Siglec-15 has been proposed as a non-classical immune checkpoint and potential therapeutic target in solid tumors. This systematic review and meta-analysis pooled observational cohort data to assess the association between tumoral Siglec-15 protein expression and survival outcomes. Twenty-seven datasets with 4,075 patients were included. High tumoral Siglec-15 expression correlated with worse overall survival (OS) and worse disease-free/recurrence-free survival (DFS/RFS), but not with progression-free survival (PFS). An inverse association with PD-L1 tumor positivity was observed. The authors conclude that the prognostic value of Siglec-15 is cancer-type dependent and recommend further prospective studies and standardized assessment strategies.
Immune checkpoint blockade targeting PD-1/PD-L1 has improved outcomes in several malignancies, yet many patients do not benefit. Siglec-15, a sialic acid-binding immunoglobulin-like lectin, is mainly expressed in osteoclasts and certain myeloid populations physiologically, but aberrant tumoral expression has been reported across multiple solid cancers. Experimental data indicate Siglec-15 can suppress T-cell activation and foster an immunosuppressive tumor microenvironment, while also promoting tumor-cell intrinsic processes linked to progression. Prior clinical studies reported inconsistent prognostic associations across tumor types. Given emerging evidence and newly published studies, an updated synthesis was performed to clarify the prognostic role of tumoral Siglec-15 and its relationship with PD-L1 and clinicopathological features.
This review followed PRISMA guidance. PubMed, Web of Science, and EMBASE were searched from inception to March 30, 2026, using terms for Siglec-15 and neoplasms. Inclusion criteria required histologically confirmed solid tumors, tissue-based measurement of Siglec-15 protein in tumor cells, comparison of high versus low/negative tumoral expression, and reporting of survival outcomes or extractable survival data. Exclusions included hematological malignancies, studies using only public databases without tissue cohorts, transcriptional-level assessments, non-tumor cellular measurements (serum, stromal cells, TAMs), and missing hazard ratios or Kaplan–Meier data.
Data extraction captured study-level details (author, year, country, tumor type, sample size), detection and scoring methods for Siglec-15 (IHC, QIF, IF; IRS/H-score/count/percentage/intensity), cutoff determination (median, fixed, optimal), survival outcomes and HR types (univariable vs. multivariable), and clinicopathological variables. When HRs were absent, they were estimated from Kaplan–Meier curves. Inter-rater agreement for extracted HRs was excellent (ICC = 0.983).
Study quality was assessed with the Newcastle–Ottawa Scale. Statistical pooling used random-effects models for HRs and ORs with 95% confidence intervals. Heterogeneity was described with I2. Prespecified subgroup analyses included country, sample size, scoring method, cutoff method, HR type, data source, study quality, and cancer type. Meta-regression and sensitivity analyses were performed when appropriate; publication bias was assessed by funnel plots and Egger’s test when ≥10 studies were available.
After screening 276 unique records, 25 studies reporting 27 eligible datasets and 4,075 patients were included. Studies spanned multiple tumor types (colorectal, gastric, lung, pancreatic, breast, osteosarcoma, glioma, renal cell carcinoma, nasopharyngeal carcinoma, thyroid, soft tissue sarcoma, bladder, and others) and varied in detection/scoring approaches (IHC, multiplex IHC, QIF, IF) and cutoff definitions.
Pooled analyses showed that high tumoral Siglec-15 expression was significantly associated with poorer OS (HR = 1.53, 95% CI 1.25–1.86, P < 0.001). High expression was also associated with worse DFS/RFS (HR = 1.30, 95% CI 1.09–1.55, P = 0.004). No significant association was observed for PFS (HR = 1.03, 95% CI 0.67–1.56, P = 0.903). For disease-specific survival (DSS), a marginal association suggested improved DSS with high Siglec-15 (HR = 0.66, 95% CI 0.43–1.00, P = 0.050), but this result derived from only three studies and was noted to require cautious interpretation.
Tumor-specific analyses revealed significantly worse OS associated with high tumoral Siglec-15 in colorectal cancer, gastric cancer, and osteosarcoma. Breast cancer datasets showed a non-significant association in the opposite direction (HR = 0.72, 95% CI 0.33–1.56). Across pooled clinicopathological comparisons, Siglec-15 expression did not show consistent associations with most variables such as gender, age, tumor differentiation, tumor size, lymph node metastasis, distant metastasis, or TNM stage.
Importantly, tumoral Siglec-15 expression was inversely associated with PD-L1 positivity in tumor cells (OR = 0.40, 95% CI 0.23–0.72, P = 0.002), supporting a largely non-overlapping expression pattern between Siglec-15 and PD-L1.
This comprehensive synthesis of tissue-based studies indicates that elevated tumoral Siglec-15 protein expression is associated with worse OS and DFS/RFS across solid tumors in pooled analyses, supporting its potential role as a prognostic biomarker and therapeutic target. The inverse relationship with PD-L1 aligns with preclinical and clinical observations that Siglec-15 and PD-L1 often show mutually exclusive tumor expression, which may have therapeutic implications for patients with PD-L1–low tumors.
Heterogeneity across studies is notable and stems from differences in tumor types, IHC/QIF/IF platforms, scoring systems (IRS, H-score, percentage/count, intensity), and cutoff definitions (median, fixed, optimal). These methodological differences limit direct comparability and underline the need for prospective studies using standardized tissue assays with fixed cutoff values and simplified scoring to facilitate clinical translation.
The marginal finding of improved DSS with high Siglec-15 must be interpreted with caution because it was based on three studies only. Likewise, tumor-specific effects varied, with some cancers (colorectal, gastric, osteosarcoma) showing consistent adverse associations while others did not.
Elevated tumoral Siglec-15 expression is associated with unfavorable overall and disease-free/recurrence-free survival in pooled analyses of solid tumors and is inversely associated with tumor PD-L1 expression. The prognostic impact appears highly cancer-type dependent. Further prospective validation and harmonization of detection and scoring methods—preferably with fixed cutoffs and simplified scoring—are needed before routine clinical application as a prognostic biomarker or to guide Siglec-15–targeted therapies.