Introduction

Teratoma is defined as a neoplasm composed of a variety of parenchymal cell types derived from three germinal layers (ectoderm, endoderm, and mesoderm) [1], and only 2% occur in the central nervous system [2]. Intraspinal teratoma was first described by Virchow in 1863 [3], and the first confirmed case of intradural spinal teratoma was reported by Gowers et al. in 1888 [4]. It is extremely rare and accounts for 0.2–0.5% of all spinal cord tumors [2, 5]. Furthermore, teratoma predominantly occurs in infants and young children [1, 2, 6, 7]. Surgery is the foundational treatment strategy for intradural spinal teratoma. Although adult-onset intradural spinal teratoma is usually mature and benign, and the recurrence rate was extremely similar between total and subtotal resection (9 vs. 11%) [8]. To avoid the immature tissue or residual malignancy, complete resection is strongly recommended [9].

There is little published information about the diagnosis and prognosis of adult-onset intradural spinal teratoma due to rarity. Most of these reports described isolated cases or fewer than three cases in a series (Table 2), and the only two series which included seven and eight adult patients older than 18 years emphasized the clinico-pathological features of the disease without discussing the diagnosis and prognosis [1, 30]. In this study, we retrospectively reviewed 18 patients with intradural teratomas who were surgically treated in our spine tumor surgery center, and compared them with the 84 reported cases in the literature since 1888 [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63]. To the best of our knowledge, this is the first largest series of adult-onset intradural spinal teratoma.

Materials and methods

Patients

From March 1998 to April 2013, 18 adult patients with intradural spinal teratomas were surgically treated in our center. They included eight males and ten females ranging from 18 to 67 years with a mean of 37.6 years. Of them, two (case #3 and #18) patients had a history of surgical treatment in other local hospitals 25 and 15 years ago, respectively, and were re-admitted in our center for local recurrence. The remaining 16 patients were considered as “intact” cases (Table 1). This study was approved by the Ethic Committee of Changzheng Hospital affiliated to the Second Military Medical University (Shanghai, China),and informed consent was obtained individually from the patients or their family members.

Table 1 Summarized data for all the cases in our study
Full size table

Radiologic studies and surgical strategies

Plain X-ray radiography, CT and MRI were performed in all patients before surgery. Radiological examinations were also required to estimate whether there was local recurrence and to observe the stability of the spine during the follow-up periods (Fig. 1). A posterior approach was performed for all patients. Finally, stability reconstruction was accomplished with posterior fixation in 13 patients of the 18 patients. The example images of reconstructed cases are presented in Fig. 1.

Fig. 1
figure 1

Images of the intradural spinal teratoma of a 32-year-old woman (Case #14). An ovoid shape, intramedullary space-occupying lesion with heterogeneous signal was presented on T1WI (a). The neoplasms extrude the conus medullaris from caudally, and tethered cord syndrome (TCS) was confirmed. Sagittal contrast-enhanced MRI showed just slightly enhancement (b). Postoperative radiographs performed 3 years after surgery and manifested that the posterior internal fixation from L1 to L3 was stable (c, d). In addition,none of local recurrence was detected through MRI (e)

Full size image

Pathologic studies

The finally pathologic diagnosis was confirmed postoperatively by two experienced pathologists independently. To avoid misdiagnosis of “teratoid cyst”, “epidermoid cyst” or “dermoid cyst”, the sections for pathologic evaluation included the entire extent of the specimen to disclose the presence of the elements from different germinal layers (Fig. 2).

Fig. 2
figure 2

Histopathology of the specimens. a (case #14) showing mature hair follicles, sweat glands, salivary glands, nervous tissues, adipose and connective tissues (Hematoxylin and eosin 200×). b (case #17) showing keratinized stratified squamous epithelium cells, hair follicles, and sebaceous glands (200×). c (case #15) revealing the cyst wall consist of ciliated columnar epithelium and squamous epithelium cells, and mature glial tissues (100 ×). d (case #10) depicting superficial layer of mature ciliated columnar epithelia, nervous, adipose, and connective tissues; adnexal structures, salivary glands, striated muscle are also present (100×)

Full size image

Neurologic evaluations

The pre- and postoperative neurologic status was evaluated by the Japanese Orthopaedic Association Scoring System (JOA score) [64]. The postoperative recovery rate was calculated by Hirabayashi’s method as follows: recovery rate = (postoperative JOA score − preoperative JOA score)/(11 − preoperative JOA score) × 100%, and classified as excellent (100–75%), good (74–50%), fair (49–25%), unchanged (24–0%), and deteriorated (decrease in score ≤0%) [64].

Follow-up evaluations

The follow-up evaluation was taken regularly through clinical visits and telephone interviews in the 3rd, 6th and 12th months after surgery in the first year and every 6 months for the next 2 years, and then annually. The recurrence status was confirmed on the basis of the clinical manifestations and imaging findings.

Literature review

The literature review was made on PubMed to sum up the clinical features of adult-onset intradural spinal teratomas from 1888 to the present. Due to the various terms assigned to the same lesion, the search was performed with the key words including “spinal’’, “intraspinal”, “intradural”, “teratoma”, “teratoid”, “teratoid cyst”, “cystic teratoma”, “teratomatous cyst”, and “cystic teratoid tumor”. “Adult” was defined as patients older than 18 years. Altogether 84 cases were filtrated and the clinical data were summarized in Table 2.

Table 2 Review of adult intradural teratoma cases reported in the literature, from 1888 to date
Full size table

Results

A total of 18 patients with adult-onset intradural spinal teratoma were included in our series, who ranged in age from 18 to 67 years with a mean of 37.6 years. The ratio of male to female was 0.8:1 vs. 1.47:1 in the literature (50 male and 34 female; age range 18–85 years, mean 40.02 years) [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63]. The long diameter of tumor ranged from 1.5 to 12 cm with a mean of 5.1 cm. In our series, tethered cord syndrome (TCS) caused by intradural teratoma occurred in two patients (case #7 and #14), and spinal bifida was observed in one patient (case #11). No other congenital anomalies such as anomalous vertebra, diastematomyelia, myelomeningocele, hemivertebrae and butterfly vertebrae, and hypoplasia of the spinous process were detected.

Intraoperative blood loss ranged from 200 to 1200 mL (mean 422.2 mL). Neither surgical complication occurred in any of our patients, nor was the incidence of spine instability observed during the follow-up periods. Histological diagnosis of mature teratoma was confirmed in all our cases, and pathologic evaluation showed well differentiated elements derived from all three germinal layers (Fig. 2). Of the 84 reported cases, 78 (92.86%) were mature teratoma, 2 cases developed with carcinoid tumor [22, 35], 2 were confirmed as malignant [30, 57], and one was immature [58].

The mean follow-up period was 79.7 months (median 60.5 months; range 27–208 months). Neurologic symptoms were improved to varying degrees postoperatively. The preoperative JOA score was 2–10 (mean 6.5) vs. 6–11 (mean 9.3) postoperation. The neurologic prognosis was excellent in 8 cases, good in 7, fair in 2, and unchanged in the remaining one. Local recurrence was only detected in 2 (11.1%) patients (case #1 and #2), who received subtotal resection and then experienced local recurrence 180 and 168 months after surgery, respectively. But both of them refused reoperation. All the four patients who received subtotal resection have still been symptomless despite the existence of the tumor until now. Meanwhile, none of local recurrence or tumor progression was detected in the rest 14 patients during the follow-up period.

Discussion

Teratoma is classified as mature, immature and malignant based on the degree of differentiation [35, 57]. There are controversies over the pathogenesis of intradural spinal teratoma. Currently, the most widely accepted hypotheses are dysembryogenic theory [31] and misplaced primordial germ cells theory [16, 30]. The former believes that teratoma originates from chaotically differentiated pluripotent cells in primitive streak or caudal cell mass due to dysfunction of several factors involved in gene function and cellular inductive interactions [31]. The latter considers that certain pluripotent primordial germ cells of the neural tube are misplaced into the dorsal midline during their normal migration from the primitive yolk sac to the gonadal ridges [16, 30], resulting in intradural teratoma formation. Associated dysraphic malformations are regarded as the strong evidence to support the dysembryogenic theory [65]. Especially in India and Turkey, the incidence of congenital anomaly or spinal dysraphism is significantly high [65,66,67]. Whereas, significant dysraphism is absent in greater percentages of adult intradural teratomas. In our study, we only found 2 cases of TCS and 1 case of spinal bifida (16.7%) vs. 35 cases (41.7%) in the literature. Therefore, some researchers insisted that adult intradural spinal teratoma without dysraphism supports the misplaced germ cell theory [1, 30, 39, 56]. However, neither of the theories can accurately elaborate the pathogenic mechanism. In our opinion, the hypothesis proposed by Makary and his colleagues [45] is most acceptable at present. They proposed that there may be a cause-and-effect relationship between spinal cord malformation and the migration of primordial germ cells. Pluripotential cells may get entrapped in an abnormal environment, resulting in teratoma formation due to chaotic differentiation induced by the dysfunction of genetic and cellular inductive interactions during embryogenesis [31].

The tumors in our series were predominantly located near the conus medullary, including 4 intramedullary cases and 14 extramedullary cases. Of the 84 cases reported in the previous literature, 59 (70.2%) cases were in the thoracolumbar junction and lumbar spine, and most of the reported cases were intramedullary (n = 53, 63.1%). The reason why the incidence of intramedullary lesions in our series was much lower than the literature was that we had classified some exophytic intramedullary teratomas (case #1, 8, 11, 13) into extramedullary cases. If these patients were categorized into intramedullary cases,the incidence of our study would be 44.4%. We believe these differences are mostly because of the limited sample size of our series or racial differences of Chinese population.

The preoperative definite diagnosis of adult-onset intradural spinal teratoma is usually difficult, because it is quite similar to other space-occupying intradural tumors, especially when the congenital anomaly is not associated. The clinical history of the adult-onset intradural spinal teratoma is usually lengthy and progressive or intermittent [29]. In our series, the duration dates from the initial symptoms ranged from 0.5 to 360 months with a mean of 50.1 months. The main presenting symptoms in our patients were pain, weakness and numbness of the lower extremities, and sphincter dysfunction, depending upon the location of the lesion. Remission of the symptoms during the course of disease may be best explained by varying activity of glandular elements or resorption or leakage of the cystic contents [9, 54]. The leakage of keratin, cholesterol and lipoid may cause recurrent aseptic meningitis [68]. This specific syndrome usually indicates the existence of intradural teratoma. Occasionally, various cutaneous markers such as skin stigmata, hairy patch and dimple, and dermal sinus can be helpful to disclose the presence of teratoma. Thus, a thorough physical examination is necessary.

X-ray radiography often demonstrates erosion of the vertebral bodies, widening of the interpedicular space, and thinned laminae, with or without significant associated vertebral anomalies (Fig. 3), such as spina bifida, vertebral body fusion, incomplete segmentation of vertebraes, diastematomyelia, butterfly vertebrae, etc. It is usually performed as the first diagnostic procedure with limited information provided. CT scanning is available to reveal mixed-density lesion that occupied the intradural space or intralesional calcifications, and assess the spinal bony structure. But it is incapable of visualizing the demarcation between tumor and spinal cord parenchyma. The variation of MRI signal is heterogeneous because of the teratoma’s multicomponent nature and dependency on the degree of cystic and solid components [63]. Lipomatous tissue presents hyperintense signal on T1 sequences while hypointense on T2 and fat suppression sequences. Calcification presents hypointense both on T1 and T2. Furthermore, contrast-enhanced MRI usually shows no enhancement. However, it is difficult to distinguish teratoma from the other intradural tumor types by CT and MR images [24, 27], except when the intratumoral calcification and the adipose tissue are present simultaneously (Fig. 4). The final diagnosis must be confirmed via histopathology [52].

Fig. 3
figure 3

The X-ray demonstrated the erosion of dorsal vertebral bodies indicating the existence of intraspinal lesion (a, case #10). Sagittal T1WI confirmed the ovoid intradural neoplasm located at the level of L2 with heterogeneous signal (b, case #10)

Full size image
Fig. 4
figure 4

T1WI showed a spindle shape, intramedullary space occupying lesion with heterogeneous signal (a, case #17). Axial CT scanning revealed the intratumoral calcification distinctly (b, case #17)

Full size image

Undoubtedly, surgical resection is an essential treatment to relieve the nerve compression in adult-onset intradural spinal teratoma at present [52, 53, 60]. It can prevent neurologic function from deteriorating effectively, and most patients have achieved preferable outcomes after surgery [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63]. Although, the recurrence rates were extremely similar for total and subtotal resection (9 vs. 11%) in the published literature [8]. In addition, adult-onset intradural spinal teratoma occurs more frequently as intramedullary. Therefore, subtotal resection is usually performed to remove the tumor as much as possible, while the integrity of neural function is preserved. Nevertheless, no patient who underwent total excision experienced recurrence. In contrast, 3 of the patients who received subtotal resection suffered relapse as reported in the literature [8, 22, 58]. In our series, two patients (50%) who received subtotal excision experienced local recurrence. To avoid misdiagnosis and decrease the incidence of relapse, total tumor resection and thorough sample examination are strongly recommended, although total resection has higher technical requirements [9].

To reduce the secondary morbidity during surgery, we applied microsurgical techniques, with monitoring somatosensory evoked potentials (SEP) and electromyography (EMG) meticulously [69, 70]. Combination of SEP and EMG can effectively reduce the mechanical injuries of conus medullaris or nerve roots, increase the success rate of tumor resection, estimate the curative effect of surgery and predict post-operative status in real time [69, 70]. Needle aspiration can not only prevent the leakage of cystic fluid into the subdural space but also diminish the tumor’s volume and be conducive to remove the neoplasm [59]. In our experience, when TCS is associated, the conus medullaris (CM) should be released sufficiently, including excising the thickening and shortening filum terminals, excising the tight tractive fibrovascular bridles and detaching the cauda equina from the teratoma and surrounding tissues. After removal of the tumor, the surgical field should be washed thoroughly with normal saline to prevent postoperative aseptic meningitis [71]. The dural sac should be closed in a watertight manner to prevent leakage of cerebrospinal fluid [72, 73].

Because of the slow-growth and indolent nature, the recurrence rate of adult-onset intradural spinal teratoma is low, and the prognosis is usually favorable. By far, adjuvant therapy for adult-onset intradural spinal teratomas remains controversial and obscure [8] due to the extremely low incidence and limited experience. Only two reported cases received radiotherapy [22, 57]. As long as malignant or immature elements are found by pathology, radiotherapy should be performed immediately even though the tumor is completely resected [39]. The dose of irradiation is always below the widely accepted spinal cord tolerance dose of 45–50 Gy [74]. Potential adverse effects of radiotherapy should also be evaluated [8, 62]. The more recent estimate of 57–61 Gy in conventional fractionation [75] may cause a 5% complication rate in 5 years. To the best of our knowledge, adjuvant chemotherapy has not been applied in adult-onset intradural spinal teratomas to date [8, 54, 62].

Conclusion

Adult-onset intradural spinal teratoma is extremely rare, and only 84 cases have been reported in the previous literature since 1888. To the best of our knowledge, this study including 18 cases is the first largest series so far. Due to the rarity and the lack of definite radiological criteria, preoperative diagnosis is usually difficult, provided intratumoral calcifications and adipose tissues are present simultaneously on CT and MR images. Recurrent aseptic meningitis or associated congenital anomaly is most valuable for diagnosis. Complete total resection is favored whenever possible [26, 35, 36, 38,39,40], but caution should be exercised when the tissue is intimately involved with the spinal cord parenchyma. The slow-growing nature of the mature forms allows for symptom-free survival even with incomplete resection [6], and malignant change is also extremely rarely seen [35]. Hence, we should not excessively pursue radical total resection, and neurologic function should be protected as much as possible. To avoid misdiagnosis, the whole excised specimen should be thoroughly sampled for histopathologic diagnosis. Once malignant or immature elements are found, radiotherapy should be performed promptly. The dose and protocol should be designed individually to avert radiation myelopathy.