A different 'diff'

Got my CBC w/diff results back today. I've been tracking the #s for 3 1/2 yrs now and though I've noticed how different labs can list the results differently, for the first time I'm finding differences in the differential from my last few tests, each taken at the same lab.

Comparing the last two - December's shows "Differential, Auto" while today it says"Differential, Manual". The Auto breaks Lymphocytes into both Lymph Abso (ALC) and Lymph % (which I ignore), while the Manual has only the Lymph %. Therefore I was forced to do math today to come up with the ALC #. That's when I realized that the Lymphocyte percentage, which had lately been in the mid-80's percentile was instead an amazingly low 69%. This % in conjunction with the number of WBC's being lower than it was in Dec. translated into the lowest ALC # I'd seen in 2 years. I was thrilled.

HOWEVER... a few hours later I noticed another Lymph category I hadn't seen during my first perusal - "Variant Lymphocytes". This was said to be 8%.

As well, as I began hunting for other diff differences, I noticed how this report displayed "Segmented Neutrophils", "Bands", and "Neut Abs Man" (the first 2 being percentages) while previously neutrophil categories had only shown "Neut %" & "Neut Abs".

Does anyone know why one test would be Auto and one would be Manual and what that implies? And what is the significance of Segmented Neutrophils vis a vis Bands and why would the manual break them down like this as opposed to one number? Most importantly how do I compare ALC from one test to the other? Do I take 8% of the WBC and add them to the 69% of the WBC for a total ALC? What are Variant Lymphocytes anyway? Are these smudge cells? [At this point I'm guessing that 8% is an estimation of the cells that were smushed and couldn't be counted and therefore are all CLL cells while the 69% of WBC are a combination of CLL cells and healthy B, T, and Natural Killer lymphocytes, though with the vast majority being CLL cells].

Thanks for the info from anyone in the know who can help me calculate my ALC given my WBC was 31.4.


7 Replies

  • Gene,

    Auto is automatic testing; your blood sample is automatically analysed by a machine that categorises your blood cells by size. That may not be an accurate method if you have a blood condition, so a manual check is sometimes done - a medical technician with the appropriate training actually counting the different cell types by looking at a blood smear sample under a microscope. Obviously there will be differences caused by the sample actually looked at and how representative that is of your blood in general, plus some inevitable human error.

    Neutrophils change in appearance as they mature, coming out of the marrow as bands (or stab cells) and later appearing segmented. I assume that the size difference isn't enough for auto counting to tell them apart, but that can obviously be done visually with a manual check.

    As always, you need to compare similar results, so you'll get a better picture of how stable your CLL is by comparing auto test results with auto test results and manual with manual (ideally from the same medical technician!) and looking for trends in the results.

    Given this paper: ncbi.nlm.nih.gov/pmc/articl...

    I wouldn't be too concerned about the variant lymphocytes - leave that to your haematologist expert to worry about. They may be familiar with the accuracy provided by the technician doing the manual test if that information is considered important.

    Hope that helps,


  • Thanks as always for contributing to my slow but steady education.

    Though I understand that sometimes manual tests are given and sometimes automated tests are given, I'm a bit confused as to when and why they are. Had 16 tests in 3 1/2 yrs and 5 have been manual (3 of first 6 and 2 of last 3). Perhaps the last two were given because my ALC had jumped the test previous to it and they were looking for confirmation, but I was not privy to why let alone that it would be manual those times around. [Susan LeClair as per Len's reply below seemed to imply that automated is always given while manual is used as a check, but this doesn't appear to be what happens in my case]

    Got the neutrophil story. Seems they are basically all Segs when manually counted anyway so the type of test doesn't really make a big diff.

    After reading the ncbi article I'm also still a little confused which might be understandable as the study basically pointed out that there is confusion among lab techs as to agreement as to what cells were on the same sample … even to the point that the same techs didn't agree with themselves when given the same sample more than once. Adding to the confusion is that terms like atypical, abnormal are at times used in different contexts -sometimes synonymous with variant and sometimes referring to other medical issues [from the article: "A confused terminology is used for abnormal lymphocytes, for example, variant lymphocytes, atypical lymphocytes and even combinations of several cells are used"]

    In any case even adding my Variant Lymphocytes (2.5) and Absolute Lymphocytes (21.7) the total is a lot less than the ALC from Dec (32.2 )… so I"m not confused at all that if I had to choose between disappointed and pleased it would be the latter.

    Thanks again Neil.

  • Gene, I'm glad to hear that you are now pleased with your results. Gaining a better understanding was worth the effort for that outcome! It will be interesting to hear why your doctor asks for the manual tests relatively frequently, as they would have to be more expensive. Len did well referencing Susan Leclair's video. I've learnt a great deal (and definitely more from her than from anyone else) from her excellent explanations on what is a very complex subject over the years.


  • Yes, I am definitely asking him about that. The transcript of the video Len provided makes it seem the automated is always given with the manual given in certain situations for quality control.

    Unfortunately it became clear to me very early on that my doc's knowledge and decision-making concerning CLL

    prognostic /diagnostic testing and treatment was not in synch with what CLL experts nor even what IWCLL standards were saying. Thanks to following CLLSA I've done a lot of 'questioning' of his decisions and at times have been able to guide him toward offering me a higher standard of care.

    Take care! Gene

  • There are a series of videos by Susan LeClair that includes a explanation of the difference between Auto Diff and Manual Diff - if you have the time watch the entire list of her videos on Patient Power.




    Here is the text of the one video that actually answers your question:

    Dr. Leclair:

    Well, hello again. This is Professor Susan Leclair from the University of Massachusetts Dartmouth, and today in this

    continuing series I'm going to talk about white cell counts and differentials. And I know several of you have written in

    about how can you tell, how do you interpret what's going on about these.

    I think we mentioned that the white cell counts, I think it was the very first of these sessions, will vary during the day, and it

    will vary during—blah-blah—and it will vary due to your emotional and physical state.

    Again if you've been driving in a rainstorm and the car on the other side of the road hits the puddle before you do and that

    puddle splashes all over your windshield—well, first you duck, but secondly the reason that you duck and the reason that

    your heart rate is going and the reason that you're having trouble breathing is that adrenalin surge that comes from fear

    and an unexpected reaction.

    So, yeah, you're not sure this is the phlebotomist that you want, you had a fight with your spouse before you got to the lab,

    the traffic was a bear, all of that stuff is going to influence the white cell count just like having done maybe 20 minutes of

    yoga and coming calm and serene is going to influence it in the other direction. So white counts will bounce around a lot.

    I also should say that laboratory counting methods are good, but they're not perfect. If I get a sample of blood from you

    and I run 10 white cell counts on that very same tube of blood, I am not going to get 10 identical answers. I'm going to get

    them close, 5.4, 5, 5.7, 6, 5.2, 5.1. They'll all be around and close to the number that is the actual number, but they will not

    be produced in a repeated identical order. There will be a range of maybe plus or minus .5 of the real one.

    So don't worry if your physician, as they frequently will do, will look at a number, well, last time it was 5.5 and this time it's

    6.0, and that's gone up you think, and he says, aah, don't worry about it. That's part of the reason he's saying that. He's

    saying that because it's not a significant rise in your white blood cell count.

    And if you did have one of those days when you were late and you were frustrated and the world was against you—I have

    those frequently—and you get up there to have your blood drawn and things don't go well, yeah, maybe it can go from a 5.0

    to maybe a 7.5 or an 8, because you've done a lot to those cells to cause them to become agitated. When you're upset,

    they're upset.

    So just kind of keep in mind that you're looking for kind of an overall trend, and big numbers are significant, less than

    maybe .5 means no change whatsoever. Less than 4.0 means a little problem, but I bet you it was something psychological

    or relatively minor to deal with. So that takes care of the total white cell count.

    Then there's two ways to figure out what cells you've got. The older, the traditional, the one I love takes a drop of blood

    smears it out onto a slide, that's why it's called a blood smear. It stains so that we can see these cells, and somebody with

    some skill and some knowledge sits down at a microscope and evaluates those cells.

    There are thousands of cells on that slide. We cannot evaluate all of them. So we randomly pick 100 cells because, well,

    that works out well for percentage. If you've got a really high white count, we might do 200 cells in order to increase the

    yield of accuracy.

    But we're concerned about quality. What do these cells look like? Do they have holes in them because something is eating

    them? Is there something wrong with your nucleus that makes them look funny? Are they bigger than they should be? Are

    they smaller than they should be? Are they staining correctly because that tells you something about the metabolism of

    the cell. So this is a real qualitative kind of assessment of these cells.

    Do you get a number? Sure. You get 60 percent polys 40 percent lymphs. Or you get 59 percent polys and 39 percent

    lymphs and 2 monos. That's the kind of differential you get. And when you look at your report form, it will have a

    percentage sign next to it to let you know.

    The problem with this percentage differential is that if you have, say, 75 or 80 percent polys—we're only counting a

    hundred cells—by definition the lymphocytes must go down, because you only have 25 left cells to count. So an increase in

    one cell line will—must—absolutely, positively must cause a decrease in the other one. So sometimes you're going to look

    at number and go, oh, my god, my lymphocytes are down. No, it's because your granulocytes are up. Or my granulocytes

    are down. No they're normal. It's because your lymphocytes are up.

    So one of the problems with the percentage differential is you can be misled sometimes by which cell is causing the

    problem. Why do I love the percentage differential? Because I get to sit at a microscope and look at the cells. I tend to root

    for them, they tend to speak to me—I know, it's a sad life, but they do, and so I like to look at them to see what's happening

    to the granules, to see what's going on inside them, because that's telling you information that you're not going to get any

    other way.

    The absolute differential count we never used to do until we got the multi-channel instruments. What the multi-channel

    instrument will do is it will count the number of cells, okay, so you get a total number of white cell count, but they will also

    evaluate the size and the shape of the cell, kind of like the mug shots you get from the police station. You get a front-on

    view, like this, so you see how fat my face is or what color my face is or something that's to do with that. And then you get

    the profile that lets you know is anything lumping out of it.

    A lymph has no granules so a lymph is going to be very flat, so you're going to see this when you look sideways, and you're

    going to see this when it looks frontways. A granulocyte's got, well, granules. So when you turn it from this to this it's got

    lots of lumps and bumps that are coming out, because it's got all those granules in them. So we can differentiate them by

    this machine.

    A machine does this very fast. So they don't count a hundred cells. These machines will count anywhere between 20,000

    and 50,000 cells. Okay. Now you get a really good sense of what cell is increased and what cell is decreased, because the

    machine is going to tell you in 50,000 cells I counted 25,461 granulocytes and whatever the math would be to finish that in

    lymphocytes, so you get an absolute cell count. And that's wonderful because that eliminates the percentage differential's

    problem with misleading information.

    What's the problem with the absolute differential? You don't look at the cells. The machine cannot look at these granules

    inside the granulocytes because remember, all they're looking for is lumps and bumps, and say, I think there are too many. I

    don't think there are enough. I think the granules that you've got are too small, or they're too big.

    In the case of the neutrophil, you have four different kinds of granules. That machine can't tell the difference between

    those granules, but sitting down at a microscope I surely can.

    So when do you want to do a manual versus an automated differential? Well, the automated is going to come all the time

    because that's what a CBC includes now. You should do a manual differential if the white count has changed a lot, you're

    not feeling well for some reason, the numbers that come out of the automated differential look suspiciously different than

    before. In a laboratory we call it a delta. What's the change between the numbers you had before and the numbers you

    have now?

    So we have on January 1 you had 60 polys, 40 lymphs and on February 1 you had 75 polys and 20 lymphs, I'd probably want

    wouldn't bother. It's not worth it. So it has to have a big enough delta, a big enough change for you to want to be able to

    see what's going on in these cells, and how can they handle it.

  • Thanks Len. I'm signed up to Andrew Schorr's Patient Power site so I now get Dr. Leclair's "lab test" videos. But I hadn't seen this one on the reasons for variability of results of white blood cell.

    From that transcript you provided I learned that the small sample size of the manual test might have something to do with my lower than expected ALC. I noticed my monocytes which had never been above 2% was 6% this time and she explained that this was the downside of the manual test given small sample size of cells. Perhaps the next slide would have had monocytes back at 2% and ALC% up the 4% lost by the mono's

    I also was able to visualize how variant lymphs are visually picked out by the technician on a manual test and assume that they are also there but just lumped in as part of the Abs Lymphs when the test is automated. Am I right to think it makes sense to view variant lymphs as another type of pretty useless cells just taking up space like CLL cells both unable to do their intended job of fighting off foreign invaders? If so I could add the 8% variants to my 69% Lymphs and that plus the extra 4% due to seemingly aberrant # of monocytes would have me back over 80% which is where I've basically been for Lymph% for the past 2 years .

    Thanks again for your time, and willingness to share your experience and expertise.

    PS I contacted Dr. Furman (your doctor and the doctor I most believe in) concerning clinical trials and to ask him about possibly becoming a patient down the road. He was open, amenable, and helpful. However this last test makes me feel that both treatment and change of doctors is even further down the road now.


  • Gemit

    I have had the same blood tests done at two different hospitals within a one hours drive... the results were freakily different...

    Blood test are a very small tip of the CLL iceberg... don't worry about the minutia, track your ALC and platelets over time...

    Dr. Susan Leclair will be a speaker at the Niagara Falls conference.. she is a wonderfully funny speaker... Andrew will have new videos from the conference in the future... stay tuned to Patient Power, a wonderful resource.

    If you can... try to attend... meeting 300 CLL patients and caregivers in one room is a truely wonderful experience...


You may also like...