Past the scalpel - post-mortem cross-sectional imaging

Although the invasive autopsy is generally considered the gold standard for the investigation of death, many practitioners and researchers have questioned this. A number of alternative approaches have been considered in order to reduce the use of the autopsy or to make it unnecessary entirely.

This thinking has been driven by a dislike for invasive autopsies, from the general public and within the medical profession; the requirements of religious doctrine and law in our multicultural society; and the need for a more cost-effective method of death investigation.

New imaging technology has been slowly implemented in autopsy practice over the past 30 years; as a result, post-mortem cross-sectional imaging has emerged as a true alternative to the traditional invasive autopsy, and is being suggested within the international medical literature as a new gold standard for the investigation of death for adults and children.

Similar to the history of X-rays themselves, investigators rapidly realised the potential of cross-sectional imaging in the form of computed tomography (CT) for forensic examination. Clinical CT brain scanning was introduced in 1972, and by 1977, its use was documented in clinical forensic practice for the investigation of cranial ballistic injury.

The first documented use of CT in autopsy practice was in 1983, as an adjunct to autopsy in the investigation of diving-related deaths. However, it was not until 1994 that it was suggested, in a paper from Israel, that CT could be used not as an adjunct but as an alternative to autopsy. This was followed in 1996 by the suggestion that magnetic resonance imaging (MRI) could provide an alternative to perinatal autopsies.

Between 1996 and 2003, there was a slow increase of academic publications exploring the use of CT in the investigation of death. However, interest only really accelerated at the beginning of this century, with international teams, such as the Virtopsy group in Switzerland, investigating and publishing cross-sectional techniques using purpose-built mortuary scanners.

Since then, we have witnessed a global expansion in research, and the routine use of post-mortem cross-sectional imaging for adults and children. Scanners have been placed into mortuaries to facilitate diagnostic services, as well as being adopted into mass-fatality investigations. In 2012, the International Society for Forensic Radiology and Imaging (ISFRI) was founded along with the first dedicated forensic imaging journal, the Journal of Forensic Radiology and Imaging. ISFRI, through its journal, annual conference and positional statements from its working groups, aims to promote, advise upon and expand the use of post-mortem cross-sectional imaging across the world.

Body of evidence

During the past 30 years, a number of different terms such as 'virtual autopsy', 'digital autopsy' and 'Virtopsy' have been used to describe the same general concept. In 2013, an international group put forward a proposed unified system of nomenclature for 'post-mortem cross-sectional imaging', to distinguish this field of practice from other forms of radiology.

This term encompasses different imaging modalities, including post-mortem computed tomography (PMCT) and post-mortem magnetic resonance imaging (PMMR), with the former being the predominate system used in adult cadaveric investigations internationally.

A plain PMCT scan through the body with no specific preparation, additional procedures or image enhancement with contrast agents, is often called a 'non-contrast' or 'native' scan. This is the only truly non-invasive form of PMCT. However, a native scan cannot visualise vascular lumens, and particularly the coronary artery lumen, and is therefore incapable of identifying coronary artery stenoses or occlusions, the most common cause of adult sudden unexpected natural death.

Therefore, like in clinical practice, most PMCT is performed with some form of contrast enhancement to depict the vasculature. This contrast can be targeted - focused on the coronary arteries only, for example - or generalised, to give a whole-body vascular enhancement by dispersing contrast in a manner similar to embalming. Furthermore, imaging can also be enhanced by expanding the lungs using a ventilator to mimic breath-hold in inspiration, as is used for most clinical CT scans.

Imaging can also be augmented by imaging guided acquisition of biopsy specimens or body fluid for toxicology and biochemistry. These techniques require invasive procedures, including inserting a needle or catheter into the blood vessels or siting a cuffed tube in the trachea using a cricothyroidectomy approach, as may be used clinically. This means that PMCT may not always be non-invasive, but generally can be considered minimally invasive when compared with autopsy.

There are two reasons to consider using PMCT as an adjunct to an autopsy or as a replacement. As an adjunct, PMCT can provide a lot of extra useful information for unnatural deaths, such as trauma, suspicious or homicide deaths, but in these circumstances, the police and courts are likely to still expect an invasive autopsy. In addition to cause of death, PMCT can also provide information to help with identification and time of death, supporting, if not enhancing, the invasive autopsy findings. PMCT may therefore be very useful in mass-fatality investigations.

PMCT is excellent for the investigation of adult skeletal trauma, and when combined with post-mortem angiography (PMCTA) and ventilation (VPMCT), can assist with the identification of specific bleeding points and penetrating trauma paths through the thoracic cavity. It has a role in the location and subsequent retrieval of evidence on or within a body. It can help identify potential health hazards, such as unsuspected infective diseases like tuberculosis or sharp foreign bodies.

It can also assist in the planning of special autopsy procedures - seeking to confirm cardiac air embolus or pneumothorax, for example. The images form a permanent record of the body before dissection, and can therefore be used for audit of practice, a second independent autopsy review and demonstration of findings to the courts.

It can be argued that PMCT is already capable of providing sufficient internal information to act as an alternative to invasive autopsy in many trauma cases, such as road traffic collisions, stabbings and gunshot/shotgun injuries. However, to date, this is a step too far for UK practitioners, and its role as an alternative to autopsy is currently confined in England and Wales to the investigation of sudden unexpected natural death by Her Majesty's Coroner (HMC).

As about 80% of these deaths are reported to be due to coronary artery disease, PMCTA with contrast enhancement of the coronary arteries can show disease in the coronary arteries and therefore diagnose natural cardiac causes of death, avoiding the necessity for an invasive autopsy. Native PMCT can also identify significant cavity bleeds such as cerebrovascular bleeds (stroke and sub-arachnoid haemorrhages) and ruptured aneurysms.

By using VPMCT, more confidence can be achieved in distinguishing true pneumonia from post-mortem respiratory changes. We must also recognise that diagnostic confidence in PMCT and autopsy is heavily influenced by the fact that an experienced autopsy practitioner could predict the autopsy findings on case history alone in up to 80% of HMC cases.

Cautious optimism

There remain limitations to consider with PMCT, though. Purchasing and siting a CT scanner, and providing trained staff is expensive and may be difficult. However, there is no specific reason why a PMCT scan should be more expensive than an autopsy for the investigation of death as long as there are enough cases to justify the upfront costs of the scanner. Purchase costs can clearly be avoided if an adjacent clinical scanner can be used out of standard clinical hours - overnight, for example.

For traumatic death, PMCT is generally better than autopsy for depicting bone trauma, particularly at the base of skull and spine, and can show areas of significant haemorrhage well, but it can be limited for some soft tissue and organ pathologies. For example, although we can see fractures in trauma, we may miss internal bruising. A full external exam is required as PMCT may fail to detect some incised wounds, lacerations and even bullet entry wounds. Soot in an airway or stomach, due to death in a fire, will be missed and are important diagnostic observations.

Even with PMCTA and VPMCT, penetrating trauma may not be seen in sufficient detail to recreate a crime scene for full legal investigation. What is becoming clear is that PMCT and autopsy are complementary; the real gold standard is now to use both tests.

For cases of natural death undergoing investigation for HMC, PMCTA with coronary artery angiography will give a confident cause of death in more than 80% of cases, on the 'balance of probabilities', the criteria required for the coroner's court.

There will always be the need for an invasive autopsy for the rest. Certain conditions remain problematic; for example, even with angiography, it is difficult to discern pulmonary thromboembolus from post-mortem clot, and this diagnosis remains difficult. Also, despite PMCT being accurate for most haemorrhages, when confined to the lumen of the gastrointestinal tract, they can be occult when death has occurred before the onset of haematemesis or melena.

These problems are likely to change as more research is undertaken and the evidence base expands. This will generally result in PMCT failing to establish a cause of death. However, in the elderly with multiple comorbidities, there is a risk that a plausible cause of death is given based on potentially fatal pathology but that the actual cause of death is missed. However, using PMCT to audit autopsies shows this is a risk for the invasive autopsy as well.

If PMCT fails to establish the cause of death, autopsy may be required. However, this autopsy can be more focused, based on the PMCT findings.

Putting autopsy in the picture

In 2012, the Department of Health (DoH) published a visionary document considering the introduction a national PMCT-based autopsy service. The underlying principle of this document, although not specifically stated, was to follow current NHS services and be a service for all, free at the point of delivery.

To date, this vision has not been realised. The principle reason for this appears to be cost - of facilitating the placement of scanners into mortuaries and of running the service, which would have to be borne by local councils. There are also manpower and training implications. It is not dead in the water, but there are problems.

Within the private sector, there has been the slow, piecemeal introduction of PMCT services within England, including the placement of scanners into mortuaries. However, introducing the service in this manner does not resolve the problems of training, accreditation, competency, audit and quality control, which should underpin any national medical service.

As these private services are paid for directly by the relatives, not HMC, then religious selection and ability to pay may become key issues. Furthermore, access will depend on locality and potentially whether some local council starts to fund a service, risking a postcode lottery of service provision. The vision of the DoH document rejected this piecemeal introduction of services for these very reasons, but we now have to accept what is happening and evolve to address any problems in quality and fairness.

We are in no doubt that, after review of case history and inspection of the body, PMCT, and PMMR for children, are the future first line for investigation of death. Associated tests, such as histology, biochemistry and toxicology, can be performed with PMCT or autopsy. The gold standard is now cross-sectional imaging in conjunction with full or limited autopsy, but there is no doubt that PMCT is potentially suitable as a replacement to invasive autopsy in the right situation. We believe this requires enhanced scanning to diagnose coronary vascular disease.

There are many areas that remain unresolved, but we believe that PMCT will soon become the commonest form of death investigation. As is the case now for many clinical imaging indications, commentators in 20 years' time will probably look back and wonder what the fuss was all about.