Functional validation of somatic variability in TP53 and KRAS for prediction of platinum sensitivity and prognosis in epithelial ovarian carcinoma patients

TL;DR

This study validates the role of TP53 and KRAS mutations in predicting platinum sensitivity and prognosis in ovarian cancer patients.

Contribution

The study functionally validates somatic variability in TP53 and KRAS and links it to clinical outcomes in epithelial ovarian carcinoma.

Findings

KRAS mutations are associated with earlier FIGO stages and non-high-grade serous subtypes.

TP53 variants disrupting the DNA binding loop correlate with longer platinum-free intervals.

Co-mutated TP53 and KRAS genes are linked to poorer overall survival.

Abstract

Concerning the dismal prognosis of chemoresistant patients with epithelial ovarian carcinoma (EOC), we aimed to follow up the findings of a previous whole-exome sequencing study using an orthogonal Sanger sequencing on the same patients and a separate set of 127 EOC patients (N = 177, all fresh frozen tumor samples). We focused on TP53 as a frequently mutated gene relevant for chemosensitivity, included KRAS as an additional therapeutically relevant target, complemented the study with transcript levels of both genes, and compared results with clinical parameters. All variants in TP53 and KRAS detected by exome sequencing were confirmed. KRAS mutated patients had significantly more frequent FIGO stages I or II (p = .002) and other than high-grade serous tumor subtypes (nonHGSCs) (p < .001), which was connected with lower KRAS transcript levels (p = .004). Patients with nonHGSC subtypes had less frequent TP53 mutations (p = .002). Carriers of TP53 variants disrupting the DNA binding loop had significantly longer platinum-free intervals than the rest (p = .037). Tumors bearing nonsense, frameshift, or splice site TP53 variants had a significantly lower TP53 transcript level, while those with missense variants had significantly higher levels than wild types (p < .001). The normalized intratumoral TP53 and KRAS transcript levels were correlated, and patients with co-mutated genes had poorer overall survival than others (p = .015). Protein levels of both genes significantly correlated with their respective transcripts (p = .028 and p = .001, respectively). Our study points to KRAS as a target for future therapy of nonHGSCs and reveals the prognostic value of TP53 variants in the DNA binding loop.

Figures

E](#f0001)) compared to those with other EOC subtypes. On the other hand, KRAS transcript level did not associate with its mutation status (*p* = .181, [Figure 1F](#f0001)). *KRAS* mutation status, spectra, or transcript level were not significantly associated with the rest of the clinical parameters (age, grade of tumor, surgical radicality, chemosensitivity status, or PFI and OS, all *p* > .05), and this was true for the association between transcript level and stage as well.Table 3.Associations between *KRAS* mutational status and stage or tumor subtype of EOC patients.Characteristics*KRAS*

A](#f0002)). Patients with co-mutated *TP53-KRAS* had significantly worse OS than wild-type patients or those with a single gene mutation (*p* = .015, [Figure 2B](#f0002)), but the PFI was not significantly affected (*p* = .065, [Figure 2C](#f0002)). Figure 2.Associations between patient survival and carriage of *TP53* or *KRAS* mutations.(A) Platinum-free interval stratified by carriage of *TP53* DNA binding loop mutations in EOC patients and (B) overall survival and platinum-free interval (C) in *TP53-KRAS* co-mutated patients compared to wild-type or single gene mutated EOC patients.

B](#f0003)). For *KRAS*, no significant association of transcript expression with mutation spectra was found (*p* > .05, [Figure 3C](#f0003)), perhaps due to the low number of observations (*n* = 3 mutated samples). A weak, non-significant correlation was observed between TP53 and KRAS transcripts (*p* = .052, [Figure 3D](#f0003)). In terms of available clinical data, neither grade (G1 or G2 versus G3 or G4) nor stage (stage I or II versus III or IV) were significantly associated with KRAS or TP53 transcript level (*p* > .05, data not shown).

Full text

Introduction

Epithelial ovarian cancer (EOC), recognized as the eighth leading cause of cancer-related death among women, stands out as one of the most lethal gynecological malignancies.^1^ Early detection of EOC is a challenge, because, in most cases, it is asymptomatic in the early stages (I or II based on The International Federation of Gynecology and Obstetrics, FIGO, guidelines). Typically, 75% of EOC cases occur at the advanced stage (III or IV), where the 5-y survival rate is approximately 20–45%, compared to 40–70% for stages I or II.^2,3^ The standard treatment for advanced EOC has been primary debulking surgery followed by chemotherapy (platinum derivatives with paclitaxel) for the majority of cases.^4^ However, most patients experience a relapse within the first 5 y after the initial diagnosis, with only 20–25% achieving cure.^5^

Morphologically, EOCs are classified into four major subtypes: serous, endometrioid, clear cell, and mucinous.^6^ Additionally, they can be divided into two primary types: type I, including endometrioid, mucinous, clear cell, and low-grade serous ovarian carcinomas (LGSCs), and type II, constituting 70% of the total and encompassing high-grade serous ovarian carcinomas (HGSCs), carcinosarcomas, and undifferentiated carcinomas. These classifications are integral to defining the aggressiveness of the cancer and its response to different chemotherapies.^3^ A significant majority, exceeding 80% of identified EOC cases, fall under the histological classification of HGSC, characterized by an aggressive phenotype that correlates with elevated mortality rate,^7^ which is attributed not only to diagnosis at the advanced stage but also to chemoresistance, where approximately 50% of the cases diagnosed at advanced stage relapse within the first 5 y.^3^

The introduction of Poly(ADP-ribose) polymerase (PARP) inhibitors like olaparib and antiangiogenic agents, such as bevacizumab or pazopanib, has led to a significant improvement in the prognosis of the patients.^8,9^ PARP inhibitors (PARPis) are primarily used for maintenance therapy for platinum-sensitive advanced EOCs.^10^ Moreover, patients with BRCA1/BRCA2 mutations demonstrate enhanced sensitivity to treatment with PARPi.^11^

Next Generation Sequencing (NGS) enables the detection of genetic variability and its linkage to multidrug resistance. Based on genomic profiling, two major EOC types have been defined. EOC of type I is characterized by mutations in the MAPK pathway (KRAS, BRAF, PTEN, and CTNNB1, etc.) and type II mutations in TP53, BRCA1, BRCA2, KIT, and EGFR.^12^ Moreover, DNA damage response and related alterations in DNA repair pathways play a crucial role in cancer development, including EOC. Germline mutations in DNA repair genes can predict hereditary forms of cancer, particularly BRCA1/2 mutations in breast and ovarian cancers.^13^ Pathogenic somatic mutations in genes from the homologous recombination DNA repair pathway, such as BRCA1/2, ATM, RAD51C, and RAD51D, were implicated in chemosensitivity and prognosis of EOC patients.^14,15^ HGSC typically shows very high frequency of somatic TP53 mutations (~90%) and genomic heterogeneity.^16^

A previous whole exome sequencing study^14^ confirmed TP53 as the most frequently mutated gene in HGSC and EOC in general and suggested its relevance for chemosensitivity. Moreover, the followed sample set enriched with clear cell, mucinous, endometrioid, and low grade serous carcinomas has quite a high frequency of mutations in KRAS (overall 12%),^14^ suggesting its relevance for therapeutic decisions. The present follow-up study uses direct Sanger sequencing of the same and an extended sample sets to demonstrate the robustness of TP53 and KRAS mutation detection, provide validation of study results, and substantiate their further use in clinical management of patients. We also complement somatic mutation screening with an assessment of both genes’ transcript levels in tumor RNA and protein levels in selected cases. We compare the results with sensitivity to EOC therapy and patient survival for evaluation of the prognostic value of these biomarkers. Our study adds another dimension to exome or genome sequencing-based EOC projects published before.^14,17–19^

Results

Patients’ characteristics

The main characteristics of all patients (N = 177) are in Table 1. The median age of patients at the time of diagnosis was 62 y (range 24–89). Most patients presented with FIGO stage III (82%), grade G3 (85%), and HGSC subtype (84%). About one-third of patients (32%) underwent preoperative chemotherapy, and half of patients (50%) had disease residuum left after surgical tumor debulking. The vast majority of patients (96%) received platinum-based chemotherapy regimens in an adjuvant setting, two received taxane monotherapy, four did not receive any adjuvant treatment due to poor performance status, and for six patients the information about therapy was not available. The median PFI and OS were 25 and 48 months, respectively. Patients with FIGO stage III or IV, residuum after surgery (R1 or R2), or with PFI < 12 months had significantly poorer OS than the rest of the patients (p < .001 for all) (Online Resource 1A-C).Table 1.Clinical characteristics of EOC patients.ParametersNumber of patientsPercentageAge at diagnosis177100Median ± SD (years)62.0 ± 11.5 FIGO stage  I127II106III13982IV85Data not available8––Histologic grade (G)  G1116G2159G314685Gx5––Tumor subtype  HGSC14384Other*2816Data not available6––Distant metastasis  Absent16195Present85Data not available8––Neoadjuvant chemotherapy  Administered5732Not administered12068Residuum after surgery  Present8550Absent (R0)8550Data not available7––Adjuvant chemotherapy  Platinum-based^#^16596Taxane monotherapy21Not administered43Data not available6––Chemosensitivity status  Resistant (PFI ≤6 months)3823Intermediate (PFI 7–11 months)2415Sensitive (PFI ≥12 months)10162Data not available14––Platinum-free interval16895Median ±95% confidence interval (months)2517.9–32.1Overall survival**16895Median ±95% confidence interval (months)4837.7–58.3Footnotes.*Other subtypes include the following carcinomas: mucinous (n = 9), clear cell (n = 10), low grade serous (n = 5), endometrioid (n = 2), and borderline (n = 2).**Includes all ratings above R0 (R1, R2, unspecified).^#^Platinum-based chemotherapy regimens include n = 147 taxane (paclitaxel/docetaxel) with platinum (carboplatin/cisplatin), n = 9 platinum monotherapy, other (n = 1 FOLFOX, n = 1 platinum with anthracycline, n = 1 platinum with paclitaxel and anthracycline, and n = 6 platinum with paclitaxel and cyclophosphamide). FOLFOX = 5-fluorouracil, leucovorin, and oxaliplatin.

Somatic genetic variability

All six KRAS variants found previously by exome sequencing (n = 50) were also detected by Sanger sequencing in the confirmation part of the study (n = 50). In the extended validation part (n = 125, two samples not assessed due to the lack of DNA), variants in a further nine samples were observed (Table 2). All variants were missense single nucleotide substitutions in exon 2 (n = 12) or 3 (n = 3). Representative chromatograms are in Online Resource 2A, B.Table 2.Molecular characteristics of EOC patients.GeneNumber of patientsPercentageKRAS mutation status*   KRAS wild-type16193 KRAS mutated158KRAS mutation spectrum   p.Gly12Asp533 p.Gly12Val533 p.Gln61His213 p.Gly12Cys17 p.Gly12Ala17 p.Gln61Arg17TP53 mutation status   TP53 wild-type7241 TP53 mutated10559TP53 mutation spectrum  Hotspots   p.Arg175His8  p.Tyr220Cys7  p.Arg273His6  p.Arg248Gln5  p.Arg248Trp4  p.Arg282Trp4  p.His214Arg3  p.His179Gln2  p.Arg249Trp2  p.Cys275Phe2  p.Gly279Glu2  p.Glu198Ter3  p.Arg213Ter3 Private missense mutations32 Private frameshift or nonsense mutations17 Private splice site mutations with pathogenic features5 TP53 mutation functional consequences^#^   Loss-of-function8792 Gain-of-function88 Not classified10––Dominant-negative effect (DNE) & loss-of-function (LOF) properties^#^  DNE_LOF6686 notDNE_notLOF57 notDNE_LOF57 Not classified29––Transactivation function^#^   non-functional6490 functional or partially functional710Not classified34––DNA binding loop affected^#^   yes9494 no66Not applicable5–Footnotes.*Result for two samples not available due to DNA of low quality/quantity.^#^Evaluated using The TP53 database of NCI (https://tp53.isb-cgc.org/.) and The Clinical Knowledgebase (https://ckb-core.genomenon.com/) and literature cited therein.

As for TP53, the confirmation set showed exactly the same variants compared to exome sequencing, i.e., 39 mutated and 10 wild-type patients. In the validation set (n = 127), an additional 66 mutated samples were identified (Table 2). Representative chromatograms are presented in Online Resource 3A-K.

Functional classifications enabled the distribution of TP53 variants to several categories: i/missense (n = 77), out of which 32 were single private mutations and the rest mutational hotspots detected in two or more patients, ii/two hotspot nonsense variants present in three patients each, and iii/private frameshifts or nonsense variants (n = 17). The last category was splice site variants with pathogenic features, which were all private (n = 5). The TP53 database of NCI and The Clinical Knowledgebase (CKB) enabled more detailed stratification of variants into loss-of-function (n = 87) versus gain-of-function (n = 8) variants. Most of the somatic variants were classified as having the following properties: dominant-negative effect or loss-of-function (n = 66), nonfunctional transactivation (n = 64), and affecting DNA binding loop (n = 94) by these databases (Table 2, Online Resource 4).

Four patients carried mutations in both TP53 and KRAS (co-mutations). One patient with the HGSC subtype had the combination of TP53-Arg282Trp with KRAS-Gln61His, chemoresistant status, and OS of 16 months. The second patient had a clear cell subtype, TP53-Arg248Gln with KRAS-Gln61Arg, chemoresistant status, and extremely short OS of 7 months. The third patient with the mucinous subtype had TP53-Arg213Ter with KRAS-Gly12Asp mutation combination, chemoresistant status, and OS of 19 months. The last patient had TP53-Cys135Trp with KRAS-Gly12Val, HGSC subtype, chemosensitive status, and OS of 43 months. Thus, carriage of TP53-KRAS co-mutations could be associated with chemoresistance and poor patient prognosis in most cases observed.

All subsequent clinical genomic analyses were performed using the combined confirmation and validation cohorts (N = 177).

Intratumoral KRAS and TP53 transcript and protein levels

To provide additional functional evidence, we analyzed by qPCR the TP53 and KRAS transcript levels in all available tumor samples together with genetic information. Five samples could not be determined due to low RNA quantity or quality and no tissue left. No extreme outliers were observed.

The carriage or type of KRAS mutations did not significantly associate with the KRAS or TP53 transcript levels (p > .05). On the other hand, a significantly lower TP53 transcript level in tumors bearing nonsense, frameshift, or splice site types of variants compared to wild-type TP53 was observed (p < .001, Figure 1A). In contrast, tumors with missense TP53 variants had significantly higher transcript levels than wild-type ones (p < .001, Figure 1A). Higher TP53 transcript level was found in tumors with TP53 variants classified as gain-of-function compared to loss-of-function (p = .027, Figure 1B). Figure 1.Associations of TP53 and KRAS normalized transcript levels in tumors with characteristics of EOC patients.(A) TP53 normalized transcript level with TP53 mutation type or (B) effect (gain-of-function vs. loss-of-function). (C) Mutual correlation between KRAS and TP53 transcript levels. (D) Four membranes representing KRAS and p53 protein levels with GAPDH loading control and Panc-1 cell line for inter-membrane normalization. (E) KRAS normalized transcript level with EOC subtype and (F) mutation status.High impact means nonsense, frameshift, or splice site functional variant classification.Footnote: *5 high impact and 4 missense variants could not be classified and are excluded from the plot.

Carriage of co-mutated TP53-KRAS did not affect transcript expression (p = .096 for TP53 and p = .306 for KRAS).

Interestingly, the normalized intratumoral TP53 and KRAS transcript levels were mutually significantly correlated (ρ = 0.384, p < .001, Figure 1C). Additionally, the protein level of KRAS (Figure 1D), analyzed by immunoblotting, significantly correlated with that of transcript (n = 52, ρ = 0.305 and p = .028) and for p53 protein (Figure 1D) with transcript such correlation was even stronger (n = 52, ρ = 0.431 and p = .001). KRAS and p53 protein levels did not correlate at all (p = .804).

Associations of somatic genetic variability and transcript levels with clinical data of patients

Afterward, we performed statistical analysis of associations between transcript levels, mutational status, spectra, and functional classifications of both genes and clinical data of patients.

Patients with FIGO stage I or II had significantly more frequently mutated KRAS compared to stage III or IV patients (p = .002, Table 3). On the other hand, patients with the HGSC tumor subtype had significantly less frequently mutated KRAS and TP53 (p < .001 and p = .002, respectively; Table 3), and they had significantly higher KRAS transcript levels (p = .004, Figure 1E) compared to those with other EOC subtypes. On the other hand, KRAS transcript level did not associate with its mutation status (p = .181, Figure 1F). KRAS mutation status, spectra, or transcript level were not significantly associated with the rest of the clinical parameters (age, grade of tumor, surgical radicality, chemosensitivity status, or PFI and OS, all p > .05), and this was true for the association between transcript level and stage as well.Table 3.Associations between KRAS mutational status and stage or tumor subtype of EOC patients.CharacteristicsKRAS wild-typeKRAS* mutated*p-valueStage I/II166.002Stage III/IV1397 HGSC1394 < .001other subtypes1810  TP53wild-typeTP53*mutated*p-valueHGSC5093.002other subtypes199 Footnotes.*Numbers of patients; for some patients clinical data or KRAS mutation status were not available.

As for TP53, its transcript level, mutation status, spectra, or functional classifications were not significantly associated with any of the clinical parameters (age, stage, grade of tumor, subtype, surgical radicality, chemosensitivity status, or OS, all p > .05). However, the PFI of patients with TP53 mutations non-disrupting the DNA binding loop (n = 6) was significantly poorer than that of patients with mutations located in the domain (p = .037, Figure 2A). Patients with co-mutated TP53-KRAS had significantly worse OS than wild-type patients or those with a single gene mutation (p = .015, Figure 2B), but the PFI was not significantly affected (p = .065, Figure 2C). Figure 2.Associations between patient survival and carriage of TP53 or KRAS mutations.(A) Platinum-free interval stratified by carriage of TP53 DNA binding loop mutations in EOC patients and (B) overall survival and platinum-free interval (C) in TP53-KRAS co-mutated patients compared to wild-type or single gene mutated EOC patients.

We further performed patient stratification into HGSC and nonHGSC subgroups, given the importance of the EOC subtype in previous analyses. Patients with nonHGSC subtypes had significantly more often less advanced stages I/II than HGSCs (p < .001, Online Resource 5) and thus less aggressive disease. However, only the PFI of patients with clear cell subtype (n = 10) was any better than that of HGSC (n = 134), while for mucinous (n = 9) or LGSC (n = 5), it was not, and endometrioid patients had worst PFI (n = 2) (Online Resource 6A). The above differences between subtypes had no prognostic consequences (Online Resource 6B).

No significant associations with clinical data were identified for KRAS or TP53 transcript levels, mutations, or their functional classifications in the HGSC subgroup (n = 143). However, patients with nonHGSC subtypes (n = 28) bearing any TP53 mutations had non-significantly poorer PFI than patients with the wild-type (p = .062). No association was found for OS or other clinical data, including chemosensitivity status.

Validation using external datasets

Finally, we attempted to validate our findings using the largest and most up-to-date publicly available EOC dataset within the GENIE project (n = 2210).

The TP53 and KRAS mutation analysis confirmed the overrepresentation of KRAS mutations in nonHGSC compared to HGSC cases. In our dataset, 32% of nonHGSC patients harbored KRAS mutations, while only 1.4% of HGSC cases had such alterations. Size of the GENIE dataset allowed the analysis of the distribution of KRAS mutations across all major nonHGSC subtypes. The frequency of KRAS mutations raised in the trend HGSC (1.2%) < < clear cell (13%) < endometrioid (27%) = LGSC (28%) < mucinous (67%). Even more interesting was the trend in the ratio of TP53/KRAS mutability among subtypes, where mucinous and clear cell cases had a 1/1 ratio, while endometrioid and LGSC subtypes had more KRAS than TP53 mutations. Patients with the HGSC subtype had a ratio close to 100/1 in favor of TP53. Most interestingly, analysis of the GENIE dataset revealed a considerable fraction of TP53-KRAS co-mutated patients, again with a high heterogeneity across subtypes. Almost half (46%) of patients with the mucinous subtype had both genes mutated. The other subtypes had a much lower proportion of such events, 4% for endometrioid and 1.5% for clear cell EOC. The occurrence of this phenomenon in LGSC and HGSC was comparable and less than 1% in both cases (Figure 3A). Figure 3.Validation of TP53 and KRAS mutational spectra and intratumoral transcript levels in EOC subtypes using external datasets.(A) TP53 and KRAS mutational spectra in EOC subtypes. (B) TP53 and (C) KRAS intratumoral transcript levels stratified their by mutation type or classification. (D) Correlation between TP53 and KRAS intratumoral transcript levels (lack of mutual correlation).

As GENIE does not contain expression data, we used the TCGA-OV dataset (n = 374) for the assessment of transcript levels. The comparison of TP53 transcript levels with main mutation classification groups confirmed the trend observed in our study, i.e., significantly higher level in tumors harboring missense mutations (p = .002) and lower in those with nonsense, frameshift, or splicing mutations of (p = .009) compared to wild-type (Figure 3B). For KRAS, no significant association of transcript expression with mutation spectra was found (p > .05, Figure 3C), perhaps due to the low number of observations (n = 3 mutated samples). A weak, non-significant correlation was observed between TP53 and KRAS transcripts (p = .052, Figure 3D). In terms of available clinical data, neither grade (G1 or G2 versus G3 or G4) nor stage (stage I or II versus III or IV) were significantly associated with KRAS or TP53 transcript level (p > .05, data not shown).

Only two KRAS-TP53 co-mutations were found in the TCGA dataset. One patient (TCGA-29–1696-01A) had KRAS Gly12Arg with TP53 frameshift co-mutation, stage IIIC, G2, and died 34 months after diagnosis. The second patient (TCGA-61–2009-01A) had KRAS Glu61Leu with TP53 missense co-mutation, stage IIIC, G3, and was alive 40 months after diagnosis. Thus, external data do not seem to corroborate our observation of the considerably poorer prognosis of the three EOC patients with such co-mutations.

Due to the absence of survival data in the public version of GENIE and the lack of histopathologically confirmed subtype stratification in the TCGA dataset (however, all tumors were serous), we could not attempt the validation of our prognostic associations.

Discussion

The present study confirmed by a gold standard direct sequencing method the presence of mutations in clinically actionable EOC oncodrivers TP53 and KRAS, reported by previous whole exome sequencing of 50 patients. We further screened both genes in the additional sample set of 127 EOC patients and complemented the somatic genotype with transcript expression data with the intent to provide deeper functional insight.

In general, results show that the functional classification of mutations and disease subtype context matters much more than carrier status alone. These aspects need careful investigation before any sensible clinical exploitation. Specifically, HGSC differs from other EOC subtypes by dominance of TP53 mutations, while KRAS mutations are more relevant to nonHGSC subtypes. This observation agrees with the generally accepted view^20^ and may have clinical consequences.^21^ The prevalence of KRAS mutations in patients with less advanced stages I/II found by us complies with the fact that some nonHGSC subtypes are more frequently diagnosed with less advanced disease than HGSC ones.^22^ Nevertheless, this was reflected by non-significantly better PFI only for clear cell subtype and just mild effect on general prognosis in terms of OS. However, the clear cell subtype is considered less sensitive to platinum-based chemotherapy than other EOCs^23^ suggesting that our observation may be due to the low number of samples evaluated (n = 10). The striking prevalence of KRAS mutations in nonHGSC subtypes calls for their integration into clinical trials with future KRAS inhibitors. Studies with KRAS-G12D (MRTX1133)^24^ or pan-KRAS^25^ inhibitors show promising results and the KRAS-G12C mutation inhibitors have already been approved for the personalization of therapy. Sotorasib has been approved for targeted therapy of non-small cell lung cancer (NSCLC)^26^ and adagrasib for NSCLC and colorectal carcinoma.^27^

The relevance of so-called loss-of-function or gain-of-function TP53 variants^28^ for disease progression and potential therapy targeting was discussed before.^29^ Therefore, we used The TP53 database of and The Clinical Knowledgebase for the stratification of patients into subgroups with variants classified according to different functional effects (Table 2). Except for variants disrupting the DNA binding loop, being predictive for prolonged PFI, other stratifications had no clinical consequences.

Perhaps, the most important result for contemporary considerations on targeted therapy and immunotherapy appeared after the analysis of KRAS-TP53 co-mutated tumors. Previous studies in NSCLC^30,31^ or pancreatic carcinoma^32^ reported contradicting results. Despite both author groups demonstrating the dismal prognosis of patients harboring such alterations, the immunologically “hot” status has been claimed for NSCLC, while the “cold” status for pancreatic carcinoma with presumed consequences for the results of eventual immune checkpoint blockade therapy. Our study shows that EOC may be subject to further research in this area as co-mutated patients in the present study (n = 4) had significantly worsened OS, moving EOC closer in this to pancreatic cancer.

The analysis of intratumoral TP53 and KRAS transcript levels has shown several significant results. First, they were mutually significantly correlated. Second, TP53 transcript level well mirrored its mutation status, namely tumors of missense variants carriers had higher, while those with other mutation types, classified as having a high predicted functional effect, lower levels compared to the wild-type (p < .001). Accordingly, variants classified as loss-of-function had lower level than those with the gain-of-function status. Finally, yet importantly, nonHGSC tumors had significantly lower KRAS transcript levels than HGSC ones, while no correlation between mutation status and transcript level was found in this subgroup suggesting involvement of a more complex mechanism, e.g., epigenetics involvement. Both KRAS and p53 protein levels correlated with their respective transcript levels, further substantiating the functional relevance of transcript analysis.

External validation using the GENIE dataset (n = 2210)^33^ helped to validate the observed correlation between TP53 mutation types and transcript expression. More importantly, it enabled a more precise evaluation of the distribution of TP53 and KRAS mutation status across EOC subtypes. This analysis revealed an increasing trend in the ratio TP53/KRAS mutability: HGSC > > clear cell ≈ mucinous > endometrioid > > LGSC. An even more striking disproportion in TP53-KRAS co-mutation frequency: mucinous (53/115) > > endometrioid (9/223) > clear cell (4/265) > HGSC (14/1467) > LGSC (1/140) was apparent. Despite the common occurrence of co-mutations in the mucinous subtype was already described,^34^ both trends suggest enormous variability among EOC subtypes, which calls for exploitation in individualized therapy.

Several limitations of the present study need to be mentioned. Firstly, the sample size precludes robust analysis of nonHGSC subtypes, which are very rare (<10% of EOC each). Unfortunately, due to serious constraints in clinical data availability in both GENIE (missing survival)^33^ and TCGA (missing EOC subtype), we could not externally validate our potentially clinically relevant results. Thus, more studies are necessary for this area and our study may contribute to meta-analyses. Second, patients analyzed in this study were untreated with PARPi or other targeted drugs. Platinum-based chemotherapy is considered the standard of care in EOC,^35^ and chemosensitivity to platinum and PARPi overlaps.^36,37^ Thus, our sample set is still relevant from this point of view. Lastly, transcript levels may not robustly correlate with protein levels as assessed by immunohistochemistry. However, while immunohistochemistry is routinely clinically used for TP53 assessment, it is not in the case of KRAS, where just the mutation status is considered for EGFR blockade therapy, and functional approval, especially for rarely occurring variants, is missing. Our study shows that this area needs further attention. Finally, the present study has clear benefits in ethnical homogeneity of the patient population, unified therapy regimen, and long-term complete clinical follow-up.

In conclusion, our study confirms previous data on KRAS as a valid and hopefully soon druggable target for nonHGSC EOCs and identifies the prognostic value of TP53 mutations in the DNA binding loop for a fraction of patients. Furthermore, we describe an intriguing enrichment of TP53-KRAS co-mutations in the mucinous subtype of EOC based on the analysis of an external dataset of 2210 samples. Our results further extend the area of precision oncology of EOC and suggest directions for future functional and preclinical studies.

Patients and methods

Patients

For this study, we used samples of surgically resected, primary EOC tumors from 50 patients (confirmation set) with available whole exome data^14^ and additional 127 EOC patients (validation set) without exome data. Patients were prospectively recruited at University Hospitals Motol, Královské Vinohrady (both in Prague, the Czech Republic), and Pilsen (the Czech Republic) between 2009 and 2020. Tumor samples were collected fresh and promptly frozen and stored at −80°C until isolation of nucleic acids. Peripheral blood samples were taken from all patients to enable tumor-normal matched analysis.

Collaborating clinicians collected the following clinical data on each patient: age at diagnosis, FIGO stage (pTNM), the histological subtype and grade of the tumor, presence of distant metastasis or residuum after surgery, oncological treatments, chemosensitivity status, and overall survival (OS) from medical records. The chemosensitivity status was based on the platinum-free interval (PFI) measured as the time from the end of the platinum-based adjuvant chemotherapy to disease recurrence or progression.^38^ Patients having PFI ≤ 6 months were considered platinum-resistant and patients with PFI ≥ 12 months platinum-sensitive. Several patients had PFI in the range of 7–12 months and were classified as partially platinum-sensitive. These patients were tentatively included in the resistant group and all association analyses were performed both with and without them. Consensual results are provided. The OS was defined as the time elapsed between surgical resection and death of any cause or patient censoring. Detailed clinical characteristics of the patients are in Table 1.

Experimental protocol of the study was approved by the Institutional Review Boards of the National Institute of Public Health in Prague (approval reference no. IGA NS9803–4 of 2 February 2008), University Hospital Motol (approval reference no. EK-890/15 of 24 June 2015), University Hospital Královské Vinohrady (approval reference no. EK-VP/40/0/2017 of 28 June 2017), and University Hospital Pilsen (approval reference no. 16-29013A of 4 June 2015). All patients included in the study read and signed the Informed Consent of the Patient.

Isolation of nucleic acids and cDNA synthesis

DNA from peripheral blood lymphocytes was isolated using the DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany). Processing tumor tissue samples involved grinding them into a fine powder using a mortar and pestle under liquid nitrogen. Subsequently, we utilized the AllPrep DNA/RNA/Protein Mini Kit (Qiagen) according to the manufacturer’s protocol for the isolation of total RNA and DNA. The quantity of the RNA and DNA samples was assessed using the Qubit 4 Nucleic Acid Fluorometric Quantification System (ThermoFisher Scientific, Waltham, MA, USA) and quality was checked by measuring the integrity number (RIN and DIN) using Agilent TapeStation 2200 (ThermoFisher Scientific). RNA was transcribed into cDNA with the help of the RevertAid™ First Strand cDNA Synthesis kit (ThermoFisher Scientific) according to the manufacturer’s protocol and checked using the previously published method.^39^

Gene expression analysis

Quantitative real-time PCR (qPCR) was performed using TaqMan® Gene Expression Assays (ThermoFisher), namely TP53 (Hs01034249_m1) and KRAS (Hs00364284_g1). PPIA (Hs99999904_m1), UBC (Hs00824723_m1), and YWHAZ (Hs03044281_g1), selected previously using NormFinder and geNorm software, served as reference genes for results normalization.^40^ The reaction mixture with volume of 5 µL contained 1 µL of 5× Hot FirePol Probe qPCR Mix Plus (ROX) (Solis BioDyne OÜ, Tartu, Estonia), 0.25 µL of 20× TaqMan® Gene Expression Assay specified above, 1.75 µL of nuclease-free water, and 2 µL of 8-times diluted cDNA. qPCR reactions were performed in a 384-well block of the ViiA7 Real-Time PCR System and evaluated using the ViiA7 System Software (Life Technologies, Carlsbad, CA, USA). Cycling parameters were initially held at 50 ^◦^C for 2 min and 10 min denaturation at 95 ^◦^C, followed by 45 cycles consisting of 15 s of denaturation at 95 ^◦^C and 60 s of annealing/extension at 60 ^◦^C. The non-template control contained water instead of cDNA and negative cDNA synthesis controls (RNA transcribed without reverse transcriptase) were employed to control carry-over contamination. All samples were analyzed in duplicates and samples with a standard deviation > 0.5 Ct between replicates were re-analyzed. The qPCR process adhered to the Minimum Information for Publication of Quantitative Real-Time PCR Experiments Guidelines (MIQE).^41^

Differences between samples and groups of patients were calculated from raw Ct values with the comparative Ct method described previously.^42^ The 2^−∆Ct^ method was used for relative quantification of gene expression, and the 2^−∆∆Ct^ method was used for fold change calculation in groups divided by clinical data or mutation carriage or classification.

Direct sequencing

Exons 2 and 3 of KRAS and 4–10 of TP53 were subjected to direct sequencing using the Sanger method. Briefly, DNA was amplified between oligonucleotide primer pairs specific for each amplicon (Online Resource 7) using regular PCR and after product length verification on agarose gel purified by ethanol precipitation. Each reaction was optimized to produce a strong single-band product. Sequencing reactions were then performed using the BigDye Terminator v3.1 Cycle Sequencing Kit (Invitrogen) with approximately 10 ng of PCR product and 2 pmol of sequencing primer in 10 µl final reaction volume according to the producer’s protocol. Separate sequencing reactions were run with both forward and reverse sequencing primers (Online Resource 7). The acquired products were purified using ExoSAP-IT™ PCR Product Cleanup Reagent (Applied Biosystems, Foster City, CA). DNA sequencing was performed by a capillary electrophoresis-based system commercially (SEQme, s.r.o., Dobris, the Czech Republic). Raw results were evaluated by BioEdit 7.2.5 program and Sequencing Analysis Software v5.2 (Applied Biosystems). Oligonucleotide primers (Online Resource 7) for sequencing were designed using the Primer3 software.^43^

Immunoblotting

Fifty-two tumor samples representing highest and lowest transcript levels and wide spectra of mutations in KRAS and TP53 were selected for protein study. Proteins isolated from patient tissues using the AllPrep DNA/RNA/Protein Mini Kit (Qiagen) were quantified using the Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific). In addition to patient tissue samples, we added a protein sample from the Panc-1 cell line to each of the four membranes. Panc-1 has been previously tested for high expression of KRAS and p53 and served as a standard for inter-membrane normalization of detected expression levels. A total of 20 μg of protein were loaded onto a 12% (w/v) polyacrylamide gel, separated by SDS-PAGE, and subsequently transferred onto a nitrocellulose membrane. Following transfer, the membrane was cut into two sections. The upper section was blocked for 60 min in Blocker™ BLOTTO in TBS buffer (Thermo Fisher Scientific) and then incubated overnight at 4°C with a multiplex combination of primary antibodies against p53 (mouse monoclonal antibody, A10610, Abclonal Science, Düsseldorf, Germany) and GAPDH (used as a loading control; rabbit monoclonal antibody, 14C10, #2118, Cell Signaling Technology, MA, USA), both diluted in Blocker™ BLOTTO in TBS containing 0.2% (v/v) Tween®20. The lower section was incubated with a primary antibody against KRAS (rabbit polyclonal antibody, GTX100636, GeneTex, CA, USA) under the same conditions. After rinsing with TBS containing 0.1% (v/v) Tween®20, the membrane sections were incubated for 60 min at room temperature with secondary antibodies. The upper section was probed with IRDye® 800CW Goat anti-Mouse IgG and IRDye® 680RD Goat anti-Rabbit IgG (Li-Cor, Lincoln, NE, USA), while the lower section was incubated with IRDye® 800CW Goat anti-Rabbit IgG. Proteins were visualized using the Odyssey® Fc Imaging System (Li-Cor) and quantified using Image Studio version 4.0.21 (Li-Cor). Membranes with areas extracted for Figure 1D preparation are in Online Resource 8.

External datasets

For validation of somatic variants in TP53 and KRAS, the American Association for Cancer Research (AACR) Genomics Evidence Neoplasia Information Exchange (GENIE) 15.0-public release dataset (released on 1 February 2024), composed of tumor panel sequencing data from multiple major cancer centers, was utilized.^33^ Only samples fulfilling the following criteria were used: EOC, primary tumor, any somatic mutation data found after matching by sample ID, and gene of interest (TP53 and/or KRAS) included in the respective sequencing panel (final dataset: n = 2210). For validation of expression levels and mutation data, we used the RNAseq gene expression (FPKM-UQ normalized) and DNAseq mutation data of the GDC TCGA-OV cohort (Mutect2 pipeline), downloaded from the University of California Santa Cruz Xenabrowser portal (https://xenabrowser.net.), which were then filtered to only primary ovarian tumors (n = 374). The dataset does not contain minority subtypes, nor detailed histopathological annotation, with all samples being classified merely as serous, and therefore, subtype-sensitive analyses were not performed using this dataset.

Statistical analysis

Associations of categorical clinical data of patients (stage, grade of tumor, residuum, chemosensitivity status) with functional classification of mutations were analyzed using the Pearson chi-square or Fisher’s exact test. For the evaluation of associations of continuous variables such as age at diagnosis or transcript expression with categorical ones, the Kruskal–Wallis test was used. Correlations among continuous variables were tested with Spearman’s rho correlation. All tests were two-sided and p-values < 0.05 were considered statistically significant. Survival curves were plotted using the Kaplan–Meier method. Expression levels were distributed by quartiles and the “optimal cut-off” was defined as the highest statistical significance by the log-rank test. All statistical analyses were performed using the SPSS v16 program (SPSS, Chicago, IL, USA). Previous version of the manuscript was posted to the ResearchSquare^44^ and the current paper presents thoroughly revised content.

List of abbreviations

AACRThe American Association for Cancer ResearchCKBThe Clinical KnowledgebaseEOCepithelial ovarian cancerFIGOThe International Federation of Gynecology and ObstetricsFOLFOX5-fluorouracil, leucovorin, and oxaliplatinGENIEGenomics Evidence Neoplasia Information ExchangeHGSChigh-grade serous ovarian carcinomasLGSClow-grade serous ovarian carcinomasMIQEthe Minimum Information for Publication of Quantitative Real-Time PCR Experiments GuidelinesNGSnext generation sequencingnonHGSCother than high-grade serous carcinomaNSCLCnon-small cell lung cancerOSoverall survivalPARPipoly(ADP-ribose) polymerase inhibitorsPFIplatinum-free intervalqPCRquantitative real-time polymerase chain reaction

Supplementary Material

KCBT_S_2025_0041R1_Online_Resouces.docx

Manuscript clean version R1.docx