Introduction

Chordoma is a rare malignant tumor with an incidence of 0.08 per 100,000, accounting for 1–4% of all bone malignancies [32]. Although reported to be rare, chordoma is the most common primary malignant tumor in the sacrum [12]. Chordoma has shown to respond poorly to radiotherapy and chemotherapy, so surgical resection represents the first choice of treatment [8, 21, 25, 35]. Because chordomas are highly recurrent, with the local recurrence rates ranging from 43 to 85%, complete tumor resection is essential for local control [21, 25]. Clinically, the tumors tend to be large at the time of treatment, posing a great challenge for surgeons to keep the balance between adequate resection and sparing of surrounding vital structures.

En bloc resection with adequate margins is an effective way of achieving long-term disease control or cure, but it is technically demanding and can be associated with a high risk of postoperative morbidity due to nerve sacrifice [13, 25]. Piecemeal resection may better preserve the nerve function, but often lead to a high rate of recurrence [27]. Sometimes surgeons and patients have to make a difficult choice between disease control and neurological function, for it is hard to have both when facing a large sacral chordoma.

In this study, 47 patients with large sacral chordoma were retrospectively reviewed. The purpose of this study was to provide some reliable and useful information contributing to surgical decision making by (1) identifying prognostic factors for large sacral chordoma and (2) comparing quality of life between patients undergoing en bloc and nerve-sparing piecemeal resection.

Materials and methods

Study design and patients

A retrospective review was performed of sacral chordoma cases between January 2010 and January 2018. The inclusion criteria were (1) patients with pathologically diagnosed chordoma in sacrum who received surgical treatment in our center, (2) patients with a tumor size ≥ 5 cm in the maximum diameter, (3) patients with tumor involved no less than two complete sacral segments, (4) patients who received en bloc or piecemeal sacrectomy instead of intralesional curettage of the lesion, (5) and patients with no metastasis at the time of surgery. In this study, sacral chordomas with a tumor size ≥ 5 cm and involving two or more segments were regarded as large sacral chordomas. The rationale for choosing large sacral chordomas as the study objects was that sacrectomy was applied in most of them. In addition, patients with sacral chordoma involving only S4 and below were excluded from the study, because the influence of surgery on nerve function and quality of life for this kind of patients is limited [37]. This study was approved by the hospital ethics committee.

The clinical and operative records, radiographic images, and pathological reports of all patients were reviewed by two individual researchers. Pre- and postoperative urinary, bowel, and ambulatory functions were recorded. Frankel Grade was used to evaluate neurological functions of the patients. The individualized surgical strategy was made for each patient based on the Weinstein–Boriani–Biagini and the Enneking system [3, 11]. After a detailed and comprehensive explanation to the patient and family about the benefits and the risks of nerve sacrifice and consequent functional cost, the patient and family decided whether or not to receive en bloc resection.

The surgical technique and protocol described by Zang et al. was followed for en bloc sacrectomy [36]. A horizontal sacral osteotomy from the back to the front was performed at the level determined on preoperative MRI with the goal of obtaining wide margins. The sacral dural sac was ligated above the sacral foraminal invasion of the tumor in the en bloc sacrectomy, while the sacral nerves were carefully preserved in piecemeal resection. In addition, cisplatin or carboplatin dissolved in distilled water was applied intraoperatively for local chemotherapy. Adjuvant therapies including postoperative radiotherapy and chemotherapy were applied based on a comprehensive consideration of the volume of the tumor, surgical method and the response to chemotherapy and/or radiation therapy by the multidisciplinary team.

Follow-up strategy

Patients were followed up at 3, 6, and 12 months after surgery, every 6 months for the next 2 years, and once a year thereafter. The diagnosis of recurrence was confirmed by postoperative pathological evaluation in patients who received a second surgery. In suspected cases without a second surgery, the diagnosis of recurrence depended on the clinical manifestation of reemerging of pain and/or lump in the original tumor site, and imaging finding of new-onset neoplasm at the surgical site evaluated by enhanced MRI. Recurrence-free survival (RFS) was defined as the interval between the date of surgery and the date of recurrence. Overall survival (OS) was defined as the interval between the date of the initial surgery and the date of death. The follow-up period was defined as the interval from the date of surgery to death, or until December 2018 for alive patients. The last status of patients was obtained from office visit or telephone interview.

Assessment of quality of life

The Functional Assessment of Cancer Therapy-General (FACT-G, version 4.0) questionnaire was applied in the assessment of quality of life in this study [19]. The FACT-G questionnaire is one of the most widely used instruments for quality-of-life evaluation and has been verified as applicable to the Chinese population [33]. The Chinese-language version of the FACT-G questionnaire was administered at 1-year follow-up for our patients. All FACT-G questionnaire data were collected and checked by two individual researchers, and missing data were minimized through telephone calls.

Statistical analysis

All statistical calculations were performed by SPSS Statistics, version 22.0 (IBM corp., New York, USA). The Kaplan–Meier method was adopted to estimate the RFS and OS time, with the log-rank test to identify the difference. Factors with a p value < 0.1 were subjected to multivariate analysis using the Cox proportional hazards model. The total FACT-G score and the subscores were compared by Student’s t test between the piecemeal resection and en bloc resection groups. A p value < 0.05 (two-sided) was considered statistically significant.

Results

Patient descriptions

The characteristics of 47 patients are described in Table 1, and the pre- and postoperative images of a representative case are shown in Fig. 1. The population comprised 33 men and 14 women, with a mean age of 54.5 years (median 54, range 22–78). The mean tumor diameter was 10.1 cm (median 10.0, range 5.0–20.0) in our series. En bloc and piecemeal resections were performed in 23 (49%) and 24 (51%) patients, respectively. The function of patients which was assessed by Frankel Grade before and after surgery is shown in Table 2. The Frankel Grade of 11 (48%) patients who received en bloc resection descended at least one grade postoperatively, while the function of 6 (25%) patients in the piecemeal group was improved 3 months after surgery. The mean follow-up period was 41.3 months (median 40, range 12–87). Recurrence and disease-related death occurred in 19 and 12 cases, respectively. The mean time from surgery to recurrence was 27.9 months, while mean follow-up for the dead patients was 42.2 months.

Table 1 Patient characteristics and univariate analysis of the prognostic factors affecting RFS and OS
Full size table
Fig. 1
figure 1

Images of a 59-year-old man with large sacral chordoma. a and b Preoperative X-ray. c Preoperative CT. d and e Preoperative sagittal and axial T2-weighted MRI. f Preoperative axial contrast-enhanced T1-weighted MRI. g The en bloc sacrectomy was performed. h Postoperative X-ray

Full size image
Table 2 Frankel Grade of patients before and after surgery
Full size table

Univariate and multivariate analysis of prognostic factors affecting RFS

The Kaplan–Meier analysis showed that the median RFS was 46 months for all 47 patients. The univariate analysis of the prognostic factors for RFS is shown in Table 1. Tumor size (p = 0.008), preoperative ambulatory status (p = 0.037), revision surgery (p = 0.093), resection mode (p = 0.004), postoperative radiotherapy (p = 0.036), and postoperative chemotherapy (p = 0.037) were potential prognostic factors affecting RFS according to the univariate analysis.

The above-mentioned six potential prognostic factors were submitted to multivariate Cox regression model (Table 3). En bloc resection significantly decreased the risk of recurrence (HR = 0.32; 95% CI, 0.10–0.99; p = 0.048). The Kaplan–Meier curves of RFS for resection mode is shown in Fig. 2a. The multivariate analysis also showed that other five factors were not independent prognostic factors for RFS.

Table 3 Multivariate analysis of the prognostic factors affecting RFS and OS
Full size table
Fig. 2
figure 2

a Kaplan–Meier curves of RFS. b and c Kaplan–Meier curves of OS

Full size image

Univariate and multivariate analysis of prognostic factors affecting OS

The Kaplan–Meier analysis showed that the median OS was 67 months with the 5-year survival rate being 52.4%. The univariate analysis of the prognostic factors for OS is shown in Table 1. Patients with a tumor size less than 10 cm had the longer OS time (p = 0.088). OS time was significantly poorer in patients with abnormal preoperative ambulatory status (p = 0.071). Patients who underwent piecemeal resection had a worse OS than those who underwent en bloc resection (p = 0.008). OS time significantly decreased in patients with postoperative recurrence (p = 0.005).

The multivariate analysis of the prognostic factors for OS is shown in Table 3. The risk of death was significantly decreased in patients with en bloc resection (HR = 0.16; 95% CI, 0.03–0.90; p = 0.037). Postoperative recurrence was significantly associated with a higher risk of death (HR = 10.01; 95% CI, 1.00–100.61; p = 0.050). Kaplan–Meier curves of OS for resection mode and postoperative recurrence are shown in Fig. 2(b and c). Multivariate analysis also showed that the tumor size and preoperative ambulatory status were not independent prognostic factors for OS.

Quality of life

The total FACT-G scores and the subscores in the four specific life domains are shown in Table 4. The total score, physical well-being score, and social/family well-being score of the piecemeal resection group were higher than those of the en bloc resection group, but the differences were not significant (p = 0.485, 0.055, and 0.259, respectively). The emotional well-being score of the en bloc resection group was significantly higher than that of the piecemeal resection group (p < 0.001), while the functional well-being score was significantly higher in the piecemeal resection group (p < 0.001).

Table 4 Quality-of-life scores measured with FACT-G
Full size table

Discussion

Chordomas constitute over 50% of primary sacral tumors, and the sacrum is the most common site for chordoma. As an indolent and slow-growing tumor, chordoma is often clinically silent until growing to the large size. [32]. Large tumor burden and sacral nerve root invasion make surgical management of sacral chordomas challenging. In this study, 47 consecutive patients with large sacral chordoma were reviewed. Survival outcomes and quality of life were analyzed and compared between patients receiving en bloc resection and piecemeal resection.

It is reported that the median RFS and OS for sacral chordoma were 44–73 months and 6–7.2 years, respectively [4, 9, 21, 26, 31]. In our series, the median RFS was 46 months, and the median OS was 67 months. A shorter survival time was observed, which may result from larger sizes of tumors included in our series. Literatures focusing on surgical treatment of sacral chordoma with at least 20 patients in the recent 10 years were reviewed, and the main results are listed in Table 5.

Table 5 Literatures focusing on surgical treatment of sacral chordoma with at least 20 patients in the recent 10 years
Full size table

Our results showed that resection mode was an independent prognostic factor affecting both RFS and OS, with en bloc resection significantly decreasing the risk of both recurrence and death. Similarly, studies in recent 10 years also addressed the importance of en bloc resection in disease control for chordoma in the sacrum. Varga et al. reviewed 167 patients with sacral chordoma in 12 spine oncology referral centers and emphasized that en bloc resection improved local recurrence-free survival significantly [31]. Dhawale et al. reviewed 21 patients with sacral chordomas treated with en bloc resection and adjuvant radiotherapy and concluded that despite the complications, increased long-term survival can be achieved with en bloc resection [9]. Ruggieri et al. reviewed 56 patients with sacral chordomas treated with surgical resection and suggested that previous intralesional surgery was associated with a higher rate of local recurrence [25]. Hsieh et al. reviewed 20 patients with sacral chordomas or chondrosarcomas and found that the mean RFS for patients with wide or marginal en bloc tumor excisions was 51 months, but that was only 17.5 months for patients who had contaminated/intralesional resections [15]. Schwab et al. reviewed 42 patients who underwent resection for sacral chordoma and emphasized that intralesional resection should be avoided as it is associated with a higher local recurrence rate and worse survival [26]. Therefore, according to both our results and previous studies, en bloc resection showed its superiority in the aspect of disease control for sacral chordomas.

Clinical management of large sacral chordoma requires a multidisciplinary approach which integrates surgeons, oncologists, radiotherapists and histologists, etc. Although chordoma was thought to be relatively radioresistant, radiotherapy has been widely used as a postoperative adjuvant therapy for treatment of spinal chordomas, especially when wide or marginal margins were difficult to obtain [22]. New advances in radiation technology have allowed for delivery of higher doses of radiation to the target volume with minimal injury to surrounding tissues. However, this technique was not widely used with a long follow-up in our center, and only 5 cases in this series received postoperative radiotherapy. Although chordomas are not sensitive to systemic chemotherapy, it is reported that direct intratumoral chemotherapy with carboplatin and epinephrine obtained a good partial response for a cervical recurrent chordoma [14]. Similar to intratumoral chemotherapy, in our center, intraoperative chemotherapy was widely used under the hypothesis that local accumulation of anticancer agents leads to enhanced efficacy with decreased systemic toxicity. In addition, enlightened by several individual successful examples [18, 28], platinum-based systemic chemotherapy was also tentatively performed in 5 cases by our multidisciplinary team. Unfortunately, according to our results, none of those adjuvant therapies showed significant improvement in the treatment of large sacral chordomas, and their efficacy needs further investigation.

The concept of quality of life is becoming increasingly important in the determination of treatment success [7, 29]. Several studies have pointed out that en bloc sacrectomy often involves transection of the sacral nerve roots and causes functional loss [20, 23]. Similarly, in our series, the function of 48% patients with en bloc resection was deteriorated in terms of Frankel Grade after surgery, while the function of 25% patients in the piecemeal group was improved 3 months postoperatively. However, the patients’ quality of life depends not only on neurologic function but also on physical, social, and emotional conditions. Therefore, we conducted a comprehensive questionnaire survey to determine whether en bloc resection seriously impairs patients’ quality of life in different domains.

In this study, the piecemeal resection group had a better functional status than the en bloc resection group, which were closely related to increases in scores of specific items, such as “I am able to work (include work at home)” and “I am enjoying the things I usually do for fun”. Although the difference did not reach statistical significance in the domain of physical well-being, the piecemeal resection group had a better score in the item of “Because of my physical condition, I have trouble meeting the needs of my family” and “I am bothered by side effects of treatment”. The sacrifice of sacral nerve in the en bloc resection group caused poorer functional (10.83 vs. 13.54) and physical (14.52 vs. 16.04) scores, but the score gaps between these two groups were relatively small. The reasons may be as follows: Firstly, some patients presented with damaged sphincter function or motor disability failed to regain normal nerve function after nerve-sparing piecemeal surgery. Secondly, re-excision of the recurrent tumor would also put the nerve roots at risk. On the other hand, the en bloc resection group had a better emotional status than the piecemeal resection group, which was closely associated with the items, such as “I am losing hope in the fight against my illness”, “I feel nervous”, “I worry about dying”, and “I worry that my condition will get worse”. Apparently, patients in the piecemeal resection group were more nervous and worry about postoperative recurrence, and patients who experienced repeated recurrence lost hope in the fight against their illness gradually. The emotional score gap (8.79 vs. 11.91) was more obvious than that of functional and physical scores. On the whole, the piecemeal resection group had no significant increase in the total FACT-G score than the en bloc resection group in our series.

There are several limitations in our study. Firstly, the retrospective nature is the main limitation. Secondly, due to the limited sample size, we did not design subgroups for en bloc resection based on the resection level in the quality-of-life evaluation. Thirdly, we failed to analyze sexual function, because more than 30% of patients in the follow-up preferred not to discuss relevant issues in detail.

In conclusion, on the one hand, resection mode was an independent prognostic factor for large sacral chordomas, with en bloc resection showing significant advantages in disease control. On the other hand, despite the unsatisfaction in functional well-being, en bloc resection did not sacrifice quality of life significantly in terms of the total FACT-G score. Therefore, we recommend en bloc resection for suitable patients as long as condition allows. All in all, decision making in sacral chordomas is a complex process and influenced by a variety of factors, such as tumor location, neurological status, and psychological condition. We are looking forward to larger and more detailed studies to provide more information concerning this issue.