Science Spotlight

A long learning process: conditioning regimen and metabolism

The figure represents cumulative incidence estimates of non-relapse mortality up to 200 days after hematopoietic cell transplantation according to the conditioning regimen and presence or absence of Gilbert's Syndrome. Gilbert's syndrome patients treated with busulfan present increased mortality incidence.
Cumulative incidence estimates of non-relapse mortality at day 200 after hematopoietic cell transplantation according to the conditioning regimen and presence or absence of Gilbert's Syndrome. Figure provided by Dr. George McDonald.

For some patients undergoing hematopoietic cell transplantation, a myeloablative conditioning regimen is administered prior to the infusion of hematopoietic cells, to eliminate the patient’s own cells and facilitate engraftment of the newly transplanted cells. Several chemotherapeutic drugs are used as part of a conditioning regimen, either alone or in combination. The metabolism of these conditioning drugs has been a recent field of investigation that is critical to refine the dosage and increase efficacy while lowering toxicity.

However, no data are available for transplanted patients with an impaired metabolism that might affect drug catabolism (elimination) and lead to toxicity issues. This is the case for Gilbert’s syndrome patients treated with irinotecan, a drug used for colorectal cancer. Gilbert’s syndrome (GS) is an inherited disorder affecting the uptake of bilirubin by the liver. Patients are characterized by disrupted bilirubin metabolism, a product of red blood cell breakdown, resulting in hyperbilirubinaemia (higher concentration of bilirubin in the blood). This disease is benign and does not need require treatment but patients occasionally develop jaundice. It is estimated that GS affects 3 to 10% of people worldwide.

Given the complete lack of data analyzing whether patients with GS have a different outcome from other people following hematopoietic cell transplantation, Drs. George McDonald, Jeannine McCune, Ted Gooley (Clinical Research Division) and other colleagues from Fred Hutch, performed a retrospective cohort study to address this question. The results were recently published in the Lancet Haematology.

The cohort included a total of 3379 patients who underwent hematopoietic cell transplantation between 1991 and 2011 at Fred Hutch. Among these participants six percent (n=211) were GS patients (defined by a high bilirubin serum concentration ≥1mg/dl). Patients were divided into groups based on autologous (same donor cells) or allogeneic (related or unrelated matched donor cells) setting of the transplant and on the conditioning regimen received: Busulfan plus Cyclophosphamide (BU/CY), Busulfan plus Mephalan plus ThioTEPA (BUMELTT), mephalan only or cyclophosphamide plus total body irradiation (CY/TBI).

The authors first compared the mortality outcomes between the transplanted patients with or without GS, but no difference was noted when analyzing all groups together. However, when looking into each group separately it was striking to observe that the GS patients conditioned with busulfan (BU/CY or BUMELTT) were significantly more at risk to die during the 200 days immediately after the transplantation. Patients undergoing other conditioning regimens, without busulfan, were not at risk, indicating busulfan as a risk factor for GS patients. The allogeneic or autologous origin of the transplanted cells did not influence the outcome. For these patients at risk, the level of unconjugated serum bilirubin was positively correlated with the overall mortality and non-relapse mortality (transplant-related deaths). Trying to understand the mechanisms responsible for busulfan worsening effect, the authors then analyzed busulfan metabolism but the pharmacokinetic was similar between GS and non-GS patients, leaving the door open for more investigations.

These observations argue against the use of busulfan as a conditioning regimen in GS patients. What are the other options for these patients? “There are alternatives to busulfan as part of conditioning regimens, but there is far less experience with these drugs in treating patients with different hematologic malignancies than there is with busulfan”, explained Dr. McDonald. More data will be acquired over time to refine their use. The investigators are now moving toward a better understanding of the underlying mechanisms: ”it would be of interest to examine the genome of a large number of past transplant recipients for the single nucleotide polymorphisms that lead to the clinical diagnosis of Gilbert's Syndrome, and to relate mutations to patient outcomes.  For our study, we used a laboratory definition of increased unconjugated serum bilirubin, in the absence of alternative causes of this finding. From a genetic point of view, Gilbert's Syndrome is a disorder that can result from over 100 different mutations in the promoter region of the UDP-glucuronosyltransferase gene. The holy grail of this work would be to identify mutations that translate to increased risk from busulfan-based conditioning”, concluded Dr. McDonald.


Funding for this study was provided by the National Institutes of Health.


McDonald GB, Evans AT, McCune JS, Schoch G, Ostrow JD, Gooley TA. 2016. Mortality outcomes after busulfan-containing conditioning treatment and haemopoietic cell transplantation in patients with Gilbert's syndrome a retrospective cohort study. The Lancet Haematolology, 3(11), e516-e525.