Skip to main content
Download PDF

Validating a clinically based MS-MLPA threshold through comparison with Sanger sequencing in glioblastoma patients

Clinical Epigenetics volume 17, Article number: 16 (2025) Cite this article

Abstract

Background

Glioblastoma is the commonest malignant brain tumor and has a very poor prognosis. Reduced expression of the MGMT gene (10q26.3), influenced primarily by the methylation of two differentially methylated regions (DMR1 and DMR2), is associated with a good response to temozolomide treatment. However, suitable methods for detecting the methylation of the MGMT gene promoter and setting appropriate cutoff values are debated.

Results

A cohort of 108 patients with histologically and genetically defined glioblastoma was retrospectively examined with methylation-specific Sanger sequencing (sSeq) and methylation-specific multiplex ligation-dependent probe amplification (MS-MLPA) methods. The DMR2 region was methylated in 29% of samples, whereas DMR1 was methylated in 12% of samples. Methylation detected with the MS-MLPA method using probes MGMT_215, MGMT_190, and MGMT_124 from the ME012-A1 kit (located in DMR1 and DMR2) correlated with the methylation of the corresponding CpG dinucleotides detected with sSeq (p = 0.005 for probe MGMT_215; p < 0.001 for probe MGMT_190; p = 0.016 for probe MGMT_124). The threshold for methylation detection with the MS-MLPA method was calculated with a ROC curve analysis and principal components analysis of the data obtained with the MS-MLPA and sSeq methods, yielding a weighted value of 0.362. Thus, methylation of the MGMT gene promoter was confirmed in 36% of samples. These patients had statistically significantly better overall survival (p = 0.003).

Conclusions

Our results show that the threshold for methylation detection with the MS-MLPA method determined here is useful from a diagnostic perspective because it allows the stratification of patients who will benefit from specific treatment protocols, including temozolomide. Detailed analysis of the MGMT gene promoter enables the more-precise and personalized treatment of patients with glioblastoma.

Background

Glioblastoma (GBM) is the commonest malignant brain tumor in adults and has a very poor prognosis, including a median overall survival (OS) of 12–17 months [1]. The current classification of GBM is based on the recently published Word Health Organization (WHO) and European Association of Neuro-Oncology (EANO) guidelines [2, 3]. These guidelines integrate the classical histological approach (nuclear expression of the ATRX chromatin remodeler, necrosis, and microvascular proliferation) with molecular testing. At the molecular level, GBM is characterized by the absence of mutations in the isocitrate dehydrogenase (IDH) genes, the presence of mutations in the telomerase reverse transcriptase gene (TERT) promoter, amplification of the epidermal growth factor receptor gene (EGFR), and/or trisomy 7 and/or monosomy 10 [2, 3].

Another important prognostic and predictive factor is the methylation of the O-6-methylguanine-DNA methyltransferase gene (MGMT) promoter, which determines the response of patients to alkylating agents such as temozolomide (TMZ) [4, 5]. This tumor suppressor gene is located on chromosome 10 (10q26.3). It is generally accepted that the downregulation of MGMT expression is mediated by the methylation of CpG islands located in the MGMT promoter. These CpG islands extend from nucleotide (nt) − 452 to nt + 308 and consist of 98 CpGs encompassing the minimal promoter, exon 1, the enhancer, and the intron 1 region [6,7,8]. The methylation of differentially methylated regions 1 and 2 (DMR1 and DMR2) containing CpGs 25–50 and CpGs 73–90, respectively, has been reported to significantly downregulate MGMT expression [8]. Several CpGs, including CpG 31 (cg12434587), CpGs 79–83, CpGs 84–87, and CpG 89 (cg12981137), have been associated with improved overall patient survival [7, 9, 10].

Several methods have been developed to detect the methylation of the MGMT promoter, and they can be divided into two groups according to the principles upon which they are based. The first group includes methods based on the bisulfite conversion of genomic DNA (gDNA), with the subsequent detection of the unconverted cytosines. Examples include pyrosequencing, methylation-specific PCR (MS-PCR), Sanger sequencing after bisulfite conversion (sSeq) [11], methylation DNA array, and other approaches [12]. The second group includes methods based on various principles, such as immunohistochemistry or methylation-specific multiplex ligation-dependent probe amplification (MS-MLPA). The MS-MLPA approach has been used since 2006 [13]. However, to date, there has been no consensus on the methods or cutoff values used to determine MGMT promoter methylation [12, 14].

In this retrospective study, we compared two methodologically distinct approaches for detecting methylation within DMR1 and DMR2: sSeq and MS-MLPA. This comparison validates the reliability of overlapping probes and establishes a consistent threshold for the MS-MLPA method. With this threshold established, we investigated the impact of the methylation detected with MS-MLPA on the survival of GBM patients.

Patients and methods

Patient cohort and biopsy processing

The 108 patients were retrospectively included in the study based on the following criteria: (i) an initial diagnosis of GBM grade 4 based on histological and molecular markers (according to WHO/EANO 2021); (ii) the patient was diagnosed in the period 2015–2020; and (iii) the acquired tumor biopsy included sufficient material for analysis with both methods.

The biopsies were obtained during the neurosurgical removal of the tumor at the Department of Neurosurgery and Neurooncology, Military University Hospital, Prague (Czech Republic). The samples were immersed in phosphate-buffered saline (PBS) with 1% heparin (Zentiva, Prague, Czech Republic) and transferred to the Center of Oncocytogenomics, General University Hospital (Prague, Czech Republic). The acquired biopsies were homogenized at medium speed for 45 s with a Minilys homogenizer (Bertin Instruments, Montigny-le-Bretonneux, France). The homogenized tumor tissue was divided into two parts. The first was used for interphase fluorescence in situ hybridization (I-FISH) analysis, and the second was centrifuged (24,000 × g, 5 min, 4 °C) and the pellet used for gDNA isolation. Peripheral blood samples taken after the neurosurgical procedure were collected into Vacuette® Tubes 2 mL K3-EDTA (Dialab, Prague, Czech Republic) and used for the isolation of gDNA. These samples were used as the negative controls.

gDNA extraction

The gDNA from the tumor biopsy samples was isolated with the DNeasy Blood and Tissue Kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s protocol. The GenElute™ Blood Genomic DNA Kit (Merck, Darmstadt, Germany) was used to isolate the peripheral blood gDNA, according to the manufacturer’s instructions. The gDNA quality and quantity were assessed with a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and/or an Invitrogen™ Qubit™ 4.0 Fluorometer (Thermo Fisher Scientific, Oregon, USA).

I-FISH

I-FISH slides were prepared with standard cytogenetic procedures. Dual-color interphase FISH with DNA probes from the Vysis LSI PTEN/CEP 10 FISH Probe Kit (Abbott Molecular, Des Plaines, IL, USA), XL EGFR amp (MetaSystems, Altlussheim, Germany), and MGMT-20-OR (Empire Genomics, Williamsville, NY, USA) was performed according to manufacturers’ recommendations. The cutoff values were defined as 5% for deletions and 2.5% for gains, as reported in Zemanova et al. study [15].

Multiplex ligation-dependent probe amplification

The Salsa® MLPA Probemix p370 BRAF-IDH1-IDH2 and Salsa® MS-MLPA Probemix ME012 MGMT-IDH1-IDH2 (version A1) (MRC Holland, Amsterdam, Netherlands) were used according to manufacturer’s instructions to detect selected copy number alterations (CNA), mutations in the IDH1, IDH2, and BRAF genes, and MGMT promoter methylation. The PCR amplicons were analyzed with the SeqStudio Genetic Analyzer System (Applied Biosystems, Waltham, MA, USA) and the fragmentation data with Coffalyser.Net (MRC Holland). The thresholds for the detection of deletions and/or gains were set as follows: MLPA probe score of ≥ 1.3 for gain/amplification, 1.15–1.29 for suspected gain/amplification, ≤ 0.7 for loss/deletion, and 0.85–0.71 for suspected loss. Only the final ratios were used for further analyses.

Bisulfite conversion and Sanger sequencing

Samples of the gDNA (100 ng) obtained from the tumor biopsies were used as the input for bisulfite conversion with the EpiTect Bisulfite Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol. The converted DNA (4 µl) was used as the input for sSeq. The 316-bp PCR product from DMR2 was prepared with the Qiagen Multiplex PCR Kit (Qiagen) using the primers described by Möllemann et al. [16]. The 297-bp PCR product covering 35 CpGs (CpG sites 22–56) in DMR1 was obtained with primers designed with the MethPrimer online tool [17]. The sequence of the forward primer was 5′-TATTTGGTAAATTAAGGTATAGAGTTTTAG-3′ and that of the reverse primer was 5′-AAAACCTAAAAAAAACAAAAAAAC-3′. The sequencing reactions were performed with the BigDye™ Terminator v1.1 Cycle Sequencing Kit (Thermo Fisher Scientific) according to the manufacturer’s protocol. The products were purified with the BigDye™ XTerminator Purification Kit (Thermo Fisher Scientific) and sequenced with the SeqStudio™ Genetic Analyzer System (Applied Biosystems). SeqScape software (Thermo Fisher Scientific) and BiQ Analyzer [18] were used to evaluate the DMR1 and DMR2 sequences. A CpG was deemed methylated when the value of C/T was ≥ 0.5. Samples were deemed to be methylated if the total number of detected methylated CpGs was > 50% [11].

Statistical analyses

The public server at usegalaxy.org (heatmap.2 tool) was used for visualization of the raw MS-MLPA data obtained from the tumor tissues and the corresponding peripheral blood samples [19]. All statistical calculations were performed with R software [20]. Receiver operator curves (ROCs) were generated with the pROC package [21]. Optimal thresholds were chosen according to the “top left” criterion. ROC curves were constructed for both DMR1 and DMR2, and the sSeq data were used as true/false discriminators. A principal components analysis (PCA) was performed to obtain one discriminant value for the MS-MLPA method and to calculate the weights of each MS-MLPA probe. The weighted sum was further used for the ROC analysis. A one-sample t-test was used to compare the MS-MLPA probe methylation status with the specific threshold obtained with the ROC analysis and the methylation of the corresponding CpGs. A Kaplan–Meier analysis was performed with survival package, R package version 3.5–3 [22], including the following clinically relevant criteria: age, sex, type of resection, Karnofsky performance score (KPS), use of 5-aminolevulinic acid imaging during surgery, initiation of the Stupp protocol, and Cox model estimates.

Results

Patient characteristics

We selected 108 patients with GBM, wild-type IDH, and WHO grade 4, based on the recently released WHO classification [2]. The median OS of all the patients was 13 months. The sex ratio slightly favored men over women (57.4% and 42.6%, respectively) (Table 1). Most patients underwent at least partial tumor resection at the time of diagnosis (n = 104). In four patients, only a biopsy sample was taken. If the patient’s health allowed, the Stupp treatment protocol was initiated (n = 86). For further information on the patients’ clinical, demographic, and histological features, see Table 1 and Table S1.

Table 1 Clinical and demographic characteristic of GBM patients in the study cohort
Full size table

Study design

To establish whether the MS-MLPA method can reliably distinguish methylated and unmethylated samples, sSeq, a well-established method for the detection of MGMT promoter methylation, was used as the discriminator in this study. The MS-MLPA kit ME012 consists of five probes distributed across the whole MGMT promoter region. The methylation-sensitive HhaI restriction enzyme recognizes the 5′-GCGC-3′ sequence present at least once in each probe. The DMR2 sequenced with sSeq contained 25 CpGs (CpG sites 74–98), whereas DMR1 contained 35 CpGs (CpG sites 22–56). Therefore, the three MS-MLPA probes (MGMT_190, MGMT_124, and MGMT_215) overlapped the sSeq-sequenced DMR1 and DMR2. The localization of the HhaI restriction site in the MGMT_190 probe corresponded to CpG 23; the MGMT_124 probe contained two restriction sites that correlated to CpG 31 and CpG 38; and the MGMT_215 probe contained one site that correlated to CpG 91 (Fig. 1).

Fig. 1
figure 1

Schematic representation of the MGMT promoter and the study design. The transcript NM_002412.4 of MGMT gene was used as reference sequence. Gray boxes encompass the 5′-untranslated region (UTR), first exon, and first intron of the MGMT gene; violet boxes encompass differentially methylated regions (DMR1 and DMR2). Red arrows indicate the locations of MS-MLPA probes; gray arrows indicate the locations of sSeq primers. Within the gray boxes, DMR1 and DMR2 sequences show CpGs and MS-MLPA probes (violet arrows)

Full size image

sSeq evaluation of DMR1 and DMR2 methylation

The methylation status of both DMR1 and DMR2 was investigated in 103 samples. The remaining samples were not sequenced due to a lack of material, poor quality gDNA, or the repeated failure of bisulfite conversion. CpGs 26–33 were repeatedly methylated in DMR1 (Fig. 2A), including the most frequently methylated dinucleotide CpG 32 (methylation detected in 29% of samples). The CpGs of DMR2 were more frequently methylated than those of DMR1. The mean proportion of methylated CpGs was 14% for DMR1 and 26% for DMR2. CpGs 79–88 were most frequently methylated in DMR2 (37% of patients), and CpG 87 was methylated in the greatest number of samples. The methylation status of both regions was determined for 97 samples. Overall, DMR1 was methylated in 12% of the investigated samples and DMR2 in 29%. For detailed information on the methylation of DMR1 and DMR2, see Table S2.

Fig. 2
figure 2

Evaluation of raw methylation data. a Methylation of individual CpGs (methylated and unmethylated) detected with sSeq in methylated and unmethylated samples. b Heatmap visualization of raw MS-MLPA data derived from peripheral blood and brain tissue. The Euclidean distance method and the complete hierarchical clustering method were used

Full size image

Evaluation of raw MS-MLPA data

The MS-MLPA method was used to detect the methylation of and deletions in the MGMT promoter of all the samples in the study (n = 108) (Table S2). The loss or suspected loss of 10q26.3 (MGMT promoter and gene) was reported in 69 samples (63.8%). The samples with suspected loss were investigated with FISH, and monosomy 10 or the loss of 10q was confirmed. The methylation values obtained for each probe were visualized with the Heatmap2 function in Galaxy software (Fig. 2). Methylation values of 0.04–0.23 were detected for probes MGMT_172 and MGMT_140, localized to the 3′ and 5′ ends of the MGMT promoter, in most of the samples, regardless of the methylation status of the remaining probes. The same pattern was observed for the data obtained from the peripheral blood of the patients (Fig. 2; Fig. S1). Therefore, the peripheral blood values were subtracted from the raw data. This modification had little effect on the sample distribution in the heatmap visualization (Fig. S1).

Coincident methylation of MS-MLPA probes and associated CpGs

The specificity and sensitivity of the MS-MLPA probes were evaluated with ROC analyses. In the analysis of DMR1, the MS-MLPA probe MGM_190 performed best, yielding the highest scores for sensitivity (91%) and specificity (88%), whereas probe MGMT_215 had the highest scores for sensitivity (72%) and specificity (89%) in DMR2 (Fig. S1). The best performance of probe MGMT_190 in DMR1 and MGMT_215 in DMR2 correlated with their locations within each region. The optimal cutoff values for the methylation detected with each MS-MLPA probe were 0.165 for MGMT_190, 0.115 for MGMT_124, and 0.17 for MGMT_215. Based on these cutoff values, the presence/absence of methylation of the MS-MLPA probes was compared with the methylation of the corresponding CpGs (sSeq). A false methylation signal obtained with the MS-MLPA method due to a polymorphism in one of the investigated probes was observed in 4.63% of samples. The MS-MLPA and sSeq methods yielded the same result (methylated or unmethylated) in > 80% of samples for the MGMT_190 (p < 0.001), MGMT_124 (p = 0.016), and MGMT_215 (p = 0.005) probes.

Calculation of MS-MLPA method threshold

The estimated linear combination of the PCA explained 55% of the variability (Fig. 3). The PCA value for the MS-MLPA analysis was calculated as follows:

$$\begin{aligned} {\text{PCA}}\,{\text{value}}\, = \, & 0.423 \times \left[ {{\text{MGMT}}\_140} \right] + 0.456 \times \left[ {{\text{MGMT}}\_190} \right] + 0.397 \times \left[ {{\text{MGMT}}\_124} \right] + 0.410 \\ & \times \left[ {{\text{MGMT}}\_160} \right] + 0.425 \times \left[ {{\text{MGMT}}\_215} \right] + 0.327 \times \left[ {{\text{MGMT}}\_172} \right] \\ \end{aligned}$$
Fig. 3
figure 3

MS-MLPA threshold set with principal components analysis (PCA) and receiver operating characteristic (ROC) analysis. a PCA of MS-MLPA data. Red arrows indicate the data trend for each MS-MLPA probe. b ROC analysis to validate the specificity and sensitivity of the PCA model for MS-MLPA data and each MS-MLPA probe

Full size image

A subsequent ROC curves analysis that included the proposed calculated values was performed separately for DMR1 and DMR2. This analysis confirmed the validity of the proposed PCA values for the MS-MLPA method and set the threshold for DMR1 as 0.536 (95% confidence interval [CI] sensitivity: 0.73–0.97; 95% CI specificity: 0.77–0.92) and that for DMR2 as 0.363 (95% CI sensitivity: 0.75–1.00; 95% CI specificity: 0.75–0.9) (Fig. 3).

Clinical and statistical validation of the MS-MLPA thresholds

The PCA values for the samples were calculated (n = 108) to enable further clinical validation of the MS-MLPA results. The threshold of 0.363 obtained for DMR2 was used, because DMR2 was more frequently methylated than DMR1. Of the GBM samples, 36% (n = 39) were methylated when this threshold was used. The Kaplan–Meier analysis of the methylation data from MS-MLPA showed that this method distinguished the methylation of the MGMT promoter and thus identified the patients with a higher probability of longer OS (p = 0.003). The Kaplan–Meier analysis of the sSeq data was inconclusive, but the trend was similar to that in the MS-MLPA results (p = 0.5) (Fig. 4).

Fig. 4
figure 4

Survival analysis of glioblastoma patients analyzed with MS-MLPA and sSeq. a Kaplan–Meier survival analysis of methylated and unmethylated MTMG promoters in patients based on the established thresholds. Pink and green areas encompassing the main curve represent standard deviation. b Kaplan–Meier survival plot for each variable

Full size image

Clinical analysis of GBM patients

The Kaplan–Meier analysis showed that the age, type of resection, KPS value, and treatment with the Stupp protocol probably affected the survival of the GBM patients (Fig. 4). These results were confirmed with a subsequent multivariant analysis using the Cox model without interactions, which determined the hazard ratios (HRs) for the individual values: HR = 2.65 (95% CI 1.6223–4.3132; p < 0.0001) for patients with a methylated MGMT promoter; HR = 3.36 (95% CI 1.7262–6.5460; p = 0.0003) for patients treated with the Stupp protocol; and HR = 1.51 (95% CI 0.9824–2.3359; p = 0.0602) for patients who underwent radical resection. Based on the Kaplan–Meier analysis and Cox’s model, the age of the patients had a significant effect on survival (HR = 1.35 [95% CI 1.0903–1.6748; p = 0.0060]), but only up to 2 years after surgery. Age did not seem an important regressor (p = 0.1951) for data collected more than 2 years after surgery (baseline). To assess whether the survival of patients was affected by other factors simultaneously with the methylation of the MGMT promoter, a Cox’s model with interactions was used, in which interaction terms of > 0.1 were omitted, as well as the main effect of sex. Only the combination of positive methylation and treatment with the Stupp protocol was statistically significant (p = 0.0045).

The patients were divided into two groups according to their OS: long-term survival (LTS, 24% of patients) and shorter survival (non-LTS, 76% of patients) based on the 2-year threshold observed in the Cox model and also reported in several publications [7]. The patients in the LTS group were younger, with a mean age at the time of diagnosis of 53.7 years (vs. 64.6 years for non-LTS patients) and the majority underwent radical tumor resection at the time of diagnosis (75% of patients in the LTS group vs. 54% of the non-LTS patients). The Stupp treatment protocol was initiated for all patients in the LTS group. For detailed information about the LTS and non-LTS patients, see Table 1 and Table S2.

Discussion

The importance of the relationship between the methylation of the MGMT promoter and better OS, probably due to the implementation of the Stupp protocol in the treatment of GBM patients, has been noted previously, especially after the new WHO classification system was implemented in 2021 [2, 23]. The new classification strictly defines the histological and molecular features of patients with GBM; consequently, the revision of previous studies of GBM may be warranted [24]. The MS-MLPA technique has been used for almost two decades and has undergone many changes from the original basic design. The first version included four probes scattered throughout the promoter and first exon of the MGMT gene [25]. The introduction of the ME011 kit followed, which comprises six probes to detect the methylation of the MGMT promoter [26,27,28]. The ME012-A1 kit, which consists of six probes but with a different design than the ME011 kit, including the discontinuation of two probes (SALSA probes ID 13716-L15582 and 14,135-L16573), has since been introduced, and the ME012-B1 kit, which contains two additional probes, has recently been released. Although the length of the original six MS-MLPA probes changed between kits A1 and B1, the design and the targeted restriction site of the HhaI enzyme remained the same. However, there has been no up-to-date clinical evaluation of the MS-MLPA technique, as previously mentioned [14]. In the present study, we compared the MS-MLPA ME012-A1 kit with sSeq performed retrospectively on a group of histologically and genetically defined GBM samples.

Several studies have investigated the frequencies of methylated CpG sites within DMR1 and DMR2 [29] and highlighted the clinical importance of CpG sites 79–87 [10], 79–82 and 86 [7], 33 and 87 [9], 79–94 [11], and 32–33 and 72–83 [30]. The majority of these CpG sites are located within DMR2. Consistent with the literature, our own evaluation of individual methylated CpGs with sSeq confirmed that the methylation hotspots were localized to CpG sites 80–87 of DMR2 and CpGs 22–33 of DMR1. The methylation of the 5′-GCGC-3′ sequence recognized by the HhaI enzyme in the MS-MLPA probe MGMT_125, located within DMR2, corresponds to the methylation of CpG site 91. The methylation of CpG sites located in DMR1, coincided for the remaining two MS-MLPA probes, MGMT_190 and MGMT_215. The MS-MLPA probes do not directly address the methylation of individual CpG sites within DMR2 or DMR1. However, as suggested by Siller et al. [11], the CpG sites within the MGMT promoter are more likely to be methylated when the bordering CpG site is also methylated, so the MS-MLPA probes may be able to detect a methylation event even if the exact hotspot CpGs are not targeted.

An MS-MLPA probe may produce a false positive result when a polymorphism occurs at the HhaI restriction site, as was observed in our study. However, higher levels of methylation were detected at the 5′ and 3′ ends of the MGMT promoter than within it with the MS-MLPA approach (probes MGMT_140 and MGMT_172) in glioma samples and even in the control samples obtained from the peripheral blood. We do not assume that the methylation detected was attributable to innate polymorphisms because the methylation signal was 5%–23%. Instead, it may have been due to nonspecific cohybridization to different genomic locations or the biological properties of the MGMT promoter. However, we found no supporting information in the literature. The observed frequencies of MGMT promoter methylation (30.8% of patients with sSeq and 35.9% with MS-MLPA) are consistent with the reported frequencies of 28%–68% [5, 10, 31]. Park et al. [26] and Trabelsi et al. [27] suggested a methylation cutoff value equivalent to the mean of all MS-MLPA probes (≥ 0.25) in the MS-MLPA ME011 kit. Using this threshold, only 25% of our samples (n = 27) would be methylated. The detection of the methylation of the MGMT promoter may be influenced by sample collection, and because the gDNA was isolated from homogenized biopsy specimens, the proportion of tumor cells may have been low. In contrast, the gDNA isolated from formalin-fixed samples is highly fragmented and thus susceptible to processing errors.

The methylation of the MGMT promoter may not be the only mechanism regulating gene expression. The involvement of the loss of 10q26 or even the biallelic deletion of the MGMT gene in the downregulation of MGMT expression has been investigated in several studies, with conflicting outcomes [32,33,34]. Neither Bady et al. (2012) nor Ramalho-Carvalho et al. (2013) observed a significant association between 10q loss and MGMT expression in GBM patients, so there was no dosage effect on MGMT expression [9, 35]. However, in our study the co-occurrence of MGMT promoter methylation and the loss of 10q were more prevalent in patients in the LTS group than in the non-LTS group (60% vs. 25%, respectively), consistent with previous findings that GBM patients with both mechanisms of MGMT inactivation had longer OS and progression-free survival [32].

The data from our MS-MLPA and sSeq analyses confirmed previously published findings that GBM patients benefit from the presence of MGMT promoter methylation, and that the prognosis is most promising when this is combined with treatment with the Stupp protocol [23]. No other associations were shown to influence the hazard ratios. Moreover, a statistical analysis (Cox’s model) of patient age at the time of diagnosis indicated that an OS of 2 years was the cutoff point between LTS and non-LTS. The definition of LTS varies in the literature, with various studies using 2–5 years as the cutoff threshold [36, 37]. Consistent with the literature [38], in this study the LTS patients were younger at the time of diagnosis, most of them underwent total resection, and had a high KPS.

While the study offers valuable insights, it also has some limitations. We did not achieve prognostic significance for sSeq data. The differences between our findings and those reported by Siller et al. [11] could be attributed to several factors, including variations in cohort characteristics and treatment protocols. In our cohort, most of patients underwent total resection (96.3% vs. 46.5% in Siller’s cohort) and received adjuvant temozolomide treatment (79.6% vs. 42.3%). These differences likely influenced the prognostic stratification observed in our study. Additionally, while our study utilized a smaller sample size, which may have limited its statistical power, we observed prognostic separation after 20 months of follow-up, although the p-value did not reach statistical significance. Our investigation also extended to the DMR1 region, complementing prior research focused on DMR2 and broadening the scope of MGMT promoter methylation assessment.

Furthermore, the following shortcomings may have also influenced the study results: the failure to determine the TERT mutation status; possible patient selection bias in this retrospective cohort; and the use of peripheral blood as the negative control for the MS-MLPA method. The manufacturer of MS-MLPA kits recommend that the same source of tissue and the same gDNA isolation kit be used for the negative controls. Several brain tissue samples from patients suffering from epilepsy were tested in our previous study with MS-MLPA kit ME011, with similar results obtained from peripheral blood [39]. However, noncancerous brain tissue is difficult to obtain, so corresponding peripheral blood samples were used. Furthermore, because these data were collected at one institution, we recommend validation of our data in a multicenter study with a larger cohort of GBM samples.

Conclusions

The results underscore the importance of methylation of the MGMT promoter in enhancing the OS of GBM patients, particularly when the Stupp protocol is used in their treatment. To our knowledge, this is the first study to thoroughly investigate the analysis of MGMT promoter methylation with the MS-MLPA method in a histologically and genetically defined group of patients with GBM, wild-type IDH, WHO grade 4 as earlier studies were conducted under the older WHO classification. Advances in MS-MLPA kits, ranging from ME011 to the latest ME012-B1, have improved the detection of MGMT promoter methylation (despite some limitations), especially at the 5′ and 3′ ends of the promoter. Our data suggest that the two-fold inactivation of the MGMT gene, due to both methylation and deletion, may improve the long-term survival of GBM patients. Finally, we have suggested a clinically validated weighted threshold for the detection of MGMT promoter methylation with MS-MLPA in patients with brain tumors. Further detailed and clinically validated analyses of the MGMT gene promoter should contribute to the more-precise and personalized treatment of patients with GBM.

Availability of data and materials

All data generated or analyzed in this study are partially included in this published article (and its Supplementary Information files). The remaining parts of the datasets used during the study are available from the corresponding author upon reasonable request.

Abbreviations

ATRX:

ATRX chromatin remodeler

DMR1/2:

Differentially methylated region 1 or 2

CI:

Confidence interval

CNA:

Copy number alteration

EGFR:

Epidermal growth factor receptor

GBM:

Glioblastoma

gDNA:

Genomic DNA

IDH:

Isocitrate dehydrogenase

sSeq:

Sanger sequencing after bisulfite conversion

MGMT:

O-6-methylguanine-DNA methyltransferase

MS-MLPA:

Methylation-specific multiplex ligation-dependent probe amplification

LTS:

Long-term survival

TMZ:

Temozolomide

TERT:

Telomerase reverse transcriptase

References

  1. Ostrom QT, Shoaf ML, Cioffi G, Waite K, Kruchko C, Wen PY, et al. National-level overall survival patterns for molecularly-defined diffuse glioma types in the United States. Neuro Oncol. 2023;25(4):799–807.

    Article  CAS  PubMed  Google Scholar 

  2. Louis DN, Perry A, Wesseling P, Brat DJ, Cree IA, Figarella-Branger D, et al. The 2021 WHO classification of tumors of the central nervous system: a summary. Neuro Oncol. 2021;23(8):1231–51.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Weller M, van den Bent M, Preusser M, Le Rhun E, Tonn JC, Minniti G, et al. EANO guidelines on the diagnosis and treatment of diffuse gliomas of adulthood. Nat Rev Clin Oncol. 2021;18(3):170–86.

    Article  PubMed  Google Scholar 

  4. Esteller M, Garcia-Foncillas J, Andion E, Goodman SN, Hidalgo OF, Vanaclocha V, et al. Inactivation of the DNA-repair gene MGMT and the clinical response of gliomas to alkylating agents. N Engl J Med. 2000;343(19):1350–4.

    Article  CAS  PubMed  Google Scholar 

  5. Hegi ME, Diserens AC, Gorlia T, Hamou MF, de Tribolet N, Weller M, et al. MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med. 2005;352(10):997–1003.

    Article  CAS  PubMed  Google Scholar 

  6. Costello JF, Futscher BW, Tano K, Graunke DM, Pieper RO. Graded methylation in the promoter and body of the O6-methylguanine DNA methyltransferase (MGMT) gene correlates with MGMT expression in human glioma cells. J Biol Chem. 1994;269(25):17228–37.

    Article  CAS  PubMed  Google Scholar 

  7. Leske H, Camenisch Gross U, Hofer S, Neidert MC, Leske S, Weller M, et al. MGMT methylation pattern of long-term and short-term survivors of glioblastoma reveals CpGs of the enhancer region to be of high prognostic value. Acta Neuropathol Commun. 2023;11(1):139.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Malley DS, Hamoudi RA, Kocialkowski S, Pearson DM, Collins VP, Ichimura K. A distinct region of the MGMT CpG island critical for transcriptional regulation is preferentially methylated in glioblastoma cells and xenografts. Acta Neuropathol. 2011;121(5):651–61.

    Article  CAS  PubMed  Google Scholar 

  9. Bady P, Sciuscio D, Diserens AC, Bloch J, van den Bent MJ, Marosi C, et al. MGMT methylation analysis of glioblastoma on the Infinium methylation BeadChip identifies two distinct CpG regions associated with gene silencing and outcome, yielding a prediction model for comparisons across datasets, tumor grades, and CIMP-status. Acta Neuropathol. 2012;124(4):547–60.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Buyuktepe M, Kaplan I, Bayatli E, Dogan H, Ugur HC. Significance of O6-methyl guanine methyltransferase promoter methylation in high grade glioma patients: optimal cutoff point, CpG locus, and genetic assay. J Neurooncol. 2023;164(1):171–7.

    Article  CAS  PubMed  Google Scholar 

  11. Siller S, Lauseker M, Karschnia P, Niyazi M, Eigenbrod S, Giese A, et al. The number of methylated CpG sites within the MGMT promoter region linearly correlates with outcome in glioblastoma receiving alkylating agents. Acta Neuropathol Commun. 2021;9(1):35.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Malmström A, Łysiak M, Kristensen BW, Hovey E, Henriksson R, Söderkvist P. Do we really know who has an MGMT methylated glioma? Results of an international survey regarding use of MGMT analyses for glioma. Neurooncol Pract. 2020;7(1):68–76.

    PubMed  Google Scholar 

  13. Jeuken J, Cornelissen S, Boots-Sprenger S, Gijsen S, Wesseling P. Multiplex ligation-dependent probe amplification: a diagnostic tool for simultaneous identification of different genetic markers in glial tumors. J Mol Diagn. 2006;8(4):433–43.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Mansouri A, Hachem LD, Mansouri S, Nassiri F, Laperriere NJ, Xia D, et al. MGMT promoter methylation status testing to guide therapy for glioblastoma: refining the approach based on emerging evidence and current challenges. Neuro Oncol. 2019;21(2):167–78.

    Article  CAS  PubMed  Google Scholar 

  15. Zemanová Z, Kramar F, Babická L, Ransdorfová S, Melichercíková J, Hrabal P, et al. Molecular cytogenetic stratification of recurrent oligodendrogliomas: utility of interphase fluorescence in situ hybridization (I-FISH). Folia Biol (Praha). 2006;52(3):71–8.

    Article  PubMed  Google Scholar 

  16. Möllemann M, Wolter M, Felsberg J, Collins VP, Reifenberger G. Frequent promoter hypermethylation and low expression of the MGMT gene in oligodendroglial tumors. Int J Cancer. 2005;113(3):379–85.

    Article  PubMed  Google Scholar 

  17. Li LC, Dahiya R. MethPrimer: designing primers for methylation PCRs. Bioinformatics. 2002;18(11):1427–31.

    Article  CAS  PubMed  Google Scholar 

  18. Bock C, Reither S, Mikeska T, Paulsen M, Walter J, Lengauer T. BiQ Analyzer: visualization and quality control for DNA methylation data from bisulfite sequencing. Bioinformatics. 2005;21(21):4067–8.

    Article  CAS  PubMed  Google Scholar 

  19. The Galaxy Community. The Galaxy platform for accessible, reproducible, and collaborative data analyses: 2024 update, Nucleic Acids Research, 2024. https://doiorg.publicaciones.saludcastillayleon.es/10.1093/nar/gkae410. Accessed Feb 2024

  20. Team RC. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/. 2023. Accessed Mar 2024.

  21. Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez JC, et al. pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinform. 2011;12:77.

    Article  Google Scholar 

  22. Therneau T. A Package for Survival Analysis in R_. R package version 3.5-3. 2023. https://CRAN.R-project.org/package=survival. Accessed March 2024.

  23. Stupp R, van den Bent MJ, Hegi ME. Optimal role of temozolomide in the treatment of malignant gliomas. Curr Neurol Neurosci Rep. 2005;5(3):198–206.

    Article  CAS  PubMed  Google Scholar 

  24. Thomas-Joulié A, Tran S, El Houari L, Seyve A, Bielle F, Birzu C, et al. Prognosis of glioblastoma patients improves significantly over time interrogating historical controls. Eur J Cancer. 2024;202: 114004.

    Article  PubMed  Google Scholar 

  25. Jeuken JW, Cornelissen SJ, Vriezen M, Dekkers MM, Errami A, Sijben A, et al. MS-MLPA: an attractive alternative laboratory assay for robust, reliable, and semiquantitative detection of MGMT promoter hypermethylation in gliomas. Lab Invest. 2007;87(10):1055–65.

    Article  CAS  PubMed  Google Scholar 

  26. Park CK, Kim J, Yim SY, Lee AR, Han JH, Kim CY, et al. Usefulness of MS-MLPA for detection of MGMT promoter methylation in the evaluation of pseudoprogression in glioblastoma patients. Neuro Oncol. 2011;13(2):195–202.

    Article  CAS  PubMed  Google Scholar 

  27. Trabelsi S, Mama N, Ladib M, Karmeni N, Haddaji Mastouri M, Chourabi M, et al. MGMT methylation assessment in glioblastoma: MS-MLPA versus human methylation 450K beadchip array and immunohistochemistry. Clin Transl Oncol. 2016;18(4):391–7.

    Article  CAS  PubMed  Google Scholar 

  28. Christians A, Hartmann C, Benner A, Meyer J, von Deimling A, Weller M, et al. Prognostic value of three different methods of MGMT promoter methylation analysis in a prospective trial on newly diagnosed glioblastoma. PLoS ONE. 2012;7(3): e33449.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Gibson D, Vo AH, Lambing H, Bhattacharya P, Tahir P, Chehab FF, et al. A systematic review of high impact CpG sites and regions for MGMT methylation in glioblastoma [A systematic review of MGMT methylation in GBM]. BMC Neurol. 2024;24(1):103.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Everhard S, Tost J, El Abdalaoui H, Crinière E, Busato F, Marie Y, et al. Identification of regions correlating MGMT promoter methylation and gene expression in glioblastomas. Neuro Oncol. 2009;11(4):348–56.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Szylberg M, Sokal P, Śledzińska P, Bebyn M, Krajewski S, Szylberg Ł, et al. Promoter methylation as a prognostic factor in primary glioblastoma: a single-institution observational study. Biomedicines. 2022;10(8):2030.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Richard S, Tachon G, Milin S, Wager M, Karayan-Tapon L. Dual MGMT inactivation by promoter hypermethylation and loss of the long arm of chromosome 10 in glioblastoma. Cancer Med. 2020;9(17):6344–53.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Bady P, Delorenzi M, Hegi ME. Sensitivity analysis of the MGMT-STP27 model and impact of genetic and epigenetic context to predict the MGMT methylation status in gliomas and other tumors. J Mol Diagn. 2016;18(3):350–61.

    Article  CAS  PubMed  Google Scholar 

  34. Low JP, Satgunaseelan L, Wright D. Biallelic MGMT loss in a case of IDH-wild-type adult glioblastoma: a case for concurrent epigenomic and molecular karyotype testing. Pathology. 2023;55(4):551–4.

    Article  CAS  PubMed  Google Scholar 

  35. Ramalho-Carvalho J, Pires M, Lisboa S, Graça I, Rocha P, Barros-Silva JD, et al. Altered expression of MGMT in high-grade gliomas results from the combined effect of epigenetic and genetic aberrations. PLoS ONE. 2013;8(3): e58206.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Miele E, Anghileri E, Calatozzolo C, Lazzarini E, Patrizi S, Ciolfi A, et al. Clinicopathological and molecular landscape of 5-year IDH-wild-type glioblastoma survivors: a multicentric retrospective study. Cancer Lett. 2024;588: 216711.

    Article  CAS  PubMed  Google Scholar 

  37. van der Meulen M, Ramos RC, Voisin MR, Patil V, Wei Q, Singh O, et al. Differences in methylation profiles between long-term survivors and short-term survivors of IDH-wild-type glioblastoma. Neurooncol Adv. 2024;6(1):vdae001.

    PubMed  PubMed Central  Google Scholar 

  38. Chehade G, Lawson TM, Lelotte J, Daoud L, Di Perri D, Whenham N, et al. Long-term survival in patients with IDH-wildtype glioblastoma: clinical and molecular characteristics. Acta Neurochir (Wien). 2023;165(4):1075–85.

    Article  PubMed  Google Scholar 

  39. Lhotska H, Zemanova Z, Cechova H, Ransdorfova S, Lizcova L, Kramar F, et al. Genetic and epigenetic characterization of low-grade gliomas reveals frequent methylation of the MLH3 gene. Genes Chromosom Cancer. 2015;54(11):655–67.

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

We would like to thank our colleagues at the Center of Oncocytogenomics for their support during the collection of data and preparation of the manuscript.

Funding

This work was supported by two projects of the Ministry of Health, Czech Republic: NU21-04-00100, the Czech Health Research Council; and the Conceptual Development of Research Organization Project, 00641655, General University Hospital in Prague.

Author information

Authors and Affiliations

  1. Center of Oncocytogenomics, Institute of Medical Biochemistry and Laboratory Diagnostics, General University Hospital and 1st Faculty of Medicine of Charles University in Prague, U Nemocnice 499/2, 128 00, Prague, Czech Republic

    Halka Lhotska, Karolina Janeckova, Tatiana Aghova, Libuse Lizcova, Karla Svobodova, Lucie Hodanova & Zuzana Zemanova

  2. Department of HLA, Institute of Hematology and Blood Transfusion, U Nemocnice 2094/1, 128 00, Prague, Czech Republic

    Hana Cechova

  3. The Clinical Trials and Research Department, General University Hospital and 1st Faculty of Medicine of Charles University in Prague, U Nemocnice 499/2, 128 00, Prague, Czech Republic

    Jaromir Macoun

  4. Department of Neurosurgery and Neurooncology, 1st Faculty of Medicine of Charles University and Military University Hospital Prague, U Vojenske Nemocnice 1200, 169 02, Prague, Czech Republic

    Dora Konecna, Filip Kramar & David Netuka

  5. Department of Pathology, 1st Faculty of Medicine of Charles University and Military University Hospital Prague, U Vojenske Nemocnice 1200, 169 02, Prague, Czech Republic

    Jiri Soukup

  6. The Fingerland Department of Pathology, Charles University, Faculty of Medicine in Hradec Kralove and University Hospital Hradec Kralove, Sokolska 581, 500 05, Hradec Kralove, Czech Republic

    Jiri Soukup

Authors
  1. Halka Lhotska
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Karolina Janeckova
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Hana Cechova
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Jaromir Macoun
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Tatiana Aghova
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Libuse Lizcova
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Karla Svobodova
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Lucie Hodanova
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Dora Konecna
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Jiri Soukup
    View author publications

    You can also search for this author inPubMed Google Scholar

  11. Filip Kramar
    View author publications

    You can also search for this author inPubMed Google Scholar

  12. David Netuka
    View author publications

    You can also search for this author inPubMed Google Scholar

  13. Zuzana Zemanova
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

H.L. and Z.Z. conceived and designed the study; H.C. and L.L. performed the MS-MLPA analyses; H.L., K.J., and T.A. performed the sSeq analysis; K.S., K.J., L.H., and L.L. performed the I-FISH and MLPA to detect CNA; J.S. performed the histopathological analysis; D.K., F.K., and D.N were responsible for neurosurgical resection and the patients’ clinical data; J.M. performed the statistical analysis of the acquired data; H.L., L.L., H.C., K.J., T.A., and Z.Z. interpreted the acquired data; H.L. and Z.Z. wrote the first version of the manuscript. All authors contributed to the final version of the manuscript. All authors have read and approved the final manuscript.

Corresponding author

Correspondence to Zuzana Zemanova.

Ethics declarations

Ethics approval and consent to participate

Written consent was obtained from all the patients in the study in accordance with the Declaration of Helsinki and the ethical standards of the local ethic committees of the General Military Hospital (Prague, Czech Republic), application no. 108/15–33/2020, and the General University Hospital (Prague, Czech Republic), application no. 262/23.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

13148_2025_1822_MOESM1_ESM.xlsx

Additional file1: Table S1. Detailed clinical and histological characteristics of GBM patients included in the retrospective study.

13148_2025_1822_MOESM2_ESM.xlsx

Additional file2: Table S2. Detailed methylation and CNA data for GBM patients included in the retrospective study. Table contains CNA and methylation data obtained with MS-MLPA, sSeq, I-FISH, and MLPA methods. Failed—bisulfide conversion repeatedly failed, so sequencing was not possible; nd—not done due to lack of material; normal—two copies of the investigated chromosome locus or gene were observed.

13148_2025_1822_MOESM3_ESM.tif

Additional file3: Detailed raw MS-MLPA data analysis. (a) ROC analysis of each MS-MLPA probe for DMR1 and DMR2. Results of sSeq analysis were used as the true/false discriminator. (b) Heatmap visualization of raw MS-MLPA and adjusted data obtained from peripheral blood and brain tissue samples. The order of samples is the same for all three heatmaps.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Lhotska, H., Janeckova, K., Cechova, H. et al. Validating a clinically based MS-MLPA threshold through comparison with Sanger sequencing in glioblastoma patients. Clin Epigenet 17, 16 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13148-025-01822-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13148-025-01822-2

Keywords

Clinical Epigenetics

ISSN: 1868-7083

Contact us