Abstract
Objective
With the development of novel antibody-drug conjugates (ADCs), folate receptor alpha (FOLR1) is a promising therapeutic target for the treatment of platinum-resistant tubo-ovarian carcinomas. The main aims of this study were to assess FOLR1 protein expression in a large cohort of ovarian carcinoma histotypes. To inform future clinical trial design we identified molecular correlates of FOLR1 expression in low-grade serous carcinoma (LGSC).
Methods
One thousand five hundred forty-seven ovarian carcinoma samples from 5 different Canadian cohorts were successfully evaluated by immunohistochemistry for FOLR1 expression using the PS2+ system. Statistical analyses with clinicopathological parameters, LGSC molecular subtypes, and overall survival (OS) were performed.
Results
High FOLR1 expression was detected in 44% of high-grade serous carcinomas, and in 30% LGSC, 8% clear cell, 6% endometrioid, and 0% mucinous and/or mesonephric-type adenocarcinomas. In 160 LGSC cases, FOLR1 expression was more frequent in cases with normal MAPK pathway status (37% MAPK wild type vs. 14% canonical MAPK pathway mutations; p=0.002), low progesterone receptor (PR) expression (41%) vs. 23% (Allred score >2; p=0.02), and p16 loss (48% p16 absent vs. 26% normal; p=0.03). Canonical MAPK mutation status and PR expression remained significant on multivariable analysis. No significant associations between OS and FOLR1 expression were observed.
Conclusion
A significant proportion of LGSC express high FOLR1 levels supporting the development of clinical trials to investigate ADCs targeting FOLR1 as novel agents for treating this disease. In LGSC, high FOLR1 expression was associated with fewer MAPK pathway alterations, low PR expression, and p16 loss.
Keywords: Ovarian Cancer, Low-Grade Serous Carcinoma, Folate Receptor Alpha, FRa, Mirvetuximab Soravtansine, Molecular Subtypes
Synopsis
FOLR1 protein levels are different across ovarian carcinoma histotypes, highest in high-grade (44%) and low-grade serous carcinoma (LGSC, 30%). In LGSC, FOLR1 expression is associated with MAPK, PR and p16 molecular subtypes. In LGSC, there is an urgent need for clinical trials targeting FOLR1 that could expand current therapy options.
Graphical Abstract
INTRODUCTION
Less common histotypes of metastatic ovarian carcinoma are resistant to chemotherapy and difficult to treat. Low-grade serous carcinomas (LGSCs) represent a prime example: LGSC accounts for only 3% of all ovarian carcinomas and are commonly diagnosed in advanced stages. The optimal primary treatment for advanced-stage LGSC remains to be defined, but usually includes surgery and maintenance anti-estrogen therapy, with or without chemotherapy. Unfortunately, over 70% of patients recur following primary treatment [1]. Given that many LGSC harbor MAPK pathway mutations, MEK inhibitors (MEKi) were studied in randomized clinical trials in patients with relapsed disease. Trametinib was shown to confer an improvement in median progressive-free survival of 5.8 months when compared to standard-of-care treatment [2]. While it was encouraging to observe this treatment benefit with a targeted therapy, MEKi have significant side effects and overall response rates in trametinib-treated trial patients was only 26% [2]. Therefore, the development of effective alternative therapeutics for LGSC remains a top priority.
Antibody-drug conjugates (ADCs) represent a therapeutic modality of intense interest for the treatment of cancer. There are recent improvements in linker chemistry and these new agents have improved therapeutic ratios. Approved ADCs for breast and urothelial cancers consist of antibodies targeting ERBB2 or TROP2 linked to cytotoxic drugs (so-called payload) such as a microtubule or topoisomerase inhibitors [3]. Mirvetuximab soravtansine (MIRV) is a novel ADC targeting folate receptor alpha (FOLR1), which is linked to a microtubule inhibitor at a drug antibody ratio of 3.5 [3]. MIRV has recently shown promising efficacy (32% overall response rate, SORAYA Trial) in advanced platinum-resistant tubo-ovarian high-grade serous carcinomas (HGSCs) with high FOLR1 expression, assessed by the Food and Drug Administration (FDA) approved companion immunohistochemistry (IHC) test (VENTANA® FOLR1 (FOLR1-2.1) RxDx Assay (Roche, Basel, Switzerland) and classified using the PS2+ scoring system [4]. FOLR1 is a membrane-bound protein encoded by the FOLR1 gene that facilitates folate uptake into cells for one-carbon metabolism important for DNA synthesis and methylation [5]. FOLR1 is overexpressed in ovarian carcinomas with limited expression in normal adult tissue making this a highly attractive therapeutic target [6].
Recognizing the activity of MIRV in platinum-resistant ovarian carcinomas, we sought to determine FOLR1 expression in a large cohort of ovarian carcinomas with a particular focus on LGSC. Using the PS2+ scoring system, 3 recent studies show that a substantial proportion of LGSC tumours express high levels (25%–40%) of FOLR1 [7,8,9]. Notably, none of the clear cell carcinoma (CCC), mucinous carcinoma (MC), or endometrioid carcinoma (EC) carcinomas expressed high levels of FOLR1 [7]. In LGSC, we now have an improved understanding of the spectrum of molecular subtypes with therapeutic relevance including MAPK pathway alterations, low progesterone receptor (PR) expression with high fraction of genome altered and p16 loss [10,11,12,13]. In LGSC, associations with MAPK mutations/activation with FOLR1 expression have been inconsistent [8,9]. Thus, an important aim of this study was to evaluate how FOLR1 expression correlates with these molecular subtypes. We sought to evaluate FOLR1 expression by IHC in a larger cohort of ovarian carcinoma cases representing 6 different histotypes (HGSC, LGSC, CCC, EC, MC, mesonephric-type adenocarcinoma [MA]) to determine the feasibility of conducting future clinical trials evaluating MIRV and/or other ADC’s targeting FOLR1 in these less common, chemotherapy-resistant ovarian carcinomas.
MATERIALS AND METHODS
1. Study cohort
One thousand seven hundred thirty-four ovarian carcinoma tissue samples were sourced from existing cohorts from Alberta and British Columbia, Canada. The Ovarian Cancer in Alberta and British Columbia (OVAL-BC) cohort recruited over 900 cases between 2001 and 2012 [14]. The Alberta Ovarian Tumor Type (AOVT) study collected over 500 cases from the Alberta Cancer Registry between 1978 and 2010 [15]. The Vancouver cohort (V2016 and V2022) consisted of 150 cases collected from 2001 to 2020 [16]. The Canadian Ovarian Experimental Unified Resource (COEUR) cohort consisted of 71 LGSC cases collected between 2010 and 2017 [17]. Additional hospital-based case series of low-grade serous and ovarian mesonephric carcinomas were included [12,18,19,20]. Tumor specimens from all cohorts were available on tissue microarrays (TMA) with each tumor represented by 2, 3, or 4, 0.6 mm diameter cores. For a subset of cases, full sections were used in addition to the TMA. The research on these cohorts received local ethics board approvals by Health and Research Ethics Board Alberta (HREBA.CC-16-0161, HREBA.CC16.0159, HREBA.CC-16.0371, HREBA.CC-21-0362) and the University of British Columbia Ethics Board (H22-00544, H17-01863, H18-00280).
2. IHC and scoring criteria
A laboratory-derived immunohistochemical assay for FOLR1 was optimized using normal fallopian tube as positive control for membranous staining with moderate intensity. A variety of normal tissues including tonsil, liver, pancreas, and colon served as negative controls (Fig. S1). IHC was performed on 4 µm tumour sections from TMA with 30 minutes of heat-induced antigen retrieval in TRIS-EDTA buffer (pH 9.0). The primary FOLR1 antibody (mouse monoclonal antibody, clone BN3.2, Leica) was incubated for 30 minutes at a dilution of 1:50, followed by a mouse linker for 10 minutes, and 30 minutes of peroxidase labelled Dako Envision+ polymer-based detection system (DAKO protocol H30-10M-30) on a DAKO Omnis auto stainer (Agilent, Santa Clara, CA, USA).
Staining was initially interpreted using the H-score system, that is a product of intensity (absent=0, weak=1, moderate=2, strong=3; Fig. S2) and % distribution, which gives a range from 0 to 300. Cores with less than 25% tumor content were deemed uninterpretable and excluded. FOLR1 expression results were categorized using the PS2+ scoring system [4]. Cases with at least moderate intensity and ≥75% distribution were categorized as high, 50%–74% expression as moderate and less than 50% regardless of intensity as low. Two pathologists (AD, MK) scored FOLR1 IHC on TMAs containing 127 samples of different ovarian carcinoma histotypes for inter-observer reproducibility. AD scored about 75% of the samples and MK the remainder.
Previously assessed biomarkers that are molecular surrogates for key molecular alterations within LGSC were used for correlative studies [12]. All but 4 cases LGSC were TP53 normal by IHC: 2 cases were not interpretable, and 2 cases showed subclonal TP53 overexpression, which was interpreted as high-grade transformation in tumors with otherwise typical features of LGSC on review [21]. P16/CDKN2A scoring on LGSC cases was categorized as normal (patchy), abnormal absent, or abnormal block expression [13]. Allred scoring was used for estrogen receptor (ER) and PR expression using the following cutoffs: ER: low Allred score <7, moderate Allred score=7–7.5, high Allred score=8; and for PR: low Allred score <2, moderate Allred score 2–5, high Allred score 6–8 [16]. USP9X: was categorized as abnormal loss or normal retained [10]. Whole exome or targeted sequencing was done to determine the presence of MAPK pathway alterations defined as oncogenic mutations in KRAS, NRAS, BRAF or IHC alteration in NF1 [12,22].
3. Statistical analysis
Cohen’s kappa coefficient was calculated to assess for paired inter-observer reproducibility of categorized PS2+ scores. Inter-observer reproducibility was compared between a 2-tier scoring system (low/moderate vs. high) vs. a 3-tier scoring system (low vs. moderate vs. high). Pearson’s χ2 test was used to assess different proportion of categorical variables. Kaplan-Meier survival analyses with left truncation at 10 years were performed and differences were assessed using the log-rank test. The primary endpoint was overall survival (OS). OS was defined as the time from diagnosis to death or last follow-up. Stage-stratified survival analyses were performed given that systemic therapy is often used to treat cases of advanced platinum-resistant disease (stages II–IV), but much less frequently used for stage I disease [23,24,25]. Multivariable cox regression models were used to estimate the hazard ratios of potential prognostic factors and evaluate their associations with OS. All tests were 2-sided and the level of significance was established at a p-value of less than 0.05. Statistical tests were run using SPSS (Version 29.0.2.0 (20); IBM Corp., Armonk, NY, USA).
RESULTS
1. FOLR1 measurement, expression, and scoring across ovarian carcinoma histotypes with survival correlations
There were 1,547 ovarian carcinomas representing the histotypes of interest, including 160 primary LGSC cases, in which FOLR1 IHC data were successfully scored (Table 1). As expected, the majority of HGSC (92%) and LGSC (91%) cases were advanced stage (II–IV) and other histotypes more frequently presented with stage I disease (CCC: 44%, EC: 54%, and MC: 75%; data not shown). Using the H-score system (Fig. 1A), FOLR1 protein expression levels were highest in HGSC (median H-score 130), followed by LGSC (median H-score 100), while the lowest expression was detected in MCs (median H-score 0). Individual H-score data is included in Table S1.
Table 1. FOLR1 expression across ovarian carcinoma histotypes based on the PS2+ scoring system.
| Histotypes | FOLR1 PS2+ scores | ||
|---|---|---|---|
| Low | Moderate | High | |
| HGSC (n=662) | 258 (39) | 116 (17.5) | 288 (43.5) |
| LGSC (n=160) | 79 (49.4) | 33 (20.6) | 48 (30) |
| CCC (n=264) | 214 (81.1) | 28 (10.6) | 22 (8.3) |
| EC (n=332) | 294 (88.6) | 17 (5.1) | 21 (6.3) |
| MC (n=117) | 116 (99.1) | 1 (0.9) | 0 (0) |
| MA (n=12) | 10 (83.3) | 2 (16.7) | 0 (0) |
| Total (n=1,547) | 971 (62.8) | 197 (12.7) | 379 (24.5) |
Values are presented as number (%). p<0.001 when comparing FOLR1 PS2+ score across all histotypes, p=0.007 (χ2 test) comparing HGSC vs. LGSC (high vs. other).
CCC, clear cell carcinoma; EC, endometrioid carcinoma; FOLR1, folate receptor alpha; HGSC, high-grade serous ovarian carcinoma; LGSC, low-grade serous ovarian carcinoma; MA, mesonephric-type adenocarcinoma; MC, mucinous carcinoma.
Fig. 1. FOLR1 protein expression levels across ovarian carcinoma histotypes. (A) Continuous data by H-score. (B) Categorical data by PS2+ score.
CCC, clear cell carcinoma; EC, endometrioid carcinoma; FOLR1, folate receptor alpha; HGSC, high-grade serous carcinoma; LGSC, low-grade serous carcinoma; MA, mesonephric-type adenocarcinoma; MC, mucinous carcinoma.
A subset of 127 cases was studied to evaluate inter-observer agreement using differing PS2+ scoring tiers. The inter-observer agreement for FOLR1 IHC interpretation achieved a Cohen’s kappa of 0.68 for the 3-tier categorization and 0.69 (87.4% concordance) for the 2-tier categorization. The main problem for distinguishing the PS2+ high from moderate FOLR1 category was the interpretation of what constitutes at least moderate intensity in cases with variable weak to moderate intensity close to the 75% cut-off (Fig. S2).
Using the PS2+ scoring system, HGSC and LGSC showed the highest proportion of high-expressing FOLR1 cases, with 44% (288/662) and 30% (48/160) being PS2+ high, respectively (Table 1). Of note, 5%–10% of EC and CCC were found to have high FOLR1 expression, and their frequency was unaffected by stage. For stage I EC cases, 7/147 (5%) had high FOLR1 expression, comparable to the 10/126 (8%) stage II–IV EC cases (p=0.28; χ2 test). Similarly, there were 10/97 (10%) stage I CCC with high FOLR1 expression which did not significantly differ from the 9/125 (7%) stage II–IV CCC cases (p=0.41; χ2 test). Remarkably, none of the MC (117 cases) or the 12 cases of the rare MA showed high FOLR1 expression (Fig. 1B, Table 1).
We examined OS associations of the patient population by FOLR1 expression. In a large cohort of 469 patients with advanced (stage II–IV) HGSC with survival outcomes data, we did not observe any differences in OS in the cohort of 212 patients with high FOLR1 expression compared to the 257 HGSC patients with lower FOLR1 expression (p=0.73; log-rank test, Fig. 2A). Additionally, survival analyses showed that FOLR1 expression was not associated with different OS outcomes in 144 patients with advanced LGSC (Fig. 2B). As expected, Cox regression analysis confirmed this finding (p=0.5) when other factors (age, stage, residuum, MAPK mutation status) were included in the model. None of the patients with high FOLR1 expressing stage I CCC (n=10) or stage I EC (n=7) died of disease, although these outcomes were not significantly different from 87 stage I FOLR1 low/moderate expressing CCC and 140 stage I FOLR1 low/moderate expressing EC (Fig. 2 C and D).
Fig. 2. Kaplan-Meier survival analyses for OS. (A) High-grade serous carcinoma, stage II–IV. (B) Low-grade serous carcinoma, stage II–IV. (C) Clear cell carcinoma, stage I. (D) Endometrioid carcinoma, stage I.
FOLR1, folate receptor alpha; OS, overall survival.
2. Spatial stability of FOLR1 expression in LGSC and expression in recurrent cases
To determine the accuracy of using TMA for assessing FOLR1 expression, we compared FOLR1 IHC scoring on whole sections from 26 LGSC with the corresponding scores on the matched TMA samples (Fig. 3). Overall, there was almost perfect agreement (93%; Cohen’s kappa 0.89) between TMA results and whole sections. There were 2 exceptions: one case showed low expression on TMA and moderate on whole section, while another case showed high expression on TMA and moderate on whole section. In addition, we assessed the FOLR1 expression on 9 recurrent LGSC cases. High FOLR1 expression was observed in 4/9 (44%) of these recurrent cases including 3 matched recurrences that were concordant with the primary.
Fig. 3. FOLR1 immunohistochemistry in LGSC. Representative images of PS2+ categories from whole sections (back frame high power inset from whole section, note round holes from coring) and corresponding tissue microarrays cores. (A) LGSC with low FOLR1 expression showing weak intensity staining in most of the tumor cells with foci of moderate staining in less than 50% of tumor cells. (B) LGSC with moderate FOLR1 expression showing moderate to strong membranous intensity in >50% but less than 75% of tumor cells. (C) LGSC with high FOLR1 expression showing strong membranous and cytoplasmic intensity in 100% of the tumor cells.
FOLR1, folate receptor alpha; LGSC, low-grade serous carcinoma.
3. Association of FOLR1 expression status in LGSC with clinico-pathological variables and molecular subtypes
Patient demographics of primary LGSC cases are summarized in Table 2. The median age of patients with primary LGSC was 55.5 (range: 21–79). 68% of patients were optimally debulked (<1 cm residual disease) and 32% were suboptimally debulked (>1 cm residual disease). As expected, most cases were advanced stage (II–IV, 92% or 144/157) at diagnosis Using PS2+ scoring, no statistically significant associations were observed between FOLR1 expression and age, stage, tumor site of origin, or treatments. A significantly higher proportion of LGSC patients with no residual disease following initial surgery (17/33 or 52%) had high FOLR1 expression compared to those with residuum (26/106 or 25%; p=0.003).
Table 2. Demographic/clinical univariate associations with FOLR1 expression in the LGSC cohort using PS2+ scoring.
| Clinico-pathologic variables | LGSC (n=160) | FOLR1 expression category | p-value* | ||
|---|---|---|---|---|---|
| FOLR1 low/moderate (n=112) | FOLR1 high (n=48) | ||||
| Age at diagnosis | 0.48 | ||||
| Mean ± standard deviation | 53.9±13.1 | 53.4±12.8 | 55.0±13.8 | ||
| Median (range) | 55.5 (21–79) | 54.5 (21–79) | 57.0 (25–79) | ||
| Stage | 0.75 | ||||
| I | 13 (8.1) | 10 (8.9) | 3 (6.2) | ||
| II–IV | 144 (90.0) | 100 (89.3) | 44 (91.7) | ||
| Unknown | 3 (1.9) | 2 (1.8) | 1 (2.1) | ||
| Residual disease | 0.003 | ||||
| Absent | 33 (20.6) | 16 (14.3) | 17 (35.4) | ||
| Optimal <1 cm | 61 (38.2) | 50 (44.6) | 11 (22.9) | ||
| Suboptimal >1 cm | 45 (28.1) | 30 (26.8) | 15 (31.3) | ||
| Unknown | 21 (13.1) | 16 (14.3) | 5 (10.4) | ||
| Tumor site | 0.67 | ||||
| Ovarian | 74 (46.3) | 42 (37.5) | 32 (66.6) | ||
| Peritoneal | 21 (13.1) | 13 (11.6) | 8 (16.7) | ||
| Unknown | 65 (40.6) | 57 (50.9) | 8 (16.7) | ||
| Treatment | 0.72 | ||||
| Chemotherapy naïve | 129 (80.6) | 92 (82.1) | 38 (79.2) | ||
| Post neoadjuvant chemotherapy | 28 (17.5) | 18 (16.1) | 10 (20.8) | ||
| Post other treatments | 3 (1.9) | 2 (1.8) | 0 (0) | ||
| Unknown | 0 (0) | 0 (0) | 0 (0) | ||
| ER | 0.30 | ||||
| Low + Moderate | 35 (21.9) | 27 (24.1) | 8 (16.7) | ||
| High | 122 (76.3) | 83 (74.1) | 39 (81.3) | ||
| Unknown | 3 (1.8) | 2 (1.8) | 1 (2.0) | ||
| PR | 0.02 | ||||
| Low | 66 (41.3) | 39 (34.8) | 27 (56.3) | ||
| Moderate + High | 91 (56.8) | 70 (62.5) | 21 (43.7) | ||
| Unknown | 3 (1.9) | 3 (2.7) | 0 (0) | ||
| P16 | 0.03† | ||||
| Normal patchy | 127 (79.4) | 94 (83.9) | 33 (68.8) | ||
| Abnormal absent | 23 (14.4) | 12 (10.7) | 11 (22.9) | ||
| Abnormal block | 8 (5.0) | 5 (4.5) | 3 (6.3) | ||
| Unknown | 2 (1.2) | 1 (0.9) | 1 (2.0) | ||
| USP9X | 0.48 | ||||
| Abnormal loss | 15 (9.4) | 10 (8.9) | 5 (10.4) | ||
| Normal retained | 81 (50.6) | 46 (41.1) | 35 (72.9) | ||
| Unknown | 64 (40.0) | 56 (50.0) | 8 (16.7) | ||
| MAPK pathway alteration | 0.002 | ||||
| Altered | 59 (36.9) | 51 (45.5) | 8 (16.7) | ||
| Normal | 75 (46.9) | 47 (42.0) | 28 (58.3) | ||
| Unknown | 26 (16.2) | 14 (12.5) | 12 (25.0) | ||
Values shown (other than age at diagnosis) are the case numbers and their corresponding frequency distribution (%) for each variable (and its categories) according FOLR1 expression (low/moderate vs. high).
ER, estrogen receptor; FOLR1, folate receptor alpha; LGSC, low-grade serous ovarian carcinoma; MAPK pathway alteration, NRAS, KRAS, BRAF, and NF1 mutations; PR, progesterone receptor; USP9X, ubiquitin specific peptidase 9 X-linked.
*Chi-square test comparing the FOLR1 expression low + moderate group vs. the FOLR1 high expression group for each clinico-pathologic variable.
†The p-value representing only the comparison between p16 normal patchy and abnormal absent.
We then analyzed FOLR1 expression in relation to key LGSC molecular characteristics and subtypes. Normal MAPK pathway status was associated with high FOLR1 expression. Thirty-seven percent (28/75) of patients without common MAPK pathway alterations expressed high levels of FOLR1 compared to 14% (8/59) of those with MAPK pathway alterations (p=0.002, Fig. 4). We also examined the association between specific MAPK pathway mutations in LGSC cases (KRAS, BRAF and NRAS) and FOLR1 expression (Table S2). High PS2+ scores were significantly less frequent in mutated KRAS cases vs. wild type (10% vs. 90%; p<0.005) and none of the 8 mutated BRAF cases (p=0.047) expressed high levels of FOLR1. Though only 20% of NRAS cases (2/10) had high PS2+ expression, this difference was not statistically significant (p=0.48). There were 86 LGSC cases with NF1 IHC results. Overall, high FOLR1 expression was observed in 38% (33/86) of cases and all 3 cases with NF1 loss had high FOLR1 expression.
Fig. 4. Associations of categorical FOLR1 protein expression with key biomarker of low-grade serous carcinoma. (A) FOLR1 expression and MAPK pathway status. (B) FOLR1 expression and PR status. (C) FOLR1 expression and p16/CDKN2A status.
FOLR1, folate receptor alpha; PR, progesterone receptor.
Low PR expression was also found to be associated with high FOLR1 expression (Table 2, Fig. 4). Forty-one percent of cases (27/66) with a PR Allred score of less than 2 expressed high FOLR1, whereas only 21/91 (23%) of tumors with a PR Allred score of 2 or greater were high FOLR1 expressors (p=0.02). Furthermore, there was a significant association between cases with p16 loss and high FOLR1 expression. Forty-eight percent (11/23) of cases with p16 loss had high FOLR1 expression while only 26% (33/127) of cases with normal p16 expression were high expressors of FOLR1 (p=0.03, Fig. 4). No significant associations were observed between FOLR1 expression and either ER expression or USP9X loss. Finally, logistic regression was performed to evaluate the impact of the 3 significant univariate molecular associations with FOLR1 expression: p16, canonical MAPK mutation status and PR expression. In 121 LGSC cases with complete data, those cases that were wild type for a canonical MAPK mutations (p<0.001) and those with low PR expression (Allred score ≤2; p=0.031), were significantly associated with a higher frequency of high FOLR1 expression (data not shown). The association between p16 expression and FOLR1 expression was not statistically significant in the multivariable model (p=0.32).
DISCUSSION
By applying the new PS2+ scoring system for the evaluation of FOLR1 IHC expression, we showed that FOLR1 protein is variably expressed across ovarian carcinoma histotypes. High FOLR1 expression has been established as a predictive biomarker in HGSC, and testing is now required to select patients for treatment with MIRV [4,26]. Previous studies showed a lack of efficacy when treating patients with FOLR1-targeting ADCs without or suboptimal measuring FOLR1 expression for patient selection [27,28]. Thus, to further explore FOLR1 as a biomarker we evaluated the prevalence of high FOLR1 expression by screening large TMA including uncommon chemotherapy resistant ovarian carcinomas histotypes. Using the PS2+ method, in our study, 30% of 160 LGSC cases showed high FOLR1 expression. This is the range of 3 previous studies reporting 30% in a total of 17 LGSC [7], 40% in 89 LGSC [8], and 25% in 281 LGSC [9]. By combining data from these 4 studies, the average frequency for FOLR1 high is 29% (158/547) in LGSC.
In addition to its established role in HGSC, our results are particularly relevant for LGSC showing nearly the same proportion of high expressing FOLR1 cases. Recognizing that LGSC are generally resistant to platinum-based chemotherapy [29], prior molecular studies were undertaken to identify biomarkers of potential therapeutic relevance for treating patients with LGSC [10,11,12]. Interestingly, we found that the frequency of high FOLR1 expression in LGSC is associated with low PR expression and/or p16 loss. In cases with low PR expression, we observed that 41% (27/66) of LGSC were FOLR1-high vs. 23% (21/91) with higher PR expression (p=0.02). Though, cases with p16 loss (n=23), were also more likely to express high levels of FOLR1 (48%; 11/23), this association was not statistically significant on multivariable analysis.
While Rushton et al. [9] reported no differences in transcriptomic MAPK pathway activation between FOLR1 low and high LGSC, we confirmed the previously reported association between MAPK pathway alterations and high FOLR1 expression [8]. In the previous study, 61% of LGSC (20/33) without MAPK alterations were positive for FOLR1 compared to 20% (9/45) of those with alterations [8]. Of the 75 patients in this study without MAPK pathway mutations, 37% (28/75) had high levels of FOLR1 expression compared to only 14% (8/59) in those cases with mutation-associated pathway activation (p=0.002). In this context, our results suggest that MIRV could potentially close a treatment option gap for LGSC that may not respond well to either anti-estrogen therapy (low PR expression with high FOLR1) or MEKi (normal MAPK pathway status with high FOLR1). Further study is also needed to determine the relative susceptibility of patients with LGSC whose tumors show p16 loss and high FOLR1 to a combination of MIRV and CDK 4/6 inhibition.
It is worth noting for the MC and MA carcinomas high expression of FOLR1 was not observed. This was also observed for MC in the report by Lawson et al. [7]. However, in contrast to their report, we noted high FOLR1 expression in 5%–10% of EC and CCC carcinomas including advanced stage cases. This may be an important finding as relapsed EC and CCC are usually platinum-resistant. Clinical trials evaluating ADCs could include MIRV in these rarer histotypes with high FOLR1 expression.
While the high frequency of high expressors is encouraging for future clinical trials, it should be kept in mind that success of ADCs depends on several factors: 1) the ability to bind to the target and enrich in tumor tissue more than normal tissue, 2) the stability of the linker in the circulation, 3) payload release and/or internalization within the tumor, 4) efficacy of the payload. Hence, the same payload (microtubule inhibitor in case of MIRV) may not be as effective in lower proliferating tumours (such as LGSC) compared to more highly proliferating cancers like HGSC [30]. As a next step, preclinical studies evaluating FOLR1 targeting agents in vitro or in vivo in LGSC models are desired before launching clinical trials.
Our results also confirm that FOLR1 expression is not prognostic [23,31,32]. Because of folate’s role in one-carbon transfer and DNA synthesis, it has been hypothesised that FOLR1 might confer a proliferation advantage for tumors [33]. However, the lack of survival association and studies highlighting the importance of alternative folate transport routes have challenged this idea [34,35]. It is interesting to note that FOLR1 shows no prognostic association in LGSC despite its association with adverse prognostic markers (low PR, normal MAPK pathway) [13,36,37]. Since FOLR1 is expressed in moderate levels in the tissue of origin (normal fallopian tube type tissue often displaced as endosalpingiosis), one might speculate that high FOLR1 expression may be a marker of differentiation and that low expression is conferring a survival disadvantage, which may offset above associations. Future omics study may shed light into differences regarding LGSC with differential FOLR1 status. While not prognostic, high FOLR1 expression is an established predictive marker for response to MIRV in HGSC [4,26].
While we interpreted the FOLR1 staining with the same PS2+ scoring method as in the SORAYA and MIRASOL trials, we did not use the FDA approved companion IHC test (VENTANA® FOLR1 (FOLR1-2.1) RxDx Assay (Roche) [26]. Instead, we optimized a laboratory-developed IHC assay with normal fallopian tube as positive control for moderate membranous (luminal) intensity as recommended in previous studies [8,32]. The almost perfect concordance of FOLR1 high prevalences for HGSC with 44% in our as well as in a recent study of 5,086 HGSC [9] and for LGSC with 30% in our study and 29% averaged across the other 3 studies [7,8,9] speaks to the robustness of our test. Since most pathology departments contract their IHC equipment with one of the dominant vendors, FDA requirement of using a specific platform may hinder wide testing. Laboratory-derived tests that are subject to quality control standards may represent an acceptable alternative to allow more patients to get tested for drug eligibility [38].
Regarding the tissue requirements for FOLR1 testing by IHC, our results indicate high concordance between small TMA cores and corresponding whole sections from the same formalin fixed paraffin embedded tissue block. Therefore, needle core biopsies, which are similar in size compared to TMA cores, should be a reasonably reliable method for evaluating FOLR1 expression in relapsed patients. Previous studies conducted reported a slightly lower spatial concordance with 82% in HGSC and over time between 69% and 71% in HGSC and LGSC, respectively [8,39]. Manning-Geist et al. [8] reported on 9 matched recurrent cases showing that 3 of the primary cases that were initially PS2+ negative were PS2+-positive on recurrence. This suggests that high FOLR1 expression is generally retained if not higher in recurrent LGSC. Though the sample size was small, 4 of 9 (44%) of our recurrent LGSC showed high FOLR1 expression. These preliminary findings suggest that trials in the recurrent setting may be promising and that retesting of FOLR1 expression should be considered on relapse. Although the interobserver agreement for categorized PS2+ scoring was substantial (kappa range 0.61–0.80), it was less than perfect (kappa range 0.81–1.00). Since the definition of high by PS2+ requires at least moderate expression intensity with at least 75% distribution in tumor cells, standardized criteria preferentially with illustrated examples should be provided to define what constitutes at least moderate FOLR1 IHC intensity.
In conclusion, both serous carcinomas, HGSC and LGSC, contain a significant proportion of tumors with high FOLR1 expression. In the context of HGSC, patients are already eligible for the ADC MIRV. However, because of potential differences in payload activity across tumor types we advocate for further preclinical models preceding clinical trials of ADC’s targeting FOLR1 in recurrent LGSC. In the context of LGSC, the co-occurrence of high FOLR1 expression with specific molecular subgroups (PR low, normal MAPK pathway, p16 loss) speaks to an urgent unmet need to proceed with clinical trials for this difficult to treat disease.
ACKNOWLEDGEMENTS
This study used resources provided by the Canadian Ovarian Cancer Research Consortium’s - COEUR biobank funded by the Terry Fox Research Institute and managed and supervised by the Centre hospitalier de l’Université de Montréal (CRCHUM). The Consortium acknowledges contributions to its COEUR biobank from Institutions across Canada (for a full list see https://www.tfri.ca/coeur). Tissue collection of the Ovarian Cancer in Alberta and British Columbia (OVAL-BC) cohort was supported by NIH P30 CA118100-11. We thank Young Ou (Anatomical Pathology Research Lab, APRL, University of Calgary) for performing immunohistochemistry and Thomas Kryton for creating composite images, and the OvCare program at University of British Columbia (UBC) for the tissue microarrays (TMA) slides.
Footnotes
Funding: Funding support was also provided by the Gynecologic Cancer Initiative at the University of British Columbia (UBC), and BC Cancer, the BC Cancer Foundation (BCCF), Vancouver General Hospital (VGH)/UBC Hospital Foundation, the Janet D. Cottrelle Foundation, and Cure our Ovarian Cancer, and from the following funding agencies: Ovarian Cancer Canada/OVCAN initiative, and the Cancer Research Society. We extend our gratitude to all the patients and their families who supported this study, including the MacKenzie, Lawler, MacRae, Ho, Luther, Ludemann, and Schmid families. Martin Köbel received internal research support from Alberta Precision Laboratory RS21-608 and RS21-609.
Conflict of Interest: Martin Köbel is consultant for Helix Biopharma, Stephanie Gaillard for Organon, Mark S. Carey is a consultant for Hexamer Therapeutics and holds securities in aiGene and Hexamer Therapeutics. All outside the scope of the submitted work.
- Conceptualization: L.M., H.J., C.M.S., K.M.
- Data curation: L.Y.Y., L.M., D.A., K.M.
- Formal analysis: L.Y.Y.
- Funding acquisition: L.M., C.L.C., K.H., C.M.S., K.M.
- Investigation: L.M., D.A., K.M.
- Methodology: D.A., K.M.
- Project administration: L.M., D.A., K.H., D.G., K.M.
- Resources: L.M., C.A., C.L.C., G.C., H.J., D.A., B.M., K.H., W.N.K.Y., T.A.V., C.M.S., K.M.
- Software: L.Y.Y., K.M.
- Supervision: L.M., C.A., C.L.C., G.C., H.J., G.S., S.J., D.A., B.M., W.N.K.Y., D.G., T.A.V., C.M.S., K.M.
- Validation: G.S., S.J., T.A.V., C.M.S.
- Visualization: L.Y.Y.
- Writing - original draft: L.Y.Y., K.M.
- Writing - review & editing: L.Y.Y., L.M., C.A., D.A., C.L.C., G.C., H.J., G.S., S.J., D.A., B.M., K.H., W.N.K.Y., D.G., T.A.V., C.M.S., K.M.
SUPPLEMENTARY MATERIALS
H-score, histotype and PS2+ categories raw data
FOLR1 expression according to canonical MAPK mutation status
Normal control tissue for folate receptor alpha immunohistochemistry optimization: Tonsil with absent staining. Placenta showing membranous staining in moderate intensity. Fallopian shows luminal membranous staining, note, some but not all ciliated cells lack staining.
Representative examples of staining intensity. Weak intensity was defined as notable staining with linear membranous accentuation. Moderate intensity was defined as linear membranous staining. Strong intensity was defined as membranous staining with accompanied with similar strong cytoplasmic staining. Weak and strong intensity represent different ends of the spectrum. Difficulties in assigning a PS2+ high category especially arises in cases with a mix of low and moderate intensity regarding the decision whether there is at least 75% distribution of the moderate staining.
References
- 1.Slomovitz B, Gourley C, Carey MS, Malpica A, Shih IM, Huntsman D, et al. Low-grade serous ovarian cancer: state of the science. Gynecol Oncol. 2020;156:715–725. doi: 10.1016/j.ygyno.2019.12.033. [DOI] [PubMed] [Google Scholar]
- 2.Gershenson DM, Miller A, Brady WE, Paul J, Carty K, Rodgers W, et al. Trametinib versus standard of care in patients with recurrent low-grade serous ovarian cancer (GOG 281/LOGS): an international, randomised, open-label, multicentre, phase 2/3 trial. Lancet. 2022;399:541–553. doi: 10.1016/S0140-6736(21)02175-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Tsuchikama K, Anami Y, Ha SYY, Yamazaki CM. Exploring the next generation of antibody-drug conjugates. Nat Rev Clin Oncol. 2024;21:203–223. doi: 10.1038/s41571-023-00850-2. [DOI] [PubMed] [Google Scholar]
- 4.Matulonis UA, Lorusso D, Oaknin A, Pignata S, Dean A, Denys H, et al. Efficacy and safety of mirvetuximab soravtansine in patients with platinum-resistant ovarian cancer with high folate receptor alpha expression: results from the SORAYA study. J Clin Oncol. 2023;41:2436–2445. doi: 10.1200/JCO.22.01900. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Antony AC. Folate receptors. Annu Rev Nutr. 1996;16:501–521. doi: 10.1146/annurev.nu.16.070196.002441. [DOI] [PubMed] [Google Scholar]
- 6.Bax HJ, Chauhan J, Stavraka C, Santaolalla A, Osborn G, Khiabany A, et al. Folate receptor alpha in ovarian cancer tissue and patient serum is associated with disease burden and treatment outcomes. Br J Cancer. 2023;128:342–353. doi: 10.1038/s41416-022-02031-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Lawson BC, Marques-Piubelli ML, Westin SN, Malpica A. Folate receptor immunohistochemical staining and gynecologic tumors: initial experience with 216 cases. Int J Gynecol Pathol. 2025;44:167–173. doi: 10.1097/PGP.0000000000001053. [DOI] [PubMed] [Google Scholar]
- 8.Manning-Geist BL, Sullivan MW, Zhou Q, Iasonos A, Selenica P, Stallworth C, et al. Folate receptor alpha expression in low-grade serous ovarian cancer: exploring new therapeutic possibilities. Gynecol Oncol. 2024;188:52–57. doi: 10.1016/j.ygyno.2024.06.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rushton T, Krause HB, Samec T, Elliott A, Karnezis AN, Toboni MD, et al. Characterizing the genomic landscape through the lens of FOLR1 status in low and high grade serous ovarian carcinoma. Gynecol Oncol. 2024;191:80–85. doi: 10.1016/j.ygyno.2024.09.021. [DOI] [PubMed] [Google Scholar]
- 10.Cheasley D, Nigam A, Zethoven M, Hunter S, Etemadmoghadam D, Semple T, et al. Genomic analysis of low-grade serous ovarian carcinoma to identify key drivers and therapeutic vulnerabilities. J Pathol. 2021;253:41–54. doi: 10.1002/path.5545. [DOI] [PubMed] [Google Scholar]
- 11.Cheasley D, Fernandez ML, Köbel M, Kim H, Dawson A, Hoenisch J, et al. Molecular characterization of low-grade serous ovarian carcinoma identifies genomic aberrations according to hormone receptor expression. NPJ Precis Oncol. 2022;6:47. doi: 10.1038/s41698-022-00288-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Da-Anoy A, Kang EY, Lee CH, Cheasley D, Llaurado Fernandez M, Carey MS, et al. Molecular surrogate subtypes of ovarian and peritoneal low-grade serous carcinoma. Int J Gynecol Pathol. 2024;43:617–625. doi: 10.1097/PGP.0000000000001043. [DOI] [PubMed] [Google Scholar]
- 13.Rambau PF, Vierkant RA, Intermaggio MP, Kelemen LE, Goodman MT, Herpel E, et al. Association of p16 expression with prognosis varies across ovarian carcinoma histotypes: an Ovarian Tumor Tissue Analysis consortium study. J Pathol Clin Res. 2018;4:250–261. doi: 10.1002/cjp2.109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Köbel M, Luo L, Grevers X, Lee S, Brooks-Wilson A, Gilks CB, et al. Ovarian carcinoma histotype: strengths and limitations of integrating morphology with immunohistochemical predictions. Int J Gynecol Pathol. 2019;38:353–362. doi: 10.1097/PGP.0000000000000530. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Köbel M, Rahimi K, Rambau PF, Naugler C, Le Page C, Meunier L, et al. An immunohistochemical algorithm for ovarian carcinoma typing. Int J Gynecol Pathol. 2016;35:430–441. doi: 10.1097/PGP.0000000000000274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Llaurado Fernandez M, Dawson A, Kim H, Lam N, Russell H, Bruce M, et al. Hormone receptor expression and outcomes in low-grade serous ovarian carcinoma. Gynecol Oncol. 2020;157:12–20. doi: 10.1016/j.ygyno.2019.11.029. [DOI] [PubMed] [Google Scholar]
- 17.Le Page C, Rahimi K, Köbel M, Tonin PN, Meunier L, Portelance L, et al. Characteristics and outcome of the COEUR Canadian validation cohort for ovarian cancer biomarkers. BMC Cancer. 2018;18:347. doi: 10.1186/s12885-018-4242-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.McIntyre JB, Rambau PF, Chan A, Yap S, Morris D, Nelson GS, et al. Molecular alterations in indolent, aggressive and recurrent ovarian low-grade serous carcinoma. Histopathology. 2017;70:347–358. doi: 10.1111/his.13071. [DOI] [PubMed] [Google Scholar]
- 19.Kommoss FK, Lee CH, Tessier-Cloutier B, Gilks CB, Stewart CJ, von Deimling A, et al. Mesonephric-like adenocarcinoma harbours characteristic copy number variations and a distinct DNA methylation signature closely related to mesonephric adenocarcinoma of the cervix. J Pathol. 2024;262:4–9. doi: 10.1002/path.6217. [DOI] [PubMed] [Google Scholar]
- 20.Köbel M, Kang EY, Lee S, Ogilvie T, Terzic T, Wang L, et al. Mesonephric-type adenocarcinomas of the ovary: prevalence, diagnostic reproducibility, outcome, and value of PAX2. J Pathol Clin Res. 2024;10:e12389. doi: 10.1002/2056-4538.12389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Murali R, Selenica P, Brown DN, Cheetham RK, Chandramohan R, Claros NL, et al. Somatic genetic alterations in synchronous and metachronous low-grade serous tumours and high-grade carcinomas of the adnexa. Histopathology. 2019;74:638–650. doi: 10.1111/his.13796. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Köbel M, Yang RZ, Kang EY, Al-Shamma Z, Cook LS, Kinloch M, et al. Survey of NF1 inactivation by surrogate immunohistochemistry in ovarian carcinomas. Gynecol Oncol. 2023;178:80–88. doi: 10.1016/j.ygyno.2023.09.016. [DOI] [PubMed] [Google Scholar]
- 23.Köbel M, Kalloger SE, Boyd N, McKinney S, Mehl E, Palmer C, et al. Ovarian carcinoma subtypes are different diseases: implications for biomarker studies. PLoS Med. 2008;5:e232. doi: 10.1371/journal.pmed.0050232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Peres LC, Cushing-Haugen KL, Köbel M, Harris HR, Berchuck A, Rossing MA, et al. Invasive epithelial ovarian cancer survival by histotype and disease stage. J Natl Cancer Inst. 2019;111:60–68. doi: 10.1093/jnci/djy071. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Köbel M, Kalloger SE, Huntsman DG, Santos JL, Swenerton KD, Seidman JD, et al. Differences in tumor type in low-stage versus high-stage ovarian carcinomas. Int J Gynecol Pathol. 2010;29:203–211. doi: 10.1097/PGP.0b013e3181c042b6. [DOI] [PubMed] [Google Scholar]
- 26.Moore KN, Angelergues A, Konecny GE, García Y, Banerjee S, Lorusso D, et al. Mirvetuximab soravtansine in FRα-positive, platinum-resistant ovarian cancer. N Engl J Med. 2023;389:2162–2174. doi: 10.1056/NEJMoa2309169. [DOI] [PubMed] [Google Scholar]
- 27.Vergote I, Armstrong D, Scambia G, Teneriello M, Sehouli J, Schweizer C, et al. A randomized, double-blind, placebo-controlled, phase III study to assess efficacy and safety of weekly farletuzumab in combination with carboplatin and taxane in patients with ovarian cancer in first platinum-sensitive relapse. J Clin Oncol. 2016;34:2271–2278. doi: 10.1200/JCO.2015.63.2596. [DOI] [PubMed] [Google Scholar]
- 28.Moore KN, Oza AM, Colombo N, Oaknin A, Scambia G, Lorusso D, et al. Phase III, randomized trial of mirvetuximab soravtansine versus chemotherapy in patients with platinum-resistant ovarian cancer: primary analysis of FORWARD I. Ann Oncol. 2021;32:757–765. doi: 10.1016/j.annonc.2021.02.017. [DOI] [PubMed] [Google Scholar]
- 29.Cobb LP, Sun CC, Iyer R, Nick AM, Fleming ND, Westin SN, et al. The role of neoadjuvant chemotherapy in the management of low-grade serous carcinoma of the ovary and peritoneum: further evidence of relative chemoresistance. Gynecol Oncol. 2020;158:653–658. doi: 10.1016/j.ygyno.2020.06.498. [DOI] [PubMed] [Google Scholar]
- 30.Kang EY, Millstein J, Popovic G, Meagher NS, Bolithon A, Talhouk A, et al. MCM3 is a novel proliferation marker associated with longer survival for patients with tubo-ovarian high-grade serous carcinoma. Virchows Arch. 2022;480:855–871. doi: 10.1007/s00428-021-03232-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Crane LM, Arts HJ, van Oosten M, Low PS, van der Zee AG, van Dam GM, et al. The effect of chemotherapy on expression of folate receptor-alpha in ovarian cancer. Cell Oncol (Dordr) 2012;35:9–18. doi: 10.1007/s13402-011-0052-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Köbel M, Madore J, Ramus SJ, Clarke BA, Pharoah PD, Deen S, et al. Evidence for a time-dependent association between FOLR1 expression and survival from ovarian carcinoma: implications for clinical testing. An Ovarian Tumour Tissue Analysis consortium study. Br J Cancer. 2014;111:2297–2307. doi: 10.1038/bjc.2014.567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Kane MA, Elwood PC, Portillo RM, Antony AC, Najfeld V, Finley A, et al. Influence on immunoreactive folate-binding proteins of extracellular folate concentration in cultured human cells. J Clin Invest. 1988;81:1398–1406. doi: 10.1172/JCI113469. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Scaranti M, Cojocaru E, Banerjee S, Banerji U. Exploiting the folate receptor α in oncology. Nat Rev Clin Oncol. 2020;17:349–359. doi: 10.1038/s41571-020-0339-5. [DOI] [PubMed] [Google Scholar]
- 35.Kamen BA, Smith AK. Farletuzumab, an anti-folate receptor α antibody, does not block binding of folate or anti-folates to receptor nor does it alter the potency of anti-folates in vitro. Cancer Chemother Pharmacol. 2012;70:113–120. doi: 10.1007/s00280-012-1890-2. [DOI] [PubMed] [Google Scholar]
- 36.Sieh W, Köbel M, Longacre TA, Bowtell DD, deFazio A, Goodman MT, et al. Hormone-receptor expression and ovarian cancer survival: an Ovarian Tumor Tissue Analysis consortium study. Lancet Oncol. 2013;14:853–862. doi: 10.1016/S1470-2045(13)70253-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Manning-Geist B, Gordhandas S, Liu YL, Zhou Q, Iasonos A, Da Cruz Paula A, et al. MAPK pathway genetic alterations are associated with prolonged overall survival in low-grade serous ovarian carcinoma. Clin Cancer Res. 2022;28:4456–4465. doi: 10.1158/1078-0432.CCR-21-4183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Terry J, Torlakovic EE, Garratt J, Miller D, Köbel M, Cooper J, et al. Implementation of a Canadian external quality assurance program for breast cancer biomarkers: an initiative of Canadian Quality Control in immunohistochemistry (cIQc) and Canadian Association of Pathologists (CAP) National Standards Committee/Immunohistochemistry. Appl Immunohistochem Mol Morphol. 2009;17:375–382. doi: 10.1097/PAI.0b013e31819adacf. [DOI] [PubMed] [Google Scholar]
- 39.Martin LP, Konner JA, Moore KN, Seward SM, Matulonis UA, Perez RP, et al. Characterization of folate receptor alpha (FRα) expression in archival tumor and biopsy samples from relapsed epithelial ovarian cancer patients: a phase I expansion study of the FRα-targeting antibody-drug conjugate mirvetuximab soravtansine. Gynecol Oncol. 2017;147:402–407. doi: 10.1016/j.ygyno.2017.08.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
H-score, histotype and PS2+ categories raw data
FOLR1 expression according to canonical MAPK mutation status
Normal control tissue for folate receptor alpha immunohistochemistry optimization: Tonsil with absent staining. Placenta showing membranous staining in moderate intensity. Fallopian shows luminal membranous staining, note, some but not all ciliated cells lack staining.
Representative examples of staining intensity. Weak intensity was defined as notable staining with linear membranous accentuation. Moderate intensity was defined as linear membranous staining. Strong intensity was defined as membranous staining with accompanied with similar strong cytoplasmic staining. Weak and strong intensity represent different ends of the spectrum. Difficulties in assigning a PS2+ high category especially arises in cases with a mix of low and moderate intensity regarding the decision whether there is at least 75% distribution of the moderate staining.