Transcranial Doppler in children with sickle cell disease: Five years of screening experience

Abstract

Introduction

Sickle cell disease (SCD) is an autosomal recessive, inherited disorder of haemoglobin. Children with sickle cell disease (SCD) are at risk of vascular occlusions. If this occurs in the brain, it can result in stroke. Stroke is the leading cause of death and disability in children with SCD.

Transcranial Doppler (TCD) has been shown to accurately predict those children at risk of stroke so appropriate preventative treatment can commence.

Method

TCD screening with a non‐imaging Doppler machine to assess stroke risk has been in operation at our institution for 5 years. Eighty‐eight patients attending the haematology clinic for SCD are scanned annually. If abnormal velocities are found, repeat scans are performed at close interval to ensure results are reproducible.

Results

No child has had a stroke since the start of our screening. Seven of the eighty‐eight patients have shown high velocities on annual screening.

Conclusion

The introduction of a TCD programme at out institution has identified children at risk of stroke so they can be closely monitored and potentially receive prophylactic treatment.

Sickle cell disease

Sickle cell disease (SCD) is an autosomal recessive inherited disorder of haemoglobin, the molecule that carries oxygen to cells in all parts of the body. Normal red blood cells containing normal haemoglobin are smooth and flexible allowing easy passage through small blood vessels to perform their function. In SCD, the red blood cells contain abnormal sickle haemoglobin which causes those cells to change to stiff sickle shapes, giving this disease its “sickle cell” name1 (Figure 1).2 Inflexible sickle‐shaped red blood cells do not travel easily through small blood vessels, getting stuck along the way, depriving oxygen to certain parts of the body, causing pain and tissue damage. Sickle‐shaped red blood cells also die prematurely (haemolysis) resulting in anaemia and the complications of haemolysis. The abnormal haemoglobin and abnormally shaped red blood cells result in a disease with multi‐systemic effects of variable severity. It is characterised by anaemia and its consequences – episodes of pain due to vaso‐occlusion, infections, bone damage, organ damage (e.g. spleen, liver, kidneys and lungs), leg ulcers and an increased risk of stroke.

Figure 1.

Normal and abnormal red blood cells.2

Sickle haemoglobin (HbS) is caused by a mutation in the beta globin (HBB) gene, where a single nucleotide mutation leads to the amino acid substitution, glutamic acid by valine, at position 6. When only one copy of the HBB gene is mutated, the individual is a healthy carrier (sickle cell trait) whose red blood cells contain both normal haemoglobin (HbA) and HbS1. Both copies of the HBB gene are abnormal in SCD, and affected individuals have red blood cells that contain predominantly HbS, some fetal haemoglobin (HbF) and no HbA.3 Carrier parents may both pass the HbS gene to their children resulting in SCD, and the risk of this happening is 1:4 per pregnancy. The co‐inheritance of HbS with other defects of the HBB gene, such as beta thalassaemia, HbD and HbC, also results in SCD. In those instances, one parent is a carrier for sickle cell (HbS) and the other parent is a carrier for the other beta globin defect, for example beta thalassaemia.

SCD is one of the most common inherited blood disorders in the world, affecting millions of people worldwide. It is most prevalent among people whose ancestors came from sub‐Saharan Africa, the Mediterranean region, the Arabian Peninsula and India.4 Sickle cell carriers have a survival advantage if infected with malaria, explaining the prevalence in tropical and sub‐tropical regions.5 It is the most common inherited blood disorder in North America, predominantly affecting people of African descent.4 The prevalence of SCD in Australia has increased in recent decades due to immigration from the Middle East and Africa.

The management of this disease centres on supportive care and specific therapies that include education, screening for and preventing complications, pharmacological therapies, blood transfusions and bone marrow transplantation.

Stroke in sickle cell disease

Stroke is one of the most deadly and most disabling complications for children with SCD and has been the focus of research for decades. Medium‐size arteries of the circle of Willis, including the carotid arteries, are particularly vulnerable to the effects of sickled red blood cells and chronic haemolysis, resulting in stenosis and formation of fragile collaterals,6 which can lead to Moyamoya syndrome. Patients are at risk of both ischaemic and haemorrhagic strokes, with ischaemic strokes occurring mostly in childhood and in adults over 30 years, while haemorrhagic strokes most commonly occur in young adult patients. Sickle cell disease is the most common cause of childhood stroke, occurring in up to 11% of children with SCD and peaking between the ages of 2 and 9 years.7 The incidence of a subsequent stroke is between 50% and 90% within 3 years of the first event.8 In patients who have had a stroke and survived, chronic transfusions have been shown to reduce the risk of recurrent stroke to 10% or less7, 8.

The pivotal Stroke Prevention Trial in Sickle Cell Anemia (STOP), that was reported by Adams et al in 1998, showed that regular blood transfusion was an effective primary prophylaxis against stroke in at‐risk children identified by transcranial Doppler ultrasound (TCD).7 A total of 3929 non‐imaging TCD scans were performed on 1934 children with SCD in the multicentre study to identify those at risk of a first stroke. Those with reproducible time‐averaged mean of the maximum velocities (TAMMv) of ≥200 cm/s were offered a place in the study. Study subjects (total 130) were randomised to receive regular blood transfusions (63) or continue standard care without transfusions (67). The study demonstrated that TAMMv ≥ 200 cm/s is associated with a 46% risk of stroke over 39 months. At the 26‐month follow up, there were 11 strokes in the non‐transfusion group and 1 in the transfusion group. The trial was terminated early as the results were so striking. Stroke risk assessment by annual TCD in patients between 2 and 16 years of age and primary stroke prevention by chronic blood transfusions (to maintain HbS < 30%) were subsequently adopted as the standard of care for children with SCD9 (Table 1).

Table 1.

TAMMv limits. Adams et al.8

Normal	Conditional	Abnormal
< 170 cm/s	170–200 cm/s	Above 200 cm/s

Our institution, The Children’s Hospital at Westmead, Sydney, Australia, looks after an increasing cohort of children with sickle cell disease (Figure 2), reflecting the increasing number of new immigrants arriving in this city from sickle cell prevalent countries (Figure 3). This made a TCD screening programme critical for patient care. It commenced in 2014.

Figure 2.

New patients with SCD per year.

Figure 3.

Patients by ethnic group.

Method

In September 2013, two experienced paediatric sonographers attended a Transcranial Doppler training course run by Pacific Vascular Imaging at Swedish Medical Centre, Cherry Hill, 550 17^th Ave, Seattle, WA 98122, USA.

A non‐imaging transcranial ultrasound machine, Multi‐Dop T (Compumedics DWL, Josef‐Schuttler‐Str.2. 78224 Singen, Germany), was purchased by the Haematology Department so imaging could be performed in accordance with the STOP protocol for obtaining velocity measurements.

The screening programme commenced in early 2014. All children with SCD attending a haematology clinic at the hospital are offered TCD screening. Initially, scanning was confined to the older children to allow consolidation of technique before adding the complication of a non‐compliant child.

The examination

The patient is scanned when well and relaxed, but not asleep. Sleep increases carbon dioxide levels leading to raised velocities. Illness, hypoxia and anaemia all raise velocities.10 Patients are asked to reschedule outpatient appointments if unwell. We do not scan children when in hospital with disease complications.

The head diameter is measured with callipers. This helps determine the distance to the midline of the brain and the expected depth of the landmark vessels especially the middle cerebral artery (MCA) (Figure 4).

Figure 4.

Calliper measurements of head.

The patient lies supine with the scanning position behind the patient. This allows scanning both sides of the head without disruption to the patient. The temporal window is utilised to image the MCA, anterior carotid artery (ACA) and the terminal internal carotid artery (TICA). TAMMv readings are recorded at close intervals along the path of each vessel. A sample size of 4‐5 mm is used. A result sheet is kept with all progress velocities able to be compared to previous examinations.

A wall‐mounted television is used as distraction, and parents are invited to bring entertainment to which the child likes to listen or watch.

Being sonographers more familiar with imaging and having access to imaging machines, we have included initial guidance with an imaging machine in our protocol.

Initial scanning is with the Toshiba (now Canon) Aplio 500 (Toshiba Medical Systems Corporation 1385, Shimoishigami, Otawara‐Shi, Tochigi 324‐8550, Japan). A low‐frequency vector probe 5–2 MHz is used with specialised TCD settings providing increased power and gain. The presence of an adequate scanning window is ascertained (Figure 5).

Figure 5.

Image of circle of Willis.

Colour and spectral Doppler are used to identify the required vessels on both sides of the brain. This is used as a guide only. No velocity readings are obtained during the screening process as the angle‐corrected peak velocities used in imaging Doppler scanning are quite different to the TAMM velocities used with a non‐imaging machine. Confirmation of depth and direction of vessels from the scanning window is useful information to lessen the overall time of the examination (Figures 6, 7, 8).

Figure 6.

Colour imaging and spectral trace of the MCA.

Figure 7.

Colour imaging and spectral trace of left ACA.

Figure 8.

Colour imaging and spectral trace of the right ICA.

Non‐imaging scanning

We next scan with the Multi‐Dop T machine. The transducer is very small and light which allows minute angle adjustment. The connecting cable is also thin and light. Although there is no imaging, there are many familiar guides for identifying desired vessels (Figures 9 and 10).

Figure 9.

Non‐imaging Multi‐Dop T machine.

Figure 10.

Scanning with non‐imaging Multi‐Dop T machine.

Vessel identification cues

• Transducer position and angle

• Flow direction

• Spatial relationship

• Relative velocity – difference in pitch

Figure 11.

Spectral trace MCA.

Information and Image display

Figure 11 displays information including the depth, gain and sample volume selected, as well as the TAMMv computation, 84 in this case. The image display shows the spectral trace, while the lower half displays the presence and direction of any flow detected along the path of the beam. The thin white line (at 62 mm) is the position of the acquired spectral trace.

TAMMv measures the time‐averaged mean of the velocities in the peak velocity window area above the diastolic (Figure 12). This method is used in non‐imaging transcranial Doppler as there is no ability to accurately determine the direction of the vessel to allow for angle correction. Therefore, obtaining peak velocities is not possible.

Figure 12.

Area of trace used for TAMMv.

The initial part of the scan is identifying the MCA. This is the easiest and most important vessel to scan as it is the most likely to be able to be reproduced in subsequent scans. The expected depth is decided from the head circumference measurement and the depth seen on the initial scans on the Aplio 500. This is usually approximately 5 cm and is our typical starting point. Hearing the high‐pitched sound of the MCA compared to other vessels is a great tool in the identification process. Once the correct vessel is identified, the depth is reduced to the most superficial point the vessel can be found. Scanning is then commenced moving deeper in minute increments checking any marked change in velocity. Images are taken at least every 2 mm with the angle of the transducer adjusted as the vessel changes direction along its course (Figure 13) (Figure 14).

Figure 13.

Multi‐Dop T display of the MCA.

Figure 14.

Bidirectional flow MCA and ACA.

As the depth increases, the vessel changes direction and becomes the anterior carotid artery. The flow is moving away from the transducer, displayed as blue below the line. At greater depth, the gain may need to be increased to allow for attenuation (Figure 15).

Figure 15.

ACA.

The TICA has a lower velocity than the MCA and ACA, with the direction towards the transducer (Figure 16). The vessel is small and seen at a very similar depth to the ACA, angling slightly forward.

Figure 16.

TICA.

This is repeated on the other side with the patients head angled slightly away from the side being scanned. We scan the patient supine to allow them to see the television, which helps to keep them calm. If the vessels cannot be found in the expected position, the imaging machine is used again to determine the best approach to the required vessel.

We have found that despite not having images, the non‐imaging machine is more sensitive in picking up these small vessels and the small, light transducer less fatiguing when scanning for a length of time. The combination of the two machines allows us to identify the vessels in most patients.

Patients begin annual screening from 2 years of age, as the rate of abnormality is highest in the 2 to 5 age groups. Adams et al.7 Some very young patients are compliant, probably due to their familiarity with the hospital setting. If the patient is not compliant or getting upset, and if no raised velocities are found, an abridged examination may be performed. Keeping the first examination non‐threatening is imperative to foster a good relationship as the child will be returning at least yearly until their late teens.

A result sheet is kept and updated at each examination. This allows an easy check of trends and change, both rising velocities and response to treatment (Figure 17).

Figure 17.

Result sheet.

There are 88 patients currently being screened for raised velocities. Each patient is scanned once per calendar year if normal velocity readings are obtained. Significant increase in velocities or other abnormal changes are immediately relayed to the treating haematologist (Table 2).

Table 2.

Current results.

Patient numbers	Consistently high velocities	Transient high velocities	Stroke	Death due to stroke
88 26 siblings	7	2	0	1 prior to screening

In the case of raised TAMMv readings above 170 m/s, a repeat examination is scheduled within a short period – one month to three months – depending how high the velocities are. The repeat scan helps to determine whether there was an aberrant reading due to illness or stress.

Apart from raised TAMMv readings, other potential abnormal signs include.

• Jumps in velocity over time

• Marked discrepancy between hemispheres

• Very low velocity may indicate post‐narrowing flow

Conclusion

Transcranial Doppler of the cerebral vessels that can be investigated through the temporal bony window can help identify which children with SCD are at risk of stroke.

Our screening programme following the STOP protocols has, over 5 years, been successful in identifying children in need of close attention and possible treatment by blood transfusion. Our rate of abnormally high TAMM velocities is approximately 8% of patients, which is consistent with the literature. Adams et al.7

Young children not able to stay still for the examination and those with a poor acoustic window are still at risk of having high velocities missed by this screening programme. There is some complacency in attending appointments in patients who have had consecutive normal scans. We are attempting to address this with letter and phone reminders. To date, we are happy there have been no stroke events in our patient population and think other centres with a population of patients with SCD should consider a similar programme.

Authorship declaration

The authorship of this paper conforms to the Journal’s authorship policy.

Funding

No funding information is provided.

Disclosure statement

There was no financial relationship with any company in the writing of this paper.