Results of the First Folate Receptor Alpha Testing Trial by the German Quality Assurance Initiative in Pathology (QuIP®)

Abstract

Background: Folate receptor alpha (FRα) is a glycosylphosphatidylinositol-anchored membrane protein encoded by the FOLR1 gene. Its overexpression in various cancers, including ovarian carcinoma, makes it a promising target for antibody-drug conjugates (ADC). Mirvetuximab soravtansine-gynx, an FRα-targeting ADC, has been approved by the FDA and EMA for the treatment of FRα-positive, platinum-resistant ovarian cancer. In the United States, patient selection is tied to the VENTANA FOLR1 (FOLR1-2.1) RxDx Assay, an immunohistochemical (IHC) test that identifies tumors with ≥75% moderate-to-strong membrane staining. However, in the European Union, no specific IHC test is mandated, and alternative antibodies are frequently used in routine pathology, necessitating validation of their diagnostic performance. Methods and Results: We report the results of the first interlaboratory proficiency trial on FRα testing conducted by the German Quality Assurance Initiative in Pathology (QuIP^®). Sixty-eight pathology institutes participated across internal and open trials using a variety of antibodies and staining platforms. The VENTANA FOLR1 RxDx Assay demonstrated the highest reliability, with 83% of participating laboratories achieving a successful result. In contrast, alternative clones such as BN3.2 (Leica/Novocastra) and EPR20277 (Abcam) showed substantially weaker staining intensity, lower concordance with reference values, and success rates of only 22-25%, while other antibodies failed entirely. Problem analysis revealed that failures with the VENTANA FOLR1 (FOLR1-2.1) RxDx Assay were mainly due to interpretative challenges, whereas weak staining was the predominant issue with alternative clones. Participation in a preparatory online seminar improved pass rates, underscoring the importance of training. Conclusions: These findings highlight the critical importance of standardized, validated assays for FRα detection to ensure accurate patient selection for targeted therapies. The study emphasizes the need for further optimization of alternative antibodies before clinical implementation.

Keywords: antibody-drug conjugates (ADC); biomarker validation; folate receptor alpha (FRα); immunohistochemistry (IHC); proficiency testing.

Figure 1

Flowchart depicting the time course of the internal and open proficiency trial.

Figure 2

(a) Results of the Internal proficiency test for cases that were later also used in the open proficiency trial. Cases 03b and 04b represent replacement cases distributed in Split 2 (disregarded for scoring of the internal proficiency test). (b) Overview of antibodies utilized by the Lead and Panel Institute participants. TP, true positive; FN, false negative; TN, true negative; FP, false positive.

Figure 3

Case-based results for (a) all submissions of participants, (b) the submissions based on the FOLR1 RxDx Assay and (c) the submissions using BN3.2 or EPR20277 antibodies (the latter was utilized by only one participant). The tables show the number of participants who selected the corresponding answer. The target values are indicated in bold. The answers rated as correct are shaded in gray. In parentheses, the number of participants is given who provided responses deviating from the target value, and who, after the review process, were awarded points (blue text) or not awarded points (red text).

Figure 4

Case 1 assessed as positive by the lead panel institutes. Immunohistochemical stainings submitted by participants with different antibodies. The BN3.2 antibody clones as well as the EPR20277 Abcam antibody resulted in weak staining intensities that led to a false negative classification in the presented cases. Magnification 5x (overview) and 40x.

Figure 5

(a) Interpretation issue, where areas with cytoplasmic staining in tumor cells were wrongly counted as positive. Magnification left panel 5x, right panel 40x. (b) Example of an interpretation issue in case 9 where dot-like expression pattern (which is supposed to be counted as positive) was disregarded. Magnification both panels 40x (c) Problem analysis for all participants (left chart) and the participants that failed the proficiency test (right chart) with indication of underlying problem.