Forty years of haematopoietic stem cell transplantation : a review of the Basel experience

The purpose of this study was to examine changes in haematopoietic stem cell transplant (HSCT) characteristics and outcome in our combined paediatric and adult programme over the past four decades, since its implementation in 1973. The total number of transplant procedures rose from 109 in the first decade (1973–82) to 939 in the last decade (2003–12). Transplant characteristics changed significantly over time: patient age increased, peripheral blood largely replaced bone marrow as stem cell source, unrelated donors became an alternative to matched siblings, and patients are increasingly transplanted in more advanced disease stages. Advances such as improved supportive care and histocompatibility typing resulted in a steady decrease of transplantrelated mortality after allogeneic HSCT (43% in the first decade, 22% in the last decade). Despite this, unadjusted survival rates were stable in the last three decades for allogeneic HSCT (approximately 50% 5–year survival) and in the last two decades for autologous HSCT (approximately 60% 5–year survival). After adjustment for covariates such as donor type, age and stage, the relative risk of treatment failure continuously dropped (for allogeneic HSCT: first decade 1.0, second decade 0.58, third decade 0.51, last decade 0.41). Collectively, these data suggest that improvements in periand post-transplant care have allowed considerable extension of transplant indications without having a negative impact on outcome.


Introduction
Haematopoietic stem cell transplantation (HSCT) replaces diseased haematopoiesis by either the patient's own haematopoietic stem cells (autologous), stem cells from a related or unrelated donor (allogeneic) or stem cells originating from an identical twin (syngeneic) after a course of chemotherapy and/or irradiation termed conditioning.A wide variety of haematopoietic and nonhaematopoietic pathologies, ranging from myeloid and lymphoid malignancies to acquired or congenital bone marrow failures, can be treated by HSCT [1,2].Since its beginnings in 1959, HSCT has seen many developments [3].With the advent of human leucocyte antigen (HLA) typing, HSCT became a legitimate treatment option in the late 1960s and early 1970s.In the 1990s, growing unrelated donor registries started to provide donors to patients lacking a compatible family member.Additional transplant sources such as mobilised autologous and allogeneic peripheral blood stem cells (PBSCs), and umbilical cord blood (CBT), mismatched and haploidentical stem cells were successfully introduced [4,5].Advances in HLA-typing and enhanced registry accrual have increased donor availability and matching performance [6].Both the prophylaxis and treatment of frequent complications of HSCT such as graftversus-host disease (GvHD) and infections have been adapted [7,8].Trends to de-escalate conditioning intensity and towards novel treatments of early relapse (e.g.donor lymphocyte infusions) have been introduced [9].HSCT may now be offered to a larger group of patients with more advanced diseases, higher age and nonpreferential donor/ host constellations [10].HSCT has become an established treatment option and in 2013 the millionth transplant worldwide was performed [11].In Basel, HSCT was first introduced in 1973 at a time when this treatment was considered experimental by the late Bruno Speck, an early pioneer in the field [12].The purpose of this study was to compare transplant characteristics over the last four decades since the beginnings of HSCT in our combined paediatric and adult programme.We also examined the changes in outcome parameters over the last four decades  and the effect of changes in disease and transplant characteristics on the outcome.

Patients and methods
The data for this analysis was taken from our unique patient number (UPN) database for the timespan November 1973 to December 2012.First autologous and first allogeneic transplants were selected, and patient, disease and transplant characteristics were compiled and compared across decades using tests for categorical or continuous data where appropriate.As patients may have received more than one transplant, such as a second HSCT in the case of failure or in the context of planned sequential HSCT, the total number of transplants exceeds the patient number as shown in tables 1 and 2. All outcome data refer to patients.The primary endpoint was overall survival (OS).Secondary endpoints consisted of progression-free survival (PFS), transplant-related mortality (TRM) and relapse incidence (RI).Both overall and progression-free survivals were estimated by Kaplan-Meier analysis.Cumulative incidences of relapse and treatment-related mortality were calculated by treating these events as competing outcomes.Variables affecting overall survival were analysed in a multivariate fashion using Cox models to compare outcome per decade.As transplant indications changed over time, diagnosis was included in multivariate models as a stratification variable.Statistical analyses were performed on SPSS version 19 (SPSS Inc, Chicago, Illinois), GraphPad Prism 5.0a (GraphPad Software, San Diego, California) and Stata/SE 12.1 (StataCorp LP, College Station, Texas).

Discussion
Stem cell transplantation, a procedure aiming to cure an array of diseases affecting haematopoiesis, was first introduced in Basel in 1973 [12].Further milestones in our centre were the first matched unrelated HSCT in 1978, the introduction of haploidentical HSCT in 1981, the beginnings of peripheral blood stem cell donation in 1991 and the first umbilical cord blood stem cell transplant in 2003.
In this study, we examined the evolution of transplant mod-  alities, patient and disease characteristics and their influence on transplant outcome during the time span between 1973 and 2012.The total number of both autologous and allogeneic HSCTs increased significantly over time.The expansion of preconditions accepted for transplant is easily reflected in the significantly higher median patient age at autologous and allogeneic stem cell transplantation and higher EBMT risk scores observed in later decades.Allogeneic transplantation is today a treatment option for patients up to 70 years of age.The recognition that graft-versus-tumour effects contribute to disease control after transplantation led to the implementation of reduced-intensity and nonmyeloablative conditioning regimens, allowing transplant procedures to be carried out in elderly and comorbid patients without excess toxicity.
Major developments in medicine translated into variations in the diseases treated by HSCT in our centre.In the era of tyrosine kinase inhibitors, the group of patients needing allogeneic HSCT for CML has narrowed down to those in an accelerated phase, blast crisis, or with proof of treatment intolerance or resistance such as second-line treatment failure or carriage of the T315I mutation [14], reflected in declining numbers of patients transplanted for CML.Similarly, the addition of high-dose chemotherapy and autologous HSCT to adjuvant chemotherapy in breast cancer was shown to have no effect on survival and autologous transplantation for many solid tumours has largely been abandoned [15].Accordingly, the numbers of patients receiving autologous HSCT for solid tumours has decreased in this survey.HLA-matched sibling donors were the sole source of allogeneic stem cell grafts in the early 1970s.As only approximately 20%-25% of patients have an HLA-identical sibling, the majority of patients in need of HSCT could not proceed to transplantation because of the lack of a donor.Over the last two decades, registries of unrelated donors have recruited more than 20 million volunteer donors, and the safety of HSCT using "alternative donors" (i.e., haploidentical and cord blood grafts) has considerably advanced.Patients in need of an allogeneic transplant without a suitable donor are, therefore, are seen only occasionally nowadays.Similarly, the last four decades have seen a significant increase in unrelated donors and a trend to mis- Overall survival after autologous haematopoietic stem cell transplantation (HSCT).
matched related donors.The primary stem cell source has shifted from bone marrow to peripheral blood and to a smaller degree umbilical cord blood.
In terms of allogeneic HSCT outcome, transplant-related mortality decreased dramatically between the first and second decade, with the reduction thereafter following a more decelerated pace, whereas relapse incidence remains unchanged over time.A significant improvement in 5-year OS and PFS between the first decade and second decade can be seen and at later dates to a lesser extent.The multivariate analysis of survival by decade suggests continued improvement by decade with a trend to decreased mortality risks when adjusted for factors such as donor choice and disease stage.Compatible with our observations, the trend to administer reduced-intensity conditioning prior to HSCT diminishes TRM, but further lessens disease control [16].
As in other forms of manual and intellectual work processes, a certain "learning curve" has been assumed in the context of HSCT [17,18].These factors may play a role in the higher survival benefit between the first and the later decades.The multitude of changes seen in the field of stem cell transplantation, such as variations in disease, patient and graft profile, as well as improved supportive measures and conditioning protocols over the decades, and their complex interplay, for instance, on the graft-versustumour-activity level, does not, however, permit us to point to a single most important factor explaining the improvement in survival parameters in multivariate analysis.
In conclusion, stem cell transplantation has seen a wide array of changes since its implementation in Basel.The patient group eligible for autologous and allogeneic HSCT in our centre has diversified towards more advanced patient ages and disease stages and to patients without a family donor, and this is reflected in a higher number of transplants performed over the years.Five-year overall and progression-free survival rose for allogeneic and autologous HSCT after the decade of their introduction.After adjustment for factors such as donor type and disease stage, survival improved in every decade.Further efforts in developing regimens that can accommodate the adverse circumstances of higher age, comorbidities and nonavailability of a sibling donor, while still providing better control over more advanced diseases, are warranted [10].To this end, specific strategies are being applied at the time of transplant, such as reduction in regimen toxicity [19], elaboration of graft-versus-host-disease prophylaxis [20] and better prevention of post-transplant relapse [21].

Figure 1 (
Figure 1 (A) Total allogeneic and autologous haematopoietic stem cell transplantation (HSCT) numbers over the decades.(B) Identical sibling and matched unrelated allogeneic HSCT patient case numbers over the decades.(C) Bone marrow and peripheral blood stem cell source patient case numbers in allogeneic HSCT.(D) Patient case numbers receiving myeloablative and reducedintensity conditioning for allogeneic HSCT.

Table 1 :
Allogeneic haematopoietic stem cell transplantation -patient and first transplant characteristics per decade.
EBMT = European Group for Blood and Marrow Transplantation Original article Swiss Med Wkly.2014;144:w13928 Swiss Medical Weekly • PDF of the online version • www.smw.ch

Table 2 :
Autologous haematopoietic stem cell transplantation -patient and first transplant characteristics per decade.

Table 3 :
Factors affecting overall survival in the Cox model.
Swiss Medical Weekly • PDF of the online version • www.smw.ch