Discussion

Sudden death cases create some of the most difficult situations for both clinicians and pathologists. They accounted for 9%-26% of all track deaths in a number of surveys (Boden et al. 2005). Wherever possible such cases should be referred to a diagnostic laboratory for detailed examination by a qualified, appropriately experienced pathologist in a facility that allows for hoisting, removal of all organs including the brain, and high quality photography. Such examinations are impossible in a stable or field. Significant problems can arise for practitioners attempting such examinations themselves, if legal action is taken.

History and submission:
The veterinarian should obtain a thorough history and submit that in writing with signature. This should include appropriate identification of the animal and indicate if death followed injections or recent exercise. Previous clinical history including recent blood results can be helpful. Photographs of the body (including identifying marks) and site of death should be provided, including presence of other horses/other animal species and samples of potential poisonous plants or chemicals. Other environmental information including, e.g. whether the body was found in water, and what the temperature has been since the animal was last seen alive (the time and date of which should be provided) can be of value in broad estimates of time of death and in attributing lesions as causative. These are forensic cases and as such a chain of custody should be established between the clinician and the pathologist and/or laboratory staff, i.e. uninterrupted control of the cadaver and any associated specimens.

Post mortem examination protocol and common lesions The protocol for post mortem examination will be reviewed, covering well-documented causes of sudden death that may be diagnosed grossly. In all surveys of sudden death the most
common causes have been cardiovascular accidents/haemorrhage
(Gelberg et al. 1985; Brown et al. 1988).

1. Skin and subcutis
It is important during the examination of superficial tissues, to determine if there are any areas of trauma (particularly that which is penetrating), and to focus on potential injection sites. In some cases of lightning strike, skin singeing may be identified. Tissues around the jugular vein and carotid artery should be examined carefully. Careful dissection of the carotid artery may be required, as inadvertent injection of various substances into it can be a cause of death. The whole body should be skinned, and subcutaneous, muscular and periarticular tissues searched. If any potential injection sites are identified they should be removed en bloc and saved in a sealed and labelling container (Rooney and Robertson 1996).

2. Cardiovascular
Important lesions include rupture of major blood vessels including the aorta, pulmonary artery, or uterine/ovarian/external iliac arteries (mares); haemorrhage from the internal carotid artery into the guttural pouch secondary to mycotic infection; cardiac

rupture; valvular lesions; myocardial lesions e.g. ischaemia or myocarditis; haemorrhage from unidentified sources; and other signs of shock including visceral congestion and petechial haemorrhages. There is no "shock organ" in the horse. Significant areas of the heart should be examined microscopically including the sinoatrial node and coronary arteries.

3. Gastrointestinal
Colic caused 36% of all deaths in one owner survey conducted in the US by the Morris Animal Foundation, and 39% of sudden deaths in one survey of 151 horses (Brown et al. 1988). Lesions may include: gastric/intestinal rupture; volvulus/entrapment/ intussusception/impaction of the small intestine; torsion/ intussusception/impaction of the caecum or colon; rectal tearing; colitis (e.g. clostridial); and diaphragmatic hernia.

4. Neurological
The skull, brain and spinal column should always be examined for signs of trauma. Slowly expanding lesions including neoplasms and abscesses may escape clinical notice but produce sudden death when they encroach on vital regions of the brain.

Found dead versus sudden death:
Strictly speaking, sudden death occurs only if the animal was closely observed and previously apparently healthy. Both acute and chronic lesions should be noted, as it is always possible that (i) clinical signs or traumatic incidents have not been noted by the owner or (ii) there is a chronic subclinical lesion that has undergone acute exacerbation, e.g. haemorrhage from a phaeochromocytoma.

Toxicology?
It is tempting to think that toxicological studies will provide an answer in the absence of obvious causes of death; however, unless there is some indication as to a likely agent, cost will generally preclude extensive investigations. It is still advisable to collect and freeze a panel of specimens including fluids, tissue specimens and ingesta/digesta in case any such testing is subsequently requested.

The likelihood of success:
Finally, it should be noted that a post mortem examination does not always allow determination of the cause of death. In multiple surveys, a definitive cause of death was only found in approximately 50-70% of cases (Platt 1982; Gelberg et al. 1985; Lyle et al. 2011). In one study a presumptive cause of death was made in further 25%, i.e. this was subject to pathologist interpretation (Lyle et al. 2011). The reality of success should be accepted by all involved parties and made clear to the owner and other concerned parties at the outset.