Infectious Diseases — New This Week

A lopsided week — CID contributed only two in-scope articles, while an entry-date sweep pulled a 17-article OFID backlog forward. The practice-shaping tier is unusually strong. Three items are worth real attention: the meta-analysis showing that no randomized trial in vaccinated, Omicron-era patients has demonstrated benefit from nirmatrelvir-ritonavir, with the NNT rising roughly an order of magnitude; a 7-country cohort re-baselining M. tuberculosis bacteremia in hospitalized people with HIV at ~1 in 20, with urine tests outperforming sputum; and a Uzbek cohort showing what life after bedaquiline resistance actually looks like — a return to injectables, 19 months of therapy, and a third of cases in people who never received the drug. Two case reports carry outsized teaching value: Q148H emerging after rifampin-based LTBI therapy (and costing the patient long-acting injectable eligibility), and a discordant two-tier Lyme serology that was actually soft-tick relapsing fever.

Clinical Infectious Diseases

CIDHIV · RandomizedNotable · surrogate

Antiviral Effect, Safety, and Tolerability of Oral VH4011499 (VH-499), a New HIV-1 Capsid Inhibitor: Proof-of-Concept Trial

In ART-naive adults with HIV-1 viremia, how much does 10 days of a new oral capsid inhibitor drop the viral load?

Population23 randomized ART-naive adults 18–65 across 5 countries; VL ≥3000 copies/mL, CD4 ≥200. Prior parenteral PrEP (cabotegravir or lenacapavir) excluded. InterventionOral VH-499 25, 100, or 250 mg on days 1 and 6 — monotherapy for 10 days, then open-label standard-of-care ART from day 11. ComparisonMatching placebo (n = 3). OutcomeMean maximum log₁₀ VL decline −1.8, −1.8, and −2.2 for 25/100/250 mg vs −0.2 for placebo. Descriptive only — no confidence intervals and no hypothesis testing, by design. 19/20 (95%) had no emergent capsid resistance mutations; one participant on the lowest dose developed a Q67H mixture by day 6 with 3.9-fold loss of sensitivity. All adverse events grade 1–2; no serious events or discontinuations. Fellow implicationA potency signal, not a practice change. The −2 log₁₀ benchmark sits alongside lenacapavir (−2.3) and dolutegravir (−2.5), and VH-499 is positioned to avoid lenacapavir's CYP3A4-inhibition interaction burden. Ten days of monotherapy in 23 people cannot speak to durability or resistance barrier. ViiV-sponsored, with the sponsor involved in design, analysis, and manuscript preparation.

Also new in CID. Stewardship ethics — an Ethics Rounds essay on an unhoused man with source-controlled MSSA septic arthritis and hardware osteomyelitis who asks to stay in hospital for IV therapy. The authors work three conflicts and land on recommending the evidence-based oral regimen while mobilizing social work and case management: “the ethical response to vulnerability is not necessarily more medical care, but more attentive care,” and “antibiotics should not be used to solve nonantibiotic problems.” The one hard number is that ~3% of central venous catheters cause a serious complication (ciag440). Screened out — four corrections, one correspondence and one letter reply on valve strands, one Voices of ID essay, and three Vol. 83(1) clinical-challenge/photo-quiz pieces. Backlog — the Vol. 83(1) issue assignment on 23 July restamped ~70 older online-first articles with an in-window publication date; these were scoped out by online date and remain available for a future pickup.

Open Forum Infectious Diseases

OFIDSTI · Phase 3 secondaryHigh yield · ID

Oral Gepotidacin for Uncomplicated Urogenital Gonorrhea: NAAT Outcomes in EAGLE-1

Does efficacy measured by NAAT — the test clinicians actually use — agree with the culture-based endpoint regulators require?

Population628 enrolled → micro-ITT 406 → 324 evaluable by NAAT; 91.6% male, mean age 33.1. Only ceftriaxone-susceptible isolates entered micro-ITT. InterventionOral gepotidacin 2 × 3000 mg, 10–12 h apart. ComparisonCeftriaxone 500 mg IM × 1 + azithromycin 1 g PO × 1. OutcomeNAAT-confirmed success, micro-ITT 82.5% (151/183) vs 75.0% (132/176), difference 8.2% (95% CI −0.1%, 16.5%); evaluable population 88.3% vs 86.3%, difference 3.8% (−3.6%, 11.2%). Both below the parent trial's culture-confirmed 92.6% vs 91.2%. The headline is the discordance: at days 4–10, 41/324 (12.7%) were NAAT-positive while 0/324 were culture-positive. Baseline NAAT–culture concordance was 98.4%. What this changesConfirms the Lancet 2025 primary result; the durable addition is the test-of-cure timing rule.

Clinical takeaway. NAAT and culture agree almost perfectly at diagnosis and diverge sharply as a test of cure — the difference is viability, not detection. A NAAT drawn at days 4–10 calls roughly one in eight cured patients a failure, which is why BASHH specifies ≥14 days. Two caveats: the 8.2% gap “favoring” gepotidacin is largely a data-completeness artifact and its CI crosses zero, and only ceftriaxone-susceptible strains were studied — precisely not the niche people will reach for this drug in. Pharyngeal disease remains the weak site (53.3% vs 81.3%; the only culture-proven persistence was pharyngeal, in the gepotidacin arm). GSK-funded, 7 of 11 authors GSK employees.

Major studies & new antimicrobials

OFIDTB/HIV · Prospective cohortHigh yield · ID

Prevalence and Mortality due to Mycobacterium tuberculosis Bloodstream Infection in Adults With HIV

In the modern ART era, how common is TB bacteremia in adults with HIV — and what does it cost them?

What changedA 7-country prospective cohort (n = 1703; Malawi, South Africa, Tanzania, Thailand, Uganda, Viet Nam, Zambia) resets a figure fellows still learn from a pre-ART-era meta-analysis. Among inpatients with confirmed TB, bacteremia was 24.2% (95% CrI 17.5–31.4) — well below the ~45% previously cited, explained by higher CD4 (median 361 vs 76) and a single blood culture here versus two. Key figuresInpatients 5.0% (95% CrI 3.4–6.8), enrolled irrespective of TB symptoms; outpatients 0.3%. Highest in South Africa, 23/195 (11.8%). Median CD4 with bacteremia 35 cells/µL (IQR 12–103) vs 372 without; 32% vs 79% on ART. Mortality 30-day HR 2.65 (95% CrI 1.06–5.01) — the 70-day estimate crosses 1.0. Diagnostic yield: urine Xpert Ultra 70.6%, urine Determine-LAM 58.8%, sputum Xpert only 52.9%; 85.3% caught by at least one rapid test. Fellow implicationTB bacteremia is essentially an inpatient phenomenon that tracks profound immunosuppression rather than symptoms — a hospitalized patient with advanced HIV can be bacteremic without classic TB symptoms. When you suspect disseminated TB there, urine-based testing carries more diagnostic load than sputum, the mirror image of non-bacteremic HIV-TB.

Clinical takeaway. Roughly 1 in 20 hospitalized adults with HIV in high-burden settings is TB-bacteremic, and it roughly doubles-to-triples early death. A single mycobacterial blood culture underdiagnoses — 42 participants with dual-positive urine Xpert and LAM had negative blood cultures — so a negative culture is not reassurance. The authors stop short of recommending routine blood cultures, but they make a strong implicit case for urine testing and for not delaying treatment.

OFIDResistance · Retrospective cohortHigh yield · ID

Characteristics, Treatment, and Long-term Outcomes of Bedaquiline-resistant Tuberculosis in Karakalpakstan, Uzbekistan

Once M. tuberculosis is bedaquiline-resistant, what regimen does the patient actually end up on — and what happens to them?

What changedOne of only two cohorts worldwide describing life after bedaquiline resistance — and in a setting with 0% HIV, which is the argument that the resistance itself, not HIV, drives the poor outcomes. Key figuresn = 50. 32% (16/50) had never received bedaquiline and 20% had never been treated for TB at all; 69% were also fluoroquinolone-resistant. The regimen reverted to injectables — linezolid 96%, imipenem 66%, amikacin 50% — a median of 5 drugs/day for 18.8 months. Median time to sustained culture conversion 5.9 months versus 27 days for bedaquiline-susceptible BPaL in the same programme. Survival 79.7% at 18 months, 63.4% at 5 years; TB-free survival 60.4%. All 40 mutations were in rv0678; tNGS was indeterminate in 81% and called 3/27 susceptible despite phenotypic resistance. Fellow implicationA bedaquiline DST result reclassifies the whole case. And because current genomics detects only 49–69% of phenotypic resistance, a molecular-only algorithm cannot exclude it — phenotypic DST at 1.0 mg/L is still required. rv0678 also confers clofazimine cross-resistance.

Clinical takeaway. Bedaquiline resistance turns a 6-month all-oral course into a ~19-month, 5-drug, injectable-containing regimen with roughly 1-in-5 mortality by 18 months. The most unsettling number is that a third of these patients were bedaquiline-naive — primary transmitted resistance is already established, so this is not purely a consequence of prior exposure.

OFIDCOVID-19 · Meta-analysisHigh yield · ID

Effectiveness of Antivirals for COVID-19 in the Post-vaccine, Omicron Era: A Systematic Review and Meta-analysis

In a vaccinated patient in 2026, does nirmatrelvir-ritonavir still prevent hospitalization and death?

What changed43 studies (8 RCTs, 35 observational). No randomized trial in a vaccinated population shows benefit — PANORAMIC and CanTreatCOVID pooled to aOR 1.00 (95% CI .49–2.03), and RECOVERY gave death HR 1.02 (.47–2.23). Key figuresObservational data still favor treatment — hospitalization OR 0.60 (.54–.67), mortality OR 0.33 (.22–.48) — but with heavy heterogeneity and confounding by indication. The NNT arithmetic is the point: ~18 pre-Omicron in the unvaccinated, versus 76 to prevent one hospitalization and 204 one death in adults ≥75, and 364 / 1394 at ages 65–74. Molnupiravir shows no benefit for hospitalization (OR 0.83, .65–1.07) and its mortality signal disappears when higher-bias studies are excluded. Fellow implicationThe 89% relative reduction from EPIC-HR does not transfer to a vaccinated 2026 patient. Benefit is plausible but rests on observational evidence, and absolute benefit has fallen roughly an order of magnitude — which argues for targeting the genuinely high-risk rather than treating broadly, and for a stewardship position that molnupiravir has no clear remaining role.

Clinical takeaway. This is a clean illustration that NNT is a function of baseline risk, and baseline risk collapsed after vaccination and Omicron. Read it as “certainty of benefit is low, and the benefit that remains is concentrated in the oldest and most immunosuppressed” — which is the authors' own conclusion. Caveats: searches ran only to December 2024 despite 2026 publication, and screening and extraction were done by a single researcher.

OFIDCongenital CMV · Meta-analysisNotable · confirmatory

Immunoglobulin Therapy for Secondary Prevention of Congenital Cytomegalovirus Infection

After primary maternal CMV, does hyperimmune globulin prevent transmission to the fetus?

Key figures13 studies (2 RCTs); controlled analysis 6 studies, 1177 participants. Pooled RR 0.73 (95% CI .54–1.00), P = .051 — a 27% point estimate that does not reach significance, with GRADE certainty low to very low. Transmission despite HIG was 27.2% (18.8–37.7). Safety is not neutral: more preterm birth in the HIG arm of Hughes et al., and a preterm-labor/preeclampsia trend in Revello (NEJM 2014). Fellow implicationDon't offer HIG outside a protocol. The number to carry into the counselling conversation is that roughly one in four pregnancies with primary maternal CMV transmits even with treatment. Note the scope limit — this is about HIG only and says nothing about valaciclovir, the intervention that does have a positive trial.

Case reports

OFIDHIV/TB · Case reportHigh yield · ID

Emergence of Integrase Strand Transfer Inhibitor Resistance Following Treatment of Latent Tuberculosis Infection in a Patient With HIV

Your patient on a single-tablet INSTI regimen needs LTBI therapy — which rifamycin, and what happens to the ART?

The caseA 42-year-old man on BIC/FTC/TAF (baseline genotype wild-type) with a positive IGRA and normal CXR. Rifapentine was denied by insurance, forcing rifampin 600 mg daily, which required two changes: TAF→TDF and dolutegravir escalated to twice daily. An insurance lapse then interrupted both therapies for a month; he later missed evening DTG doses. He finished with persistent low-level viremia (47–161 copies/mL). The lessonA protocolized long-acting-injectable eligibility review triggered a proviral archive genotype before any regimen change → Q148H, which disqualified CAB/RPV. He was switched to darunavir/cobicistat/TAF/FTC → <20 copies/mL. Q148H confers high-level resistance to raltegravir, elvitegravir, and cabotegravir, but only ~0.8-fold reduction for DTG when it stands alone. Fellow implicationStarting rifampin converts a one-pill once-daily regimen into a multi-tablet twice-daily one — and that complexity change is itself the resistance risk. The pharmacology: DTG is chiefly UGT1A1-metabolized so doubling the dose rescues exposure; bictegravir is metabolized roughly equally by CYP3A and UGT1A1, so rifamycins induce both and it must be swapped, not escalated. Rifabutin needs no DTG adjustment; 3HP pairs with once-daily DTG but not with BIC.

Clinical takeaway. Two durable points from one patient. Choose the rifamycin and the ART pairing together, and verify rifapentine coverage before committing — the formulary decision drove the whole cascade. And treat persistent low-level viremia as a signal, not noise: it predicts non-sustained suppression on long-acting injectables and may mark archived resistance, so get the genotype before switching. Caveat: n = 1, and because proviral genotyping under-detects, a clean result would not have been reassurance.

OFIDVector-borne · Case reportHigh yield · ID

Human Exposure to Spirochete-Infected Ornithodoros turicata in a Residential Setting in the Austin, Texas Metropolitan Area

The Lyme EIA is positive but the confirmatory immunoblot is negative — what does that pattern actually mean?

The caseThe diagnosis was missed three times: day 1 fever 39.4–40 °C called viral; day 4 viral plus post-concussion; day 5 a rash prompted doxycycline for presumed rickettsiosis, which the patient self-stopped after 4 days; day 10 he was admitted with left facial paralysis attributed to murine typhus. On day 17 a live engorged tick was found on his ankle — Ornithodoros turicata carrying Borrelia turicatae, confirmed by xenodiagnosis in a naive mouse and hybrid genome sequencing. The lessonSerology was the trap. First-tier Lyme EIA positive with a negative confirmatory immunoblot is the classic cross-reaction pattern of relapsing-fever Borrelia, and the CDC immunoblot was positive for rGlpQ — the antigen that discriminates relapsing-fever group from Lyme. The 17-kDa antigen response was undetectable, consistent with a day-11 draw: serology can be falsely negative <14 days from onset. Fellow implicationIn central Texas, recurrent fevers ± cranial neuropathy should put soft-tick relapsing fever on the differential, and a discordant two-tier Lyme result should open the question rather than close it. Soft ticks feed for minutes, painlessly — no recalled bite has essentially no negative predictive value — and this exposure happened inside an urban apartment, not a rural cabin.

Clinical takeaway. Facial palsy occurs in both Lyme and relapsing fever, so it does not discriminate — the serology does, and only if you order the right test. When serology is equivocal early, microscopy, PCR, or culture will settle it. Note the ecology too: Ornithodoros live >10 years, feed >15 times, and transmit transovarially, so a household or park focus persists once established. Causality here is inferential (n = 1, no convalescent serum), and the doxycycline frequency is never stated — don't infer a regimen from this paper.

OFIDTB meningitis · Case + PKNotable · n = 1

CSF Concentrations of Bedaquiline, Pretomanid, and Linezolid During Curative Treatment of Fluoroquinolone-Resistant Pre-XDR Tuberculous Meningitis

Do BPaL drugs actually reach the CSF — and how would you know if you measured it wrong?

The caseA 35-year-old HIV-negative woman with pre-XDR TB meningitis (gyrA D94G), diagnosed on retained products of conception after sputum and BAL culture and Xpert Ultra were all negative. Cured with BPaL + clofazimine + cycloserine over 43 weeks (29 weeks of pretomanid), plus VP shunts, steroids, and thalidomide — no recurrence at 40 months, but severe residual disability. Key figuresCSF:plasma unbound AUC₂₄ 0.99 for linezolid and 1.48 for pretomanid. Bedaquiline concentrations were higher at week 21 than at weeks 9–12 — CNS accumulation over months. Linezolid AUC₂₄/MIC was 171 at an MIC of 0.5 mg/L but only 85.5 at 1.0 mg/L, below the target of 125. Fellow implicationTwo durable pharmacology points. For highly protein-bound drugs, raw CSF:plasma ratios systematically understate CNS penetration unless normalized for CSF's ~10-fold lower protein — the same reason clofazimine's penetration was historically underestimated. And target attainment fails as MICs rise, which is the practical argument for TDM in CNS TB and a specific caution for lineage 1 strains. Note: the abstract's claim that therapeutic CNS bedaquiline concentrations were achieved sits in tension with the Results, where unbound bedaquiline stayed below the critical concentration throughout.

Also new this week. Teicoplanin vs standard of care — in 102 propensity-matched pairs, cure 78.4% vs 72.5% (P = .31) and 90-day mortality HR 0.89; the transferable pearl is that omitting the loading dose predicted failure (OR 3.50) given teicoplanin's 40–100 h half-life (ofag462). QuantiFERON IFN-γ and TB progression — TB2 reached AUROC 0.84 with 80%/78% sensitivity and specificity, clearing WHO's target product profile, but PPV ≈ 1% — a trial-enrichment tool, not a clinical test (ofag437). Point-of-care pharyngitis panel — ≥1 pathogen in 74.5% of 200 urgent-care patients; antibiotics were prescribed in 25.5% when the panel was completely negative vs 7.0% when only a virus was found, so a positive viral result reassures more than a negative panel (ofag441). Deploying new gonorrhea drugs — a transmission model found that with low background resistance, combination therapy preserved both drugs best, against the reflexive “reserve the new agent” instinct; ceftriaxone resistance is ~30% in parts of East Asia vs sporadic elsewhere (ofag447). Deprivation and antibiotic concordance — across 1.37 million ED encounters, antibiotic-decision quality mediated only ~1–11% of deprivation-associated outcome gaps; overuse (9.5%) and underuse (7.4%) were nearly equally common (ofag455). Directly observed therapy — in 278 007 Brazilian notifications, DOT was associated with treatment success (OR 1.46) via reduced loss to follow-up (OR 0.49), with no effect on death (OR 1.02) (ofag434). Norovirus and gastroenteritis hospitalization — cardiovascular disease carried the highest adjusted risk (aRR 1.67); the transferable finding is that relative risk from comorbidity is larger in adults 18–64 (aRR 4.19) than ≥65 (1.90) despite lower absolute risk (ofag449). Persistent 9vHPV in MSM — commercial sex (HR 3.50) and versatile (2.86) or receptive (1.91) roles predicted persistent anal infection; prediction model AUC 0.727, not externally validated (ofag432). HBV coinfection by birth cohort — HBsAg 6.7% overall; 2.6% in those born 1980–1999 vs 7.1% earlier, but still well above the general population, so birth year is no reason to skip screening or vaccination (ofag400).

Summaries are original paraphrases prepared for educational use; figures are drawn from the published full text. Read the full articles at the DOIs linked above — Clinical Infectious Diseases requires a subscription; Open Forum Infectious Diseases is open access. Pull window July 22–31, 2026 (19 articles archived: 2 CID, 17 OFID; the OFID set includes a backlog tail with online dates back to 7 July). Citations: Clin Infect Dis 2026 (advance articles) and Open Forum Infect Dis 2026;13(7). DOIs: 10.1093/cid/ciag452, 10.1093/cid/ciag440, 10.1093/ofid/ofag339, 10.1093/ofid/ofag445, 10.1093/ofid/ofag449, 10.1093/ofid/ofag462, 10.1093/ofid/ofag428, 10.1093/ofid/ofag437, 10.1093/ofid/ofag455, 10.1093/ofid/ofag448, 10.1093/ofid/ofag441, 10.1093/ofid/ofag405, 10.1093/ofid/ofag447, 10.1093/ofid/ofag432, 10.1093/ofid/ofag438, 10.1093/ofid/ofag434, 10.1093/ofid/ofag431, 10.1093/ofid/ofag400, 10.1093/ofid/ofag424.