Interventional procedure overview of radiofrequency ablation as an adjunct to balloon kyphoplasty or percutaneous vertebroplasty for palliation of painful spinal metastases

Systematic review and metanalysis

UK

Databases searched (Ovid MEDLINE, Embase, CENTRAL) until July 2020. Reference lists of included studies for additional records was also done. Trial registries were also searched.

n= 15 studies (non-randomised: 5 prospective and 10 retrospective studies)

2 of these were comparative studies ( RFA + cement augmentation versus RFA + radiotherapy; RFA alone versus RFA + cement augmentation)

with 725 adult patients with spinal metastases.

Mean age across studies ranged from 59 to 69.6 years.

50% (364/361) female

Inclusion criteria: both randomised and non-randomised comparator study designs with patients aged over 18-years-old; presenting with spinal metastases and have undergone treatment with RFA alone or RFA combined with another modality; reporting pain, disability, HRQoL, complications, tumour control and mortality.

Exclusion criteria: Studies that only included data for primary spinal tumours, animals and RFA assisted open surgery were excluded.

Patients were treated with radiofrequency ablation (RFA) and majority had an additional vertebroplasty treatment.

Variety of ablation systems were used in the studies; most commonly used was the STAR® Tumor Ablation System (temperature used for ablation 50 degrees), other system used were the CAVITY SpineWand (cold energy temperature 42 degrees), OsteoCool ablation device, RFA-I type multipolar cancer ablation system, Radionics system, Cool-Tip RF ablation system, and CelonPower system.

mean follow-up periods varied across studies (range of 24–48 hours [1 study], 2-4 weeks [2 studies] to 60 months)

None

8 studies (n=286)

3-5 weeks

2.24

1.55–2.93

89%

4 studies (n=98)

3-4 months

3.00

1.11–4.90

95%

4 studies (n=144)

5-6 months

3.54

1.96–5.11

88%

Bagla 2016 (n=50)

FACT-G7 (0-28)

Mean score 11

15.8 (p<0.0001)

16.2 (p<0.0001)

Bagla 2016 (n=50)

FACT-BP (0-60)

Mean score 22.6

37.3 (p<0.0001)

38.9 (p<0.0001)

Bagla 2016 (n=50)

Modified Oswestry Disability Index (0-100)

Mean score 52.9%

40% (p<0.01)

37% (p<0.01)

Sayed

FACT-G7

NS

–

–

Baseline

3-6 months

15-36 months

Proschek 2009 (n=16)

Oswestry Disability Questionnaire score

RFA group 64% (range 38-84%)

RFA plus vertebroplasty group

66% points (range 39-86%)

33%, range 23-38%; p=0.06

(RFA group)

35%, range 26-38%; p=0.071 (RFA plus vertebroplasty group)

Gervagez

Pain disability index score

% decrease at 6 weeks (p<0.015)

4% decrease (p=0.002)

10% decrease over 6 months (p=0.003)

Zhao

EORTC QLQ-C30 scale

Physical function (p=0.03)

Emotional function (p=0.003)

% (n)

Cement extravasation after vertebroplasty

10.3% (72 occurrences)

In 1 patient it caused pain and needed surgical removal.

All-cause mortality* (in 10 studies) at mean 1 year follow up

23.6% (109/462)

Postoperative sepsis resulting in death (RFA done despite having a subclinical paravertebral abscess which was misdiagnosed)

n=1

Deaths unrelated to the procedure (in 2 studies)

n=13

Systematic review

Austria

Databases searched (Ovid MEDLINE, Embase, the Cochrane Library, CRD and PubMed); until December 2016. Manual search for additional records done.

n= 9 studies (4 prospective and 5 retrospective studies)

with 583 patients with painful vertebral metastases.

Mean age across studies ranged from 61 to 69.6 years.

Sex not reported.

Inclusion criteria: patients with solitary fracture-related vertebral metastases unresponsive to previous curative or symptomatic treatments; RFA with or without vertebroplasty or other add-on therapies (for example, radiation); randomised controlled trials, non-randomised controlled trials, prospective and retrospective case series (with more than 30 patients); published in English or German.

Exclusion criteria: multiple publications including subgroup analysis published by the same investigator in a previous article were excluded from the final analysis.

Patients were treated with radiofrequency ablation (RFA) and 72% (437/583) had an additional vertebroplasty treatment. Vertebroplasty was done if there was a risk of fracture and instability of the bone structure due to tumour removal.

Variety of ablation systems were used in the studies; most commonly used were the STAR® Tumor Ablation System (temperature used for ablation 50degrees) and the CAVITY SpineWand (cold energy temperature 42 degrees)

mean follow-up periods varied across studies (range of 24–48 hours [1 study], 2-4 weeks [2 studies] to 60 months)

None

Bagla 2016 (n=50)

NPRS (0-100)

Mean score 5.9

2.6 (p<0.0001)

(n=40)

2.1 (p<0.0001)

(n=33)

Nakatsuka 2009

(n=10)

VAS (0-10)

Mean score 7.5 ± 2.7

1 week

2.7 ± 2

(P=.00005)*

Proschek 2009 (n=16)

VAS (0-10)

RFA group

Mean score 7.6

RFA+ cement

Mean score 7.9

After treatment

RFA group [n=8]

5.5 (p=0.018) RFA+ cement (n=8)

5.0

15-36 months

RFA group [n=8]

4.0 (p<0.008)

RFA+ cement (n=8)

3.5 (p<0.005)

Georgy 2007 (n=15)

VAS (0–10-point scale)

Pain scores range 6 to 10

Range 0-5

Pain relief in 87% (13/15) at 2-4 weeks

Bagla 2016 (n=50)

FACT-BP (0-60)

Mean score 22.6

37.3 (p<0.0001)

38.9 (p<0.0001)

Bagla 2016 (n=50)

Modified Oswestry Disability Index (0-100)

Mean score 52.9%

40% (p<0.01)

37% (p<0.01)

–

–

Baseline

3-6 months

15-36 months

Proschek 2009 (n=16)

Oswestry Disability Questionnaire score

RFA group 64% (range 38-84%)

RFA plus vertebroplasty group

66% points (range 39-86%)

33%, range 23-38%; p=0.06

(RFA group)

35%, range 26-38%; p=0.071 (RFA plus vertebroplasty group)

Major complications (procedure related)

0

Overall complications

30.2 (78/583)

Ranged from 4.3 to 40%

Adverse events (procedure-related or non-procedure-related)

18 (105/583)

ranged from 5.6% to 11.1% in each study.

Procedure-related adverse events

2.74 (16/583) range 0 to 11%

Non-procedure-related adverse events

15.2 (89/583) range 4.3 to 73%

Increased pain and numbness

1 (6/583)

Post-procedure radicular symptoms and pain

0.8 (5/583)

Rate of adverse events not RFA-related (but vertebroplasty-related)

ranged from 4.3% to 73.0%

Cement extravasation after vertebroplasty

15.3% (67/437)

Mortality (reported in 2 studies)

5 to 10 deaths

Systematic review

France

Databases searched (Ovid MEDLINE, Embase, the Cochrane CENTRAL and PubMed); until March 2017. conference abstracts were also searched.

n= 8 studies (4 prospective and 4 retrospective studies)

with 261 patients with painful spinal metastases (340 vertebral lesions)

Lesions were located mainly in lumbar or thoracic area. Metastases origin from breast, lung, and renal cancers. previous treatments (n=4) chemotherapy, radiation therapy, surgery, or a combination.

Mean age across studies ranged from 59 to 69.6 years.

Sex not reported.

Inclusion criteria: randomised controlled or non-randomised studies with a prospective or retrospective design; adults with spinal metastasis; treated with RFA alone or in combination/ comparison with other treatments; studies reporting patients' pain before and after RFA; and English-language studies.

Exclusion criteria: multiple publications were excluded.

Patients were treated with radiofrequency ablation (RFA) and 239 had cement augmentation (during RFA in 40-60%, or in 95.8% of treated vertebrae in the same treatment session). 4 patients underwent cement augmentation after a few days or months.

Variety of ablation systems were used in the studies (STAR Tumor Ablation System, the CAVITY SpineWand, RITA, CELON Power System, Cool-Tip RF)

Lumbar ones were the most commonly treated. Mean ablation time ranged between 6 and 9.75 min. across 4 studies. Ablation zone ranged from 1-8 cm. Bipolar RFA was applied in 4 out of the 8 studies.

mean follow-up periods varied across studies (ranged from 2-4 weeks [2 studies] to mean 20 months)

Authors are proctors for Medtronic and Galil Medical.

Bagla 2016 (n=50)

RFA + cement augmentation

NRS (0-100)

5.9

3.7 (discharge)

2.2 (<0.0001)

37

-

Bagla 2016 (n=50)

RFA + cement augmentation

NRS (0-100)

5.9

2.6 (1 month, n=40)

3.3 (<0.0001)

56

-

Bagla 2016 (n=50)

RFA + cement augmentation

NRS (0-100)

5.9

2.1 (3 months, n=34)

3.8 (<0.0001)

64

-

Gronemeyer 2002 (n=10)

RFA+ cement augmentation

VAS (0-10)

5.9

2.6 (mean 5.8 months)

3.3

56

-

Nakatsuka 2009 (n=10)

RFA (all patients)

VAS

7.5

2.7 (1 week)

4.8 (0.00005)

64

-

Nakatsuka 2009 (n=10)

RFA alone (n=4)

4.3

1.7 (1 week)

2.6 (0.0004)

60

Nakatsuka 2009 (n=10)

RFA + cement augmentation (n=6)

6.6

1.7 (1 week)

4.9 (0.003)

74

-

Proschek 2009 (n=16)

RFA alone (n=8)

VAS

7.9

4 (15-36 months)

3.9 (0.008)

49

-

Proschek 2009 (n=16)

RFA +cement augmentation (n=8)

7.6

3.5 (15-36 months)

4 (0.005)

52

-

Greenwood 2015 (n=21)

RFA +cement augmentation

NRS

8

2.9 (4 weeks)

5.1 (<0.0003)

63

-

Anchala 2004 (n=92)

RFA+ vertebral augmentation

VAS

7.5

2.2 (1 month, n=83)

5.26 (<0.0001)

70

-

Anchala 2004 (n=92)

RFA+ vertebral augmentation

VAS

7.5

1.7 (6 months, n=9)

5.7 (0.009)

76

-

Georgy 2009 (n=37)

RFA+ vertebral augmentation

VAS

8

4 (2-4 weeks)

4

50

89.5

Yang 2017 (n=25)

RFA+ cement augmentation

VAS

100

Mean 7.8 months

Anchala 2004

76.9% (70/92)

at average 92 days after RFA

23.1% (22/92) at average 82 days

Greenwood 2015

92.3 (12/13) at 3 months

100% (at 6 months)

–

Yang 2017

66.67% (at 2 years)

–

Bagla 2016 (n=50)

FACT-G7 (0-28)

Mean score 11

15.8 (p<0.0001)

16.2 (p<0.0001)

Bagla 2016 (n=50)

FACT-BP (0-60)

Mean score 22.6

37.3 (p<0.0001)

38.9 (p<0.0001)

Bagla 2016 (n=50)

Modified Oswestry Disability Index (0-100)

Mean score 52.9%

40% (p<0.01)

37% (p<0.01)

Proschek 2009 (n=16)

Oswestry Disability Questionnaire score

Baseline

3-6 months

15-36 months

Proschek 2009 (n=16)

Oswestry Disability Questionnaire score

RFA group 64% (range 38-84%)

RFA plus vertebroplasty group

66% points (range 39-86%)

33%, range 23-38%; p=0.06

(RFA group)

35%, range 26-38%; p=0.071 (in both RFA alone and RFA plus vertebroplasty groups)

Grade IV-V complications

0

Transient neural damage (grade II) related to high temperature rise during RFA (resolved 2 days after the procedure with intravenous administration of steroids)

n=1 (Nakatsuka 2009)

Contralateral lower limb pain and numbness during RFA -grade I (2 spontaneously resolved with temperature decrease and 2 needed steroids, but 1 developed heaviness in legs 1 day after treatment which resolved at 1 week [grade II-IIIa])

16% (4/25)

(Yang 2017)

Bone cement leakage outside the vertebral body (Gregory 2009)

70.5% (26/37)

31 levels

Radicular pain due to cement leakage needing selective nerve block (grade IIIa)

(1/26)

Systematic review

USA

PubMed was searched and bibliographies of selected articles were examined for additional studies. Search dates not reported.

n= 8 retrospective studies with 265 patients with palliative treatment of vertebral metastasis.

Combined tumour treatment with vertebral stabilisation techniques

not reported.

Inclusion criteria: randomised controlled trials and retrospective studies in English using a multidisciplinary approach of tumour treatment and vertebral stabilisation.

Exclusion criteria: nonhuman studies, case reports, narrative reviews, clinical reports without technical outcomes, and studies involving patients with osteoporotic, traumatic, or metastatic vertebral compression fractures were excluded.

RFA followed by PKP (n=4 studies, 110 patients, 142 vertebrae)

Mean RFA procedure time ranged from 9 to 47 minutes.

Mean ablation temperature 95^°C

Mean cement used ranged from 6.1 to 7.9 ml.

RFA followed by vertebroplasty (n=4 studies, 155 patients, 172 vertebrae)

Mean RFA time ranged from 4.1 to 8.6 minutes.

Mean ablation temperature ranged 71.8°C to 95°C.

Mean cement used ranged from 2.9 to 9.1 ml

mean follow-up periods varied across studies (3 days to 15 months).

No conflicts of interest.

Zheng 2014 (n=26)

7.69 ±1.12

6.62 ±1.02

2.96± 0.92 (6 months)

4.73

Lane 2011 (n=36)

7.2±1.69

NM

3.4±1.6 (1 week)

3.8

Munk 2009 (n=19)

7.9±1

NM

3.82±3.82 (6 weeks)

4.08

Burgard 2014 (n=29)

NM

NM

NM

NM

Wallace 2015 (n=105)

8.0

3.9

2.9 (4 weeks)

5.1

Toyota 2005 (n=17)

6.3

NM

2.4 (3 days)

3.9

Madaelil 2016 (n=11)

8

NM

3 (1 month)

5

Hoffman 2008 (n=22)

8.5

5.5

3.5 (15 months)

5

Extravasation

19 (Lane 2011)

Anterior leaks

2 (Lane 2011)

Cement leakage

7 (Munk 2009)

Local hematoma

2 (Burgard 2014)

Pain

2 (Burgard 2014)

Oxygen saturation decrease

1 (Burgard 2014)

Increased paresis

1 (Burgard 2014)

Postoperative radicular pain

4 (Wallace 2015)

Hematoma

4 (2 in Toyota 2005, 2 in Hoffman 2008)

Cement leaks

8 (Madaelil 2016)

Prospective case series (NCT03249584: OPuS One Study)

Global multicentre study (USA, Canada, Europe)

October 2017- March 2019

n=100 patients with painful metastatic bone disease

metastatic tumour location: 87 involving thoracolumbar spine and 13 around pelvis and/or sacrum.

Mean age 64.6 years (range 30-89 years); 56% (56/100) female

Inclusion criteria: patients at least 18 years of age with metastatic tumours of the thoracic and/or lumbar vertebral body/bodies, peri-acetabulum, iliac crest, and/or sacrum and were candidates for radiofrequency ablation (RFA); had osteolytic bone metastases confirmed by imaging or biopsy.

Exclusion criteria: pure osteoblastic tumours, worst pain rated as < 4 on a scale from 1 to 10 in the past 24 hours, more than 2 painful sites requiring treatment, or Karnofsky performance score < 40.

Radiofrequency ablation (RFA) for palliative treatment with the OsteoCool RF Ablation System done using manufacturer algorithm. In the thoracolumbar spine, the vertebral bodies were accessed via a transpedicular or parapedicular approach. A total of 134 ablations were done under imaging guidance. 68 patients had a single target site treated and 32 had multiple sites treated. 85% of RFA approaches were bilateral (2 probes) in the thoracic and lumbar vertebrae. Polymethyl methacrylate [PMMA] augmentation was done in the majority of cases (97%; 130/134). After RFA patients received chemotherapy (43%), steroids (39%), and osteoporosis medications (38%).

6 months

Medtronic sponsored the study and analysed the data collected by individual sites. Few authors were paid consultants for different companies and received personal fees or research funding.

BPI worst pain (mean score ± SD), 95% CI, p value

8.2 ± 1.7

5.6 ± 2.7

(–3.1 to –1.9, p<0.001)

4.7 ± 2.9

(-3.9 to –2.7, p<0.001)

3.9 ± 3.0

(-4.7 to –3.1, p<0.001)

3.7 ± 2.9

(-5.1 to –3.1, p<0.001)

3.5 ± 3.2

(-6.2 to –3.5, p<0.001)

Subjects with ≥ 2-point change (%)

-

59

66

75

83

86

BPI average pain (mean score ± SD), 95% CI, p value

6.0 ± 2.1

4.0 ± 2.3

(–2.5 to –1.4, p<0.001)

3.3 ± 2.3

(–3.3 to –2.1, p<0.001)

2.8 ± 2.2

(–3.4 to –2.1, p<0.001)

2.8 ± 2.4

(–3.6 to –2.1, p<0.001)

2.9 ± 2.5

(–4.4 to –2.1, p<0.001)

Subjects with ≥ 2-point change (%)

-

51

62

67

74

77

BPI pain interference score (mean score ± SD), 95% CI, p value

6.1 ± 2.3

4.1 ± 2.8

(–2.4 to –1.4, p<0.001)

3.1 ± 2.7

(–3.5 to –2.3, p<0.001)

2.9 ± 2.5

(–3.5 to –2.2, p<0.001)

2.8 ± 2.7

(–3.9 to –2.1, p<0.001)

2.5 ± 2.5

(–4.7 to –2.6, p<0.001)

EQ-5D index (mean score ± SD), 95% CI, p value

0.48 ± 0.32

0.58 ± 0.33

(0.03–0.15, p=0.0018)

0.64 ± 0.28

(0.10–0.21, p<0.001)

0.69 ± 0.21

(0.10–0.24, p<0.001)

0.66 ± 0.26

(0.05–0.25, p=0.0021)

0.69 ± 0.24

(0.09–0.34, p=0.0006)

Retrospective cohort study

USA (single centre)

January 2012 to August 2019

N=166 patients with spinal osseous metastases (266 tumours).

Location of treated tumours: lumbar 51.5% (137/266), thoracic 41.3% (110/266), sacral 6.8% (18/266), and cervical 0.4% (1/266)

Radiotherapy before RFA: 69 patients (108/266 tumours) had EBRT or SBRT ranging from 17 months to 1 month before RFA

Median age 61 years (range 52-69); 53.6% (89/166) female

Inclusion criteria: patients unable to undergo radiation therapy, or tumour progression at sites previously treated with radiation therapy, resistance to chemoradiotherapy, and substantial pain (brief pain inventory score of at least 4); selected by a multidisciplinary team (of radiation and medical oncologists, interventional musculoskeletal radiologists, and oncologic spine surgeons) and treated to achieve local tumour control and pain palliation.

Exclusion criteria: correctable coagulopathy, active infection, entirely osteoblastic metastases, pathologic compression fracture with spinal instability, or metastases resulting in spinal cord compression.

Percutaneous image guided radiofrequency ablation (RFA) combined with or without vertebral augmentation of vertebral metastases was done mainly using conscious sedation. STAR Tumor ablation system used. Vertebral body was accessed via a bipedicular approach. RFA was done according to a sequential bipedicular protocol (in 71% [189/266]) or a simultaneous bipedicular protocol (in 28.9% [77/266]). For both, the ablations were done anteriorly and then the posterior vertebral body and pedicles were treated aligning with the international spinal radiosurgery consortium consensus recommendation.

Vertebral augmentation was done with the StabiliT vertebral augmentation system. Cement was injected using the same working cannulas used for RFA in 91% (242/266) of tumours. Vertebral augmentation was not done if they were small sacral tumours (<2-3 cm) or involved only in the posterior vertebral elements or lower sacral segments.

Passive thermal protection with a decrease of moderate to mild sedation during ablation along the posterior vertebral body was applied in cases done under conscious sedation

6 months (median imaging follow up was 202 days, IQR 142-466 days)

One author received consultation fees and the other authors declared that there are no conflicts of interest

Sequential RFA group

76.% (126/165 tumours)

Simultaneous RFA

85.7% (54/63 tumours)*

BPI score

Pre-procedural score (median)

8±1

Postprocedural score (median, at 1 week, 1,3, 6 months)

3±1^

Total complication rate

3% (8/266)

Major complication rate

0.4% (1/266)

Bilateral lower extremity weakness, difficulty in urination, inability to have an erection 3 days after the procedure (grade 3, attributed to spinal cord thermal injury during the procedure, treated with high-dose intravenous steroids resulting in only slight improvement of symptoms)

n=1

Minor complication rate

2.6% (7/266)

Periprocedural transient radicular pain in the adjacent nerve distribution (grade 2, resolved after transforaminal steroid injections)

n=4

Delayed secondary lumbar vertebral fracture (due to osseous weakening in which vertebral augmentation was not done, grade 2, treated with analgesics)

n=1

Asymptomatic spinal cord oedema along treated area (noted on MRI at 3-4 months, resolved without treatment)

n=2

Other complications not related to RFA

Deaths (related to other causes without spinal cord compression)

27

Progression of metastatic disease (entered hospice care)

5

Surgery (posterior spinal fusion) after RFA

3% (5/166)

Retrospective comparative case series

China

January 2016 to December 2018

N=87 patients with spinal metastatic tumours (125 vertebral bodies).

Group A (treated with RFA combined with bone cement PVP/PKP) n=35 (with 47 vertebral segments [26 thoracic and 21 lumbar])

Group B (treated with bone cement only) n=52 patients (with 78 vertebral segments [42 thoracic and 36 lumbar])

Group A: mean age 51.4 ± 9.3 years; 40% (14/35) female

Group B: mean age 52.2 ± 8.5 years; 38% (20/52) female

Inclusion criteria: definite diagnosis of spinal metastatic cancer (pathological or cytological); structurally intact posterior margin of the vertebral body without nerve root symptoms; thoracic and lumbar vertebral body lesions (of osteolytic or mixed destruction); willingness to undergo the procedure and relatively treatment compliance.

Exclusion criteria: incomplete structure of the posterior margin of the vertebral cortex or infiltration of tumour into the dura, accompanied by nerve root symptoms; osteogenic lesions; terminal patients; severe cardiopulmonary disease or coagulation dysfunction.

Group A (Radiofrequency ablation (RFA) combined with PKP or PVP): RFA was done on the anterior half of the vertebral body at 90°C, and on the posterior part at 75°C and the treatment time was 4–6 min. It was done from different angles, and the electrode was removed slowly. After RFA, PVP or PKP was done.

Group B (Bone cement alone): directly performed PVP or PKP surgery and injected bone cement into the diseased vertebra.

6 months

The authors declared that there are no conflicts of interest.

VAS scores

–

–

–

Baseline

7.52 ± 1.44

7.63 ± 1.52

0.736

3 days

2.79 ± 0.53*

2.88 ± 0.51*

0.429

1 month

2.14 ± 0.40*

2.28 ± 0.43*

0.130

6 months

2.23 ± 0.46*

3.15 ± 0.52*

<0.001

ODI scores

–

–

–

Baseline

77.52 ± 8.84

76.65 ± 8.12

0.638

3 days

48.79 ± 6.45*

49.42 ± 6.94*

0.671

1 month

43.23 ± 5.69*

45.08 ± 6.43*

0.172

6 months

46.46 ± 6.46*

52.15 ± 7.52*

< 0.001

Anterior height of vertebral body

–

–

–

Baseline

18.53 ± 3.84^

18.66 ± 3.24

0.840

Postoperative

24.23 ± 4.25^

23.89 ± 4.34

0.670

Intermediate height of vertebral body

–

–

–

Baseline

24.12 ± 3.88^

24.32 ± 3.52

0.768

Postoperative

28.18 ± 4.25^

27.33 ± 4.39

0.291

Bone cement leak

6.4% (3/35)

20.5% (16/52)

0.033

Tumour recurrence

11.4% (4/35)

30.8% (16/52)

0.036

Retrospective cohort study

USA (one centre)

2011-2017

n= 64 patients with painful spinal metastases

RFA (Spine star) with vertebral augmentation (n=22)

RFA (OsteoCool) with vertebral augmentation (n=12)

Kyphoplasty alone (n=30)

Mean age 62.6 years; male 56% (36/64)

Inclusion criteria: patients greater than 18 years old having metastatic vertebral compression fracture involving the thoracolumbar spine.

Exclusion criteria consisted of non-pathologic osteoporotic compression fractures, metastasis in cervical spine, or previous radiofrequency ablation (RFA) treatment. patients with coagulopathy greater than 1.4 and platelet count less than 50,000, or an active infection either systemically or locally.

Patients were treated with radiofrequency ablation (RFA) using 2 ablation systems (STAR Tumor Ablation System, the CAVITY SpineWand). RFA was done at 70°C for 5–15 min with subsequent cement injection.

Kyphoplasty alone was done in 30 cases.

2 weeks

Authors declare that they have no conflict of interest.

Preoperative pain score

6.9

6

6.3

Postoperative pain score

2.7

1.7

2.3

Between 7-14 days pain score

3.3

3.28

3.69

Opioid use

50% (11/22)

41.7 (5/12)

30% (9/30)

Postoperative day 0

–

–

Kyphoplasty versus OsteoCool

0.49 (1.08)

0.99

Kyphoplasty versus SpineStar

1.63 (0.49)

0.79

OsteoCool versus SpineStar

0.86 (1.12)

0.99

Day 1-14

–

–

Kyphoplasty versus OsteoCool

0.17 (1.12)

1

Kyphoplasty versus SpineStar

1.76 (1.02)

0.72

OsteoCool versus SpineStar

1.67 (1.12)

0.76

30 days and above

–

–

Kyphoplasty versus OsteoCool

1.07 (1.33)

0.98

Kyphoplasty versus SpineStar

2.76 (1.17)

0.14

OsteoCool versus SpineStar

1.23 (1.48)

0.95

Retrospective cohort study

USA (one centre)

2016-2017

n= 26 patients with 28 painful spinal metastases (in thoracic/lumbar spine) treated with

RFA plus cement augmentation (n=17 [17 lesions]) versus combined RFA with radiation therapy (RT, n=10 [11 lesions])

Median age 63 years (no significant difference between 2 groups; p=0.57); male

Inclusion criteria: all patients who underwent RFA for painful spinal metastases, regardless of metastatic disease burden.

Exclusion criteria: patients treated with RT alone.

RFA (n=11 lesions) OsteoCool ablation device was used at more than 60 C. Vertebral augmentation was done after RFA using balloon-assisted techniques with the Kyphon system. Polymethyl methacrylate was injected. At a median of 11 months post-RFA, 3 lesions were treated with radiotherapy for local progression.

Radiotherapy (n=11 lesions) majority of the patients were treated using 3-dimensional conformal radiotherapy (3D-CRT, with a median dose of 30 Gy in 3 Gy daily fractions). Some were treated using volumetric-arc therapy, or stereotactic body radiotherapy (SBRT) (in 2 cases at 28 days post-RFA, both having 35 Gy in 5 fractions). 10 lesions were treated at a median of 28 days and 1 lesion was treated 1 day before RFA.

12 weeks

Authors declare that they have no conflict of interest.

RFA plus cement augmentation

4.2

2.7±2.9

(p<0.0001)

2.1

RFA plus radiotherapy

4.5±3.07

2.7±2.9

(p<0.0001)

1.6±2.6

(p<0.0001)

Local failure^ %

47 (8/17)

9 (1/11)*

0.049

Time to local failure (weeks)

44

Not reached

0.016

Distant failure

NR

NR

0.70

Time to distant failure (weeks)

11.3

36.3

0.15

Survival (median, weeks)

31.9

55.3

0.0045

Retrospective cohort study

China

2010-2013

n= 169 patients with painful spinal metastases (in thoracic/lumbar spine) treated with

Group A: percutaneous vertebroplasty (PVP) combined with ¹²⁵I seed implantation (n=49)

Group B: PVP combined with RFA (n=51)

Group C PVP combined with Zoledronic acid (n=38) and

Group D: PVP combined with radiotherapy (n=31).

Median age 56.9 years.

(95/169) male

not reported

Group A: underwent PVP combined with ¹²⁵I seed implantation (n=49)

Based on patient condition the dose and distribution of ¹²⁵I seed implantation was done and further evaluated using the dosimetry protocol. Then, bone cement was injected into the patient's diseased vertebra.

Group B: underwent PVP combined with RFA (n=51)

radiofrequency ablation (RFA) was done and bone cement in a semi-solidified state was injected.

Group C: underwent PVP combined with Zoledronic acid (n=38)

PVP procedure was under local anaesthesia, then bone cement was slowly injected into the vertebral body until reaching the edge. The patients were then treated with intravenous drip of 4 mg zoledronic acid for 15 minutes and once every 3–4 weeks.

Group D: underwent PVP combined with radiotherapy (n=31)

PVP was done under local anaesthesia and then bone cement was separately injected into the thoracic and lumbar region, respectively. Three or 4 days after PVP, patients had radiotherapy, with the diseased vertebra as the central point of radiotherapy, 10 times in 2 weeks and total radiation dose of 30 Gy.

All patients were supervised for 24 hours after operation. CT was done to investigate the distribution of bone cement in the diseased vertebra, and antibiotics were given to prevent postoperative infection.

6 months

Authors state that they have no conflict of interest to declare.

PVP combined with 125 I seed implantation

–

–

–

Baseline

8.16±1.06

68.19±0.89

-

24 hours

3.91±1.01^

49.92±1.01^*

73.47%+

1 month

3.15±1.16^

48.26±0.99^

71.43%

6 months

2.39±0.89^

47.99±0.89^

67.35%

PVP combined with RFA

–

–

–

Baseline

8.07±0.79

71.04±0.83

–

24 hours

4.61±0.75^#

37.03±0.76^

76.47%+

1 month

4.38±0.61^

36.84±0.91^

74.51%

6 months

4.34±0.31^

36.61±1.04^

72.55%

PVP combined with zoledronic acid

–

–

–

Baseline

8.02±0.93

67.85±0.88

–

24 hours

4.43±1.19^#

49.21±0.87^*

73.68%+

1 month

4.27±0.76^

48.94±0.44^

71.05%

6 months

3.99±0.41^

48.87±0.54^

65.79%

PVP combined with radiotherapy

–

–

–

Baseline

7.91±0.92

70.22±0.92

–

24 hours

4.72±0.21^#

41.01±0.37^*

90.32%

1 month

4.66±0.31^

40.83±0.43^

87.10%

6 months

4.63±0.10^

40.74±0.54^

83.87%

Case report

USA

Not reported

N=1 patient with spinal metastases

61-year-old woman

RFA and kyphoplasty using polymethylmethacrylate for spinal metastases using a bipolar cooled RFA system to treat lesions at L1 and L3

1 year

Not reported