Abstract
The overall unfavorable prognosis of CD10 negative (CD10−) precursor B-cell acute lymphoblastic leukemia (BCP-ALL) is well known. We analyzed 4473 pediatric patients (pts) <18 years (y) with BCP-ALL and immunophenotyping of CD10 enrolled from 1986 to 2000 in three consecutive ALL-BFM trials to explore prognostic factors in the CD10− subset. CD10 negativity was defined by CD10 expression in <20% of blasts. 233 pts (5.2%) were CD10−. In comparison to CD10 positive (CD10+) BCP-ALL pts, CD10− pts comprised more infants (age <1y 34% vs. 1%, p(X2)<0.001), more cases with hyperleukocytosis (WBC ≥100/nl 43% vs. 6%, p<0.001), more CNS involvement (CNS positive 10% vs. 2%; p<0.001) and an impaired treatment response (prednisone poor response (PPR) 22% vs. 5%, p<0.001; induction failure 6% vs. 1%, p<0.001). Estimated probability of 5 years event free survival (5y-pEFS) was significantly lower in pts with CD10− as compared to CD10+ BCP-ALL (49±3% vs 81±0.6%, p(log-rank)<0.001). Cox regression analysis including age, WBC, prednisone response (PR) and MLL/AF4 status as covariables revealed CD10 negativity as independent prognostic factor (RR 1.5, 95% confidence interval (CI) 1.1–2.1, p=0.01). Further analyses were performed within the CD10− group: 83% of infants and 60% of pts ≥1y were successfully analyzed for MLL/AF4. MLL/AF4 was detected in 55% of pts <1y and 27% of pts ≥1y. The well known risk factors for BCP-ALL (sex, age, WBC, CNS involvement, MLL/AF4 and PPR) also had prognostic impact within the CD10− group:
. | . | n* . | 5y-pEFS* (%) . | SE (%) . | p (log-rank) . |
---|---|---|---|---|---|
*5 pts w/o reinduction were excluded | |||||
sex | female | 109 | 55 | 5 | 0.022 |
male | 119 | 40 | 5 | ||
age | <1y | 78 | 25 | 5 | <0.001 |
≥1y | 150 | 62 | 4 | ||
WBC | <100/nl | 128 | 62 | 4 | <0.001 |
≥100/nl | 100 | 33 | 5 | ||
CNS | neg | 181 | 54 | 4 | 0.011 |
pos | 21 | 33 | 10 | ||
MLL/AF4 | neg | 95 | 53 | 5 | 0.001 |
pos | 61 | 29 | 6 | ||
PR | good | 170 | 57 | 4 | <0.001 |
poor | 50 | 30 | 6 |
. | . | n* . | 5y-pEFS* (%) . | SE (%) . | p (log-rank) . |
---|---|---|---|---|---|
*5 pts w/o reinduction were excluded | |||||
sex | female | 109 | 55 | 5 | 0.022 |
male | 119 | 40 | 5 | ||
age | <1y | 78 | 25 | 5 | <0.001 |
≥1y | 150 | 62 | 4 | ||
WBC | <100/nl | 128 | 62 | 4 | <0.001 |
≥100/nl | 100 | 33 | 5 | ||
CNS | neg | 181 | 54 | 4 | 0.011 |
pos | 21 | 33 | 10 | ||
MLL/AF4 | neg | 95 | 53 | 5 | 0.001 |
pos | 61 | 29 | 6 | ||
PR | good | 170 | 57 | 4 | <0.001 |
poor | 50 | 30 | 6 |
Out of a number of immunophenotypic markers, analyzed at different expression cut-off points, CD24 at missing or weak expression of <40% and CD65 at high expression of ≥40% were significantly correlated with unfavorable clinical characteristics and worse outcome within the CD10− group. Significant correlation with PR could only be demonstrated for expression of CD24, which is presumed to act as negative regulator in B-cell development through mediation of apoptosis.
. | age<1y* . | WBC ≥100/nl* . | MLL/AF4 pos* . | PPR# . | pEFS§ . |
---|---|---|---|---|---|
. | n/total (%) . | n/total (%) . | n/total (%) . | n/total (%) . | % ±SE . |
* all p(X2)<0.01, #CD24 p=0.01, CD65 n.s., §all p(log-rank]<0.001 | |||||
CD24 <40% | 37/77 (48) | 52/77 (68) | 33/56 (59) | 24/73 (33) | 32 ±5 |
CD24 ≥40% | 34/122 (28) | 35/122 (29) | 18/76 (24) | 20/119 (17) | 59 ±5 |
CD65 <40% | 56/180 (31) | 67/180 (37) | 39/120 (33) | 39/175 (22) | 30 ±7 |
CD65 ≥40% | 22/44 (50) | 32/44 (73) | 21/34 (62) | 10/41 (24) | 54 ±4 |
. | age<1y* . | WBC ≥100/nl* . | MLL/AF4 pos* . | PPR# . | pEFS§ . |
---|---|---|---|---|---|
. | n/total (%) . | n/total (%) . | n/total (%) . | n/total (%) . | % ±SE . |
* all p(X2)<0.01, #CD24 p=0.01, CD65 n.s., §all p(log-rank]<0.001 | |||||
CD24 <40% | 37/77 (48) | 52/77 (68) | 33/56 (59) | 24/73 (33) | 32 ±5 |
CD24 ≥40% | 34/122 (28) | 35/122 (29) | 18/76 (24) | 20/119 (17) | 59 ±5 |
CD65 <40% | 56/180 (31) | 67/180 (37) | 39/120 (33) | 39/175 (22) | 30 ±7 |
CD65 ≥40% | 22/44 (50) | 32/44 (73) | 21/34 (62) | 10/41 (24) | 54 ±4 |
Including age, WBC, PR and MLL/AF4 status as covariables, out of the analyzed markers only CD65 proved to be an independent prognostic factor in CD10− BCP-ALL (Cox regression analysis: RR 1.5, 95% CI 1.1–2.9, p=0.018). The identification of additional prognosis associated immunophenotypic markers may contribute to further refinement of treatment strategies for CD10− BCP-ALL pts.
Author notes
Corresponding author
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal