HIV-1 and TB DNA linked in spinal lesions

When tuberculosis reaches the spine it can cause serious disease, and scientists have long suspected that HIV-1 infection can make TB worse — and vice versa — by helping each pathogen replicate or hide in tissues. To probe that interaction directly in diseased tissue, a team led by Anna K. Coussens conducted a prospective study of adults being investigated for spinal tuberculosis (STB) in South Africa. The study enrolled 25 adults, 13 (52%) of whom were people living with HIV-1 (PLWH) and were on antiretroviral treatment. From those patients the researchers collected 25 open surgery or CT-guided biopsies and divided them into 93 segments to allow detailed, within-sample analysis. The goal was to look for DNA from both pathogens in the same tissue locations, and to compare pathogen abundance with the number of human cells present and with measures of HIV activity in the blood. By focusing on actual spinal lesions, the team aimed to move beyond blood tests and understand how the two infections coexist in the places where disease is happening.

The researchers used droplet digital PCR (ddPCR) to detect and quantify Mycobacterium tuberculosis complex (MTBC) targets (rpoB, IS6110), HIV-1 targets (pol, gag) and the human gene RPP30 as a measure of cell number. ddPCR sensitivity for Mtb was validated against Xpert-Ultra and culture. ddPCR detected MTBC DNA in biopsies from 10/10 (100%) culture-confirmed STB, 5/6 (83%) Xpert-Ultra-confirmed STB, and 4/9 (44%) biopsies clinically diagnosed Not STB (all four of those had previous pulmonary TB). Detected MTBC ranged from 8 to 59,144 rpoB copies per biopsy. When normalized to human cells, rpoB copies per million human cells were higher in biopsies from PLWH (p=0.0096) and positively correlated with matched-segment HIV-1 pol copies per million cells (r=0.40; p=0.0003), but did not correlate with peripheral viral load (VL). HIV-1 DNA was detected in all PLWH biopsies, including four with undetectable VL. HIV-1 pol copies per million cells were higher in segments where MTBC DNA was co-detected (p=0.011) and pol copy numbers also correlated strongly with VL (r=0.91; p=0.0003).

These results show a reciprocal relationship between Mtb and HIV-1 abundance in spinal tissue, suggesting that the two pathogens can coexist and potentially promote each other locally in lesions. The presence of HIV-1 DNA in all PLWH biopsies — including samples from patients with undetectable blood VL — highlights how tissue-level reservoirs can persist despite antiretroviral treatment and undetectable plasma virus. The strong correlations between local HIV-1 pol copies and MTBC rpoB burden in the same tissue segments support the idea that co-infected microenvironments could foster reservoir persistence or even expand both infections. The study demonstrates the value of sensitive tools like droplet digital PCR (ddPCR) for mapping pathogen DNA in biopsies and argues for future work that focuses on TB–HIV co-infected tissue segments to understand how the immune microenvironment affects long-term persistence and interactions of these pathogens.