Circulating Inhibitors, Lupus Anticoagulant and Heparin Contamination: How to Interpret Abnormal Coagulation Tests

Consider this scenario: inpatient with bruising has tests for coagulation. The APTT is prolonged at 40 s (normal <35). Is this abnormal? Is it due to the prophylactic clexane he is receiving? How do you tell? Heparin contamination of inpatient samples is common. However, the causes of a prolonged APTT with normal PT are varied, and must be considered in an (usually) elderly patient with ecchymoses and bruises. Aetiologies include heparin, LMWH, low dose dabigatran, inherited deficiencies of factor VIII, IX and XI or VWF in young persons(with bleeding) or factor XII, HMWK or prekallikrein (without bleeding) and ciculating inhibitors to Factor VIII (commonest), seen typically in postpartum women, in autoimmune disease such as SLE and RA, and in cancer, or inhibitors to factor IX or XI.

If heparin contamination is suspected in a repeatedly prolonged APTT (always repeat the test), the best test is a combination of thrombin time and reptilase time. Thrombin time (TT), performed by using the patient's plasma with human or bovine thrombin, will be prolonged in heparin contamination, while reptilase time (RT), performed by mixing the patient's plasma with a venom obtained from Bothrops snake remains normal. The venom, unlike thrombin, cannot be inhibited by heparin or LMWH or by direct thrombin inhibitors.

Alternatively, use an inhibitor of heparin such as protamine or a heparin binding resin such as Heparsorb and repeat the test. APTT will normalise if it was prolonged due to heparin.

In cases of hypofibrinogenemia or dysfibrinogenemia, both TT and RT will be prolonged.

A circulating inhibitor is often suspected in adults with ecchymoses or bruising, occasionally with associated joint swelling due to haemarthrosis, prolonged APTT, normal PT, and no alternative causes such as liver failure, DIC (usually pronlongs both PT & APTT) or concurrent anticoagulant adminstration. The commonest circulating inhibitor is to factor VIII. While this is typically seen in haemophiliacs, it can also be seen in postpartum women, usually primigravidas, and as stated above, with autoimmune disease or cancer.

After ruling out heparin contamination as above, the patient's plasma is incubated 1:1 with normal plasma for 1 to 2 hours at 37 degrees celsius (to allow stable binding of IgG antibodies, which work best at body temperature) . APTT will correct with factor VIII, IX, or XI deficiencies or with vWD. If the APTT does not correct, it's either a circulating inhibitor, usually to Factor VIII, or lupus anticoagulant. The latter presents with thrombosis rather than bleeding and can be easily ruled out by one of two tests. First, adding phospholipid (usually cephalin from brain tissue) normalises the APTT in subjects with lupus anticoagulant. Here, one must be careful in performing a hexagonal phase assay rather than a lamellar assay. Hexagonal phase assay includes adding phosphatidylethanolamine rather than phospatidylcholine, as is found in lamellar (bilipid layer) assays. Although the latter is found in physiological membranes, circulating lupus anticoagulant is only inhibited by phospholipids present in the hexagonal form, comprising packed phospatidylethanolamine in a six-cylinder configuration. Using lipids in a lamellar form will lead to a false negative test.

The second test is to simply repeat the clotting with the dilute Russell Viper Venom Time (dRVVT). This bypasses all factors upstream of Factor X. Since Factor X still needs phospholipids to activate thrombin, it is abnormal with antiphospholipid syndrome, but not with circulating Factor VIII inhibitor.

It is highly likely that a person with prolonged APTT that does not correct with normal plasma, and with a normal dRVVT has a circulating inhibitor to Factor VIII. The next step is the Bethesda assay. Here, increasing dilutions of the patient's plasma is added to normal plasma, and assay is performed for Factor VIII activity, as in haemophilia. The reciprocal of the dilution that produces 50% of the normal Factor VIII activity is expressed in Bethesda Units (BU). More than 5 BU is considered abnormal.

Therefore, I'd suggest the following algorithm when one comes across an unexplained rise in APTT with normal PT.

1. Repeat test.

2. If APTT still prolonged, rule out heparin contamination by checking Thrombin Time and Reptilase Time. If TT is prolonged and RT is normal, heparin is present. If both TT and RT are prolonged, proceed to the next step.

3. Dilute the patient's plasma 1:1 with normal plasma and incubate for 60 minutes at 37 degrees Celsius. If APTT normalises, the patient has deficiency of one of the following factors: Factor VIII, IX, XI, VWF, Factor XII, HMWK, or prekallekrein. Deficiency of one of the last three prolongs the APTT, but does not lead to bleeding.

4. If the APTT does not correct with addition of normal plasma, the patient either has lupus anticoagulant or a circulating inhibitor to one of Factor VIII (commonest), or Factor IX or XI. Lupus anticoagulant leads to clotting rather than bleeding in vivo.

5. Test for lupus anticoagulant by doing one of these two tests: Do the dilute Russell Viper Venom Time. If dRVVT is prolonged, patient has Lupus anticoagulant.

Alternatively, add phospholipid to the mixture of normal and patient plasma and repeat APTT. If the latter returns to normal, lupus anticoagulant is present.

6. If the dRVVT is normal or if APTT is not corrected by adding phospholipid, a circulating inhibitor to Factor VIII, IX or XI is present. Inhibitors to Factor VIII are by far the commonest, and should particularly be suspected in postpartum women, usually after their first child, or older subjects. In such cases, proceed directly to the Bethesda assay. Presence of more than 5 BU is consistent with presence of Factor VIII inhibitor.

Graft Versus Host Disease Associated With Blood Transfusion

GVHD typically occurs after HSCT as the donor lymphocytes attack antigen presenting cells with a different HLA haplotype in host tissues, typically in the skin, liver and gastrointestinal tract. But can you get GVHD in non-transplant subjects?

Normally, blood transfusions would not be expected to cause GVHD, as the recipient's immune system would destroy lymphocytes in the donor blood.

However, GVHD has been demonstrated after blood transfusion in subjects who have not had a transplant. This typically occurs in two settings.

Firstly, the subject is immunosuppressed, either because of an underlying T-cell disorder, as in children, or because of chemotherapy or radiotherapy for haematologic or solid cancer. Hence the donor lymphocytes are free to run amok.

The second setting is unique and occurs in immunocompetent subjects. This typically happens when a subject who is heterozygous for a HLA antigen receives a blood transfusion from a donor who is homozygous for that HLA antigen. Thus, the host's lymphocytes do not recognise the blood donor's lymphocytes as foreign, but the donor lymphocytes recognise the "mismatched" HLA antigen in the host tissues as alien. This kind of mishap is likely to occur when there is limited genetic variability in the population, as in Japan, or when the blood donor is a family member, particularly if there is a high rate of consanguinity, thus causing shared HLA antigens, such as in rural Turkey or on the subcontinent.

Unlike transplantion associated GVHD, transfusion associated GVHD is almost always fatal. This is because apart from attacking the skin, liver and GI tract, lymphocytes in the donor blood also attack the host's bone marrow, causing aplasia. Remember, that in HSCT, the marrow has been ablated, and replaced by the donor marrow, which will therefore not be attacked by the graft lymphocytes.

Transfusion associated GVHD starts 4-30 days after transfusion, is usually mild and insidious in onset and is therefore often not recognised, and is usually attributed to the underlying disease. It starts with fever and rash, which can become erythrodermatous in severe cases, and proceeds to vomiting, severe diarrhoea, right upper quadrant pain, deranged LFTs, and pancytopenia.

Transfusion associated GVHD can be caused by transfusing whole blood, packed RBC, platelets, leucocytes and fresh unfrozen plasma. It does not occur when transfusing deglycerolated, frozen RBC or cryoprecipitate or FFP.

Diagnosis can be made by a skin biopsy and by demonstrating that the circulating lymphocytes are the donor's rather than the host's.

This devastating complication can be prevented by gamma-irradiating donor blood in the following 4 instances- in immunosuppressed subjects, in subjects who have had HCT (although GVHD here would not affect the marrow), in instances where the blood donor is a relative and in HLA-matched platelet transfusions.

Diagnosing Renal Artery Stenosis- Look for Cardio-renal Syndrome and Flash Pulmonary Oedema

We all know renovascular hypertension exists. But how does it present clinically? Are their any distinguishing features that set it apart from other causes of hypertension and from subjects with essential hypertension?

Renovascular hypertension has two principal aetiologies. By far the commonest aetiology is atherosclerotic narrowing of the proximal renal artery. This can be predominantly unilateral or bilateral.

A much less common cause is fibromuscular dysplasia, affecting young females often in the 30-50 year range. This typically affects the distal renal artery, and for some reason, tends to be predominantly right sided.

So how does renovascular disease present? There are two terms often used to describe this condition, without realising that this is in fact the underlying aetiology. The first is cardio-renal syndrome and the second is flash pulmonary oedema.

Probably the most characteristic presentation of atherosclerotic renal artery stenosis is rapidly worsening hypertension, worsening renal impairment and incremental heart failure. This is often erroneously interpreted as cardio-renal syndrome, particularly in subjects with pre-existing heart failure, but the key finding is the hypertension. Subjects with longstanding heart failure, the cohort who develop cardio-renal syndrome, often have low or low-normal BP. They should not show a trend for a recent increase in BP.

This classical presentation has been somewhat tempered in the last decade or two by the widespread use of ACE inhibitors and direct angiotensin blockers for treating both hypertension and heart failure, but when present, is strongly suggestive of severe renal artery stenosis.

Some physicians measure serum renin and aldosterone levels to make a suggestive diagnosis of renovascular hypertension, before considering imaging studies such as CT angiogram. However, this could be misleading. While renin and aldosterone levels would be expected to be high in predominantly unilateral renal artery stenosis, these hormones are normal in bilateral renal stenosis and in subjects who have renal artery stenosis in a single surviving kidney.


The other, albeit, less common presentation is flash pulmonary oedema. Far too often, clinicians invoke the rare diagnosis of phaeochromocytoma in such subjects, when the blame lies with a much more common condition. Renal artery stenosis should always be suspected in subjects who have one or more episodes of "flash" pulmonary oedema.

Trautmann's Triangle & Otogenic Brain Abscess

A 64 year old woman presents with fever, headache and a discharging right ear.

A CT scan shows appearances consistent with a large right cerebellar abscess. This is later confirmed on a MR scan. You start antibiotics to cover S. aureus, gram negatives and anaerobes and given the discharging ear, decide to send off an ENT consult.

What putative diagnosis would you put on the consult request? Acute otitis media? Chronic suppurative otitis media?

Surprising though it may seem, in a case series of 40 otogenic abscesses, acute otitis media was not implicated even once. In all but one case, the predisposing factor was a cholesteatoma.

With otogenic brain abscesses affecting the temporal lobe, spread occurs through the tegmen tympani.

For abscesses affecting the cerebellum, spread from the affected ear occurs through the Trautmann's triangle, an area bounded by the sigmoid sinus posteriorly, bony labyrinth anteriorly, superior petrosal sinus superiorly and the internal jugular vein inferiorly.

On CT, an abscess may be confused with a necrotic tumour. The best arbiter is a diffusion weighted MR scan.

In the above case, a combined neurosurgical and ENT approach would be favoured. Most experts would drain the cerebellar abscess first, followed by a radical mastoidectomy.

Reference

Mandalà M, Muzzi E, Trabalzini F. Giant Cerebellar Lesion in a Patient With Purulent Ear Drainage. JAMA Otolaryngol Head Neck Surg 2016. doi:10.1001/jamaoto.2016.0014

Postural BP- You Are Looking At The Wrong Metric

How common is this? An elderly man is admitted to an inpatient facility following a fall at home. During his consultation, you find out that he felt "giddy" before falling. It's happened a few times in the past as well.

Quite reasonably, you instruct the nurse to measure supine and standing blood pressure. You specifically request the nurse to measure the BP after the patient has been lying or standing for a full three minutes.

The nurse informs you that the BP fell by 10/5 mm Hg when the gentleman stood. You are reassured and leave it at that.

But have you missed the diagnosis here?

The normal response to being stood is for the systolic BP to fall slightly and for the diastolic BP to rise slightly. Therefore, the pulse pressure narrows.

More importantly though, the pulse rate speeds up slightly. Apart from the vasoconstriction involved in raising the standing BP, the heart beats slightly faster to maintain the systolic BP at a nearly constant level.

In autonomic failure, be it peripheral, i.e. diabetes, amyloid, or central, as in Parkinsons and Parkinkons plus syndromes and much less commonly in primary autonomic failure, both systolic and diastolic BP fall. More significantly, the pulse rate, instead of rising, remains invariant. It neither rises nor falls. Most patients with autonomic failure also have postprandial hypotension.

Vasovagal syncope is more common than autonomic failure, with two principal types- cardioinhibitory- due to excessive parasympathetic activity, and vasodepressor- due to interruption of sympathetic outflow.

In cardioinhibitory syncope, the pulse rate actually falls on standing. The BP falls too. Some patients may develop asystole.

In vasopdepressor syncope, there is sinus tachycardia. The BP falls.

In postural orthostatic tachycardia sydrome (POTS), the patient's pulse rate increases by >30 beats per minute upon standing. The BP does not fall.

Therefore, the following is a guide to using the blood pressure & pulse rate to differentiate between various types of syncope- standing BP and pulse compared with supine.

Old subject- BP falls by >20/10, pulse rate remains unchanged- autonomic failure

Old or young subject- BP falls, pulse rate falls or becomes asystolic (ECG may show 1st degree block in other cases)- cardioinhibitory syncope

Young subject- BP falls, becomes tachycardic- vasodepressor syncope

Young subject- BP does not fall, pulse rate increases by >30/minute- POTS

Any age- BP does not fall, pulse rate increases by <30/minute- normal response.

Which Lung Nodules Need Follow-up?

A 65 year old man who is a lifelong smoker undergoes low dose CT scan of the chest given recent onset cough and night sweats. The scan picks up a 8 mm nodule in the right upper lobe. Thoracic lymph nodes are not enlarged and there are no other lesions. How does one follow up?

The incidentally discovered solitary lung nodule must be one of the commonest problems faced by clinicians. Further, in the USA in particular, there is insurance coverage for low dose CT scannning of current smokers or ex-smokers aged 50-75 who have quit within the last 15 years, who have at least a 30-pack year smoking history. Studies have shown that such screening reduces lung cancer related mortality by 20%.

But which nodules do you ignore and which do you follow-up?

Studies conducted in smokers show that several factors predict risk of a cancerous lung nodule. Size is probably the most important. Nodules less than 5 mm carry <1% risk of cancer. This risk rises to 50% with nodules greater than 20 mm. A lesion bigger than 30 mm is classified as a "mass" rather than as a nodule. Other factors which increase risk of a nodule being cancerous are older age, female gender, family history of lung cancer, location in the upper lobe, emphysema, a part-solid rather than pure solid or ground glass (also called sub-solid) nodule and spiculation. The likelihood of cancer decreases when the number of nodules crosses four. Other factors also influence the likelihood of cancer being present. While most patterns of calcification are thought to favour a benign nodule, eccentric calcification increases the risk that the nodule is cancerous. Popcorn calcification is characteristic of hamartomas, while central calcification is usually benign. Peri-fissural nodules are almost always benign. Nodules which have been shown to grow on successive CT scans are clearly suspicious. However, the pace of growth matters. Most cancerous nodules have a "tumour-doubling time" between 20 and 400 days. Nodules which double in less than 20 days are likely to be infective. While most nodules taking more than 400 days to double in diameter on CT are likely to be benign, this does not apply to ground glass (subsolid) or part solid nodules. These lesions, particularly the subsolid nodules can only be picked up on CT, but not on chest Xray. Subsolid or part solid nodules can represent in situ adenocarcinoma, minimally invasive adenocarcinoma, lepidic adenocarcinoma, or carcinoid, are particularly common in non-smokers, and take longer to increase in size, with a tumour doubling time of up to 800 days. While solid nodules need only be followed up for 2 years, subsolid or part solid nodules need follow-up for 3 years. There is an on-line resource, available from Brock University, Canada, based on the Pan Can study which was then validated in a British Colombian cohort, which incorporates the above factors into a risk model. http://www.uptodate.com/contents/calculator-solitary-pulmonary-nodule-malignancy-risk-brock-university-cancer-prediction-equation?source=see_link&utdPopup=true Using the calculator, subjects are classified into 3 groups in terms of the likelihood that lung cancer is present <5% 5-65% >65%

With the lowest risk group, it is only necessary to do surveillance CT scans at algorithm determined intervals. The intermediate risk group should have a PET scan to assess the likelihood of cancer (>2.5 SUV being suspicious), while the highest risk group would have a direct non-surgical or surgical biopsy.

Normally, nodules less than 5 mm need no further follow-up unless they are part-solid. Larger solid nodules less than 8mm (called sub-centimetre nodules although the threshold is 8 mm rather than 1 cm), would need follow up up to 2 or 3 years depending on whether they are solid or subsolid. Nodules larger than 8 mm would ordinarily be scanned after 3-6 months, with repeat scans in 9-12 months and 18-24 months. If the nodule has grown more than 50% between scans or by more than 20% in two dimensions, it will need a biopsy.

Any lymph node involvement would mandate a biopsy straightaway.

Just a word of caution. These guidelines were designed for smokers and have not been validated in non-smokers. The predominant cancer in non-smokers is adenocarcinoma, which tend to be peripherally located ground glass nodules. Small cell cancers are solid nodules, also peripherally located, while squamous cell cancers are usually central lesions.

Endocarditis Is Not Always Infective

A 31 year old lady presents to you with a history of recent recurrent "bruising". Although she has no active lesions at the moment and looks quite well, pictures taken on her mobile phone show a substantial dark area on the left ring finger and larger, incomplete dark circular lesions on her upper shins. The GP has checked her ANA. It's negative. Inflammatory parameters are normal. She has two healthy children.

A 71 year old man presents with transient dark lesions on his fingers. On his second visit, you are lucky enough to catch sight of some of these dark areas on the fingers. They are tender. The patient is well and apyrexial without audible murmurs. You diagnose Osler's nodes and ask for blood cultures and echo.

All perfectly reasonable. Except that it may not be enough.

The prospect of embolic lesions from the heart valves is terrifying, but not all such cases are infective endocarditis. There are other possibilities.

Young women with SLE often have Libman Sacks lesions on left sided heart valves, particularly the mitral valve. This is three times more common in those with anti-phospholipid antibodies, which can exist independently of lupus. Although subjects with SLE and Libman Sack's lesions tend to have active lupus, often associated with lupus nephritis, subjects with aPL alone may have been hitherto asymptomatic. While it is important to dip the urine for blood and protein to screen for lupus nephritis, immune glomerulonephritis in infective endocarditis can cause an active urinary sediment.

In fact, Glomerulonephritis is one of the four immune manifestations of infective endocarditis, the other three being Osler's nodes, Roth spots and a false positive Rheumtoid factor (aid memoire GORR).

Treatment would be immunosuppression and anticoagulation, although there are no RCTs that support the use of the former in Libman Sacks endocarditis.

The old man with "Osler nodes" may have cancer, with thrombotic, non infective endocarditis. Clots, present on the heart valves (again more common on the left), are thought to be due to hypercoagulability. Such patients may often have occult DVTs or pulmonary emboli. Gastric or pancreatic cancers may have associated portal or mesenteric vein thrombosis. Apart from treatment of the underlying cancer, these patients need clexane. Warfarin would be ineffective.

Cancer and infective endocarditis can co-exist due to Streptococcus bovis septicaemia. These patients often have underlying colon cancer. Others have cirrhosis.

Subjects with left atrial myxoma can present exactly like subacute bacterial endocarditis with fever, lassitude, murmurs and embolic phenomena. Echo may pick up pedunculated lesions on the mitral valve.

A final non-infective cause of cardiac emboli is eosinophilia. Regardless of the underlying cause of eosinophilia, prolonged or high eosinophil counts go through three phases- cardiac microabscesses, a flabby, poorly contractile myocardium and endomyocardial fibrosis. It is in the second stage that clots build up, often in the left ventricle, which can then embolise and cause strokes, renal or splenic infarcts, just like infective endocarditis.

Eosinophilia can also occur secondarily in subjects with cholesterol emboli, usually from the aorta. These often present in the elderly with renal impairment and livedo in the lower limbs following an invasive procedure such as cardiac catheterisation. There are reports of cholesterol emboli occurring after anticoagulation.

It is important to consider these non-infective causes of endocarditis in subjects with cardiac vegetations on echo, who are culture negative and do not have the expected response to antibiotics.