Clinicopathological characteristics and prognostic factors in axial chondroblastomas: a retrospective analysis of 61 cases and comparison with extra-axial chondroblastomas

A total of 132 patients were included in the study, comprising 61 patients with ACB and 71 patients with EACB. Recently, our research group communicated the patient characteristics [12]. Comprehensive patient and tumor characteristics, treatment history, and clinical outcome data were obtained from the patients’ medical records. The collected clinical information encompassed patient demographics (age and sex), clinical features (including duration of symptoms and preoperative/postoperative sensorimotor status), and treatment modalities (type of surgery and adjuvant radiotherapy). Preoperative magnetic resonance imaging was employed to evaluate the extent of tumor invasion into the surrounding tissues. The pathological diagnosis was independently confirmed by two neuropathologists based on examination of hematoxylin and eosin–stained sections and pathology findings, including the presence of secondary aneurysmal bone cyst (ABC) and chicken-wire calcification. The primary events of interest were LRFS and OS. LRFS was defined as the duration between tumor resection and the first recurrence, while OS referred to the time from surgical resection to the patient’s death from any cause [12]. The type of surgical resection was determined based on a previously reported method [13], distinguishing between wide resection (such as gross total or en bloc resection with negative margins) and non-wide resection (including intralesional or marginal resection).

Immunohistochemical analysis was conducted following the previously described protocol [12]. In brief, 4-μm paraffin-embedded tissue sections were deparaffinized in xylene and gradually rehydrated using a series of graded ethanol solutions, followed by rinsing in distilled water. Antigen retrieval and blocking were performed, after which the sections were incubated overnight at 4 °C with primary antibodies: anti-vimentin [Vim] (Abcam company) at a dilution of 1:400, anti-S100 (Abcam company) at a dilution of 1:100, and anti-cytokeratin [CK] (Abcam company) at a dilution of 1:20. Subsequently, the sections were treated with secondary biotinylated antibodies against rabbit or mouse immunoglobulins, followed by incubation with streptavidin-peroxidase conjugate (Auragene, Changsha, Hunan, China). Visualization was achieved using a solution of 3,3-diaminobenzidine, and the sections were counterstained with hematoxylin for preservation.

The immunohistochemical staining results were independently evaluated by two pathologists. Positive expression for S100, Vim, and CK was determined by the presence of yellow or brownish-yellow granules at the corresponding site. For each section, five randomly selected high-magnification fields were observed. The staining proportion in each field was quantified, and the mean value was calculated. The expression level of each immunohistochemical marker was assessed in the hematoxylin and eosin (HE) sections and categorized as absent (0), rare (1), moderate (2), or significant (3) according to a previously established method [14]. Tissue samples were considered negative if a score of 0–1 was observed and positive if the score was higher.

The X-tile software version 3.6.1 (https://​medicine.​yale.​edu/​lab/​rimm/​research/​software.​aspx) was utilized to determine the threshold values for age, duration of symptoms, and tumor size in the survival analysis, with OS as the outcome parameter [15]. The threshold point represents the value that yields the minimum p-value from the corrected log-rank test [16]. Based on this threshold, patients were categorized into two subgroups: those with values less than or equal to the cutoff, and those with values greater than the cutoff. The cutoff point was specifically defined as the value that produced the minimum p-value from the log-rank test, which was duly corrected [17]. Statistical analyses were performed using the SPSS 26.0 software (SPSS, IBM, Armonk, NY). Descriptive statistics were presented as mean ± standard deviation, and comparisons were conducted using t-tests or ANOVA for continuous data. Categorical data were expressed as frequencies or proportions, and the chi-square test was employed for statistical analysis. Univariate Kaplan–Meier curves and log-rank tests were employed for survival analysis, exploring the associations between clinicopathological parameters and patient outcomes. A multivariate Cox proportional hazards model was employed to identify independent risk factors for LRFS and OS. Only variables that demonstrated statistical significance in the univariate survival analysis were included in the multivariate analysis. All hypothesis tests were two-sided, and a significance level of P < 0.05 was applied to determine statistical significance.

A total of 132 patients diagnosed with CB were enrolled in this study, with 61 patients diagnosing ACB and 71 patients diagnosing EACB (Fig. 1). The patient characteristics are summarized in Table 1. Significant differences were observed between ACB and EACB patients in terms of age, tumor size, surrounding tissue invasion, preoperative sensory or motor dysfunction, and Vim expression (P < 0.001, P < 0.001, P < 0.001, P = 0.021, and P = 0.002, respectively). All patients underwent surgery, with 65 patients receiving wide resections and 67 patients receiving non-wide resections. None of the patients received chemotherapy. Postoperative adjuvant photon radiotherapy was administered to 34 patients. For the survival analysis of OS, age, tumor size, and symptom duration were used as cutoffs for subgroups in patients with ACB and EACB, as shown in Additional file 1: Supplemental Digital Content 1 and Supplemental Digital Content 2. Representative images of immunohistochemical markers can be seen in Fig. 2.

When comparing the ACB and EACB cohorts, significant differences were observed between the two groups. Patients with ACB were found to be younger compared to those with EACB (P < 0.001, Table 1). Additionally, ACB patients had larger tumor sizes than EACB patients (P < 0.001, Table 1). Moreover, ACB patients exhibited a higher proportion of surrounding tissue invasion and a higher incidence of preoperative neuromotor dysfunction (P < 0.001 and P = 0.021, respectively, Table 1). Furthermore, ACB patients showed a higher frequency of high expression of vimentin (Vim) (P = 0.002, Table 1).

The univariate Kaplan–Meier analysis revealed significant associations between several factors and LRFS and OS. Specifically, the type of resection and the presence of chicken-wire calcification were found to be significantly associated with LRFS (P < 0.001 and P = 0.001, respectively, Additional file 1: Supplemental Digital Content 3 and Fig. 3). Patients who underwent wide resection and exhibited chicken-wire calcification had better LRFS outcomes. Moreover, surrounding tissue invasion, type of resection, and chicken-wire calcification significantly influenced OS (P = 0.001, P = 0.044, and P = 0.017, respectively, Additional file 1: Supplemental Digital Content 3 and Fig. 4). Patients without surrounding tissue invasion, those who underwent wide resection, and those with chicken-wire calcification had longer OS, indicating a more favorable prognosis. Subsequently, in the multivariate Cox analysis, the type of resection and the presence of chicken-wire calcification emerged as independent predictors of LRFS (P = 0.002 and P = 0.003, respectively, Table 2), while the type of resection alone could independently predict OS (P = 0.027, Table 2).

The univariate Kaplan–Meier analysis showed that the type of resection, adjuvant radiotherapy, and surrounding tissue invasion were associated with LRFS (P = 0.003, P = 0.013, and P = 0.032, respectively, Additional file 1: Supplemental Digital Content 4 and Fig. 5). Patients who underwent wide resection had better LRFS, while those with surrounding tissue invasion and without adjuvant radiotherapy had poorer LRFS. Additionally, the type of resection, surrounding tissue invasion, adjuvant radiotherapy, and postoperative sensory or motor dysfunction were associated with patient OS (P = 0.003, P = 0.016, P = 0.014, and P = 0.014, respectively, Additional file 1: Supplemental Digital Content 4 and Fig. 6). Patients who underwent not-wide resection, had surrounding tissue invasion, did not receive radiotherapy, and experienced postoperative sensory or motor dysfunction had shorter OS, indicating a worse prognosis. In the multivariate Cox regression model, it was found that the type of resection and surrounding tissue invasion could independently predict LRFS (P = 0.019 and P = 0.041, respectively, Table 3), while the type of resection alone could independently predict OS (P = 0.039 and P = 0.032, respectively, Table 3).

In this study, we conducted a comprehensive analysis of the largest cohort of ACB patients and examined the relationship between clinicopathological characteristics and patient survival. Additionally, we compared the patient characteristics and prognostic factors between ACB and EACB patients. Our findings revealed several noteworthy observations. Firstly, ACB patients exhibited larger ages and tumor sizes compared to EACB patients. Furthermore, the incidence of surrounding tissue invasion and postoperative sensory or motor dysfunction was higher among ACB patients. Notably, high expression of Vim was more commonly observed in ACB patients. Regarding survival outcomes, the type of resection was associated with LRFS in both ACB and EACB cohorts, while the type of resection and surrounding tissue invasion were associated with OS. However, it is important to note that most other factors demonstrated inconsistent survival outcomes between the two groups. These findings suggest that ACB may possess distinct molecular and clinical characteristics compared to EACB. This comprehensive understanding of prognostic factors in ACB allows for the implementation of reasonable prognostic risk stratification, ultimately leading to improved patient survival.

This study aimed to compare the patient characteristics and prognostic patterns between ACB and EACB, shedding light on their biological differences. While most parameters exhibited similar expression in ACB and EACB, the expression of Vim was found to be higher in ACB. Vim is a major intermediate filament protein in mesenchymal cells and is associated with accelerated growth, infiltration, and poor prognosis in various cancers [18‐21]. Based on this finding, we hypothesized that ACB may exhibit greater biological aggressiveness and a higher recurrence rate compared to EACB. This aligns with previous reports suggesting that spinal CB is more aggressive and prone to recurrence than bone CB in extremities [9‐11]. Consistent with our hypothesis, we observed a larger tumor size and a higher incidence of surrounding tissue invasion in ACB patients, which indicate tumor aggressiveness and poor prognosis [22, 23]. Moreover, ACB patients were more likely to experience sensory or motor dysfunction, which can be attributed to the proximity of ACB tumors to neurovasculature in the spine and skull, increasing the risk of nerve damage compared to EACB. Additionally, we noted that the average age of ACB patients was higher than that of EACB patients. Interestingly, literature reports indicate that CB primarily affects individuals under 50 years of age, with a peak incidence in the 20–30 age range [5, 24], while cranial CB patients tend to be around 40 years old [7]. Furthermore, CB commonly occurs at the ends of long bones in young patients [4], whereas in older individuals, the tumor can arise at various sites, including the craniofacial skeleton [24, 25]. This could explain the higher average age observed in ACB patients. However, further comparative analysis with large sample data is warranted for future research.

Given the aggressive nature of CB, surgical intervention is considered essential for treatment [5, 10, 26]. Our study supports this notion by demonstrating that wide tumor resection leads to favorable LRFS outcomes in patients, consistent with previous findings [5, 10, 26]. It is widely recommended by experts to aim for the complete removal of tumor tissue during surgery to minimize the risk of postoperative recurrence and achieve optimal disease control [5, 10, 26]. This finding is further supported by a recent study specifically focusing on spinal CB [27]. Notably, it has been reported that the presence of residual lesions after surgery increases the likelihood of tumor recurrence [7]. Furthermore, our analysis revealed that patients with surrounding tissue invasion experienced shorter OS, which aligns with previous reports. This observation can be easily understood, as extensive tumor infiltration or the proximity of the tumor to critical nerves, blood vessels, and other tissues can hinder achieving wide resection during surgery, thereby increasing the likelihood of postoperative recurrence in patients [7, 27]. Additionally, it is worth noting that prolonged infiltration and tissue damage caused by tumor growth, as well as the potential impact of surgical intervention on important neurovascular structures and organs, may contribute to the worsening of symptoms and a subsequent decline in the patients’ anti-tumor immune function. These factors further create conditions that facilitate tumor recurrence and subsequently lead to an increased recurrence rate [28‐30].

Chicken-wire calcification is a commonly observed feature in the eosinophilic mechanism of CB and can serve as a diagnostic tissue marker for CB [5, 31, 32]. Our study aligns with a previous integrated analysis, demonstrating that spinal CB patients expressing chicken-wire calcification have a more favorable prognosis [17]. Previous research has shown that patients with calcification in tumor tissue experience significantly longer median progression-free survival and overall survival compared to those without calcification [33]. Calcification primarily involves the deposition of calcium salts and minerals, and bone-bridging proteins play a role in regulating this process [34]. On the other hand, osteopontin has been found to promote malignant tumor invasion, growth, and metastasis [35]. Thus, we hypothesize that the downregulation of osteopontin expression in ACB tumors may contribute to their reduced aggressiveness. Interestingly, studies have even demonstrated that among different types of calcification present in CB tumor tissue, patients with chicken-wire calcification have a better prognosis than those with non-chicken-wire calcification [36]. This difference may be attributed to distinct biological behaviors resulting from the spatial arrangements of calcifications, warranting further investigation into these theories.

Another significant finding of this study revealed that EACB patients who underwent adjuvant radiotherapy had a poorer prognosis. This observation aligns with previous reports suggesting that radiotherapy may lead to the transformation of CB into a more malignant sarcoma [2, 7, 27]. In fact, some studies have even indicated that any form of adjuvant therapy should be avoided in CB patients [37]. On the contrary, radiotherapy has been shown to reduce tumor recurrence rates and can be beneficial for patients with postoperative recurrence or those deemed inoperable, leading to a favorable prognosis [9]. Consequently, the prognostic role of radiotherapy in CB remains controversial, and future studies involving larger sample sizes and detailed information on patient radiotherapy, combined with in vivo and in vitro experiments, are necessary to comprehensively evaluate the effects of adjuvant radiotherapy in CB patients. Current studies propose that radiation promotes epithelial-mesenchymal transition and induces the generation of new cancer stem cells from non-stem cells in various human cancers [38, 39]. This concept could serve as a theoretical foundation for the adverse prognosis observed in CB patients receiving adjuvant radiotherapy. Therefore, the detection and proteomic study of newly formed cancer stem cells may aid in identifying the precise mechanisms underlying disease progression in these CB patients.

In addition, numerous other factors significantly influence the prognosis of tumor patients, including preoperative frailty [40]. Frailty represents one of the most pressing global public health challenges, characterized by an accelerated decline in functional reserve associated with aging, rendering individuals more vulnerable following surgical procedures [41]. As a preoperative evaluation index, frailty has demonstrated remarkable prognostic prediction ability and risk stratification potential [41‐43]. Studies have confirmed the correlation between frailty and an elevated risk of complications, as well as increased postoperative mortality. Timely identification of frail patients allows for the mitigation of vulnerable areas, thereby reducing the occurrence of complications and promoting favorable outcomes [44]. Therefore, based on our research findings, we recommend the implementation of frailty assessments in a timely and accurate manner for CB patients requiring surgical treatment. Integrating this research variable in the future can effectively assist CB patients in avoiding potential risks and enhance our ability to guide their clinical management more effectively.

Although this study is retrospective in nature, it is important to acknowledge the need for future prospective studies with larger sample sizes and comprehensive data records to further validate the findings presented here.