ABSTRACT

Objective:

This study aimed to evaluate radiotherapy (RT) doses, RT effects on neurocognitive functions, and possible factors that may affect recurrence or death in patients with glioblastoma multiforme (GBM).

Methods:

The data of 21 patients with GBM were retrospectively analyzed. RT treatment plans and doses and hippocampus ipsilateral and contralateral doses were recorded. The Mini-Mental State Examination (MMSE) is used to assess neurocognitive functions. The time of recurrence and death, if any, of the patients was recorded. Factors such as gender, age, patient performance status, tumor size, tumor localization, type of surgery, and time between surgery and RT were analyzed to determine any effect on the risk of recurrence or death.

Results:

The median planning target volume dose was 59.86 gray (Gy). The maximum ipsilateral hippocampus dose was 51.85 Gy, and the maximum contralateral hippocampus dose was 46.25 Gy. With the MMSE, 3 of 4 patients had cognitive impairment. At the end of follow-up, 16 patients had recurrence and died. The median disease-free survival was 10 months [95% confidence interval (CI): 5.7-14.2], and the median overall survival was 24 months (95% CI: 16.0-31.9). Only poor performance status increased the risk of recurrence (hazard ratio: 4.31, 95% CI: 1.26-14.70, p=0.02).

Conclusion:

Because hippocampus shielding was not performed, our hippocampus doses were high. Hippocampal-sparing RT is essential for the preservation of neurocognitive functions. The increased risk of recurrence in patients with poor performance status is possibly related to treatment dose reduction, delay, or discontinuation.

INTRODUCTION

Primary brain tumors constitute approximately 2% of all cancers. In adults, glioblastoma multiforme (GBM) and grade 3 anaplastic astrocytoma (AA) are the two most common histological types (1,2). GBM, the most common and aggressive primary malignant intracranial tumor in adults, accounts for 50-60% of all astrocytic tumors. The patients are usually between 45 and 70 years of age. Diagnosis in childhood is rare. The male/female ratio is approximately 1.5/1 (3-5).

GBM is an infiltrating tumor located in the cerebral hemisphere. It is commonly at the borders of the parietal, occipital, and frontal lobes. They usually occupy more than one functional brain region because of their diffuse and deep localization, which is one of the main factors that make surgery difficult. They can also be localized outside the cerebral hemispheres. Almost half of brainstem glial tumors exhibit high malignancy characteristics. Furthermore, approximately 10% of GBMs are not deeply localized and may mimic cerebral metastases by localizing on the white-gray matter border (6-9).

The standard treatment is the Stupp protocol, which involves surgical resection of the tumor, followed by regional radiotherapy (RT) with concurrent temozolomide (TMZ) chemotherapy and six adjuvant cycles of TMZ (10). A randomized phase III study revealed that RT and TMZ combination prolongs survival in GBM. In this study, 573 patients were randomized to receive RT only or RT concurrent with TMZ, followed by adjuvant TMZ. RT was administered in 30 fractions [2 gray (Gy)/day] totaling 60 Gy. Concurrent TMZ was administered at a dose of 75 mg/m²/day continuously during RT, including weekends. Adjuvant TMZ was initiated 28 days after the completion of RT and was administered for 6 cycles (150-200 mg/m²/day, on 1-5 days every 28 days). In this study, the median overall survival (OS) was 12 months in the RT-alone arm and 15 months in the RT-TMZ arm. The 2-year survival rates were 10% and 26%, respectively. The median disease-free survival (DFS) was 5 months and 6.9 months. The 1- and 2-year DFS rates were 9% and 2% in the RT-alone arm and 27% and 11% in the RT-TMZ arm (p<0.0001). In this phase III study, the RT-TMZ combination demonstrated a remarkable improvement in survival (11).

In a phase II study by Athanassiou et al. (12), among 110 patients, RT alone (60 Gy) was compared with RT-TMZ treatment. The results of the study favored the combined treatment arm: the median DFS was 5.2 months vs. 10.8 months; the 1-year DFS rates were 7.7% vs. 36.6%; the median OS was 7.7 months vs. 13.4 months; and the 1-year OS rates were 15.7% vs. 56.3%, respectively. Toxicity was mostly hematological, and 1 patient was reported to have died due to febrile neutropenia.

The hippocampus plays an important role in emotional learning and memory consolidation (13,14). Whole-brain RT can lead to various side effects, such as the development of cerebellar and neurocognitive dysfunctions and impaired short-term memory and learning ability (15,16). Interruption of neurogenesis in the subgranular region can lead to memory impairment (17,18). The function of the hippocampus is negatively affected by radiation. Therefore, improving techniques to protect the hippocampus from radiation is vital. Dosimetric studies performed with intensity-modulated radiotherapy (IMRT) or tomotherapy in the literature show promise.

In our study, we evaluated RT doses and hippocampal neurocognitive effects in patients with GBM treated with the Stupp protocol. We also evaluated treatment tolerance, disease-free and OS times, and possible factors that may affect recurrence or death.

MATERIALS AND METHODS

In our study, we retrospectively evaluated 21 patients treated with a histologically confirmed diagnosis of GBM between April 2015 and October 2020. Inclusion criteria were as follows: being over 17 years of age, having Eastern Cooperative Oncology Group (ECOG) performance score ≤3, and having normal bone marrow, renal, and liver function (defined as hemoglobin ≥10 g/dL, thrombocyte count ≥100,000/µL, absolute neutrophil count ≥1,500/µL; serum creatinine level <1.5 mg/dL; alanine aminotransferase and aspartate aminotransferase levels <2.5 times the upper limit of normal, total bilirubin <1.5 times the upper limit of normal). Patients with poor medical condition due to other comorbidities or infections were excluded from the study.

The treatment protocol was TMZ concurrent RT with a median of 60 Gy (2 Gy-60 Gy) with standard 6 mV photon energy applied according to the Stupp protocol. In RT, we used VMAT and IMRT techniques in 30 fractions as planning target volume 1 (PTV1)=GTV+2 cm=40 Gy, PTV2=GTV+1 cm=60 Gy as standard. TMZ was administered concurrently with RT as 75 mg/m²/day throughout RT; and it was administered 150-200 mg/m²/day, on 1-5 days every 28 days for 6 cycles, following.

When evaluating neurocognitive functions, we first drew the hippocampus retrospectively to determine RT doses. We further evaluated the ipsilateral and contralateral hippocampus doses. Subsequently, the Mini-Mental State Examination (MMSE) was performed on 4 alive patients in the outpatient setting. The MMSE evaluates orientation, registration (short-term memory), recall, attention and calculation, language, and the ability to understand and follow verbal and written commands. A score equal to or greater than 24 indicates normal cognition. Scores below 24 indicate cognitive impairment (19-23 points: mild, 10-18 points: moderate, ≤9 points: severe impairment). The MMSE test was obtained from the internet (19).

Disease recurrence and death data were obtained from patient files. DFS was defined as the length of time between the end of primary treatment for cancer and the time of disease recurrence or death. OS was defined as the length of time from GBM diagnosis to death.

This study was conducted in accordance with the Declaration of Helsinki. Ethics committee approval was obtained from the Bursa City Hospital Institutional Clinical Research Ethics Committee for the study (approval number: 2019-KAEK-140, 2021-19/7, date: 20.10.2021).

Statistical Analysis

Statistical analysis was performed using SPSS software version 28.0 (IBM Corp., Armonk, NY, USA). The variables’ distribution is evaluated using the Shapiro-Wilk test, and the variables are given as median (minimum-maximum) values and frequency values. Categorical variables were compared using the chi-square test and Fisher’s exact test. The Kaplan-Meier test, with comparisons made with the log-rank test, was used for survival analyses. Cox regression analysis was used to evaluate possible factors on the risk of recurrence and death. A p-value <0.05 is considered statistically significant.

RESULTS

Twenty-one patients were involved in the study. After the median follow-up period of 66 months, the median DFS was 10 months [95% confidence interval (CI): 5.7-14.2 months] and the median OS was 24 months (95% CI: 16.0-31.9 months). The patients’ clinical characteristics are given in Table 1.

We administered TMZ to 21 patients concomitantly with RT. 4 patients could not complete RT because of ECOG performance status deterioration. Two patients died during RT. TMZ-induced pancytopenia developed in one patient, which required treatment delay. In the first-month control MRI after RT, 2 patients had residual/recurrent mass images and were referred for surgery. Adjuvant TMZ could not be administered to 5 patients because of ECOG performance status deterioration. In 1 patient, the dose was reduced by 10% because of grade 2 thrombocytopenia. Four patients had recurrence on adjuvant treatment. Three of them were referred for surgery. One patient underwent reirradiation and was administered second-line chemotherapy following. Seventeen of 21 patients who died during the follow-up period, 4 patients remained alive.

In the statistical analysis, we examined the factors that may alter the risk of recurrence or death. Upon evaluation for the risk of recurrence, being older than 50 years of age, being male, having poor ECOG performance score, smaller tumor size, total excision, tumors on right cerebral hemisphere or non-frontal location, and the interval between surgery and RT being more than 4 weeks were found as factors increasing the risk, with only poor ECOG performance score being statistically significant (p=0.02). Factors increasing the risk of death were the same as those for the risk of recurrence, except that having a left cerebral tumor seemed to increase the risk of death, and none of the factors were statistically significant. The Cox regression analysis results are given in Tables 2, 3.

RT critical organ doses are given in Table 4. We further examined the ipsilateral and contralateral hippocampus doses of 19 patients. The median ipsilateral maximum hippocampus dose was 61.21 Gy (minimum-maximum: 26.53-63.06), and the median contralateral maximum hippocampus dose was 48.01 Gy (minimum-maximum: 16.03-62.89). The ipsilateral and contralateral hippocampus doses are given in Table 5.

In our evaluation of 4 living patients with MMSE, we found 1 patient with normal cognitive function, one with mild cognitive function, one with moderate cognitive function, and one with severe cognitive dysfunction. The MMSE is shown in Figure 1.

DISCUSSION

In high-grade astrocytomas, the 5-year survival after treatment with surgery and RT is less than 10-20%; the median OS is 27 months in AA and 8 months in GBM (20).

The treatment of GBM consists of surgery, followed by RT and chemotherapy. Chemotherapy has been used in GBM treatment since the 1970s. Nitrosoureas are alkylating agents and are the oldest drugs used in the treatment of central nervous system malignancies. Carmustine (BCNU) is still one of the most effective chemotherapeutics, with a response rate of approximately 40%. In recent years, TMZ, a second-generation alkylating agent, has been used in GBM and AA and has been found to be as effective as BCNU, and has been included in standard post-RT treatment protocols (21,22). In a multicenter study, in 525 patients with recurrent GBM, with TMZ, a 6-month DFS rate of 46% and a 12-month OS rate of 24% were observed (23).

TMZ is a per oral applied, rapidly absorbed agent and shows nearly 100% bioavailability. TMZ can cross the blood– brain barrier and reach effective concentrations in the central nervous system, with a cerebrospinal fluid/plasma ratio of approximately 30-40% (24). TMZ improves the quality of life and prolongs DFS and OS (25). In older studies, the median OS was around 1-1.5 years. However, with improvements in cancer care, some patients with GBM have an OS of nearly 2 years. For example, in the GEINO 14-01 trial published in 2020, the median OS of patients who received 6 months of adjuvant TMZ was 23.3 months (26). Similarly, in a study comparing different RT techniques, the median OS of patients was around 18-25 months (27). In our study, the patients’ median DFS was 10 months, and the median OS was 24 months, which is consistent with recent data. In contrast, isocitrate dehydrogenase 1 (IDH1) mutation and O6-methylguanine-DNA methyltransferase (MGMT) methylation are known prognostic factors for GBM. The OS of class 1 patients (class 1=MGMT methylated/IDH1 mutant or MGMT methylated/IDH1 wild type/Gross Total Resection/Karnofsky Performance Status ≥90) reaches 67 months (28). The mutation status of our patients is unknown. The longer OS in our study could be the result of the high MGMT methylated percentage of patients in the study group.

While administering RT to the patients, we tried to keep our PTV 95 (PTV enclosed by the 95% isodose) and risk organs within the RTOG 0933 range, as shown in Table 2. However, we did not draw the hippocampus. In recent studies, it has been reported that deficits in learning and memory, especially in patients receiving whole-brain RT, are associated with the radiation-affected hippocampus (29,30). In a study using linac-based IMRT with hippocampus sparing for whole-brain RT, Gondi et al. (31) reported that they delivered 30 Gy to the whole brain. They determined the shape of the hippocampus and created hippocampal avoidance zones using a volumetric expansion of 5 mm around the hippocampus. According to their results, the maximum dose received by the hippocampus was 15.3 Gy, whereas the mean dose was 7.8 Gy. They reported that modern IMRT techniques provide hippocampus preservation with acceptable target coverage and homogeneity (31). According to RTOG 0933, the hippocampus maximum dose should be ≤16 Gy. Because we did not perform hippocampus protection, our hippocampus RT doses were high.

While it is possible to protect both hippocampi in whole brain RT, it is difficult to protect the same side of the hippocampus in tumors to which higher doses are administered, such as GBM. Wee et al. (32) showed that hippocampus preservation does not increase the risk of relapse.

Study Limitations

One of the major limitations of this study is the limited number of patients. Statistical significance could not be reached for the factors of disease recurrence or death. In addition, the genetic profile of patients for prognostic changes is unknown. The retrospective design of the study is another limitation. MMSE could not be performed on all patients because 16 patients included had already died at the time of evaluation.

CONCLUSION

In our study, the Stuppe protocol for the treatment of GBM was well tolerated. Because we did not preserve the hippocampus during RT, our patients showed loss of neurocognitive functions. We have thereby understood that we should pay more attention to protecting at least the opposite hippocampus according to the tumor location.