- Email: [email protected]
Costs and Outcomes with Once-Daily versus Every-6-Hour Intravenous Busulfan in Allogeneic Hematopoietic Cell Transplantation
ARTICLE IN PRESS Biol Blood Marrow Transplant 000 (2019) 1 5
Biology of Blood and Marrow Transplantation journal homepage: www.bbmt.org
Costs and Outcomes with Once-Daily versus Every-6-Hour Intravenous Busulfan in Allogeneic Hematopoietic Cell Transplantation Surbhi Singhal1,*, Ted Kim2, Patricia Jenkins2, Brandon Bassett2, D. Kathryn Tierney2,3, Andrew R. Rezvani2,3 1
Department of Medicine, Stanford University, Stanford, California Stanford Hospital and Clinics, Stanford, California 3 Department of Medicine, Division of Blood and Marrow Transplantation, Stanford University, Stanford, California 2
Article history: Received 28 June 2019 Accepted 6 September 2019 Keywords: Busulfan Hematopoietic stem cell transplantation Value-based care
A B S T R A C T The high cost of healthcare in the United States has not been consistently associated with improved health outcomes or quality of care, necessitating a focus on value-based care. We identified busulfan dosing frequency during allogeneic hematopoietic cell transplantation (HCT) conditioning as a potential target for optimization. To improve patient convenience and to decrease the cost of busulfan-based conditioning regimens, our institution changed busulfan dose frequency from every 6 hours (q6h) to once-daily (q24h). We compared costs and patient outcomes between these 2 dosing schedules. In June 2017, our institution transitioned from q6h to q24h busulfan dosing. We compared patients who received busulfan/cyclophosphamide conditioning regimens (BU/CY) for allogeneic HCT in the year before the dosing change (q6h cohort) and those who did so in the year after the dosing change (q24h cohort). The primary outcomes were differences in cost, day +90 mortality, and day +90 relapse. Between June 1, 2016, and June 1, 2018, 104 patients (median age 49 years; range, 20 to 63 years) received BU/CY before allogeneic HCT. Fifty-nine patients (57%) received q6h busulfan and 45 (43%) received q24h busulfan. There were fewer men in the q24h busulfan cohort compared with the q6h busulfan cohort (42% versus 64%; P = .024), but there were no other significant differences between the groups. There was an average annual cost savings of $19,990 per patient with q24h busulfan compared with q6h busulfan, and an annual busulfan cost savings of $899,550. There was a significantly lower day +90 mortality in the q24h busulfan cohort compared to the q6h busulfan cohort (0% versus 10%; P = .028). There were no significant differences in relapse at day +90 or in hospital length of stay. Our data indicate that i.v. busulfan dosing for allogeneic HCT conditioning is a target for improved value-based care. At our institution, patients who received q24h busulfan dosing had similar or superior outcomes compared with those receiving q6h dosing, with an average annual cost reduction of $19,990 per patient and an overall annual reduction in busulfan cost of approximately $900,000. These data support the adoption of q24h i.v. busulfan dosing as a standard of care to improve value-based care in allogeneic HCT. © 2019 American Society for Transplantation and Cellular Therapy. Published by Elsevier Inc.
INTRODUCTION The alkylating agent busulfan was initially used as an oral therapy in chronic myelogenous leukemia [1]. In 1983, busulfan was described in combination with cyclophosphamide (BU/CY) as a novel conditioning agent for hematopoietic stem cell transplantation (HCT) [2], and this combination has since been adopted as a standard HCT conditioning regimen [3]. The initial BU/CY conditioning regimen incorporated oral busulfan, the only form available at the time. Given the higher busulfan doses required in the setting of transplantation conditioning, there was concern surrounding oral bioavailability,
Financial disclosure: See Acknowledgments on page 5. *Correspondence and reprint requests: Surbhi Singhal, MD, 300 Pasteur Drive, Lane 154, Stanford, CA 94305. E-mail address: [email protected] (S. Singhal).
gastrointestinal tolerance, and patient adherence with large daily doses. Therefore, oral busulfan doses in the BU/CY regimen were split and administered throughout the day, typically 4 times daily [4]. As the BU/CY regimen gained traction, i.v. busulfan was marketed to address the concerns with oral bioavailability [5]. For what appear to be historical reasons, the i.v. busulfan formulation was also initially administered every 6 hours (q6h), despite typical daily dosing with other alkylating agents [6]. Intravenous busulfan is approved by the US Food and Drug Administration for q6h dosing, but oncedaily dosing (q24h) has been adopted by some transplant centers. In a separate literature review, we previously summarized evidence suggesting comparable efficacy and safety with q24h versus q6h i.v. busulfan regimens [5]. At our institution, we found q6h busulfan dosing to be inconvenient for patients and nursing staff, while simultaneously
https://doi.org/10.1016/j.bbmt.2019.09.008 1083-8791/© 2019 American Society for Transplantation and Cellular Therapy. Published by Elsevier Inc.
ARTICLE IN PRESS 2
S. Singhal et al. / Biol Blood Marrow Transplant 00 (2019) 1 5
increasing drug waste from vial rounding with multiple daily doses. Intravenous busulfan is supplied as a powder in 60-mg vials, must be reconstituted immediately before administration, and cannot be safely stored after reconstitution. Doses are determined to the milligram level based on body weight and pharmacokinetics and do not fall evenly into 60-mg increments. Therefore, each time a busulfan dose is reconstituted and prepared for administration, excess busulfan must be discarded. For example, for a hypothetical patient weighing 90 kg, an initial q6h i.v. busulfan dose of 72 mg (.8 mg/kg) would require the use of two 60-mg vials, with most of the second vial’s contents discarded. By administering busulfan q24h rather than q6h, drug wasting would occur only once rather than 4 times daily, with a theoretical reduction in drug waste. Value-based care is an increasing priority in our healthcare system, and we identified the q6h i.v. busulfan dose schedule in the BU/CY regimen as a potential target to reduce drug waste and associated costs. A previous Italian study found that interventions to reduce waste in the administration of 6 i.v. cancer drugs saved a total of 24,347 Euros over the course of a year, with persistent savings over the subsequent 2 years of the study period [7]. We sought to assess the comparative efficacy, safety, and total cost of q6h and q24h busulfan dosing. To our knowledge, this is the first study describing once-daily busulfan conditioning regimens to improve value-based care. METHODS Patient Cohort We conducted a retrospective preintervention and postintervention study to evaluate demographic differences, patient outcomes, and potential cost savings between the q6h and q24h busulfan regimens. This study was reviewed and approved by Stanford University’s Institutional Review Board. On June 1, 2017, our institutional standard practice for targeted i.v. busulfan in the BU/CY regimen changed from q6h to q24h dosing. High-level patient data of all patients who underwent allogeneic HCT after BU/CY conditioning regimens in the study time frame were obtained from Stanford University’s Division of Blood and Marrow Transplantation database. We included all patients who underwent allogeneic HCT after receiving a BU/CY conditioning regimen between June 1, 2016, and June 1, 2018—that is, from 1 year before through 1 year after the change in dosing schedule. Intravenous busulfan is not used as a conditioning regimen for autologous HCT at our institution. The primary outcomes examined were mean busulfan cost per patient and the incidences of overall mortality and of relapse at day +90 after allogeneic HCT. Secondary outcomes included inpatient length of stay, mean peak serum bilirubin concentration, and the incidences of sinusoidal obstruction syndrome (SOS), acute graft-versus-host disease (GVHD), and seizure. The primary clinical team diagnosed SOS based on the modified Seattle criteria. Post hoc analysis revealed differences in the number of fosaprepitant doses given in each cohort, and thus we also examined the associated cost of fosaprepitant. A manual chart review was performed to determine the size of each weight-based and pharmacokinetic-guided busulfan dose administered, the number of fosaprepitant doses given to each patient during the inpatient admission, and the incidences of relapse and mortality at day +90 after transplantation.
Conditioning Regimen Intravenous busulfan was pharmacokinetically targeted in all patients, using weight-based and pharmacokinetic-guided doses [8]. Actual body weight was used to calculate the weight-based doses, while the pharmacokinetic-guided doses were adjusted for hepatic clearance and drug interactions to achieve a prespecified target plasma concentration-time area under the curve, with the assistance of a busulfan pharmacokinetics laboratory. For logistic reasons, our institutional standard is to use 5 weight-based doses and 11 pharmacokinetic-guided doses of busulfan (total of 16 doses) in the q6h regimen, whereas the q24h regimen required only 2 weight-based doses and 2 pharmacokinetic-guided doses (total of 4 doses). Patients receiving the q6h regimen were conditioned with busulfan .8 mg/kg i.v. q6h on days -7 through -4 before allogeneic HCT (with pharmacokinetic targeting of doses 6 through 16), followed by cyclophosphamide 60 mg/kg/day on days -3 and -2 before allogeneic HCT. Patients receiving the q24h regimen were conditioned with busulfan 3.2 mg/kg i.v. on days -7 through -4 (with pharmacokinetic targeting of doses 3 and 4), followed by cyclophosphamide 60 mg/ kg/day on days -3 and -2 before allogeneic HCT.
For patients receiving the q6h regimen, pharmacokinetic samples were drawn following dose 1 at the end of the infusion, at 15 minutes and 30 minutes after the end of the infusion, and at 4, 5, and 6 hours after the start of the infusion. For patients receiving q24h busulfan, pharmacokinetic samples were drawn following dose 1 at the end of the infusion, at 15 minutes after the end of the infusion, and at 4, 5, 6, and 8 hours after the start of the infusion. The target plasma concentration-time area under the curve was 4.8 to 5.4 mg*h/L after 1 q6h dose and 19.2 to 21.6 mg*h/L after 1 q24h dose. Our institutional standard is to give ursodeoxycholic acid as SOS prophylaxis and phenytoin as seizure prophylaxis, using the same dose schedule in both cohorts. Anecdotally, the nurses administering q24h busulfan noted increased nausea in their patients, and thus we added prophylactic fosaprepitant to the q24h regimen; fosaprepitant for nausea was available as needed for both the q24h and q6h cohorts. Prophylaxis against GVHD consisted of tacrolimus and methotrexate (15 mg/m2 on day +1 and 10 mg/m2 on days +3, +6, and +11). Actual body weight was used for dosing unless it exceeded ideal body weight by 15 kg, in which case-adjusted body weight was used. Antimicrobial prophylaxis and treatment and supportive care were administered according to our institutional standard practice and did not differ between the 2 cohorts. The cost of each busulfan regimen was calculated using the April 2017 average wholesale price of US $2498.75 for a 60-mg vial of busulfan [6]. Statistical Analysis Continuous variables were analyzed using the 2-sided Student t test, whereas categorical variables were analyzed using Pearson’s chi-squared test.
RESULTS Demographics and Transplantation-Related Characteristics Between June 1, 2016, and June 1, 2018, 104 patients (age 49, 20-63 years, 55% male) received conditioning BU/CY regimens before allogeneic HCT at Stanford University (Table 1). Fifty-nine of these patients (57%) received q6h busulfan, and 45 (43%) received q24h busulfan. The most common indication for allogeneic HCT in these cohorts was acute myelogenous leukemia (AML; 40%), followed by myelodysplastic syndrome (35%). Fifty-seven patients (55%) were in complete remission at the time of allogeneic HCT. Nineteen of 25 patients with AML in the q6h cohort had achieved complete remission (76%), compared with 16 of 17 patients with AML who achieved remission in the q24h cohort (94%; P = .122). Growth factor-mobilized peripheral blood was the stem cell source for 60 patients (58%), and the other 44 patients (42%) received bone marrow allografts. Forty-one patients (39%) had an HLAidentical sibling donor, and 2 patients had an HLA-mismatched related donor, whereas the remaining patients had HLAmatched or HLA-mismatched unrelated donors. There were fewer men in the q24h busulfan cohort compared with the q6h busulfan cohort (42% versus 64%; P = .024), but there were no other significant differences between the treatment groups (Table 1). Transplantation-Related Toxicities and Outcomes All but 1 patient engrafted following allogeneic HCT, with 1 graft failure in the q6h busulfan cohort. Mortality at +90 after allogeneic HCT was significantly lower in the q24h busulfan cohort compared with the q6h busulfan cohort (0% versus 10%; P = .028) (Table 2). Of the 6 patients who died before day +90 (all in the q6h cohort), 3 died of acute GVHD, 2 died of recurrent malignancies, and 1 died due to SOS. There were no significant differences between the q6h and q24h cohorts in mean peak bilirubin concentration (2.6 versus 1.7; P= .268) or in the incidence of SOS (9% versus 18%; P = .324). Acute GVHD grade II-IV occurred in 24 patients (41%) in the q6h cohort and in 9 patients (20%) in the q24h cohort, a difference that did not reach statistical significance. There were no significant differences in the incidence of relapse at day +90 after HCT or in mean length of stay during HCT inpatient admission (31.3 days versus 29.1 days; P = .304) (Table 2).
ARTICLE IN PRESS S. Singhal et al. / Biol Blood Marrow Transplant 00 (2019) 1 5
3
Table 1 Demographic and Transplantation-Related Characteristics Characteristics Number of patients Age, yr, mean (SD)
Total
q24h Busulfan
q6h Busulfan
104
45
59
P Value
49.0 (10.9)
49.4 (11.6)
48.7 (10.5)
.765
Male sex, n (%)
57 (55)
19 (42)
38 (64)
.024
BMI, mean (SD)
25.7 (3.2)
25.7 (3.6)
25.7 (2.9)
.989
5 (8.5)
.425 .636
Obesity (BMI 30), n (%)
11 (11)
6 (13.3)
Hematologic disease, n (%) AML
42 (40)
17 (38)
25 (42)
Myelodysplastic syndrome
36 (35)
17 (38)
19 (32)
.554
Acute lymphoblastic leukemia
13 (13)
4 (9)
9 (15)
.331
Chronic myelogenous leukemia
9 (9)
5 (11)
4 (7)
.436
Hodgkin lymphoma
1 (1)
1 (2)
0 (0)
.250 .380
Non-Hodgkin lymphoma
1 (1)
0 (0)
1 (2)
Idiopathic erythrocyte aplasia
1 (1)
0 (0)
1 (2)
.380
Other
1 (1)
1 (2)
0 (0)
.250
Complete remission
57 (55)
24 (53)
33 (57)
.718
No response/stable disease
23 (22)
12 (27)
11 (19)
.352
Other disease status
23 (22)
9 (20)
14 (24)
.617
Disease status at transplantation, n (%)
Product, n (%) Peripheral blood
60 (58)
30 (67)
30 (51)
.106
Bone marrow
44 (42)
15 (33)
29 (49)
.106
41 (39)
21 (47)
20 (34)
.187
2 (2)
1 (2)
1 (2)
.846
URD identical
50 (48)
21 (47)
29 (49)
.802
URD mismatched
11 (11)
2 (4)
9 (15)
.076
Donor, n (%) HLA-identical sibling HLA-mismatched relative
Significant P values are in bold type. BMI indicates body mass index; URD, unrelated donor.
Cost-Savings Analysis The mean cost was significantly lower for the q24h busulfan conditioning regimen compared with the q6h regimen ($46,976.50 per patient versus $66,966.50 per patient; P < .001) (Figure 1). This reduction in busulfan waste resulted in a mean cost difference of $19,990 per patient. Over the course of 1 year, use of the q24h regimen in 45 patients resulted in a total busulfan cost savings of $899,550.00. On average, patients in the q24h cohort received 1 more dose of fosaprepitant than those in the q6h cohort (2.49 versus 1.49; P < .001), associated with a mean cost difference of $400.50 per patient. Despite the increased use of fosaprepitant, there continued to be a statistically significant
cost savings of $19,589.50 (P < .001) with the q24h busulfan regimen compared with the q6h regimen (Figure 1). DISCUSSION The transition from q6h to q24h busulfan dosing in the BU/ CY regimen for allogeneic HCT is straightforward to implement, results in substantial cost savings, and is neutral or positive regarding patient outcomes. In this sequential cohort study, we observed a reduction in busulfan cost of approximately $20,000 per patient and $900,000 annually after switching to q24h busulfan dosing, as well as a significant decrease in day +90 mortality. To our knowledge, this is the
Table 2 Transplantation-Related Toxicities and Outcomes Outcomes Number of patients Total bilirubin peak, mean (SD) SOS, n/N (%) Seizure, n (%)
Total 104
q24h Busulfan
q6h Busulfan
45
59
2.2 (4.6)
1.7 (3.0)
7/65 (11)
2/11 (18)
2.6 (5.4)
P Value .268
5/54 (9)
.324 .212
2 (2)
0 (0)
2 (4)
57 (66)
14 (52)
43 (73)
.056
Grade I
24 (42)
5 (36)
19 (44)
—
Grade II
22 (39)
7 (50)
15 (35)
—
Grade III
8 (14)
2 (14)
6 (14)
—
Grade IV
3 (5)
0 (0)
3 (7)
—
30.4 (10.4)
29.1 (12.4)
31.3 (8.5)
.304
+90-day relapse, n (%)
6 (6)
1 (2)
5 (8)
.175
+90-day mortality, n (%)
6 (6)
0 (0)
6 (10)
.028
Acute GVHD, n (%)
Length of stay, d, mean (SD)
Significant P values are in bold type.
ARTICLE IN PRESS 4
S. Singhal et al. / Biol Blood Marrow Transplant 00 (2019) 1 5
Figure 1. Median cost of busulfan and fosaprepitant per patient by month. Cost of busulfan was calculated based on the April 2017 average wholesale price (AWP) of $2498.75 per vial, and the cost of fosaprepitant was calculated based on the April 2018 AWP of $401.56 per dose. In June 2017, our institution transitioned from q6h to q24h dosing. This transition was associated with a significantly increased number of mean doses of fosaprepitant per patient (from 1.5 to 2.5 doses per patient; P < .001), but there continued to be a significant cost savings of $19,589.50 per patient (P < .001).
first published study describing the cost and value benefit of q24h busulfan compared with q6h busulfan. The cost savings with q24h busulfan dosing schedule are driven primarily by reduced drug waste. The q24h regimen uses fewer total doses of busulfan than the q6h regimen (4 versus 16), which minimizes the number of potential doses between 60-mg intervals and reduces the number of partially unused vials. We noted a significant difference in the number of fosaprepitant doses between the 2 cohorts, likely driven by the addition of prophylactic fosaprepitant to the q24h regimen but not to the q6h regimen. Although the inclusion of fosaprepitant increased the cost of the regimen, the increased fosaprepitant cost was dwarfed by the cost savings associated with the q24h busulfan regimen (Figure 1). Beyond the direct drug-related cost savings, there are additional less tangible but very real benefits for the q24h regimen over the q6h regimen. Although our study was not designed to capture the net impact of less frequent drug administration on nursing and pharmacy resources, we postulate the q24h regimen would require less pharmacy compounding time and less time for nursing administration and patient assessment. This could be particularly meaningful considering that a previous study of nurses found an association between nursing workload and likelihood of medication error [9]. Given its ease of administration, the q24h busulfan regimen has practical implications in reducing provider workload and improving patient satisfaction. We also noted a lower day 90+ mortality in the q24h busulfan group compared with the q6h busulfan group. Previous studies did not demonstrate differences in relapse rate, disease-specific survival, or overall survival between the q24h and q6h busulfan regimens [10,11]. Given our study’s retrospective clinical review design, we are unable to determine whether the q24h busulfan caused the lower mortality rates, and we must acknowledge the possibility that the lower
mortality rate may reflect unmeasured confounders that we were unable to identify. Nonetheless, the switch to q24h busulfan was at least neutral, and possibly beneficial, regarding patient outcomes. A 2016 study of adults with AML, chronic myelogenous leukemia, and myelodysplastic syndrome compared q6h and q24h busulfan, in combination with either cyclophosphamide or fludarabine, and found an increased incidence of acute GVHD grade II-IV and chronic GVHD in the q24h busulfan group. However, this finding was attributed to differences in graft source and donor type between the groups, rather than to the conditioning regimens [3]. In our study, we observed lower rates of acute GVHD in the q24h busulfan group that approached statistical significance, but as with the lower morality in the q24h cohort, this may reflect confounders not accounted for in the study. Several other studies have evaluated q6h versus q24h i.v. busulfan with cyclophosphamide conditioning regimens and found no differences in toxicity, including no difference in acute GVHD [10-12]. The limitations of this study are those inherent to a retrospective preintervention and postintervention design. This limitation is especially important when interpreting the observed mortality difference between the groups, which we speculate may be secondary to other confounders and not necessarily from the change in busulfan dosing regimen. One potential confounder is the difference in the interval between the last dose of busulfan and the first dose of cyclophosphamide, with some evidence suggesting that this interval may affect transplantation outcomes [13]. At our institution, the interval between the start of the final busulfan dose and the start of the first cyclophosphamide dose was approximately 12 hours with q6h busulfan and approximately 30 hours with q24h busulfan. In this retrospective study, we did not have a reliable means of assessing the intangible benefits of the q24h
ARTICLE IN PRESS S. Singhal et al. / Biol Blood Marrow Transplant 00 (2019) 1 5
regimen or comparing the total transplantation cost between the 2 cohorts. In addition, a portion of the cost savings could be accounted for by differences in body weight and pharmacokinetics between the 2 cohorts. Although the April 2017 average wholesale price of busulfan was used to extrapolate costs between the 2 groups, this does not reflect potential changes in drug costs over time. Future interventions could include i.v. busulfan administration in the outpatient setting and institutional development of a population pharmacokinetic model to allow for pharmacokinetic-guided busulfan dosing, as discussed by the American Society for Blood and Marrow Transplantation busulfan practice guidelines [8]. Future studies could seek to explore a potential cost impact from less tangible benefits and potential clinical impact on SOS, GVHD, and overall survival. In summary, this retrospective study strongly supports the adoption of q24h busulfan dosing for the BU/CY regimen as a means of improving value-based care, reducing drug waste, and reducing provider workload in allogeneic HCT. These data indicate that once-daily i.v. busulfan is comparable or superior to the Food and Drug Administration-approved standard q6h busulfan dose schedule in both safety and efficacy, and results in substantial cost savings.
ACKNOWLEDGMENTS The authors thank Linda Elder and her team, who maintain the patient database for the Stanford Division of Blood and Marrow Transplantation. Financial disclosure: A.R.R. participated on one-time ad hoc scientific advisory boards for Nohla Therapeutics and for Kaleido, both in 2018. He receives research support from AbbVie for an ongoing clinical trial and has served as a medical expert witness for the US Department of Justice. His brother is an employee of Johnson & Johnson. Conflict of interest statement: There are no conflicts of interest to report.
5
REFERENCES 1. Chronic granulocytic leukaemia: comparison of radiotherapy and busulfan therapy. Report of the Medical Research Council’s working party for therapeutic trials in leukaemia. Br Med J. 1968;1:201–208. 2. Santos GW, Tutschka PJ, Brookmeyer R, et al. Marrow transplantation for acute nonlymphocytic leukemia after treatment with busulfan and cyclophosphamide. N Engl J Med. 1983;309:1347–1353. 3. Pasquini MC, Le-Rademacher J, Zhu X, et al. Intravenous busulfan-based myeloablative conditioning regimens prior to hematopoietic cell transplantation for hematologic malignancies. Biol Blood Marrow Transplant. 2016;22:1424–1430. 4. Chandy M, Balasubramanian P, Ramachandran SV, et al. Randomized trial of two different conditioning regimens for bone marrow transplantation in thalassemia the role of busulfan pharmacokinetics in determining outcome. Bone Marrow Transplant. 2005;36:839–845. 5. Gong CL, Rezvani AR, Kim T, Studdert AL. Busulfan dosing literature review. J Blood Lymph. 2018;8:204. 6. Micromedex. Busulfan. Truven Health Analytics. https://www-microme dexsolutions-com.laneproxy.stanford.edu/micromedex2/librarian/PFDe faultActionId/evidencexpert.DoIntegratedSearch?navitem=topHome&is ToolPage=true#. Published 2019. Accessed January 25, 2019. 7. Fasola G, Aprile G, Marini L, Follador A, Mansutti M, Miscoria M. Drug waste minimization as an effective strategy of cost-containment in oncology. BMC Health Serv Res. 2014;14:57. 8. Palmer J, McCune JS, Perales MA, et al. Personalizing busulfan-based conditioning: considerations from the American Society for Blood and Marrow Transplantation Practice Guidelines Committee. Biol Blood Marrow Transplant. 2016;22:1915–1925. 9. Holden RJ, Scanlon MC, Patel NR, et al. A human factors framework and study of the effect of nursing workload on patient safety and employee quality of working life. BMJ Qual Saf. 2011;20:15–24. 10. Hill BT, Rybicki L, Carlstrom KD, et al. Daily weight-based busulfan with cyclophosphamide and etoposide produces comparable outcomes to fourtimes-daily busulfan dosing for lymphoma patients undergoing autologous stem cell transplantation. Biol Blood Marrow Transplant. 2016;22: 1588–1595. 11. Ryu SG, Lee JH, Choi SJ, et al. Randomized comparison of four-times-daily versus once-daily intravenous busulfan in conditioning therapy for hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2007;13: 1095–1105. 12. Zhang H, Graiser M, Hutcherson DA, et al. Pharmacokinetic-directed highdose busulfan combined with cyclophosphamide and etoposide results in predictable drug levels and durable long-term survival in lymphoma patients undergoing autologous stem cell transplantation. Biol Blood Marrow Transplant. 2012;18:1287–1294. € n O, Hassan Z, Nilsson C, Oberg 13. Hassan M, Ljungman P, Ringde G. The influence of busulphan on the pharmacokinetics of cyclophosphamide and its 4-hydroxy metabolite: time interval effect on therapeutic efficacy and toxicity. Ther Drug Monit. 1999;21:434.
Biology of Blood and Marrow Transplantation journal homepage: www.bbmt.org
Costs and Outcomes with Once-Daily versus Every-6-Hour Intravenous Busulfan in Allogeneic Hematopoietic Cell Transplantation Surbhi Singhal1,*, Ted Kim2, Patricia Jenkins2, Brandon Bassett2, D. Kathryn Tierney2,3, Andrew R. Rezvani2,3 1
Department of Medicine, Stanford University, Stanford, California Stanford Hospital and Clinics, Stanford, California 3 Department of Medicine, Division of Blood and Marrow Transplantation, Stanford University, Stanford, California 2
Article history: Received 28 June 2019 Accepted 6 September 2019 Keywords: Busulfan Hematopoietic stem cell transplantation Value-based care
A B S T R A C T The high cost of healthcare in the United States has not been consistently associated with improved health outcomes or quality of care, necessitating a focus on value-based care. We identified busulfan dosing frequency during allogeneic hematopoietic cell transplantation (HCT) conditioning as a potential target for optimization. To improve patient convenience and to decrease the cost of busulfan-based conditioning regimens, our institution changed busulfan dose frequency from every 6 hours (q6h) to once-daily (q24h). We compared costs and patient outcomes between these 2 dosing schedules. In June 2017, our institution transitioned from q6h to q24h busulfan dosing. We compared patients who received busulfan/cyclophosphamide conditioning regimens (BU/CY) for allogeneic HCT in the year before the dosing change (q6h cohort) and those who did so in the year after the dosing change (q24h cohort). The primary outcomes were differences in cost, day +90 mortality, and day +90 relapse. Between June 1, 2016, and June 1, 2018, 104 patients (median age 49 years; range, 20 to 63 years) received BU/CY before allogeneic HCT. Fifty-nine patients (57%) received q6h busulfan and 45 (43%) received q24h busulfan. There were fewer men in the q24h busulfan cohort compared with the q6h busulfan cohort (42% versus 64%; P = .024), but there were no other significant differences between the groups. There was an average annual cost savings of $19,990 per patient with q24h busulfan compared with q6h busulfan, and an annual busulfan cost savings of $899,550. There was a significantly lower day +90 mortality in the q24h busulfan cohort compared to the q6h busulfan cohort (0% versus 10%; P = .028). There were no significant differences in relapse at day +90 or in hospital length of stay. Our data indicate that i.v. busulfan dosing for allogeneic HCT conditioning is a target for improved value-based care. At our institution, patients who received q24h busulfan dosing had similar or superior outcomes compared with those receiving q6h dosing, with an average annual cost reduction of $19,990 per patient and an overall annual reduction in busulfan cost of approximately $900,000. These data support the adoption of q24h i.v. busulfan dosing as a standard of care to improve value-based care in allogeneic HCT. © 2019 American Society for Transplantation and Cellular Therapy. Published by Elsevier Inc.
INTRODUCTION The alkylating agent busulfan was initially used as an oral therapy in chronic myelogenous leukemia [1]. In 1983, busulfan was described in combination with cyclophosphamide (BU/CY) as a novel conditioning agent for hematopoietic stem cell transplantation (HCT) [2], and this combination has since been adopted as a standard HCT conditioning regimen [3]. The initial BU/CY conditioning regimen incorporated oral busulfan, the only form available at the time. Given the higher busulfan doses required in the setting of transplantation conditioning, there was concern surrounding oral bioavailability,
Financial disclosure: See Acknowledgments on page 5. *Correspondence and reprint requests: Surbhi Singhal, MD, 300 Pasteur Drive, Lane 154, Stanford, CA 94305. E-mail address: [email protected] (S. Singhal).
gastrointestinal tolerance, and patient adherence with large daily doses. Therefore, oral busulfan doses in the BU/CY regimen were split and administered throughout the day, typically 4 times daily [4]. As the BU/CY regimen gained traction, i.v. busulfan was marketed to address the concerns with oral bioavailability [5]. For what appear to be historical reasons, the i.v. busulfan formulation was also initially administered every 6 hours (q6h), despite typical daily dosing with other alkylating agents [6]. Intravenous busulfan is approved by the US Food and Drug Administration for q6h dosing, but oncedaily dosing (q24h) has been adopted by some transplant centers. In a separate literature review, we previously summarized evidence suggesting comparable efficacy and safety with q24h versus q6h i.v. busulfan regimens [5]. At our institution, we found q6h busulfan dosing to be inconvenient for patients and nursing staff, while simultaneously
https://doi.org/10.1016/j.bbmt.2019.09.008 1083-8791/© 2019 American Society for Transplantation and Cellular Therapy. Published by Elsevier Inc.
ARTICLE IN PRESS 2
S. Singhal et al. / Biol Blood Marrow Transplant 00 (2019) 1 5
increasing drug waste from vial rounding with multiple daily doses. Intravenous busulfan is supplied as a powder in 60-mg vials, must be reconstituted immediately before administration, and cannot be safely stored after reconstitution. Doses are determined to the milligram level based on body weight and pharmacokinetics and do not fall evenly into 60-mg increments. Therefore, each time a busulfan dose is reconstituted and prepared for administration, excess busulfan must be discarded. For example, for a hypothetical patient weighing 90 kg, an initial q6h i.v. busulfan dose of 72 mg (.8 mg/kg) would require the use of two 60-mg vials, with most of the second vial’s contents discarded. By administering busulfan q24h rather than q6h, drug wasting would occur only once rather than 4 times daily, with a theoretical reduction in drug waste. Value-based care is an increasing priority in our healthcare system, and we identified the q6h i.v. busulfan dose schedule in the BU/CY regimen as a potential target to reduce drug waste and associated costs. A previous Italian study found that interventions to reduce waste in the administration of 6 i.v. cancer drugs saved a total of 24,347 Euros over the course of a year, with persistent savings over the subsequent 2 years of the study period [7]. We sought to assess the comparative efficacy, safety, and total cost of q6h and q24h busulfan dosing. To our knowledge, this is the first study describing once-daily busulfan conditioning regimens to improve value-based care. METHODS Patient Cohort We conducted a retrospective preintervention and postintervention study to evaluate demographic differences, patient outcomes, and potential cost savings between the q6h and q24h busulfan regimens. This study was reviewed and approved by Stanford University’s Institutional Review Board. On June 1, 2017, our institutional standard practice for targeted i.v. busulfan in the BU/CY regimen changed from q6h to q24h dosing. High-level patient data of all patients who underwent allogeneic HCT after BU/CY conditioning regimens in the study time frame were obtained from Stanford University’s Division of Blood and Marrow Transplantation database. We included all patients who underwent allogeneic HCT after receiving a BU/CY conditioning regimen between June 1, 2016, and June 1, 2018—that is, from 1 year before through 1 year after the change in dosing schedule. Intravenous busulfan is not used as a conditioning regimen for autologous HCT at our institution. The primary outcomes examined were mean busulfan cost per patient and the incidences of overall mortality and of relapse at day +90 after allogeneic HCT. Secondary outcomes included inpatient length of stay, mean peak serum bilirubin concentration, and the incidences of sinusoidal obstruction syndrome (SOS), acute graft-versus-host disease (GVHD), and seizure. The primary clinical team diagnosed SOS based on the modified Seattle criteria. Post hoc analysis revealed differences in the number of fosaprepitant doses given in each cohort, and thus we also examined the associated cost of fosaprepitant. A manual chart review was performed to determine the size of each weight-based and pharmacokinetic-guided busulfan dose administered, the number of fosaprepitant doses given to each patient during the inpatient admission, and the incidences of relapse and mortality at day +90 after transplantation.
Conditioning Regimen Intravenous busulfan was pharmacokinetically targeted in all patients, using weight-based and pharmacokinetic-guided doses [8]. Actual body weight was used to calculate the weight-based doses, while the pharmacokinetic-guided doses were adjusted for hepatic clearance and drug interactions to achieve a prespecified target plasma concentration-time area under the curve, with the assistance of a busulfan pharmacokinetics laboratory. For logistic reasons, our institutional standard is to use 5 weight-based doses and 11 pharmacokinetic-guided doses of busulfan (total of 16 doses) in the q6h regimen, whereas the q24h regimen required only 2 weight-based doses and 2 pharmacokinetic-guided doses (total of 4 doses). Patients receiving the q6h regimen were conditioned with busulfan .8 mg/kg i.v. q6h on days -7 through -4 before allogeneic HCT (with pharmacokinetic targeting of doses 6 through 16), followed by cyclophosphamide 60 mg/kg/day on days -3 and -2 before allogeneic HCT. Patients receiving the q24h regimen were conditioned with busulfan 3.2 mg/kg i.v. on days -7 through -4 (with pharmacokinetic targeting of doses 3 and 4), followed by cyclophosphamide 60 mg/ kg/day on days -3 and -2 before allogeneic HCT.
For patients receiving the q6h regimen, pharmacokinetic samples were drawn following dose 1 at the end of the infusion, at 15 minutes and 30 minutes after the end of the infusion, and at 4, 5, and 6 hours after the start of the infusion. For patients receiving q24h busulfan, pharmacokinetic samples were drawn following dose 1 at the end of the infusion, at 15 minutes after the end of the infusion, and at 4, 5, 6, and 8 hours after the start of the infusion. The target plasma concentration-time area under the curve was 4.8 to 5.4 mg*h/L after 1 q6h dose and 19.2 to 21.6 mg*h/L after 1 q24h dose. Our institutional standard is to give ursodeoxycholic acid as SOS prophylaxis and phenytoin as seizure prophylaxis, using the same dose schedule in both cohorts. Anecdotally, the nurses administering q24h busulfan noted increased nausea in their patients, and thus we added prophylactic fosaprepitant to the q24h regimen; fosaprepitant for nausea was available as needed for both the q24h and q6h cohorts. Prophylaxis against GVHD consisted of tacrolimus and methotrexate (15 mg/m2 on day +1 and 10 mg/m2 on days +3, +6, and +11). Actual body weight was used for dosing unless it exceeded ideal body weight by 15 kg, in which case-adjusted body weight was used. Antimicrobial prophylaxis and treatment and supportive care were administered according to our institutional standard practice and did not differ between the 2 cohorts. The cost of each busulfan regimen was calculated using the April 2017 average wholesale price of US $2498.75 for a 60-mg vial of busulfan [6]. Statistical Analysis Continuous variables were analyzed using the 2-sided Student t test, whereas categorical variables were analyzed using Pearson’s chi-squared test.
RESULTS Demographics and Transplantation-Related Characteristics Between June 1, 2016, and June 1, 2018, 104 patients (age 49, 20-63 years, 55% male) received conditioning BU/CY regimens before allogeneic HCT at Stanford University (Table 1). Fifty-nine of these patients (57%) received q6h busulfan, and 45 (43%) received q24h busulfan. The most common indication for allogeneic HCT in these cohorts was acute myelogenous leukemia (AML; 40%), followed by myelodysplastic syndrome (35%). Fifty-seven patients (55%) were in complete remission at the time of allogeneic HCT. Nineteen of 25 patients with AML in the q6h cohort had achieved complete remission (76%), compared with 16 of 17 patients with AML who achieved remission in the q24h cohort (94%; P = .122). Growth factor-mobilized peripheral blood was the stem cell source for 60 patients (58%), and the other 44 patients (42%) received bone marrow allografts. Forty-one patients (39%) had an HLAidentical sibling donor, and 2 patients had an HLA-mismatched related donor, whereas the remaining patients had HLAmatched or HLA-mismatched unrelated donors. There were fewer men in the q24h busulfan cohort compared with the q6h busulfan cohort (42% versus 64%; P = .024), but there were no other significant differences between the treatment groups (Table 1). Transplantation-Related Toxicities and Outcomes All but 1 patient engrafted following allogeneic HCT, with 1 graft failure in the q6h busulfan cohort. Mortality at +90 after allogeneic HCT was significantly lower in the q24h busulfan cohort compared with the q6h busulfan cohort (0% versus 10%; P = .028) (Table 2). Of the 6 patients who died before day +90 (all in the q6h cohort), 3 died of acute GVHD, 2 died of recurrent malignancies, and 1 died due to SOS. There were no significant differences between the q6h and q24h cohorts in mean peak bilirubin concentration (2.6 versus 1.7; P= .268) or in the incidence of SOS (9% versus 18%; P = .324). Acute GVHD grade II-IV occurred in 24 patients (41%) in the q6h cohort and in 9 patients (20%) in the q24h cohort, a difference that did not reach statistical significance. There were no significant differences in the incidence of relapse at day +90 after HCT or in mean length of stay during HCT inpatient admission (31.3 days versus 29.1 days; P = .304) (Table 2).
ARTICLE IN PRESS S. Singhal et al. / Biol Blood Marrow Transplant 00 (2019) 1 5
3
Table 1 Demographic and Transplantation-Related Characteristics Characteristics Number of patients Age, yr, mean (SD)
Total
q24h Busulfan
q6h Busulfan
104
45
59
P Value
49.0 (10.9)
49.4 (11.6)
48.7 (10.5)
.765
Male sex, n (%)
57 (55)
19 (42)
38 (64)
.024
BMI, mean (SD)
25.7 (3.2)
25.7 (3.6)
25.7 (2.9)
.989
5 (8.5)
.425 .636
Obesity (BMI 30), n (%)
11 (11)
6 (13.3)
Hematologic disease, n (%) AML
42 (40)
17 (38)
25 (42)
Myelodysplastic syndrome
36 (35)
17 (38)
19 (32)
.554
Acute lymphoblastic leukemia
13 (13)
4 (9)
9 (15)
.331
Chronic myelogenous leukemia
9 (9)
5 (11)
4 (7)
.436
Hodgkin lymphoma
1 (1)
1 (2)
0 (0)
.250 .380
Non-Hodgkin lymphoma
1 (1)
0 (0)
1 (2)
Idiopathic erythrocyte aplasia
1 (1)
0 (0)
1 (2)
.380
Other
1 (1)
1 (2)
0 (0)
.250
Complete remission
57 (55)
24 (53)
33 (57)
.718
No response/stable disease
23 (22)
12 (27)
11 (19)
.352
Other disease status
23 (22)
9 (20)
14 (24)
.617
Disease status at transplantation, n (%)
Product, n (%) Peripheral blood
60 (58)
30 (67)
30 (51)
.106
Bone marrow
44 (42)
15 (33)
29 (49)
.106
41 (39)
21 (47)
20 (34)
.187
2 (2)
1 (2)
1 (2)
.846
URD identical
50 (48)
21 (47)
29 (49)
.802
URD mismatched
11 (11)
2 (4)
9 (15)
.076
Donor, n (%) HLA-identical sibling HLA-mismatched relative
Significant P values are in bold type. BMI indicates body mass index; URD, unrelated donor.
Cost-Savings Analysis The mean cost was significantly lower for the q24h busulfan conditioning regimen compared with the q6h regimen ($46,976.50 per patient versus $66,966.50 per patient; P < .001) (Figure 1). This reduction in busulfan waste resulted in a mean cost difference of $19,990 per patient. Over the course of 1 year, use of the q24h regimen in 45 patients resulted in a total busulfan cost savings of $899,550.00. On average, patients in the q24h cohort received 1 more dose of fosaprepitant than those in the q6h cohort (2.49 versus 1.49; P < .001), associated with a mean cost difference of $400.50 per patient. Despite the increased use of fosaprepitant, there continued to be a statistically significant
cost savings of $19,589.50 (P < .001) with the q24h busulfan regimen compared with the q6h regimen (Figure 1). DISCUSSION The transition from q6h to q24h busulfan dosing in the BU/ CY regimen for allogeneic HCT is straightforward to implement, results in substantial cost savings, and is neutral or positive regarding patient outcomes. In this sequential cohort study, we observed a reduction in busulfan cost of approximately $20,000 per patient and $900,000 annually after switching to q24h busulfan dosing, as well as a significant decrease in day +90 mortality. To our knowledge, this is the
Table 2 Transplantation-Related Toxicities and Outcomes Outcomes Number of patients Total bilirubin peak, mean (SD) SOS, n/N (%) Seizure, n (%)
Total 104
q24h Busulfan
q6h Busulfan
45
59
2.2 (4.6)
1.7 (3.0)
7/65 (11)
2/11 (18)
2.6 (5.4)
P Value .268
5/54 (9)
.324 .212
2 (2)
0 (0)
2 (4)
57 (66)
14 (52)
43 (73)
.056
Grade I
24 (42)
5 (36)
19 (44)
—
Grade II
22 (39)
7 (50)
15 (35)
—
Grade III
8 (14)
2 (14)
6 (14)
—
Grade IV
3 (5)
0 (0)
3 (7)
—
30.4 (10.4)
29.1 (12.4)
31.3 (8.5)
.304
+90-day relapse, n (%)
6 (6)
1 (2)
5 (8)
.175
+90-day mortality, n (%)
6 (6)
0 (0)
6 (10)
.028
Acute GVHD, n (%)
Length of stay, d, mean (SD)
Significant P values are in bold type.
ARTICLE IN PRESS 4
S. Singhal et al. / Biol Blood Marrow Transplant 00 (2019) 1 5
Figure 1. Median cost of busulfan and fosaprepitant per patient by month. Cost of busulfan was calculated based on the April 2017 average wholesale price (AWP) of $2498.75 per vial, and the cost of fosaprepitant was calculated based on the April 2018 AWP of $401.56 per dose. In June 2017, our institution transitioned from q6h to q24h dosing. This transition was associated with a significantly increased number of mean doses of fosaprepitant per patient (from 1.5 to 2.5 doses per patient; P < .001), but there continued to be a significant cost savings of $19,589.50 per patient (P < .001).
first published study describing the cost and value benefit of q24h busulfan compared with q6h busulfan. The cost savings with q24h busulfan dosing schedule are driven primarily by reduced drug waste. The q24h regimen uses fewer total doses of busulfan than the q6h regimen (4 versus 16), which minimizes the number of potential doses between 60-mg intervals and reduces the number of partially unused vials. We noted a significant difference in the number of fosaprepitant doses between the 2 cohorts, likely driven by the addition of prophylactic fosaprepitant to the q24h regimen but not to the q6h regimen. Although the inclusion of fosaprepitant increased the cost of the regimen, the increased fosaprepitant cost was dwarfed by the cost savings associated with the q24h busulfan regimen (Figure 1). Beyond the direct drug-related cost savings, there are additional less tangible but very real benefits for the q24h regimen over the q6h regimen. Although our study was not designed to capture the net impact of less frequent drug administration on nursing and pharmacy resources, we postulate the q24h regimen would require less pharmacy compounding time and less time for nursing administration and patient assessment. This could be particularly meaningful considering that a previous study of nurses found an association between nursing workload and likelihood of medication error [9]. Given its ease of administration, the q24h busulfan regimen has practical implications in reducing provider workload and improving patient satisfaction. We also noted a lower day 90+ mortality in the q24h busulfan group compared with the q6h busulfan group. Previous studies did not demonstrate differences in relapse rate, disease-specific survival, or overall survival between the q24h and q6h busulfan regimens [10,11]. Given our study’s retrospective clinical review design, we are unable to determine whether the q24h busulfan caused the lower mortality rates, and we must acknowledge the possibility that the lower
mortality rate may reflect unmeasured confounders that we were unable to identify. Nonetheless, the switch to q24h busulfan was at least neutral, and possibly beneficial, regarding patient outcomes. A 2016 study of adults with AML, chronic myelogenous leukemia, and myelodysplastic syndrome compared q6h and q24h busulfan, in combination with either cyclophosphamide or fludarabine, and found an increased incidence of acute GVHD grade II-IV and chronic GVHD in the q24h busulfan group. However, this finding was attributed to differences in graft source and donor type between the groups, rather than to the conditioning regimens [3]. In our study, we observed lower rates of acute GVHD in the q24h busulfan group that approached statistical significance, but as with the lower morality in the q24h cohort, this may reflect confounders not accounted for in the study. Several other studies have evaluated q6h versus q24h i.v. busulfan with cyclophosphamide conditioning regimens and found no differences in toxicity, including no difference in acute GVHD [10-12]. The limitations of this study are those inherent to a retrospective preintervention and postintervention design. This limitation is especially important when interpreting the observed mortality difference between the groups, which we speculate may be secondary to other confounders and not necessarily from the change in busulfan dosing regimen. One potential confounder is the difference in the interval between the last dose of busulfan and the first dose of cyclophosphamide, with some evidence suggesting that this interval may affect transplantation outcomes [13]. At our institution, the interval between the start of the final busulfan dose and the start of the first cyclophosphamide dose was approximately 12 hours with q6h busulfan and approximately 30 hours with q24h busulfan. In this retrospective study, we did not have a reliable means of assessing the intangible benefits of the q24h
ARTICLE IN PRESS S. Singhal et al. / Biol Blood Marrow Transplant 00 (2019) 1 5
regimen or comparing the total transplantation cost between the 2 cohorts. In addition, a portion of the cost savings could be accounted for by differences in body weight and pharmacokinetics between the 2 cohorts. Although the April 2017 average wholesale price of busulfan was used to extrapolate costs between the 2 groups, this does not reflect potential changes in drug costs over time. Future interventions could include i.v. busulfan administration in the outpatient setting and institutional development of a population pharmacokinetic model to allow for pharmacokinetic-guided busulfan dosing, as discussed by the American Society for Blood and Marrow Transplantation busulfan practice guidelines [8]. Future studies could seek to explore a potential cost impact from less tangible benefits and potential clinical impact on SOS, GVHD, and overall survival. In summary, this retrospective study strongly supports the adoption of q24h busulfan dosing for the BU/CY regimen as a means of improving value-based care, reducing drug waste, and reducing provider workload in allogeneic HCT. These data indicate that once-daily i.v. busulfan is comparable or superior to the Food and Drug Administration-approved standard q6h busulfan dose schedule in both safety and efficacy, and results in substantial cost savings.
ACKNOWLEDGMENTS The authors thank Linda Elder and her team, who maintain the patient database for the Stanford Division of Blood and Marrow Transplantation. Financial disclosure: A.R.R. participated on one-time ad hoc scientific advisory boards for Nohla Therapeutics and for Kaleido, both in 2018. He receives research support from AbbVie for an ongoing clinical trial and has served as a medical expert witness for the US Department of Justice. His brother is an employee of Johnson & Johnson. Conflict of interest statement: There are no conflicts of interest to report.
5
REFERENCES 1. Chronic granulocytic leukaemia: comparison of radiotherapy and busulfan therapy. Report of the Medical Research Council’s working party for therapeutic trials in leukaemia. Br Med J. 1968;1:201–208. 2. Santos GW, Tutschka PJ, Brookmeyer R, et al. Marrow transplantation for acute nonlymphocytic leukemia after treatment with busulfan and cyclophosphamide. N Engl J Med. 1983;309:1347–1353. 3. Pasquini MC, Le-Rademacher J, Zhu X, et al. Intravenous busulfan-based myeloablative conditioning regimens prior to hematopoietic cell transplantation for hematologic malignancies. Biol Blood Marrow Transplant. 2016;22:1424–1430. 4. Chandy M, Balasubramanian P, Ramachandran SV, et al. Randomized trial of two different conditioning regimens for bone marrow transplantation in thalassemia the role of busulfan pharmacokinetics in determining outcome. Bone Marrow Transplant. 2005;36:839–845. 5. Gong CL, Rezvani AR, Kim T, Studdert AL. Busulfan dosing literature review. J Blood Lymph. 2018;8:204. 6. Micromedex. Busulfan. Truven Health Analytics. https://www-microme dexsolutions-com.laneproxy.stanford.edu/micromedex2/librarian/PFDe faultActionId/evidencexpert.DoIntegratedSearch?navitem=topHome&is ToolPage=true#. Published 2019. Accessed January 25, 2019. 7. Fasola G, Aprile G, Marini L, Follador A, Mansutti M, Miscoria M. Drug waste minimization as an effective strategy of cost-containment in oncology. BMC Health Serv Res. 2014;14:57. 8. Palmer J, McCune JS, Perales MA, et al. Personalizing busulfan-based conditioning: considerations from the American Society for Blood and Marrow Transplantation Practice Guidelines Committee. Biol Blood Marrow Transplant. 2016;22:1915–1925. 9. Holden RJ, Scanlon MC, Patel NR, et al. A human factors framework and study of the effect of nursing workload on patient safety and employee quality of working life. BMJ Qual Saf. 2011;20:15–24. 10. Hill BT, Rybicki L, Carlstrom KD, et al. Daily weight-based busulfan with cyclophosphamide and etoposide produces comparable outcomes to fourtimes-daily busulfan dosing for lymphoma patients undergoing autologous stem cell transplantation. Biol Blood Marrow Transplant. 2016;22: 1588–1595. 11. Ryu SG, Lee JH, Choi SJ, et al. Randomized comparison of four-times-daily versus once-daily intravenous busulfan in conditioning therapy for hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2007;13: 1095–1105. 12. Zhang H, Graiser M, Hutcherson DA, et al. Pharmacokinetic-directed highdose busulfan combined with cyclophosphamide and etoposide results in predictable drug levels and durable long-term survival in lymphoma patients undergoing autologous stem cell transplantation. Biol Blood Marrow Transplant. 2012;18:1287–1294. € n O, Hassan Z, Nilsson C, Oberg 13. Hassan M, Ljungman P, Ringde G. The influence of busulphan on the pharmacokinetics of cyclophosphamide and its 4-hydroxy metabolite: time interval effect on therapeutic efficacy and toxicity. Ther Drug Monit. 1999;21:434.