Depressive and anxiety disorders worsen the prognosis of glioblastoma

Questionnaires reveal significant depressive and anxiety disorders in GBM patients

First, we used the PHQ-9 and GAD-7 scales (Tables 1 and 2) to assess depression and anxiety in our enrolled patients and a group of healthy controls. The basic information of all enrolled patients was provided as Supplementary Table 1. As shown in Table 3 and Figure 1, after learning about their diagnoses, the GBM patients all exhibited significantly greater depression and anxiety than their control conterparts. We then examined the severity distribution of the PHQ-9 and GAD-7 scores in GBM patients. As shown in Table 4, 90% of GBM patients displayed minimal to mild depressive disorder, and 97.6% displayed minimal to mild anxiety disorder. Eight (10%) and two (2.4%) patients exhibited moderate depression and anxiety, respectively. The more detailed PHQ-9 and GAD-7 scores are provided in Supplementary Table 2.

The relationship between the questionnaire results and prognosis in GBM patients

To evaluate the relationship between the PHQ-9/GAD-7 scores and prognoses of our GBM patients, we determined their overall survival times using telephone calls or chat tools. All the patients had undergone brain surgery with the same group of doctors. As shown in Figure 2A, the PHQ-9 and GAD-7 scores both exhibited significant negative correlations with the overall survival time.

To further examine this relationship, we divided GBM patients into “high” and “low” groups according to whether their PHQ-9 scores were higher or lower than the average. Likewise, we divided the patients into “high” and “low” groups according to whether their GAD-7 scores were higher or lower than the average. In total, 22 patients had higher-than-average scores on both the PHQ-9 and GAD-7 scales (Supplementary Figure 1), while 27 patients had lower-than-average scores on both scales (Supplementary Figure 2). Cox regression analyses were used to evaluate the mortality of patients in the high and low groups for each scale. As shown in Figure 2B, GBM patients with higher PHQ-9 or GAD-7 scores exhibited poorer survival than their lower-score counterparts.

It is well established that men and women react significantly differently to similar psychological factors. As shown in Figure 2C, male patients had significantly lower PHQ-9 scores than female patients, while there was no significant difference in GAD-7 scores between the genders. The effects of depression/anxiety on the overall survival time differed between the genders. Male patients with higher PHQ-9 and GAD-7 scores exhibited shorter overall survival times than male patients with lower scores (Figure 2D). However, female patients with higher PHQ-9 scores did not exhibit significantly shorter overall survival times than female patients with lower scores, while female patients with higher GAD-7 scores did display shorter overall survival times than female patients with lower scores (Figure 2E).

MRI patterns of the high and low PHQ-9 and GAD-7 groups

Our previous research demonstrated that different MRI patterns reflect the prognoses of GBM patients [19]; thus, we collected MRI images from patients in the present study (Supplementary Figure 3). As shown in Table 5, a significantly greater proportion of patients presented with large peritumor edemas, irregular edema shapes and necrotic tumors in the high PHQ-9 group than in the low PHQ-9 group; however, there was no significant difference in the proportion of patients whose tumor or edema crossed the midline of the brain, who exhibited contrast enhancement or who had a tumor larger than the median size. As for the GAD-7 scoring groups, we observed a significantly greater proportion of patients with necrotic tumors in the high GAD-7 group than in the low GAD-7 group. Radar charts were produced to more vividly depict the MRI characteristics of the different groups (Table 5, Figure 3A and 3B).

Differences in cell biology between the high and low PHQ-9/GAD-7 groups

GBM cells with more aggressive proliferation and invasion abilities may be more likely to cause necrosis within the tumor itself [18, 20]. Our MRI images revealed that the proportion of patients with necrotic tumors was greater in the high PHQ-9 and GAD-7 groups than in the corresponding low groups. To further explore this phenomenon, we conducted cellular experiments using primary GBM cells from patients who obtained high or low scores on both the PHQ-9 and GAD-7 questionnaires.

Sphere formation assays are commonly used to detect differences in cell proliferation. As shown in Figure 4A–4C, the sphere formation ability was significantly greater in the high PHQ-9/GAD-7 score group than in the low group. Although the mean diameter of the spheres did not differ significantly between the high and low groups on day 3 of cell culture, the growth rate of the diameter of the spheres within seven days was significantly faster in the high group than in the low group (high group: day 3, 23.67 ± 1.77 mm, day 5, 101.12 ± 6.25 mm, day 7, 127.33 ± 7.89 mm, n = 3; low group: day 3, 25.33 ± 1.77 mm, day 5, 74.00 ± 3.61 mm, day 7, 98.61 ± 4.06 mm, n = 3). After seven days of cell culture, we also calculated the number of spheres that developed from single cells in both groups. The number of spheres was significantly greater in the high group than in the low group (high group: 4.6 ± 0.51, n = 5; low group: 2.8 ± 0.37, n = 5). In addition, we observed a significantly greater proportion of 5-ethynyl-2-deoxyuridine (EdU)-positive cells in the high group than in the low group (Figure 4D and 4E). These results indicated that primary GBM cells from the high group had stronger proliferation abilities than those from the low group.

We also conducted a wound healing assay to compare the invasive abilities of primary GBM cells from the two groups. After 24 hours, we observed significantly better wound healing in the high group than in the low group (Figure 4F). These results indicated that GBM cells from the high group had greater invasive capacities than those from the low group.

Expression patterns of mesenchymal and epithelial markers in both groups

GBM has been divided into different subgroups, namely proneural, classical and mesenchymal subgroups, of which the mesenchymal subgroup displays the most aggressive and proliferative cell characteristics both in vivo and in vitro, as well as the worst patient prognosis [21]. Stimuli such as inflammatory responses, immune factors, etc. can induce the transformation of other GBM subgroups (especially the proneural subgroup) into the mesenchymal subgroup [22]. This process is called the proneural-mesenchymal transition, which is essentially the epithelial-mesenchymal transition (EMT) of GBM [23]. We examined the expression of mesenchymal (Vimentin and CD44) and epithelial markers (E-cadherin and SRY-Box Transcription Factor 1 [SOX1]) in cultured GBM cells from the high and low PHQ-9/GAD-7 score groups. As shown in Figure 5A and 5B, Vimentin and CD44 levels were significantly higher in the high-score group than in the low-score group, while E-cadherin and SOX1 levels were significantly lower in the high-score group than in the low-score group.

Subjects

We enrolled a total of 84 patients (40 men and 44 women) ranging from 32 to 68 (44.11 ± 1.09) years old with pathological diagnoses of GBM in our project. All the enrolled GBM patients had Karnofsky Performance Status scores greater than 70. For the controls, we randomly selected 45 healthy adults from different backgrounds (26 men and 19 women) ranging from 33 to 63 (46.02 ± 1.12) years old. This study was approved by the local Ethics Committee and conformed to the principles outlined in the Declaration of Helsinki. All patients provided written informed consent.

Instruments

In order to accommodate the different educational backgrounds of patients of different ages, we selected the PHQ-9 and GAD-7 scales to assess the presence of major depressive disorder and anxiety disorder. These scales are widely used as reliable measures of depression and anxiety [10, 40]. All patients were asked to complete the two questionnaires after being diagnosed with glioma (even though the diagnosis of GBM had not yet been pathologically verified). For the control group, a community survey was conducted. The detailed items of the PHQ-9 and GAD-7 surveys are presented in Tables 1 and 2.

MRI feature definition

The definitions of MRI features in GBM patients were detailed in our previous article [19]. In short, for all patients, the unidimensional maximum diameter in centimeters was used to measure the tumor size on T1-W images. The median tumor size was 4.8 cm (ranging from 2.2 to 9.9 cm). In accordance with the method of Hartmann [41] peritumor edema was morphologically classified based of T2-W images, and the severity was determined based on whether the edema was ≥ 1 cm or < 1 cm. Necrosis was assessed on axial contrast-enhanced T1-W images, and was established when a region had a high signal on T2-W images, a low signal on T1-W images and an irregular enhancing border on contrast-enhanced images. A cyst was defined as a rounded region with a low T1-W signal, a very high T2-W signal matching a cerebrospinal fluid signal, and a thin, smooth, regular and slightly enhancing or non-enhancing wall. Contrast enhancement of a tumor was classified as not obvious (less than the signal of fat) or obvious (similar to that of fat). The specific classification of imaging features is described in Supplementary Table 2. Using these classification methods, two experienced radiologists without knowledge of our research independently analyzed the imaging data of all patients.

Culture of primary GBM cells

Briefly, jelly-like tumor tissue was obtained during surgery and placed in a sterilized 50-mL centrifuge tube containing ice-cold phosphate-buffered saline. Then, the tumor tissue was carefully transported from the operating room to the laboratory in an icebox. The supernatant was discarded, and the tissue sample was placed in a sterile dish and cut into 1-mm³ pieces with sterilize scissors and tweezers. The tissue pieces were transferred to a 15-mL centrifuge tube, and phosphate-buffered saline containing 1% penicillin-streptomycin was added. The sample was mixed and shaken up and down three times to remove any remaining red blood cells as thoroughly as possible. Once the upper layer of liquid became clear, the supernatant was carefully removed and about 3 mL of trypsin was added for every 2 cm³ of tissue. The sample was incubated at 37 °C for 10 min and shaken every 2 min to ensure that the tissue was fully digested. After the digestion, the tube was allowed to stand for 2 min before the supernatant was transferred to an Eppendorf tube and centrifuged for 5 min at 1000 rpm.

After the centrifugation, the pelleted cells were resuspended in Dulbecco’s Modified Eagle’s Medium containing 10% fetal bovine serum, and were transferred to a culture dish. The cells were incubated at 37 °C with 5% CO₂, and the medium was changed every two days. The primary GBM cells appeared as irregular spindle-like cells under a microscope. Primary GBM cells were cultured in stem cell culture medium – either Dulbecco’s Modified Eagle’s Medium or neurobasal medium (Life Technologies Corporation) supplemented with 20 ng/mL fibroblast growth factor 2 (PeproTech), 20 ng/mL epidermal growth factor (PeproTech), B27 and N2 supplemental factors (Gibco) and antibiotics (penicillin and streptomycin). Isolated GBM cells grew into spheres after one to two months of continuous culturing. GBM stem cells were tested for their capability to self-renew using the sphere-formation assay described below.

EdU assay

The EdU assay was conducted according to the manufacturer’s instructions. Briefly, primary GBM cells from the different groups were plated and cultured in a 12-well culture plate with a cover glass. The cells were then treated with 50 mM EdU for 2 h, followed by Azide-555 staining and DNA staining. Images were acquired and analyzed with FLUOVIEW FV 10i (Olympus).

Sphere formation assay

GBM spheres were harvested and dissociated with 0.25% trypsin. After centrifugation, the suspensions of expanded GBM cells were seeded in a 96-well plate (three to four cells per well). On days 3, 5 and 7 of incubation in GBM stem cell proliferation medium, the long diameter and shape of each sphere was measured. A digital still camera (Axio Observer A1) was used to convert the fixed area (10 mm²) at the center of each well into a digital image, and Image-Pro Plus 6.0 (Media Cybernetics) was used to count the number of GBM cell spheres.

Western blotting

Western blotting was performed on primary GBM cells. Briefly, 50 mg of total protein from each group was separated on a 10% sodium dodecyl sulfate polyacrylamide gel and transferred to a 0.22-mm polyvinylidene difluoride membrane (Millipore). The membrane was blocked with 5% skim milk at room temperature for 2 h and then incubated with specific primary antibodies at 4 °C overnight. The membrane was then incubated with the appropriate horseradish peroxidase-conjugated secondary antibody (diluted 1:5000; Boster) at 37 °C for 1 h. Protein bands on the membrane were visualized with an enhanced chemiluminescence kit (Millipore) using a FluorChem FC system (Alpha Innotech Corporation).

Statistical tests

All variables are presented as the mean ± standard error. Since the variables were normally distributed, we used Student’s t-test or one-way analysis of variance for statistical analysis. GraphPad Prism software was used (Version 5.01; GraphPad Software, Inc). P < 0.05 was considered statistically significant.