Xpert MTB/XDR Surfaces Extensive Multiclass Tuberculosis Resistance in Northern Morocco
Routine deployment of Xpert MTB/XDR across a regional tuberculosis diagnostic network in northern Morocco produced a dense molecular portrait of drug resistance over four years. In 258 testing episodes collected from April 2022 through July 2026, the assay did more than flag rifampicin resistance: it identified layered patterns of isoniazid, fluoroquinolone, and broader multiclass resistance that are difficult to capture with rifampicin-only screening alone.
The surveillance message is immediate. In a setting where tuberculosis control depends on early recognition of resistant disease, a single molecular platform generated enough resistance detail to separate patients with isolated rifampicin susceptibility from those carrying deeper and more complex resistance burdens.
Resistance Burden Extended Well Beyond Rifampicin
Mycobacterium tuberculosis complex was detected in 203 of the 258 testing episodes, or 78.7%. Among those positive episodes, 188 had an interpretable rifampicin result. Molecular rifampicin resistance appeared in 75 of 188 episodes, equal to 39.9%, while isoniazid resistance was detected in 71 of 192 episodes, or 37.0%.
The pattern becomes more concerning when paired results are considered together. Molecular multidrug-resistant tuberculosis was identified in 51 of 184 episodes with both rifampicin and isoniazid information, accounting for 27.7%. In other words, a substantial fraction of patients entering this diagnostic stream were not carrying a single resistance trait but a resistance phenotype already aligned with intensified treatment needs and closer follow-up.
One sentence captures the operational value of the assay: it did not merely confirm tuberculosis, it differentiated resistance architecture.
Multiclass Resistance Was Common Even Among Rifampicin-Susceptible Episodes
Among 186 episodes with interpretable rifampicin and additional resistance results, 75 episodes, or 40.3%, showed resistance to at least one additional drug class. The split within that group is clinically revealing. Resistance to at least one other drug class was found in 21 of 112 rifampicin-susceptible episodes, or 18.8%, and in 54 of 74 rifampicin-resistant episodes, or 73.0%.
Those figures matter because they expose the ceiling of rifampicin-only surveillance. A patient can look manageable by the oldest molecular gatekeeper and still harbor resistance to companion drugs that shape regimen selection and outcomes. The assay therefore expands the surveillance lens from a binary rifampicin result to a broader resistance map that can influence both programmatic reporting and downstream clinical decision-making.
Among 92 resistant episodes with complete profiles, 66 episodes, or 71.7%, showed resistance to two or more drug classes. This was not a scattered collection of isolated mutations. It was a heavy clustering of multiclass resistance in the same diagnostic network that is intended to detect and contain it.
Mutation Patterns Pointed to Expected but Important Molecular Targets
The assay output also exposed which molecular targets were carrying the resistance burden. rpoB3– and rpoB4-associated signals predominated among rifampicin-resistant episodes. katG-associated signals were most common among isoniazid-resistant episodes, while gyrA targets dominated the fluoroquinolone-resistant group.
These patterns fit the known biology of drug resistance in tuberculosis. rpoB mutations alter the rifampicin target site, katG changes impair activation of isoniazid, and gyrA alterations reduce fluoroquinolone binding to DNA gyrase. The assay’s value lies in turning those mechanisms into actionable surveillance data at the point where patients first enter the testing system.
Mixed wild-type and mutation-associated signals also appeared across rpoB and gyrA. Those mixed signals raise the possibility of heteroresistance, mixed infection, or borderline molecular states that deserve careful interpretation rather than reflexive simplification. In surveillance terms, they signal a need for confirmatory methods when molecular outputs are discordant, atypical, or likely to alter regimen choice.
Routine Molecular Surveillance Can Reveal More Than Case Counts
The four-year programmatic experience suggests that routine Xpert MTB/XDR testing can function as a surveillance instrument, not just a diagnostic add-on. A regional network that captures multiclass resistance patterns in real time can begin to describe the local burden of difficult-to-treat tuberculosis with far more precision than rifampicin-only testing allows.
This matters in practical public-health terms because drug-resistant tuberculosis is not only about finding resistance; it is about classifying it correctly early enough to shape therapy and limit transmission. When multiclass resistance is concentrated among rifampicin-resistant episodes, the need for expanded molecular coverage becomes even clearer. The assay gives programs a way to identify higher-risk resistance profiles before culture-based results or full sequencing are available.
Still, molecular surveillance cannot stand alone. The resistance signatures reported here need phenotypic susceptibility testing and genomic sequencing to resolve discordant patterns, confirm unusual profiles, and define how often mixed signals reflect true biologic complexity. Molecular tools can accelerate detection, but they do not fully settle the clinical question in every patient.
Programmatic Gains Depend on Confirmatory Testing and Regional Infrastructure
The data support integrating Xpert MTB/XDR into routine drug-resistance surveillance in northern Morocco, but they also define the limits of a purely molecular approach. Programs that rely on this platform will still need laboratory pathways capable of handling borderline results, adjudicating discrepancies, and linking molecular resistance with treatment history and outcome data.
The broader value of this work lies in its programmatic scale. Over four years, routine testing generated enough information to characterize both the frequency and structure of resistance across a regional network. That kind of dataset is exactly what tuberculosis control programs need when they are trying to move from suspicion-based testing to systematic resistance monitoring.
The preprint provides the dataset, assay-level detail, and codebook needed to reproduce the analyses, but it remains a preprint and has not undergone peer review. The main clinical question now is not whether molecular surveillance can detect resistance—it clearly can—but how best to combine it with phenotypic and genomic methods so the resistance pattern seen on the instrument matches the resistance problem in the patient.