A 70 year old female was admitted into another hospital in Japan, following 3 episodes of gastrointestinal haemorrhage.
She had first noticed dark, black stool the day before admission and she had slight lower abdominal discomfort.
However, by the next day, this was followed by fresh red blood on defecation. There was no information about how much bloody stool was produced.
The patient had no other symptoms including no nausea, haematemesis, vomiting, epigastric pain, dizziness, dyspnoea, palpitations etc...
The patient had known severe coronary artery disease and had had a previous myocardial infarction resulting in 100% stenosis of the RCA, LCX arteries and a patent LAD which housed a stent to maintain its patency. Not only that, an apical thrombus had been diagnosed due to dyskinesis of the left ventricle, several months before for which anticoagulation was used.
The patient was also receiving aspirin and ticlopidine but no proton pump inhibitor (PPI).
The patient was not using alcohol. There was no history of any previous GI complaint.
Previous medical history apart from the above also included Diabetes Mellitus (Type 2) for which a sulphonylurea was used for control.
She was a non-smoker.
On examination the patient looked unwell and pale. The blood pressure was 80/40 despite having received 3 litres of crystalloid fluid. Pulse was 120/min and regular. Temp 37.0
Resp Rate 14/min and O2 sats 97% on room air.
CVS: Pulse quality was reduced at the radial artery; it was difficult to find. Left hand middle and ring fingers were cold to touch. JVP was not elevated. No heaves or thrills. Heart sounds were normal. No peripheral oedema was found.
RESP: Trachea was central. Hyper-expanded chest. Slight crackles evident at the right base only.
ABDOMEN: Soft, non-distended, no organomegaly, palpable colon from faeces only. No rebound or guarding. Bowel sounds increased. Rectal examination: Fresh Blood and no masses.
Bloods revealed a raised BUN/Creatinine ratio, Hb 6.0, INR 3.0. Normal CK and normal Liver Function Tests (three hours before, the Hb was 11g.dl)
One area with highest risk from bleeding was the Upper GI tract such as the stomach or duodenum in view of the history of anti-platelet agents and malaena, although lower GI haemorrhage could not be totally excluded because of the lower abdominal discomfort that was non-specific. Sometimes, a caecal focus of bleeding can produce stool that looks similar to malena. A raised BUN to Creatinine ratio can signify upper GI haemorrhage but in this case it might also be as a result of renal hypoperfusion from shock-- there is simply no way to accurately tell unless a gastroscopy is performed !
However, this patient was not straight forwards because despite the bleeding, by reversing the coagulation might result in either stent thrombus or embolisation of the apical thrombus. This is a so-called Catch 22 situation i.e. you cannot win either way. You are damned if you do and damned if you don't..... :-(
It was clear that by not reversing the coagulation disturbance the patient would exsanguinate.
A senior doctor who reviewed the patient was suspicious that the combination of warfarin plus two anti-platelet agents was the cause for the bleeding. It was suggested that the following things should be instituted:
- Blood replacement to aim for an Hb>10g/dl especially as coronary heart disease
- Reduce the INR with Vitamin K (low dose e.g. 0.5-1.0mg only), plus Fresh Frozen Plasma (FFP) to aim for an INR of 1.5 or thereabouts
- Intravenous Proton Pump Inhibitor
- Stop anti-platelet agents (temporarily)
- Urgent gastroscopy and colonoscopy. The former was suggested because of the raised BUN to Creatinine ratio.
- Insert a Foley catheter to measure hourly urine output
- Aim for systolic BP >100mmHg
- Repeat echocardiogram
A gastroscopy was not performed and hence, an upper GI bleed was never excluded here.
Repeat transthoracic echocardiogram revealed that the thrombus located at the apex had become smaller in size as one would expect on treatment.
In this case, what one has to decide is which problem is going to be immediately life threatening, in this case, GI haemorrhage. Hence, once the bleeding had been stopped e.g. by clipping bleeding vessels and stopping harmful drugs temporarily e.g. aspirin, ticlopidine and warfarin and reversing coagulation disturbance, then treatment to prevent stent thrombosis / thromboembolisation would be the next step.
In view that the thrombus had reduced in size (and was more likely to be organising) and that with time, stent thrombosis is relatively less compared to the time when they are initially inserted, the relative risk of thrombosis/embolisation versus uncontrolled haemorrhage tipped the scales in favour of concentrating on stopping bleeding first. Remember the emergency ABC- without a Circulation the heart will have no blood supply for itself or for the rest of the body! Hence, stopping bleeding and maintaining the circulation is essential.
There were certain areas with this cases where management could have been optimised, for example:
- Performing Gastroscopy to exclude an upper GI source of haemorrhage (remember, there was malaena too and raised BUN/creatinine ratio!)
- Use of FFP-- essential. Reversing the INR (in this case, probably an incomplete reversal of INR would suffice) until all bleeding had ceased then followed by intravenous heparin for cardiac protection. Use of heparin is preferred in such circumstances because if bleeding recurs, the heparin can be immediately stopped and if need be, protamine sulphate can be used to reverse it. The only concerns here would be recurrent bleeding and the potential Heparin-Induced Thrombocytopaenia. HIT does not happen immediately and the risk increases the longer the heparin is used, usually over several days to weeks.
Hence, the idea that hoping to maintain a high INR for coronary protection and at the same time clipping bleeding areas in the face of generalised mucosal bleeding that could result in multiple torrential sites of bleeding at any time and anywhere needs a re-evaluation. It is like trying to plug holes in a kitchen sieve- a self-defeating objective. Not performing a gastroscopy again needs reconsidering because this patient had generalised mucosal bleeding and was probably bleeding from every part of the GI tract.
Remember, it is not just GI tracts that bleed on warfarin with dysfunctional platelets, anywhere can bleed, but especially the brain. The GI tract is a good guide (and the skin too) to know what is going on with coagulation in the rest of the body and with mucosal bleeding then be warned because your patient could have a intracerebral bleed. The important thing here was to control the clotting first and thereby control the bleeding to make the patient haemodynamically stable and then image the GI tract and then consider starting the heparin to protect the heart.
In this case, the patient's INR was reduced to 1.4 on Vitamin K alone and the repeat colonoscopy reconfirmed multiple areas of previous haemorrhage.
These kinds of case are not described in your average medical text book as no case is the same as the next. Doctors should always expect the unexpected and hope for the best, but plan for the worse. An element of uncertainty is always going to occur in a case such as this, and sometimes, keeping your nerve and following long-upheld evidence based rules but at the same time taking a leap of faith into the unknown is required in order to help your patient.