Introduction

The speed of haemopoietic engraftment after autologous SCT is generally accepted to be a function of the number of stem cells infused at the time of graft administration. Rapid and consistent engraftment of all lineages is observed when a CD34⁺ cell dose over 10 10⁶/kg is given, but transplants utilizing this cell dose are a minority in clinical practice. At doses below this, engraftment is less reliable with one quarter of patients receiving <10 10⁶/kg failing to achieve a platelet count of 20 10⁹/l by day 20 in one study¹ and an even higher percentage at the most common CD34⁺ cell doses infused of 1–5 10⁶ cells/kg. Other factors associated with slower platelet engraftment include the use of BM rather than PBSCs and transplants performed for acute myeloid leukemia.² Delayed platelet engraftment (defined hereafter as an unsupported platelet count <20 10⁹/l 21 days following autologous SCT) presents a particular clinical problem due to the associated requirement for supportive transfusion and the increased resource utilization associated with the effects of haemorrhage, the requirement for blood products and the morbidity associated with blood and platelet transfusion. Moreover, failure of platelet recovery by day 60 after autologous SCT is associated with a significant increase in the incidence of both relapse and non-relapse mortality.²

To date, no method has emerged for specifically predicting delayed platelet engraftment after SCT. A logical approach would be to analyse the number of platelet progenitor cells infused at the time of graft administration. However, accurate assessment of platelet or megakaryocyte progenitor cells has proved difficult since no specific phenotype for such cells exists. The thrombopoietin receptor c-mpl (CD110) is expressed on a subpopulation of CD34⁺ cells that includes but is not restricted to megakaryocyte progenitors.³ Binding of thrombopoietin to c-mpl induces megakaryocytic proliferation and differentiation.⁴ We analysed a series of patients undergoing autologous SCT to determine whether the dose of CD34⁺CD110⁺ cells infused would predict speed of platelet engraftment. We focused on autologous rather than allogeneic transplants to exclude the effects of factors not restricted to graft progenitor cell number (such as CMV infection, GVHD and myelosuppressive medications including methotrexate, ganciclovir and trimethoprim/sulphamethoxazole) on platelet recovery.

Materials and methods

Patients

We retrospectively analysed the graft composition of 39 patients who underwent autologous transplantation for various haematological and non-haematological diseases in the Blood and Marrow Transplant Unit at Westmead Hospital between March 2001 and July 2005 (Table 1). Institutional Ethics approval was given for the analysis. Patients were transplanted for the following diseases: non-Hodgkins lymphoma, n=16; multiple myeloma, n=11; Hodgkins lymphoma, n=4; Ewings sarcoma, n=3; neuroblastoma, n=2; amyloid, n=1; rhabdomyosarcoma, n=1; medulloblastoma, n=1.

Table 1 - Clinical details of 39 patients undergoing autologous stem cell transplantation.

Full table

Mobilization

PBSCs were mobilized with 4 g/m² CY and 10 g/kg/day G-CSF for 4 days before initial stem cell harvest. Leucapheresis was performed on a Cobe Spectra using a 12 l exchange according to standard methods. Collection on 1 day was performed in 34 patients; two consecutive collections were performed in five patients. One patient received BM stem cells in combination with PBSCs.

Immunophenotyping

Absolute CD34⁺ cells and CD34 subsets expressing c-mpl were enumerated using an in-house single platform viable CD34 flow cytometry assay.⁵ Briefly, for fresh samples 10 l of haemopoietic progenitor cell (HPC) (obtained by reverse pipetting) product was stained with 5 l of CD34 PE, 5 l CD45 FITC, 5 l CD110 APC and 10 l of 7AAD in a TRUCOUNT tube (antibodies supplied by BD Bioscience (San Jose, CA, USA) were titred in-house using an HPC product) and incubated for 15 min at room temperature. There were no wash steps or lysis agent used in the procedure. A total of 450 l of PBS was added to the tube immediately before analysis by flow cytometry. Initially, we used fresh HPCs to establish the coexpression of CD110 on CD34⁺ cells. To determine the cursor placement on our dot plots, we chose CD3 APC (same isotype as CD110) as a negative control. CD3 is not expressed on CD34⁺ cells but stains mature T cells within our samples giving rise to specific staining rather than non-specific staining, which is often experienced with an irrelevant isotype control. We found virtually no membrane staining of platelets using the CD110 antibody at the titration utilized in these experiments. For this study, the analysis was performed on cryopreserved samples. At thawing, a small pilot vial was removed from the liquid nitrogen storage container, immediately placed in a water bath at 37°C gently shaken and removed as soon as thawing was complete. A 10 l aliquot was stained as described above. Post-thaw samples were not washed to remove DMSO nor diluted to achieve the desired cell count, to minimize potential errors. Data acquisition was performed within 30 min of thawing.

Flow cytometry

List mode data was acquired on a FACSCalibur flow cytometer and analysed using Cell Quest 3.1 software (BD Bioscience). A modification of the ISHAGE gating strategy described by Sutherland⁶ was used to obtain the absolute number of CD34⁺/CD110⁺ cells per microliter.

Gating strategy for the in-house no-lyse single-platform protocol

Figure 1 shows the gating strategy for CD34 analysis on a thawed PBSCH product, using the in-house no-lysis protocol. FL1 threshold was set on CD45 expression and adjusted to minimize debris (Figure 1a). Viable leukocytes were identified as 7 AAD-negative events (Figure 1b). TRUCOUNT beads were enumerated in R6 (Figure 1f). Acquisition continued until a minimum of 100 CD34⁺CD110⁺ events were collected (this ensured the number of beads acquired always exceeded 3000). The number of CD34⁺CD110⁺ cells/l was calculated as follows: no. of viable CD34⁺/CD110⁺ cells (Figure 1h) multiplied by the number of beads per TRUCOUNT tube divided by the number of beads counted (R6) multiplied by the sample volume.

Figure 1.

In-house no-lyse viable CD34 gating strategy. An example of a PBSCH product. Histogram (a) shows all CD45⁺ leukocytes region 1 (R1). Histogram (b) shows viable leukocytes gated on R1. Histogram (c) is gated on viable CD45⁺ events (that is, R1 and not R2) and CD34⁺ events are identified in R3. Histograms (d) and (e) are sequentially gated from R3 and R4 respectively. Histogram (h) shows the viable CD34⁺CD110⁺ events identified in R5. Histogram (f) shows ungated events and the bead count is obtained from R6. Histogram (g) shows control staining of cells in R5 with anti-CD3 antibody.

Full figure and legend (237K)

Haemopoietic recovery after BMT

Autologous transplantation was performed following therapy using conditioning-chemotherapy regimens at standard doses (Table 1). Leukocyte engraftment was defined as the first of three consecutive days that neutrophils exceeded 0.5 10⁹/l. Platelet engraftment was taken as the first day of platelet count 20 10⁹/l on at least three occasions, 7 days after the most recent platelet transfusion (as defined by the International Bone Marrow Transplant Registry). Platelet infusions were given as single units of pooled platelets. Infusions were given when platelet counts decreased below 15 10⁹/l or according to clinical needs determined by treating clinicians.

Statistical analysis

The statistical software packages SPSS for Windows Version 1.5 and Prism were used to analyse the data. Mann–Whitney and ²-tests were used to compare distributions of continuous or biological variables by subgroups of interest. Two-tailed tests with a significance level of 5% were used throughout. Receiver operator characteristics (ROC)⁷ curves were used to illustrate the performance of CD110 in predicting time to engraftment (<21 vs >21 days).

Results

Details of cell numbers contained in autologous blood or marrow stem cell products

All patients undergoing autologous transplantation received at least 2 10⁶ CD34⁺ cells/kg (Table 1). Median CD34⁺ cell number/kg infused was 3.9 10⁶ (range: 2–27.4 10⁶/kg). The median number of CD34⁺CD110⁺ cells/kg infused was 13.8 10⁴/kg (range: 2.3–130.3 10⁴/kg). There was a weak correlation between the number of CD34⁺ and the number of CD34⁺CD110⁺ cells contained within grafts, r²=0.48 (Figure 2).

Figure 2.

Correlation between absolute number of CD34⁺ and absolute number of CD34⁺110⁺ cells/kg infused at the time of transplant.

Full figure and legend (11K)

Platelet reconstitution post-autologous transplantation

Seven of 39 patients failed to achieve an unsupported platelet count of >20 10⁹/l before day 21, achieving platelet independence between days 22 and >60 (Figure 3). Six of the seven received a graft containing <6.0 10⁴ CD34⁺CD110⁺ cells/kg. The remaining patient received a dose of 43 10⁴ CD34⁺CD110⁺ cells/kg. He underwent autologous transplantation for a lymphoproliferative disorder associated with cold agglutinins with significant autoagglutination at room temperature that may have caused artefactual thrombocytopaenia.⁸ Platelet engraftment occurred at day 39. Thirty-two patients became platelet transfusion independent within 21 days of transplantation. All but two received >6.0 10⁴ CD34⁺CD110⁺ cells/kg (P<0.001 compared with those showing delayed platelet transfusion independence). There was no obvious association between high doses of CD34⁺CD110⁺ cells/kg and rapidity of platelet engraftment (Figure 3). Two patients receiving <6.0 10⁴ CD34⁺CD110⁺ cells/kg displayed platelet engraftment within 21 days of transplantation but one of these two required over 6 months to sustain a platelet count greater than 100 10⁹/l. A ROC curve was generated using the absolute number of CD34⁺CD110⁺ cells/kg for each patient, which showed that >6.0 10⁴ CD34⁺CD110⁺ cells/kg was both highly sensitive (93.8%) and highly specific (85.7%) for achieving platelet engraftment within 21 days (Figure 4).

Figure 3.

Number of days to platelet engraftment plotted against absolute number of CD34⁺CD110⁺ cells infused at the time of transplant. Each point represents one patient. Horizontal line indicates threshold value of 6.0 10⁴ CD34⁺CD110⁺ cells/kg and vertical line 21 days post transplant.

Full figure and legend (15K)

Figure 4.

ROC analysis of the threshold dose of CD34⁺/CD110⁺ cells required for platelet engraftment within 21 days. More than 6.0 10⁴ CD34⁺/CD110⁺ cells/kg gave the highest degree of sensitivity 93.8 and specificity 85.7.

Full figure and legend (48K)

Comparing patients displaying platelet engraftment within 21 days of transplant with those displaying engraftment beyond that time, the former group received a median of 16.4 10⁴ CD34⁺CD110⁺ cells/kg compared with 5.2 10⁴ CD34⁺CD110⁺ cells/kg for the latter group (P=0.003). This highly significant difference in absolute number of CD34⁺CD110⁺ cells infused was not observed when the percentage of CD34⁺ cells expressing CD110 was analysed (median 6.0 vs 5.5% for rapid vs slow platelet engrafters respectively, P=0.45). Three of the seven patients with delayed engraftment received CD34⁺ cell doses above 5 10⁶/kg. Patients with >21 days to platelet engraftment received platelet transfusions more often than those with <21 days to platelet engraftment (median 5 vs 2 transfusions, P<0.0001).

Coexpression of CD110, CD41 and CD61 on CD34⁺ cells

Infusion of a large number of differentiated megakaryocyte progenitors might predict early but unsustained platelet engraftment post transplant. In an attempt to explain the rapid platelet engraftment observed in two patients with <6 10⁴ CD34⁺CD110⁺ cells/kg, we studied the coexpression of CD41 and CD61 on CD34⁺CD110^- cells and CD34⁺CD110⁺ cells. There was no difference in the percentage or absolute numbers of CD34⁺CD110⁺CD41/61⁺ cells or in the percentage or absolute numbers of CD34⁺CD110^-CD41/61⁺ cells between the two patients with <6 10⁴ CD34⁺CD110⁺ cells/kg and two patients from this cohort receiving >6 10⁴ CD34⁺CD110⁺ cells/kg who engrafted at the same time (data not shown).

Discussion

Although the content of CD34⁺ cells in the graft is the single most important determinant of neutrophil and platelet engraftment after autologous blood or marrow transplant, many patients receiving a CD34⁺ cell dose widely considered adequate for transplantation (>2 10⁶/kg) suffer delayed platelet engraftment. At CD34⁺ cell doses below 5 10⁶/kg, there is a 10–15% chance of delayed platelet transfusion independence, and even 5% of patients receiving 5–10 10⁶/kg CD34⁺ cells experience the same complication.⁹ Most patients undergoing autologous transplantation for haematological malignancy receive CD34⁺ cell doses in this range. In our series of 39 consecutive patients undergoing autologous transplant for haematological malignancy and solid tumours, over 60% received a CD34⁺ cell dose below 5 10⁶/kg while another quarter received a CD34⁺ cell dose between 5 and 10 10⁶/kg. Only a minority (more than 10%) of patients in our series received a CD34⁺ cell dose above 10 10⁶/kg at which the risk of delayed platelet engraftment is negligible.

We analysed the correlation between recovery to platelet independence and the infused dose of cells coexpressing CD34 and CD110 (c-mpl, the thrombopoietin receptor). Our data show a strong correlation between infusion of a threshold dose (6 10⁴/kg) of CD34⁺CD110⁺ haemopoietic progenitor cells and achievement of platelet transfusion independence by day 21 post-transplant. Achievement of the threshold dose was a highly sensitive (93.8%) and highly specific (85.7%) test for rapid achievement of platelet transfusion independence following autologous transplantation. Patients receiving less than the threshold dose had a high chance of remaining platelet transfusion dependent beyond 21 days post transplant. For these patients, there was a significantly greater requirement for platelet transfusion. This measure has a superior correlation with platelet transfusion independence compared with the infused dose of CD34⁺ cells/kg. Indeed, three of seven patients in our series with delayed platelet transfusion independence received a CD34⁺ cell dose above 5 10⁶/kg. Above and below the threshold dose we identified, there was no correlation between the dose of CD34⁺CD110⁺ haemopoietic progenitor cells and the time to platelet transfusion independence. We could not identify over-representation of any specific disease state or conditioning regimen in the group of patients with delayed platelet transfusion independence.

Previous attempts to identify progenitor cell populations associated with platelet engraftment have focused on both immature CD34⁺ subsets lacking CD33 or CD38 and committed platelet progenitors expressing the platelet fibrinogen receptor gp IIb/IIIa (CD41/CD61). The number of CD34⁺CD33^- cells and the number of CD34⁺CD38^- cells infused at transplant do not correlate well with platelet transfusion independence.^{10, 11} Dercksen¹² reported a threshold value of 0.45 10⁶ CD34⁺CD41⁺ cells/kg for rapid platelet engraftment following autologous transplant but half of the patients receiving less than this number also engrafted before day 21 limiting the value of this measure in clinical transplantation. Another study showed only weak correlation between the number of CD34⁺CD41a⁺ cells infused and time to platelet recovery to 20 10⁹/l¹³. Similarly, the number of CD34⁺CD61⁺ cells has not proved useful clinically in predicting slow platelet engraftment.^{1, 14} When we analysed expression of CD41/CD61 in conjunction with CD34 and CD110, we could not identify a more accurate method of predicting platelet engraftment based on expression of more mature platelet markers.

In summary, we have identified a threshold dose for a progenitor cell subpopulation that predicts platelet transfusion independence within 21 days of autologous SCT in patients mobilized with CY and G-CSF. The assay is flow-based, can be easily incorporated into the enumeration of CD34⁺ cells at the time of harvest and gives rapid and reproducible results without the need for prolonged colony culture. Future studies will need to address whether exceeding the threshold value in poor mobilizers using repeated collections or alternative mobilization strategies will allow for a directed approach to cell collection to avoid the morbidity and mortality associated with delayed platelet engraftment.

References

1. Meldgaard Knudsen L, Jensen L, Jarlbaek L, Hansen PG, Hansen SW, Drivsholm L et al. Subsets of CD34+ hematopoietic progenitors and platelet recovery after high dose chemotherapy and peripheral blood stem cell transplantation. Haematologica 1999; 84: 517–524. | PubMed | ChemPort |
2. Nash RA, Gooley T, Davis C, Appelbaum FR. The problem of thrombocytopenia after hematopoietic stem cell transplantation. Stem Cells 1996; 14 (Suppl 1): 261–273. | PubMed |
3. Ramsfjell V, Borge OJ, Cui L, Jacobsen SE. Thrombopoietin directly and potently stimulates multilineage growth and progenitor cell expansion from primitive (CD34+ CD38-) human bone marrow progenitor cells: distinct and key interactions with the ligands for c-kit and flt3, and inhibitory effects of TGF-beta and TNF-alpha. J Immunol 1997; 158: 5169–5177. | PubMed | ISI | ChemPort |
4. Kaushansky K. Thrombopoietin: the primary regulator of megakaryocyte and platelet production. Thromb Haemost 1995; 74: 521–525. | PubMed | ISI | ChemPort |
5. Sartor M, Antonenas V, Garvin F, Webb M, Bradstock KF. Recovery of viable CD34+ cells from cryopreserved hemopoietic progenitor cell products. Bone Marrow Transplant 2005; 36: 199–204. | Article | PubMed | ISI | ChemPort |
6. Gratama JW, Kraan J, Keeney M, Sutherland DR, Granger V, Barnett D. Validation of the single-platform ISHAGE method for CD34(+) hematopoietic stem and progenitor cell enumeration in an international multicenter study. Cytotherapy 2003; 5: 55–65. | Article | PubMed | ISI | ChemPort |
7. Altman DG, Bland JM. Statistics notes: diagnostic test 3: receiver operating characteristic plots. Br Med J 1994; 309: 188. | ISI | ChemPort |
8. Crowther H, Collins D, Antonenas V, McGurgan M, Kerridge I, Gottlieb D et al. Successful autologous peripheral blood stem cell harvest and transplant in a patient with cold agglutinins. Bone Marrow Transplant 2006; 37: 329–330. | Article | PubMed | ISI | ChemPort |
9. Weaver CH, Hazelton B, Birch R, Palmer P, Allen C, Schwartzberg L et al. An analysis of engraftment kinetics as a function of the CD34 content of peripheral blood progenitor cell collections in 692 patients after the administration of myeloablative chemotherapy. Blood 1995; 86: 3961–3969. | PubMed | ISI | ChemPort |
10. Millar BC, Millar JL, Shepherd V, Blackwell P, Porter H, Cunningham D et al. The importance of CD34+/CD33- cells in platelet engraftment after intensive therapy for cancer patients given peripheral blood stem cell rescue. Bone Marrow Transplant 1998; 22: 469–475. | Article | PubMed | ISI | ChemPort |
11. Robinson SN, Freedman AS, Neuberg DS, Nadler LM, Mauch PM. Loss of marrow reserve from dose-intensified chemotherapy results in impaired hematopoietic reconstitution after autologous transplantation: CD34(+), CD34(+)38(-), and week-6 CAFC assays predict poor engraftment. Exp Hematol 2000; 28: 1325–1333. | Article | PubMed | ISI | ChemPort |
12. Dercksen MW, Rodenhuis S, Dirkson MK, Schaasberg WP, Baars JW, van der Wall E et al. Subsets of CD34+ cells and rapid hematopoietic recovery after peripheral-blood stem-cell transplantation. J Clin Oncol 1995; 13: 1922–1932. | PubMed | ISI | ChemPort |
13. Feng R, Shimazaki C, Inaba T, Takahashi R, Hirai H, Kikuta T et al. CD34+/CD41a+ cells best predict platelet recovery after autologous peripheral blood stem cell transplantation. Bone Marrow Transplant 1998; 21: 1217–1222. | Article | PubMed | ISI | ChemPort |
14. Maharaj D, Steinberg JP, Gouvea JV, Gieser PW. Changes in endogenous TPO levels during mobilization chemotherapy are predictive of CD34+ megakaryocyte progenitor yield and identify patients at risk of delayed platelet engraftment post-PBPC transplant. Bone Marrow Transplant 1999; 23: 539–548. | Article | PubMed | ISI | ChemPort |

Acknowledgements

We thank the medical and nursing staff of the Blood and Marrow Transplant Service of Westmead Hospital and Children's Hospital Westmead for their assistance with this work.