We read with great interest the study by Williams et al.1 on the association among HtrA1 concentrations, AMD, and genetic risk at the Chr10 locus as determined using an in-house luminescent ELISA assay (SCTM). This method differs from previously published colorimetric assays (MBS1 and MBS2).2,3 Although the authors used reliable methods to validate four ELISA kits—two in-house luminescent (SCTM and R&M) and two commercial colorimetric (MBS1 and MBS2)—several concerns arise regarding their approach.
Detection range: The linear sensitivity ranges of MBS1 (2–20 ng/mL) and MBS2 (0.2–2 ng/mL) do not overlap, making direct correlation challenging, as the same sample cannot be used for both kits. The dilution methods (MBS1, 1/100; MBS2, 1/10) further complicate this, resulting in different detection ranges (MBS1, 200–2000 ng/mL; MBS2, 2–20 ng/mL). This issue extends to other kits, as well, with many results falling outside their detection ranges.
Standard curves: The standard curves for MBS1 and MBS2 follow a log–log model, yet the authors applied a linear model for all kits, potentially compromising accuracy.
Detection mode: Colorimetric and chemiluminescence ELISAs have different detection modes, measuring optical density and relative light units, respectively.4 Both MBS1 and MBS2 diluent solutions significantly reduced the relative light unit of the SCTM standard compared to the SCTM diluent solution. The stabilizers and preservatives in the colorimetric assay may affect the luminescent reaction, raising questions about the validity of cross-standard curves.
Despite claims of high specificity and sensitivity for the SCTM ELISA, its consistency and detection range are ambiguous. For example, 57% of HtrA1 WT standard inputs had intra-assay CVs exceeding 10%, and 29% exceeded 40%, indicating inconsistency.5,6 Additionally, the HtrA1 levels measured by SCTM ELISA in 65 serum samples (200–600 ng/mL) were significantly higher than those in 248 serum samples assessing AMD risk (32.4 ng/mL for males; 32.5 ng/mL for females). This discrepancy, along with dilution concerns (10- to 25-fold dilution), suggests that many samples fell outside the optimal linear range of the assay for serum (1.8–11.45 ng/mL).
References
- 1. Williams BL, Zouache MA, Seager NA, et al.. Levels of the HtrA1 protein in serum and vitreous humor are independent of genetic risk for age-related macular degeneration at the 10q26 locus. Invest Ophthalmol Vis Sci. 2024; 65(4): 34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Tosi GM, Caldi E, Neri G, et al.. HTRA1 and TGF-β1 concentrations in the aqueous humor of patients with neovascular age-related macular degeneration. Invest Ophthalmol Vis Sci. 2017; 58(1): 162–167. [DOI] [PubMed] [Google Scholar]
- 3. Pan Y, Iejima D, Nakayama M, et al.. Binding of Gtf2i-β/δ transcription factors to the ARMS2 gene leads to increased circulating HTRA1 in AMD patients and in vitro. J Biol Chem. 2021; 296: 100456. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Cox KL, Devanarayan V, Kriauciunas A, et al.. Immunoassay methods. In: Markossian S, Grossman A, Arkin M, et al., eds. Assay Guidance Manual [Internet]. Bethesda, MD: Eli Lilly & Company, National Center for Advancing Translational Sciences; 2004. [PubMed] [Google Scholar]
- 5. Hurley IP, Coleman RC, Ireland HE, Williams JHH.. Measurement of bovine IgG by indirect competitive ELISA as a means of detecting milk adulteration. J Dairy Sci. 2004; 87(3): 543–549. [DOI] [PubMed] [Google Scholar]
- 6. Reed GF, Lynn F, Meade BD.. Use of coefficient of variation in assessing variability of quantitative assays. Clin Vaccine Immunol. 2002; 9(6): 1235–1239. [DOI] [PMC free article] [PubMed] [Google Scholar]