The surgical treatment of intramedullary spinal cord tumors: A retrospective analysis of 76 patients.

Abstract

Intramedullary spinal cord tumors (IMSCTs) comprise about 20% of all intraspinal tumors (ISTs), and the two most common IMSCTs are ependymomas and astrocytomas, respectively.1 Low-grade ependymomas predominantly occur in adults, while low-grade astrocytomas are more likely to occur in children.2 With low incidence, high-grade astrocytomas are characterized by infiltrating, poorly circumscribed, hard to completely remove by surgery, and unfavorable prognosis. IMSCTs usually occur in cervical thoracic spinal cord with a slow onset and no specific symptoms in the early stage, and when the disease progresses to a certain extent, symptoms such as pain, dys- and paresthesias, spasticity, weakness and loss of bowel and bladder function may occur.3 Due to the specificity of the spinal cord in anatomical and physiological functions, surgical treatment of intramedullary tumors is difficult and has high risks. In the past, the treatment of intramedullary tumors is relatively conservative. With the rapid developments of microsurgical techniques and neuroimaging, the increasing popularity of minimally invasive surgery, and the improvement and renewal of surgical procedures, more and more patients have achieved good prognosis through early diagnosis and surgical treatment.4 To further summarize the clinical features, the experience of diagnosis and treatment of IMSCTs, following institutional review board approval, we identified 76 patients, including 40 males and 36 females with IMSCTs who were admitted to Neurosurgery Department, the First Hospital of China Medical University between January 2012 and June 2016. The average age of onset is 41.2 years (16-71 year). IMSCTs in some patients stretched across cervical thoracic and thoracolumbar spinal cord, and we chose the main section of lesion for statistical analysis. Of all the lesions, 32 were encountered in the cervical segment, 29 in the thoracic segment, and 15 in the lumbosacral segment. The pathological diagnosis was classified according to the classification criteria of WHO nervous system tumor (2007).5 Tissue type of the 76 patients was shown in Table 1. The mean course of disease of 76 patients with IMSCTs was 2 years (ranging from 1 month to 10 years). The clinical manifestations of the patients on admission were pain (48 patients, 63.2%), dyskinesia (45 patients, 59.2%), dysesthesia (57 patients, 75.0%), and sphincter dysfunction (29 patients, 38.2%). The preoperative spinal cord function of patients was evaluated using the McCormick scale.6 The score categories were as follows: I = neurologically intact; mild lesion damage, but no effect on limb function; normal gait; II = sensory deficit or mild motor; mild to severe gait disturbance; severe pain affecting the quality of life, but functional independence; III = moderate neurologically deficit; gait with support (wheelchair or crutch) or limitation of function of both upper limbs; can or cannot live by himself/herself; IV = severe motor or sensory deficit, wheelchair or crutch needed and limitation of function of both upper limbs; cannot live by himself/herself. Of the 76 patients, 6 patients were evaluated as grade I, 28 were grade II, 24 were grade III, and 18 were grade IV. There were 43 patients with ependymoma and the preoperative McCormick Scale grade was I in 3, II in 19, III in 15, IV in 6 patients. Preoperative magnetic resonance imaging (MRI) was performed on 76 patients to determine the location, size, shape of the lesion, and the lesion's relationship with the surrounding tissues and to infer the lesion character according to the different signals on the MRI. Surgeries were performed on patients in prone position under general anesthesia using posterior median incision approach. After the operation, the patients may undergo symptomatic treatments such as hormone therapy, dehydration therapy, anti-inflammation therapy, acesodyne therapy, hemostasis therapy, fluid replacement therapy. Early rehabilitation, functional exercise, and hyperbaric oxygen therapy, etc. should be performed on patients with spinal cord dysfunction after surgery. According to the histological types and the extent of lesion resection, the radiotherapy and chemotherapy can be considered. The location, tissue type, symptoms and signs, extent of tumor resection, and outcomes of surgery (outpatient or telephone follow-up) of IMSCTs were reviewed. Analysis was performed using IBM SPSS 22.0, and a P value < 0.05 was considered to be significant. Verla et al7 believe that with the development of microsurgical instruments and improvement of surgical techniques, the success rate of total resection of IMSCTs has been higher. In this study, all of the 76 patients were operated successfully without any operative death. Among them, total removal of the tumor was achieved in 49 patients (64.5%), subtotal excision in 17 patients (22.4%), and partial resection in 10 patients (13.2%). The extent of resection of different pathological types of tumors was shown in Figure 1A. Among IMSCTs in 76 patients, the total resection rate of ependymoma was significantly higher than that of nonependymoma (88.4% vs 33.3%, P < 0.05). The extents of excision of tumors in different locations were shown in Figure 1B. Among IMSCTs in 76 patients, the total resection rates of cervical tumors and thoracic tumors were significantly higher than that of lumbosacral tumors, respectively (81.3% vs 20.0%, 69.0% vs 20.0%, P< 0.05). According to the changes of symptoms and signs after the operation, outcomes of surgery were evaluated into four levels: improvement, stabilization, aggravation, and death. We reviewed the short-term outcomes (2 weeks after surgery) and long-term outcomes (6 months follow-up after surgery). Bedside assessments of the 76 patients with IMSCTs two weeks after surgery showed that tumor symptoms and signs were improved in 15 patients (19.7%), stable in 25 patients (32.9%), aggravated in 36 patients (47.4%), and no patient died (Figure 1C). The statistical significance of the short-term outcome of surgery performed on 76 patients with IMSCTs of McCormick Scale grade was discussed. They were evaluated as “improved and stable”: Grade I + II (64.7%), Grade III (54.2%),Grade IV (27.8%); “aggravation”: Grade I + II (35.3%),Grade III (45.8%),Grade IV (72.2%)(P< 0.05). The short-term outcome of surgery performed on 43 patients with ependymoma of different McCormick Scale grade showed no significant differences (P> 0.05). The prognosis data of 72 patients (4 patients lost) were obtained by outpatient or telephone follow-up with the period exceeding 6 months. The outcomes of long-term follow-up were as follows: tumor symptoms and signs were improved in 21 patients (29.2%); stable in 25 patients (34.7%); aggravated in 23 patients (31.9%); and 3 patients died (4.2%) (Figure 1D). Postoperative pathology of 3 dead patients revealed that there were one patient with anaplastic astrocytoma, one with glioblastoma, one with metastatic carcinoma; total tumor resections were not achieved; tumor recurrence and metastasis were occurred 7-15 months after surgery. Comparisons among the long-term follow-up outcome of surgery performed on 72 patients evaluated as preoperative McCormick Scale Grade I + II, Grade III, Grade IV showed a significant difference. The long-term outcome of surgery performed on 72 patients with IMSCTs is evaluated as “improved and stable”: Grade I + II (26 patients, 78.8%), Grade III (14 patients, 63.6%), Grade IV (6 patients, 35.3%); “aggravated and died”: Grade I + II (7 patients, 21.2%), Grade III (8 patients, 36.4%), Grade IV (11 patients, 64.7%) (P < 0.05). Outcome of long-term follow-up of 43 patients with ependymoma of different McCormick ScaleGrade (no death and no patient lost) showed no significant differences (P> 0.05)(Figure 1 C and D). In the study, we found that the most common intramedullary tumor was ependymoma, followed by astrocytoma, dermoid cyst, hemangioblastoma; the extent of tumor resection was related to histological type and lesion location; early diagnosis and surgical treatment were the best choice for most patients. Magnetic resonance imaging examination is currently the preferred method for the preoperative diagnosis of IMSCTs, by which most patients can get initial diagnosis. Surgery is the first choice for the treatment of intramedullary tumors so far, but the choice of the optimal time to perform surgery remains controversial.4, 8 In this study, postoperative short-term and long-term curative effect in the 76 patients with IMSCTs showed that when tumors were at McCormick Scale Grade I + II, the operation effects were good; when tumors progressed to Grade IV, the operation effects were poor. In terms that the comparisons among the short-term follow-up outcome of surgery performed on 43 patients with ependymoma of different preoperative McCormick Scale Grade showed no significant difference, neither did the long-term follow-up outcome, one reason is that the patient number is small and more data of patients need to be collected to perform the statistical analysis. Considering that complications such as edema of the spinal cord, small local blood flow, and microcirculatory disturbance may occur after surgery, some patients had transient neurological deficits after surgery, while most of them were gradually relieved within six months after the operation. Due to the limited number of patients, the groups with small number of patients were merged when designing grouping and performing statistical analysis and single factor analysis have been performed to investigate the surgical outcome of surgery performed on patients with tumors of different preoperative McCormick grades, and thus, confounding bias may exist. In the future, we intend to further explore 1.whether age (older age), tumor size, histological subtypes of glioma, preoperative nutritional status are related to the extent of tumor resection and tumor prognosis; 2. the clinical application value of regarding preoperative McCormick scale grade as an independent predictor of surgical outcome of IMSCTs on the basis of the enlarged patient number. The authors declare no conflict of interest.