In this study, we compared 61 patients with AJCC stage III STS who survived for more than 5 years with 36 patients who died of the disease within 5 years. Lower ASA scores, smaller tumor sizes, and lower histological tumor grades were each independently associated with better outcomes. However, these risk factors did not preclude long-term survival, as 22 of 43 patients (51.2%) with tumor sizes larger than 10 cm and 29 of 55 (52.7%) patients with grade 3 tumors were long-term survivors.
The tumor size is subdivided and emphasized in the new 8th AJCC STS staging system. Grade 2 or 3 sarcomas larger than 5 cm are classified as stage III and those larger than 10 cm are subclassified as stage IIIB. Likewise, in the present study, tumors larger than 10 cm were associated with poor long-term survival. Contrary to the new AJCC staging system in which grade 2 or 3 does not alter the tumor stage, grade 3 was significantly associated with poor survival in this study.
The ASA score, ranging from 1 to 6, was developed to predict the operative risk of patients with certain physiological statuses.[10] ASA 1 stands for a group of healthy patients without any systemic disease including hypertension and diabetes mellitus, among others. Many previous studies[14, 15] have suggested a relationship between the ASA score and oncological outcomes, including those of STS. Likewise, in this study, low ASA scores were significantly associated with long-term survival. This finding suggests that the individual’s general physiological status can independently affect the oncological outcome.
The actual survival rate in our study was 62.9% (61 of 97). The estimated survival rate of patients with AJCC stage III STS was 50% in a previous study.[8] However, this previous study was based on the 7th AJCC staging system. In a relatively new study[16] using the 8th AJCC staging system, the disease-specific survival rates for stage IIIA and IIIB STS were 77% and 62%, respectively, which were higher than the findings of this study (72% and 51%, respectively). The authors expected the actual survival rate to be somewhat lower than the estimated survival rate since the Kaplan–Meier method tends to overestimate the survival probability for cancers with poor prognosis.[9] However, in our study, these were very similar (62.9% vs. 62.8%). In retrospect, the prognosis of patients with stage III STS may not be as poor as that implied by the exaggerated estimated survival rate, unlike those corresponding to pancreatic cancer, hepatocellular carcinoma, and adrenocortical carcinoma.[9, 17, 18]
Leiomyosarcoma and malignant peripheral nerve sheath tumor (MPNST) were poor prognostic histology according to a previous study regarding prognostic factors of STS.[4] We observed a similar tendency in 7 of 10 patients with leiomyosarcoma and 4 of 7 patients with MPNST who were short-term survivors. In contrast, 9 of 10 patients with myxofibrosarcoma survived for more than 5 years in our study. Previous studies[19, 20] have shown better prognosis of myxofibrosarcoma in comparison to other types of sarcomas, although with higher local recurrence rates.
The administration of adjuvant radiotherapy was associated with long-term survival in the univariate analysis. Fifteen of the 36 short-term survivors did not undergo adjuvant radiotherapy; 4 presented with early distant metastasis, 3 had a poor general condition, and one of them had prolonged wound-healing problem. The remaining 7 patients did not receive adjuvant radiotherapy, as decided by radiation oncologists considering the tumor histology. The survival advantage of adjuvant radiotherapy, in part, reflects the difference in tumor biology apart from the benefit of radiotherapy itself.[21]
Unplanned excision of STSs leads to poor outcomes compared to those of planned excision.[22, 23] It is without doubt that unplanned excision causes high morbidity and may lead to poor functional outcomes; however, its relation with disease-specific survival remains unclear. A previous study[24] has shown non-inferior oncological outcomes in patients who underwent aggressive re-excisions after unplanned excision of stage III STS. No conclusion could be made in our study because only a small number of patients with stage III STS (n = 6) were referred to our institute after unplanned excision. However, five of six patients who underwent unplanned excisions survived for more than 5 years in our study. The effect of unplanned excision on oncological outcomes in patients with AJCC stage III STS remains to be validated.
While all of the short-term survivors developed tumor recurrence in our study, only about 10% of the long-term survivors developed recurrence. Comparing the recurrence timing between the two groups, early recurrence within a year after resection was more common in the short-term survival group (67% vs. 14%). This finding may imply the importance of tumor surveillance, especially during the first year after resection, since early recurrence may lead to poor survival.[25] However, instead of early recurrence, micrometastasis may already have been present at the time of initial surgery, thus leading to early recurrence and poor survival outcomes.[26] Developing this idea, patients exhibiting factors of “poor” prognosis may benefit more from adjuvant chemotherapy, the only effective treatment against micrometastasis, because these patients may already have micrometastasis before resection. The higher rate of metastatic recurrence in the short-term survivor group in our study supports this hypothesis. Selection of these patients with “poor” prognosis for adjuvant chemotherapy remains to be studied in the future.
Surgical resection after tumor recurrence may improve the survival rate in patients with stage III STS according to our study. Although many other factors including individual comorbidities, age, tumor location, and the presence of distant metastasis must be considered in deciding whether to perform surgical resection of recurred tumors, our finding suggests that long-term survival is possible with adequate treatment including surgical resection even in cases of recurrence. More studies with larger patient groups must be conducted in the future to verify this finding.
The results of our study must be carefully interpreted considering some limitations. First, we performed a retrospective study of patients treated at a single tertiary referral hospital. Future validations are needed using external databases in a prospective setting. Second, the drop-out rate might have been higher in the short-term survivor group because the deceased patients may have discontinued visiting the hospital without notice. For the purpose of our study, the authors excluded the censored data from the analysis, and this may have caused an overestimation of the actual survival rate.