We observed no difference in local recurrence and distant metastasis between the JR group and the JP group. Meanwhile, overall patient survival was also similar in both groups. These comparative outcomes suggested that joint-preserved limb sparing surgery with the aid of adjuvant co-treatment is not detrimental to local and systemic oncologic control when compared with joint-sacrificed limb sparing surgery. The major complications were similar in both groups. Given the high MSTS score and limb survival in the JP group, retaining native joint is preferable, particularly in children and adolescents whose growth plates are open. Furthermore, we identified marginal surgical margin of soft tissue was an independent risk factor in predicting local recurrence and found that limb diameter increase and disappearance of perivascular fat plane during neoadjuvant chemotherapy predict distant metastasis and poor overall survival.
There were limitations to our study. First, it was a retrospective study not allowing us to capture some important information but only what was listed in the medical records. For instance, we did not evaluate some known prognostic factors such as tumor necrosis rate due to its unavailability for some patients who underwent reconstruction with recycled tumor-bearing bone. Regarding histologic response to preoperative chemotherapy, tumor necrosis rate remains the most powerful treatment related prognostic factors[20]. However, it is generally available a week after surgery and preoperative evaluation thus depends upon clinical and radiographic assessment. Therefore, we introduced clinical and image evaluation as complement assessing chemotherapy response. The reason why we utilized MRI study observing perivascular space is that the intact fat will be a final barrier before tumor sheathing on the vascular side. Tumor enlargement has been regarded as a predictor for poor prognosis[21], therefore, limb diameter increase was a potential prognostic factor similar to tumor volume change. However, defining limb diameter increase using a cutoff of 10% as indictor of volume enlargement was sort of exploratory.
The second limitation to our study was introducing the concept of ablative margin. In attempt to preserve the joint with tumor involving epiphysis, we resected tumor through in situ-ablated tumor-bearing bone. There was no described margin for this surgical resection previously. Although pathologic examination of specimen revealed no live tumor at the osteotomy site, we could not obtain direct histologic evidence of no live tumor in the residual epiphysis[14]. Therefore, we admit the concept of ablative margin is path breaking attempt in nature due to the utilization of the in situ ablation technique and we hope that future researches will verify the efficacy of this surgical procedure.
Surgical margin has been a fundamental issue related to local recurrence[22]. In order to differentiate the influence of the different margin status on the safety of joint preservation, we separated the axial margin and longitudinal margin in this study. Axial margin is mainly relevant to the soft tissue. We found that marginal margin in the soft tissue was a risk prognostic factor for local recurrence in multivariate analysis, which was in agreement with previous studies[23]. However, whether joint end could be safely preserved is mostly determined by the extent of longitudinal resection margin of the bone. The crucial issue is how close we can approach tumor to save the nearby joint end safely. Most surgeons believed that epiphysis preservation should only be reserved for tumor at least 2cm away from the articular surface or tumor not invading the epiphysis[24]. Previous studies have revealed that the prevalence of transepiphyseal spread of the tumor is around 81%[1,2], suggesting only approximate 20% of osteosarcomas are eligible for joint preservation surgery, which have been verified in early limb-salvage reports[5]. In the current series, the prevalence of tumor invading epiphysis is 78.6%, which is similar to previously reported rates[1,2,5]. Differently, almost half of these patients with epiphyseal tumor involvement had their native joint preserved. This is mainly due to tumor resection via the ablated tumor-bearing bone. It subsequently raises the query: whether ablated surgical margin would lead to increase of local recurrence? We found there was no statistical difference in the effect of the wide and ablated margin on local recurrence. Therefore, ablated bone margin is adequate and comparable to wide bone margin in terms of its influence on local relapse.
Our second question was to determine prognostic factors for metastasis and Patient’s survival. In multivariate analysis, we found that limb diameter increase and perivascular fat plane disappearance during neoadjuvant chemotherapy were independent predictors of metastasis and poor survival. Regarding tumor volume change, it has received little attention as a viable prognostic factor because usually there is not a marked volumetric response during neoadjuvant chemotherapy[24]. Recently, some report suggested reduced or stable tumor size cannot guarantee a good response, increase in tumor volume is well correlated with a poor histologic response[25,26]. Therefore, this factor is regarded as one of the parameters that may predict poor histologic response to preoperative chemotherapy[27]. Our results are consistent with previous report in which the authors found that tumor enlargement after chemotherapy has a greater relative effect on survival than any other factor, including initial tumor size, surgical margin, and histologic response[28].
Conceptually, distant metastasis could be dependent on direct vascular or lymphatic pathways. The major vessels invasion would be a main culprit for subsequent metastasis and worse survival[29]. MRI can reveal gross encasement of a vessel readily, but it usually cannot differentiate mere contact, adherence or subtle invasion if no tissue plane is evident between the tumor and the vessels[30]. We are not fully certain whether tumor closely abutting major vessel represent possible metastatic path or seeding mechanism, however, gradual perivascular fat disappearance during chemotherapy was an independent prognostic factor in predicting metastasis and worse survival in the current study. Therefore, we suspect the disappearance of perivascular fat during chemotherapy may lead to progressive vessel or lymphatic invasions and subsequent tumor seeding by circulating tumor cells. However, we did not have histopathologic proof of this correlation and it should be under further investigation in the future.