks post-menstrual age, lack of BPD data (21)), thus 147 neonates were finally included and analyzed. Table 2 shows the basic population data: infants with and without BPD have similar baseline characteristics, but BPD infants are more preterm. BPD was mild, moderate and severe in 24 (16.3%), 32 (21.8%) and 16 (10.9%) neonates, respectively. A subgroup of 115 neonates (57 in BPD and 58 in the control cohort, respectively) underwent both LUS and eLUS calculations. Basic patients' characteristics were similar between recruiting centers. Neonates were stable during lung ultrasound and data collection and no problem was noticed.
TABLE-US-00002 TABLE 2 Basic population details. Data are expressed as mean (standard deviation) or median [interquartile range] or number (%). Whole BPD Control population cohort cohort (147) (72) (75) p Gestational age 27.3 (1.9) 26.3 (1.9) 28.1 (1.5) <0.001 (weeks) Birth weight (g) 954 (289) 812 (252) 1089 (257) <0.001 SGA neonates 37 (25%) 22 (30.6%) 15 (20%) 0.140 Male sex 77 (52%) 38 (52.8%) 39 (52%) 0.925 Antenatal steroids 123 (84%) 61 (84.7%) 62 (82.7%) 0.736 Cesarean section 96 (65%) 44 (61.1%) 52 (69.3%) 0.295 Abbreviations: BPD: bronchopulmonary dysplasia; NICU: neonatal intensive care unit; SGA: small for gestational age.
[0089] Both LUS and eLUS significantly vary between the timepoints (both overall p<0.0001, within subjects' contrast) and post-hoc tests show that they are lower at 36W than at DO (p=0.003 for LUS; p=0.05 for eLUS), D7 (both p<0.001), D14 (both p<0.001) and D28 (both p<0.001). Diff33w-28 LUS and eLUS are also different between BPD and control cohorts (both overall p<0.0001, between subjects' contrast).
[0090] LUS significantly correlate with oxygenation metrics and with PtcCO2 (except on D7) and this is confirmed upon adjustment for gestational age and BPD diagnosis. LUS is also correlated to Silverman's score at D0 (ρ=0.432, p<0.0001; β=0.18, p<0.0001), D7 (ρ=0.45, p<0.0001; β=0.1, p=0.001), D14 (p=0.394, p<0.0001; β=0.07, p=0.014), D28 (ρ=0.544, p<0.0001; β=0.13, p<0.0001) and 36W (ρ=0.59, p<0.0001; β=0.16, p<0.0001). Similar correlations were found using eLUS, except that for PtcCO.sub.2, which did not correlate with eLUS at any time-point (not shown).
[0091] For the prediction of BPD, a preliminary analysis showed that AUC of LUS and eLUS were slightly higher for the scores calculated on the seventh and fourteenth day of life BPD. When these scores were adjusted for gestational age in multivariate logistic models a significant interaction between lung ultrasound scores and gestational age was evident, with increasing gestational age and lung ultrasound score being associated with reduced and augmented BPD occurrence, respectively.
[0092] Thus, gestational age-adjusted lung ultrasound scores were calculated as lung ultrasound score-to-gestational age ratio for early (D7) or late (D14) prediction of BPD. They resulted significantly associated to BPD, both at D7 (adjusted LUS: OR: 803 (95% CI: 55-11625), p<0.0001; adjusted eLUS: 861 (64-11586), p<0.0001) and at D14 (adjusted LUS: OR: 917 (95% CI: 69-12183), p<0.0001; adjusted eLUS: 815 (6704-9484), p<0.0001).
[0093]
[0094]