Reply: Relationship between neonatal screening results by high-performance liquid chromatography and the number of alpha-thalassaemia gene mutations: consequences for the cut-off

We have made some comparisons between the work from the Netherlands cited by Bouva et al. ¹ and the Michigan (USA) experience with this BioRad Variant™nbs system.

From June 2010 to January 2012 the Michigan Department of Community Health (MDCH) Newborn Screening Program sent 62 samples with elevated FAST peak percentage (FAST) on high-performance liquid chromatography (HPLC), with the presence of Bart's haemoglobin confirmed by isoelectric focusing, to the Children's Hospital Oakland Research Institute (CHORI) in California, for molecular confirmation. FAST peak percentage on HPLC is made up of Bart's hemoglobin as well as possible other fast moving hemoglobins. Initially, only samples with FAST ≥25% were sent for confirmation testing. This was then expanded to better characterize lower concentrations of FAST and genotype. In 2010 and early 2011, FAST on HPLC and molecular results obtained from CHORI resembled the Netherlands data, especially in α traits, both cis and trans and in Hb H disease cases. There were no samples that were without an alpha thalassemia mutation from the CHORI DNA panel.

In the June of 2011, we received BioRad Variant™nbs reagents with a new resin lot 01490, and observed a dramatic drop in FAST values on HPLC. Also at this time, we began sending samples with FAST ≥15% on HPLC to CHORI. The value of 15% FAST on the new resin lot 01490 of reagents is equal to ∼21.8% FAST on the previous lot 62700 of reagents used in 2010 and early 2011. During this time frame, we had one sample that was without an alpha thalassaemia mutation from the CHORI DNA panel (Table 1).

Table 1

MDCH results (NEW LOT 01490) (only values of 15% and above sent for testing at CHORI) (DNA panel of –α ^3.7, −α ^4.2, −α ^cs –THAI, –SEA, –MED and –FIL)

Category	Phenotype	Median FAST peak(range)	Median FAST peak(range) converted to previous lot 62700 values	Numberofsamples
1	No mutation found	16.9*	24.4	1
2	Silent carrier	17.3 (15.2–17.4)	25.0 (22.1–25.1)	3
3	α Trait (trans)	15.9 (15.7–17.7)	23.0 (22.8–25.6)	9
4	α Trait (cis)	16.4 (14.8–19.5)	23.9 (21.1–28.1)	14
5	Hb H disease	31.5 and 40.2†	44.9 and 57.2	2

MDCH, Michigan Department of Community Health

*Only one sample in this category. Absolute result is depicted

†Only two samples in this category. Absolute results are depicted

The BioRad Variant™nbs system is not Food and Drug Administration approved for the detection of haemoglobin Bart's, and validations should be done to confirm that the cut-off for FAST provides consistent clinical data from one resin lot to another. An absolute cut-off for FAST cannot be assumed.

MDCH, with input from its physician advisory committee, will attempt to take the process of screening for Haemoglobin H disease a step further and identify the majority of two gene cis deletions for alpha thalassaemia. When cis deletions are detected, genetic counselling will be provided to assess the risk in future pregnancies of Haemoglobin H disease or Bart's hydrops fetalis. Bart's haemoglobin detected by HPLC/FAST that is above the detect limit, instead of above the three gene deletion cut-off, would be reported with the newborn screening results, as there is considerable overlap with FAST values with Bart's haemoglobin detected by Isoelectric Focusing (IEF) for 0, 1 and 2 gene deletions.

We would be interested to hear comments from other Newborn Screening programmes on this expanded mission for screening.