Abstract
Background/Aim: Plasma D-dimer levels have been known to be associated with tumor progression; we, therefore, investigated whether the D-dimer levels during preoperative systemic chemotherapy can be prognostic indicators in patients with high-grade musculoskeletal sarcoma. Patients and Methods: We investigated 28 cases of high-grade sarcomas, and evaluated the utility of D-dimer levels for preoperative evaluation of the effects of systemic chemotherapy. Results: Among the candidate parameters determined based on D-dimer levels at several time-points of neoadjuvant chemotherapy and several oncological outcomes, the plasma D-dimer level completion of the second course of chemotherapy and the ratio of plasma D-dimer levels at completion of preoperative chemotherapy to the level of plasma D-dimer on referral, could significantly predict patient prognosis (p=0.049 and p=0.02, respectively). Conclusion: D-dimer level changes could be a helpful marker for preoperative evaluation of the effect of systemic chemotherapy in terms of prognosis prediction in high-grade musculoskeletal sarcoma patients.
Recent advances in systemic chemotherapy have improved the oncological outcomes of high-grade bone and soft tissue sarcomas (1-3), and systemic chemotherapy is now standard-treatment for this tumor entity. Most systemic chemotherapy protocols for bone and soft tissue sarcomas require high-dose chemotherapeutic agents such as methotrexate, cisplatin, adriamycin, and ifosfamide, that may lead to severe side-effects (4). Moreover, sensitivity to chemotherapy differs among cases, despite diagnosis and histological characteristics being similar (1, 2). If patients with low sensitivity are administered a particular regimen, treatment may not be effective. Thus, a useful and simple modality to confirm sensitivity to specific regimens is in urgent need.
Radiological evaluation of tumor shrinkage is a standard modality to evaluate the chemotherapy response in solid tumors. For this purpose, the revised Response Evaluation Criteria in Solid Tumors (RECIST) criteria have been published (5). Application of these criteria in bone and soft tissue tumors, however, has brought to surface several problems. For example, intra-cortical bone tumors refractory to systemic chemotherapy may not increase in size because the area occupied by the tumor itself remains unchanged. Moreover, it is difficult to evaluate the effects of chemotherapy by shrinkage status in high-grade soft tissue tumors, as these sometimes enlarge despite chemotherapy being effective, which is seen as central necrosis. The chemotherapy effect in these cases is instead represented by decreases in the gadolinium-enhanced area. Thus, discrepancies between the tumor shrinkage status and chemotherapy effect have been reported (6, 7). Currently, a combination of several modalities, including 201Tl and 99mTc-hexakis2-methoxyisobutylisonitrile scintigraphy, 18F-Fludeoxyglucose positron emission tomography/computed tomography, and diffusion-weighted magnetic resonance imaging, is recommended to predict the chemotherapy effect; however, some of these are not universally used and/or expensive (8, 9). Hence, a simple evaluation tool in addition to these radiological evaluation modalities is required.
Previously, we confirmed that the plasma D-dimer levels in bone and soft tissue tumors could reflect the tumor malignancy stage, and prognosis, and demonstrated a close relationship between the plasma D-dimer levels and malignancy (10), its usefulness in the prediction of oncological outcomes of musculoskeletal tumor treatment (11), and its significance in the differential diagnosis between large lipoma and well-differentiated liposarcoma (12). Herein, we hypothesized that changes in the D-dimer levels may be an indicator of the chemotherapy effect, as represented by the oncological outcome, in high-grade malignant bone and soft tissue tumors during treatment with systemic chemotherapy. To address this hypothesis, we analyzed the D-dimer levels at several time points during the treatment course in order to identify the plasma D-dimer level parameters that are useful for the prediction of overall survival.
Materials and Methods
This retrospective uncontrolled study employed data from medical records. Inclusion criteria were a pathological diagnosis of high-grade bone and soft tissue sarcoma; treated at our Institution between 2007 and 2013; application of preoperative systemic chemotherapy and surgical resection; at least 12 months of follow-up, except in cases of death within 12 months; and available clinical, pathological, and radiological data. Patients were excluded if any of the following were identified at presentation: pre-existing hyper-coagulopathy; recent anti-coagulant therapy, including prophylaxis of thromboembolic complications; recent trauma; inflammatory diseases; or another recent, major surgery.
Finally, 28 patients were included (17 males and 11 females; mean age=31 years). The bone tumors were osteosarcoma in 9 and Ewing sarcoma in 3 cases, and the soft tissue sarcomas were undifferentiated pleomorphic sarcoma in 5, liposarcoma in 4, epithelioid sarcoma in 2, and others in 5 cases. At the final follow-up, 11 patients had a continuous disease-free survival status, 1 was alive with disease, 5 had no evidence of disease, and 11 patients had died of disease.
The drugs and doses (total dose/course) of chemotherapy regimens were as follows (4): methotrexate 12 g/m2, doxorubicin 60 mg/m2, and cisplatin 120 mg/m2 (MAP); methotrexate 12 g/m2 (patients younger than 19 years) or 10 g/m2 (patients older than 20 years) (M); doxorubicin 60 mg/m2 and cisplatin 120 mg/m2 (AP); ifosfamide 10 g/m2 and doxorubicin 60 mg/m2 (AI); ifosfamide 15 g/m2 (I); vincristine 2 mg/m2 (maximal dose, 2 mg), doxorubicin 75 mg/m2, and cyclophosphamide 1,200 mg/m2 (VDC); and ifosfamide 9 g/m2 and etoposide 500 mg (IE) (1, 3, 13). For patients with osteosarcoma, the MAP, M, AP, AI, or I regimen was used (13); the MAP or AP regimen was routinely used at the initial stage of treatment. If radiological examination suggested disease progression after 2 or 3 courses of systemic chemotherapy, neoadjuvant chemotherapy was interrupted and wide resection or amputation was performed for local control. If postoperative pathological evaluation revealed that preoperative chemotherapy was not effective, a regimen including ifosfamide (AI or I) was introduced as adjuvant chemotherapy (2, 13).
For patients with Ewing's sarcoma, the VDC and IE regimens were used (3). If radiological examination suggested disease progression after 2 or 3 courses, neoadjuvant chemotherapy was interrupted and wide resection, amputation, or radiotherapy was performed for local control. The postoperative regimen was not changed even if the postoperative pathological evaluation revealed that the preoperative chemotherapy was not effective.
For patients with non-small round-cell soft tissue sarcoma, the AI or I regimen was used (1). If radiological examination suggested disease progression after 2 courses, neoadjuvant chemotherapy was interrupted and wide resection or amputation was performed. The postoperative regimen was not changed even if the postoperative pathological evaluation revealed that the preoperative chemotherapy was not effective.
Preoperative chemotherapy was interrupted in 7 cases because of progressive disease; these cases were classified as “interrupted.” The remaining 21 cases with completion of the intended courses of chemotherapy were classified as “completed”.
Radiological evaluation of the effect of chemotherapy was based on RECIST (5). Responses were classified as “partial response,” “stable disease,” or “progressive disease”. In the present study, patients with a “partial response” or “stable disease” were defined as good responders (GR) to chemotherapy, while those with “progressive disease” were defined as poor responders (PR). Histological evaluation of chemotherapy was based on the degree of necrosis: grades IV (100% necrosis), III (90-99% necrosis), II (50-89% necrosis), and I (0-49% necrosis) (14). Grades III and IV were defined as GR to chemotherapy; grades I and II were defined as PR. The plasma D-dimer level parameters were defined as follows: “DD on referral” was defined as the plasma D-dimer level on referral assessed before performing any intervention, including chemotherapy, radiotherapy, open biopsy, or tumor resection. “DD 1w” was defined as the plasma D-dimer level assessed 1 week from the start point of the first chemotherapy course. “DD post-2nd course” was defined as the plasma D-dimer level after completion of the second course of chemotherapy, at which time the decision to continue or interrupt chemotherapy was based on radiological findings. “DD preop” was defined as the plasma D-dimer levels at completion of preoperative chemotherapy. The DD ratio 1w, DD ratio 2nd course, and DD ratio preop were defined as DD 1w/DD on referral, DD post-2nd course/DD on referral, and DD preop/DD on referral, respectively. To measure the D-dimer levels, a latex agglutination assay (STA Liatest® D-Di; Roche Diagnostics AG, Rotkreuz, Zug, Switzerland) was performed using a STA-R® coagulation analyzer (Diagnostica Stago, Inc., Parsippany, New Jersey, USA), as previously described (10-12). Based on the sensitivity of this assay, levels <0.2 μg/mL were treated as 0.2 μg/mL.
The patients were divided into groups based on treatment outcomes, as follows: (i) chemotherapy completion (interrupted vs. completed), (ii) radiological evaluation of preoperative chemotherapy (PR vs. GR), (iii) histological evaluation of chemotherapy (PR vs. GR), and (iv) oncological outcome (dead vs. alive). The plasma D-dimer levels in each group were compared using Mann-Whitney U tests to extract the useful parameters for evaluation of treatment outcomes. The significant sets of treatment outcome and plasma D-dimer level parameters were subjected to receiver operating characteristic (ROC) curve analysis to establish the optimal cut-off values for each parameter. Finally, the usefulness of the cut-off values for overall survival analysis was evaluated using Kaplan-Meier curves and the Log-rank test. p-Values <0.05 were considered significant. Statistical analyses were performed with JMP (version 8; SAS Institute Inc., Cary, NC, USA).
The study was approved by the institutional review board of the authors' institution (authorization number 541).
Results
In order to select for prognostic factors, we first collected data of the plasma D-dimer levels at several time-points during preoperative chemotherapy (as described above). Moreover, we calculated the ratios of the D-dimer levels from different time-points under the hypothesis that effective chemotherapy can down-regulate the plasma D-dimer level, which is a marker of tumor progression (Table I). In order to extract the cut-off values for discriminating the prognostic events for each parameter, we set four categories of treatment outcomes: completion of chemotherapy, radiological evaluation, histological evaluation, and patient status. In seven cases, the regimen was not completed because of tumor progression at the completion of the second course, as assessed by radiological evaluation. Among the total 28 cases, 9 and 19 cases were evaluated as PR radiologically and histologically, respectively. Death occurred in 10 cases.
Summary of plasma D-dimer level parameters.
Using the above outcomes, each D-dimer-related parameter was evaluated (Table II). Among these, DD post-2nd course and DD preop differed significantly between the “interrupted” and “completed” groups (p=0.01 and p=0.03, respectively), the DD preop and DD ratio preop differed significantly between the PR and GR groups, radiologically (p=0.049 and p=0.02, respectively), and the DD ratio preop differed significantly between the PR and GR groups, histologically (p=0.006).
The above five parameters were considered useful in predicting prognosis and were subjected to ROC analyses to determine the optimal cut-off values (Table III). The cut-off value of DD post-2nd course in terms of protocol completion was 0.60, as were those of DD preop in terms of protocol completion and radiological evaluation. The cut-off values of the DD ratio preop for the radiological and histological evaluations were 1.15 and 1.33, respectively.
Finally, we confirmed the usefulness of these values in determining oncological outcomes preoperatively by performing survival analyses (Figure 1A-D). Among the candidate parameters, the cut-off values of the DD post-2nd course in terms of protocol completion and of the DD ratio preop in terms of the histological evaluation could significantly predict patient prognosis (p=0.049 and p=0.02, respectively).
D-dimer parameters and treatment outcomes.
Kaplan-Meier curves for the confirmation of usefulness of the cut-off values of plasma D-dimer (DD)-related parameters selected by the receiver operating characteristic curve analyses. Cut-offs of (A) DD of 0.60 after the 2nd course, (B) DD of 0.60 preoperatively, (C) DD ratio of 1.15 preoperatively, and (D) DD ratio of 1.33 preoperatively, were examined.
Results of the receiver operating characteristic curve analyses.
Discussion
Plasma D-dimer is a degradation product of fibrinolysis that is broadly used in screening for venous thromboembolism (10, 15). Several reports have suggested this parameter as a biomarker for tumor malignancy status, stage, and prognosis (16-24). Likewise, we previously established associations between the plasma D-dimer levels and malignancy/prognosis of malignant bone and soft tissue tumors. The pre- and postoperative D-dimer levels are significantly higher in malignant than in benign musculoskeletal tumors (10), and elevated D-dimer levels indicate poorer prognoses for patients with malignant musculoskeletal tumors (11). We also pointed-out the usefulness of this parameter in the differential diagnosis between lipoma and well-differentiated liposarcoma, two representative lipogenic tumors that present similar clinical, radiological, and histological characteristics (12). Furthermore, this parameter could be potentially used to evaluate the treatment effect (16, 20, 21, 25). Altiay et al. showed a significant association between the plasma D-dimer levels and chemotherapy response in locally advanced or metastatic lung cancer patients (20). Antoniou et al. detected the D-dimer plasma levels in advanced lung cancer patients before, during, and after chemotherapy, and found that a significant number of patients with complete or partial response showed a reduction in the D-dimer plasma values, whereas patients with progression or recurrent disease showed significantly increased values (21). Khoury et al. analyzed the changes in D-dimer levels in patients with refractory castration-resistant prostate cancer to assess their concordance during the treatment course and showed that increases in D-dimer levels were associated with increased risks of progressive disease (16). These studies showed not only the predictive ability of the pre-treatment D-dimer level for the treatment outcomes but also a significant relationship between the D-dimer level changes and the treatment effects.
In the present study, we collected several D-dimer-related parameters from various time-points during preoperative chemotherapy to predict the chemotherapy effect. First, D-dimer levels at four representative time-points, including the beginning of the preoperative chemotherapy, one week after the initial course, after the completion of the second course, and just before the operation, were analyzed, based on previous reports that changes in D-dimer levels before and after chemotherapy can predict treatment response (20, 21). Next, the changes in the D-dimer levels (DD ratios) were examined, based on the hypothesis that effective chemotherapy can down-regulate the D-dimer levels.
Theoretically, the oncological event examined in the ROC analysis should be patient status (dead or alive), because the survival analysis end-point in the present study was the overall survival. However, we failed to confirm a significant relation between the patient status and any D-dimer parameters (Table II). Nonetheless, although not significantly different, the mean DD preoperatively in surviving patients was unexpectedly higher than that in patients who had died. The patient status at a particular time-point might be changed during the follow-up because of the wide individual variation in the follow-up periods and the retrospective study design. Thus, we added several other end-points that enabled us to identify several significant candidate data sets for predicting treatment outcomes (Table II). Using these candidate data sets, we finally identified two significant parameters, a cut-off of 0.60 for the DD post-2nd course and a cut-off of 1.33 for the DD ratio preoperatively (Figure 1).
If the tumor did not respond to the protocol, in most cases, the protocol was interrupted and surgery was immediately performed. If our hypothesis that D-dimer levels represent tumor activity is true, early interruption of chemotherapy would enhance the differences in D-dimer-related parameters between cases with protocol completion and interruption. Hence, in estimating the clinical significance of the data sets in the present study, a certain bias by early interruption in refractory cases should be noted. Additionally, the limitations of this retrospective study, such as the considerable small sample size and wide heterogeneity of tumor types, regimen types, regimen completion status, and patient background factors, such as age, that may affect the D-dimer levels, should be noted. A prospective study with strict inclusion criteria in terms of the particular disease and regimen is warranted in the future.
In conclusion, D-dimer level changes may be a helpful marker for preoperative evaluation of the effect of systemic chemotherapy, in terms of prognosis prediction in high-grade musculoskeletal sarcoma patients.
Acknowledgements
This research was (partially) supported by the Practical Research for Innovative Cancer Control from Japan Agency For Medical Research and development, AMED.
Footnotes
Conflicts of Interest
The authors declare that they have no conflicts of interest.
- Received September 13, 2015.
- Revision received October 2, 2015.
- Accepted October 19, 2015.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
