Prognostic Value of TTM with TTM Distribution and Ball Score in Relapsed/Refractory Ibrutinib Treated B-CLL - in the Real-World Setting

There is a high number of clinical and biological parameters with impact on prognosis in B-CLL and number of successful prognostic models were developed (clinical stages, CLL IPI, MDACC score, etc.). However, rapidly changing therapeutic landscape with more successful targeted terapines with different modes of action, render most of these models developed in era of chemo and chemoimmunotherapy less useful. Recently simple prognostic score (BALL score) based on 4 parameters (LDH>UNL, B2microglobulin>5mcg/l, Hemoglobin<120 g/l in males and 110g/l in females, and time from start of last therapy- 0-1 parameters - low risk, 2-3 - intermediate risk and 4 high risk) was developed and validated in number of cohorts (Soumerai et al, Lancet Hematol).

High cost of novel drugs in less affluent countries led to a higher criterion for reimbursement. Until recently in Croatia ibrutinib was reimbursed only for early relapse (less than 24 months from last therapy) or refractory disease with additional criteria for high risk disease including Rai stages III and IV or high tumor burden demonstrated by TTM score >15.

TTM score (www.b-cll.org) is old, simple and continuous parameter useful for prognosis and response assessment (Jaksic B et al BJH 1980) and because it cover tumor mass in all major lymphoid compartments and allow tumor distribution assessment (Jaksic O et al, Haematologica 2001) it can be very useful for response assessment to novel agents were redistribution of lymphocytes can be significant (Jaksic O BJH 2014).

In other to evaluate usefulness of novel BALL score and old TTM score in real life setting where the novel drug is available only to patients with higher risk defined by respective scores, we have evaluated series of 42 RR CLL patients treated with ibrutinib at our institution since March 2015. There were 15 females and 27 males, median age 70 years (range 53 to 82), 9 patients had 17p deletion, median TTM was 14 (range 1.2-28), 20 patients had Rai stage III/IV. Median follow up was 24 months, and maximal 64 months. Since all patients were in early relapse or refractory, we have actually used simplified BALL score based on only 3 parameters. There was no significant relationship between BALL and TTM in our patient population.

In our patient cohort low risk BALL score had only 3 patients, 28 patients had intermediate risk and 11 high risk and it showed only marginal discriminative power (p=0.054) While TTM>15 did not show discriminative power (p=0.14).

When we added TTM>18 as an additional parameter (point) to BALL score, this modified score (0-3 low/intermediate - 23 patients and 3-4 high risk - 19 patients) showed significant discriminative power (p=0.002) (Figure 1).

Observed results indicate that: 1) performance of prognostic models may be significantly dependent on discriminative power of its components and if these components are criteria for initiation of therapy with novel agent(i.e. only high risk group defined by these parameters) it significantly lower its prognostic power, and vice versa 2) such criteria for therapy selection/initiation until these criteria are met may actually mean that we treat patients with novel agents when these are less effective.

We have shown that addition of TTM significantly improves BALL score.

Refinement of prognostic scores with parameters that may have predictive power for novel drugs, as well as adaptation criteria for reimbursement in future studies (including pharmacoeconomics) may further improve quality of care of B-CLL patients.

View largeDownload slide

View largeDownload slide

Close modal