Finding Answers in Angiography

The complete diagnostic autopsy has changed little since its inception. It involves invasive surgery to examine the three body cavities – head, thorax and abdomen – and works on the assumption that it will provide enough information to identify a cause of death on a balance of probabilities.

But that examination omits many parts of the body, meaning it’s possible to miss significant information – especially if the person conducting the autopsy lacks training, experience, or sufficient time for a thorough investigation. That’s where post-mortem cross-sectional imaging – computed tomography (CT) in adults and magnetic resonance imaging (MRI) in children – enters the picture.

Post-mortem computed tomography (PMCT) alone can identify 60–70 percent of natural causes of death; that alone isn’t sufficient, so the technique is most commonly used as an adjunct to invasive autopsy. Why does PMCT fail to provide an answer in so many cases? The principal reason is that you can’t confidently diagnose coronary artery disease – a leading cause of natural death – with PMCT alone. To see that, you have to add in angiography (PMCTA), which increases the diagnostic success rate to about 92 percent of all natural causes of death, as well as 100 percent of non-suspicious trauma cases (1). So if you want to know absolutely everything about a particular case, you do both – but if you need a more volume-based system, you can save time and resources by starting with PMCTA, triaging out those cases that need to be autopsied.

How it’s done

We use a triage system to determine the investigations needed for each case – the idea being that you don’t go any farther through the process than you need to answer your specific question (2).

We begin with a clinical history, information on the place of death, and establishment of “the question.” An external examination follows to ensure there are no suspicious aspects to the death; at that stage, we also prepare the body for PMCTA – and, if necessary, take toxicology samples.

The PMCTA procedure itself begins with the insertion of a catheter into the left common carotid artery through a small (~2 cm) incision at the base of the neck. We advance the catheter to just above the aortic valve of the heart, inflate a balloon to secure it, and at the same time insert a small tube into the airway to ventilate the lungs and enhance their imaging. Then we perform a native CT scan – that is, we scan the body from head to toe without contrast or ventilation to check the position of the catheter and, if possible, identify the main pathology.

At that stage, if there is a life-threatening pathology – for example, a ruptured abdominal aortic aneurysm – we can stop the process, because we have answered the question. If no such finding is present, we then inject a small quantity of both air (negative) and then positive contrast medium into the coronary arteries for a double-contrast examination. To distribute the contrast, we have to roll the body; we do that five times and image the heart and vessels each time. We also ventilate and image the lungs. Finally, we remove the catheters, close the incision, and place the body back into storage.

We then report those images – and if, on a balance of probabilities, there is a cause of death that is consistent with the external examination and the history, then we provide a cause of death and go no further. If not, we decide whether to proceed with a limited or a full autopsy (triage), which would hopefully yield any missing information. Laboratory tests would be next in the chain; we simply keep investigating until we answer the question.

Where it came from

About 100 years ago Upton Sinclair said, “It’s difficult to get a man to understand something if his income depends on him not understanding it.” I think this could be one of the reasons why PMCT has not been widely adopted to date.

The technique has been used in autopsy practice since the early 1980s, but always as an adjunct to traditional methods. In 1994, a group in Israel undertook a study of suspicious deaths and concluded that, in trauma cases, PMCT could replace autopsies. However, the outcome was not very well received by the general pathology community.

Since then, though, the field has undergone transformational change: improved technology and software – and a more open view to the concept. A number of research groups around the world – in Switzerland, Japan, and the United Kingdom – have looked at different aspects of minimally invasive autopsy. The result? In 2017, we (my co-authors and I, and undoubtedly a wider number of researchers, practitioners and authors) agree with the 1994 research that the vast majority of death investigation could be done purely with PMCT. So in essence, we’ve come full circle; we’ve established an evidence base for the early work and come up with the same answer.

Why is PMCT now at the forefront? Firstly, I think there is a perception that members of the public don’t want traditional autopsies; there are cultural, religious, ethical, and even aesthetic concerns – and minimally invasive techniques are often acceptable where a complete diagnostic autopsy may not be. Secondly, we use the technique more than ever now, and I myself know that it yields equivalent – if not better – findings than actually dissecting the body. It makes me question my own practice, and that’s a good thing. If I know that there’s an equivalent or better system to achieve the same result, why aren’t I using it?

To PMCT or not to PMCT?

We use PMCTA whenever we need to image blood vessels. Our standard technique targets the coronary arteries, but there’s another method – multi-phase PMCTA – that looks at the other vessels of the body. So it comes back to asking the right question; you must decide what you want to learn from your investigation, and then you can decide which of the two techniques is better suited to your needs.

There are also situations where PMCTA is not the solution. At the moment, we can’t confidently diagnose sepsis, meningitis, gastrointestinal bleeds, or pulmonary thromboemboli with these minimally invasive techniques. You might still use them, but only as an adjunct to a limited or even a full autopsy. We’re trying to improve PMCTA for those scenarios where it is not yet ideal, though; at the 2017 meeting of the International Society for Forensic Radiology and Imaging, our unit presented work on a system that identifies most, though not all, pulmonary thromboemboli (publication in development). It’s a system I’m optimistic will increase the diagnostic ability of PMCTA further toward the 100% goal for natural death.

At the Victorian Institute of Forensic Medicine in Australia, I understand they perform approximately 6,000 of these examinations every year. In Scandinavia, most mortuaries have their own CT scanners and use them routinely. Large areas of continental Europe have adopted PMCTA for death investigation, and it’s now even making inroads into the United States. Penetration into the UK has been comparatively slow thus far, but it’s gaining momentum. All of the research proving its suitability has already been successfully completed, so all that’s left is to engage in the process – and, of course, to overcome issues of cost and bureaucracy. But I envisage that it will become the standard form of autopsy examination in the not-too-distant future – and that invasive autopsy will be limited to specific cases.

Who is responsible?

Right now, there is some debate as to who should report the images when undertaking PMCTA. In most countries, the pathologist reports the images under the guidance of a radiologist – but some people argue that it should be the other way around.

My colleagues and I at Leicester run the only UK-based training programs for people who want to work on PMCTA (3). We train the mortuary staff – the APTs and the radiographers – to prepare and scan the bodies. We also help educate radiologists, because they know the anatomy and the disease processes, but need minor additional training in relation to the changes that occur after death and the medico-legal questions they haven’t previously dealt with. Pathologists, too, need additional training if they plan to report scans – I, for instance, report all my own scans under the guidance of a radiologist, but I’ve had to learn how to read CT scans and use diagnostic software. Personally, I don’t think either way – having pathologists report under radiologists’ guidance or vice versa – is objectively better. All that matters is that it’s done properly, and that the person who does it understands the changes that occur after death and the questions being asked.

Importantly, though, I’d like pathologists to know that they don’t need to worry about this kind of change. I’ve heard concerns – “Does this mean we aren’t going to do autopsies anymore?” “Are we no longer involved in death investigation?” – but that’s not the case at all. Rather, it gives us a completely new perspective on the body. We can see and appreciate things that just can’t be seen or appreciated in a traditional autopsy. Modern scanners and software generate detailed images and even allow users to virtually manipulate the body in 2D and 3D, to the point where I truly believe that CT should be the gold standard of clinical practice in the dead – just as it is in the living. So if you really want to fully understand the human body and the process of death, then my advice is: get involved. It will change your entire autopsy practice, revitalize your interest in it, and take you into a world that you just can’t fully appreciate any other way.