One Lesion, Not Several: Toxoplasmosis or Lymphoma in Advanced HIV
A solitary ring-enhancing brain lesion in a man with untreated, far-advanced HIV sits right where the usual empiric-toxoplasmosis playbook stops being an easy call — a solitary lesion is exactly the finding the literature flags as tilting the odds toward the diagnosis empiric treatment could miss.
J.O., a 46-year-old man, worked construction until back pain forced him onto disability two years ago and now lives alone since his divorce. He was diagnosed with HIV eleven years ago but stopped taking antiretroviral therapy roughly three years ago after losing his insurance and, by his own account, “just gave up going back.” He has not seen a doctor regularly since. His sister brought him in after two weeks of worsening headaches, and over the past several days a noticeable weakness in his left hand and arm that has made it hard for him to hold a coffee cup. He is alert and appropriately oriented, but the left-sided weakness is now clearly measurable on exam, roughly 4/5 strength throughout the left arm, and his sister reports he has seemed “a step behind” mentally over the past week that neither of them thought much of at the time.
His CD4 count returns at 32 cells/mm³ — profound immunosuppression, consistent with years off therapy. MRI shows a single ring-enhancing lesion in the right basal ganglia with moderate surrounding vasogenic edema; no other lesions are seen anywhere else in the brain. Toxoplasma IgG serology is positive, confirming prior exposure. In a patient this immunosuppressed with a positive Toxoplasma IgG, a ring-enhancing lesion is the classic picture for reactivation toxoplasmic encephalitis — except that classic picture is usually multiple lesions, and this is one. A solitary lesion doesn't rule out toxoplasmosis, which does present this way in a real minority of cases, but it is also the single MRI finding most consistently named in the literature as shifting the odds toward primary CNS lymphoma instead, a point usually traced to Ciricillo and Rosenblum's series on solitary lesions in AIDS. The counterweight is that the presumptive-diagnosis convention the NIH/CDC/IDSA opportunistic-infection guidelines still endorse does not turn on lesion count at all: it turns on positive serology, profound immunosuppression, and a documented response to empiric therapy, and he satisfies the first two outright. Neither reading settles him. What they establish is narrower, and more useful to the people standing at his bedside: his scan has moved him out of the population the empiric-trial-first convention was built on, without moving him into one where biopsy is clearly the right answer.
On the ward, reading one lesion instead of several
Start pyrimethamine-sulfadiazine with leucovorin today, and don't let the fact that this is one lesion instead of several talk us out of it. His Toxoplasma IgG is positive, his CD4 count is 32, and the appearance and location are entirely consistent with toxoplasmic encephalitis, which does present as a solitary lesion in a real minority of cases — not the majority, but real. The standard approach is an empiric trial with repeat imaging in ten to fourteen days, and I'd rather give that trial the chance to work than send a man this immunosuppressed for a brain biopsy before we've actually tried the far less invasive option first.
One thing I want settled before it drifts: we are not holding his ART for a month while we sort this out. People reach for the cryptococcal and tuberculous meningitis rule — wait four to six weeks, IRIS will kill him otherwise — and it does not transfer. Toxoplasmosis-associated IRIS runs around five percent, and the timing studies have not tied it to how early ART starts. The NIH/CDC/IDSA opportunistic-infection guidelines say two to three weeks, resting it on the 282-patient trial of opportunistic infections other than tuberculosis where the early arm started at a median of twelve days against the deferred arm's forty-five, and had significantly less AIDS progression and death. At a CD4 of 32, the thing most likely to hurt him over the next year is untreated HIV.
The solitary lesion is exactly the finding the literature most consistently names as tilting the odds toward primary CNS lymphoma rather than toxoplasmosis — it's not a minor detail, it's the single strongest imaging predictor either diagnosis has going for it. If this is lymphoma and we spend two weeks on an empiric antiparasitic trial before finding out, that's two weeks of a lesion left to grow in a disease where definitive therapy is what changes its course, and nothing about an antiparasitic trial is doing anything for him in the meantime.
I take the point that toxoplasmosis can present as a solitary lesion — it just makes this presentation less classic for toxo, not unlikely, and “less classic, not unlikely” is exactly the situation where I'd want more information before committing two weeks to the wrong diagnosis.
There's a way to get more information without exposing him to either the drug trial's two-week delay risk or an invasive biopsy outright: functional imaging. Toxoplasmosis lesions are typically metabolically cold on FDG-PET or thallium SPECT, and lymphoma lesions are typically hot — that distinction meaningfully re-stratifies the odds here without committing to either of your positions. There is a second one nobody has mentioned: EBV DNA on the CSF. The NIH/CDC/IDSA opportunistic-infection guidelines flag quantitative CSF EBV above roughly ten thousand copies per milliliter as raising real concern for CNS lymphoma — not diagnostic on its own, but it is a tube of fluid rather than a burr hole, and between it and functional imaging we could re-stratify him twice before anyone touches his brain. If we genuinely can't get the scan within a reasonable window, an empiric trial is still reasonable, but the solitary-lesion flag is real enough that I'd tighten the reassessment to seven days instead of the usual ten to fourteen, with a low threshold to go straight to biopsy if he isn't clearly better by then.
Agreed: empiric pyrimethamine-sulfadiazine-leucovorin started today. Functional imaging (FDG-PET or thallium SPECT) was requested urgently to help re-stratify while treatment is already underway. Antiretroviral therapy was scheduled to start inside the two-to-three-week window the NIH/CDC/IDSA opportunistic-infection guidelines name, rather than deferred on IRIS grounds: TE-associated IRIS is reported at roughly five percent, and the long deferral that is correct in cryptococcal and tuberculous meningitis has never been shown to apply here.
Not agreed, and left open: the exact reassessment interval if functional imaging isn't available in time. The oncologist wants the tighter seven-day window given the solitary-lesion flag; the infectious disease physician is comfortable with the standard ten to fourteen days given the positive serology and CD4 count, describing the tighter window as, in his words, “treating one lesion as more menacing than the numbers actually support.” Neither voice's position was overruled before the team moved to arranging the imaging.