Folate receptor-α (FRα, encoded by the FOLR1 gene) is overexpressed in several solid tumor types, including epithelial ovarian cancer (EOC), making it an attractive biomarker and target for FRα-based therapy in ovarian cancer.
To describe the development, analytic verification, and clinical performance of the VENTANA FOLR1 Assay (Ventana Medical Systems Inc) in EOC.
We used industry standard studies to establish the analytic verification of the VENTANA FOLR1 Assay. Furthermore, the VENTANA FOLR1 Assay was used in the ImmunoGen Inc–sponsored SORAYA study to select patients for treatment with mirvetuximab soravtansine (MIRV) in platinum-resistant EOC.
The VENTANA FOLR1 Assay is highly reproducible, demonstrated by a greater than 98% overall percent agreement (OPA) for repeatability and intermediate precision studies, greater than 93% OPA for interreader and greater than 96% for intrareader studies, and greater than 90% OPA across all observations in the interlaboratory reproducibility study. The performance of the VENTANA FOLR1 Assay in the SORAYA study was evaluated by the overall staining acceptability rate, which was calculated using the number of patient specimens that were tested with the VENTANA FOLR1 Assay that had an evaluable result. In the SORAYA trial, data in patients who received MIRV demonstrated clinically meaningful efficacy, and the overall staining acceptability rate of the assay was 98.4%, demonstrating that the VENTANA FOLR1 Assay is safe and effective for selecting patients who may benefit from MIRV. Together, these data showed that the assay is highly reliable, consistently producing evaluable results in the clinical setting.
The VENTANA FOLR1 Assay is a robust and reproducible assay for detecting FRα expression and identifying a patient population that derived clinically meaningful benefit from MIRV in the SORAYA study.
Epithelial ovarian carcinoma (EOC) accounts for approximately 90% of all cases of ovarian cancer, which remains a leading cause of gynecologic cancer mortality in women worldwide.1,2 Late-stage diagnosis of EOC is an important contributing factor to its lethality, as evidenced by a 5-year survival rate of 29% for women with advanced disease.3 The foundation of standard-of-care treatment for newly diagnosed EOC involves primary debulking surgery followed by combination chemotherapy using a platinum-based regimen, most typically the platinum/paclitaxel doublet. EOC is highly chemosensitive, and most women achieve remission with front-line therapy; however, most patients with advanced EOC ultimately relapse.4 Determination of therapeutic options for patients with recurrent disease has traditionally relied on the platinum-free interval (PFI), defined as the time since stopping platinum based-chemotherapy to disease progression; patients may be broadly classified as either platinum-sensitive (PFI >6 months) or platinum-resistant (PFI ≤6 months). Patients with platinum-sensitive disease have a high likelihood of responding to additional platinum-based chemotherapy—unfortunately, almost all will eventually develop acquired (secondary) platinum resistance.5 Patients with platinum-resistant recurrences typically receive non–platinum-based, single-agent chemotherapy (eg, paclitaxel, gemcitabine, carboplatin, or poly ADP-ribose polymerase [PARP] inhibitors).6
More recently, the incorporation of molecularly targeted agents into treatment paradigms for patients with EOC has improved outcomes and resulted in an increased prevalence of patients living with this disease. For example, use of the anti–vascular endothelial growth factor (VEGF) monoclonal antibody bevacizumab alongside traditional chemotherapy followed by bevacizumab maintenance has resulted in meaningful improvements in survival for women in the first line and recurrent, platinum-sensitive EOC settings.7,8 Similarly, the integration of PARP inhibitors into standard-of-care therapy, including escalation into earlier lines of treatment, has dramatically altered the therapeutic landscape for EOC.9 Despite this welcome progress, there remains an urgent need for new targets and therapies to treat EOC, particularly in light of the high rate of recurrence and resistance to primary therapy.
Folate receptor-α (FRα) is a cell surface glycoprotein encoded by the FOLR1 gene that binds and internalizes reduced folates and folic acid.10,11 FRα aids in the cellular uptake of folate, which is an essential cofactor in 1-carbon metabolism, a metabolic pathway required for DNA and RNA synthesis that in turn supports cell growth and proliferation.12 In line with this role, FRα is highly expressed during periods of rapid cell growth and division, such as pregnancy and embryogenesis. However, this receptor typically shows a restricted distribution pattern in normal tissues, with low-level expression that is limited to a variety of polarized epithelia, such as those seen in the choroid plexus, thyroid, salivary glands, kidney, lung, and bladder.13–15 In contrast, FRα upregulation and overexpression have been reported in several solid tumor types, including EOC, triple-negative breast cancer, endometrial cancer, and non–small cell lung cancer.14,16,17 Mechanistically, overexpression of FRα has been suggested to support the increased metabolic demands of tumor cells, although a defined function in carcinogenesis is yet to be confirmed.18 FRα has now emerged as an attractive candidate for novel molecularly targeted approaches designed to exploit this differential distribution pattern between normal and malignant tissues as a potential avenue for therapeutic intervention. In this regard, several FRα-targeting approaches are being developed, including small molecules, monoclonal antibodies, antibody-drug conjugates (ADCs), folate-drug conjugates, and immunotherapy.18
For EOC specifically, a number of FRα-targeting agents are under clinical investigation. Monoclonal antibodies, such as MOv18 and farletuzumab, have both previously demonstrated limited single-agent activity in patients with recurrent, platinum-sensitive EOC.18–20 Newer therapies include ADCs, which use a highly specific monoclonal antibody that is directed toward a tumor-specific antigen to deliver a cytotoxic payload directly to tumor sites.21 High-affinity binding of the ADC to its target on the tumor cell results in internalization of the ADC complex, which in turn triggers the release and functional activation of the cytotoxic payload within the tumor cells themselves. Capitalizing on the specificity of earlier FRα-targeting antibodies, the ADCs MORAb-202 and mirvetuximab soratansine (MIRV) are both being investigated in phase 1 to 3 clinical trials (NCT05613088, NCT04300556, and NCT05041257). Folate-drug conjugates are also designed to provide site-specific delivery of cytotoxics to tumors, but unlike ADCs, they instead attach cytotoxic agents directly to folate itself in order to deliver the payload to FRα-expressing tumors. The folate-drug conjugate EC1456 is currently being investigated in a phase 1 trial in patients with solid tumors, including recurrent EOC (NCT01999738).
Importantly, the cytotoxic activity of FRα-targeting agents, specifically MIRV and MORAb-202, has been shown to correlate with FRα expression preclinically, suggesting that determination of patient tumor FRα levels may be important for the success of experimental therapies in clinical trials.22,23 Several past and ongoing clinical trials of FRα-targeted therapies have implemented routine immunohistochemistry (IHC) assays (NCT02546921, NCT04300556, NCT03748186, NCT05041257, and NCT05445778) to determine FRα protein expression for selection and/or stratification of the EOC patient trial population based on expression level.
Here we describe the analytic performance and scoring method for the VENTANA FOLR1 (FOLR1-2.1) RxDx Assay (hereafter, the VENTANA FOLR1 Assay, Ventana Medical Systems Inc) and its clinical use to determine patient FRα expression levels for treatment with MIRV, both of which received recent US Food and Drug Administration (FDA) approval in the United States. MIRV is an ADC that is composed of a humanized anti-FRα monoclonal antibody to which the potent microtubule-destabilizing compound DM4 is attached via a cleavable linker.23 Preclinical studies demonstrated antitumor activity of MIRV as both a single agent and when combined with other agents in ovarian cancer cell and mouse models.23,24 Accordingly, MIRV has now undergone advanced clinical evaluation, demonstrating promising clinical activity in the setting of platinum-resistant EOC.25
The SORAYA study (NCT04296890), which enrolled 106 patients, was a pivotal single-arm study to determine the efficacy and safety of MIRV in patients with platinum-resistant, advanced high-grade epithelial ovarian, primary peritoneal, or fallopian tube cancers (hereafter classified as EOC in relation to the data reported in this publication) with high FRα expression, determined using the VENTANA FOLR1 Assay. Results from the SORAYA study demonstrated meaningful antitumor activity in patients treated with MIRV, including an objective response rate (ORR) of 32.4% (95% CI, 23.6%–42.2%) and a median duration of response of 6.9 months (95% CI, 5.6–9.7 months).26
The VENTANA FOLR1 Assay was developed for use as a companion diagnostic to aid pathologists in identifying FRα-positive EOC patients. It was analytically validated using industry standard verification studies and clinically validated in the SORAYA trial as a reproducible IHC assay for detecting FRα expression in EOC.26 We examined the repeatability and reproducibility of the assay, interreader and intrareader precision, and interlaboratory reproducibility, as well as the staining performance of the assay in the SORAYA trial.
MATERIALS AND METHODS
Tissue Specimens, Tissue Processing, and IHC Staining
For all analytic performance studies described, EOC tumor specimens were sourced from commercial vendors as formalin-fixed, paraffin-embedded tissues. All specimens were provided absent of identifying data, except for the histologic diagnosis. Resections of EOC were used in analytic verification of the VENTANA FOLR1 Assay. Formalin-fixed, paraffin-embedded blocks of EOC were sectioned at ∼4 μm and mounted onto positively charged glass slides and stored at room temperature. All tissue sections were stained with the VENTANA FOLR1 Assay on the BenchMark ULTRA staining platform using a standardized staining procedure that encompassed steps for deparaffinization, antigen retrieval, primary antibody, and detection and counterstaining with hematoxylin and bluing reagent (Supplemental Table 1, see supplemental digital content containing 3 Tables, 2 Figures, and method details for the Western blots, at https://meridian.allenpress.com/aplm in the November 2024 table of contents). All staining used the VENTANA FOLR1 Assay, OptiView DAB IHC Detection Kit, and VENTANA ancillary reagents (buffers, hematoxylin, bluing reagent). All samples were reviewed by a pathologist to confirm the histologic diagnosis and ensure the presence of morphologically intact tumor.
Tissue specimens used to determine the clinical performance of the VENTANA FOLR1 Assay (staining acceptability, acceptable background, and acceptable morphology rates) were from the ImmunoGen-sponsored SORAYA study, described below.
Assay Development and Optimization
The anti-FOLR1 (FOLR1-2.1) mouse monoclonal primary antibody was optimized by Ventana Medical Systems Inc for use as a fully automated IHC assay on the BenchMark ULTRA staining platform (Ventana Medical Systems Inc, Tucson, Arizona) using the OptiView DAB IHC Detection Kit (Ventana Medical Systems). Assay optimization included evaluation and determination of antibody titer, antibody diluent, multiple antigen retrieval methods, antibody incubation conditions, and counterstain conditions. Western blot analysis was used to assess sensitivity and specificity of the anti-FOLR1 (FOLR1-2.1) mouse monoclonal primary antibody to FRα and compared to folate receptor isoforms (Supplemental Table 3 and Supplemental Figures 1 and 2).
FRα Scoring Method
FRα expression in tumor cells of EOC tissues exhibits both cytoplasmic and membranous staining patterns with varying ranges of stain intensity. However, only membrane staining is evaluated to determine FRα expression status. The membrane staining pattern may be apical or circumferential (partial or complete) and can be categorized into 4 staining intensities: negative, weak, moderate, and strong.27 A stain intensity reference slide containing normal fallopian tube tissue that was representative of having a moderate staining intensity was used to aid pathologists in interpretation of stained slides. FRα expression in EOC tissues was evaluated using both staining intensity and percentage of stained viable tumor cells. EOC was classified as positive for FRα expression when 75% or more of viable tumor cells were stained at the moderate and/or strong intensity (Table 1). The scoring method was established based on data from previous MIRV clinical trials.25,28
Analytic Verification Studies
Intermediate Precision and Repeatability
Intermediate precision and repeatability studies for the VENTANA FOLR1 Assay in EOC samples were completed to demonstrate interantibody lot precision, interdetection kit lot precision, interinstrument precision, interday precision, and intrarun repeatability. For all studies, 24 unique samples were used and evaluated for FRα expression. For interantibody lot, 1 lot of detection kit was used and each case was stained in duplicate on 1 BenchMark ULTRA instrument with 3 different lots of antibody. For interdetection kit lot, 1 lot of antibody was used and each case was stained in duplicate on 1 BenchMark ULTRA instrument with 3 different lots of detection kit. For interinstrument, 1 lot of antibody and 1 lot of detection kit were used to stain duplicate slides of each case on 3 BenchMark ULTRA instruments. For interday precision, replicate slides from 24 unique samples were stained with VENTANA FOLR1 Assay on a single BenchMark ULTRA instrument across 3 nonconsecutive days. Finally, for intrarun repeatability, all cases stained for interantibody lot, interdetection kit lot, interday, and interinstrument were pooled. An equal number of FRα-positive and FRα-negative samples were used in the intermediate precision and repeatability studies.
Reader Precision
The reader precision study encompassed interreader and intrareader precision and assessed 100 EOC samples spanning a range of FRα expressions, including samples around the cutoff for EOC used in this study (75% cutoff). Samples stained with the VENTANA FOLR1 Assay were scored twice by each of 3 pathologists with a minimum 2-week washout period between reads. For each read, slides were blinded, randomized, and evaluated independently by each pathologist.
Before inclusion in the reader precision study, a consensus score was established for each case by a separate set of 2 pathologists to ensure that the case enrollment included an even distribution of positive and negative FRα cases. Consensus scores were not used as a reference standard for any analyses. For interreader precision, for each pair of readers, the result from each case from the first reader was compared to its respective result from the second reader and deemed concordant or discordant. Average positive agreement (APA) and average negative agreement (ANA) were calculated for each pair of readers and aggregated across reader pairs to determine the overall APA and ANA. For intrareader precision, the result from each case from the first read was compared to its respective result from the second read and deemed concordant or discordant. APA and ANA were calculated for each reader and aggregated across readers to determine the overall APA and ANA.
Interlaboratory Reproducibility
An interlaboratory reproducibility study for the VENTANA FOLR1 Assay was conducted to demonstrate the reproducibility of the assay in determining the FOLR1 status in EOC tissue samples. In this study, 28 unique samples (equal number of cases above and below the cutoff) were stained at 3 external College of American Pathologists (CAP)–accredited laboratories on 5 nonconsecutive days during a period of at least 20 days. Staining was conducted on a single BenchMark ULTRA instrument at each site using 1 lot of VENTANA FOLR1 Assay and Optiview DAB IHC Detection Kit. Before staining, slides were blinded and randomized. The stained slides were independently evaluated by 12 pathologists (4 pathologists at each site). The sample set comprised a total of 420 case slides (140 slides per site, 3 sites) and 1680 reads (420 case slides, 4 pathologists) generated from 28 unique samples. However, 1 of the 1680 reads (0.06%) was not evaluable, making the total number of reads 1679.
Clinical Performance
Pathologist Training
Pathologist training was performed at multiple sites for the purposes of clinical trial testing. Training consisted of a review of the comprehensive interpretation guide, didactic training on FOLR1 staining characteristics, and multihead scope session review of example cases. The training was followed by proficiency testing with a set of glass slides from at least 40 cases and a review of discrepancies. Each pathologist had to achieve 85% or more positive percent agreement (PPA) and 85% or more negative percent agreement (NPA) with the consensus FOLR1 score.
ImmunoGen Study IMGN853-0417 (SORAYA Study)
The SORAYA study was a global, single-arm pivotal study sponsored by ImmunoGen that evaluated the safety and efficacy of MIRV in patients with platinum-resistant, advanced high-grade epithelial ovarian, primary peritoneal, or fallopian tube cancers with high FRα expression. The VENTANA FOLR1 Assay was used to select patients for enrollment based on FRα expression. The intent-to-diagnose (ITD) population consisted of all screened patients for the SORAYA study for whom at least 1 of their samples was tested and an evaluation was attempted with the VENTANA FOLR1 Assay. The ITD population consisted of 438 patients, and 431 of these patients had a final evaluable FRα status by the VENTANA FOLR1 Assay.
Statistical Analyses for VENTANA FOLR1 Assay Performance in the SORAYA Study
Evaluation of FRα protein expression for the SORAYA study was performed at a central laboratory using the VENTANA FOLR1 Assay. Pathologists assessed morphology and level of background staining on each case. Morphology was considered acceptable if cellular elements of interest were visualized, allowing clinical interpretation of the stain. Background was considered acceptable if nonspecific staining did not interfere with the interpretation. Pathologists also evaluated each slide for the presence of staining artifacts or other observations that might compromise the interpretation of the slide and determined if the overall staining was acceptable. The data related to the VENTANA FOLR1 Assay were monitored in accordance with federal regulations, Good Clinical Practice guidelines, and Ventana Medical Systems standard operating procedures.
Statistical Analyses for Precision Studies
Interantibody, interdetection, interinstrument, interday, and intrarun studies were evaluated by PPA, NPA, and overall percent agreement (OPA). The acceptance criteria for all 5 studies were defined as PPA, NPA, and OPA of at least 90%. Two-sided 95% CIs were calculated using the percentile bootstrap method from 2000 bootstrap samples. For PPA, NPA, or OPA observations of 100%, the CIs were calculated using the Wilson score method.
Interreader and intrareader precision were evaluated using APA, ANA, and OPA. The acceptance criteria for interreader and intrareader precision was defined as an APA and ANA of at least 85%. Two-sided 95% CIs were calculated using the percentile bootstrap method from 2000 bootstrap samples. For APA, ANA, or OPA observations of 100%, the CIs were calculated using the Wilson score method.
For the interlaboratory reproducibility studies, PPA, NPA, and OPA were calculated for the overall, intrasite, and intrareader agreement rates. For the overall agreement analysis, modal FRα status (status most frequently assigned by multiple independent readers, also referred to as the majority status) was used as the reference status (positive or negative). The acceptance criteria for the overall agreement rate were defined as PPA and NPA of at least 85%. For intrasite analysis, each observation was compared to the respective intrasite mode for each case. For intrareader analysis, each observation was compared to the respective intrareader mode for each case. APA, ANA, and OPA were calculated for the intersite and interreader agreement rates using pairwise comparison. The intersite analysis was calculated by pooling all results from all possible pairs of observations per case between sites (28 cases × 16 reader pairs per day between any 2 sites × 25 day pairs × 3 site pairs). The interreader analysis was calculated by pooling all results from all possible pairs of observations per case within each day at each site (28 cases × 6 reader pairs × 5 days × 3 sites). The 95% CIs for agreement were calculated using the percentile bootstrap method from 2000 bootstrap samples selected with stratification on the diagnostic score bin determined during study screening.
A definition of the analytical studies and their agreement rate acceptance criteria are summarized in Supplemental Table 2.
RESULTS
FRα IHC Analysis
EOC tissues stained with the VENTANA FOLR1 Assay display both cytoplasmic and membranous tumor cell (TC) staining; however, only membrane staining contributes to assessment of FRα status. The assessment requires visual estimation of the percent TCs staining at a negative, weak, moderate, and strong intensity (Table 1). Examples of negative, weak, moderate, and strong FRα intensities are shown in Figure 1. Specimens are considered FRα positive if 75% or more of TCs exhibit moderate and/or strong staining intensities.
Representative staining of the VENTANA FOLR1 Assay in epithelial ovarian cancer (EOC). A, Low magnification of folate receptor-α (FRα) immunohistochemistry (IHC) staining detected using the OptiView DAB IHC Detection Kit (see Materials and Methods). EOC can contain all levels of staining intensity. Boxes indicate zoomed-in insets from A (B through D). Representative images of weak (1+; B), moderate (2+; C), and strong (3+; D) staining. E, Low magnification of representative FOLR1-negative EOC case shows FRα membrane staining in EOC (original magnification ×20).
Representative staining of the VENTANA FOLR1 Assay in epithelial ovarian cancer (EOC). A, Low magnification of folate receptor-α (FRα) immunohistochemistry (IHC) staining detected using the OptiView DAB IHC Detection Kit (see Materials and Methods). EOC can contain all levels of staining intensity. Boxes indicate zoomed-in insets from A (B through D). Representative images of weak (1+; B), moderate (2+; C), and strong (3+; D) staining. E, Low magnification of representative FOLR1-negative EOC case shows FRα membrane staining in EOC (original magnification ×20).
Repeatability and Intermediate Precision
The intrarun repeatability and the interday precision of the VENTANA FOLR1 Assay in EOC samples showed an OPA of 99.1% (214 of 216) and 98.6% (142 of 144), respectively. Similarly, the interinstrument, interantibody lot, and interdetection lot precision studies showed an OPA of 99.3% (143 of 144), 100.0% (144 of 144), and 99.3% (143 of 144), respectively. The PPA, NPA, and OPA for the repeatability and intermediate precision studies are shown in Figure 2.
Interantibody lot, interdetection lot, interinstrument, interday, and intrarun agreements. All precision and repeatability studies used replicate slides from 24 unique samples. For each study, 1 pathologist evaluated folate receptor-α expression, and the positive percent agreement (PPA), negative percent agreement (NPA), and overall percent agreement (OPA) rates were calculated. Greater than or equal to 75% tumor cell membranes staining at moderate and/or strong intensities was considered positive.
Interantibody lot, interdetection lot, interinstrument, interday, and intrarun agreements. All precision and repeatability studies used replicate slides from 24 unique samples. For each study, 1 pathologist evaluated folate receptor-α expression, and the positive percent agreement (PPA), negative percent agreement (NPA), and overall percent agreement (OPA) rates were calculated. Greater than or equal to 75% tumor cell membranes staining at moderate and/or strong intensities was considered positive.
Reader Precision
For interreader and intrareader precision studies, 3 internal pathologists interpreted the VENTANA FOLR1 Assay in 100 EOC tissues. Interreader and intrareader precision showed OPAs of 93.3% (280 of 300) and 97.0% (291 of 300), respectively. The APA, ANA, and OPA for the interreader and intrareader precision studies are shown in Figure 3.
Interreader and intrareader agreements. One hundred epithelial ovarian cancer samples spanning a range of folate receptor-α expressions were used in the reader precision studies. Samples were evaluated by 3 pathologists, and the average positive agreement (APA), average negative agreement (ANA), and overall percent agreement (OPA) were calculated. Greater than or equal to 75% tumor cell membranes staining at moderate and/or strong intensities was considered positive.
Interreader and intrareader agreements. One hundred epithelial ovarian cancer samples spanning a range of folate receptor-α expressions were used in the reader precision studies. Samples were evaluated by 3 pathologists, and the average positive agreement (APA), average negative agreement (ANA), and overall percent agreement (OPA) were calculated. Greater than or equal to 75% tumor cell membranes staining at moderate and/or strong intensities was considered positive.
Interlaboratory Reproducibility
The interlaboratory reproducibility study was conducted at 3 external CAP-accredited laboratories, and 12 external pathologists (4 pathologists at each site) reviewed cases. The PPA, NPA, and OPA were calculated for the overall, intrasite agreement, and intrareader agreement rates (Figure 4). The OPA across all observations for all cases compared against the majority result for each case was 90.6% (1521 of 1679). The OPAs for the overall intrasite and intrareader agreement were 91.2% (1531 of 1679) and 95.6% (1605 of 1679), respectively. APA, ANA, and OPA were calculated for the overall intersite and interreader agreement (Figure 4). Intersite and interreader OPAs were 84.0% (28 188 of 33 560) and 85.3% (2146 of 2517), respectively. Note that intersite and interreader counts indicate the number of paired observations being compared and do not represent the number of unique cases.
Interlaboratory reproducibility study. Twenty-eight epithelial ovarian cancer samples spanning a range of folate receptor-α expression were used in the interlaboratory reproducibility studies. Samples were evaluated by 12 pathologists at 3 unique sites, and the positive percent agreement (PPA), negative percent agreement (NPA), and overall percent agreement (OPA) were calculated for the overall agreement, overall intrasite agreement, and overall intrareader agreement. The average positive agreement (APA), average negative agreement (ANA), and OPA were also calculated for the overall intersite and overall interreader agreements. Greater than or equal to 75% tumor cell membranes staining at moderate and/or strong intensities was considered positive. a Overall agreement for all observations across sites, readers, and days. Not to be confused with OPA.
Interlaboratory reproducibility study. Twenty-eight epithelial ovarian cancer samples spanning a range of folate receptor-α expression were used in the interlaboratory reproducibility studies. Samples were evaluated by 12 pathologists at 3 unique sites, and the positive percent agreement (PPA), negative percent agreement (NPA), and overall percent agreement (OPA) were calculated for the overall agreement, overall intrasite agreement, and overall intrareader agreement. The average positive agreement (APA), average negative agreement (ANA), and OPA were also calculated for the overall intersite and overall interreader agreements. Greater than or equal to 75% tumor cell membranes staining at moderate and/or strong intensities was considered positive. a Overall agreement for all observations across sites, readers, and days. Not to be confused with OPA.
Clinical Performance in the SORAYA Study
The SORAYA study was a pivotal study to determine the efficacy and safety of MIRV in patients with platinum-resistant, advanced high-grade epithelial ovarian, primary peritoneal, or fallopian tube cancers with high FRα expression. The VENTANA FOLR1 Assay was used to determine the FRα status of patients, and only patients whose tumors had high FRα expression were enrolled in the SORAYA study. SORAYA met its primary end point of ORR, as demonstrated by an ORR of 32.4% (95% CI, 23.6%–42.2%) in the VENTANA FOLR1 Assay selected patient population.26
In total, 438 patient tumor samples were tested with the VENTANA FOLR1 Assay during enrollment screening for SORAYA. These 438 patients comprise the ITD population. The ITD population was defined as all patients for whom a diagnostic assessment was attempted with the VENTANA FOLR1 Assay as part of screening for SORAYA. Of these 438 ITD patients, 431 patients had a final evaluable FRα status by the VENTANA FOLR1 Assay.
The sample characteristics from the patients tested with the VENTANA FOLR1 Assay are described in Table 2. Most of the samples tested with the VENTANA FOLR1 Assay were high-grade serous EOC (86.3%; 378 of 438). Samples were either from excision/resection (75.6%; 331 of 438) or biopsy (24.4%; 107 of 438). Most samples were primary tumors (63.7%; 279 of 438) that were located in the ovary (48.6%; 213 of 438), peritoneum (10.0%; 44 of 438), and fallopian tube (3.9%; 17 of 438).
In the SORAYA study, trained pathologists determined the FRα status of patient samples using the VENTANA FOLR1 Assay Scoring method in Table 1. The distribution of raw FRα scores for all patients in the ITD population is illustrated in Figure 5. Of the 431 patients tested with the VENTANA FOLR1 Assay who had an evaluable score, 155 (36.0%) were FRα positive. In the ITD population, the VENTANA FOLR1 Assay exhibited a final staining acceptability rate of 98.4% (431 of 438), a final acceptable background rate of 99.8% (434 of 435), and a final acceptable morphology rate of 99.1% (434 of 435).
Distribution of folate receptor-α (FRα) scores from patient specimens stained with VENTANA FOLR1 Assay in the intent-to-diagnose (ITD) population of the SORAYA study. Violin plot of the combined % tumor cell (TC) at 2+ and %TC at 3+ for the ITD population in the SORAYA study. Asterisk indicates FOLR1-positive was defined as 75% or greater of viable tumor cells with 2+ and/or 3+ membrane staining. Blue indicates the proportion of patients in the ITD group who had an evaluable VENTANA FOLR1 Assay result that was positive (36.0%).
Distribution of folate receptor-α (FRα) scores from patient specimens stained with VENTANA FOLR1 Assay in the intent-to-diagnose (ITD) population of the SORAYA study. Violin plot of the combined % tumor cell (TC) at 2+ and %TC at 3+ for the ITD population in the SORAYA study. Asterisk indicates FOLR1-positive was defined as 75% or greater of viable tumor cells with 2+ and/or 3+ membrane staining. Blue indicates the proportion of patients in the ITD group who had an evaluable VENTANA FOLR1 Assay result that was positive (36.0%).
DISCUSSION
Here, we describe the development, analytic verification, and clinical performance of the VENTANA FOLR1 Assay in EOC. Using industry standard verification studies, we show that the VENTANA FOLR1 Assay is highly reproducible, demonstrated by a greater than 98% PPA, NPA, and OPA for the repeatability and intermediate precision studies. Furthermore, all interreader agreements (APA, ANA, and OPA) were greater than 93%, and all intrareader agreements were 96.0% or greater. In addition, the interlaboratory reproducibility of the VENTANA FOLR1 Assay was tested at 3 unique sites, and the OPA across all observations was greater than 90%, while the OPA between all reading pathologists was greater than 85%. Collectively, these data support the VENTANA FOLR1 Assay as the first robust and reproducible non–laboratory-developed test (LDT) IHC assay for the detection of FRα expression in EOC.
The VENTANA FOLR1 Assay was the IHC assay used in the ImmunoGen-sponsored SORAYA study to determine patient FRα expression levels and select patients for MIRV treatment. The SORAYA trial was a simultaneous investigation of the (1) safety and efficacy of MIRV and (2) clinical performance of the VENTANA FOLR1 Assay and its ability to identify patients for MIRV therapy. The patient population that was identified using the VENTANA FOLR1 Assay and treated with MIRV demonstrated a clinically meaningful benefit to MIRV treatment.26 More recently, results from the ImmunoGen-sponsored MIRASOL study (NCT04209855) were reported.29 Like the SORAYA study, MIRASOL used the VENTANA FOLR1 Assay to select FRα-positive patients for treatment with MIRV. In this open-label trial, MIRV demonstrated substantial clinical efficacy in the FRα-positive patient population selected by the VENTANA FOLR1 Assay, compared with standard chemotherapy. These results confirmed the findings of the SORAYA trial and further demonstrate the clinical utility of the VENTANA FOLR1 Assay in selecting EOC patients for treatment with MIRV.
Within the SORAYA trial, the clinical performance of the VENTANA FOLR1 Assay was also evaluated by the overall staining acceptability rate, which was calculated using the number of patient specimens that were tested with the VENTANA FOLR1 Assay that had an evaluable result. The overall staining acceptability rate in the ITD population was 98.4%, demonstrating that the VENTANA FOLR1 Assay is highly reliable and consistently produces evaluable results in the clinical setting. Additionally, for a pathologist to properly evaluate FRα expression with the VENTANA FOLR1 Assay, the background staining and tissue morphology must both be acceptable on the slide stained with the assay. The overall background acceptability and morphology acceptability rates in the ITD population were both 99.8%.
The efficacy results from the SORAYA trial combined with the robust staining performance in the clinical setting support the safe and effective use of the VENTANA FOLR1 Assay for selecting patients who may benefit from MIRV therapy.
The efficacy of inhibitory FRα-targeting antibodies and ADCs relies on their ability to recognize and bind to FRα on the surface of cancer cells. In vitro studies demonstrated that the IC50 of MIRV in different cell lines scaled with the number of surface FRα molecules.23 Consistent with these preclinical data, selection of EOC patients with high FRα expression using the VENTANA FOLR1 Assay led to identification of a patient population that derived clinically meaningful benefits from MIRV treatment in the SORAYA study.26 In line with these promising results, additional investigations are underway to examine the safety and efficacy of FRα-targeting therapies in different cancer indications (clinicaltrials.gov). Future analytic studies examining the VENTANA FOLR1 Assay’s ability to detect FRα in other cancers that upregulate FRα, such as renal, triple-negative breast, endometrial, and lung cancer, would be of value.
LDTs are clinical diagnostic tests developed specifically for use in a single laboratory. With the widespread availability of commercial antibodies for IHC, many clinical laboratories choose to develop IHC LDT tests, in lieu of verification of analytically and clinically validated IHC assays. Although organizations such as the CAP provide guidance for how IHC LDTs should be validated by laboratories, this process can be both time- and resource-consuming. Moreover, each newly purchased antibody must be revalidated and closely monitored for drifts in LDT performance to account for lot-to-lot variances in reagents. Analytically and clinically validated assays are uniformly produced and scaled to support clinical trials and widespread clinical use across the globe, whereas LDTs are best suited in the single-laboratory setting.
In this study, we show the analytic and clinical performance of a diagnostic assay developed according to the current FDA guidelines for companion diagnostics.30 Using accepted studies, we show the repeatability and reproducibility of an IHC assay that is intended for patient selection in routine clinical practice. Moreover, deployment of this assay within a pivotal clinical trial26 demonstrated the clinical performance and enabled establishment of the clinical utility for the VENTANA FOLR1 Assay.