Oral metal immunotoxicity in rodents: a recoverability-aware systematic review and within-study quantitative reanalysis of splenic SRBC-TDAR.

Abstract

Introduction: Public toxicology databases are increasingly used for evidence synthesis, but their value depends not only on the availability of studies but also on whether endpoint-level data can be recovered for reuse. Using the Chemical Effects in Biological Systems (CEBS) database as the primary source, we reviewed oral heavy-metal immunotoxicity in rodents.

Methods: We assessed whether splenic sheep red blood cell (SRBC) T-dependent antibody response data, measured as antibody-forming/plaque-forming cells (TDAR AFC/PFC), were sufficiently recoverable for structured synthesis and quantitative reanalysis. Recoverability was defined as the availability of endpoint information needed for reuse, including group size, mean, and variance when quantitative reanalysis was attempted.

Results: Fifteen records were screened, and 12 records were retained across vanadium, chromium, tungsten, cadmium, and lead; one of these 12 records contained separate lead and cadmium exposure arms. Three retained records were CEBS package-level studies, yielding 30 module-level entries spanning primary AFC, serum anti-SRBC IgM, natural killer (NK) cell function, splenic phenotype, hematology, body weight, and organ weights. Only one retained study, which examined sodium metavanadate in female B6C3F1/N mice exposed via drinking water for 28 days, provided a fully recoverable primary dataset with extractable group sizes, means, and variances, enabling quantitative reanalysis of five AFC contrasts from 31.3 to 500 ppm. AFC/10⁶ spleen cells shifted from +5.2% at 31.3 ppm to -29.6% at 500 ppm (lnRR -0.351, 95% CI -0.634 to -0.067), and benchmark-like thresholds for 10%, 20%, and 25% AFC decline were estimated at approximately 84, 184, and 283 ppm, respectively. The same package revealed marked assay- and normalization-dependent divergence, with plaque-based AFC declining, ELISpot-based IgM AFC increasing, and serum IgM and NK responses showing limited concordance.

Discussion: Overall, CEBS module-level data supported a stronger endpoint-centered synthesis than narrative review alone. However, incomplete endpoint recoverability, rather than study scarcity, remained the main barrier to broader quantitative integration across oral metal studies.