Surgical treatment of sacral chordomas: monocentric cohort analysis

Full Text

INTRODUCTION

Sacral chordoma is a slow-growing malignant neoplasm characterized by locally invasive growth, a pronounced tendency for recurrence, and the development of distant metastases. The incidence is less than 0.8 cases per 1 million adult population per year [1, 2]. Chordomas occur in the structures of the skull base and the axial skeleton. Chordoma of the skull base was first described in 1846 by Rudolf Virchow [3]. According to the international classification proposed in 2020 by C. Fletcher et al., the fol-lowing tumor types are distinguished based on the degree of malignancy: conventional, poorly differentiated, and dedifferentiated chordoma [4]. It should be noted that the most aggres-sive variant — dedifferentiated chordoma — most often develops from conventional chordoma with increasing duration of carcinogenesis [5].

The tumor pathomorphology is characterized by polymorphic cellular architecture, formation of a dense capsule, and intraosseous spread. The tumor originates from the remnants of the axial skeleton, the notochord [6, 7]. In the overall structure of the disease, sacral localization accounts for 50-60% of cases; less commonly, the tumor occurs in the skull base (up to 30%) and in the vertebrae (up to 15%) [8, 9].

Difficulties in early diagnosis are caused by the slow expansile growth of the tumor. The ini-tial symptoms are nonspecific and most often present as vertebrogenic pain in the lumbar region, as well as a sensation of a mass in the gluteal area [10]. Pelvic organ dysfunction (POD) developing due to tumor compression of the sacral spinal nerve roots is detected on average 12 months after disease onset and represents the first specific symptom [11].

Differential diagnosis of the tumor should be performed with certain primary lesions, including giant cell tumor, chondrosarcoma, myxo-papillary ependymoma, and plasmacytoma, as well as with metastatic and infectious destruc-tion of the sacrum. The gold standard of invasive diagnosis is trephine biopsy followed by histological and immunohistochemical exami-nation of the surgical specimen [5].

The global consensus on the treatment of chordomas, reached in 2017, indicates low tumor sensitivity to radio- and chemotherapy, with prio-rity given to surgical methods, particularly enbloc resection [12]. Despite the improvements in surgical techniques, such as the implementation of stable osteosynthesis principles, use of a multidisciplinary approach for the reconstruc-tion of post-resection defects, prevention of sur-gical site infection (SSI), the rate of postoperative complications reaches 50%, and in case of intralesional resection, recurrence develops within the first 12 months in the majority of patients [13, 14, 15, 16].

The possibilities of pharmacological therapy in the treatment of chordomas are currently extremely limited. Among all antitumor agents, the most well-established efficacy has been demonstrated for imatinib (a protein tyrosine kinase inhibitor) and sorafenib (a protein kinase inhibitor), and their use is justified in palliative treatment to slow disease progression and alleviate symptoms [17, 18, 19].

Thus, the method of choice in the treatment of sacral chordoma, allowing the best outcomes to be achieved, remains the surgical removal of the tumor within healthy tissue margins. It should be noted that, at present, the domestic literature lacks cohort studies analyzing the long-term outcomes of the surgical treatment of sacral chordomas [20, 21], which prompted us to systematize and present our own experience.

The aim of the study — to analyze the short-term and long-term outcomes of surgical treat-ment of sacral chordomas using en-bloc resection.

METHODS

Study design

The study was a consecutive single-center cohort study corresponding to level III evidence. The co-hort included 9 consecutively operated patients treated at the Scientific Clinical Center for Spinal Pathology of the St. Petersburg State Research Institute of Phthisiopulmonology from 2020 to 2024 inclusive.

Patients were included in the study according to the following criteria:

• presence of sacral chordoma at the time of primary surgery confirmed by histological examination, including immunohistochemistry;
• resection volume — en-bloc;
• patient age ≥ 18 years;
• follow-up duration of 12 months or more.

The following parameters were assessed: duration of the diagnostic interval (time from the onset of first symptoms to morphological verification of the disease) and therapeutic interval (time from morphological verification to en-bloc resection); intensity of vertebrogenic pain according to the visual analog scale (VAS) before surgery and at 12 months; characteristics of neurological impairment according to the Frankel scale (types A-E) before surgery and at 12 months; degree of social maladaptation before surgery and at 12 months according to the Oswestry Disability Index (ODI) and Short Form-12 (SF-12, physical component) scales; severity of comorbidities according to the Charlson Comorbidity Index (CCI) [22, 23, 24, 25].

The extent of tumor spread was determined according to the tumor staging system proposed by W.F. Enneking [26]. For topographic and anatomical description of tumor boundaries, the Fourney classification was used: type I — S4-Co; type II — S3-Co; type III — S2-Co; type IV — S1-Co; type V — L5-Co [27]. The radicality of resection was confirmed by the histological examination of the surgical specimen, with an R0 resection margin considered the target outcome.

The clinical characteristics of the patients are presented in Table 1.

 

Table 1

Clinical characteristics of the patients

Age/sex	Tumor type 1	Disease stage 2	CCI 3	Diagnostic interval, mo	Therapeutic interval, mo
68 / m	I	I	6	6	18
43 / m	I	II	5	4	14
51 / f	II	II	3	9	7
41 / m	II	III	3	11	9
67 / m	II	III	2	7	2
57 / f	II	III	4	11	6
63 / m	III	III	4	13	4
72 / f	III	III	3	12	2
48 / f	IV	III	6	8	3

¹ — according to Fourney; ² — according to Enneking; ³ — Charlson comorbidity index.

 

Surgical technique

Histological verification of the tumor was performed by trephine biopsy under fluorosco-pic (C-arm) guidance. During the procedure, a Jamshidi needle was inserted along a transpe-dicular trajectory (for chordomas of types II-IV), while for type I tumors a transcoccygeal approach was used. The tissues along the biopsy needle tract were excised en-bloc during the definitive surgery.

Depending on the type of chordoma, four vari-ants of en-bloc resection according to the Fourney classification were distinguished: variant 1 for type I tumors (2 cases), variant 2 for types II and III (6 cases), variant 3 for type IV (1 case), and variant 4 for type V (0 cases) (Figure 1).

 

Figure 1. Variants of chordoma resection

 

Chordoma resection was performed via an iso-lated posterior approach in 8 patients, involving a linear skin incision followed by sharp dissection of the posterior supporting complex. Hemostasis was achieved using bipolar coagulation. In one case, a combined (anterior-posterior) approach was used to allow the mobilization and ligation of the presacral venous plexus (type IV chordoma). The lateral boundaries of dissection were limited by the sacroiliac joints, and the cranial boundary was set one motion segment above the planned resection level. At the first stage, laminectomy was performed using a high-speed drill and Kerrison rongeurs, followed by the mobilization of the dura mater and radiculolysis of the sacral nerve roots (depending on the chordoma type, proximal intact roots were preserved). After circumferential mobilization of the dura mater below the level of the origin of intact nerves, the dural sac was continuously sutured with non-absorbable 3-0 suture material, followed by the resection of the caudal segments (Figure 2).

 

Figure 2. Intraoperative view after dura mater resection and S1 roots neurolysis

 

At the second stage, the sacrum was mobilized in the caudal direction by transecting the anococcygeal ligament, and the ventral compo-nent of the tumor was dissected extracapsularly. Proximal osteotomy of the sacrum was performed using a Misonix ultrasonic bone scalpel (Figure 3).

 

Figure 3. Intraoperative view after total sacrectomy: a — visualization of S1 nerve roots and the posterior wall of the rectum; b — macroscopic specimen

 

After the removal of the specimen, the opera-tive field was irrigated three times with Lava-sept solution, followed by exposure to an iodine-containing antiseptic solution for 10 minutes. In the presence of visual signs of the instability of the posterior pelvic ring (one clinical case, type IV chordoma), spinopelvic stabilization at the L3-S1-ilium level was performed. The supporting elements were connected using pre-contoured rods in a three-rod configuration, followed by posterior spinal fusion (Figure 4).

 

Figure 4. Final stage of surgery: a — intraoperative view after instrumental fixation; an autologous fibular bone graft has been placed between the iliac crests; b — control X-ray showing correct positioning of the screws

 

For the reconstruction of the post-resection defect of the posterior pelvic wall, a vascularized muscle flap harvested from the m. gluteus maximus was used (in 8 of 9 patients). In one case, a biological mesh implant for hernia repair was used, followed by coverage with a gluteus maximus flap. A drainage tube was placed in the resection area (active suction drainage according to Redon). Wound closure was performed in layers using interrupted sutures. The criterion for drain removal was an output of up to 50 ml per day. Intraoperatively, blood loss (gravimetric method, total loss in sponges and suction) and the duration of surgery were recorded.

Outcome assessment

Follow-up was evaluated at 3, 6, and 12 months after surgery and annually thereafter based on in-person examinations (3 cases) and remote telemedicine consultations (6 cases), including the assessment of follow-up MRI and CT scans of the surgical area, as well as ODI and SF-12 questionnaires. Overall survival and recurrence-free survival were assessed. In cases of complications in the surgical area, the standardized Clavien-Dindo classification was used [28]. The timing of infectious complications was determined according to V. Prinz and P. Vajkoczy (2020) [29].

Statistical analysis

Statistical analysis was performed using SPSS software, version 22.0 (IBM, USA). The normality of distribution of quantitative variab-les was assessed using the Shapiro-Wilk test. The significance of differences in quantitative parameters (VAS, ODI, and SF-12 scores) before surgery and at follow-up time points was evaluated using the Wilcoxon signed-rank test, while differences between quantitative variables were assessed using the Mann-Whitney U test. Results are presented as M±SD, Me (min-max). Qualitative variables are presented as absolute values and percentages. Differences were considered statistically significant at a two-sided significance level of p < 0.05.

RESULTS

The mean age of patients at the time of surgery was 56 years 6 mos ± 11 years 4 mos; Me = 57 years (41-72). The cohort included 5 (55%) men and 4 (45%) women. Follow-up duration was 2 years 4 mos ± 5 mos; Me = 2 years 6 mos (1 year 2 mos - 4 years 10 mos). The preoperative CCI score was 4.0±1.4 points; Me = 4.0 (3.0-5.0). No significant differences were observed between the duration of the diagnostic and therapeutic intervals, which were 7.2±5.6; Me = 6 (3-9) and 9±3; Me = 9 (7-11), respectively (p = 0.250).

In the postoperative period, all patients demonstrated a significant reduction (p = 0.003) in the intensity of vertebrogenic pain: pre-operative VAS scores were 5.4±1.5 points; Me = 6 (3-7), decreasing to 2.8±0.8 points at 12 months; Me = 3 (2-4). A similar positive trend with improvement in the quality of life was observed both in the social maladaptation index, which decreased from 32.5±5.1 points; Me = 35 (23-38) preoperatively to 15.3±2.2 points; Me = 15 (12-19) postoperatively (p = 0.007), and in SF-12 scores: 22.0±1.8; Me = 22.2 (19-25) before surgery and 32.7±2.9; Me = 32.2 (29.2-39.6) after surgery (p = 0.003). These data are presented in Table 2.

 

Table 2

Indicators of pain syndrome, neurological deficit, and quality of life in patients before en-bloc resection and 12 months after

Patient’s number	VAS before/after	POD before/after	ODI before/after	SF-12 before/after
1	3 / 2	- / -	27 / 14	20.0 / 32.7
2	5 / 3	- / -	29 / 12	21.0 / 31.6
3	4 / 2	+ / +	23 / 15	25.0 / 39.6
4	7 / 3	- / -	36 / 16	23.0 / 34.4
5	6 / 4	+ / +	36 / 19	22.2 / 33.1
6	7 / 2	+ / +	38 / 16	19.0 / 30.8
7	4 / 2	+ / +	35 / 13	24.0 / 32.2
8	7 / 3	+ / +	37 / 18	21.6 / 31.2
9	6 / 3	+ / +	32 / 15	22.4 / 29.2

 

Neurological deficits prior to surgery were observed in three cases (33%) and were characterized by lower paraparesis of the Frankel grade D. Pelvic organ dysfunction of varying severity was present in the majority of patients (77%). No deterioration in neurological status was observed postoperatively.

Intraoperative blood loss and duration of surgery are presented in Table 3.

 

Table 3

Indicators of intraoperative blood loss and duration of surgery

Chordoma type according to Fourney	Blood loss, ml	Duration of surgery
I (2; 22%)	215 ± 21 Me = 215	2 hrs ± 14 mins Me = 2 hrs
II (4; 44%)	322 ± 78 Me = 310	2 hrs 45 mins ± 12 mins Me = 2 hrs
III (2; 22%)	640 ± 84 Me = 640	4 hrs 15 mins ± 21 mins Me = 4 hrs 15 mins
IV (1; 11%)	800	5 hrs 10 mins
In total (9; 100%)	422 ± 220 Me = 350	3 hrs 11 mins ± 67 mins Me = 2 hrs 50 mins

 

A clinical example of en-bloc resection of a ty-pe III chordoma in a 72-year-old female patient, along with follow-up imaging data (MRI at 1 month and 4 years 10 months), is presented in Figure 5.

 

Figure 5. Multiplanar MRI reconstruction: a — before surgery; b — 1 month after surgery; c — 4 yrs and 10 mos after surgery, with no signs of local recurrence

 

Complications

In the postoperative period, 4 cases (44%) of Clavien-Dindo grade III b complications were identified, requiring revision surgical interventions. In 2 (50%) patients, revision procedures were performed in the early period (on days 9 and 16), and in 2 (50%) cases in the late period (after 6 and 9 months). Surgical site infection predominated among complications (3 cases; 75%), including one case (33%) associated with an implan-ted spinopelvic fixation system (type IV chordoma, 9 months after the primary surge-ry). Among orthopedic complications, a vertical fracture of the S1 vertebra was identified 6 months after the initial intervention.

The strategy for revision surgical treatment of these patients was based on accepted recommendations in the spinal surgery commu-nity and included surgical debridement of the operative site (3 cases; 75%), in one case supplemented by the use of negative pressure wound therapy (implant-associated late deep SSI, type IV chordoma). Negative pressure wound therapy system replacement in this patient was performed 3 times at 4-day inter-vals. Definitive plastic closure of the gluteal defect was performed after obtaining two consecutive negative bacteriological cultures of wound discharge. For soft tissue reconstruction, a technique of the transposition of two full-thickness V-shaped m. gluteus maximus flaps was used.

In case of a vertical S1 fracture (type III chordoma), the restoration of pelvic ring stabi-lity was achieved by spinopelvic stabilization using a transpedicular system at the L3-ilium level in a three-rod configuration (Figure 6).

 

Figure 6. Multiplanar MRI reconstruction: a — before surgery; b — 3 months after surgery; c — 6 months after surgery, showing a vertical S1 fracture; CT after revision surgery

 

An R0 resection margin was confirmed in all cases based on the histological examination of the surgical specimen. No recurrences (local tumor progression or distant metastases) were observed during the follow-up period. Overall and recurrence-free survival at follow-up was 100%.

DISCUSSION

The surgical treatment of sacral chordomas represents a technically demanding task due to the need for en-bloc resection of the tumor in accordance with oncological principles [30, 31, 32]. In many cases, performing such an intervention without compromising the patient’s quality of life is not possible because of the necessity to resect intracanal neural struc-tures, which carries a high risk of neurologi-cal deficit, predominantly affecting pelvic organ function. Despite these neurological consequences, surgery allows for sustained regression of vertebrogenic pain and reduction in the level of social maladaptation, enabling patients to maintain self-care, including intermittent bladder catheterization. It should be noted that 77% of patients included in the present study had POD even before surgery, and in the postoperative period none of them required cystostomy.

A key aspect in evaluating the effectiveness of chordoma resection is achieving local tumor control, as well as overall and recurrence-free survival. According to the literature, 5-year survival reaches 92%, 10-year survival — 45%, and 15-year survival — 36%. These outcomes fully justify the use of an extensive surgical procedure [33]. Nevertheless, the rate of local tumor recurrence at various time points after surgery reaches 42-67% [34, 35, 36]. The most probable mechanism of local recurrence in the postoperative period is microscopic satellite involvement of the muscles of the posterior trunk and gluteal region, which from a surgical standpoint necessitates sharp dissection of tissues during tumor mobilization [37]. The extent of chordoma also influences the risk of recurrence. Thus, in a cohort study by C. Xie et al., patients with local tumor recurrence were analyzed. It was shown that most recur-rences occur in initially large tumors with axial dimensions exceeding 8 cm. The time to recurrence ranges from 3 months to 13 years, which inherently limits the possibility of comparison with our data, given the maximum follow-up duration of 4 years 10 months [38].

The technical aspects of en-bloc resection are based on a multidisciplinary approach, with particular attention to two key points: (1) dif-ficulty in mobilizing the ventral component of the tumor in patients who have undergone neoadjuvant radiotherapy, which we encounte-red in one case (type C chordoma); (2) the lack of necessity to reconstruct the post-resection defect of the posterior pelvic wall using synthetic materials, including mesh implants, due to the formation of an extensive soft-tissue defect and the need for prolonged wound drainage. In one case, the use of this approach proved to be a significant factor in the development of deep SSI.

Study limitations

The present study is limited by the small sample size and the absence of follow-up exceeding 10 years.

CONCLUSION

The analysis of single-center clinical experience allowed us to determine the feasibility, effec-tiveness, and safety of en-bloc resection for sacral chordomas. This surgical approach ensured 100% resectability, a high level of local control, and improvement in patients’ quality of life in the postoperative period; however, it is associated with the risk of postoperative complications of Clavien-Dindo grade III b exceeding 40%. Further studies are required to validate the obtained results.

DISCLAIMERS

Author contribution

Naumov D.G. — study concept and design, literature search and review, data analysis and interpretation, drafting the manuscript.

Vishnevsky A.A. — data analysis and interpreta-tion, editing the manuscript.

Babich A.I. — data analysis and interpretation, editing the manuscript.

Semenov D. Yu. — data analysis and interpretation.

Yablonsky P.K. — study concept and design, editing the manuscript.

All authors have read and approved the final version of the manuscript of the article. All authors agree to bear responsibility for all aspects of the study to ensure proper consideration and resolution of all possible issues related to the correctness and reliability of any part of the work.

Funding source. This study was not supported by any external sources of funding.

Disclosure competing interests. The authors declare that they have no competing interests.

Ethics approval. Not applicable.

Consent for publication. Written consent was obtained from the patients for publication of relevant medical information and all of accompanying images within the manuscript.

Use of artificial intelligence. No generative artificial intelligence technologies were used in the preparation of this manuscript.

About the authors

Denis G. Naumov

St. Petersburg State Research Institute of Phthisiopulmonology; St. Petersburg State University

Author for correspondence.
Email: dgnaumov1@gmail.com
ORCID iD: 0000-0002-9892-6260
SPIN-code: 1514-5424

Cand. Sci. (Med.)

Russian Federation, St. Petersburg; St. Petersburg

Arkadiy A. Vishnevsky

St. Petersburg State Research Institute of Phthisiopulmonology

Email: vichnevsky@mail.ru
ORCID iD: 0000-0002-9186-6461
SPIN-code: 4918-1046

Dr. Sci. (Med.)

Russian Federation, St. Petersburg

Aleksandr I. Babich

St. Petersburg State Research Institute of Phthisiopulmonology

Email: ai.babich@spbniif.ru
ORCID iD: 0000-0003-2734-0964
SPIN-code: 5754-3035

Dr. Sci. (Med.)

Russian Federation, St. Petersburg

Dmitry Yu. Semenov

St. Petersburg State Research Institute of Phthisiopulmonology

Email: dy.semenov@spbniif.ru
ORCID iD: 0000-0003-2845-1703
SPIN-code: 2839-7241

Dr. Sci. (Med.), Professor

Russian Federation, St. Petersburg

Petr K. Yablonsky

St. Petersburg State Research Institute of Phthisiopulmonology; St. Petersburg State University

Email: piotr_yablonskii@mail.ru
ORCID iD: 0000-0003-4385-9643
SPIN-code: 3433-2624

Dr. Sci. (Med.), Professor

Russian Federation, St. Petersburg; St. Petersburg

References

Supplementary files