Total IgA, IgG, and IgM before and after Evusheld and COVID infection

[SeymourB]

I've seen people speculate here and elsewhere how Evusheld or COVID infection might affect total immunoglobulin values.

I just got my latest blood tests back prior to my upcoming hemo/onco visit, and thought I'd share my N=1 results. I've had more testing than usual because I have a new hemo/onco, and my trend has increased. But I have very strange CBC's and differentials - with unusual monocytes, eosinophils, and basophils that might be due to reactive lymphocytes being mistaken by both automated analyzers and pathologists. That's for a future post after I see the hemo/onco and an infectious disease specialist next week.

All tests were done in the U.S. through Quest Diagnostics, a huge laboratory testing company with massive labs in a few U.S. cities. I like Quest because they're huge, and have a big quality control infrastructure. They use expensive, high, throughut blood analyzers, and provide nice graphs in their portal. But I think the manual review of odd results is lacking. In the results below, there's nothing odd - unlike my CBC's and differentials.

---

Below are my test results - type of IgG, the measurement, the range of the instrument, and a flag showing whether it was Low or not.

January 28, 2022:

Immunoglobulin A57mg/dL70320L

Immunoglobulin G451mg/dL6001540L

Immunoglobulin M19mg/dL50300L

February 28, 2022 Evusheld first half dose

May 9, 2022 Evusheld second half dose

May 12, 2022:

Immunoglobulin A62mg/dL70320L

Immunoglobulin G466mg/dL6001540L

Immunoglobulin M18mg/dL50300L

May 16, 2022 COVID first symptoms

June 6, 2022 last positive later flow COVID antigen test and diagnosis of bacterial sinus infection

June 30, 2022:

Immunoglobulin A74mg/dL70320

Immunoglobulin G469mg/dL6001540L

Immunoglobulin M21mg/dL50300L

---

For those more familiar with Ig in grams per liter, you can slide a decimal points to the left a few places.

Evusheld has an average half life of 90 days based on early studies, and is a modified IgG. So my first half dose had decline some between March and April. Then I got another half dose, which is why you shouldn't apply this arithmetic below to your own results. But I think they show how big the stadium is.

Above, you can see that IgG rose from 451 mg/dL to 466 mg/dL after Evusheld - 15mg/dL. I did not expect that much change, to be honest. My IgG does vary a bit, but almost always trends downward. The biggest upward change was 5 years ago, when it increased by 11 mg/dL over a 6 month period. So it could be a combination of a rise due to some undiagnosed infection plus Evusheld.

49 days from the previous total Ig test, 45 days after first COVID symptoms, and 24 days after last positive lateral flow COVID antigen test, IgG increased from 466 mg/dL to 469 mg/dL - 3mg/d. Human IgG does have an average half life of about 28 days, depending on which type of IgG. There are 4 known types of IgG, which the lab test lumps together in a single number, and the average half life varies between them. Inherited genetics also may affect half life. Mixed with all that easy arithmetic is unfathomable complexity of B-cell response and development. Plasma B-cells continue to secrete Ig for month, slowly descending, though, so unlike MABs, I think all the half life stuff doesn't matter so much within weeks infections.

SARS-CoV-2 doesn't stimulate much IgG, though. The SARS-CoV-2 attacks cells the lining of respiratory tract, where IgG just cannot reach easily. Those cells are protected by IgA, though. So we can see IgA increased from 62 mg/dL to 74 mg/dL - 12mg/dL over that same period. IgA doesn't make it into blood that much, though. The vast majority of it is right where we want it in mucosal tissues.

Viral infections also stimulate IgM, which increased about like the IgG. IgG went from 18 mg/dL to 21 mg/dL. IgM also doesn't make it into the blood much, either, and can be found in lymph nodes, spleen and tissues as a first responder to help display antigen to T-cells and neutrophils.

Complicating the analysis is a secondary Staphylococcus Lugdunensis bacterial infection that I imagine was always there (it's a commensal bacteria), but took off with the COVID sinus infection. I kick myself for not asking for a sinus culture sooner, because I've had several long bacterial sinus infections before. It was not diagnosed until June 6. It's not as bad as it was, but just keep hanging around. Like IgG and IgM, antibiotics don't make it through to mucosa easily

But as I look back at Ig test results before and after previous long sinus infections, and my IgG actually declined by 63 mg/dL, my IgA declined by 9mg/dL, and IgM increased by 1gm/dL, but the tests did not bracket the infections that well. I ended up on SCIG for a couple of years as a result of those, but went off it in 2020, because of the pandemic, masking, distancing, and low plasma donations in the U.S. I have not gotten back on. We'll see how that works out.

I hope this satisfies some curiousity and concern for some of you.

=seymour=

[gardening-girl]

Interesting to see real-life clinical numbers Seymour. As always, I couldn’t help compare your results with published results to see if the increase that you noticed in your IgG level was within the realm of what has been published. In the plots above if you consider that you are at the high end of reported Evusheld serum concentrations, (considering the error bars), with a second dose of Evusheld (600 mg total) you might expect to see an increase in serum IgG of somewhere in the range of 10 mg/dL based on data in Figure 5A. The 15 mg/dL that you report is certainly likely to be at least predominately attributable to Evusheld. (15 mg/dL = 15,000 ug/dL = 150 ug/ml -not too far off from extrapolating the published values)

Loo et al. science.org/doi/10.1126/sci...

SCIENCE TRANSLATIONAL MEDICINE 25 Jan 2022 Vol 14, Issue 635

ema.europa.eu/en/documents/... European Medicines Agency EMA/205600/2022

nejm.org/doi/full/10.1056/N...

N Engl J Med June 9, 2022

Plot of Evusheld serum concentration

[gardening-girl in reply to gardening-girl]

A second plot.

Plot of Evusheld serum concentration

[gardening-girl in reply to gardening-girl]

A third plot of Evusheld serum concentration.

Plot of Evusheld serum concentration

[bennevisplace in reply to gardening-girl]

Thanks. Any idea why this third plot seems to show a much longer half life than the other two? AZ stated that they engineered AZD7442 for extended protection of at least 6 months, which is not compatible with a half life of 90 days as mentioned by Seymour.

[gardening-girl in reply to bennevisplace]

Actually the third plot doesn't show a longer half-life. Half-life is calculated from the time of the [Peak] serum concentration of Evusheld to the time when the concentration reaches 1/2[Peak]. So it does still come out to be ~90 days in this study. Clinical trial data has shown that at 6 months (and maybe even longer) there is still enough Evusheld to provide protection.

In this third graph, after the first antibody half-life there would be ~12 ug/ml of Evusheld antibody, 90 days later, 6 months post-peak level there would be ~6 ug/ml Evusheld antibodies, 3 months later (9 months post peak) there would be ~3 ug/ml. I’ve seen it reported that Evusheld can still be effective at a serum concentration of ~2 ug/ml. I’ll have to try to track down that reference.

Serum concentration of Evusheld showing half-life determination.

[bennevisplace in reply to gardening-girl]

Fair enough. I'd better arrange another eye test!

[SeymourB in reply to gardening-girl]

bennevisplace -

BTW, that graph is from the Supplementary Appendix, which has a link in the NEJM paper itself.

Although the NEJM article mentions the 600mg revised dose and Omicron variant, the Supplement says that the neutralization assay data was from the Science Translational Medicine article, which was from January 22, which says:

"Although further testing is in progress for the omicron VOC that emerged in November 2021, data published elsewhere show that AZD7442 retains neutralizing activity against the omicron variant, with reported IC50 values that range from 51 to 277 ng/ml (36–39)"

To be safe, I would pick the 277ng/mL. The references cited (36-39) are all from December. Two of them have yet to be published.

Now, all I need is a graph that shows neutralization for BA.2, BA.4, and BA.5. In the meantime, I found multiple preprints regarding Evusheld and those subvariants listed on:

opendata.ncats.nih.gov/vari...

Much to digest. Always there's a caveat that the cell line and pseudovirus used for neutralization assays can create diverging pictures. I've seen at least one effort to define a standard assay, but scientists like their suppliers. We need leadership on this from WHO, I think.

=seymour=

[bennevisplace in reply to SeymourB]

Thanks again Seymour. The opendata looks like a useful resource, except my phone is playing up and I'm on breakfast duty, so yet another postponed study hour. To think, at school my attitude to those was quite the opposite.

[SeymourB in reply to gardening-girl]

gardening-girl -

My results were a mere 3 days after Evusheld.

I think the upper, colorful graph shows much less than 150ug/mL for the 300mg IM dose. It looks to me like about 37ug/mL (halfway between 25 and 50) at 14 days - the 14 day label is really poorly done there. That graph is from Fig. 5. of the Science Translational Medicine study.

That's close to the black and white graph from Figure 1 of the European Medicines Agency Assessment report, which I judge to be 23ug/mL for the production cell line at 29 days. It's a shame they only sampled at 8 and 29 days. The other 2 production data points look to me like about 55 and 90 days, but I couldn't easily find them in the text. I wish that all graphs would have accompanying tables.

I don't know the repeatability error of the Quest Ig analyzer, but it's likely to be as much as +/-5%. That might be +/-22mg/dL, which could alone account for the difference. I'll see what I can dig up without paywall access to instrument journals or from FDA. That could take me awhile, but might be very instructive to my fellow patients.

=seymour=

[gardening-girl in reply to SeymourB]

Seymour, I definitely should not have doubled your Evusheld dose when making my calculations since the second dose would not be contributing much yet to serum levels! When I looked at your data again, considering what the serum concentration might be 73 days after your first dose, subtracting the 14 days to reach peak concentration with a blood draw 59 days later (~2 months) which is 2/3 of the way to the reported half-life, the number looks more like this: 2.1 mg/dL possible Evusheld contribution.

I based my calculations on this sentence which was stated in the Loo et al.

paper:

“AZD8895 (16.5 μg/ml) and AZD1061 (15.3 μg/ml) were similar and reached a median Tmax after 14 days.”

Therefore ,a serum concentration at day 14, reported to be 31.8 ug/ml could contribute 3.18 mg/dL to the IgG pool. Two months later, 2/3 of the way to its half-life the concentration of Evusheld would be 2.1 mg/dL, so my estimate of a possible contribution of 10 mg/dL to your IgG levels was pretty far off!

E-Lynn has probably made a good point in her post below in questioning our rationality in trying to extract data without knowledge of statistical significance!

I definitely agree with you Seymour, that it is frustrating to try to extract data from graphs without actually knowing the numbers.

[SeymourB in reply to gardening-girl]

gardening-girl -

I could have had the S. Lugdunensis infection before the COVID diagnosis, although I don't recall the distinctive smell of Lugdunensis. I may also have other infections that could have increased IgG, as well. We don't get diagnosed with an infection based on such slight changes, though.

=seymour=

[E-Lynn]

Perhaps I'm missing something, but I do not see a statistically significant change in your IgG, IgM, or IgA, and the small differences might be explained by how hydrated you were, the way your blood was drawn, or other operational differences. Sometimes I wonder if we ponder our labs like astrologers did the stars, perhaps trying to read more into them than is there. But as I say, I could be missing something and wish you all the best.

[SeymourB in reply to E-Lynn]

E-Lynn -

Agreed. I just wanted to share actual data for the curious and anxious. It's a hobby. I do plan to investigate the science of Total Immunoglobulin tests in more depth to try to show how meaningful or meaningless small changes are.

I do think we have a lot of time to gaze at the navel compared to other cancers. It's the perfect disease for science nerds.

But we really should address the elephant in the room - anxiety from numeric changes. I still really like Paul Simon's When Numbers Get Serious:

youtube.com/watch?v=cYLH0rP...(Not working as expected?)

=seymour=

[Ernest2]

Awesome. Many thanks Seymour for that.

I need to go away and think about this for some time. With my primary care doctor I've just decided to skip the 5th BioNtech ( The 4th one in March gave me 6 weeks of Nausea and my digestive system is still under investigation for the effect of Flucloxacillin last year. Yes they still prescribe Flucloxacillin in the UK as nobody wants to do the tests on the better alternatives). So I need to give Evusheld some thought, and with the rest of the UK CLLer's write to our politicians about that.

The thing I can't get a handle on is my IgA/G/M are low, BUT my practical immunity appears good (but that is confused I guess by factors such as all my friends being too scared to meet me, so post lockdown we still communicate on WhatsApp etc.)

My initial dumb question is how noisy are IgA/G/M results?

The other issue is sometimes I'm only given overall Ig results that I can't relate. From:

labtestsonline.org.uk/tests...

. . . I guess the single Ig figure might also include IgD and IgE ?

I made my first foray into private testing this year (B12 not Ig), but that is a conundrum because we know different labs produce different results, and NHS in the UK will fairly generally only look at their own regional lab.

Thanks again for your post.

Best wishes to all,

Ernest

[Ibru]

In looking at Evusheld “protection”over time, is it meaningful at all to look at the SARS-CoV-2 Semi–Quant total antibodies against the spike protein from Quest Laboratories?

[SeymourB in reply to Ibru]

Ibru -

I think many of us do like to see the spike IgG for emotional confirmation of effect. The Leukemia Lymphoma Society is now gathering spike IgG results from patients who are about to get the follow-up dose of Evusheld. So their scientists are interested, too. It will take months for them to publish the results, though.

I haven't seen any other study show how effective a given spike IgG value is against current Omicron variants for anyone else, either. By the time they can collect such data, the variants will probably have changed.

For me, I wanted to show that a person with CLL who did get infected had a substantially high reading. It demonstrates very well that Evusheld is not a substitute for a mask.

=seymour=

[Ibru in reply to SeymourB]

Thank you. Very well said. N95's rule!!

[Unglorious]

I think Evusheld plus an N95 mask, a good combination. You may still get Omicron but hopefully you won't land up in the hospital, because you have extra protection. I am hoping to get the shots, I am desperate for have some immunity. I take Acalabrutinib, does not allow for antibodies.