|Year : 2015 | Volume
| Issue : 2 | Page : 264-267
Prognostic role of pathologic fracture in osteosarcoma: Evidence based on 1,677 subjects
Department of Orthopedics, The First Affiliated Hospital of Chongqing Medical University, Chongqing, China
|Date of Web Publication||7-Jul-2015|
Department of Orthopedics, The First Affiliated Hospital of Chongqing Medical University, Chongqing - 400 016
Source of Support: None, Conflict of Interest: None
Introduction: The prognostic role of pathologic fracture in osteosarcoma has not been settled down, as previous researches have reached contradictory results. This study is prepared to clarify whether pathologic fracture predicts a poor prognosis of patients with osteosarcoma.
Materials and Methods: After completely retrieved databases including China National Knowledge Infrastructure (CNKI), PubMed, Highwire, EMBASE and Science Direct, all eligible articles were included based on inclusion/exclusion criteria. And we performed a meta-analysis to calculate the pooled prognostic role of pathologic fracture in osteosarcoma.
Results: Finally, a total of 8 articles met the inclusion/exclusion criteria, involving 1,677 subjects. We found that the pooled hazard ratios (HRs) with 95% confidential interval (CI) of pathologic fracture in osteosarcoma for overall survival (OS) and disease-free survival (DFS) are 2.13 (1.43, 3.15) and 1.58 (1.11, 2.24), respectively.
Conclusion: This meta-analysis demonstrated that presentation with a pathologic fracture in osteosarcoma was correlated with a poor prognosis. And pathologic fracture might be a poor predictor of survival in osteosarcoma.
Keywords: Meta-analysis, osteosarcoma, pathologic fracture, prognosis
|How to cite this article:|
Yang M. Prognostic role of pathologic fracture in osteosarcoma: Evidence based on 1,677 subjects. J Can Res Ther 2015;11:264-7
| > Introduction|| |
Historically, a popular hypothesis insisted that osteosarcoma complicated by a pathologic fracture would result in "a hematoma which may spread and contaminate the adjacent soft tissue, the neurovascular bundle and an adjacent joint",  and limb salvage surgery has been assumed to increase the risk of local recurrence. Thus, amputation has been a recognized surgical strategy for an osteosarcoma patient complicated by pathologic fracture.  However, the concept was gradually shifting based on some reports , which found that the ratio of local recurrence between limb salvage and amputation in osteosarcoma complicated by a pathologic fracture has no significant difference.
As a controversial clinical issue, the prognostic role of pathologic fracture in osteosarcoma is not well determined until now. Because osteosarcoma is an infrequent disease, and osteosarcoma accompanying a prognostic fracture is even rarer. Researchers found that 5-10% osteosarcoma patients present pathologic fracture at diagnosis or during adjuvant chemotherapy.  Even though some studies had been designed to explore the survival impact of pathologic fracture in osteosarcoma patients, a relatively small number of subjects might influence the reliability of their conclusions. In addition, the difference of study design, inclusion/exclusion criteria, variables assessment and therapeutic protocols in these clinical studies finally results in contradictory conclusions.
According to this, the prognostic role of pathologic fracture in osteosarcoma has not been settled down. Thus, the aim of our meta-analysis was designed to investigate the prognostic relationship between pathologic fracture and osteosarcoma.
| > Materials and methods|| |
Our search strategy included the following keywords variably combined by "osteosarcoma", "fractures, spontaneous", "fractures", "spontaneous", "spontaneous fractures", "pathologic", and "pathologic fracture". Databases that we have searched included CNKI, PubMed, Highwire, EMBASE and Science Direct.
Studies were regard as qualified if they met all of the following inclusion criteria: (I) The tumor type is osteosarcoma; (II) the research should be a case/control study; III) the prognostic associations between pathologic fracture and osteosarcoma, including overall survival (OS), progression-free survival or disease-free survival (PFS/DFS) were measured. Study was excluded based on any of the following criteria: (I) the following article types were excluded: Reviews, letters, case reports, surgery associative articles, or animal experiments; (II) the articles were not related to information of osteosarcoma complicated by pathologic fracture; (III) the research was a cohort study or have no control group; (IV) it lacked critical data for hazard ratio (HR) analysis.
Two researchers were assigned to filter the initial included articles based on the inclusion/exclusion criteria, independently. Any disagreement was resolved by consensus. Baseline data and survival information were extracted by the other two researchers independently. The original survival data, including the Kaplan-Meier (K-M) survival curves or HR and 95% confidence interval (CI) of survival outcomes. Multivariate Cox hazard regression analysis data is our priority, if not, univariate Cox hazard regression analysis or K-M survival curves with log-rank P value of survival outcomes were instead. The data of K-M survival curves was extracted by Engauge Digitizer 4.1 (http://sourceforge.net). Moreover, we extracted baseline data of eligible articles, including first author, publication year, subjects (case/control), match factors, tumor types, follow-up, types of survival outcomes, therapy strategy, and altitude of authors.
This meta-analysis was carried out using STATA 11.0 (STATA Corporation, College Station, TX). Heterogeneity was defined as P < 0.05 or I 2 ≥ 60%. If heterogeneity was fine (P ≥ 0.05, I 2 < 60%), a fixed effect model was used to pool the HRs.  On the contrary, a random effect model was used to calculate the prognostic role of pathologic fracture in osteosarcoma, if the subgroup analysis was unable to decrease the heterogeneity. If the pooled HR > 1 and the 95% CI exclude 1, it indicated a significant worse outcome for the case group compared to the control group. The publication bias was tested by Begg's test and Egger's test, with P < 0.05 indicating that potential publication bias existed.
| > Results|| |
Eligible articles filtration
After initial retrieval, 303 articles were included, involving 205 articles in PubMed, 20 articles in CNKI, 36 articles in High wire, 17 articles in EMBASE and 25 articles in Science Direct. Following careful screened titles and abstracts of each article, 280 articles were excluded. 23 articles were retrieved for further filtration. After screened the entire parts of articles, 15 articles were further excluded. Eight articles ,,,,,,, were finally included as eligible articles, including a total of 1,677 subjects with a median number of 209.75 subjects per study. The concrete step of our filtration is presented in [Figure 1], and the characteristics of included articles are listed in [Figure 2]. We found that several survival outcome variants were used in eligible articles involving overall survival (OS), DFS,  metastasis-free survival (MFS),  and local recurrence-free survival (LRFS).  For easy to analysis, we defined DFS to include disease-free survival, MFS and LRFS.
Prognostic role of pathologic fracture in osteosarcoma
The heterogeneity has existed among these 8 studies for pooled HR with 95% CI of OS (P = 0.020, I 2 = 57.9%). However, the heterogeneity has not existed among 4 articles when combined HR with 95% CI of DFS. Thus, we have used random effect model and fixed effect model to calculate the pooled HR with 95% CI of OS and DFS, respectively. We found that the pooled hazard ratios (HRs) with 95% confidential interval (CI) of pathologic fracture in osteosarcoma for OS and DFS are 2.13 (1.43, 3.15) [Figure 3]a and 1.58 (1.11, 2.24) [Figure 3]b, respectively.
|Figure 3: The pooled prognostic role of pathologic fracture in osteosarcoma. A, the pooled HR with 95% CI of overall survival (OS) based on random effect model; B, the pooled HR with 95% CI of disease-free survival (DFS) based on fixed effect model|
Click here to view
Begg's test and Egger's test were used to evaluate the publication bias. The P value of Egger's test (P = 0.822) and Begg's funnel plot [Figure 4] did not detect any evidence of publication bias.
|Figure 4: Evaluation of the publication bias based on Begg's funnel plot|
Click here to view
| > Discussion|| |
In this study, we primarily conducted a meta-analysis to evaluate the relationships between pathologic fracture and osteosarcoma. For eliminating the between-study heterogeneity, we just included case/control studies, because these control groups were matched with case groups in terms of age, sex, tumor size at presentation, tumor location, and treatment and so on. In addition, this meta-analysis combined the 8 independent studies and could fill a gap of limitation of included subjects in each study. Finally, we discovered that osteosarcoma patients complicated by pathologic fracture have a poorer prognosis compared with osteosarcoma patients without pathologic fracture, either in OS or DFS.
There are several potential reasons that might explain the prognostic effect of pathologic fracture in osteosarcoma patients. First, researchers found that tumor size is a factor that might impact survival of osteosarcoma patients complicated by pathologic fracture. Bacci et al.  discovered that the tumor volume in patients with a pathologic fracture is higher than patients without pathologic fracture, even though the difference is not significant. Furthermore, they found that tumor volume is an independent negative survival predictor in osteosarcoma  in another study. In addition, the other researchers also found similar relationship between tumor size and poor prognosis in patients with a pathologic fracture. Kim et al.  demonstrated that the tumor size of patients with a pathologic fracture was significantly larger compared with patients without a pathologic fracture. Therefore, patients with a pathologic fracture usually have larger tumor size, which could partly explain the poorer prognosis of osteosarcoma complicated by pathologic fracture.
The local recurrence might be another reason. Based on a case/control study, researchers  found that pathologic fracture was a significant independent risk factor of local recurrence, and local recurrence was a multivariate risk factor for death. They further analyzed the correlation between the limb salvage ratio and local recurrence ratio. The operative stabilization of pathologic fracture by limb salvage surgery might increase the local recurrence ratio. However, other researchers  have not found any relationship between pathologic fracture and local recurrence.
However, the poor prognosis of patients with pathologic fracture might not be correlated with chemotherapy response. Scully and co-workers  discovered that fractures that heal after chemotherapy are associated with a significantly higher ratio of tumor necrosis than those that do not. But they did not found a significant correlation between fracture union per se and prognosis of osteosarcoma patients. Furthermore, the other three clinical control studies ,, independently obtained a similar conclusion that there is no significant difference in chemotherapy response between patients with pathologic fracture and without pathologic fracture.
In addition, the other clinical characteristics should be considered, such as tumor location and histological subtypes. And researchers  suggested that pathologic fracture previously occurred in telangiectatic subtype, and another study  reported that a better prognosis in telangiectatic subtypes than ordinary osteosarcoma, even though no significance.
| > Conclusion|| |
This study is a first mate analysis, which evaluates the prognostic effect of pathologic fracture in osteosarcoma. We demonstrated that patients with pathologic fracture have a worse survival outcome, including OS and DFS. And pathologic fracture might be a poor predictor of survival in osteosarcoma.
| > References|| |
Abudu A, Sferopoulos NK, Tillman RM, Carter SR, Grimer RJ. The surgical treatment and outcome of pathological fractures in localised osteosarcoma. J Bone Joint Surg Br 1996;78:694-8.
Jaffe N, Spears R, Eftekhari F, Robertson R, Cangir A, Takaue Y, et al
. Pathologic fracture in osteosarcoma: Impact of chemothearpy on primary tumor and survival. Cancer 1987;59:701-9.
Natarajan MV, Govardhan RH, Williams S, Raja Gopal TS. Limb salvage surgery for pathological fractures in osteosarcoma. Int Orthop 2000;24:170-2.
Bacci G, Longhi A, Versari M, Mercuri M, Briccoli A, Picci P. Prognostic factors for osteosarcoma of the extremity treated with neoadjuvant chemotherapy: 15-year experience in 789 patients treated at a single institution. Cancer 2006;106:1154-61.
Simon MA. Limb salvage for osteosarcoma. J Bone Joint Surg Am 1988;70:307-10.
Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003;327:557-60.
Zuo D, Zheng L, Sun W, Hua Y, Cai Z. Pathologic fracture does not influence prognosis in stage IIB osteosarcoma: A case-control study. World J Surg Oncol 2013;11:148.
Kim MS, Lee SY, Lee TR, Cho WH, Song WS, Cho SH, et al
. Prognostic effect of pathologic fracture in localized osteosarcoma: A cohort/case controlled study at a single institute. J Surg Oncol 2009;100:233-9.
Scully SP, Ghert MA, Zurakowski D, Thompson RC, Gebhardt MC. Pathologic fracture in osteosarcoma: Prognostic importance and treatment implications. J Bone Joint Surg Am 2002;84-A: 49-57.
Ferguson PC, McLaughlin CE, Griffin AM, Bell RS, Deheshi BM, Wunder JS. Clinical and functional outcomes of patients with a pathologic fracture in high-grade osteosarcoma. J Surg Oncol 2010;102:120-4.
Jeon DG. Prognostic effect of pathologic fracture in localized osteosarcoma. Asia Pac J Oncol Hematol 2010;2:131-5.
Qi WX, He AN, Tang LN, Shen Z, Yao Y. COX proportional-hazard regression analysis of prognostic factors in 52 adult patients with osteosarcoma. Tumor 2011;31:846-50.
Huang YJ, He AN, Qi WX, Sun YJ, Shen Z, Yao Y. Retrospective analysis of clinical features and prognostic factors for pathological fracture in 39 patients with extremity osteosarcoma. Tumor 2012;12:1015-20.
Tian C, Hang YJ, Tang LN, Lin F, Min DL, Yao Y. Retrospective analysis of clinical characteristics and prognostic factors for pathologic fracture in 44 patients with extremity osteosarcoma. Chinese-German J Clin Oncol 2014;1:35-41.
Bacci G, Ferrari S, Longhi A, Donati D, Manfrini M, Giacomini S, et al
. Nonmetastatic osteosarcoma of the extremity with pathologic fracture at presentation: Local and systemic control by amputation or limb salvage after preoperative chemotherapy. Acta Orthop Scand 2003;74:449-54.
Bramer JA, Abudu AA, Grimer RJ, Carter SR, Tillman RM. Do pathological fractures influence survival and local recurrence rate in bony sarcomas? Eur J Cancer 2007;43:1944-51.
Huvos AG, Rosen G, Bretsky SS, Butler A. Telangiectatic osteogenic sarcoma: A clinicopathologic study of 124 patients. Cancer 1982;49:1679-89.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]