Abstract
Abstract 4047
Stem Cell collection via apheresis is the preferred method of collecting stem cells for hematopoietic stem cell transplantation. Accurate prediction of stem cell yield is important for the planning of apheresis procedures and for the collection of adequate stem cells. Pierelli et. al. (Vox Sanguinis 2006, 91; 126) proposed a mathematical formula to estimate the CD34+ dose collected on 1st day of apheresis based on the peripheral blood (PB) CD34+ concentration pre-apheresis and the blood volume processed (BVP). Patients and Methods : We tested the predictive value of this formula in a multicenter sample of 1608 apheresis procedures performed at 5 centers (SCSP n=85, CAS n=195, HIAE n=305, China n=172, MDACC n=851). A 50% randomly selected sample of the MDACC study population was included in this analysis. Each contributing institution selected all consecutive apheresis procedures performed over a designated time frame. Baseline patient characteristics were collected including age, gender, diagnosis, mobilization regimen, complete blood counts on day of collection, and the absolute PB CD34+ counts within 24 hours of the first apheresis procedure. Subjects who did not have data on the PB CD34+ counts were not included in the analysis. Information on total BVP and CD34+ cells collected/kg was also collected. There were 694 males (59%), median age was 50 years. To facilitate comparison of results, we used the same statistical methods reported by Pierelli et. al. to assess the correlation between the actual (ay) and predicted (py) CD34+ yields, including assessment of the linear correlation between these measures and the distribution of the ay/py ratio. Results: Data on both actual and predicted yields were available for 1148 (97%) records. Overall, Pearson's correlation coefficient (r) between ay and py was 0.67, ranging from 0.67 for MDACC to 0.86 for China and CAS. Median ay/py ratio was 1.1 (0.02–433) ranging from 0.99 for HIAE to 1.5 for China and CAS. To characterize the correlation between ay and py and facilitate the clinical application of our findings, we classified the actual and predicted yields as falling below (<2), within (2–5), or above (>5) the conventionally acceptable collected CD34+ doses (x106/Kg). Positive predictive value (PPV) of py was estimated considering the distribution of ay as the “gold standard”. PPV was relatively high for py >5 [85% (95% CI 81–89%)], average for py <2 [72% (95% CI 68–76%)], and low for py between 2 and 5 [56% (95% CI 51–62%)]. This pattern was consistent across institutions. PPV ranged from 55% (SCSP) to 80% (China) for py<2; from 37% (CAS) to 68% (MDACC) for py between 2 and 5, and from 73% to 94% for py>5, exceeding 80% at all institutions except SCSP. Overall, 13% of cases predicted to have a CD34+ yield between 2 and 5, had an actual yield <2. This proportion varied across institutions ranging from 5% at SCSP to 28% at HIAE. Notably, the distribution of BVP (ml/Kg) was comparable across the 3 categories of ay defined above with a median of 204 (range 74–263). Consistent pattern was also observed within institutions. Conclusion: Our data indicate that the formula of Pierelli et.al. is associated with high PPV for predicted CD34+ doses >5, acceptable PPV for doses<2, and relatively low PPV for doses falling between 2 and 5. The data also suggests that CD34+ yields correlate with pre-apheresis CD34+ count and are independent of BVP
| N . | Total 1182 . | range . | SCSP 85 . | range . | CAS 195 . | range . | China 172 . | range . | HIAE 305 . | range . | MDACC 425 . | range . |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Absolute PB CD34+ × 106/μl (median, range) | 28 | 0.3–2735 | 31 | 4.2–279 | 25 | 1.3–1750 | 14 | 0.3–411 | 31 | 0.6–604 | 31 | 2–2735 |
| Weight, Kg (median, range) | 72 | 9–169 | 67 | 38–143 | 61 | 9–110 | 63 | 22–100 | 73 | 13–143 | 84 | 11–169 |
| Day 1 Blood volume processed, L (median, range) | 15 | 3–131 | 14 | 8–25 | 12 | 3–29 | 9.5 | 4.5–12.5 | 19 | 5–131 | 16 | 8–27 |
| Actual Day 1 CD34+ dose (x 106/Kg), (median, range) | 2.9 | 0.01–109 | 2.9 | 0.2–29 | 3.8 | 0.01–109 | 1.8 | 0.01–50 | 3.5 | 0.1–44 | 2.8 | 0.2–101 |
| <2 | 36% | 27% | 33% | 53% | 34% | 33% | ||||||
| 2–5 | 31% | 40% | 28% | 20% | 27% | 39% | ||||||
| >5 | 33% | 33% | 38% | 27% | 39% | 28% | ||||||
| Predicted Day 1 CD34+ dose (× 106/Kg), (median, range) | 2.4 | 0.02–211 | 2.6 | 0.3–24 | 2.2 | 0.1–73 | 0.96 | 0.02–25 | 3.2 | 0.1–67 | 2.4 | 0.1–211 |
| <2 | 44% | 45% | 44% | 63% | 33% | 42% | ||||||
| 2–5 | 30% | 25% | 37% | 17% | 28% | 35% | ||||||
| >5 | 27% | 30% | 19% | 20% | 38% | 23% | ||||||
| Ratio CD34+ dose collected/predicted | 1.1 | 0.02–433 | 1.1 | 0.2–11 | 1.5 | 0.03–6.3 | 1.5 | 0.02–433 | 0.99 | 0.04–12 | 1.1 | 0.1–9.9 |
| Pearson's correlation coefficient | 0.67 | 0.83 | 0.86 | 0.86 | 0.77 | 0.67 | ||||||
| Pearson's within Predicted Day 1 dose | ||||||||||||
| <2 | 0.42 | 0.08 | 0.63 | 0.5 | 0.57 | 0.28 | ||||||
| 2–5 | 0.28 | 0.25 | 0.28 | 0.2 | 0.36 | 0.25 | ||||||
| >5 | 0.56 | 0.77 | 0.81 | 0.7 | 0.63 | 0.6 |
| N . | Total 1182 . | range . | SCSP 85 . | range . | CAS 195 . | range . | China 172 . | range . | HIAE 305 . | range . | MDACC 425 . | range . |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Absolute PB CD34+ × 106/μl (median, range) | 28 | 0.3–2735 | 31 | 4.2–279 | 25 | 1.3–1750 | 14 | 0.3–411 | 31 | 0.6–604 | 31 | 2–2735 |
| Weight, Kg (median, range) | 72 | 9–169 | 67 | 38–143 | 61 | 9–110 | 63 | 22–100 | 73 | 13–143 | 84 | 11–169 |
| Day 1 Blood volume processed, L (median, range) | 15 | 3–131 | 14 | 8–25 | 12 | 3–29 | 9.5 | 4.5–12.5 | 19 | 5–131 | 16 | 8–27 |
| Actual Day 1 CD34+ dose (x 106/Kg), (median, range) | 2.9 | 0.01–109 | 2.9 | 0.2–29 | 3.8 | 0.01–109 | 1.8 | 0.01–50 | 3.5 | 0.1–44 | 2.8 | 0.2–101 |
| <2 | 36% | 27% | 33% | 53% | 34% | 33% | ||||||
| 2–5 | 31% | 40% | 28% | 20% | 27% | 39% | ||||||
| >5 | 33% | 33% | 38% | 27% | 39% | 28% | ||||||
| Predicted Day 1 CD34+ dose (× 106/Kg), (median, range) | 2.4 | 0.02–211 | 2.6 | 0.3–24 | 2.2 | 0.1–73 | 0.96 | 0.02–25 | 3.2 | 0.1–67 | 2.4 | 0.1–211 |
| <2 | 44% | 45% | 44% | 63% | 33% | 42% | ||||||
| 2–5 | 30% | 25% | 37% | 17% | 28% | 35% | ||||||
| >5 | 27% | 30% | 19% | 20% | 38% | 23% | ||||||
| Ratio CD34+ dose collected/predicted | 1.1 | 0.02–433 | 1.1 | 0.2–11 | 1.5 | 0.03–6.3 | 1.5 | 0.02–433 | 0.99 | 0.04–12 | 1.1 | 0.1–9.9 |
| Pearson's correlation coefficient | 0.67 | 0.83 | 0.86 | 0.86 | 0.77 | 0.67 | ||||||
| Pearson's within Predicted Day 1 dose | ||||||||||||
| <2 | 0.42 | 0.08 | 0.63 | 0.5 | 0.57 | 0.28 | ||||||
| 2–5 | 0.28 | 0.25 | 0.28 | 0.2 | 0.36 | 0.25 | ||||||
| >5 | 0.56 | 0.77 | 0.81 | 0.7 | 0.63 | 0.6 |
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
