Limits...
Performance status is the most powerful risk factor for early death among patients with advanced soft tissue sarcoma: the European Organisation for Research and Treatment of Cancer-Soft Tissue and Bone Sarcoma Group (STBSG) and French Sarcoma Group (FSG) study.

Penel N, Glabbeke MV, Mathoulin-Pelissier S, Judson I, Sleijfer S, Bui B, Schoffski P, Ouali M, Marreaud S, Brouste V, Duhamel A, Hohenberger P, Blay JY - Br. J. Cancer (2011)

Bottom Line: The 90-day mortality rate was 8.6 and 4.5% in both cohorts.The CHAID analysis retained PS as a major discriminator followed by histological grade (only for patients with PS ≥2).In both models, PS was the most powerful PF for 90-day mortality.

View Article: PubMed Central - PubMed

Affiliation: Department of General Oncology, Centre Oscar Lambret, Regional Comprehensive Cancer Centre, 3, rue F Combemale, 59020, Lille, France. n-penel@o-lambret.fr

ABSTRACT

Background: We investigated prognostic factors (PFs) for 90-day mortality in a large cohort of advanced/metastatic soft tissue sarcoma (STS) patients treated with first-line chemotherapy.

Methods: The PFs were identified by both logistic regression analysis and probability tree analysis in patients captured in the Soft Tissue and Bone Sarcoma Group (STBSG) database (3002 patients). Scores derived from the logistic regression analysis and algorithms derived from probability tree analysis were subsequently validated in an independent study cohort from the French Sarcoma Group (FSG) database (404 patients).

Results: The 90-day mortality rate was 8.6 and 4.5% in both cohorts. The logistic regression analysis retained performance status (PS; odds ratio (OR)=3.83 if PS=1, OR=12.00 if PS ≥2), presence of liver metastasis (OR=2.37) and rare site metastasis (OR=2.00) as PFs for early death. The CHAID analysis retained PS as a major discriminator followed by histological grade (only for patients with PS ≥2). In both models, PS was the most powerful PF for 90-day mortality.

Conclusion: Performance status has to be taken into account in the design of further clinical trials and is one of the most important parameters to guide patient management. For those patients with poor PS, expected benefits from therapy should be weighed up carefully against the anticipated toxicities.

Show MeSH

Related in: MedlinePlus

CHAID algorithms. (A) STBSG data set and (B) FSG data set.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3101912&req=5

fig1: CHAID algorithms. (A) STBSG data set and (B) FSG data set.

Mentions: The CHAID analysis provided a very simple algorithm. In the decision tree, the most powerful discriminator (splitter) was the PS; three subsets of patients were discriminated with increasing risk of early death: patients with PS=0 (early death rate: 3.3%), patients with PS=1 (early death rate: 9.4%) and patients with PS ⩾2 (early death rate: 25.5%). There was no discriminator able to split the two first categories of patients. In the development data set, among patients with PS ⩾2, the histological grade was able to individualise two subsets of patients; when the grade was 3, the rate of early was 36.3% and in the other situations, the rate of early death was 19.5% (Figure 1).The area under the receiver operator curve was 0.67 (0.64–0.71). The optimal classification was based on the separation of patients with PS=(0–1) from other patients (Table 4). Using this classification, the prognostic accuracy was 86.2% (84.5–87.4), the positive predictive value was 25.3% (20.4–30.1) and the negative predictive value was 93.3% (92.0–94.6).


Performance status is the most powerful risk factor for early death among patients with advanced soft tissue sarcoma: the European Organisation for Research and Treatment of Cancer-Soft Tissue and Bone Sarcoma Group (STBSG) and French Sarcoma Group (FSG) study.

Penel N, Glabbeke MV, Mathoulin-Pelissier S, Judson I, Sleijfer S, Bui B, Schoffski P, Ouali M, Marreaud S, Brouste V, Duhamel A, Hohenberger P, Blay JY - Br. J. Cancer (2011)

CHAID algorithms. (A) STBSG data set and (B) FSG data set.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3101912&req=5

fig1: CHAID algorithms. (A) STBSG data set and (B) FSG data set.
Mentions: The CHAID analysis provided a very simple algorithm. In the decision tree, the most powerful discriminator (splitter) was the PS; three subsets of patients were discriminated with increasing risk of early death: patients with PS=0 (early death rate: 3.3%), patients with PS=1 (early death rate: 9.4%) and patients with PS ⩾2 (early death rate: 25.5%). There was no discriminator able to split the two first categories of patients. In the development data set, among patients with PS ⩾2, the histological grade was able to individualise two subsets of patients; when the grade was 3, the rate of early was 36.3% and in the other situations, the rate of early death was 19.5% (Figure 1).The area under the receiver operator curve was 0.67 (0.64–0.71). The optimal classification was based on the separation of patients with PS=(0–1) from other patients (Table 4). Using this classification, the prognostic accuracy was 86.2% (84.5–87.4), the positive predictive value was 25.3% (20.4–30.1) and the negative predictive value was 93.3% (92.0–94.6).

Bottom Line: The 90-day mortality rate was 8.6 and 4.5% in both cohorts.The CHAID analysis retained PS as a major discriminator followed by histological grade (only for patients with PS ≥2).In both models, PS was the most powerful PF for 90-day mortality.

View Article: PubMed Central - PubMed

Affiliation: Department of General Oncology, Centre Oscar Lambret, Regional Comprehensive Cancer Centre, 3, rue F Combemale, 59020, Lille, France. n-penel@o-lambret.fr

ABSTRACT

Background: We investigated prognostic factors (PFs) for 90-day mortality in a large cohort of advanced/metastatic soft tissue sarcoma (STS) patients treated with first-line chemotherapy.

Methods: The PFs were identified by both logistic regression analysis and probability tree analysis in patients captured in the Soft Tissue and Bone Sarcoma Group (STBSG) database (3002 patients). Scores derived from the logistic regression analysis and algorithms derived from probability tree analysis were subsequently validated in an independent study cohort from the French Sarcoma Group (FSG) database (404 patients).

Results: The 90-day mortality rate was 8.6 and 4.5% in both cohorts. The logistic regression analysis retained performance status (PS; odds ratio (OR)=3.83 if PS=1, OR=12.00 if PS ≥2), presence of liver metastasis (OR=2.37) and rare site metastasis (OR=2.00) as PFs for early death. The CHAID analysis retained PS as a major discriminator followed by histological grade (only for patients with PS ≥2). In both models, PS was the most powerful PF for 90-day mortality.

Conclusion: Performance status has to be taken into account in the design of further clinical trials and is one of the most important parameters to guide patient management. For those patients with poor PS, expected benefits from therapy should be weighed up carefully against the anticipated toxicities.

Show MeSH
Related in: MedlinePlus