Prognostic value of CLL-IPI in R/R CLL

An evaluation of the chronic lymphocytic leukemia (CLL) International Prognostic Index (IPI) in idelalisib phase III trials was published on 8 November 2018 in Leukemia Lymphoma, by lead author Jacob Soumerai from Massachusetts General Hospital, MA, USA.

The CLL-IPI is commonly used as a prognostic marker in naïve CLL but its use in relapsed or refractory (R/R) CLL has not yet been determined. The CLL-IPI system consists of five adverse prognostic factors, including age (> 65 years; one point), Rai I–IV or Binet B/C (one point), β2-microglobulin (B2M, > 3.5 mg/L; two points), unmutated immunoglobulin heavy chain variable region (IGHV; two points), and TP53 mutation and/or deletion (four points). Depending on the scoring from these five factors, patients are separated into four CLL-IPI risk groups: low (score 0–1), intermediate (score 2–3), high (score 4–6), and very high (score 7–10). The aim of this study was to evaluate results from three phase III trials (115 [NCT01569295], 116 [NCT01539512], and 119 [NCT01659021]) using idelalisib treatment for R/R CLL, so as to determine the prognostic value of CLL-IPI for this population. The primary endpoint of the study was overall survival (OS).

Study design

• N = 897 R/R CLL patients randomized and treated from three phase III studies (115 [N = 416], 116 [N = 210], and 119 [N = 171])
• Of those, n = 864 (96.3%) patients were included in the complete analysis dataset
• Patients considered in the idelalisib arm (n = 491)
  • Study 115 (NCT01569295):
    • Patients who received idelalisib plus rituximab and bendamustine
  • Study 116 (NCT01539512):
    • Patients who received idelalisib plus rituximab
  • Study 119 (NCT01659021):
    • Patients who received idelalisib plus ofatumumab
  • Patients considered in the comparator arm (n = 406)
    • Study 115:
      • Patients who received placebo plus rituximab and bendamustine
    • Study 116:
      • Patients who received placebo plus rituximab
    • Study 119:
      • Patients who received placebo plus ofatumumab
    • Median age (range) = 66 (32–92) years
    • Median number of prior lines (range) = 3 (1–13)
    • Median follow-up (range) = 21.4 (0.03–45.3) months (290 deaths)

Results

Whole cohort (idelalisib and comparator arms):

• CLL-IPI scoring risk groups:
  • Low: 2.2%
  • Intermediate: 12.8%
  • High: 48.7%
  • Very high: 36.2%
• The CLL-IPI risk score was prognostic for OS in R/R CLL patients (log rank test across risk groups P < 0.0001; C-statistic of 0.706 [95% CI, 0.587–0.813])
• 24-month OS rates per CLL-IPI risk group:
  • Low: 93.3%
  • Intermediate: 88.3%
  • High: 69.8%
  • Very high: 52.6%
  • Statistical significant differences in OS were observed between intermediate and high (HR = 3.044), and between high and very high risk groups (HR = 1.835) but not between low and intermediate risk groups (HR = 1.846)

Idelalisib versus comparator arm:

• The CLL-IPI was prognostic for OS in the comparator subgroup (log rank test across risk groups P < 0.0001; C-statistic of 0.736; 95% CI, 0.565–0.879) and for OS in the idelalisib subgroup (log rank test across risk groups P < 0.0001; C-statistic of 0.682; 95% CI, 0.509–0.834)
• The CLL-IPI was also prognostic for OS in patients who received chemo-immunotherapy (study 115; C-statistic 0.742; 95% CI, 0.563–0.889) or CD20 antibody therapy (studies 116 and 119; C-statistic 0.670; 95% CI, 0.503–0.819)

Multivariate analysis for CLL-IPI variables:

• From the five factors used to calculate the CLL-IPI, prognostic for OS were:
  • Age > 65 (HR = 1.5; [95% CI, 1.18–1.91]; P = 0.001)
  • B2M > 3.5 mg/L (HR = 2.64; [95% CI, 1.77–3.94]; P < 0.0001)
  • IGHV unmutated (HR = 1.48; [95% CI, 1.06–2.07]; P = 0.02)
  • TP53 mutation and/or deletion (HR = 2.05; [95% CI, 1.63–2.58]; P < 0.0001)
• From the five factors used to calculate the CLL-IPI, non-prognostic for OS were:
  • Rai I–IV or Binet B/C stage (HR = 1.29; 95% CI, 0.32–5.2, P = 0.72)
• Only age significantly affected OS when considering with idelalisib treatment (P = 0.0359) but no other CLL-IPI factor was associated with idelalisib treatment

The results from this analysis demonstrate that the CLL-IPI scoring system is prognostic for OS in R/R CLL patients. Nevertheless, the multivariate analysis suggested that the contribution of each CLL-IPI scored factor might be different in R/R CLL. Thus, the authors recommend a slightly modified CLL-IPI scoring system that will provide a good prognostic factor for outcomes in R/R CLL.