Test rivals the Fibrosure/Fibroscan ?

[copyman]

this test is very interesting, it is called the FIB-4 test. from what i read it is as good as the fibrosure and almost as good as the fibroscan. get your recent bloodwork ouit and give it a try:

FIB-4 = Age (years) x AST (IU/mL) / Platelets (x1000) x ALT (IU/mL)1/2



Results

• The FIB-4 index correctly identified patients with severe fibrosis or cirrhosis (stage F3-F4), with areas under the receiver operating characteristic curve (AUROCs) of 0.85 and 0.91, respectively.

• A FIB-4 index score below 1.45 had a negative predictive value of 94.7% and a sensitivity of 74.3% for excluding severe fibrosis.

• A FIB-4 score higher than 3.25 had a positive predictive value of 82.1% and a specificity of 98.2% for confirming the existence of severe fibrosis.

• Using these cut-offs, 72.8% of the 847 liver biopsies were correctly classified.

• The FIB-4 index was strongly correlated with FibroTest results for scores below 1.45 or above 3.25 (P < 0.01).

• A FIB-4 value below 1.45 or above 3.25 (which accounted for 64.6% of all cases) was concordant with FibroTest results in 92.1% and 76% of cases, respectively.

Conclusion

In conclusion, the authors wrote, "For values outside [the] 1.45-3.25 [range], the FIB-4 index is a simple, accurate, and inexpensive method for assessing liver fibrosis and proved to be concordant with FibroTest results."

This study confirms previous data showing that while non-invasive methods perform well at distinguishing between absent or mild fibrosis and extensive fibrosis or cirrhosis, they do less well at distinguishing between intermediate stages.

[jboyhk]

something not right with formula. Take for example a 40yr old man with normal test scores

40X25X275X25X.5= 46.6

[Upbeat]

Thanks for the info.  Now if I can make heads or tales of it.  I don't know what my numbers are, all I was ever told was slightly elevated.  The formula seems kinda strange.  How do you get between 1.45 and 3.25 by multiplying age in my case 60 x ast lets say 60,  lets say 300 platelets and the alt lets say also 60.  These are numbers taken out of the air but what would the fib-4 show?

                                                                                                         Ron

[copyman]

upbeat, your score would be good, 0.60  here is the formula in your case:

60 x 60 = 3600 divided by 300000 ( 300 platelets x 1000) =0.01 x 60 =0.60

jboy, you forgot to divide this symbol / means divide.

[jboyhk]

That X was a typo, I actually did divide it in the calcs.

Do you mean age*AST/(plat*ALT*.5) ?

The difference is what is in the divisor

[Upbeat]

Thanks.  Kinda simple when you see it like that.  Jboys question on the 1/2 or .5 at the end of the formula is confusing.

                                                                                                          Ron

[copyman]

LOL, i'm not that good at math. i think if you are below 1.46 that would correlate with like F0-F1. do a google on "FIB-4 test" and you will see a lot of info in it.

[Janann]

Thanks for this! I'm very math-challenged, but your example in post C3 helped a lot - I'm with you till the last step. Where does the 1/2 come in? We are using the whole ALT value, right? or do we use half of it?.......or do we divide at the end by 1/2?

LOL this happens every time I do math. So far my results are looking good, but I haven't done anything with the 1/2.....:(  I sure do like further research into non-invasive testing!!!!!!

Janann  

[Valtod]

Actually, the formula is:
FIB-4=(age*AST)/(PLT*ALT^1/2), or
FIB-4=(age*AST)/(PLT*sqrt(ALT))

That is ALT to a power of 1/2, or square root of ALT.

For example, age=45, AST=40, ALT=64, PLT=300, so sqrt(ALT)=8.
Then FIB-4=(45*40)/(300*8)=1800/2400=0.75

The Serum Fibrosis Test FIB-4 results range from 0.2 (mild fibrosis) to 10 (cirrhosis).

Important cutoff values are 1.45 and 3.25.
FIB-4 < 1.45 has a NPV (negative predictive value) of 94% to exclude severe fibrosis with a sensitivity of 74%.
FIB-4  > 3.25 has a PPV (positive predictive value) of 82% to confirm the existence of a significant fibrosis (F3-F4) with a specificity of 98%.

For more detailed correspondence between fibrosis scores (Metavir F stages) and FIB-4 index, see the box plots of Figure 1 at http://www.retroconference.org/2007/PDFs/913.pdf

See also:
http://www.medscape.com/viewarticle/501435
ttp://www.retroconference.org/2007/PDFs/912.pdf
http://www.cdlpharma.com/laboratoire_central/pdf/aasld_2006_cacoub_et_al.pdf
http://www.hepatitis-central.com/mt/archives/2007/06/non-invasive_an.html
http://www.aegis.org/conferences/croi/2006/867b.pdf

[Valtod]

Ron, your FIB-4 index is NOT 0.60 but 1.55.

Here's why: age=60, AST=60, ALT=60, PLT=300, so sqrt(ALT)=7.746
Then FIB-4=(60*60)/(300*7.746)=3600/2324=1.55

I hope this helps.

[Lilla]

what does it mean if you are >1.45, but <3.25?

my blood work for Fibrospect II test was
  age 55xAST105=5775 /  platelets 229 x ALT sqrt 168 (12.961)=2968.07
            5775/2968.07 =  1.9457

Last blood work showed
  age 55xAST34=1870 / platelets 115 x ALT sqrt 23 (4.796)=551.54
             1870/551.54  =  3.39

with that formula, my score looked better pre-treating
Am I figuring something not correct?

[Upbeat]

Its my understanding that during treatment your platelets can drop alot.  It seems to be the deciding factor on the formula.

                                                                                                       Ron

[copyman]

LOL, like i said my math is terrible. Val can you please tell me my score? my stats follow:

age-50, AST-26, Platelets-200, ALT-50

Thanks

[Valtod]

Lilla, your calculations are correct. Your pre-treatment FIB-4 index was 1.95. However, the number during treatment is meaningless because your platelets and enzymes (AST/ALT) are in flux. All the biochemical scores should be calculated using average and consistent bloodwork lab results. If you're on ANY medication or exposed to toxicity that affects your platelets and enzymes, the formulas won't produce reliable results. For example, even if you're not on IFN/RBV but only taking antibiotics, your enzymes may increase.

So you have to wait several months (3 or even 6) after the end of your current treatment, so your platelets could come back to (or get even higher than) the pre-treatment value. Also, the AST/ALT should stabilize on levels closer to normal range (if the treatment is successful).

You also asked "what does it mean if you are >1.45, but <3.25?" Unfortunately, not much. Your FIB-4 index of 1.95 suggests that you were probably F2. However, with just a SLIGHTLY lower probability you could be anywhere between F0 and F4 as well.

If you look at the FIB-4 box plot of Figure 1 at http://www.retroconference.org/2007/PDFs/913.pdf , you'll see that the correspondence between fibrosis scores (Metavir F stages) and FIB-4 index is quite fuzzy and overlapping in the middle range.

Half of the patients with F0 are clustered in the 1.0-1.45 range of FIB-4.
Half of the patients with F1 are clustered in the 0.8-1.25 range of FIB-4.
Half of the patients with F2 are clustered in the 1.2-1.8 range of FIB-4.
Half of the patients with F3 are clustered in the 1.1-1.3 range of FIB-4.
Half of the patients with F4 are clustered in the 1.3-3.1 range of FIB-4.

That's why 1.45-3.25 is really the "fuzzy" zone where the FIB-4 biochemical scores do NOT have much predictive value.

As you can see, FIB-4 is really useful for scores > 3.25, where we can EXPECT with very high probability cirrhosis or severe fibrosis (F4).

At the lower end, where the FIB-4 scores < 1.45, we can EXCLUDE with very high probability cirrhosis or severe fibrosis.

And that's pretty much all you can learn from biochemical scoring such as FIB-4.

Hope this helps.

[Valtod]

Assuming the lab results are NOT during treatment:
Age=50, AST=26, ALT=50, PLT=200, so sqrt(ALT)=7.07
Then FIB-4=(50*26)/(200*7.07)=1300/1414=0.92
Which corresponds to F1 (see all the caveats in my previous post).

All the best,
Val

[copyman]

thanks for the calculation. seems this is similar results that you get with the fibrosure & fibrospect tests. fairly accurate on the low F0-1 & high end F4 but much less for the middle F2 & F3. only difference is the fibro tests cost around $300+

[Janann]

Thank you for the enlightenment! This clarifies it even for a math dunce such as myself. :) Your info on how to interpret the result was also really helpful. Much obliged...

Janann :)