1 Introduction

There are two types of cancers that occur in the bones: “metastatic bone tumors,' where cancer originating in other organs spreads to the bones, and “primary malignant bone tumors,' where cancer originates directly from the bones.^1,2 The latter is mainly comprised of tumors known as sarcomas. Primary bone sarcomas are a rare entity, constituting only 0.2% of all cancers.^3–5 Published data indicates that the five-year survival rate for patients with localized bone sarcoma ranges approximately between 40% and 55%.¹ It is documented that sarcomas arising from bone predominantly affect younger populations in contrast to soft tissue sarcomas, thereby presenting a significant societal challenge.^6,7

Standard therapeutic protocols for individuals diagnosed with localized bone sarcoma generally involve comprehensive surgical resection aimed at achieving oncological margins, frequently augmented by adjunctive radiotherapy or preoperative chemotherapy.^8,9 Securing a negative surgical margin remains the cornerstone goal in treating localized bone sarcoma. Moreover, there are instances where managing a positive macroscopic margin with subsequent radiotherapy is considered acceptable. Innovations in reconstructive surgery techniques and advancements in prosthetic technology have expanded the capabilities for repairing significant tissue defects post-tumor excision. This progress has subsequently broadened the patient criteria for limb-sparing procedures.¹⁰ Nevertheless, in scenarios where essential neurovascular structures or bones are implicated, limb-sparing surgery may not be a viable option, thereby requiring the consideration of amputation as a surgical intervention.

In specific cases of epithelioid carcinomas, the presence of tumor invasion into principal vascular or osseous structures is indicative of a more advanced cancer stage.^11–13 Furthermore, previous report indicates that patients with localized soft tissue sarcoma undergoing amputation surgery demonstrate reduced survival rate.¹⁴ Conversely, the effect of direct invasion into vital structures, including vessels and bones, on the survival outcomes of patients with localized bone sarcoma is still not definitively understood.

This investigation hypothesizes that individuals diagnosed with localized extremity bone sarcoma, when subjected to amputation, might exhibit less favorable survival outcomes as opposed to those who undergo limb-sparing treatments. Our analytical efforts leveraged a cohort derived from the Surveillance, Epidemiology, and End Results (SEER) registry. The primary aim was to explore the association between amputation for localized extremity bone sarcoma and survival rates, additionally seeking to pinpoint patient demographics especially impacted by survival discrepancies following amputation.

2 Methods

2.1

2.1 Patient population

The SEER program, encompassing data from 18 regional cancer registries, captures about 28% of the U.S. demographic, serving as an extensive databank for studying cancer epidemiology. Our study honed in on bone sarcoma cases identified between 2000 and 2019. These instances were pinpointed using the criteria for Rare Cancers as defined by the Surveillance of Rare Cancers in Europe (RARECARE), falling explicitly within the ‘51: Bone Sarcoma' category.¹⁵ By applying these selection criteria, our research covers a wide and representative cohort of bone sarcoma patients, allowing for a precise evaluation of survival impacts stemming from amputation compared to limb-sparing treatments. Our focus was specifically on primary tumor sites located in the extremities, as categorized by the International Classification of Diseases for Oncology (ICD-O codes C40.0–C40.3, C40.8, C40.9). The SEER database utilizes an original histological grading system, which we converted to Federation Nationale des Centres de Lutte le Cancer (FNCLCC) grades,¹⁶ following previous reports^17–19: The FNCLCC grading system aligns directly with the SEER grading framework, where FNCLCC Grade 1 corresponds to SEER Grade 1, FNCLCC Grade 2 to SEER Grade 2, and FNCLCC Grade 3 matches SEER Grade 3. Within the SEER grading methodology, Grade 1 tumors are classified as low grade. Conversely, tumor grades ranging from 2 to 4 are categorized under the high-grade classification. By employing SEER × Stat software (version 8.3.9.1, developed by the National Cancer Institute in Bethesda, MD), our study performed data extraction to identify an initial cohort of 7271 patients diagnosed with bone sarcoma. This cohort was narrowed down by eliminating cases with metastatic disease (1290 cases), involvement of lymph nodes (322 cases), cancers that were not staged (350 cases), or deaths attributed to undetermined causes (29 cases), resulting in a refined count of 5280 patients with localized bone sarcoma. An additional step excluded 1515 patients who did not have documented amputation or limb-sparing surgeries, yielding a final study group of 3765 patients (Fig. 1).

Close

Fig. 1 A diagram detailing the selection process for patients in this study, represented through a straightforward flowchart.

Export to PPT

3 Result

3.2

3.2 Impact of amputation on survival in localized extremity bone sarcoma

After adjusting for demographic and clinical variables through propensity score matching, the analysis encompassed 1698 patients (849 in the Amputation cohort and an equal number in the Limb-sparing group). Post-adjustment, the Amputation cohort showed a significant decrease in CSS, with a HR of 1.28 and a 95% Confidence Interval (CI) of 1.05–1.55, as shown in Fig. 2A. Similarly, a diminished OS was observed in the Amputation cohort, with an HR of 1.37 and a 95% CI of 1.15–1.63, detailed in Fig. 2B.

Close

Fig. 2 Graphs showing the differences in cancer-specific survival (A) and overall survival (B) among surgical groups, adjusted for similarities in patient characteristics through propensity score matching.

Export to PPT

3.3

3.3 Subgroup evaluations of survival based on tumor resection

Our data analysis revealed that overall, patients subjected to amputation for localized extremity bone sarcoma displayed inferior survival outcomes. In patients diagnosed with high-grade sarcomas, a noticeable reduction in OS was observed following amputation when compared to patients who underwent surgery aimed at conserving the limb. However, this observation did not extend to CSS, where the differences did not reach statistical significance (CSS: HR = 1.20, 95% Confidence Interval [CI]: 0.96–1.50; OS: HR = 1.26, 95% CI: 1.04–1.53), as illustrated in Fig. 3A and B. In contrast, for individuals with low-grade sarcomas, the survival rates post-amputation did not significantly deviate from those observed in patients who underwent limb-sparing surgeries (CSS: HR = 1.56, 95% CI: 0.49–4.95; OS: HR = 1.54, 95% CI: 0.74–3.17), as shown in Fig. 3C and D.

Close

Fig. 3 Survival curves for cancer-specific (A, C) and overall (B, D) survival based on tumor grade after adjusting for patient characteristics, with high-grade tumors shown in A and B, and low-grade tumors in C and D.

Export to PPT

Further examinations into tumor size effects demonstrated that for patients with tumors exceeding 8 cm, amputation was linked to reduced survival rates (CSS: HR = 1.32, 95% CI: 1.02–1.71; OS: HR = 1.39, 95% CI: 1.09–1.76), depicted in Fig. 4A–E. Similarly, individuals with tumors measuring 8 cm or smaller saw a decrease in OS, while CSS remained unaffected by amputation (CSS: HR = 1.31, 95% CI: 0.90–1.90; OS: HR = 1.50, 95% CI: 1.11–2.03), as presented in Fig. 4C–D.

Close

Fig. 4 Survival analysis curves for different tumor sizes post-adjustment, highlighting cancer-specific (A, C) and overall survival (B, D) for tumors larger than 8 cm (A, B) and up to 8 cm (C, D).

Export to PPT

Analysis focusing on tumor location indicated that amputation of tumors in the upper extremities resulted in lower OS but did not significantly impact CSS when compared to limb-sparing procedures (CSS: HR = 1.31, 95% CI: 0.85–2.51; OS: HR = 1.50, 95% CI: 1.11–2.42), as evidenced in Fig. 5A–B. This trend was also noted for tumors in the lower extremities, with noted decreases in survival rates (CSS: HR = 1.25, 95% CI: 1.01–1.54; OS: HR = 1.33, 95% CI: 1.11–1.61), detailed in Fig. 5C–D.

Close

Fig. 5 Depiction of survival outcomes, split by the tumor's location, with cancer-specific (A, B) and overall survival (C, D) outcomes for tumors in the upper (A, C) and lower extremities (B, D).

Export to PPT

3.4

3.4 Histological subtype and survival

Our findings highlighted the significant role of histological subtypes on the survival rates of patients. Individuals with osteogenic sarcoma faced a marked decline in OS though CSS was not significantly impacted (CSS: HR = 1.25, 95% CI: 0.97–1.61; OS: HR = 1.36, 95% CI: 1.08–1.71), as shown in Fig. 6A–D. Conversely, survival rates for patients with chondrogenic sarcoma and Ewing sarcoma were not markedly affected, indicating no significant survival disadvantage (Chondrogenic Sarcoma – CSS: HR = 1.38, 95% CI: 0.88–2.16; OS: HR = 1.36, 95% CI: 0.96–1.92; Ewing Sarcoma – CSS: HR = 1.77, 95% CI: 0.82–3.85; OS: HR = 1.87, 95% CI: 0.87–4.04), as demonstrated in Fig. 6B, C, E, and F.

Close

Fig. 6 Curves illustrating survival based on the type of bone sarcoma, with cancer-specific (A–C) and overall survival (D–F) for Osteogenic sarcoma (A, D), Chondrogenic sarcoma (B, E), and Ewing sarcoma (C, F), after matching for patient similarities.

Export to PPT

4 Discussion

Our study's key conclusions highlight the following: (1) Around 25% of the patients in our study population found themselves in a situation where choosing amputation as a treatment option was unavoidable.; (2) The choice of amputation, often deemed unavoidable, was more prevalent among male patients and those with tumors identified as high-grade; (3) Individuals with localized extremity bone sarcoma who underwent amputation exhibited poorer survival rates; and (4) Those with tumors larger than 8 cm or located in the lower limbs particularly experienced diminished survival outcomes. In our study, it's crucial to acknowledge that variables such as the Eastern Cooperative Oncology Group (ECOG) performance status and the specifics of treatment protocols were not incorporated into our analysis. This omission could impact the clarity and applicability of our results. Thus, further investigations are necessary to validate our results and investigate these relationships in more depth.

Scholarly discussions regarding the impact of amputation compared to limb-sparing surgery on survival outcomes in individuals with localized malignant bone tumors show considerable diversity. K. Godley et al. reported findings suggesting that survival outcomes post-amputation were similar to those seen with limb-sparing surgeries. Their study, which took a retrospective, population-based approach, evaluated 84 patients who had undergone pathological fractures in the femur as a result of primary bone sarcoma.²⁰ Moreover, findings from a single-institution study focusing on bone tumors indicated that amputation surgery did not significantly affect the overall survival of patients diagnosed with primary bone sarcoma and associated pathological fractures.^21,22 Conversely, various descriptive epidemiological studies, inclusive of meta-analyses based on histological subclassifications, have suggested that patients requiring amputation surgery face reduced survival durations.^5,23,24 Utilizing the SEER database, the most extensive public cancer data repository, our research incorporated a comprehensive analysis of 3765 patients with primary bone sarcoma. It's worth noting that earlier studies failing to demonstrate a survival disadvantage with amputation might not have had adequate statistical power for thorough analysis.

Our focused analysis revealed that amputation, when it became an unavoidable choice, significantly adversely affected survival, particularly for tumors larger than 8 cm. Utilizing propensity score matching to adjust for baseline discrepancies in patient demographics revealed a distinct decrease in survival among individuals with localized extremity bone sarcoma who received amputation. This indicates that localized tumors that invade crucial structures may considerably affect survival rates. For epithelioid tumors, along with hepatocellular carcinomas ¹¹and breast cancers¹³ that infiltrate essential anatomical features, such malignancies are classified into a more advanced stage. While additional research is necessary, these preliminary findings indicate that patients with localized bone sarcoma, who exhibit invasion into vital anatomical structures, may require classification into a higher stage of cancer. This reevaluation could lead to more tailored treatment strategies and potentially improve prognostic accuracy.

Interpreting the outcomes of this investigation necessitates acknowledging its inherent limitations. Firstly, the SEER database does not furnish exhaustive details on specific patient treatments, including comorbidities, the status of surgical margins, and ECOG performance status, which may introduce bias into our findings. It is plausible that patients selected for amputation might have been battling more aggressive disease manifestations or severe complications. Secondly, the granularity regarding chemotherapy and radiation therapy within the SEER database is limited. This lack of detail means that some patients in the Amputation cohort could have been candidates for neoadjuvant chemotherapy that ultimately did not yield the anticipated effectiveness, necessitating the choice of amputation. Moreover, the determination to proceed with limb-sparing surgery or amputation is inherently a clinical decision made by the treating oncologist, with criteria for reconstructive surgery to remedy soft tissue deficits potentially varying across different institutions. This variation can influence the surgical options presented to patients. Finally, SEER database represents the ‘initial course of treatment' outcome, which includes scenarios where initial attempts at limb-sparing surgery were unsuccessful, leading to amputation, unless such amputation occurred due to a later recurrence. These considerations underscore areas ripe for future inquiry to deepen our understanding and the overall completeness of related studies. Despite these constraints, we assert that the knowledge obtained from our study provides significant contributions to predicting outcomes for patients with primary malignant bone tumors.

In conclusion, our research supports the hypothesis that amputation surgery, when it becomes an unavoidable course of action, adversely affects survival rates in patients with localized extremity bone sarcoma. However, caution must be exercised in interpreting these results due to the omission of factors like ECOG performance status and specific treatment approaches, introducing the possibility of selection bias. Future investigations should aim for randomized clinical trials that cover detailed treatment protocols and performance status indicators to validate our findings further. This future research is crucial for not only confirming our conclusions but also potentially informing treatment plans for individuals with primary malignant bone tumors.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

5 Guardian/patient's consent

N/A.

6 Institutional review board approval

In the course of our study, we employed de-identified data sets sourced from the SEER database, each tagged with a distinct identifier to aid in our analysis. Considering the anonymized nature of the data and in line with the current regional regulations and institutional norms for conducting research on human subjects, it was determined that the need for ethical consent was not applicable.

CRediT authorship contribution statement

Nayuhito Yanagisawa: assisted with data interpretation and revised the manuscript for important intellectual content. Masatake Matsuoka: was involved in the design of the study, performed the clinical assessment, analysis, and interpretation of data, and drafted and revised the manuscript. Tomohiro Onodera: assisted with data interpretation and revised the manuscript for important intellectual content. Koji Iwasaki: were involved in data acquisition and revised the manuscript critically for important intellectual content. Masanari Hamasaki: were involved in data acquisition and revised the manuscript critically for important intellectual content. Taku Ebata: were involved in data acquisition and revised the manuscript critically for important intellectual content. Yoshiaki Hosokawa: were involved in data acquisition and revised the manuscript critically for important intellectual content. Eiji Kondo: were involved in data acquisition and revised the manuscript critically for important intellectual content. Norimasa Iwasaki: were involved in data acquisition and revised the manuscript critically for important intellectual content, All authors have read and approved the final manuscript.