Analysis of long-term outcome of image-guided volumetric modulated arc therapy (VMAT) for primary malignant tumor of the cervical spine

ABSTRACT

Objective

Retrospective analysis of the long-term clinical outcome and acute toxicity of the primary malignant tumor of cervical spine receiving CBCT image-guided VMAT.

Methods

Thirty patients with primary malignant tumor of the cervical spine included in our center, from December 2013 to January 2016, 28 patients were retrospectively studied. The prescription dosage 95% PTV volume dose was 44 Gy, 2.0 Gy/fraction, and a total of 22 times. The median PGTV synchronized volume dose was 60 Gy (45–62.1 Gy), median 2.5 Gy (2–2.7 Gy)/fraction. In volumetric modulated, two arc volumetric modulated arc therapy (VMAT) was used, with spinal cord dosage D_MAX< 45 Gy. Early response rate and acute toxicities were analyzed.

Result

The follow-up duration was 6–76 months (median 53 months). At the end of follow-up of June 1, 2019, 78.6% (22/28) patients were still alive. 3 and 5-y local control rates were 67.3% and 56.5% while 3 and 5-y OS were both 78.6% in the whole group of patients, respectively. Fourteen patients with chordoma 5-y local control rates and OS were 57.1% and 85.7%, respectively. Nine patients with giant-cell tumor of bone had a 5-y local control rate and OS were 77.8% and 85.7%, respectively. The response rate for moderate pain or above was 80% (8/10). Eleven patients (39.3%) suffered from grade 1 acute skin toxicity. Twenty-four patients (85.7%) had grade 1/2 mucositis. No radiation-induced spinal cord injury was found.

Conclusion

The image-guided VMAT for primary malignant tumor of the cervical spine provided a satisfactory long-term local control rate.

KEYWORDS:

• Cervical spine
• chordoma
• giant-cell tumor of bone
• VMAT
• EBRT

Introduction

The incidence rate of primary malignant bone tumors is low, accounting for only 0.2% of all malignant tumors. There are many pathological types of primary malignant bone tumors, of which osteosarcoma (about 35%), chondrosarcoma (30%) and Ewing’s sarcoma (16%) are the three most commonly seen ones. Some of them can be cured by proper treatment. Undifferentiated pleomorphic sarcoma (UPS), fibrosarcoma, chordoma and giant cell tumor of bone account for 1–5%¹ of all primary malignant bone tumors. Chordoma is more common in men, with a high incidence from 50 to 60 y old.² In recent years, neoadjuvant or adjuvant multidrug combined with chemotherapy has significantly improved the prognosis of osteosarcoma and Ewing’s sarcoma.^3,4 There are different comprehensive treatment methods according to different pathological types. The development of radiotherapy technology and equipment makes it possible for malignant primary bone tumors such as chondrosarcoma, osteosarcoma, chordoma and giant cell tumor of bone to receive radical dose. The National Comprehensive Cancer Network (NCCN) Guidelines recommend that postoperative residual or radical doses for malignant tumors such as chondrosarcoma and chordoma be ≥70 Gy (routine), showing that the improvement of the efficacy of treating primary bone tumors depends on the increase in radiotherapy dose. The tumor dose in the spine is subject to the strict dose-limiting of the spinal cord by its unique anatomical characteristics, which is extremely demanding for radiotherapy equipment and technology and requires special equipment and technology. It takes time to popularize protons, therefore, the development of intensity modulated radiation therapy (IMRT), the dose distribution carving the tumor away from normal tissues in the area, conformation, the high accuracy of the positioning, the high repeatability of the posture and the precision of the target area, are the key steps to achieve higher dose or hypofraction in current IMRT.

The precise treatment of image-guided IMRT has generated clinical benefits. At present, there is no research or report on precise radiation therapy for primary malignant tumor of the cervical spine. This paper has used image guidance combined with VMAT technology to have preliminary study on the primary tumor of the cervical spine, so as to lay a theoretical and clinical foundation for improving the curative effect of radiotherapy for tumors of the cervical spine and even the spine.

Material and method

Clinical data

Thirty patients with primary malignant tumor of the cervical spine included in our center, from December 2013 to January 2016, were retrospectively studied. Of these, 2 cases were lost to follow-up and 28 cases were included in the retrospective analysis. There were 13 male (46.4%), 15 female (53.6%), aged 12–71 (median 42 years old). 14 cases (50.0%) were chordoma, 9 cases (32.1%) were giant cell tumor of bone, 4 cases (14.3%) were sarcoma, and 1 case (3.6%) was schwannoma. Fourteen cases (50%) received radiotherapy alone and 14 cases (50%) were treated after postoperative recurrence or postoperative residual. The follow-up lasted 6–76 months (median 53 months). No patient received local radiotherapy previously (Tables 1 and 2). All patients signed informed consents prior to treatment.

Table 1. General information of patients (n = 28).

	Quantity	Percent (%)
Median age	42 (12～71)
Gender
Male	13	46.4
Female	15	53.6
KPS score
70	6	21.4
80	13	46.4
90	8	28.6
100	1	3.6
Pathological type of tumors
Chordoma	14	50.0
Giant cell tumor of bone	9	32.1
Sarcoma	4	14.3
Schwannoma	1	3.6
Prior treatment
Primary	14	50
Postoperative residual	5	17.9
Postoperative recurrence	9	32.1
Median follow-up time (month)	53 (6～76)

Table 2. General information of patients (n = 28).

No.	Gender	Age	Pathological type	Position	PGTVRT dose	NRS	Frankel Grade
1	M	12	Ewing sarcoma	C2	RT45 Gy/20 f	2-3	E
2	M	39	Hemangiosarcoma	C1-4	S+ RT 66 Gy/33 f	2	E
3	F	21	Hemangiosarcoma	C2-3	RT50 Gy/20 f	4-5	E
4	F	71	Schwannomaaganglioma	C5-6	RT60 Gy/30 f	4-6	E
5	M	45	Chordoma	C2	RT62.1 Gy/27 f	3-4	E
6	M	47	Chordoma	C3-5	S+ RT62.1 Gy/27 f	2	D
7	M	31	Chordoma	C1-3	RT60 Gy/22 f	4	E
8	M	23	Giant cell tumor of bone	C2-3	S+ RT60 Gy/22 f	2	E
9	F	24	Giant cell tumor of bone	C4-T1	RT60 Gy/22 f	5-6	E
10	F	38	Giant cell tumor of bone	C1-6	RT60 Gy/22 f	3-4	E
11	F	42	Giant cell tumor of bone	C2-7	RT50 Gy/20 f	0	E
12	F	56	Chordoma	C2	RT62.1 Gy/27 f	3	E
13	M	60	Chordoma	C4	RT60 Gy/22 f	1	E
14	M	50	Chordoma	C2	RT62.1 Gy/27 f	2	E
15	M	44	Chordoma	C2	S+ RT66 Gy/33 f	0	E
16	F	35	Chordoma	C1-4	S+ RT60 Gy/25 f	0	E
17	F	32	Chordoma	C2-6	S+ RT62.1 Gy/27 f	5-7	E
18	F	52	Chordoma	C2	RT62.1 Gy/27 f	3-4	E
19	F	44	Chordoma	C1-2	RT60 Gy/22 f	5	E
20	F	50	Chordoma	C2	S+ RT60 Gy/22 f	2	E
21	F	14	Chordoma	C4-7	S+ RT60 Gy/22 f	2	E
22	F	39	Chordoma	C2	RT62.1 Gy/27 f	5-6	E
23	M	65	Osteosarcoma	C5-7	S+ RT60 Gy/25 f	2	E
24	M	29	Giant cell tumor of bone	C2-5	S+ RT60 Gy/22 f	1	E
25	F	25	Giant cell tumor of bone	C2	S+ RT60 Gy/22 f	0	E
26	F	58	Giant cell tumor of bone	C3-T1	S + S+ RT60 Gy/22 f	3	E
27	M	23	Giant cell tumor of bone	C5-7	S+ RT55 Gy/22 f	0	E
28	M	42	Giant cell tumor of bone	C1-2	S+ RT55 Gy/22 f	0	E

S = surgery, RT = radiotherapy, M = male, F = female, C = cervical vertebra.

Positioning method

All patients were in supine position, head on C/B pillow, fixed by five-fixed-sites, neck and shoulder low-temperature thermoplastic membrane composite plate . Large aperture CT machine (Philips, Netherlands, Brilliance. 1MBigBore CT) was applied to carry out simulation positioning enhanced scan. The scanning range was from the supraorbital margin to the thoracic spine 4 level, including the whole cervical spine and accessories (layer thickness 3 mm, 120 KV, 200 mAs). The scanned image was transmitted to our Eclipse (Varian Medical Systems, Inc.) treatment planning system for outlining target area.

Target area outline and plan design

Among all sequence images of MRI scan of PACS system patients in our hospital, T1 weighted, T1 enhanced and T2 weighted were selected. Sequence Water: AX1T1 C+, aXT2FRFSE was transmitted to Varian planning system, in the same patient’s CT positioning image ID number. The positioning CT plain scan MRI aXT2FRFSE sequence image was fused under the image fusion software of Varian Eclipse planning system to outline the target area. The physician from the radiation oncology department followed the rules of ICRU: GTV was defined as the range of malignant lesions visible in image with a certain shape and size. CTV-included GTV and appropriate external expansion plus lesion involving vertebral body plus accessories. PTV included CTV with 0.5 cm external expansion. PGTV-included GTV with 0.5 cm external expansion. The spinal cord outlining range included CTV with 6 mm up and down. The image after outlining the target area was transmitted through the MOSAIQ network to Oncentra (Nucletron, External Beam v4.3) treatment planning system for outlining target area and designing treatment plan. The prescription dosage 95% PTV volume dose was 44 Gy, 2.0 Gy/fraction, and a total of 22 times. The median PGTV synchronized volume dose was 60 Gy (45–62.1 Gy), median 2.5 Gy (2–2.7 Gy)/fraction, and a median total of 22 (20–33) times. In volumetric modulated, 2 arc VMAT was used, simultaneous integrated PGTV dose boost and a total of 90 dose control points, which was reverse dose optimization, to meet the requirements of the prescription. Moreover, the spinal cord dosage was D_MAX< 45 Gy.

IGRT with HexaPOD evo RT 6D couch radiation therapy

After examination and approval by a senior physicist and a senior physician from the radiation oncology department, the patient started treatment. The routine radiotherapy was implemented by the Elekta Axesse Accelerator in the hospital. Airborne CBCT position verification was carried out on a daily basis to adjust the six-dimensional set-up errors of HexaPOD evo RT6D couch.

Follow-up and statistics analysis

Clinical and imaging assessments were conducted 4 weeks after the radiation therapy and evaluated every 2 months for the first year and then every 6 months thereafter. Local control time was calculated from the beginning of the treatment. According to WHO According to WHO Response Evaluation Criteria in Solid Tumors, the maximum diameter of tumor in vertical direction was measured on follow-up CT images. Complete response (CR) was defined as a complete disappearance of the tumor for over 4 weeks; partial response (PR) as tumor shrinkage ≥50% for more than 4 weeks; stable disease (SD) is between PR and PD: tumor enlargement <25%, shrinkage <50%; progress of disease (PD) refers to new tumor or lesion increase ≥25%.

The Numeric Rating Scale (NRS) was selected for the pain score: 0 point: no pain; 1–3 point: mild pain; 4–6 point: moderate pain, influencing one’s sleep; 7–10 point: severe pain, unbearable. Spinal cord injury was assessed by the Frankel Grade. Lower limbs function was evaluated according to Frankel Grade of Spinal Cord Injury Classification.

The post radiotherapy acute reaction was assessed by scored by the Common Terminology Criteria for Adverse Events V4.0. Kaplan-Meier method was applied to calculate local control rate and tumor-specific survival. SPSS 16.0 software was used to analyze data.

Results

Analysis of survival and local control

The follow-up duration was 6–76 months (median 53 months). At the end of follow-up of June 1, 2019, 78.6% (22/28) patients were still alive, and 21.4% (6/28) were dead. Two patients (1 of chordoma, 1 of giant cell tumor of bone) died of local progression. Of these, three patients with chondrosarcoma, hemangiosarcoma and schwannoma, respectively, died of systemic metastasis and one with chordoma died of heart disease. Three- and 5-y local control rates were 67.3% and 56.5% while 3 and 5-y OS were both 78.6% in the whole group of patients, respectively. Fourteen patients with chordoma had a local evaluation of tumor as SD after EBRT. Three- and 5-y local control rates were both 57.1%, respectively. The median local control time was 44 months (6–64 months) and 3- and 5-y OS rate was 85.7%. Nine patients with giant-cell tumor of bone had a 3- and 5-y local control rate of 77.8%, and the median local control time was 60 months (11–69 months) with 3- and 5-y OS rate of 85.7%. As for four patients with sarcoma, two of them relapsed, only one of them still alive without relapse at the end of follow-up. One case with schwannoma got local recurrence 16 months after the radiation therapy.

2. Pain assessment

The whole group of patients was accompanied by different degrees of pain before the radiation therapy, with NRS scores ranged from 0 to 9 point. The median pain response time was 2 (1–6)months after RT. The partial response rate for moderate or above pain was 80% (8/10), of which three patients (30%) had complete response, five patients (50%) had partial response, and two patients (20%) had no response. (Table 3).

Table 3. NRS score of 22 patients before and after RT.

NRS score before RT	Number of cases	NRS score at the last follow-up
		CR	Improved	Not-improved	Aggravated
Mild (1–3 point)	12	8	4	0	0
Moderate (4–6 point)	9	3	4	2	0
Severe (7–9 point)	1	0	1	0	0

3. Frankel grade of spinal cord injury classification

In 28 patients, only 1 case was graded as D before RT, who got pain relief postoperative and after RT, and the improvement of limbs allowed the return to daily activities.

4. Acute complications

One patient (3.6%) had grade 4 oropharyngeal mucosa reaction (delayed healing in postoperative retropharyngeal) after RT. Eleven patients (39.3%) suffered from grade 1 acute skin toxicity. Twenty-four patients (85.7%) had grade 1/2 mucositis. 5 patients (17.9%) showed grade1/2 myelosuppression. No radiation-induced spinal cord injury was found. (Table 4)

Table 4. Grading of RT acute complications for 28 patients (CTCAE V4.0).

	Grade 1	Grade 2	Grade 3	Grade 4	Grade 5
Myelosuppression	3	2	0	0	0
Radiodermatitis	11	0	0	0	0
Nausea	6	4	0	0	0
Anorexia	6	4	0	0	0
Dysphagia	15	4	0	0	0
Oropharyngeal mucosa reaction	15	9	0	1	0

Discussion

At present, due to the intervention of multidisciplinary diagnosis and treatment, nearly 75% of osteosarcoma can be cured and 90–95% patients can preserve organ function.⁵ Even if there is distant metastasis, patients still have the possibility of being cured.^6,7 According to the cancer statistics database of SEER (Surveillance, Epidemiology, and End Results) in the United States, the overall 5-y survival rate of primary malignant bone tumors reached 66.6%. Most of the giant cell tumors of bone are benign, but because of its growth invasiveness and the strong destructiveness to the vertebral body, generally, it is considered as malignant for the treatment. In this study, the treatment and follow-up of patients with giant cell tumor of bone were included.

In 2014, Wilartratsami et al.⁸ reported the epidemiological characteristics of primary spine tumors from 1996 to 2010. Common malignant tumors were chordoma, chondrosarcoma, and osteosarcoma, while the most commonly-seen benign tumors were giant cell tumor of bone and hemangioma. This was similar to the report of MSKCC. The average onset age of primary spine tumors was 44.68, with women accounting for 53%. About 74% of malignant tumors and 52.953% of benign tumors were accompanied by pain. The prognostic study of primary malignant spine tumor was confined to a small sample and control experiment in one single research center, and there was a lack of large and comprehensive clinical data report. In recent years, along with the wider application of imaging examination technology, especially MRI, the detection rate of spine tumors has increased.⁹ Total spondylectomy is effective in the treatment of malignant spine tumors, but when diagnosed, many patients had epidural invasion, large paravertebral tumor or concomitant bone-related diseases, which made them not suitable for total spondylectomy. In modern times, the treatment of primary malignant bone tumor emphasize on multi-disciplinary and multi-measures therapy including surgery, radiotherapy, chemotherapy, or multimodal combination.

Radiotherapy has always been an adjuvant treatment or initial treatment for malignant bone tumors, but due to the limitation of the maximum tolerance dose of the spinal cord, the curative effect of traditional radiotherapy technology is barely satisfactory. The modern radiotherapy technology has taken a series of measures to increase the dose of spine tumors while reducing spinal cord toxicity and protecting the esophagus, oropharynx, lung, intestinal tract, and other peripheral organs at risk.

The cervical spine has curvature segments and obvious physiological kyphosis and lordosis are present in all spinal components. Unlike other spinal parts, the primary tumor, especially the upper cervical spine (C1, C2), is close to the medulla, which can compress the medulla or cause local instability of the cervical spine and lead to high cervical spinal cord damage to endanger the patient’s life. In this case, there is a surgical risk and many complications in perioperative period, difficult to have thorough surgical resection, and easy to relapse after the operation. Furthermore, it is difficult to rebuild stability. Therefore, adjuvant therapy, such as radiotherapy, chemotherapy or immunotherapy, is a necessary means for the treatment of tumors of the cervical spine. IMRT, protons, SRS or FSRT are recommended by clinical guidelines for its ability to enable high dose so as to provide maximum protection of peripheral normal organs.^10,11 Chordoma is relatively common in primary tumors of the cervical spine.¹² Chordoma occurs in the residual spinal cord tissue during the embryonic period, high occurrence in the elderly, mainly involved in the axial skeleton, of which the spine (15%), sacrum (50–60%), and basion (25–35%) are the three common primary sites. Cervical chordoma can produce local tumor compression and induce airway obstruction, dysphagia or other symptoms. The preferred treatment is radical surgery with tumor-free principle. A study of 400 patients with chordoma in the SEER database showed that the median survival was 6.29 y, with 5-, 10- and 20-y survival rates of 67.6%, 39.9%, and 13.1%, respectively.² A number of studies^12-14 reported that surgical margin is an important prognostic factor for the survival of spinal chordoma without recurrence, and the study of Boriani¹² found that only patients with total vertebrectomy with tumor-free incisal edge had 5-y survival, while all follow-up patients undergoing resection within the tumor foci had recurrence within 2 y. The previous view was that chordoma was insensitive to radiotherapy. However, with the continuous improvement of radiotherapy equipment and technology in recent years, studies on surgery combined with radiotherapy for the treatment of spinal chordoma^15-17 showed a result that it can improve the local control of tumor as well as disease-free survival. A retrospective study in 2014¹⁵ included 50 patients with primary tumors of the spine (29 cases of chordoma, 14 cases of chondrosarcoma, and 7 others). Five- and 8-y local control rate for patients receiving high doses of proton/photon radiotherapy for primary tumor were 94% and 85%, respectively. In the 8th year of follow-up, there was 13% of grade 3–4 radiotherapy adverse reaction. A study of 126 patients with spinal/sacral chordoma reported¹⁷ that 5-y local control and survival rates were 81% and 62%, respectively, after receiving high doses of proton radiotherapy. The domestic research on the radiotherapy curative effect of primary tumor of the spine is rare. This study included 28 cases of primary tumor of the cervical spine. Innovatively (photon, non-proton) intensity modulated local lesion boost technology was applied. Three- and 5-y local control rate were 67.3% and 56.5%, respectively, for all patients. The tumor local evaluation in 14 cases of chordoma after radiotherapy was SD and their 3- and 5-y local control rates were both 57.1%, with median 44 months. Three- and 5-y OS were both 85.7%, and no severe radiation-induced myelopathy was found after 5-y observation, with early toxic and adverse effects acceptable. Our initial results indicated that local control rate remained to be improved, and the survival rate was similar to that reported in foreign literature. However, it was needed to extend the follow-up time to get data for more than 5 y.

The incidence rate of giant cell tumor of bone took up 3–5% of that of all primary tumor of bone. The age of high occurrence was 20–40 and its characteristics were prone to recurrence and high pulmonary metastasis tendency.^18,19 Extensive surgical resection and lesion scraping are two main methods of operation. The recurrence rate of the two methods had a significant difference: the recurrence rate of the former was 0–12%, and the latter 12–65%.^20–24 Function loss and high postoperative complication of extensive resection limited the choice of this method.^25,26 For inoperable patients to receive radiotherapy, it can be the preferred treatment method. For patients who had incomplete surgery or positive incisal edge, surgery can be assisted by receiving radiotherapy, to improve the local control rate and disease-free survival for patients with giant cell tumor of bone.^27–29 In a recent retrospective clinical study,²⁸ the efficacy of radiotherapy in 58 cases of osteosarcoma (45 cases of primary tumors and 13 cases of recurrent tumors) was analyzed, with a median follow-up time of 8 y. The 5-y local control rate and survival rates were 85% and 94%, respectively. The multi-factor analysis found out that age was the only prognostic factor affecting the radiotherapy local control rate and survival rate of giant cell tumor of bone (local control rate of the young 96%, the elderly 73%; survival rate of the young 100%, the elderly 65%). Other studies have also reported that tumor diameter >4 cm, tumor recurrence, and radiotherapy dose <40 Gy were poor prognostic factors for tumor local control rate.^30,31 Ma Qingjun et al.³² of Peking University 3rd Hospital in China reported 36 cases of spinal giant cell tumor of bone, of which three patients were treated with single radiotherapy and seven patients were treated with surgery combined with radiotherapy. The follow-up confirmed that the treatment efficacy in these 10 patients was good. In this study, nine cases of giant cell tumor of bone had a median follow-up for 60 months (11–69 months), with 5-y local control rate of 77.8%, 5-y OS of 85.7%, with satisfactory curative effect. Among these nine patients (three cases of primary, six cases combined with surgery), no case of severe radiotherapy acute complication was found.

Radiotherapy is also one of the important adjuvant methods of surgical treatment for malignant tumors such as primary osteosarcoma of the spine or chondrosarcoma derived from mesenchymal tissues. A retrospective analysis of 60 patients with chondrosarcoma indicated that surgery combined with radiotherapy could benefit the local control rate of tumors unable to extended resection.³³ Recent studies have concluded that postoperative radiotherapy can reduce local recurrence for chondrosarcoma derived from mesenchymal.³⁴ Radiation therapy plays an increasingly significant role in the treatment of malignant tumors of bone.

This study reviewed and analyzed the short-term clinical efficacy for 28 cases of primary tumor of the cervical spine with the application of MRI positioning fusing with CT accurate target area outlining. It utilized new VMAT, applied with IGRT, under the precise radiation therapy after HexaPOD evo RT6D couch six-dimensional set-up errors correction. It can also treat complications which is acceptable.

Conclusion

The image-guided VMAT for primary malignant tumor of the cervical spine provided a satisfactory long-term local control rate. Organs can be preserved and pain can be effectively alleviated. Injury in neurological function and motor function can be improved with low incidence of complications. Further follow-up studies are required.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.