Abstract
Background : The chronic lymphocytic leukemia international prognostic index (CLL-IPI) was first reported in 2015 and published in 2016 (Lancet Oncology 17:779). The utility of this prognostic tool has been confirmed in independent validation studies across countries and in different practice settings (i.e., clinical trials, academic medical centers, national population based cohorts). However, there are subtle differences of median survival of each risk category across series and limited information is available on the utility of the tool to predict time from diagnosis to first treatment (TTFT). To increase understanding of these important clinical questions, we conducted a systematic review and meta-analysis which includes all published studies that have used CLL-IPI to predict clinical outcome of CLL.
Material & Methods : A systematic literature search was performed using PubMed to identify full length reports on the utility of the CLL-IPI reported prior to June 30, 2017. The search strategy used both Medical Subject Headings (MESH) terms and free text words to increase the sensitivity. The electronic search yielded 9 records including 7 full-text articles assessed for eligibility. A complementary manual search of the ASH, EHA and ASCO conference proceedings identified 3 additional citations.
Results: In total 12 series including 7909 patients analysed the impact of CLL-IPI on overall survival (OS). The patient distribution into the CLL-IPI risk categories was as follows: low-risk [LR], median, 45.9%[range, 1.3-58%]; intermediate-risk [IR], median, 30% [range,8.2-39%]; high-risk [HR], median, 16.5% [range, 12.6-56%]; very high-risk [v-HR], median, 3.6% [range, 2.2-40.8%]. The wide range of distribution across CLL-IPI risk categories reflects differences of the point in the course of the disease at which CLL-IPI was assessed in the different studies (i.e., at diagnosis [8 studies], at time of first treatment [3 studies] and at relapse [one study]). The 5-year survival probability was 92% for LR CLL-IPI (95% CI, 90-93%), 81% for IR CLL-IPI ( 95% CI, 80-83%), 60% for HR CLL-IPI (95% CI, 58-63%) and 34% for v-HR CLL-IPI (95% CI, 29-40%)(Fig 1a). Values of the "Q" or "I2" test suggest a certain degree of heterogeneity across different studies (i.e., [LR], Q=63.9, P< 0.001; I2= 84%; [IR], Q=59.8, P<0.001, I2= 83%; [HR], Q=49.7, P<0.001,I2= 80%; [v-HR], Q=27.4, P=0.001, I2= 67%).
TTFT was evaluated in 9 studies including 5206 patients. The probability of remaining free from therapy at 5 years on pooled meta-analysis was 82% (95% CI: 80-83%) in the LR group, 45% (95% CI: 43-48%) in the IR group, 30% in the HR group (95% CI: 26-33%) and 16% in the v-HR group (95% CI: 10-22%)(Fig 1b). Notably, the heterogeneity in TTFT across different studies was greatest among patients of LR (Q=38.8,P<0.001;I2 =79%) and IR CLL-IPI (Q=27.0, P<0.001; I2 =70%) but significantly decreased among patients of HR (Q=14.4,P=0.57; I2 = 45%) and v-HR CLL-IPI (Q=1.65,P=0.95; I2 =0%) indicating more consistent TTFT outcomes among the HR and v-HR groups across studies.
Conclusions: In this comprehensive review and meta-analysis of studies published on CLL-IPI to date, we confirmed the value of the CLL-IPI as a robust tool to predict OS and TTFT. Although CLL-IPI simply prognosticates CLL and integrates multiple factors into a single risk score, further prospective evaluation of this tool in the era of B-cell receptor and bcl-2 inhibitors are needed.
Molica: Roche: Other: Advisory Board; Jansen: Other: Advisory Board; AbbVie: Other: Advisory Board; Gilead: Other: Advisory Board. Shanafelt: Jannsen: Research Funding; GlaxoSmithKline: Research Funding; Genentech: Research Funding; Celgene: Research Funding; AbbVie: Research Funding; Pharmacyclics: Research Funding; Hospira: Research Funding.
