Kaposi’s sarcoma (KS) is a mesenchymal tumour that requires infection with Human Herpesvirus 8 (HHV-8) for pathogenesis. Uncontrolled HIV infection is a major risk factor for KS, but forms of KS occurring with well-controlled HIV can resemble classic KS. This study aimed to compare clinical features and the composition of the tumour microenvironment (TME) between HIV-associated KS in patients with well-controlled infection and non-HIV-associated (classic) KS without clinically evident immunosuppression.
The primary objectives were to characterise transcriptomic and immune-cell differences in archival tumour samples and to investigate relationships between intra-tumoural HHV-8 gene expression and immune markers.
Patients were selected from a prospectively maintained dataset at the National Centre for HIV malignancies, Chelsea and Westminster Hospital. The study included 40 consecutive patients with histology-proven KS diagnosed between January 2008 and February 2015 and with available tissue samples. Cases with uncontrolled HIV (HIV RNA >200 copies/ml or CD4 <200 cells/mm3) or other causes of immunosuppression were excluded. The non-HIV cohort comprised classic KS cases only.
Ethics and tissue governance were addressed through the National Centre for HIV malignancies and the Imperial College Healthcare Tissue Bank under existing approvals. Informed consent was obtained from participants.
After pathology quality checks, RNA was extracted from three 10 μm scrolls from formalin-fixed paraffin-embedded (FFPE) tumour blocks using a standard kit. Gene expression profiling used the NanoString nCounter PanCancer IO 360 Panel. A custom CodeSet of 10 HHV-8 genes was added to the panel to permit intra-tumour viral gene expression assessment. Further details of data processing and normalisation are reported in the study supplementary methods.
Multiplex immunofluorescence (mIF) and immunohistochemistry were performed on 12 FFPE tissue blocks following protocols developed by the ICR/RMH Integrated Pathology Unit. Two multiplex panels, each comprising six biomarkers plus DAPI, were developed and applied using the PhenoImager HT system. Antibody staining was automated on the BOND RX 3.0 Fully Automated Slide Stainer. The mIF approach aimed to characterise immune cell populations and viral protein expression within the TME. Specific panel marker lists and staining details are provided in the supplementary tables referenced by the source.
Plasma HHV-8 DNA was measured at the time of tissue biopsy using a quantitative nested PCR method targeting the ORF26 gene (minor capsid protein) of KSHV. The source reports plasma HHV-8 results for baseline comparisons between cohorts.
The cohort comprised 21 HIV-associated KS and 19 non-HIV-associated (classic) KS patients. Most patients were male (37/40). All HIV-associated patients had been on cART for at least three months at sampling and met the criteria of HIV-RNA <200 copies/ml and CD4 >200 cells/mm3.
Twenty-four patients had prior loco-regional therapy, with excision most common, and 22 had received at least one line of systemic therapy. A higher proportion of classic KS cases received loco-regional treatments (73.7% vs 43.5%, p = 0.049). The proportion receiving systemic therapy was not significantly different between groups (62.5% vs 42.1%, p = 0.183). Doxorubicin-based regimens were used as first-line systemic therapy in most treated patients.
Patients with HIV-associated KS were younger at first diagnosis, had higher peripheral blood CD8 counts and lower plasma HHV-8 levels compared with non-HIV-associated cases. Other baseline features were largely balanced between groups.
Targeted transcriptomic analysis revealed distinct TME features between groups. The TME of non-HIV-associated (classic) KS demonstrated transcriptional upregulation of pathways linked to adaptive and innate immunity and angiogenesis relative to the HIV-associated KS cohort. Conversely, HIV-associated KS tumours were associated with a lower degree of inflammation and evidence of more pronounced epithelial-to-mesenchymal transition.
Multiplex immunofluorescence showed reduced infiltration of activated CD4 cells in HIV-associated KS compared with classic KS. The source reports these immune infiltration differences but refers readers to supplementary materials for full marker-level data and analytic details.
Across both cohorts, intra-tumoural HHV-8 gene expression correlated positively with markers of activated CD4, CD8 and NK cells, as well as with immune checkpoint gene expression. The authors interpret this association as evidence that intra-tumoral viral gene expression is a driver of TME composition in KS.
After a median follow-up of 15.3 months (95% CI: 12.0–30.3), median progression-free survival (PFS) for the whole cohort was 17.3 months (95% CI: 8.0–NA). PFS did not differ significantly by HIV status (hazard ratio 1.3; 95% CI: 0.50–3.48; p = 0.6). The source reports these outcome data; detailed subgroup survival curves and supplementary figures are referenced in the original article.
In patients with well-controlled HIV infection, KS tumours show a lower degree of inflammation and features consistent with epithelial-to-mesenchymal transition compared with non-HIV-associated (classic) KS. Importantly, intratumoural HHV-8 gene expression was associated with immune-cell infiltration and checkpoint gene expression in both groups, suggesting viral transcriptional activity influences TME composition. These observations support further investigation of immune-targeted therapies in KS and highlight the relevance of viral gene expression when considering the immunobiology and treatment of Kaposi sarcoma.
Note: The source reports additional methodological and analytic details in supplementary methods and tables; those specifics are cited but not reproduced here.