Cohort population
Overall, 114 consecutive patients were operated for spinal ependymomas between January 2009 and June 2020 in our department. We excluded 21 foreign patients living abroad due to missing follow-up reports. Thirteen patients declined participation (reasons included the personal aspect of the questions or non-willingness to participate in clinical trials), and 13 patients did not respond to our contact attempt. Two patients were deceased, leaving 65 patients included in the study assessment. Complete data and answered questionnaires were available for all patients (flowchart, Figure 1).
Median age was 45 years (IQ range 36 to 57 years), and most patients were female (34/65, 52.3%). All patients underwent surgery for spinal ependymomas: WHO°1 (19/65, 29.2%), WHO°2 (45/65, 69.2%) or WHO°3 (1/65, 1.5%), according to the WHO 2016 classification.[14] Multifocal lesions were diagnosed in 10/65 patients (15.4%), and four patients suffered from tumor recurrence (6.2%). Tumors were classified as intramedullary in 41.5% (27/65) and extramedullary in 58.5% of the cases (38/65).
Most patients underwent a unilateral approach for tumor resection (41/65 patients 63.1%) or laminoplasty (20/65 patients, 30.8%) in one (38.5%, 25/65), two (26.2%, 17/65) or more segments (35.4%, 15/65). Mean surgery duration (LOS) was 183 minutes (IQ range 120-219 minutes), depending significantly on tumor location (extramedullary: mean 160 minutes, intramedullary: mean 215 minutes, p = 0.009).
Neurological status
Preoperative neurological deficits were diagnosed in 28/65 patients (43.1%), including mostly mild sensory deficits (21/65, 32.3%), motor deficits (12/65, 18.5%) and bladder dysfunction (4/65, 6.2%). Median preoperative modified McCormick grades showed only mild impairment: 75.4% had intact neurological function or, at most, minimal sensory deficits according to McCormick Grade I; 20% had mild neurological impairment, according to Grade II. Early postoperative deficits were observed in 37/65 patients (56.9%) with impaired sensory (31/65, 47.8%) and motor function (15/65, 23.1%). Regarding median postoperative McCormick grades, Grade II accounted for 24.6%, and 23.1% of the patients were classified as McCormick Grade III after surgery. At follow-up, 52.3% of the patients described persistent neurological deficits, again most commonly affecting sensory function (47.7%), 80% of the patients were functionally independent (modified McCormick grade I or II; Figure 2).
Quality of life
The analyzed questionnaires included the EQ-5D assessment and SF-36 patient-reported survey. The assessment was performed after a median interval of 5.4 years after surgery (IQ range 3-8.5 years).
The mean utility (u) evaluated by the EQ-5D questionnaire was 0.676 (range 0-1), indicating a high burden of disease and self-perceived disability. Most patients were restrained in the pain category, demonstrating the lowest scores in this part of the questionnaire. Utility differed significantly between patients concerning difficulties returning to physical activities (u: 0.58 vs. 0.81, p = 0.03). We also found a strong dependence on sex: female patients demonstrated significantly lower levels of overall quality of life, with a mean u of 0.573 vs. 0.788 in male patients (p = 0.006, correlation coefficient 0.287). Health utilities were significantly lower in patients suffering from intramedullary ependymoma (u = 0.50) compared to patients operated on for extramedullary tumors (u = 0.80, p = 0.000). Furthermore, a correlation was found between the overall quality of life and the modified McCormick grade at follow-up (I: 0.828, II: 0.627, III: 0.432, p = 0.001, correlation coefficient -0.447) as well as the presence of postoperative neurological deficits at follow-up (u: 0.554 vs. 0.809, p = 0.001, correlation coefficient -0.417). Age did not affect health-related quality of life (p = 0.364).
Assessing the SF-36 general health questionnaires, the lowest scores were observed for vitality (mean 49.8%, range 10-95%) and role limitations due to physical constraints (mean 51.4%, range 0-100%). The highest scores were reported in the categories of social functioning (mean 72.9%, range 12.5-100%) and emotional role limitation (mean 67.7%, range 0-100%). Patients suffering from difficulties returning to sport activities after surgery had significantly lower scores in all categories on the SF-36 survey compared to patients denying any problems in returning to daily living (Table 1). Patients with intramedullary ependymoma scored significantly lower compared to patients with extramedullary ependymoma (p = 0.001 to 0.005 in all subgroups).
Reintegration into professional employment
Assessing the pre- and postoperative occupations of our patients, we found a shift of full-time to part-time employment and patients who retired/were unable to work. Table 2 describes the findings focusing on pre- and postoperative education as well as pre- and postoperative occupation. Before surgery, the majority of patients were full-time employees (34/65, 52.3%). After surgery, the amount of full-time employment diminished to 21/65 (32.5%), and the number of patients who retired early increased from 16.9% to 29.2%. Reasons for failure or difficulties encountered during reintegration were persistent pain (13/65, 20%), physical stress (11/65, 16.9%), impaired accessibility of work (1/65, 1.5%) and motor deficits in two cases (3.1%). Age, WHO grade, and sex did not significantly influence the ability to return to work (age p = 0.240, WHO grade p = 0.595 and sex p = 0.621).
Sports and daily living activities
Before surgery, 66% of the patients performed individual sports and 15% trained and participated in team sports. After surgery, 63% continued to perform individual sports, but only 6% continued in team sports (Figure 2). Assessing the frequency of sport activities, most patients attended sport activities 2 to 4 times/week before surgery, whil 24.6% trained 5 to 7 times/week. After surgery, the amount of intensive training (5-7 times/week) diminished to 16.9% (11/65), while other patients remained active 2-4 times/week (47.7%, 31/65) (Figure 3).
When asked if the encountered problems whilst performing sport activities related to their surgical treatment, 67.7% (44/65) of the patients confirmed pain (12/44, 27.3%), coordination problems (8/44, 18.2%), fear of injuries (5/44, 11.4%), motor deficits (9/44, 20.4%) and fatigue symptoms (5/44, 11.4%) to be related with the intervention. Following beneficial exercise aspects were stated by 30 patients: increased mobility (15/65%, 23.1%), more health awareness (12/65, 18.5%) and social inclusion and participation (3/65, 4.6%).
Prognostic factors
With regards to difficulties returning or performing sports related to medical issues, we found a significant association between occurrence of difficulties and preoperative as well as postoperative McCormick grades (p = 0.012 and p = 0.002, respectively), chosen approach (unilateral approach 56.2% vs. laminectomy 100% and laminoplasty 85%, p = 0.023) and number of operated segments (monosegmental approach 52% vs. 5 or more segments 100%, p = 0.022).
Tumor dignity, sex and age did not affect participation in sport activities after surgery (p = 0.92, p = 0.994 and p = 0.510, respectively).
With regards to difficulties with professional reintegration, we found no significant association between early retirement or inability to work and tumor WHO grade, number of segments or sex (p = 0.595, p = 0.244 and p = 0.571, respectively). Patients reporting constraints returning to work tended to be younger (median age 42 years vs. 50 years, p = 0.056), but this finding was only suggestive significance.