Editor–We read with interest the report by Sen Gupta and colleagues on a challenging presentation of antineutrophil cytoplasm antibody (ANCA)-associated vasculitis (Clin Med August 2011, pp 368–71). However, we are concerned by their recommendation that thromboprophylaxis should be instigated in such cases.
Gastrointestinal (GI) bleeding is a well-recognised complication of acute kidney injury (AKI), occurring in approximately 15% of patients in one series, where it was associated with prolonged hospital admission and an increased risk of death.1 Factors such as uraemic platelet dysfunction and stress (peptic) ulceration secondary to critical illness are thought to contribute to this risk. For this reason, gastro-protection with H2 receptor antagonists or proton pump inhibitors is often advised in AKI (although there is no randomised control data to support this recommendation). The recent exposure to non-steroidal anti-inflammatory drugs was an additional risk factor for GI bleeding in this case.
This particular patient also demonstrated features suggestive of systemic vasculitis at first presentation (including AKI with an active urinary sediment, thrombocytosis, severe anaemia, reduced alveolar-arterial gradient with infiltrates on the chest radiograph) and the possibility of life-threatening pulmonary haemorrhage (present in 12–29% of patients with microscopic polyangiitis2) should be considered at the outset. No reference was made in the report to assessing gas transfer factor which, if increased, would be a useful indicator of pulmonary haemorrhage in this setting. The severe anaemia may also have alerted to the possibility of gut vasculitis.
Finally, this patient required a renal biopsy to secure a diagnosis and allow the initiation of definitive treatment; this procedure can often be delayed or complicated in patients who have received anticoagulants inappropriately. For these reasons, we would caution against the routine use of thromboprophylaxis in patients with AKI, particularly in those with features suggestive of coincident pulmonary or gut haemorrhage. It should also be noted that anticoagulants, such as low-molecular weight heparins, may accumulate unpredictably in renal failure and thus require dose-modification in accordance with estimated glomerular filtration rate.
Footnotes
Please submit letters for the editor's consideration within three weeks of receipt of Clinical Medicine. Letters should ideally be limited to 350 words, and sent by email to: clinicalmedicine{at}rcplondon.ac.uk
- © 2011 Royal College of Physicians
Acute renal failure in diabetes: looking beyond diabetic retinopathy
We thank McAdoo and Pusey for their interest in our case report and for their comments on the bleeding risks in AKI. In response to the points raised:
Our standard policy in AKI is to use gastroprotective proton pump inhibitor therapy as they suggested. In this case we used lansoprazole 30 mg once daily considering the associated GI bleeding risk given the patient's recent use of non-steroidal anti-inflammatory drugs and the steroid therapy for treatment of his ANCA-positive vasculitis.
Pulmonary haemorrhage was considered and gas transfer factor was measured (64%). As it was not elevated, we did not mention it in the case report.
We agree that low-molecular weight heparins may accumulate in renal failure and that their use in AKI is controversial. However, this patient undoubtedly had a higher than usual risk of thrombosis and we therefore selected to use a low dose of daily enoxaparin (20 mg).
On balance, although there is a significant GI and pulmonary haemorrhage risk in AKI and in ANCA-positive vasculitis, this man also had a sizeable risk of thrombosis given his diabetes with hyperosmolality on presentation, AKI and sepsis. We therefore maintain that thromboprophylaxis should be considered in similar cases, given the trial evidence1,2 but that this should be balanced against the calculated bleeding risk which should be assessed for each case.
Footnotes
Please submit letters for the editor's consideration within three weeks of receipt of Clinical Medicine. Letters should ideally be limited to 350 words, and sent by email to: clinicalmedicine{at}rcplondon.ac.uk
- © 2011 Royal College of Physicians
Jump to section
Related Articles
Cited By...
- No citing articles found.