The Monocyte Monolayer Assay: Asking Cells to Predict the Future

A patient shows up needing blood. Her antibody screen comes back positive for something rare, something we can find antigen-negative units for, but not easily, and not fast. The question on the table isn't "is this antibody there." It clearly is. The question is: will it actually hurt her if we transfuse anyway?
This is the part of transfusion medicine that doesn't make it into the textbooks as often as it should. We're very good at detecting alloantibodies. We are, frankly, less good at predicting what they'll do once red cells carrying the corresponding antigen show up in circulation. Some antibodies are clinically vicious. Others sit there, technically present, doing essentially nothing. Serology alone can't always tell you which kind you're dealing with. So instead of asking the antibody directly, we ask a different question: what will a monocyte do when it meets a red cell coated in this antibody?
That's the monocyte monolayer assay, or MMA.
What the assay actually does
The setup is simpler than it sounds. You isolate monocytes from a normal donor and let them settle into a monolayer, a single sheet of cells, on a glass or plastic surface. You then take red cells sensitized with the patient's antibody (antigen-positive donor cells incubated with the patient's serum or plasma) and layer them on top. Then you wait, and you look.
Monocytes have Fc receptors. If the antibody coating those red cells is one the monocyte's receptors recognize and respond to, the monocyte will do one of two things: engulf the red cell whole (phagocytosis) or grab onto it without fully swallowing it (adherence, sometimes called rosetting depending on the exact readout). Count how many red cells are being eaten or grabbed out of a set number of monocytes, and you get a percentage. That percentage is the assay's answer to the question serology couldn't fully address: is this antibody the kind that provokes extravascular destruction in vivo, or not.
Higher phagocytic/adherence index, more concern for clinically significant hemolysis. Lower index, more reassurance that the antibody may be biologically quiet, at least as far as this in vitro proxy can tell you.
Where it earns its keep
The MMA isn't run casually. It shows up in specific, high-stakes scenarios:
Deciding whether to transfuse antigen-positive units when antigen-negative units are scarce or unobtainable, particularly for high-prevalence antigens where "just find negative blood" isn't a real option
Assessing antibodies of uncertain clinical significance, especially ones with a spotty or contradictory track record in the literature
Supporting decisions in hemolytic disease of the fetus and newborn, where the stakes of both over- and under-intervention are considerable
In all of these, the assay is functioning as a stand-in for a question we can't ethically or practically answer by simply transfusing and watching what happens.
Where it stops being reassuring
Here's the part I think about more than the mechanics. The MMA correlates reasonably well with clinical hemolysis for a number of antibody specificities, and that correlation is exactly why the assay gets used. But correlation built from case series and reference lab experience is not the same as a validated, universally reliable predictive test. Interlaboratory variability exists. Cutoffs for "significant" versus "not significant" differ by lab and by antibody specificity, sometimes for reasons of accumulated institutional experience rather than a single settled cutoff in the literature. And a normal donor's monocytes, doing their job in a dish, are still one step removed from what actually happens inside a specific patient's spleen, liver, and vasculature.
So when a clinician calls asking whether the MMA result means it's "safe" to transfuse, the honest answer is more hedged than that word implies. A reassuring MMA lowers the estimated risk. It does not zero it out. We are still, in the end, making a probabilistic judgment call dressed up in a percentage.
I don't think that's a flaw specific to this assay. It's the condition of most predictive testing in medicine: we build proxies because we can't ask the real question directly, and then we have to be honest with ourselves about how far the proxy actually gets us.
The MMA gets transfusion medicine further than serology alone would. Whether it gets us far enough, in any individual case, is a judgment call every time. I'm not sure there's a version of this test, or maybe any test, that removes that part of the job.



