Factors associated with local recurrence after ... - Wiley Online Library

0 downloads 0 Views 549KB Size Report
Sep 19, 2017 - Jong Woong Park MD1. | Ilkyu Han MD, PhD1. 1 Department of Orthopaedic Surgery, Seoul. National University Hospital, Seoul, Korea.
Received: 27 June 2017

|

Revised: 9 August 2017

|

Accepted: 19 September 2017

DOI: 10.1002/jso.24880

RESEARCH ARTICLE

Factors associated with local recurrence after surgery for bone metastasis to the extremities Chandra Kumar Krishnan MS1,2 | Han-Soo Kim MD, PhD1 | Ji Yeon Yun BS1 | Hwan Seong Cho MD, PhD3 | Jong Woong Park MD1 | Ilkyu Han MD, PhD1 1 Department

of Orthopaedic Surgery, Seoul National University Hospital, Seoul, Korea

Background and Objectives: With increasing life expectancy of patients with bone

2 Department

metastasis, durable surgical stabilization of bone metastasis is necessary. Local

of Surgical Oncology, Cancer Institute (WIA), Adyar, Chennai, India

recurrence (LR) can compromise surgical stabilization and necessitate retreatment. We

3 Department

of Orthopedic Surgery, Seoul National University Bundang Hospital, Bundang-gu, Seongnam, Korea

analyzed LR rate and factors associated with LR in patients undergoing surgery for bone metastasis. Methods: Patients (n = 301) who underwent surgery for bone metastasis to the

Correspondence Ilkyu Han, MD, PhD, Department of Orthopaedic Surgery, Seoul National University Hospital, 101 Daehak-ro Jongnogu, Seoul 03080, Korea. Email: [email protected]

extremities were reviewed. Possible factors that might be associated with LR were investigated. Results: LR rate was 16% (49/301). Surgical margin was associated with LR, as patients with en-bloc resection had significantly less LR than patients who underwent curettage (5/66 vs 44/235, P = 0.03). Prostate cancer had lowest rate (0%) of LR and colon cancer had highest rate (31%). Interval from surgery to LR differed among primary cancer types (4.5 ± 3.9 months [lung cancer], vs 12.3 ± 12.9 months [other cancers], P = 0.041). In multivariate analysis, en-bloc surgical margins (HR = 0.372, P = 0.036) and primary cancers of breast or prostate (HR = 0.391, P = 0.049) were independent factors associated with longer LR-free survival. Conclusions: LR after surgery for bone metastasis to extremities is affected by surgical margin and primary cancer type. These factors, along with expected patient survival, need to be considered when planning surgery for bone metastasis to extremities. KEYWORDS

bones of lower extremity, bone neoplasms, bones of upper extremity, neoplasm metastasis

1 | INTRODUCTION

of patients with bone metastases.3–5 Given the increasing life expectancy of patients with bone metastases, durable surgical

Although most patients with bone metastases to the extremities have a

stabilization of bone metastases, which provides lifelong pain relief

limited life expectancy of only a few months, a considerable proportion

and functional recovery, is necessary.6 Durability of surgical stabiliza-

1,2

of patients live longer, even a few years.

Moreover, ongoing

tion is determined not only by the adequacy of the surgical

improvements in systemic therapy are prolonging the life expectancy

construction, such as the choice of implant or the length of the fixation, but also by the effectiveness of local tumor control. Local

Investigation performed at Department of Orthopaedic Surgery, Seoul National University Hospital, 101 Daehak-ro Jongno-gu, Seoul 03080, Korea.

J Surg Oncol. 2017;1–8.

tumor recurrence can compromise surgical stabilization and necessitate re-treatment.7,8 Although there are many reports in the orthopedic

wileyonlinelibrary.com/journal/jso

© 2017 Wiley Periodicals, Inc.

|

1

2

KRISHNAN

|

literature regarding adequate surgical construction, few studies have

TABLE 1

been published on local tumor control after surgical stabilization of bone

Variable

metastasis to the extremities.7–15

Previous radiotherapy

To provide durable surgical stabilization and select optimal surgical options with regard to resection and reconstruction, identification of the risks of local recurrence (LR) and its associated factors is needed. Therefore, this study sought to identify the LR rate and to analyze the clinicopathologic factors associated with LR in patients undergoing surgery for bone metastasis to the extremities.

ET AL.

Treatment of bone metastasis N

%

Not received

247

83

Received

49

17

En-bloc resection

66

22

Curettage

235

78

Intramedullary nail

154

52

Bipolar hip hemiarthroplasty

52

17

Megaprosthesis

51

17

Surgical Margin

Reconstruction used

2 | M ATERIA LS AN D METH ODS A retrospective review of a prospectively collected institutional database was performed and identified 437 patients who underwent surgery for bone metastasis to the extremities between January 2001 and December 2015. The institutional review board of the hospital approved this study. For the purpose of analysis, the following exclusion criteria were applied: 1) patients with hematolymphoid malignancy of the bone

Plate

24

8

Cement with screws and pins

16

5

No reconstruction

4

1

Not received

174

59

Received

121

41

Postoperative radiotherapy

(n = 64), 2) patients who had undergone operation on the small bones of the hands or feet (n = 11), 3) patients whose tumors were not surgically removed (n = 41), and 4) patients who underwent amputa-

applied to the main lesion. Stems of endoprostheses were fixed with

tion (n = 1). Of the remaining 320 patients, 19 patients with less than

methylmethacrylate cement. Forty-nine (17%) patients had undergone

1 month of follow up for LR were excluded, which left us with 301

previous radiation therapy for the main lesion. When postoperative

patients for analysis. For patients with multiple surgeries on the same

radiation therapy was administered, the entire bone was included in the

lesion, only the index surgery was counted (n = 28). Among the 301

field.

patients, 249 (83%) died during the follow-up period. The actuarial

In general, postoperative follow-up evaluations were performed at

survival rate calculated through Kaplan-Meier analysis at 6 months,

2 weeks and then at 3-6 month intervals. For LR surveillance, imaging

1 year, and 2 years was 62.1 ± 2.8%, 40.0 ± 2.9%, and 22.2 ± 2.6%,

such as plain radiographs, MRI scans, CT scans, PET scans, or bone

respectively. The mean follow-up period of the entire cohort was

scans were examined. The follow-up schedule and modality varied

14 months (range, 1-130 months).The median follow-up time in

according to each patient's condition.

survivors was 14 months (range, 1-130 months).

LR-free survival was the endpoint of the study. LR was defined as

Surgery was performed for impending or existing pathologic

radiological identification of recurrence in the operated tumor bed.

fractures. All surgeries were performed by two senior orthopedic

Tumor progression in a separate lesion where the tumor had not been

oncologists (HSK and IH). Surgical options regarding resection and

removed was not regarded as LR. Patients with no signs of LR at the

reconstruction were determined by (1) location of the main lesion in the

last imaging were considered censored, meaning that the outcome did

bone; (2) degree of cortical destruction; (3) radiological pattern of the

not occur by the last imaging. LR-free survival was defined as the

lesion (lytic vs sclerotic); (4) presence of separate lesions in the operated

period between the date of surgery and the date of LR or date of last

bone; (5) response to adjuvant radiation therapy or chemotherapy; (6)

imaging. We only considered the first identification of LR in our time-

expected survival time; and (7) patient preference. Resection was

to-event analysis. We obtained data on the following variables for

performed as en-bloc resection or curettage (Table 1). En-bloc resection

association with LR: (1) patient demographics, (2) primary cancer type,

was done by resecting the lesion as a whole with surrounding normal

(3) cancer burden, (4) characteristics of the bone metastatic lesion

tissue. Curettage was performed under direct vision to remove all

operated on, and (5) treatment of the bone metastasis.

grossly visible tumors. Electrocauterization or high-speed burrs were

For patient demographics, patients’ sex and age were retrieved.

used to remove possible remaining tumor when possible. Reconstruc-

One hundred and ninety-five patients were men (65%) (Table 2). The

tion was performed using osteosynthetic devices or endoprostheses.

mean age was 60 years (range, 19-91). For the purpose of analysis,

Osteosynthetic devices were predominantly used after curettage and

patients were dichotomized as ≤60 years (n = 150, 50%) and >60 years

endoprosthetic reconstruction after en-bloc excision. The most

(n = 151, 50%). The most common primary cancers in the study group

common reconstruction methods used were intramedullary nails

were lung cancer (n = 73, 24%), hepatocellular carcinoma (n = 68, 23%),

(n = 154, 52%), bipolar hemiarthroplasty of the hip (n = 52, 17%), and

renal cell carcinoma (n = 51, 17%), and breast cancer (n = 35, 11%). For

megaprosthesis (n = 51, 17%). When osteosynthetic devices were used

the purpose of analysis, patients were grouped into the breast or

after curettage, methylmethacrylate cement augmentation was also was

prostate cancer group, which respond well to the medical treatment, and

KRISHNAN

ET AL.

TABLE 2

| TABLE 2

Patient characteristics N

%

≤60

150

50

>60

151

50

Variable

3

(Continued) N

%

≤6.8

178

59

>6.8

123

41

Variable Size, cm

Age, years

Time from cancer diagnosis to bone metastasis, months

Sex Male

195

65

≤8

152

51

Female

106

35

>8

148

49

Lung cancer

73

24

Hepatocellular carcinoma

68

23

other cancers group.16 Histological confirmation was obtained in all

Renal cell carcinoma

51

17

cases. For cancer burden, time from cancer diagnosis to bone metastasis,

Breast cancer

35

11

Colon cancer

13

4

Prostate

8

3

Sarcoma

6

2

patients (10%) presented with solitary bone metastases at the time of

Bladder cancer

5

2

surgery, defined as metastasis to a single bone without metastasis to

Gastric cancer

5

2

other bones or organs. For characteristics of the bone metastatic lesion

Metastasis of unknown origin

4

1

operated on, the bone operated on, the location of the metastasis within

Others

33

11

the bone, the presence of pathological fractures, and the size of the bone

Primary cancer type

presence of visceral metastases, and presence of solitary metastases were investigated. The median time from the diagnosis of the cancer to the diagnosis of bone metastasis was 8 months (range, 0-248). At the time of surgery, 220 patients (73%) had visceral metastases. Thirty

metastasis were investigated. The femur (n = 208, 70%) and humerus

Involvement of other bones

(n = 70, 23%) were the most common bones operated on. The ends of

Not present

62

21

Present

239

79

Not present

80

27

as the largest diameter on preoperative imaging or pathology report, was

Present

220

73

6.8 cm (range, 2.0-36.0). For treatment of the bone metastasis, surgical

No

270

90

administration of postoperative radiation were investigated. The

Yes

30

10

surgical margin was categorized into en-bloc resection or curettage

the bone were involved in 209 patients (70%) and the shafts in 89 Visceral metastasis

patients (30%). One hundred and forty-six patients (49%) presented with pathological fractures. The mean size of the metastasis, measured

margin, previous radiation to the bone metastatic lesion, and the

Solitary bone metastasis

(Table 1). En-bloc resection was performed in 66 patients (22%) and

Bone operated on

curettage in 235 patients (78%). All patients with en-bloc resection had

Femur

208

70

Humerus

70

23

Tibia

16

5

(range, 8-60 Gy). Postoperative radiation therapy was considered in

Radius

4

1

patients with longer life expectancy, radiation-sensitive primary cancer

Fibula

3

1

or whose tumors were removed with curettage. However, no

bone was undertaken in 121 patients (41%) with a median dose of 30 Gy

prospective criteria was used to administer postoperative radiation

Part of bone Proximal

200

67

Shaft

89

30

Distal

negative margins. Postoperative radiation therapy covering the entire

9

3

therapy. We presented continuous variables as a mean with SD and categorical variables as frequencies with percentages. We used a Chisquare test for categorical variables and analysis of variance for

Pathological fracture

continuous variables. Various clinicopathological factors were analyzed

Absent

155

51

Present

146

49

271

90

variables, multivariate analysis was performed using the variables with P

Sclerotic

19

6

values of 60

91

81

73

55 0.052a

Primary cancer type Lung cancer

82

78

74

74

Hepatocellular carcinoma

90

78

59

29

Renal cell carcinoma

95

85

74

51

Breast cancer

94

85

85

85

Colon cancer

69

35

35

0

Prostate cancer

100

100

100

100

Sarcoma

80

80

80

NA

Bladder cancer

80

80

NA

NA

Gastric cancer

100

NA

NA

NA

Absent

90

86

80

71

Present

88

77

72

62

Visceral metastasis

0.228

Solitary bone metastasis

0.135

No

89

80

73

60

Yes

90

85

85

85

Time from cancer diagnosis to bone metastasis, months

0.375

≤8

88

77

72

59

>8

90

85

78

73

Pathological fracture

0.933

Absent

89

81

74

69

Present

88

80

76

58

Size, cm

0.606

≤6.8

91

83

75

64

>6.8

85

78

74

67

Previous radiotherapy

0.347

Not received

87

82

75

64

Received

100

72

72

72

En-bloc resection

96

87

87

74

Curettage

87

79

72

63

Not received

86

78

70

70

Received

91

84

79

63

Surgical margin

0.037

Postoperative radiotherapy

NA, not available. Breast or prostate cancer vs. others.

a

P-value

0.512

5

6

KRISHNAN

|

ET AL.

(4.5 ± 1.1 months vs 12.2 ± 2.1 months, P = 0.006) (Table 4). On Kaplan-Meier analysis of factors associated with the development of LR, primary cancer type showed a trend toward significance, as patients with prostate cancer or breast cancer had a longer LR-free survival than patients with other cancers (101.7 ± 13.7 months vs 72.2 ± 7.5 months, P = 0.052) (Figure 1) (Table 3). The actuarial local control rate of prostate cancer or breast cancer was higher than that of other cancers (at 1 year, 88.4 ± 5.5% vs 78.8 ± 3.5%; at 2 year, 88.4 ± 5.5% vs 70.8 ± 4.7%, respectively, P = 0.052).

3.3 | Multivariate analysis On multivariate analysis of variables associated with LR-free survival, en-bloc surgical margin (HR = 0.372, P = 0.036) and primary cancers of the breast or prostate (HR = 0.391, P = 0.049) were independent factors associated with a lower likelihood of LR (Table 5). None of the factors related to patient demographics, cancer burden, or characterFIGURE 3 Kaplan-Meier analysis of local recurrence-free survival according to the surgical margin. Patients with en-bloc resection had significantly better local recurrence-free survival rates than patients with curettage (P = 0.037, log rank test)

istics of the bone metastatic lesion were associated with LR-free survival.

4 | DISCUSSION after en-bloc resection was higher than that after curettage (at 1 year, 86.6 ± 6.6% vs 78.9 ± 3.4%; at 2 years, 86.6 ± 6.6% vs 71.1 ± 4.4%, respectively).

The prevalence of bone metastasis is rising due to a higher rate of diagnosis, better systemic treatment, patients with cancer living longer, and a higher disease burden rate.3,17,18 With advances in treatment, patients with cancer are living longer, and thus there is a greater responsibility on the orthopedic surgeon to preserve the

3.2 | Primary cancer type

quality of life by preventing and managing problems related to bone

LR rate varied according to the primary cancer type. The lowest LR rate

metastasis. Surgery for bone metastasis should not fail to outlive the

was observed in patients with prostate cancer (0%, 0/8), followed by

patient, since retreatment may not be possible or safe with advancing

breast cancer (14%, 5/35), and the highest LR rate was observed in

disease.3,7 Causes of failure after bone metastasis surgery can be

patients with colon cancer (31%, 4/13) (Table 4). In addition, the

categorized into either non-oncologic reasons or oncologic reasons,

interval from surgery to LR differed among primary cancer types, as

that is, local recurrence or tumor progression. Most existing literature

lung cancer had significantly a shorter interval than other cancers

on failure after bone metastasis surgery has reported non-oncologic

TABLE 4

Comparison of local recurrence and survival among primary cancer types

Primary cancer type

Number

Number with LR

LR rate

Interval to LRa, months (range)

Lung cancer

73

13

18%

5

Postoperative survivalb, months (95%CI) (1-14)

8

(6-10)

Hepatocellular carcinoma

68

10

15%

11

(2-27)

7

(5-9)

Renal cell carcinoma

51

9

18%

14

(1-34)

9

(7-19)

Breast cancer

35

5

14%

15

(2-52)

16

(8-26)

(1-36)

4

(2-6)

17

(0-43)

Colon cancer

13

4

31%

14

Prostate cancer

8

0

0%

NA

Sarcoma

6

1

17%

NA

14

(3-27)

Bladder cancer

5

1

20%

NA

4

(3-5)

LR, local recurrence; CI, confidence interval; NA, not applicable. a Mean interval from surgery to LR. b Estimated median post-operative survival.

KRISHNAN

ET AL.

TABLE 5 recurrence

|

Multivariate analysis of prognostic factors of local

Prognostic factor

7

renal cell carcinoma.4,21,22 Of note, time to LR showed a trend toward significance as being shorter in the curettage group than in the en-bloc

Risk ratio

P-value

95%CI

group (8.9 ± 10.7 months vs 15.5 ± 14.8 months, P = 0.113). However, we found no difference between the curettage group and the en-bloc

Primary cancer type Breast or prostate

0.391

Others

1

0.154-0.995

0.049

group in terms of the rate of re-treatment (54% [21/39] vs 44% [4/9], respectively, P = 0.933) or re-operation (26% [10/39] vs 22% [2/9], respectively, P = 0.711). This is similar to the observation made by

Surgical margin

Langerhuizen et al.21 No difference in complication rates between the

En-bloc resection

0.372

Curettage

1

0.147-0.939

0.036

two groups was seen; 18% (12/66) of patients who underwent wide resection developed complications compared to 13% (30/235) of patients who underwent curettage.

CI, confidence interval.

LR was affected by the primary cancer type. The LR rate differed according to the primary cancer type, as colon cancer showed the reasons such as implant breakage or infection,8,9,19 and little literature

highest LR rate (31%), while prostate and gastric cancers showed

has focused on oncologic reasons for failure.8 Identification of

lowest LR rate (0%) (Table 4). Moreover, the time to LR differed

associated factors of LR after bone metastasis surgery would be

according to the primary cancer type, as lung cancer showed the

helpful in planning surgery. Although there are studies reporting LR

shortest interval to LR (Table 4). These findings, along with the known

after surgery for bone metastasis to the extremities, they have

prognostic impact of the primary cancer type23 highlight the

included fewer patients and are most pertaining to a single primary

importance of primary cancer type when planning surgery for bone

cancer type or a single bone or a part of a bone. The authors believe

metastasis.

that this study is the first to focus on LR after surgery for bone

No significant difference in LR rate was found between the

metastasis to the extremities in a large cohort of patients with various

patients who received postoperative radiation therapy and those who

primary cancer types. This study investigated a variety of possible

did not (22/121 vs 27/174, respectively, P = 0.545). This finding was

associated factors with respect to LR after surgery for bone metastasis

consistent across all primary cancer types. These findings may be due

to the extremities and identified the surgical margin and the primary

to selection bias, as postoperative radiation could have been

cancer type as independent factors associated with LR.

administered preferentially in patients with a high risk of LR, such as

This study had several limitations to be noted. First, the relatively

an insufficient surgical margin or a larger extent of the tumor. Whether

small number of patients limits the interpretation of our results. Given

or not the administration of postoperative radiation therapy reduces

the different LR patterns observed according the primary cancer types,

LR in bone metastasis to the extremities needs to be validated in larger

a larger number of patients is needed to address the impact of each

cohort, possibly in a prospective setting.

primary cancer type on LR after surgery for bone metastasis. Second,

Factors representing the cancer burden were not associated with

treatment for bone metastasis may not have been standardized, as this

LR. The presence of solitary metastasis (P = 0.135), time from cancer

study was performed over a relatively long period of time. Moreover,

diagnosis to bone metastasis (P = 0.375), and the presence of visceral

medical management of bone metastases, such as chemotherapy, anti-

metastasis (P = 0.228) were not associated with LR. These findings may

bone resorbing drugs, and hormone therapy was not considered.

imply that local control after surgery for bone metastasis is affected

However, our treatment paradigm did remain largely unchanged over

more by local factors, such as the surgical margin, than systemic

the study period. Third, the treatment of the primary cancer and

factors.

metastases to other organs were not considered. Fourth, the testing of

As expected, patients with a longer postoperative survival time

multiple (ie, 11) variables might have increased the chance of finding a

had a higher rate of LR. Twenty-six percent of patients who survived

statistically significant association. Finally, the surveillance protocol for

more than one year after surgery for bone metastasis developed LR as

LR was not uniform. Majority of the patients had only plain radiographs

compared to 4% and 16% of patients who survived for 3 months and

for follow up, which might have led to the underestimation of LR of

12 months, respectively. This is in line with previous studies, where

osteosclerotic metastases such as the metastasis of prostate cancer.

failure after surgery for bone metastasis increased with sur-

The LR rate of this study was 16%, similar to the LR rate of two

vival.3,8,24,25 Of note, the association between survival time and LR

9,20

previous studies, which reported LR rates of 12% and 17%.

Of note,

rate was dependent upon the primary cancer type, as this association

for the purpose of analysis, we counted only the LR in the original

was not evident with lung cancer, possibly due to the shorter interval

tumor bed. There were patients in whom tumor progression was seen

to LR.

in a nearby lesion, which had not been treated. We found a significant difference in the LR rate based on surgical

Of the 49 patients who developed LR, 26 (53%) were symptomatic and had either pain or instability which necessitated retreatment in the

margin, as the LR rate was significantly higher in the curettage group

form of repeat surgery (n = 12), radiation (n = 12), or systemic therapy

than in the en-bloc group (P = 0.03). This finding was consistent across

(n = 2), while 23 (47%) patients received no intervention for their LR.

the various primary cancer types in this study. The importance of

The mean survival time of patients who underwent retreatment for

surgical margin on LR after bone metastasis surgery has been shown in

their LR was significantly longer than that of patients who were not

8

KRISHNAN

|

treated (25.6 ± 20.2 months vs 12.5 ± 13 months, P = 0.020). Patients with a better general condition and lower oncological burden underwent retreatment, giving rise to this significant difference in the survival time between these groups.

5 | C ONC LU SI ON S In conclusion, LR after surgery for bone metastasis to the extremities is affected by the surgical margin and the primary cancer type. These factors, along with the expected patient survival, need to be considered when planning surgery for bone metastasis to the extremities.

ACKNOWLEDGMENTS This work was not supported by any fund.

ORCID Ilkyu Han

http://orcid.org/0000-0001-9841-3775

REFERENCES 1. Fottner A, Szalantzy M, Wirthmann L, et al. Bone metastases from renal cell carcinoma: patient survival after surgical treatment. BMC Musculoskelet Disord. 2010;11:145. 2. Rove KO, Crawford ED. Metastatic cancer in solid tumors and clinical outcome: skeletal-related events. Oncology (Williston Park). 2009;23:21–27. 3. Agarwal MG, Nayak P. Management of skeletal metastases: an orthopaedic surgeon's guide. Indian J Orthop. 2015;49:83–100. 4. Ratasvuori M, Wedin R, Hansen BH, et al. Prognostic role of en-bloc resection and late onset of bone metastasis in patients with boneseeking carcinomas of the kidney, breast, lung, and prostate: sSG study on 672 operated skeletal metastases. J Surg Oncol. 2014;110:360–365. 5. Ratasvuori M, Wedin R, Keller J, et al. Insight opinion to surgically treated metastatic bone disease: scandinavian Sarcoma Group Skeletal Metastasis Registry report of 1195 operated skeletal metastasis. Surg Oncol. 2013;22:132–138. 6. Damron TA, Sim FH. Surgical treatment for metastatic disease of the pelvis and the proximal end of the femur. Instr Course Lect. 2000;49:461–470. 7. Miller BJ, Soni EE, Gibbs CP, Scarborough MT. Intramedullary nails for long bone metastases: why do they fail? Orthopedics. 2011;34:274. 8. Wedin R, Bauer HC, Wersall P. Failures after operation for skeletal metastatic lesions of long bones. Clin Orthop Relat Res. 1999;358: 128–139. 9. Alvi HM, Damron TA. Prophylactic stabilization for bone metastases, myeloma, or lymphoma: do we need to protect the entire bone?. Clin Orthop Relat Res. 2013;471:706–714.

ET AL.

10. Baloch KG, Grimer RJ, Carter SR, Tillman RM. Radical surgery for the solitary bony metastasis from renal-cell carcinoma. J Bone Joint Surg Br. 2000;82:62–67. 11. Capanna R, Campanacci DA. The treatment of metastases in the appendicular skeleton. J Bone Joint Surg Br. 2001;83:471–481. 12. Chan D, Carter SR, Grimer RJ, Sneath RS. Endoprosthetic replacement for bony metastases. Ann R Coll Surg Engl. 1992;74:13–18. 13. Chandrasekar CR, Grimer RJ, Carter SR, et al. Modular endoprosthetic replacement for metastatic tumours of the proximal femur. J Orthop Surg Res. 2008;3:50. 14. Sarahrudi K, Greitbauer M, Platzer P, et al. Surgical treatment of metastatic fractures of the femur: a retrospective analysis of 142 patients. J Trauma. 2009;66:1158–1163. 15. Uptodate: Evaluation and management of complete and impending pathological fractures in patients with metastatic bone disease, multiple myeloma, and lymphoma. In Timothy A Damron M, Jeffrey A Bogart M, Mark Blisky M (eds), 2016. 16. Bauer HC, Wedin R. Survival after surgery for spinal and extremity metastases. Prognostication in 241 patients. Acta Orthop Scand. 1995; 66:143–146. 17. Blank AT, Lerman DM, Patel NM, Rapp TB. Is prophylactic intervention more cost-effective than the treatment of pathologic fractures in metastatic bone disease?. Clin Orthop Relat Res. 2016;474:1563–1570. 18. Harrington KD. Orthopedic surgical management of skeletal complications of malignancy. Cancer. 1997;80:1614–1627. 19. Steensma M, Boland PJ, Morris CD, et al. Endoprosthetic treatment is more durable for pathologic proximal femur fractures. Clin Orthop Relat Res. 2012;470:920–926. 20. Epstein-Peterson ZD, Sullivan A, Krishnan M, et al. Postoperative radiation therapy for osseous metastasis: outcomes and predictors of local failure. Pract Radiat Oncol. 2015;5:e531–e536. 21. Langerhuizen DW, Janssen SJ, van der Vliet QM, et al. Metastasectomy, intralesional resection, or stabilization only in the treatment of bone metastases from renal cell carcinoma. J Surg Oncol. 2016; 114:237–245. 22. Les KA, Nicholas RW, Rougraff B, et al. Local progression after operative treatment of metastatic kidney cancer. Clin Orthop Relat Res. 2001;390:206–211. 23. Nathan SS, Healey JH, Mellano D, et al. Survival in patients operated on for pathologic fracture: implications for end-of-life orthopedic care. J Clin Oncol. 2005;23:6072–6082. 24. Helwig U. Multiple operations in patients with bony metastases and a prolonged course of disease. Eur J Surg Oncol. 1997;23:59–63. 25. Yazawa Y, Frassica FJ, Chao EY, et al. Metastatic bone disease. A study of the surgical treatment of 166 pathologic humeral and femoral fractures. Clin Orthop Relat Res. 1990;251:213–219.

How to cite this article: Krishnan CK, Kim H-S, Yun JY, Cho HS, Park JW, Han I. Factors associated with local recurrence after surgery for bone metastasis to the extremities. J Surg Oncol. 2017;1–8. https://doi.org/10.1002/jso.24880

SYNOPSIS There are many reports in the orthopedic literature regarding adequate surgical construction, few studies have been published on local tumor control after surgical stabilization of bone metastasis to the extremities. Therefore, this study sought to identify the local recurrence (LR) rate and to analyze the clinicopathologic factors associated with LR in patients undergoing surgery for bone metastasis to the extremities. LR after surgery for bone metastasis to extremities is affected by surgical margin and primary cancer type.