A debate over how best to monitor the immune-blocking drug ravulizumab is raising an important warning for clinicians: routine laboratory tests may not reliably show whether the medicine is fully suppressing its target in people with generalized myasthenia gravis. In a published critique of a recent study, researchers argue that commonly used complement activity assays—known as CH50 and AH50—could give a misleading picture of pharmacological inhibition, particularly when blood samples are processed outside the body. The concern centers on a deceptively simple question: does a test measure the drug’s actual effect in the patient, or does it measure what happens after the drug and its target separate in a laboratory tube? The distinction could affect treatment decisions for patients receiving ravulizumab, a long-acting monoclonal antibody designed to block complement component 5, or C5. C5 is a central protein in the terminal complement cascade, an immune pathway that can damage tissues when activated inappropriately. Ravulizumab’s therapeutic effect depends on binding C5 tightly enough to prevent the protein from being converted into fragments that ultimately help form the membrane attack complex.
The disputed study examined classical and alternative complement pathway activity as possible biomarkers for therapeutic drug monitoring. CH50 assesses the capacity of serum to activate the classical complement pathway, while AH50 measures activity through the alternative pathway. In both assays, patient serum is diluted and mixed with specialized reagent systems containing complement-sensitive targets. The amount of residual pathway activity is then inferred from how much cellular damage occurs. A low result may suggest that complement is blocked, whereas a high result may indicate that the pathway remains active. But the critique argues that these tests have not been analytically or clinically validated for monitoring C5 inhibitors. That gap matters because ravulizumab is not simply removing C5 from circulation; it is forming a drug–target complex. If that complex dissociates during sample dilution, incubation or handling, previously occupied C5 may become available again. The test could then report restored complement activity even though the drug is suppressing C5 effectively inside the patient. In other words, the assay may be measuring the stability of a molecular complex under laboratory conditions rather than the patient’s true pharmacodynamic state.
This phenomenon is a familiar challenge in bioanalysis of therapeutic antibodies. A “free target” assay attempts to measure the fraction of a protein that is not bound by a drug, while a “total target” assay measures both bound and unbound forms. Neither measurement is automatically equivalent to biological activity. The balance can shift depending on binding affinity, drug and target concentrations, dilution factors, temperature, pH and the length of time between blood collection and analysis. Ravulizumab binds C5 with high affinity, but high-affinity interactions can still be disturbed when samples are diluted or exposed to assay conditions that differ from those in the bloodstream. The authors of the critique point to clinical studies in which complement inhibition was assessed using an optimized serum free-C5 test. They also cite direct comparisons suggesting that CH50 results do not consistently agree with free-C5 measurements. This discordance creates a practical risk: two tests performed on the same patient could support different conclusions about whether complement blockade is adequate, potentially prompting unnecessary dose changes, additional testing or an incorrect interpretation of treatment failure.
The issue is further complicated by the way ravulizumab concentrations were measured. In the phase 3 CHAMPION MG trial, investigators identified a serum ravulizumab concentration above 175 micrograms per milliliter as a pharmacokinetic threshold and quantified total ravulizumab—both antibody bound to C5 and antibody circulating freely—using a validated liquid chromatography–tandem mass spectrometry method. The newer analysis used the same numerical threshold but measured free ravulizumab with an enzyme-linked immunosorbent assay, or ELISA. Although both approaches produce concentration estimates, they are not interchangeable. Mass spectrometry separates and identifies molecules according to their chemical properties and can be designed to quantify total drug with high specificity. ELISA relies on antibody recognition and can be configured to detect free or bound forms, but its result depends heavily on reagent design, sample preparation and the molecular species accessible to the detection antibodies. Applying a threshold established for one assay to results generated by another can therefore create an apples-to-oranges comparison. The critique suggests that this methodological mismatch may help explain why 22.7 percent of patients in the analysis appeared to have ravulizumab levels below the 175-microgram-per-milliliter benchmark.
That percentage is potentially attention-grabbing because a concentration below a defined threshold might be interpreted as underexposure, loss of complement control or a need to adjust treatment. Yet a threshold has meaning only within the analytical system and clinical context in which it was validated. A total-drug concentration is not equivalent to a free-drug concentration: total ravulizumab includes molecules already engaged with C5, while free ravulizumab represents antibody available to bind additional target. Conversely, measuring free drug alone does not directly prove that enough C5 is inhibited to prevent downstream complement activation. Pharmacokinetics describes how drug concentrations change over time, whereas pharmacodynamics describes what the drug does to its biological target. Effective therapeutic monitoring requires a validated bridge between the two. That bridge should establish the relationship among dose, drug concentration, free C5, complement activity and clinical outcomes. Without it, an apparently low concentration or an unexpectedly high CH50 value could reflect assay behavior rather than inadequate treatment. The researchers calling for clarification say details about the optimization, timing and analytical performance of the CH50 and AH50 procedures are needed before the tests can be interpreted confidently.
The clinical data presented in the disputed analysis also appear less straightforward than the laboratory findings. Patients who continued ravulizumab reportedly had lower median AH50 and CH50 values than those who discontinued treatment, a pattern that might initially suggest stronger complement inhibition among those remaining on therapy. However, the median change from baseline in the Myasthenia Gravis Activities of Daily Living, or MG-ADL, score was zero in both groups at follow-up. MG-ADL is a patient-centered measure covering everyday functions affected by myasthenia gravis, including speaking, chewing, swallowing, breathing and limb activity. A stable score in both groups does not readily support a conclusion that one group experienced clinically meaningful worsening. Another measure, the Quantitative Myasthenia Gravis score, moved in the opposite direction: it improved by 0.5 points in patients who discontinued ravulizumab but worsened by 2.0 points in those who continued. Such small or discordant changes are difficult to interpret without repeated assessments, information about treatment timing and a prespecified definition of clinical deterioration. They also conflict with other reports describing stable improvements over time among patients treated with ravulizumab.
The critique emphasizes that myasthenia gravis is a fluctuating disease, making isolated measurements particularly vulnerable to misinterpretation. The disorder is caused by impaired communication between nerves and muscles, most often because autoantibodies disrupt proteins at the neuromuscular junction. In some patients, complement activation contributes to damage at the postsynaptic membrane, providing the rationale for C5 inhibition. But symptoms can vary with exertion, infection, medication changes, sleep and the interval since the last treatment. A blood test taken at one time point may therefore fail to capture the patient’s functional state, even if the assay itself is technically sound. Likewise, a clinical score collected only at baseline and study end may miss transient deterioration or gradual recovery. The authors argue that trends across several validated measures should be considered together with the patient’s own account of how long the treatment’s benefit lasts. Current management guidelines recommend regular clinical examinations and disease-specific scoring, but they do not recommend routine therapeutic drug monitoring for ravulizumab. That absence reflects a broader consensus: until laboratory biomarkers are rigorously validated against outcomes, repeated clinical evaluation remains the more dependable way to determine whether complement inhibition is translating into meaningful benefit.
The exchange does not show that CH50 and AH50 are useless, nor does it establish that ravulizumab is ineffective. Instead, it illustrates how a test can be biologically plausible yet insufficiently validated for a high-stakes clinical purpose. Before complement activity assays are adopted for therapeutic monitoring, researchers would need to define standardized sample handling, quantify the extent of drug–C5 dissociation during testing, compare results with validated free-C5 and total-drug methods, and demonstrate that assay values predict symptoms, exacerbations or other meaningful outcomes. Studies would also need to account for dosing intervals and differences between patients in drug clearance and complement biology. The critique was authored by F.S. and R.P.; the source reports that the work was funded by Alexion, AstraZeneca Rare Disease, and that one author is employed by Alexion and may hold AstraZeneca stock or stock options. Those disclosures do not resolve the scientific questions, but they are important context when interpreting the debate. For now, the message for clinicians is clear: a surprising CH50, AH50 or ravulizumab concentration should not be treated as a standalone verdict on treatment success. The most reliable signal remains the patient’s sustained clinical response, assessed systematically and repeatedly alongside any laboratory evidence.
