High performance liquid chromatography

Although confusing at first - these are actually quite easy to read once you've looked at a few - and probably simpler than antibody panels! We recommend you go to the lab and read the weekly runs a few times with the haematology scientists who authorise the reports - there will be regular runs of these on pregnant women as part of the antenatal sickle and thalassaemia screening programme in each hospital you work at. Once you have seen a number the pattern recognition should come easily.

The nomenclature of each different type of haemoglobin generally relates to the way that it elutes on HPLC or the way it runs on the electrophoresis gel and the place of discovery - and has nothing to do with the genetic or phenotypic mutation - for example, you may have a Hb J which could be either an alpha-chain or beta-chain variant. Without looking up the specific genotype it will be impossible to tell. For all our examples we will be using the 'Bio-Rad' printouts - other machines may use different formats so please be aware that the exam printout may look slightly different from that which you are used to.

Below is a normal HPLC plot. A few rules:

• Ensure that there are a million 'events' in the total area - if it is below this level then potentially there is not enough sample to provide an accurate result.
• The elution times are shown on the graph - these are not the percentages of haemoglobin present - for that you have to look at the table (listed as being the area %). The size of the peak does offer a rough estimate of the proportions of haemoglobin in the blood.
• Each Hb has its own elution time, plus a denatured and glycated haemoglobin. Be aware that different haemoglobins' elution times may intersect with each other (for example the denatured portion of HbS may elute at the same time as HbA2). HbA2 also has its own denatured and glycated peaks but they are too small to pick up and be displayed on the result.
• In paediatrics, you expect HbF to have been replaced with normal adult haemoglobin by six months of age (and certainly by 9-10 months).

Hb F, A2 and A (recorded as 'A0') are usually clearly marked. The order of elution for all haemoglobins is:

• glycated form; denatured form; then actual haemoglobin.

So at 1.33mins the glycated form of HbA elutes, followed by the denatured form at 1.69mins and then the actual HbA at 2.46 mins.

Variation in HbA2 level

There are a number of causes of variations in HbA2, of which the main causes that you need to know are here:

Delta-beta haemoglobin abnormalities

There are two worth remembering:

1. Delta-beta thalassaemia trait = one gene deletion of the two inherited copies means a failure of production of delta and beta globin chains in that copy (results in slightly raised HbF and normal HbA2 - i.e. if delta chain production is decreased, this will also decrease the overall amount of HbA2 present). This is the classic 'silent' beta thalassaemia trait as the HbA2 in a standard beta thalassaemia trait would be elevated at >3.5%.

2. Hb Lepore = formation of a fusion between the delta and beta globins resulting in a beta chain variant. This is not a deletion but it does mean that beta chain production is reduced and may give the phenotypic expression of a beta thalassaemia trait. Equally here HbA2 is not elevated as in standard cases of beta thalassaemia trait as the delta globin gene is affected.

Other general rules in interpreting HPLC plots:

• HbA2 elution times can represent a beta thalassaemia trait or HbE (especially consider if HbA2 is >10%)
• In co-inheritance of variant haemoglobins (ie. HbA plus anything else apart from HbA) the percentage will not be 50:50 because other haemoglobins are not as stable as HbA. Characteristically a ratio is more like 60:40 weighted in favour of HbA. A 60:40 ratio is usually indicative of a beta chain abnormality.
• Co-inheritance of a alpha chain variant/deletion might lead to a ratio of 75:25 in favour of HbA. Inheritance of both alpha and beta chain variants can sometimes give unusual ratios which don't fall within the normal expected numbers.
• Consider whether this might be a picture of a transfused patient (there might also be transfusion of variant haemoglobins as well - ie. donors don't have to be HbAA to donate blood! They can also be HbAS (hence sickle patients have the requirement of sickle negative blood).
• Elevated HbF might be a suggestion that there is a beta thalassaemia gene inheritance as well as other causes, such as HPFH or patients on hydroxycarbamide

It's easiest to learn as we go along so there will be a number of examples to follow. Try to work out what is going on before looking at the answer as you go so you get the most out of the plots. If in real life you have a plot which the scientists can't identify, Bio-Rad for example have a variant library and you can also search online for what might give the characteristic peak in question. Please also read the two small sections in the haemoglobinopathy screening guidelines from BCSH (under headings 'Assessment of abnormal antenatal screening results' and 'High performance liquid chromatography' under Laboratory methods) before the next section.