Malaria Relapse: Distinguishing Recurrence From Reinfection

When malaria returns after treatment, the recurring infection falls into one of three biologically distinct categories: a relapse from dormant liver parasites, a reinfection from a new mosquito bite, or a recrudescence from blood-stage parasites that survived the original drug course. Telling these apart matters enormously for the patient and for public health, yet doing so in practice remains one of the trickiest problems in tropical medicine. A systematic review estimated that at least 79% of Plasmodium vivax recurrences are attributable to relapse rather than reinfection, which reshapes how clinicians and malaria-control programs should respond.¹ The difficulty is that all three types look identical under a microscope and produce the same fever.

Three Distinct Paths to the Same Fever

A relapse happens when dormant forms of the parasite, called hypnozoites, wake up in the liver weeks, months, or even years after the initial infection and release a new wave of parasites into the bloodstream. Only two human malaria species do this: Plasmodium vivax and Plasmodium ovale. Reinfection, by contrast, requires a fresh bite from an infected mosquito delivering entirely new parasites. Recrudescence is different again: the original blood-stage parasites were never fully cleared, often because the drug used was not effective enough against that particular strain, and they multiply back to detectable levels.

For P. falciparum, the deadliest malaria species, there are no hypnozoites, so recurrence is either reinfection or recrudescence. For P. vivax, all three possibilities are on the table at once, and the treatment implications diverge sharply. A relapse demands radical cure with drugs that target the liver stage. A reinfection calls for renewed blood-stage treatment and better mosquito-bite prevention. A recrudescence signals possible drug resistance and the need for a different antimalarial regimen.

The Hypnozoite Problem

Hypnozoites are the biological root of relapse, and they have been frustratingly difficult to study. These are tiny, single-nucleus forms of the parasite that sit quietly inside liver cells while their siblings grow into full-blown schizonts, burst, and flood the bloodstream. Research using liver-chimeric mouse models has shown that the two fates diverge as early as three days after infection: developing schizonts balloon in size and replicate their DNA, while hypnozoites stay small and show only a single nucleus, with minimal DNA replication.² Yet they are not inert. In one model system, hypnozoites modestly increased in size over 49 days, suggesting ongoing metabolic activity rather than true suspended animation.³

The membrane surrounding hypnozoites has unique molecular features, including a polarized prominence of a protein called UIS4 that does not appear on growing schizonts.² Another host protein, aquaporin-3, is recruited to the hypnozoite membrane in a pattern distinct from its recruitment to schizonts, hinting that dormant and developing liver forms have different relationships with the host cell.⁴ All of this matters because any future drug or vaccine targeting hypnozoites specifically will need to exploit these biological differences.

What Triggers a Relapse

Nobody fully understands what flips the switch that wakes a hypnozoite. One hypothesis proposes that systemic infections with other parasites or bacteria can trigger reactivation. A review of clinical observations found that parasitic and bacterial febrile illnesses appeared capable of provoking P. vivax relapse, while viral infections did not seem to have the same effect, suggesting that specific components of the inflammatory response, rather than fever alone, may be involved.⁵ A more speculative idea is that proteins introduced by a new Anopheles mosquito bite could provide the activation signal, essentially that the hypnozoite recognizes a cue associated with its original vector.⁶

From an evolutionary standpoint, the timing of relapse is not random. Tropical strains tend to relapse quickly, within three to four months, while strains from temperate regions delay for seven to eight months.⁷ This geographic pattern aligns neatly with mosquito seasonality: in the tropics, where mosquitoes are available year-round, fast relapse maximizes the parasite’s chances of being picked up and transmitted. In temperate zones, mosquitoes disappear during winter, so the parasite is better off waiting until the following transmission season.⁸ The parasite appears to have evolved its dormancy timing to match local vector ecology.

Why Telling Relapse From Reinfection Is So Difficult

In a patient living in an area where malaria is common, a second episode of fever and parasites could be any of the three types of recurrence, and the clinical presentation gives you no way to distinguish them. The traditional approach has been genotyping: compare the genetic profile of the parasites from the first episode to those from the recurrence. If the parasites are genetically identical, the logic goes, it is a relapse or recrudescence. If they look different, it is a new infection.

The reality is far messier. A study using whole-genome sequencing found that two-thirds of recurrent P. vivax infections were heterologous relapses, meaning the returning parasites were genetically different from the initial ones because they came from a different batch of hypnozoites deposited during the same or a prior infection.⁹ This fundamentally undermines the simple “same genotype equals relapse” approach: a relapse can produce parasites that look like a brand-new infection.

For P. falciparum, where the question is recrudescence versus reinfection, genotyping faces its own challenges. A comparison of three commonly used genetic markers in Uganda found that classification depended heavily on which markers were used and how mixed infections were interpreted. One marker, msp-2, performed better than others in that setting.¹⁰ More recent work has evaluated targeted amplicon deep sequencing panels, which showed high concordance between markers, correctly distinguishing recrudescence from new infection in 90% of patients across all markers in one study.¹¹

Genomic and Statistical Advances

Researchers have been pushing beyond traditional genotyping toward whole-genome approaches and probabilistic models. In the Peruvian Amazon, whole-genome sequencing of P. vivax allowed investigators to distinguish homologous relapses (same parasite clone returning) from heterologous relapses and reinfections with improved accuracy, identifying 12 homologous relapses and 3 likely heterologous relapses among recurrent cases.¹²

A complementary approach combines genetic data with the timing of the recurrence. A model called Pv3R uses parasite genetic similarity alongside time-to-event data and identity-by-descent analysis to assign each recurrence a probability of being a relapse, reinfection, or recrudescence.¹³ The key insight is that no single data type is sufficient. The genetics tell you how related the parasites are, but the timing tells you how plausible each category is given local transmission intensity and known relapse patterns. Combining both yields a probability estimate rather than a forced binary classification.

For antimalarial drug trials specifically, where determining whether treatment failures are true recrudescences matters enormously, nanopore sequencing is emerging as a rapid option. One study demonstrated that a nanopore-based amplicon sequencing assay could distinguish recrudescence from new infection in 85% of paired P. falciparum samples across all six markers tested.¹⁴ The advantage of nanopore technology is speed: results come back within hours rather than weeks, which could accelerate drug efficacy assessments in the field.

When the Drug Works but the Patient’s Genes Do Not

Primaquine is the main drug used for radical cure, meaning it targets hypnozoites in the liver to prevent relapse. But primaquine is a prodrug: it needs to be metabolized by a liver enzyme called CYP2D6 before it becomes active against hypnozoites.¹⁵ People who carry genetic variants that reduce CYP2D6 activity metabolize primaquine poorly, which means the drug may not clear their hypnozoites even at standard doses.

A meta-analysis pooling data across multiple studies found that individuals with reduced CYP2D6 activity had roughly 1.8-fold higher odds of P. vivax recurrence compared to those with normal or fast metabolizer status.¹⁶ Individual studies from different populations paint a consistent picture. In South Korea, intermediate metabolizers had about 2.3 times the odds of recurrence.¹⁷ In the Brazilian Amazon, those with reduced CYP2D6 activity had about 1.9 times the risk.¹⁸

This has a practical implication that is easy to overlook: when a patient on primaquine has a recurrence, it might not be a reinfection or a drug-resistant relapse. It might simply be that the patient’s genetic makeup prevented the drug from doing its job. Without CYP2D6 testing, that patient could be misclassified as having a new infection when the real problem is inadequate drug activation.

The G6PD Safety Barrier

Primaquine’s other major limitation is its potential to cause severe hemolysis, the destruction of red blood cells, in people with glucose-6-phosphate dehydrogenase (G6PD) deficiency, the most common inherited enzyme deficiency in humans.¹⁹ G6PD deficiency is concentrated in malaria-endemic populations, which is an evolutionary irony: the very people who most need primaquine’s anti-relapse properties are disproportionately at risk of its worst side effect.

This creates a genuine clinical dilemma. In many settings, G6PD testing is unavailable or only qualitative (positive/negative), which can miss women who are heterozygous and have intermediate enzyme activity. A quantitative point-of-care G6PD test has been developed that can provide enzyme activity measurements in basic clinical laboratories, potentially allowing clinicians to identify patients who can safely receive primaquine and at what dose.²⁰

Pharmacometric modeling has suggested that ascending-dose primaquine regimens, which start at a low dose and gradually increase, can achieve radical cure in G6PD-deficient individuals while keeping hemolysis within manageable bounds. In Southeast Asian G6PD-deficient male volunteers, standard primaquine doses reduced the lifespan of deficient red blood cells by about 30 days, but the hemoglobin drops were predictable and recoverable.²¹ Meanwhile, a safety study of high-dose, short-course primaquine after point-of-care G6PD testing found that while about two dozen participants experienced hemolysis-related adverse events, none developed severe clinical symptoms, and nearly all completed their treatment course with hemoglobin recovery within a week.²² These findings suggest the safety barrier, while real, may be more navigable than previously feared when proper testing and dose management are in place.

Serological Markers for Finding Hidden Carriers

One of the biggest gaps in malaria elimination is the inability to identify people carrying hypnozoites before they relapse. You cannot see hypnozoites on a blood test. There is no liver biopsy that health systems could realistically deploy at scale. Researchers have been working on a workaround: measuring antibody responses to P. vivax proteins as a proxy for recent infection, on the theory that someone who was infected recently is likely still carrying hypnozoites.

A panel of eight P. vivax proteins was shown to classify individuals with recent infections (within nine months) with about 80% sensitivity and specificity in validation cohorts from Thailand, Brazil, and the Solomon Islands. Modeling suggested that a serological testing and treatment strategy could reduce P. vivax prevalence by 59 to 69%.²³ However, when this same eight-antigen classifier was tested in a subsequent field study, sensitivity dropped to about 70%, below the 80% threshold established in the original validation.²⁴ The concept remains promising, but the gap between controlled validation and real-world performance is a familiar story in diagnostics development.

How Other Malaria Species Complicate the Picture

Plasmodium malariae presents a different puzzle. It does not form hypnozoites, so it should not relapse. Yet it can recur years or even decades after exposure, sometimes in people who have not returned to malaria-endemic areas. The leading explanation is that P. malariae persists at extremely low levels in the blood, below the threshold of detection, essentially hiding in plain sight rather than hiding in the liver. The slow metabolism of some trophozoites may shield them from antimalarial drugs, enabling recrudescence long after the initial infection was supposedly cleared.²⁵ Whether P. malariae might also harbor some form of liver dormancy remains unresolved: the failure to find hypnozoites in liver biopsies does not conclusively rule them out, since biopsies sample only a tiny fraction of liver tissue.²⁶

This matters because a P. malariae recurrence years after travel to an endemic area will almost certainly be misidentified at first as a new infection or mistaken for a different illness entirely. Clinicians in non-endemic countries who see a returned traveler with malaria should consider the possibility that the infection dates from the original trip, even if it was years ago.

Children and Recurrence

The burden of recurrence falls heavily on children in endemic areas. In Papua, Indonesia, the cumulative risk of having a recurrent P. vivax episode within one year was about 50% in children under five, compared to roughly 31% in children aged five to fourteen.²⁷ The same diagnostic challenge applies: molecular markers for distinguishing relapse from reinfection have not been fully validated in pediatric populations, and seasonal analysis of case patterns remains one of the only indirect methods to estimate which type of recurrence is occurring.

Children are also a complication for radical cure. Tafenoquine, a newer alternative to primaquine that requires only a single dose, is currently approved only for patients aged 16 and older in many settings. Clinical trial protocols for mass drug administration in Peru, for instance, specify tafenoquine for those 16 and up but default to a seven-day primaquine course for younger patients.²⁸ A seven-day course is harder to complete than a single dose, and adherence in children depends heavily on caregivers.

Mass Drug Administration and the Elimination Puzzle

From an elimination standpoint, hypnozoites are the single biggest obstacle to getting rid of P. vivax. You can spray mosquitoes and distribute bed nets, but as long as a reservoir of people carry dormant liver parasites, the infection keeps coming back. Simulation modeling indicates that multiple rounds of mass drug administration with a high-efficacy radical cure drug are needed to have a substantial chance of eliminating P. vivax from a given area.²⁹

Real-world evidence is starting to accumulate. Focal primaquine mass drug administration in a low-endemicity area of northern Myanmar reduced both clinical and subclinical P. vivax infections.³⁰ But containment efforts face the recurring challenge that some patients relapse because of antimalarial drug resistance or non-compliance, while others are ineligible for primaquine due to G6PD deficiency.³¹ Each of those patients becomes a potential source of ongoing transmission, perpetuating a cycle that no amount of mosquito control alone can break.