The kids aren’t alright - optimising radiation dose for young patients

Over recent decades, the use of computed tomography (CT) has soared. Valued for its speed, diagnostic image quality and accuracy, CT has become one of the most popular imaging modalities, with use climbing year on year. It has been estimated that UK hospitals performed a record five million scans in 2013-14.

While this has been good news for diagnosis, there are concerns regarding the potential risks. CT uses high-energy X-rays, therefore exposing the patient to a significant dose of ionising radiation. According to FDA, the typical effective dose for a chest X-ray is 0.02mSv, whereas an adult abdominal CT scan is 8.00mSv, 400 times more potent. To put this in perspective, the average person living in the UK will be exposed to 2.50mSv of natural background radiation every year.

For many experts, this presents real cause for concern. High radiation doses, such as those endured by victims of Hiroshima and nuclear accidents, are known to increase long-term cancer risk. And while the increased risk posed by medical radiation is small, on a population level, the numbers start to add up. In decades to come, as many as 1.5-2.0% of cancers in the US may be traced to the recent CT boom.

This issue is particularly salient when it comes to paediatric CT: because children are smaller and more radiosensitive than adults, they are especially vulnerable to harm. Deciding whether or not to perform a CT scan at all can therefore be a tricky question.

"Every time an examination is requested, you have to justify the need," points out Dr Catherine Owens, paediatric radiologist at Great Ormond Street Hospital, London. "Firstly, you have to ask is the examination necessary at all? "Secondly, if it is indicated, is there a better way of getting to the clinical solution - can you use something, like ultrasound or MRI, that doesn't use ionising radiation and thus has no radiation burden?"

Owens is president of the European Society of Radiology and a reader at University College London, as well as working on various EU projects addressing CT use within Europe. She advocates being extremely careful within paediatric imaging, not only by providing a robust justification for the exam but also ensuring the radiation dose is optimised by prudent use of the scanners on the market.

"We have to acquire images that are fit for diagnostic purpose - that is, diagnostic but at a radiation dose that is as low as reasonably achievable," she explains. "Putting it simply, there's an inverse relationship between the image noise and the dose, so you need to have a clear enough image to see any abnormality but at a radiation dose that is as low as possible for the child."

How low can you go?

There is no simple formula for tackling these issues; they must be addressed on a case-by-case basis. Take the question of whether to perform CT or MRI: while MRI is theoretically safer, insofar as it involves no radiation at all, examination acquisition times tend to be around 40 minutes as opposed to several seconds for CT. As a result, a child will need to be sedated or anaesthetised before the scan can occur, otherwise there will be significant motion-degradation artefact.

Then there's the question of dose management. In any given scenario, radiologists will need to work with medical physicists and radiographers to determine how low their effective doses can go, as they must retain diagnostic image quality while keeping radiation exposure to a minimum. That will depend in part on the type of abnormality they are expecting to locate.

"For example, if you have a patient who has a malfunctioning ventriculoperitoneal shunt in the brain - hydrocephalous imaging - you don't need to have the best-quality images for the brain parenchyma. Low-dose or 'noisy' images will still allow good assessment of the cerebral ventricles," explains Owens. "But if you're looking at tiny anatomical structures, your doses will be higher because you need to use more power from the machine to look at subtle differences in tissue density."

If a practitioner is focused first and foremost on image quality, the temptation will be to administer a high radiation dose. After all, overexposed CT looks visually attractive - subject a child to the same radiation levels as an adult and the diagnostic image will not suffer for it; unfortunately, the child's health may well do.

"Ten years ago, it was more difficult to scan children because you had to keep them still and the older scanners took much longer, so children were often sent to specialist centres for sedation and anaesthesia in order to undertake CT," Owens explains. "Now, because the scanners are so quick, it's very easy for adult units to scan children, and they may not be aware that the scanner parameters, and hence the exposures you require, are much lower."

Child of light

While the effective doses may vary hugely, appropriate parameters for an adult are around ten times higher than those recommended for a child. There are various reasons for this disparity. The first pertains simply to body size - use the same energy on a smaller patient and that patient will receive a higher radiation dose per unit of mass.

Secondly, children have more-sensitive cell replication. This means radiation exposure is potentially more hazardous to their DNA, particularly in areas such as the ovaries, bone marrow and breast tissue.

These risks are compounded by the fact that children have longer left to live. Because radiation-induced cancers don't appear until further down the line, younger patients are more likely to reach the end of that potential latency period and suffer from cancer in the long term.

Owens points out that it is difficult to quantify the hazards exactly, noting that the available data is frequently confusing and inconclusive. "Data is often based on people exposed to huge radiation incidents, like nuclear war or power station accidents," she says. "We try to extrapolate that data back into cancers induced from medical diagnostic radiation exposure, which is very difficult and imprecise."

That said, however, it seems clear that the attributable mortality risk from a single dose of radiation decreases dramatically with advancing age. According to research from the International Commission on Radiation Protection, the risk is 20% greater in those exposed aged 30 than those exposed aged 50, rising to 200% for 20-year-olds and a staggering 500% greater for under-10s.

On a similar note, a 2012 study in The Lancet reported that children who underwent multiple brain CT scans tripled their risk of brain cancer and leukaemia later in life.

The law of exposure

There is evidently a need for strict referral and practice guidelines to ensure paediatric CT is used judiciously. While new techniques have been developed to assist radiation reduction - not least ultra-fast multidetector CT scanners and newer post-processing techniques, such as iterative reconstruction, for example - it is still up to practitioners to ensure each test is suited to the individual patient's requirements.

And with so many pieces to the puzzle, the whole team, spanning many disciplines, needs to chip in.

"The production of good-quality CT scans is a very interactive job," says Owens. "The companies that make the machines, they have a responsibility, the radiographers who scan the patients have a responsibility and the radiologists reporting it have a responsibility.

"The medical physicist plays a pivotal role in quality assurance and has in-depth knowledge of the CT machines, which are becoming more complex and have enhanced capabilities to provide super-quality images with lower exposure."

She contends that developing clearer protocols will come down, in the first instance, to sharing data between colleagues. At Great Ormond Street Hospital, there are regular multidisciplinary meetings, designed to educate clinicians about when CT is and isn't necessary. For instance, MRI or ultrasound might often be preferable, but when a child can't be sedated or has a pacemaker, MRI can be contraindicated.

More broadly, she believes it will be important to harmonise guidelines across institutions and even across countries. If all radiologists are to stay informed about the latest dose-reduction strategies, it will be crucial to instigate and continue with an open dialogue.

"We have empirical guidelines and recommendations, but they're not set out universally across Europe, and we're trying to do that with the EU," Owens says. "I think it's very difficult to legislate across Europe, as the new machines are much more efficient than the older machines and some countries cannot afford this expensive equipment.

"We have to somehow allow different countries to operate the machinery they have to the best of their ability," she adds. "It would also be useful to have robust legislation for manufacturers, where they're required to issue specific weight or diameter-based protocols for their CT machines."

For the time being, however, paediatric CT remains a heterogeneous discipline, with different institutions using different machines in different ways. Owens would like to see a change in the culture. After all, with 100,000 scans performed on children in England in 2012, CT is here to stay - at least for the near future. The need for standardised recommendations is more salient than ever.

"Some of our clinical colleagues in different disciplines are surprised that there isn't more upfront legislation stopping people from using CT in children unless they comply to a certain standard," she says. "So maybe we're not quite strict enough."