Prognostic factors for response and survival in primary central nervous system lymphoma (PCNSL) from a randomized phase III trial (G-PCNSL-SG-1)

Reporter: Lara Bonner Millar, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: June 5, 2011

Presenter: K. Jahnke
Presenter's Institution: Charité Hospital, Campus Benjamin Franklin, Berlin, Germany


  • CNS lymphoma is rare but incidence is rising in both immunocompetent and imuunocompromised patients. For reference, the incidences in 1973 and 1992 were:
    • 1973: 2.5/10,000,000
    • 1992: 30/10,000,000
  • The greatest rise has occurred among the elderly. The median age at diagnosis is around 60 yrs in immunocompetent patients.
  • The International Extranodal Lymphoma Study Group established a prognostic index for CNS lymphoma based on 105 patients. The independent negative predictors of survival are age > 60, ECOG PS <1, elevated LDH, elevated CSF protein, and deep location within the brain. Each factor was scored as one point if present, resulting in 3 different prognostic groups (Risk groups: 0-1 factors, 2-3 factors, 4-5 factors).
  • Another model has been developed at MSKCC: in this model age and Karnofsy PS were predictive of survival. 3 different prognostic groups were generated from this model.
  • The purpose of this study was to establish prognostic factors in newly diagnosed primary central nervous system lymphoma (PCNSL) specifically in immunocompetent patients. It is important to generate prognostic factors based on phase III data because prior prognostic factors were mainly generated from small phase II or retrospective series. Differences in treatment outcome may be related to inhomogeneous distribution of prognostic factors.


  • Four-hundred and eleven patients (median age 63 years, median Karnofsky performance status (KPS) 70%) enrolled in a German multicenter randomized phase III trial (G-PCNSL-SG-1) between May 2000 and May 2009 were analyzed as part of this study. Univariate and multivariate logistic regression analyses were performed to identify prognostic factors for overall response (OR) and progression-free (PFS) and overall survival (OS).
  • The aim of the phase III non-inferiority study was to deliver primary systemic high-dose methotrexate (ifosphamide was added in 2006) treatment to PCNSL patients and to define the role of whole brain irradiation (WBI) in primary therapy, specifically to assess whether patients who have undergone primary chemotherapy can postpone irradiation until recurrence without decrements in progression-free survival or overall survival. This is the first multicenter RCT to ask this question.
    • In this study, one arm was given "standard therapy" with six cycles of high-dose MTX and subsequent whole brain irradiation; the second arm gave the same MTX but reserved irradiation until recurrence.
    • Those who recurred underwent a second randomization for salvage therapy with AraC vs. irradiation to assess equivalency with regard to time to progression and overall survival.


  • The results of the study, which was analyzed according to intention-to-treat, N = 411, found that up front whole brain radiotherapy improves PFS but not OS.
    • Median overall survival was 32.4 months in patients receiving whole brain radiotherapy and 37.1 months in those not receiving whole brain radiotherapy (p=0·71).
    • Median progression-free survival was 18.3 months in patients receiving whole brain radiotherapy, and 11.9 months in those not receiving whole brain radiotherapy (p=0·14).
  • The median age was 62 and the KPS was 80%
  • In order to identify prognostic factors for response, PFS, and OS, simple log regressions and Cox regressions were done.
  • Several favorable prognostic factors for overall response were identified on multivariate analysis:
    • These included female gender (p= 0.003), no corticosteroid use at entry (p=.0033), < 2 lesions (p= .009), and HDMTX plus ifosfamide vs. HDMTX monotherapy (p=SS).
  • Age (<60 vs. ?60 years; p = 0.007), KPS (?70 vs. <70%; p <0.001), and BMI ? 25 were strongly associated with OS.

Author's Conclusions

  • This is the first study that established prognostic factors for OR and survival in PCNSL using a multicenter, prospective, randomized phase III trial. The data, especially the strong association of age and KPS with OS, should be taken into consideration in the design of future studies in PCNSL.
  • Some of the prognostic factors require further explanation. For example, patients not on steroids may have been in better overall health and therefore could get the full course of MTX, resulting in improved outcomes. In terms of gender, female gender may have been a prognostic factor because females in this population had fewer comorbidities than males. Body mass index likewise could be reflective of the burden of comorbidities.

Clinical Implications

  • The strong association of age and PS justifies treatment stratification according to these risk factors in future trials.
  • Further assessment of the application of these prognostic factors, identified in immunocompetent patients, to those who are immunocompromised would be very useful. In addition, further study of other factors that may be relevant in the immunocompromised population would be of great interest.
  • We need a prognostic scoring system for clinical use based on simple clinical features; such a system should be based on the results of a large study population, such as this. It will be interesting to see if all previously identified prognostic indices which were based on smaller numbers of patients not treated in a randomized fashion will maintain validity.
  • The authors' findings of improved PFS with addition of whole brain radiotherapy support the continued use of this modality in certain subsets of patients with primary CNS lymphoma; whole brain radiotherapy has previously been associated with severe neurocognitive toxicity in patients over 60, and analysis of neurocognitive outcomes for patients treated as part of this study would be of great interest.
  • There are no established prognostic factors for MTX induced toxicity and the optimal dose of MTX in PCNSL is not established. One major problem is the MTX dose is often reduced a prioi in elderly patients (to reduce risk of acute renal failure), which may lead to worse outcome. The use of prognostic factors to guide clinical decision making may avoid toxicity.
  • Overall, this study represents an important contribution to the literature, and may prove practice changing if prognostic factors identified here are adopted and utilized during clinical decision-making.


  1. Abrey LE, Ben-Porat L, Panageas KS, et al: Primary central nervous system lymphoma: The Memorial Sloan-Kettering Cancer Center prognostic model.
    J Clin Oncol 24:5711-5715, 2006
  2. Ferreri AJM, Blay JY, Reni M, et al: Prognostic scoring system for primary CNS lymphomas: The International Extranodal Lymphoma Study Group experience.
    J Clin Oncol 21:266-272, 2003
  3. Thiel E, Korfel A, Martus P, et al. High-dose methotrexate with or without whole brain radiotherapy for primary CNS lymphoma (G-PCNSL-SG-1): a phase 3, randomised, non-inferiority trial. The Lancet 11: 1036–1047, 2010.