This is a new guideline from the BCSH. It is quite a chunky guideline and it covers a lot of detail. I will give a few 'top learning points' then a more detailed summary beneath. At the end I have attached some algorithms, which are really helpful in knowing how to investigate / manage people.
Top learning points
How good is the test?
There is a lot of inter-lab variation. Using a cut-off of 148 pmol/L (200 ng/L) it has a sensitivity of 97% - so pretty good in that sense. However, some assays are affected by high anti-intrinsic factor antibody (IFAB) and can give a false normal result. If there is folate deficiency, you can also sometimes get a false normal result.
What should you do on looking at any B12 result?
What is the clinical scenario? If there was a high degree of suspicion (eg megaloblastic anaemia, neurological symptoms, glossitis), then you may not trust a 'normal' result and you may want to look further.
What tests should you do looking for the cause of the B12 deficiency?
IFAB for pernicious anaemia - but only in those at high risk (see summary below for more details)
What else can effect B12 levels?
See the summary below for more details, but the following may impact on the results:
- Metformin (may lower levels - but the clinical significance is uncertain)
- COC / HRT (may lower levels - but the clinical significance is uncertain)
- Pregnancy (levels drop physiologically in the 3rd trimester, but a true deficiency is harmful to the baby, so you would need to treat it)
- Vegan (at much higher risk of deficiency due to inadequate intake)
- Malabsorption (higher risk - eg gastric surgery including banding)
- PPI (can reduce absorption)
- Age (can reduce absorption)
What treatment should I use?
Follow the BNF advice. If there is neurological involvement, patients need injections every other day until symptoms resolve (rv after 3/52).
If there were no initial neurological symptoms, give injections every 3m.
If there were neurological symptoms, give injections every 2m.
Should we give them more frequently if patients ask? They sit on the fence and advise that they can't make a recommendation. There is no evidence of harm, but no studies looking at if there are benefits.
Can we use oral treatment? See the algorithms below. Low dose oral B12 may be suitable for some patients who are asymptomatic with borderline levels.
What about folate deficiency?
There is nothing new here for us.
How do you diagnose vitamin B deficiency?
You may have thought that the answer to this one was pretty obvious - you just check the levels... but it isn't that simple.
Firstly the test isn't perfect. It has a sensitivity of 97%, so 3 out of 100 cases will be missed. If one of those cases is a patient with subacute degeneration of the cord, that could be disastrous.
Secondly there are false normal readings, especially with high IFAB levels (eg in pernicious anaemia), but also sometimes with folate deficiency.
Thirdly there is a large grey area of results, where management may not be certain (eg 110 to 148 pmol/L). In some cases this may represent true deficiency, but in others it may not.
There are other tests that may help in the future, but they're not readily available yet, so not really relevant at present.
When should you consider testing for vitamin B12 levels?
- Megaloblastic anaemia
- Neurological symptoms (see patient.co.uk for a description)
- Screening eg tiredness / dementia / raised MCV without anaemia etc.
When you get a Vitamin B12 result back, what should you consider?
Why was this test done? What was the pre-test scenario? Am I surprised the result is normal / abnormal? What other medical conditions does the patient have (eg malabsorptive states / medications / pregnancy)? What is their diet like?
If the pre-test probability was high - eg megaloblastic anaemia, neurological symptoms suggestive of deficiency or glossitis, then have a low threshold for investigating further. The algorithms I've attached at the end show how to do this.
Beware that 25% of people with neurological symptoms and Vitamin B12 deficiency, will have a normal MCV, so don't be reassured by that.
Should we test everyone with Vitamin B12 deficiency for Pernicious Anaemia?
Basically no. In pernicious anaemia, you produce antibodies to Intrinsic Factor, so you can't take B12 up out of the food. However, it is actually quite uncommon. IFAB testing has a high Positive Predictive value (ie 95% of positive cases on testing will actually have pernicious anaemia), but a low sensitivity (only 40 - 60% of cases will be picked up on testing with IFAB).
So who should you test?
- Highly suspicious presentation for B12 deficiency (megaloblastic anaemia, glossitis and neurological symptoms), even with a normal B12 level (remember that high IFAB levels can give a false normal B12 result).
- Other autoimmune disorders, or a family history of autoimmune disorders.
- No other obvious cause (eg no malabsorption / medications)
How does malabsorption affect B12 levels?
There are various different conditions that can affect B12 absorption in different ways:
- Surgery - eg gastrectomy / gastric banding / removal of small bowel (eg Crohn's)
- Medical conditions - eg Crohn's
- 'Food Bound malabsorption'. For B12 to bind to Intrinsic Factor and then to be absorbed, it has to be separated from the food by the stomach acid. Various things can stop this separation. Age can (it reduces stomach acid) and PPI use can too. These are probably big causes of sub-clinical deficiency.
How can various drugs affect B12 levels?
This is quite complicated as there is a lot of uncertainty about the mechanisms involved and also about when a low B12 is actually relevant in the context of medications (eg the combined pill can lower B12 levels, but when you do other tests to check deficiency, they are OK, so the low B12 may not be that important). The algorithm below suggests various ways of managing this.
What about pregnancy?
B12 levels naturally drop in the 3rd trimester, so it is difficult to interpret a low level. The BCSH suggest covering the pregnancy with 3 B12 injections as a low B12 is dangerous to the baby, then investigating mum post-natally.
NB Algorithm 2 is attached at the end of the summary.
Can we use oral treatment?
High dose oral B12 (1000 to 2000 mcg) is available outside of the UK and on the internet, but isn't licensed here. At this high dose, even if there are IFAB, enough is absorbed to meet daily requirements. A recent Cochrane review suggested that it is as effective as im B12. Given it isn't currently available here and that it's cost-effectiveness and efficacy in a wider population isn't certain yet, the BCSH doesn't currently suggest that it should be used. However, they do say that you can consider it in maintenance or correction of sub-optimal levels in asymptomatic patients.
Low dose oral B12 can be used in treating asymptomatic patients with subclinical deficiency. See the algorithms below.
Algorithms for investigation and management
The following 2 algorithms from the guideline are really helpful and guide what investigations to do and how to manage people. Algorithm 1 is for patients where there is strong clinical suspicion of deficiency (eg megaloblastic anaemia / glossitis / neurolgical signs). Algorithm 2 is where there are no objective features of deficiency (eg tiredness or dementia screening). MMA, tHey and HoloTC are the other blood tests that you can do looking for deficiency, but they aren't readily available yet, so I haven't talked about them.
Latest posts by Louise Hudman (see all)
- Nice | Who should we be testing for cirrhosis? - October 3, 2016
- Nice | Non-alcoholic fatty liver disease NAFLD - August 26, 2016
- Evolocumab and Alirocumab for treating dyslipidaemia - August 25, 2016