Biochemical Bone Markers for Early Detection of Osteopenia of Prematurity

Osteopenia of prematurity (OOP) imposes the risk of fractures and growth failure to premature infants. Studies have investigated the validity of biochemical markers of osteopenia but till date it is not established. So, this study was intended to examine the diagnostic performance of biochemical markers in early detection of osteopenia of prematurity. This prospective study was conducted in the Neonatal Intensive Care Unit (NICU), Department of Neonatology, Bangabandhu Sheikh Mujib Medical University during June 2013 to February 2014. A total of 100 premature infants with gestational age ≤ 34 weeks were consecutively included over 9 months period. Serum alkaline phosphatase, serum calcium and serum inorganic phosphates were measured from 1 week of chronological age until corrected term age. At corrected term age, radiologic examination was done for the assessment of osteopenia. Of the enrolled infants, 36/78 (46%) developed radiological evidence of osteopenia. Serum inorganic phosphate level was significantly less in osteopenic infants than non-osteopenic infants throughout first two months of life (p <0.001). The area under ROC curve for serum inorganic phosphate was 85% (p = 0.001). If the cut off value of serum inorganic phosphate was set at 3.6 mg/dl, then a sensitivity of 86% and a specificity of 49% were obtained. Low serum inorganic phosphate at 3 weeks of life can be used as a marker for early detection of osteopenia of prematurity.


Introduction
In the face of advancement in antenatal and perinatal care, every year, 15 million babies are born prematurely around the world of which more than 1 million die. 1 However, the continuous developments in intensive care have led to a progressive decline in mortality. 2The achievement through an intensive care has not always ensured a favourable outcome of premature newborn.Osteopenia of prematurity (OOP) is an important co-morbidity which expose them to the risk of fractures and growth failure. 3Premature infants are at risk of suboptimal bone mineralisation in early life, as the later part of gestational period is vital for mineral accretion. 4Their bones mineral content at term corrected age has been found to be 25-70% lower than in term. 5Most of the previous studies reported that incidence of osteopenia of prematurity was as high as up to 75%. 6,7Radiologic evidence of osteopenia has been reported in 23% of very low birth weight infants and in 55% to 60% of extremely lowbirth weight infants.Fractures were reported in nearly 10% of VLBW infants at 36 to 40 weeks postmenstrual age. 8Possibility of declination of incidence is anticipated over past two decades because of nutritional management. 9Recent data support its existence especially in extremely low birth weight population. 3Bangladesh is among the top 10 countries with the highest numbers of preterm birth. 1 Dealing prematurity related complications are the great challenges for the neonatal physicians.Proportion of osteopenia of prematurity was 28% according to a study done in a Bangladeshi population during 2003-2004. 10o date, enough data are not available regarding risk factors, useful biochemical markers which can detect premature babies at risk.

Contemporary diagnosis of overt osteopenia has
been based on characteristic radiological changes 11 .However, the radiological changes of osteopenia are not easily detected until bone mineralization is reduced by at least 20%. 12Dual energy X-ray absorptiometry (DEXA) is considered to reflect most accurately the state of bone mineralization.Because of the rarity of DEXA instruments, a biochemical marker would be a good alternative in resource limited centers. 13Serum alkaline phosphatase, serum calcium and serum inorganic phosphates have been suggested to be a way of identifying preterm infant with osteopenia of prematurity. 14here is scarcity of evidences that any of the frequently used serum markers are valid marker of osteopenia of prematurity.Most of the authors suggested 3 weeks of life as early period of screening as alkaline phosphatase level rises in all newborns in the first 2-3 weeks of life and increases further if there is insufficient mineral supply. 12Hung et al concluded that serum ALP exceeding 700IU/L at 3 weeks post natal age can predict the risk of osteopenia of prematurity. 12ackstrom et al showed that high serum alkaline phosphatase activity and low serum phosphate level are the best available screening method for low bone mineral density. 13A systematic review compared the results of frequently used serum and urine measurements to the results of number of imaging techniques, which showed that increased urinary calcium concentration, may be a valid biochemical marker, but more research was recommended to confirm this. 11Currently, there is neither standard recommendation for screening of osteopenia of prematurity nor data are available describing the usual values of bone markers in preterm Infants. 15Scarcity of evidences regarding screening and diagnostic tool might have led premature infants undiagnosed subsequently subject them to the risk of growth failure.Therefore, this study was intended to examine the diagnostic performance of biochemical markers in early detection of osteopenia of prematurity.

Materials and Methods
This prospective study was conducted in the Neonatal Intensive Care Unit (NICU), Department of Neonatology, Bangabandhu Sheikh Mujib Medical University over a period of 9 months (June 2013 to February 2014).After obtaining informed written consent from parents, a total of 100 neonates were enrolled consecutively.Infants with haemodynamically unstable prematurity, congenital anomalies and suspected congenital bone or muscle disease were excluded.Baseline demographics including gestational age (GA), birth weight (BW), small for gestational age (SGA), gender and perinatal characteristics including place of delivery, mode of delivery, Apgar score at 5 minutes were collected.Birth weight was recorded from labour room or neonatal referral sheet in case of out born infants.Weight at admission was taken by a digital weighing scale (SALTER, Model-914).Gestational age was assessed by LMP, sonographic findings and modified new ballard scoring.Postnatal medical records (respiratory distress syndrome, sepsis, and retinopathy of prematurity, time to full feeding) were recorded.
Blood samples for biochemical tests were done biweekly from 1 week postnatal age onwards up to corrected term age.During study period, these biochemical markers were incorporated into routine biochemical follow-up of premature infants and sample were taken aseptically adjusting with other purposes of venepuncture.Frequency was at least 3 in infants with gestational age ≥ 33 weeks, 4 in ≥ 31 weeks, 5 times in ≥ 29 weeks.Serum alkaline phosphatase activity was measured by colorimetrically by pnitrophenyle phosphorus acid radical quality method.It was reported in U/L and normal reference value for adult was 50-136 U/L.A prefeed blood sample in a plain tube was used.Serum calcium was examined by the complexed cresolephthalein test.A Level of < 7 mg/dl was taken as low calcium value for the newborn.Serum inorganic phosphate was measured using a direct phosphomolibdate reaction, which can be measured colorimetrically.Normal reference value was taken as 3.0-4.5 mg/dl.Enrolled premature infants underwent for radiologic examination of forearm with portable X-ray machine at corrected term age.Radiologic examination of forearm with portable X-ray machine was performed by single radiographer.Infants were exposed to radiation dose at 50-55 kVp (kilovolt peak) and 0.5 mAs (milliamp seconds).Diagnostic reference level was 68 kVp / 0.5mAs.Comments on radiographic findings were made by a radiologist who was blinded to laboratory findings.The scoring method of Koo et al 17 was used to assess osteopenia.Grade 0 has been defined as findings of normal bone, grade 1 as mineral rarefaction, grade 2 as fraying and cupping of the metaphyses and grade 3 as fractures.Osteopenia was considered when there is radiographic evidence of diminished bone density defined as > grade 1. Patients' outcomes including presence of osteopenia, length of hospital stay, gestational age at radiologic examination, gestation were documented.
Study was approved by Institutional Review Board (IRB) of Bangabandhu Sheikh Mujib Medical University.Data were compared between osteopenic and non-osteopenic using Chi Squared test for categorical variables and Independent-t test for continuous variables.Receiver operator characteristics (ROC) curves were used to determine the individual diagnostic performance of biochemical bone markers.The statistical software IBM SPSS Statistics version 19 (SPSS Chicago, IL) was used for the statistical analysis.Results were considered statistically significant at p < 0.05.

Results
During the study period, 190 eligible preterm infants ≤ 34 weeks gestation were admitted in the Neonatal Intensive Care Unit.After taking informed written consent, a total of 100 neonates were enrolled.Three parents denied giving the blood for serum bone biochemical markers estimation and withdrew consent, 9 patients died during study period (6 during hospital stay, 3 after discharge from hospital) and 10 patients did not complete the follow-up and radiological diagnosis.The remaining 78 infants were ultimately retained for analysis.Demographic and perinatal characteristics of studied infants are presented in table I.Of the 78 infants, 36 (46%) developed radiological evidence of osteopenia at corrected term age.Of them, 22 (28%) infants were grade 1, 13 (17%) were grade 2, remaining 1 (1%) was grade 3 osteopenia according to Koo's criteria 17 .Mean gestational age of radiologic detection of osteopenia of prematurity was 39 ± 1.3 weeks.According to the gestational age category, osteopenia was significantly more frequent (22/31, 71%) in less than 32 week premature infants in comparison to infants with gestational age ≥32 weeks.Comparison of demographic and clinical characteristics in premature infants who had osteopenia and those without osteopenia at corrected term age is shown in table II.Compared with non-osteapenic infants, infants with osteopenia were significantly smaller (p value 0.001) and more premature (p value <0.001).Infants with osteopenia suffered from sepsis significantly more frequently than non-osteapenic infants (p value 0.01).Premature infants with osteopenia tended to have significant delay in achieving full feed (p value 0.001) than that of infants without osteopenia.Significant difference was also found in case of need for mechanical ventilation (p value 0.028).Hospital stay was also significantly prolonged in osteopenic preterm infant than that of non-osteopenic infants (p value 0.001).The result of this study showed, serum alkaline phosphatase level was significantly elevated only in 5-8 weeks of life in case of osteopenic infants when compared with non osteopenic.There was no significant difference in serum alkaline level between two groups in 1-4 weeks of life, suggesting that serum ALP level is not a good marker of osteopenia in first 4 weeks of life.Serum inorganic phosphate level was significantly less in osteopenic infants than nonosteopenic group in first two months of life, indicating that a persistently low phosphate level contributes to the development of osteopenia of prematurity.A fall in plasma phosphate concentration has been described in very low birth weight infants who developed disturbances in the mineralization, especially among those receiving human milk. 13,16This might be the explanation of the finding regarding changes in the serum phosphate in the present study.A retrospective study of 230 conducted by Sreekanth et al demonstrated similar findings with that of present study with the exception that they have investigated biochemical markers among extremely low birth weight population. 3o date, role of biochemical markers in detection of osteopenia of prematurity are questionable.Utility of serum alkaline phosphatase and serum phosphate in the detection of osteopenia of prematurity were challenged by J Faerk et al as they could not demonstrate any association between bone mineral content and mean alkaline phosphatase and serum phosphate. 14The role of urinary calcium and phosphorus levels were investigated along with serum calcium and phosphorus levels by catache and Leone. 19They argued that serum calcium and phosphorus levels are not good markers in early detection of mineral deficiency, rather urinary calcium level may be helpful in early detection of mineral deficiency.
In this study of a recent cohort of premature babies', 46% (36/78) had radiological evidence of osteopenia at corrected term age.The incidence is similar to that has been previously reported in literature.In 1989, Koo et at 17 observed 30% incidence of fractures/rickets in first 1 year of life while following 78 preterm infants.Mitchell et al 15 selectively screened 32 extremely low birth weight infants with ALP level > 800 IU/L for radiological rickets.All had evidence of osteopenia and 18/32 (56%) had radiological rickets.In a recent study by Yi-Li Hung et al concluded that osteopenia of prematurity among ≤ 34 weeker was 39%. 12 A retrospective review by Burm seok et al between 2009 and 2011 revealed 43% (112/258) incidence of osteopenia. 20ring the study period radiology was the only tool to make definitive evidence of osteopenia.So, radiation hazard of X-ray could not be avoided in studied infants.In standard radiograph, the bone changes of osteopenia of prematurity are a late sign as these appear only after 20-30% of bone mineral content is lost.So, it was possible to miss the premature infants with mild osteopenia.

Conclusion
Radiologic evidence of Osteopenia was 46% among premature infants gestational age ≤ 34 weeks.Low serum inorganic phosphate at around 3 weeks of life can be used as a predictor of osteopenia of prematurity.Future research is recommended to find out a reliable investigation tool to make correct and early diagnosis.

Table I :
Baseline characteristics of enrolled premature infants (n= 78)

Table II :
Demographic and clinical characteristics in osteopenic infants compared with non-osteopenic infants

Table III :
Serum alkaline phosphatase, serum inorganic phosphate, serum calcium in osteopenic infants compared with non-osteopenic infants Biochemical data were available for 255/295 (87%) of the study population.Of 255 sets of data, biochemical data of 78 infants were available in 1 week of age, 63 data were collected at 3 weeks of age, 60 data were available at 5 weeks of age and 54 data were documented at 7 weeks of age.As biochemical tests could not be done strictly on biweekly basis, biochemical data were summarized in two time period; 1-4 weeks and 5-8 weeks.The proportion of missing biochemical data were similar between the two groups (26/150 [17.3%] vs 18/100 [18%], p=0.39).Serum alkaline phosphatase level was significantly elevated in 5-8 weeks of life (p <0.001) in osteopenic infants when compared with non-osteopenic infants.No significant difference in serum alkaline phosphatase level could be demonstrated between two groups in 1-4 weeks of life.Serum inorganic phosphate level was significantly less in infants with osteopenia than without osteopenia both in1-4 weeks (p <0.001) as well as 5-8 weeks of life (p <0.001).No differences were found between osteopenic and nonosteopenic infants regarding serum calcium level in both time periods.accuracy of inorganic phosphate was more in 3 weeks curve as the area under ROC curve for serum inorganic phosphate was 85% (p = 0.001) at 3 weeks of age.At 3 weeks age the areas under At 1

weeks of age At 3 weeks of age ROC
curve for serum calcium and serum alkaline phosphatase were 58% and 34% respectively (p 0.29 and 0.06 respectively).If the cut off value of serum inorganic phosphate was set at 3.6 mg/dl, then a sensitivity of 86% and a specificity of 49% were obtained.