Newborn Screening for Congenital Hypothyroidism, Congenital Adrenal Hyperplasia, and Glucose-6-Phosphate Dehydrogenase Deficiency for Improving Health Care in India

Abstract

Newborn screening (NBS) aims toward early detection of treatable congenital disorders. From January 2008 through December 2017, 13,376 newborns were screened for congenital hypothyroidism (CH), congenital adrenal hyperplasia (CAH), and glucose-6-phosphate dehydrogenase (G6PD) deficiency at Sir Ganga Ram Hospital, India, by measuring G6PD activity, thyroid-stimulating hormone, and 17-hydroxyprogesterone on dried blood specimens. The birth prevalence of 1:2,000 for CH, 1:2,500 for CAH, and 1:125 for G6PD deficiency indicates the latter as the most prevalent. Performance evaluation of testing reveals a robust screening program with 100% sensitivity and >99% specificity. Hence, we recommend NBS for early diagnosis and treatment to prevent adverse outcomes.

Introduction

Newborn screening (NBS), a comprehensive preventive public health program, has salvaged thousands from severe disabilities or death through early diagnosis and treatment. ^{1 2} The program has come a long way since its discovery by Robert Guthrie in the 1960s to screen for phenylketonuria in heel prick blood samples on filter paper. ³ Currently, NBS is a mandatory practice that has become an integral part of neonatal evaluation in many countries worldwide ( https://www.cdc.gov/newbornsscreening ). Screening for glucose-6-phosphate dehydrogenase (G6PD) deficiency was initiated in 1965 ⁴ ; congenital hypothyroidism (CH) and congenital adrenal hyperplasia (CAH) were first screened for in 1970. ⁵ Presently, many countries have established extensive NBS programs that screen for as many as 54 inherited conditions including enzyme deficiencies, hormone deficiencies, lysosomal storage disorders, organic acidurias, and amino acidurias among others. ⁶

A recent study using meta-analysis cites a 1:1,964 global prevalence of inborn errors of metabolism. ⁷ Until 1990, the incidence of the CH was estimated as 1:3,000 to 1:4,000; however, advanced technology, changes in screening algorithms, and increased awareness have resulted in a higher reporting rate of 1:2,000. ⁸ The estimated prevalence of CAH is 1:10,000 with the annual incidence ranging from 1:5,000 to 1:15,000 ( https://www.orpha.net/consor/cgi-bin/OC Exp.php? Expert = 418). The worldwide incidence of classical CAH varies along ethnic and geographic lines with the highest rates of incidence being reported among the Yupic Eskimos of Alaska (1:280) and the inhabitants of the French island of Reunion (1:2,100). ⁹ G6PD deficiency, the only enzyme defect that was screened for in this study, affects ∼400 million people worldwide. ^{10 11}

An estimated 30,000 children are born in India everyday of which 1 out of every 1,000 is affected with a neonatal disorder. ¹² The reported incidences of CH (1:1,103), CAH (1:5,762), and G6PD deficiency (1:66–1: 1,000) indicate a high prevalence in the Indian population. ^{13 14 15 16} The major contributing factors for the same include a high birth rate, high rates of consanguinity in marriages, as well as inadequate availability of diagnostic and genetic counseling facilities. Undiagnosed and untreated individuals suffering from these disorders, especially those who have incurred permanent damage remain a financial burden on their families and society. Consequently, an effective NBS program is the need of the hour in our country especially in light of the facts that diagnostic tests for these disorders are cheap and treatments fairly straightforward with good prognosis.

The present study, therefore, intended to screen neonates for CH, CAH, and G6PD deficiency to expedite diagnosis and subsequent treatment, as well as facilitate prevention of adverse outcomes. The study also aimed to determine birth prevalence and assess the usefulness of neonatal screening for these disorders in the North Indian population.

Materials and Methods

Blood Sampling

Blood samples of 13,376 newborns were collected on neonatal cards (Whatman 903) at the neonatal unit of Sir Ganga Ram hospital, New Delhi in satisfactory condition with requisite details including neonate's name, mother's full name, father's full names, address and phone number, sex, date of birth, gestation (preterm: delivered < 37 weeks; term: delivered ≥ 37 weeks), birth weight (low < 2.5 kg; normal > 2.5 kg), transfusion status, breast/bottle fed, delivery status (normal vaginal/cesarean), single/twin, and physician's details. The preferred time of blood collection was 24 to 48 hours of age by heel prick method for term neonates. Collection was never postponed beyond 7 days of age where any form of inpatient neonatal care was provided regardless of prematurity, illness, feeding history, or drug treatment. The appropriate strategy for premature infants was to always collect samples close to 7 days of age and immediately before any transfusion for sick neonates irrespective of age.

Informed Consent

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki declaration of 1975, as revised in 2000 of the World Medical Association. Parents were counseled on the benefits of NBS and informed consent for the screening tests to be performed on DBS of their neonates was obtained before including them in the study.

Screening Procedures and Strategy for Diagnosis

G6PD enzyme activity, thyroid-stimulating hormone (TSH) levels, and 17-hydroxyprogesterone (17-OHP) levels were measured semiquantitatively in dried blood specimens using time resolved fluoroimmunoassays as per kit manufacturer's instructions (Perkin Elmer, Turku, Finland) and MultiCalc software ( www.perkinelmer.com ) on DELFIA Victor 2D fluorometer. Reference ranges for normal and abnormal cases were established by analyzing 1,000 newborn samples prior to commencement of the screening program. Average, median, standard deviation, and ranges were determined for all three parameters and strictly monitored on a regular basis for statistical analysis. Concentration of TSH up to 9 μU/mL of blood was categorized as normal. Values between 9 and 18 μU/mL blood and beyond 18 μU/mL blood were considered intermediate and positive, respectively, for CH. The normal biological reference interval for G6PD enzyme activity was >2.0 U/g hemoglobin and any sample with a measured activity below this cutoff was reported as positive. In the case of 17-OHP, concentrations <30 nmol/L blood and <60 nmol/L blood for term and preterm neonates, respectively, were classified as normal. Samples with concentrations ranging from 30 to 90 nmol/L blood for term and 60 to 90 nmol/L blood for preterm infants were treated as intermediate cases, while all samples with a concentration beyond 90 nmol/L blood were reported as positive for CAH. Intermediate and positive cases were repeated in a second test using a separate blood spot on the same card before communicating results to patients. These cases were further recalled for confirmatory testing on another blood draw/serum sample. G6PD enzyme deficiency was confirmed by quantitative analysis using an enzyme kinetic procedure ( www.trinitybiotech.com ). Quantification of serum TSH, free T3, and free T4 was performed by enzyme-linked immunosorbent assay (ELISA) ( www.biocheckinc.com ) followed by ultrasonography of thyroid gland for ectopic/agenesis of thyroid for all cases that screened positive for CH. Serum 17-OHP and cortisol (8 am and 4 pm) radio immunoassays ( www.cisbio.com ) as well as dehydroepiandrosterone ELISA assays were performed to confirm cases of CAH.

To ensure accurate results and correct analysis, the laboratory continuously participates in the Newborn Screening Quality Assurance Program run by Centers for Disease Control and Prevention (CDC), Atlanta, United States and Preventive Medicine Foundation (PMF), Taiwan for proficiency testing. ¹⁷ The laboratory is also accredited at the national level under ISO 15189:2012. Clinical outcome was recorded by collecting data on the quantitative confirmatory results and telephonic communication with the concerned patients. To assess the significance of birth weight and gestational age in NBS, patient data were categorized into male/female, term/preterm, and low birth weight/normal birth weight.

Results

The demographic characteristics of the neonates screened ( n  = 13,376) along with the positive screening results for CH ( n  = 9), CAH ( n  = 15), and G6PD deficiency ( n  = 138) are tabulated in Table 1 . The status of the cases that were screened positive as either true- or false-positive postconfirmatory testing is given in Table 2 .

Table 1. Demographic characteristics of newborns screened for CH, CAH, and G6PD deficiency.

Total newborns screened ( n )	Sex	Gestation	Birth weight	Screening results (positive) for the given disorders
				CH ( n )	CAH ( n )	G6PD deficient ( n )
13,376	Male ( n  = 7,301)	Term ( n  = 6,032)	LBW ( n  = 938)	1	2	13
			NBW ( n  = 5,094)	4	4	77
		Preterm ( n  = 1,269)	LBW ( n  = 1,003)	0	2	15
			NBW ( n  = 266)	0	0	7
	Female ( n  = 6,075)	Term ( n  = 5,278)	LBW ( n  = 1,000)	1	2	6
			NBW ( n  = 4,278)	3	3	18
		Preterm ( n  = 797)	LBW ( n  = 672)	0	2	1
			NBW ( n  = 125)	0	0	1
13,376			Total positive cases	9	15	138

Abbreviations: CAH, congenital adrenal hyperplasia; CH, congenital hypothyroidism; G6PD, glucose-6-phosphate dehydrogenase; LBW, low birth weight; NBW, normal birth weight.

Table 2. Status of screen positive newborns after diagnostic confirmation for CH, CAH, and G6PD deficiency.

S. no.	Disorder	Screen positive	Recalled cases for confirmation	True positive	False positive (rate)
1	CH	9	9	7	2 (0.015%)
2	CAH	15	15	5	10 (0.07%)
3	G6PD deficiency	138	138	108 (male: 84) (female: 24)	30 (0.2%)
	Total confirmed		162	120	42

Abbreviations: CAH, congenital adrenal hyperplasia; CH, congenital hypothyroidism; G6PD, glucose-6-phosphate dehydrogenase.

Statistical Analysis of Screening Results

Patients were divided into subgroups based on gender, gestational age, birth weight, and finally screening results. Mean values of the cases that were screened as positive or negative in each group were compared using the unpaired t -test. A p -value of < 0.05 was taken as statistically significant while evaluating results and comparison using statistical methods was not possible in certain groups where the number of cases was <3. A statistical significance ( p -value < 0.0001) was observed between the positive and negative cases. However, no significant effect of gestational age or body weight on screening results was seen.

Confirmation of Screen Positives and Follow-up

Identical results were obtained in 90 to 95% cases that were screened positive in the first attempt on repeat testing with a separate blood spot. These were 138/152 cases that were deficient in G6PD enzyme activity, 9/10 with elevated TSH levels, and 15/16 with high 17-OHP levels. Thus, a total of 162 (1.21%) cases were recalled for confirmatory testing, of which 120 (74%) were true positives, while 42 (26%) were false positives. The maximum number of false positives (30/162) was reported in G6PD deficiency (18.5%). The overall false-positive reporting rate of CH was markedly lower (<0.015%) in comparison to CAH (0.07%) and G6PD deficiency (0.2%). No false negative was found in our study during 3 months of follow-up (clinical assessment was performed by a qualified clinician during the follow-up period).

The entire process from sample collection, screening, and confirmation of diagnosis to commencement of treatment has been completed within 15 days. All confirmed cases are currently doing well on prescribed medication and by following the requisite dietary advice except for two of CH (acute intercurrent illness) and one of CAH (salt wasting crisis), who died due to noncompliance.

Diagnostic Evaluation of the Outcome of Newborn Screening Program

The clinical significance of NBS is highlighted by the 100% sensitivity and >99.6% specificity observed in this study. Using the above-mentioned cutoffs for the three different parameters that were measured in the NBS program, the positive prediction rate for G6PD deficiency and CH was determined to be 78.2 and 77.7%, respectively. The same was significantly lower for CAH and was found to be 33.3%. Within the population included in this study, the frequency of CH and CAH was <0.05%, while that of G6PD deficiency was 0.8% (1.15% males and 0.4% females). Further, based on these frequencies, the overall birth prevalence in the Indian population is 1:125 for G6PD deficiency, 1:2,000 for CH, and 1:2,500 for CAH ( Table 3 ).

Table 3. Diagnostic evaluation/performance of the newborn screening for CH, CAH, and G6PD deficiency.

Disorders	Negative predictive value	Positive predictive value	Sensitivity (%)	Specificity (%)	Frequency (%)	Birth prevalence
CH	100	77.7	100	99.9	0.05 (7/13,376)	1:2,000
CAH	100	33.3	100	99.9	0.04 (5/13,376)	1:2,500
G6PD deficiency	100	78.2	100	99.7	Male: 1.15 (84/7,301)	1:87
					Female: 0.4 (24/6,075)	1:250
					Total: 0.8 (108/13,376)	1:125

Abbreviations: CAH, congenital adrenal hyperplasia; CH, congenital hypothyroidism; G6PD, glucose-6-phosphate dehydrogenase.

A diagnostic accuracy of >99% for CH, CAH, and G6PD deficiency in proficiency testing (CDC, Atlanta, United States and PMF, Taiwan) along with rigorous monitoring of intra- and interassay coefficients of variation within acceptable limits (<20%) for each of these assays ensured excellent testing quality.

Discussion

NBS for three congenital disorders, that is, CH, CAH, and G6PD deficiency aimed at and has successfully resulted in early detection and intervention while simultaneously establishing their incidence in North Indian neonates.

The overall incidence of CH in our study was found to be 1:2,000, which is consistent with earlier studies that reported incidences ranging from 1:500 to 1:5,263 from India and other Asian countries. ^{18 19} The incidence of CH appears to be higher in the Asian population when compared with the United States as reported by Harris and Pass. ²⁰ Another report by Padilla and Therrell analyzed NBS data from the Asia Pacific region and concluded that certain countries in the region do not screen all neonates for CH thus increasing the burden of this disorder. ²¹ This is especially significant since presymptomatic diagnosis and the subsequent initiation of treatment can avert impairment of neurological development and the consequent deficit of intelligence.

CAH is an autosomal recessive disorder which accounts for significant mortality and morbidity in India. It has a reported incidence of 1 in 28,000 in Taiwan but ranges from 1 in 10,000 to 1 in 15,000 in most populations. ²² Multiple studies on the Indian population report the incidence of CAH to range from 1:4,000 to 1:12,500. ²³ Rama Devi and Naushad in 2004 reported a frequency of 1 in 2,575 neonates, ²⁴ which is comparable to our high estimate of incidence, that is, 1:2,500. NBS for CAH is of paramount importance given the high incidence in our population as well as the fact that lifesaving treatment is available for this potentially fatal disorder. The American Academy of Pediatrics reported a high false-positive rate (FPR) for CAH due to the effect of gestational age, birth weight, early collection, prematurity, and stress due to illness. ²⁵ In contrast, our study found no significant effect of gestational age or birth weight on the screening results analyzed for all three disorders.

G6PD deficiency is the most common enzyme defect worldwide that occurs with a frequency >5% throughout the Asia Pacific region. However, many countries in the same region have inefficient or nonexistent programs for neonatal G6PD screening. ^{26 27} On account of the multicultural nature of our country, the prevalence of G6PD deficiency varies between regions, states and even communities. Earlier studies have reported an incidence of G6PD deficiency ranging from 2 to 27% across various Indian communities. ^{13 15 28} Our study has estimated a frequency of 1:125 (1:87 in males and 1:250 in females). Goyal et al reported a prevalence of 1.5% in northern India, ¹⁵ while the present study that included a mixed population from the same region has estimated an overall prevalence of 0.8%. The difference may be explained by the possible over representation of a particular community in the former study.

On account of being an X-linked disorder, G6PD deficiency is expected to be rarer in females; however, the male-to-female ratio in India is highly variable. While we are reporting a ratio of 3.5:1, others have reported a ratio of 5.4:1 from the same geographic region. ¹⁵ It is reported that in areas of high prevalence, homozygous females may approach frequencies much nearer to that of hemizygous males. For instance, while the male-to-female ratio of G6PD affected individuals is 3:1 in the Vataliya Prajapati community in Western India, ²⁶ it is considerably closer in western Bengal which has a ratio of 1:1. ²⁸

G6PD is a manageable disorder which manifests primarily in males due to its X-linked pattern of inheritance. Heterozygous females, however, may be symptomatic due to a skewed degree of lyonization resulting in a red blood cell population that is largely deficient with respect to an active enzyme. ²⁹ Early detection will result in the prevention of clinical manifestation and aid physicians while counseling patients regarding contraindicated drugs, dietary restrictions, as well as avoidance of environmental factors that may trigger jaundice and kernicterus. ^{4 30} While the World Health Organization recommends screening in areas where prevalence is >3% in males, ³¹ we are of the opinion that screening programs for G6PD deficiency should be in effect in areas with lower prevalence due to the earlier reasons.

We observed a higher overall FPR of 0.2% in G6PD deficiency screening in comparison to the other two disorders. We speculate that this could be accounted for by improper time of sampling since samples drawn during or shortly after acute hemolysis will result in false-positive results. Thus, a stronger scrutinization of sample timing and condition may be useful in reducing the FPR.

Recent years have seen a significant improvement in the NBS programs operating in India. However, there remains an urgent need to establish cost-effective screening procedures as well as efficient systems for quality control, patient recall, initiation of treatment, and follow-up. These measures along with effective counseling and communication with families regarding the benefits of NBS including diagnosis prior to clinical presentation, prompt treatment to prevent manifestation of symptoms, and consistent compliance with treatment and follow-up will result in successful execution of NBS programs.

We, therefore, recommend screening for CH, CAH, and G6PD deficiency in neonates to exert a positive impact on the mortality and morbidity due to complications associated with these disorders.