Patients with glioblastoma multiforme (GBM) often have poor prognoses with a 5-year overall survival (OS) rate of approximately 5% (1,2). Improvement of the patients’ prognoses may be achieved with novel systemic agents and modern techniques in the fields of surgery and radiation oncology. Another option that may lead to improved outcomes is the concept of treatment personalization considering the expected treatment results and the patient’s individual situation including performance status, co-morbidity, age, and survival prognosis (3). Prognostic factors that help estimate OS and progression-free survival (PFS) likely facilitate the selection of a personalized treatment approach.

Treatment of GBM generally includes resection (as extensive as safely possible) followed by concurrent chemoradiation with temozolomide (TMZ) and subsequent (=adjuvant) TMZ alone (4). This regimen may be supplemented by treatment with tumor-treating fields (TTF) (5). In a previous study of 167 patients, we identified several patient-, tumor-, and treatment-related factors that were associated with OS, namely single GBM lesion, higher Karnofsky performance score (KPS), O⁶-methylguanine-DNA methyltransferase (MGMT) promoter methylation, maximum cumulative diameter of the GBM lesions, upfront resection, and chemotherapy in addition to radiotherapy (6).

In addition to these factors, the predictive value of the inflammatory markers platelet-to-lymphocyte ratio (PLR) and neutrophil-to-lymphocyte ratio (NLR) has received more attention. The prognostic significance of these factors has been described for different cancer types (7-20). Moreover, the role of pre-treatment PLR and NLR as prognostic factors is increasingly investigated in patients with GBM (2,21-30). The vast majority of the available studies, systematic reviews and meta-analyses investigated the inflammatory markers prior to resection or biopsy of GBM (2,21-30). One meta-analysis investigating the role of NLR differentiated between pre-treatment and pre-operative NLR (31). However, “pre-treatment” was not clearly defined. In the present study, PLR and NLR were assessed prior to or at the beginning of radiotherapy, which was performed following resection or biopsy of GBM. Our study aimed to provide additional information regarding the impact of PLR and NLR on the patient’s prognosis, in order to support radiation oncologists when selecting the appropriate dose-fractionation regimen for an individual patient. This is important, since patients with more favorable prognoses appear good candidates for longer-course radiotherapy using conventional fractionation (e.g., 60 Gy in 30 fractions over 6 weeks or 59.4 Gy in 33 fractions over 6.5 weeks) (4,32). Patients with poor or intermediate prognoses may receive hypo-fractionated radiotherapy with 40 Gy in 15 fractions over 3 weeks, and patients with very poor prognoses may be considered for ultra-hypo-fractionated radiotherapy with 25 Gy in 5 fractions over 1 week (6,33-35).

Eighty patients with GBM receiving conventionally fractionated radiotherapy plus concurrent TMZ (4) following resection or biopsy between 2014 and 2022 were included in this retrospective study (original protocol and amendment approved by the Ethics Committee of the University of Lübeck, file 2022-509). Dose-fractionation regimens included 59.4 Gy in 33 fractions over 6.5 weeks in 57 patients (71%) and 60.0 Gy in 30 fractions over 6 weeks in 18 patients (23%). The other five patients received 57.6 Gy in 32 fractions (n=2), 55.8 Gy in 31 fractions (n=2), and 54 Gy in 30 fractions (n=1). All patients received concurrent chemotherapy with TMZ (4,6), and 64 patients (80%) received adjuvant chemotherapy starting with TMZ. In seven of these patients, TMZ was replaced after 1-5 courses by procarbazine/lomustine or lomustine. Seven of the 80 patients received additional treatment with TTF.

PLR was calculated by dividing the absolute platelet count by the absolute lymphocyte count and NLR by dividing the absolute neutrophil count by the absolute lymphocyte count. PLR (≤150 vs. >150), NLR (≤3 vs. >3), which were both assessed prior to or at the beginning of the radiotherapy course, and 10 additional factors were investigated for potential associations with PFS and OS. Additional factors included number of GBM lesions (1 vs. ≥2), main site of GBM (temporal and/or parietal vs. other sites), maximum cumulative diameter of GBM lesions (<40 vs. ≥40 mm), Karnofsky performance score (KPS) (≤80 vs. 90-100), sex (female vs. male), O6-methylguanine-DNA methyl-transferase (MGMT) promoter methylation (no vs. yes), extent of upfront resection (gross total resection=GTR vs. subtotal resection=STR vs. biopsy only), and adjuvant chemotherapy with TMZ following concurrent chemoradiation (no vs. yes). The distribution of these factors is shown in Table I.

PFS and OS were referenced from the first day of radiotherapy. Univariable analyses were performed using the Kaplan–Meier method and differences between the Kaplan–Meier curves were evaluated using the log-rank test. After Bonferroni adjustment (11 tests), p-values <0.0045 represented an alpha level <5% and were, therefore, considered significant. A p-value <0.05 indicated a strong trend and a p-value <0.10 a trend for an association with PFS or OS. Those factors that achieved significance or showed a strong trend were additionally analyzed in a multivariable manner using a Cox proportional hazards model. In the multivariable analyses, p-values of <0.05 and <0.20, respectively, were considered indicating significance or at least a trend.

The median follow-up times were 16 (3-114) months in the entire study population and 24 (4-114) months in those 28 patients still alive at the last contact. On univariable analyses, better PFS was significantly associated with a single GBM lesion (p<0.001) and KPS 90-100 (p<0.001) (Table II). PLR ≤150 (p=0.007, Figure 1), maximum cumulative diameter <40 mm (p=0.037), age ≤59 years (p=0.030), and MGMT promoter methylation (p=0.007) showed a strong trend for better PFS. Moreover, trends were found for NLR (p=0.162) and GTR (p=0.101). In the multivariable analysis of PFS, PLR ≤150 (p=0.045), KPS 90-100 (p=0.034), and MGMT promoter methylation (p=0.001) were significant. Trends were observed for a single GBM lesion (p=0.071) and age ≤59 years (p=0.086). The complete results of the multivariable analysis of PFS including hazard ratios and 95% confidence intervals are shown in Table III.

On univariable analyses, improved OS was significantly associated with a single GBM lesion (p<0.001), KPS 90-100 (p<0.001), and GTR (p=0.002) (Table IV).

Strong trends were found for PLR ≤150 (p=0.005, Figure 2) and adjuvant chemotherapy (p=0.005). Moreover, NLR ≤3 (p=0.052), maximum cumulative diameter <40 mm (p=0.101), and MGMT promoter methylation (p=0.188) showed a trend toward better OS. In the multivariable analysis of OS, a single GBM lesion (p=0.030), KPS 90-100 (p=0.017), and adjuvant chemotherapy (p=0.004) were significant, and PLR ≤150 showed a trend (p=0.195) (Table V).

The prognosis of patients with GBM is generally poor and requires improvement (1,2). Outcomes may be improved with the use of modern drugs and techniques or by personalizing the treatment to an individual patient situation and needs. Optimal treatment personalization should always consider the patient’s survival prognosis and the expected results of the GBM therapy. Estimation of these endpoints can be facilitated by using prognostic factors of PFS and OS. Several patient- and tumor-related factors have already been identified as predictors of outcomes in patients irradiated for GBM (6). During the last decade, the prognostic value of PLR and NLR has been increasingly investigated in patients treated for GBM.

In 2015, Han et al. presented a retrospective study of 152 patients with GBM (21). Patients received neurosurgical resection (114 patients) or biopsy (38 patients) of GBM followed by chemoradiation with TMZ according to the protocol of the trial by Stupp et al. (4). Pre-treatment NLR and PLR were used as continuous and dichotomous variables. In the latter situation, cut-off values were 135 for PLR and 4 for NLR (22). On univariable analyses, both PLR <135 and NLR <4 were significantly associated with better OS, whereas on multivariable analysis, only NLR remained significant. In 2017, Wang et al. reported the retrospective data of 166 patients with primary (141 patients) or secondary (25 patients) GBM (22). PLR and NLR were assessed prior to surgery of GBM. On multivariate analysis, both PLR (≤175 vs. >175) and NLR (≤4 vs. >4) were identified as independent prognostic factors of OS. However, the prognostic value of NLR was limited to the subgroup of patients with isocitrate dehydrogenase wild type (IDH-wt). In the retrospective study of Yersal et al. that included 80 evaluable patients, pre-operative PLR and NLR were not significantly associated with PFS or OS (23). Lv et al. compared the pre-operative inflammatory markers PLR, NLR and the systemic immune inflammation index (SII) in a cohort of 192 patients with GBM from China (2). The optimal cut-off values for PLR, NLR, and SII were defined to be 87, 2.7, and 718, respectively. On univariable analyses, lower NLR and SII were significantly associated with improved OS, and lower PLR showed a trend (p=0.068) for better OS. In addition, NLR was significant in the multivariable analysis as well and considered superior to SII in predicting OS. In 2020, Marini et al. investigated the prognostic role of PLR (<175 vs. ≥175) and NLR (<4 vs. ≥4) in 124 patients receiving surgery for GBM (24). Lower PLR was significantly associated with better PFS and OS on univariable analyses, and lower NLR with better OS on univariable and multivariable analyses. Yang et al., who investigated associations between pre-operative inflammatory markers and OS in 187 patients with gliomas including 88 patients with GBM, used cut-off values different from other studies, namely <213 vs. ≥213 for PLR and <2 vs. ≥2 for NLR (25). In the subgroup of patients with GBM, lower NLR was significantly associated with worse OS on univariable and multivariable analyses, whereas PLR was not significant. In the retrospective study of Duan et al. from 2023, higher PLR (≥93.5) and higher NLR (≥2.12), both assessed prior to surgery, were identified as risk factors of shorter OS in a series of 281 patients with GBM (26). In contrast, a retrospective study from the same year including 89 patients with GBM from Romania did not find sufficient evidence that pre-operative inflammatory markers may serve as reliable prognostic factors in these patients (27).

In addition to these retrospective studies, very few meta-analyses and/or systematic reviews on the prognostic significance of PLR and NLR in patients with GBM are available (28-31). In 2020, Yang et al. performed a comprehensive meta-analysis and found a significant association between NLR and OS in patients with GBM (28). For PLR, such an association was not observed. In the meta-analysis of Guo et al. from 2022, higher pre-operative and pre-treatment NLR (cut-off value=4) were associated with worse OS (31). In the systematic review of Bispo et al. from 2023 that evaluated the prognostic role of pre-operative PLR in patients with GBM, seven of the 11 suitable studies found an association between PLR and OS and one study an association between PLR and PFS (29). In the systematic review and meta-analysis presented by Jarmuzek et al. in 2023, both higher pre-operative PLR and NLR were significantly associated with shorter OS (30).

Since the results of the available studies, systematic reviews and meta-analyses were heterogeneous with respect to the prognostic value of PLR and NLR in patients with GBM, additional studies are required. Moreover, in the vast majority of these articles, PLR and NLR were assessed pre-operatively. The present study investigated the role of PLR and NLR assessed prior to or at the beginning of the post-operative radiotherapy course. According to its results, PLR proved to be an independent predictor of PFS (multivariable analysis). Moreover, lower PLR showed a strong trend for an association with improved OS on univariable analysis and a trend in the subsequent multivariable analysis. In addition, trends were found for associations between lower NLR and treatment outcomes in terms of PFS and OS. The cut-off values used in our study, namely 150 for PLR and 3 for NLR, were well in the ranges of 87-213 (median=142.5) for PLR and 2-4 (median=3.57) for NLR used in previous studies (2,21-27). The prognostic significance of PLR can be used for the design of future clinical trials.

Moreover, in combination with other prognostic factors, such as number of GBM lesions, KPS, MGMT promoter methylation, maximum cumulative diameter of the GBM lesions, upfront resection, and addition of chemotherapy to radiotherapy, it may guide radiation oncologists when selecting the most appropriate dose-fractionation regimen for an individual patient (6). Since patients with low PLR have more favorable prognoses, they may receive conventionally fractionated radiotherapy with higher total doses and lower doses per fraction over 6 to 6.5 weeks (4,32). Higher doses will lead to better long-term control of GBM, and lower doses per fraction are generally associated with less radiation-related late toxicity (32,36,37). Patients with higher PLR and estimated poor or intermediate survival times appear good candidates for hypo-fractionated radiotherapy with 40 Gy in 15 fractions of 2.66 Gy given over 3 weeks (33,34). In the randomized trial of Roa et al., conventionally fractionated and hypo-fractionated radiotherapy were compared in patients with GBM aged ≥60 years (33). Median OS was poor, less than 6 months, in both groups. In another randomized trial of Roa et al. performed in elderly (age ≥65 years) and/or frail (KPS 50-70) patients with GBM, 25 Gy in 5 fractions of 5 Gy over 1 week (ultra-hypo-fractionation) was not inferior to 40 Gy in 15 fractions over 3 weeks with respect to PFS, OS, and quality of life (35). Thus, elderly and frail patients with high PLR and other factors suggesting very poor survival prognoses may be considered for ultra-hypo-fractionated radiotherapy. Before following these recommendations, the retrospective nature of the present study and previous studies need to be considered, since these studies bear the risk of hidden selection biases. Prospective trials are required to overcome this limitation.

In summary, given the limitations of our study, lower PLR prior to or at the beginning of radiotherapy was associated with improved treatment outcomes in patients irradiated for GBM. This factor should be considered when designing future clinical trials. Since the studies investigating the prognostic role of PLR and NLR are retrospective in nature, prospective trials are urgently required to properly define the predictive value of these inflammatory markers.

The Authors report no conflicts of interest related to this study.

The study was designed by all Authors. Data were collected by O.Z. and D.R., and analyzed by D.R. and N.Y.Y. The article was drafted by D.R. and finally approved by all Authors.

O.Z. received a scholarship from the University of Lübeck within the framework of the Emergency Aid Program for the Support of Refugee Academics from Ukraine.