Technique for PBSC harvesting in children of weight under 10 kg
Correspondence to: Dr Z Ko Department of Internal Medicine Hematooncology, University Hospital Brno, Brno, Jihlavska 20, 639 00 Czech RepublicPeripheral blood stem cell (PBSC) harvesting in the smallest children (weight kg), with newly diagnosed tumours were scheduled to be treated with three to six sequential courses of high dose chemotherapy, each followed by PBSC support. About 50 ml of blood was drawn from a venous catheter, injected into a transfer bag containing ACD A, and centrifuged. The buffy coat obtained was pooled in a collection bag, remaining plasma and erythrocytes were immediately reinfused and a subsequent cycle started. From three to 13 cycles were performed in 1 days and 18.0 106 CD34+cells were collected. We did not detect any bacterial contamination or any notable complications. Fifteen PBSC reinfusions have been performed to date, each with rapid engraftment taking between 7 and 13 days. Patients are in very good PR (18 months from diagnosis) or in CR (6 and 8 months). The aim of new protocols is to use several cycles of a regimen containing a small number of agents with known activity at maximum tolerated doses. T asics running he use of PBSC rescue after each high dose chemotherapy appears to be the most favourable way to avoid delays between courses and to minimize risks from prolonged neutropenia and thrombocytopenia. PBSC have some advantages compared to bone marrow, including minor invasivity,1 and more rapid reconstitution of both haematopoiesis and immune function.2 However, limited data are available on PBSC collection in the smallest pediatric patients weighing less than 10 kg.3,4,5,6 PBSC collections using automated or semi automated devices (eg COBE Spectra) in these patients are much more complicated and risky than in older children or adults. Small children are very sensitive to hypocalcaemia. The volume of the disposable apheretic set is more than 25% of the total blood volume (TBV), and flow capacity of the most frequently used permanent tunnelled central venous catheter (CVC) is usually not sufficient for leukapheresis.
In our department, PBSCs have been collected from 40 children since 1998 and a total of more than 1100 aphereses have been performed since 1995. In spite of our previous experience in older children7 and considering the increased risk of leukapheresis in the smallest children, we have developed a simple manual technique for PBSC harvesting in children weighing less than 10 kg. Our aim was to collect a sufficient number of haematopoietic stem cells (HSC) for several PBSC supports.
Three girls with weight under 10 kg, aged less than 12 months, with newly diagnosed high r asics running isk malignancies were subjected to PBSC collection between December 2000 and March 2001, and subsequently treated with sequential courses of high dose chemotherapy, each followed by PBSC support. Details considering patients’ data are shown in Table 1. Children with brain tumours (UPN 00018 and UPN 00096) were treated according to the Baby Brain UKCCSG protocol, but with PBSC support instead of autologous whole blood reinfusions. This protocol consisted of six courses of chemotherapy (cyclophosphamide 120 mg carboplatin AUC = 6.63, vincristine 0.05 mg per course) with G CSF support and whole blood reinfusions. Chemotherapy was delivered biweekly, irrespective of the peripheral blood count. One girl (UPN 01030) with rhabdomyosarcoma was treated according to the SIOP MMT 98 protocol with a reduced cyclophosphamide dose in the first course. A combination of cyclophosphamide and growth factor in two daily doses was chosen to attempt to reach a maximal available concentration of mobilized progenitors in the peripheral blood without major toxicity.9,10,11
The percentage of CD34+ cells among total leukocytes was determined as described previously,12 according to the ‘Milan protocol’.13 The number of CFU GM was evaluated in routine colony assay systems using complete methylcellulose based medium MethoCult HCC 4434 (StemCell Technologies, Vancouver, Canada). CFU GM colonies were scored after 14 days of incubation (37 5% CO2 humidified atmosphere).
Harvest (150 ml) and collection (600 ml) bags were prepared using 150 ml and 600 ml transfer bags (Baxter, Fenwal Division, Deerfield, IL, USA) about a week before the expected harvest. Spikes of tubings of transfer bags were cut off and connected with the male inlets of three way stopcocks (Ohmeda, Helsingborg, Sweden). Prepared bags were then sterilized.
Harvests were initiated when the CD34+ cell count in a patient’s peripheral blood exceeded at least 50 106 (50 per microlitre). All manipulations were performed in a laminar flow box (CA 4; Clean Air, Woerden, The Netherlands) except withdrawal and centrifugation.
Prior to collection, one of the prepared harvest bags was primed with 5 ml ACD A (IVEX Pharmaceuticals, Larne, UK) and a 60 ml plastic disposable syringe (B Braun Melsungen, Melsungen, Germany) was connected to the three way stopcock. Approximately 50 ml of blood (8 ml 10% of TBV) was drawn using the previously connected syringe and then injected into the harvest bag containing ACD A. After centrifugation (15 min, 120 g; centrifuge GR 4.22; Jouan, St Herblain, France), the harvest bag was carefully transported to the laminar flow box and placed into a plasma extractor.
At first, plasma was drawn from the harvest bag and retained in the syringe. The buffy coat was then carefully obtained using a new 10 ml disposable syringe and injected into the collection bag. Finally, plasma was injected back into transfer bag, mixed with remaining erythrocytes, and slowly (1 ml reinfused to the patient. The next cycle started after the reinfusion. Buffy coats were pooled and stored (no longer than 24 h) in the collection bag until cryopreservation. Samples were drawn from the collection bag usually after each procedure and rapidly analysed for blood count and CD34+ cell content to assess the number of progenitors in the collection bag. An entire cycle lasted about 60 min and it was possible to complete up to five cycles a day without taking special measures in the laboratories. Cryopreservation was performed as described previously.14 Briefly, pooled buffy coats were brought to the required volume by 4% human serum albumin (Instituto Grifols) in Hanks’ salt solution (Sigma Aldrich) and mixed. Samples for CFU GM assay and asics running microbial contamination test were drawn. The cell suspension was then mixed with DMSO (Sigma Aldrich), divided into freezing bags (Baxter, Fenwal Division; Nexell Therapeutics, Irwine, CA, USA) according to the number of collected progenitors, cryopreserved and stored in liquid nitrogen. Thawed PBSC were washed with 4% human serum albumin (Instituto Grifols) in Hanks’ salt solution (Sigma Aldrich) before reinfusion to avoid toxicity associated with DMSO.
During mobilization chemotherapy, no complication was observed. All patients developed leukopenia ( 109 Slight thrombocytopenia (>80 109 was seen in UPNs 00018 and 00096. The pretreated patient (UPN 01030) developed moderate thrombocytopenia (>35 109 but no substitution was required. Harvests started on day 10 (patients 00018 and 00096) and on day 15 (patient 01030) from the first cyclophosphamide infusion. Details of peripheral blood WBC and CD34+ cell counts at harvests are shown in Table 2.
PBSC collections lasted 3 days in patients 00018 and 00096. We performed nine and 13 harvest cycles, respectively, to obtain sufficient numbers of stem cells for six supports and a back up. Seven bags containing 2.67 and 1.79 106 CD34+ cells were frozen for these patients. Three harvest cycles were more than sufficient to collect an adequate number of CD34+ cells for three supports and a back up in patient 01030. We waited for higher CD34+ cell counts for logistical reasons and based on previous results with patients 00018 and 00096. On day 15, the CD34 count reached more than 2000 CD34+ cells per l and 32.20 106 CD34+ cells were collected. Four bags, each containing 8.05 106 CD34+ cells were frozen for this patient according to the protocol. Details about mobilizations and harvests are shown in Table 2. fluids were administered to patients during procedures except for heparin infusion. In two patients without pretreatment, haematocrit ranged between 0.35 and 0.36 respectively, both before and after collections. The haematocrit values of the pretreated child ranged between 0.29 and 0.26 (before and after collections). Blood pressure and other vital signs were monitored during collection days and no deviations from normal range for age were observed. We did not detect any microbial contamination of the collected cell suspensions. Children were housed with their mothers in a quiet ward, and no complications were observed.
To date, all 15 PBSC reinfusions have been performed in these three children. Rapid engraftment was observed after 7 to 13 days upon each PBSC reinfusion. Patient 00018 is in a very good PR 18 months after diagnosis. Patients 00096 and 01030 are in first CR lasting 6 and 8 months, respectively. Setting up leukapheresis can be time consuming and can delay the start of chemotherapy. There are many contentious issues involving PBSC collections in low weight infants. To date, only limited data are available.3,4,5,6,17 Complications related to leukapheresis using separation devices were described elsewhere.17,18 Citrate induced hypocalcaemia may represent serious side effects with a multifaceted spectrum of clinical manifestations, including convulsions and severe hypotension. Small children are very sensitive to hypocalcaemia and they cannot report incipient paresthesias. Using heparin instead of citrate and bedside monitoring of ionized calcium levels with calcium gluconate substitution can alleviate this problem.3,4 Another problem is the very low total blood volume (TBV) which is less then 800 ml. Priming the apheretic set with leukocyte depleted and irradiated erythrocytes usually resolves this issue. On the other hand, a transfusion unit of erythrocytes also contains citrate solution and initiation of leukapheresis should be performed very carefully. The reinfusion procedure at the end of leukapheresis must be carried out cautiously to avoid volume overload.
Insufficient venous access is a problem for all practitioners attempting to collect PBSC from smaller children because peripheral veins are inadequate in these patients. A suitable solution could be usage of the same catheter for both collections of PBSC and for therapeutic purposes. The use of standard tunnelled double lumen CVC for leukapheresis in children was advocated by several groups.4,5,17,18,19 The CVC size should correspond with the patient’s weight and, therefore, the CVC flow capacity could create difficulties in maintaining an interface in very small patients.4 Tunnelled silicon based catheters tend to collapse, their flow rate is significantly lower, and the ports are side by side which causes recirculation of drawn and returned (processed) blood, decreasing collection efficacy. Another possibility is to use an additional venous access only for the leukapheresis.
The p asics running rocedure described offers a higher level of stem cell support compared to autologous whole blood transfusions (UKC CSG Baby Brain protocol) and we believe that rescue using higher numbers of progenitors after high dose chemotherapy is more advantageous for the patients despite increased technical difficulties. This technique does not depend on the flow capacity of a venous access and it is not accompanied by complications arising from an apheretic procedure. As compared with leukapheresis,3,4,7,20,21,22,23 we obtained sufficient numbers of progenitors to perform from three to six PBSC supports after high dose therapy: 26.8 106 (UPN 00018), 18.0 106 (UPN 00096), and 32.2 106 (UPN 01030) CD34+ cells per kg of body weight. We succeeded in collecting and cryopreserving more than the minimum safe number of hematopoietic progenitors (>1 106 CD34+ cell for each stem cell support, and a large number of progenitors was collected in three cycles in patient 01030.
We did not observe any notable complication and the risks for children were trivial. To date, all planned PBSC reinfusions have been performed with rapid engraftment. Patient 00018 is in a very good PR 18 months after diagnosis. Patients 00096 and 01030 are in first CR lasting 6 and 8 months, respectively. We conclude that the procedure is feasible and safe. This technique seems to be an effective way for PBSC harvesting in very small children in order to provide high dose chemotherapy with intent to cure. However, this technique requires a potent mobilization regimen.