Patient blood management is an evidence-based perioperative multidisciplinary and multimodal patients-specific team approach, which consists of three pillars that optimize the volume of blood transfusions [14, 15]. The preoperative and intraoperative steps of the second and third pillars include the identification of bleeding risks, assessment and optimization of the patient’s physiological reserve and risk factors, and meticulous hemostasis and surgical techniques. Because extra-abdominal high-grade STS are rare heterogeneous tumors with variable presentations, localizations, behaviors, and outcomes [16], it is challenging to implement the steps mentioned above of evidence-based knowledge. This was also demonstrated in the results of our retrospective analysis, which revealed a very high coefficient of C/T ratio and a high economic loss.

Therefore, we attempted to solve this problem by analyzing our clinical data to reveal the potential prognostic factors for blood transfusion and to create a risk prediction score model that can predict blood transfusion to patients with high-grade extra-abdominal STS with high probability. The TRANSAR score can be deployed anywhere and managed effortlessly.

The results of our study indicate that in most cases, cross-matching would be needed for patients with coronary heart disease, which could be associated with perioperative antiplatelet management strategy, including continuation until surgery, or was not stopped during surgery. In addition, this could also be associated with an increased volume of transfusion and blood loss. Moreover, an increased risk of re-operation because of postoperative bleeding, as well as an increased length of hospital stay [17,18,19]. In this study, the preoperative anemia in patients with STS varied between 20 and 30%. The cause of tumor-related anemia may depend on the dysfunction of iron metabolism, inadequate production of erythropoietin, reduced number of erythroid progenitor cells in the bone marrow, and the production of inflammatory cytokines [20, 21]. Additionally, low preoperative hemoglobin concentration could be explained by the tumor’s paraneoplastic effects or the influence of neoadjuvant chemotherapy as a part of multimodal oncological treatment [22, 23].

The tumor size and grading could be associated with higher blood loss during tumor resection because of possible hypervascularity, surgical challenges, and extended operating time [24, 25]. Our study demonstrates that neo-adjuvant radiation therapy or radiation therapy for a previous tumor can increase the risk of excessive blood transfusions and thus can be associated with fibrosis or soft tissue edema due to radiation toxicity, further complicating the surgical procedure [26, 27].

The findings of this study have to be interpreted in light of some limitations. The first limitation is the small size of the patient group and retrospective study design because of the rarity of these malignancies in the population. Because of these reasons, it was challenging to create a score that could predict the number of necessary blood units to be cross-matched. A multicenter prospective study utilizing this training model’s data and even the possibility of real-time constant model retraining could solve this problem. The second limitation concerns the differences in the expertise of operating surgeons, which can affect the amount of blood loss. Therefore, this study’s results and applications could apply only to high-volume sarcoma centers with surgical expertise in levels IV and V.

Thus, the TRANSAR score is the first attempt to design a prediction model for a blood transfusion by patients with extra-abdominal high-grade STS using a combination of demographic and clinical variables on admission. After summarizing all scores of five prognostic factors, the patients with a score of more than six points have a risk of being transfused more than 50%. However, the clinician makes the final decision to perform a cross-match.

TRANSAR score is a part of evidence-based medicine, which could provide tremendous and advantageous information for clinical practice in treating rare diseases. As the colleagues, Kohane IS et al. published in the New England Journal of Medicine, “biomedical research, data technologies, and clinical care all require resources, but the era of shifting more and more economic resources toward healthcare is going to end.” Therefore, in the future, there could be an increased focus on more efficient use of resources to deliver the best care to the patient at the lowest cost [28].