Intranasal GM-CSF enhances efficacy of local ACNU delivery rendezvousing with TMZ plus irradiation in glioblastoma

This study aims to investigate the efficacy and safety of intranasal granulocytemacrophage colony stimulating factor (GM-CSF) treatment rendezvousing with chemoradiotherapy for post-operative glioblastoma patients. A total of ninety-two patients were randomized into two groups: control group (n = 46), patients who received radiotherapy with concomitant and adjuvant local delivery of nimustine hydrochloride (ACNU) rendezvousing with systemic administration of temozolomide (TMZ); observation group (n = 46), patients who received intranasal GM-CSF prior to each cycle of adjuvant chemotherapy based on the control group. Karnofsky performance status (KPS) scores, progression-free survival (PFS), overall survival (OS) and adverse effects were compared between these two groups. Two patients in the control group were excluded due to grade 3 hematologic toxicity. Furthermore, the observation group was superior to the control group with regard to PFS (7.8 months vs. 6.9 months, P = 0.016) and OS (19.2 months vs. 17.1 months, P = 0.045 without adjustment for interim analyses). KPS scores was higher in the observation group than in the control group after six months (84.35 ± 8.86, 80.65 ± 7.72; t = 4.552, P = 0.036). Neutropenia and thrombocytopenia decreased in the observation group, with incidences of 8.7% and 8.7%, respectively, when compared with the control group (29.5% and 18.2%, respectively; P = 0.012); while other adverse events were similar in both groups. Most adverse events were grade I-II and resolved spontaneously. Intranasal GM-CSF enhances the efficacy of the local delivery of ACNU rendezvousing with oral TMZ chemotherapy associated with significantly improved survival and quality of life in glioblastoma patients after surgery. This therapy could relieve chemotherapy related neutropenia, and does not increase the adverse events of other aspects. www.impactjournals.com/oncotarget/ Oncotarget, Advance Publications 2018


INTRODUCTION
Glioblastoma multiforme (GBM) is the most aggressive and most common type of primary brain tumor. Despite recent advances in surgery, radiotherapy and chemotherapy, this strategy yields a median survival of 14.6 months, with a 2-and 5-year survival rate of 27.2% and 9.8%, respectively [1]. Recurrence is the main reason for poor prognosis [2]. Cancer stem cells (CSCs) are known to be chemoradiotherapy resistant. Thus, therapeutic strategies against CSCs are essential for the accomplishment of cancer eradication [3]. Other researchers have found that CSCs have a potential to create their own niche, which helps maintain the CSC phenotype and promote tumor progression [4,5].
Unlike low-grade gliomas, the complete surgical removal of glioblastoma is impossible due to the infiltrative nature of the disease and functional vulnerability of the brain [6]. Postoperative radiotherapy with the administration of temozolomide (TMZ) and local delivery of ACNU could remove most of the residual glioma cells, but not glioma stem cells. The cytotoxicity of TMZ is related to DNA methylation and the subsequent formation of O6methylguanine (O6-MeG), followed by cell cycle arrest at the G2/M phase [7,8]. Glioma stem cells (GSCs) have been considered to be the less differentiated populations in malignant tissues, and considered as cells responsible for the maintenance of tumor tissues, as well as for the relapse of tumors after conventional treatment [9]. GBM are among the first solid cancers, in which tumor cells with stem celllike features, such as the so-called CSCs, were identified. These cells are slow-dividing in vivo, suggesting that cell cycle quiescence underlies the chemotherapy resistance of CSCs, leading to glioma relapse.
Recently, there are many methods for influencing the growth of CSCs. From this viewpoint, the inhibition of CSCs is a promising strategy for glioma eradication. With the use of dose-intensive chemotherapy, GM-CSF has been widely applied to minimize chemotherapy-induced myelosuppression. In addition, GM-CSF has also been specifically used as an anti-tumor agent with varying degrees of success [11,12]. The expression of GM-CSF and its receptor genes within human glioma specimens have been previously reported. These genes were overexpressed in most malignant tumors [14]. Yoriko Saito et al. [10] found that granulocyte colony-stimulating factor (G-CSF) cytokine treatment induces quiescent human AML LSC entry into the cell cycle, significantly enhances chemotherapy sensitivity, and increases the elimination of LSCs. In vivo studies have shown that GM-CSF has potent anti-tumor effects via immune stimulation [13]. Although GM-CSF is commonly used for treating chemoradiotherapy-related hematological toxicity, its impact on the outcome of glioblastoma patients remains unclear.
With the attempt to improve the efficacy of radiotherapy with the addition of chemotherapy, the aim of this study was to evaluate an approach that would allow the safe and feasible addition of GM-CSF to ACNU interstitial chemotherapy rendezvousing with TMZ chemotherapy plus radiotherapy in glioblastoma patients.

Patient characteristics
Ninety two patients were enrolled between 2009 and 2012. However, two patients in the control group were excluded due to grade 3 hematologic toxicity. These patients received more than two days of GM-CSF treatment to relieve the toxicity. Furthermore, they were treated with antibiotics, antifungal agents, GM-CSF and red blood cell transfusions. The clinical characteristics of these patients are shown in Table 1. The final patient population consisted of 90 patients, including 52 male and 38 female patients. The median age of these patients was 51.9 years old (range, 19-65 years old). There were no significant differences between these two groups with respect to baseline characteristics (Table 1).

Survival outcome PFS and OS
The progression-free survival was 7.8 months for the observation group and 6.9 months for the control group, and the difference was statistically significant (P = 0.016). The observation group was superior to the control group with regard to overall survival (17.1 months vs. 19.2 months), and this difference revealed a borderline significance (P = 0.045) (Table 2, Figure 1A and 1B).

Prognostic factors
Survival times were compared between these two treatment groups on the basis of six prognostic factors (Table 1). Progression-free survival was better in patients with the methylated MGMT promoter than in patients with the unmethylated MGMT promoter (P = 0.008). Similarly, the methylated MGMT promoter was predictive of improved overall survival (P = 0.023). However, no other factors were significantly predictive of the outcome. Although there was no statistical difference, compared with the high protein levels of MGMT, low MGMT protein was predictive of better progression-free survival (8.05 months vs. 6.61 months, P = 0.104) and overall survival (19.24 months vs. 16.29 months, P = 0.062) ( Table 1).

Adverse events
The most common adverse effects were tolerable gastrointestinal reactions (manifested as anorexia, nausea, vomiting, diarrhea and constipation), hematological suppression (expressed as neutropenia, thrombocytopenia and hemorrhage), liver and kidney dysfunction, and electrolyte imbalance during the observation period (Table 3). Adverse effects were similar in the incidence of gastrointestinal toxicities, liver and kidney dysfunction, electrolyte imbalance, infection, hypersensitivity reaction and fatigue between the two groups. However, neutropenia and thrombocytopenia incidence significantly decreased when GM-CSF was used during each cycle of adjuvant chemotherapy, with incidences of 8.7% and 8.7%, respectively, when compared with the control group (29.5% and 18.2%, respectively) (X 2 = 6.381, P = 0.012). In addition, no III/IV grade chemoradiotherapy related adverse events occurred in the observation group during the follow-up. Furthermore, no patient died of adverse events in either of the groups. All other adverse effects were self-limited and resolved soon after the cessation of treatment, and well before the beginning of the subsequent treatment.

DISCUSSION
This randomized clinical trial evaluated the application of the GM-CSF regimen administered during the rendezvous chemoradiotherapy for glioblastoma, in order to enhance chemotherapy sensitivity and increase the elimination of GSCs. These results indicate that patients in the observation group had significantly superior progression-free and overall survival rates, and better performance status, indicating that GM-CSF therapy may prevent late relapse and improve quality of life.
Treatments that substantially reduce the tumor mass by removing proliferating cells fail to cure patients, because cancer stem cells are usually slow cycling cells,  and are thereby insensitive to these treatments [19]. B Auffinger et al. [23] provided evidence that glioma cells exposed to chemotherapeutic agents are able to convert into stem-like cells, replenishing the original tumor population, and leading to enhanced chemoresistance. According to our previous in vitro studies, the application of GM-CSF targeting glioma stem cells could sensitize chemotherapy, and thereby increase the clearance rate for GSCs. Bao et al. [16] postulated that cell-cycle delay might represent a mechanism for genome protection in glioma-initiating cells. Analogously, Yoriko Saito et al. [10] found that G-CSF cytokine treatment induces  quiescent human AML LSC entry into the cell cycle, significantly enhances chemotherapy sensitivity, and increases the elimination of LSCs. The study conducted by Malgorzata Sielska et al. [20] revealed that autocrine GM-CSF had little or no effect on cell viability or the proliferation of glioma cells. There were also conflicting data implicating GM-CSF in the progression of a number of malignancies. However, we used GM-CSF aimed to glioma stem cells, and not to normal glioma cells. All patients in this clinical study underwent craniotomy microsurgical total resection of the brain tumor, and received three-dimensional conformal radiotherapy rendezvous chemotherapy. After the first phase of chemoradiotherapy, most tumor cells were eradicated; but the remnants of GSCs are resistant to chemoradiotherapy, and were not easy to clear. GM-CSF cytokine treatment may induce quiescent glioma stem cell entry into the cell cycle, and significantly enhance chemotherapy sensitivity. In a therapeutic setting, the proportion of CD133+ cells (normally 5-30%) was enriched after irradiation in vivo and in vitro, resulting in the increased tumorigenicity of remaining cells [16].
In addition, the efficacies of systemically administered GM-CSF on glioma have previously been studied. In vivo studies have shown that GM-CSF has a potent anti-tumor effect via immune stimulation (Galea and Cogne, 2002). GM-CSF plays a critical role in the development and maturation of dendritic cells (DCs), as well as the proliferation and activation of T cells, linking innate and acquired immune response [15], and increasing DC-mediated responses to tumor cells [17,18]. More recently, this has been used in the treatment of a wide range of malignancies [13]. Nebiker et al. [11] further unraveled an important paradoxical colorectal cancer feature, which is represented by the favorable prognostic role of GM-CSF. Immune-dependent and immune-independent antitumor activities of GM-CSF in human colorectal cancer have been suggested. The hematopoietic cytokine GM-CSF has been investigated as a monotherapy, and as a component of combination therapies for melanoma [16]. Our results are consistent with the views of the above studies. In the present study, progression-free survival and overall survival were significantly extended for patients who received GM-CSF therapy, and their quality of life was significantly improved afterwards.
In clinical medicine, we used GM-CSF with the minimum recommended dose, which is safe. Patients received intranasal GM-CSF treatment, which is a noninvasive and practical alternative to other forms of administration. Studies have shown that the nasal route could be used to successfully deliver drugs to the central nervous system (CNS) [22]. Joseph Scafidi et al. [21] provided direct evidence that intranasal treatment is a plausible route to introduce sufficient HB-EGF into the brain and WM of critically ill VPT infants. This approach allows the GM-CSF cytokine to target the brain more directly, leading to less impact on other parts of body. Hence, it does not increase adverse events in patients.
The present study revealed that both the methylated MGMT promoter and low MGMT expression were independent prognostic markers for longer survival in patients with newly diagnosed glioblastoma. Our study results confirm that MGMT status is a strong and independent prognostic factor for survival in patients with glioblastoma. IDH1 gene mutation did not present a significant survival advantage, which is probably due to the limited number of patients.
In summary, this pilot study indicates that intranasal GM-CSF with rendezvous chemoradiotherapy is a novel therapeutic approach for the prevention of glioma relapse, which could enhance the efficacy of the local delivery ACNU rendezvousing with oral TMZ chemotherapy in glioblastoma patients who received GM-CSF prior to each cycle of adjuvant chemotherapy after surgery.

Patients
Patients who underwent full resection by microsurgery, had histologically confirmed glioblastoma (WHO class IV), were 18-65 years old, and had a good performance status (Karnofsky performance score, KPS ≥70) at two weeks post-operation were eligible for this study. Blood chemistry and hepatic and renal function were as follows: white blood cell (WBC) count ≥4 × 10 9 /L, hemoglobin level ≥100 g/L, platelet count ≥100 × 10 9 /L, aspartate transaminase (AST) level ≤40 IU/L, alanine transaminase (ALT) level ≤40 IU/L, and serum creatinine level ≤140 umol/L. MGMT protein, methylation states, IDH1/IDH2 gene sequences and markers of glioma stem cells (CD133+/Nestin) were detected. Baseline medication was recorded in all patients. Patients with multiple or disseminated tumors were excluded. In addition, the following patients were classified as ineligible: pregnant patients, patients who received insulin injection, patients who had myocardial infarctions and unstable angina pectoris within the last two months, and patients with mental disorders, a history of pulmonary fibrosis or interstitial pneumonia, or other forms of cancer that occurred within five years of the treatment period. All patients provided a written informed consent. This study was approved by the Life Science Ethics Review Committee of Zhengzhou University.

Study design and treatment
In the present study, a total of ninety-two patients were randomized into two groups: control group (n = 46), patients who received radiotherapy with concomitant and adjuvant local delivery of ACNU rendezvousing with systemic administration of TMZ; observation group (n = 46), patients who accepted GM-CSF combined with the above rendezvous chemoradiotherapy. Gross tumor volume (GTV) was defined as primary tumors with or without enhancement on magnetic resonance imaging (MRI). Patients in the control group received ACNU interstitial chemotherapy (2.5 mg/d, 3 d/week, intracapsular injection) rendezvous with TMZ chemotherapy (75 mg/(m 2 •d), 7 d/ week) for six consecutive weeks during radiotherapy. A total dose of 60.0-61.2 Gy was applied to the gross tumor volume, followed by six cycles of adjuvant ACNU interstitial chemotherapy (2.5 mg/d for three days during each 28-day cycle) and TMZ chemotherapy (150-200 mg/ (m 2 •d) for five days each 28-day cycle). The observation group received the same standard protocol, as described. The only difference between these two groups was the intranasal application of GM-CSF (3 μg/kg/d) on the first day and third day of each cycle of adjuvant chemotherapy in the observation group.

Surveillance and follow-up
Baseline and follow-up examinations included vital signs, subjective symptoms, neurologic examination, MRI and full blood count, and hepatic and renal function. All examinations were performed before the beginning of each cycle, and every two weeks or when they were clinically indicated during radiation and concomitant chemotherapy. MRI was performed every 4-8 weeks. KPS scores were recorded at post-operative 2, 4 and 6 months. Treatment was delayed for one or two weeks for patients with a neutrophil level of <1.5 × 10 9 /L or a platelet level of <100 × 10 9 /L. If a cycle was delayed for two weeks due to hematological toxicity, the drug dose was reduced by 25%. Treatment response was assessed by regular enhanced MRI scans. Progression-free survival was measured from the date of the initial operation to the date of tumor progression, death, or end of follow-up. A 25% or greater increase in tumor size, the development of new tumors, or patient worsening was treated as disease progression. Toxic effect data were also collected and included in the analysis. These toxic effects were graded in accordance with the NCI CTCAE version 3.0.

Statistical analysis
The Kaplan-Meier method was used to estimate the progression-free survival and overall survival distributions. Correlations between two independent variables were analyzed by t-tests, X 2 -tests, or Fisher's exact test. P < 0.05 (two-sided) was considered statistically significant. These analyses were performed using SPSS version 17.0.

CONFLICTS OF INTEREST
None.

FUNDING
This work was supported by the Key Programs of Science and Technique Foundation of Henan Province