Recursive partitioning analysis of prognostic factors for glioblastoma patients aged 70 years or older

Jacob G Scott, Luc Bauchet, Tyler J Fraum, Lakshmi Nayak, Anna R Cooper, Samuel T Chao, John H Suh, Michael A Vogelbaum, David M Peereboom, Sonia Zouaoui, Hélène Mathieu-Daudé, Pascale Fabbro-Peray, Valérie Rigau, Luc Taillandier, Lauren E Abrey, Lisa M DeAngelis, Joanna H Shih, Fabio M Iwamoto, Jacob G Scott, Luc Bauchet, Tyler J Fraum, Lakshmi Nayak, Anna R Cooper, Samuel T Chao, John H Suh, Michael A Vogelbaum, David M Peereboom, Sonia Zouaoui, Hélène Mathieu-Daudé, Pascale Fabbro-Peray, Valérie Rigau, Luc Taillandier, Lauren E Abrey, Lisa M DeAngelis, Joanna H Shih, Fabio M Iwamoto

Abstract

Background: The most-used prognostic scheme for malignant gliomas included only patients aged 18 to 70 years. The purpose of this study was to develop a prognostic model for patients ≥70 years of age with newly diagnosed glioblastoma.

Methods: A total of 437 patients ≥70 years of age with newly diagnosed glioblastoma, pooled from 2 tertiary academic institutions, was identified for recursive partitioning analysis (RPA). The resulting prognostic model, based on the final pruned RPA tree, was validated using 265 glioblastoma patients ≥70 years of age from a data set independently compiled by a French consortium.

Results: RPA produced 9 terminal nodes, which were pruned to 4 prognostic subgroups with markedly different median survivals: subgroup I = patients <75.5 years of age who underwent surgical resection (9.3 months); subgroup II = patients ≥75.5 years of age who underwent surgical resection (6.4 months); subgroup III = patients with Karnofsky performance status of 70 to 100 who underwent biopsy only (4.6 months); and subgroup IV = patients with Karnofsky performance status <70 who underwent biopsy only (2.3 months). Application of this prognostic model to the French cohort also resulted in significantly different (P < .0001) median survivals for subgroups I (8.5 months), II (7.7 months), III (4.3 months), and IV (3.1 months).

Conclusions: This model divides elderly glioblastoma patients into prognostic subgroups that can be easily implemented in both the patient care and the clinical trial settings. This purely clinical prognostic model serves as a backbone for the future incorporation of the increasing number of potential molecular prognostic markers.

Figures

**Figure 1**
Recursive partitioning analysis (RPA) trees for the 437 patients in the Memorial Sloan-Kettering Cancer Center (MSKCC) + Cleveland Clinic Foundation (CCF) data set (All patient and tumor characteristics, as well as extent of surgery, (Table 1) were evaluated as potential split points. Nine terminal nodes were pruned to generate four prognostic subgroups using the endpoint of overall survival. Abbreviations: KPS=Karnofsky performance status; PR=partial resection; GTR=gross total resection; N=number of patients in subgroup.

**Figure 2**
Kaplan-Meier curves showing overall survival for (A) the Memorial Sloan-Kettering Cancer Center (MSKCC) + Cleveland Clinic Foundation (CCF) data set split according to subgroups derived from its RPA; (B) the French data set split according to subgroups derived from MSKCC + CCF RPA. Abbreviations: KPS=Karnofsky performance status; PR=partial resection; GTR=gross total resection

Source: PubMed

Recursive partitioning analysis of prognostic factors for glioblastoma patients aged 70 years or older

Abstract

Figures

Sponsorer og samarbeidspartnere

Medisinsk tilstand

Legemiddelintervensjoner