Hematopoietic Cell Transplantation Trends and Outcomes in Canada: A Registry-based Cohort Study

1. Introduction

Hematopoietic cell transplantation (HCT) is an established therapeutic procedure for hematological malignancies and serious non-malignant hematological disorders^1,2. It encompasses two major modalities: autologous HCT, which involves administering high doses of myelotoxic chemo- or radiotherapy followed by the reinfusion of autologous hematopoietic progenitor cells to recover the patient’s hematopoietic system and allogeneic HCT, which includes myelotoxic or immunosuppressive therapy followed by infusion of hematopoietic progenitor cells derived from a related donor, volunteer unrelated donor, or placental cord blood donor³.

Despite its curative potential, HCT is a technically complex and resource-intensive procedure, often associated with considerable toxicity. In Canada, HCT is offered by a limited number of specialized transplant centers, all based at major hospitals. Over time, advances in donor selection⁴, pre-transplant therapy, post-transplant supportive care^5–7, and graft-versus-host disease (GVHD) management ^8,9 have influenced HCT trends and outcomes. Several national¹⁰ and international¹¹ registries have described HCT trends and outcomes and demonstrated improvements in clinical outcomes over time^12,13. However, the specific quantity, trends, and clinical outcomes of HCT over time in Canada is not known.

To address this knowledge gap, we examined the characteristics of HCT in Canada using registry data. Our primary aim was to analyze the trends in HCT procedures and characteristics of recipients, as we hypothesized that the HCT landscape has evolved over time in line with other developed countries. Additionally, we examined the hypothesis that survival among transplant recipients have improved over time, potentially attributable to advances in supportive care leading to a reduction in HCT-related toxicities and associated non-relapse mortality.

2. Methods

2.1. Study Design, Participants and Setting:

We conducted a retrospective cohort analysis of autologous and allogeneic HCTs at Canadian transplant centers performed between January 1, 2000, and December 31, 2019. To determine whether patient characteristics have changed and whether outcomes have improved over time, patients were grouped into two periods: 2000 to 2009 (early era) and 2010 to 2019 (later era). We included children and adults who received a first HCT at any Canadian transplant center reporting to the Cell Therapy Transplant Canada (CTTC) registry over this period.

Separate analyses of clinical outcomes were performed on patients with non-malignant disease indications. Additionally, a subgroup analysis of allogeneic HCTs was performed on adults with indication of acute leukemia: acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), acute leukemia of ambiguous lineage, myelodysplastic syndrome (MDS) and myeloproliferative neoplasia (MPN). This subgroup analysis was intended to remove the potential confounding effect of HCT indications which were more frequently performed in the earlier HCT era and whose clinical outcomes may differ substantially from the other allogeneic HCT indications.

2.2. Data Sources:

Through a collaboration between CancerCare Manitoba and CTTC, patient, donor, and outcomes data from these centers were voluntarily collected and analyzed. The CTTC registry is a research consortium consisting of 15 adult and pediatric HCT centers across Canada. The CTTC registry collects transplant essential data (TED) forms prior to transplantation, and at 100 days, 6 months, and 1-year post-transplant and annually thereafter. Socio-demographic data, medical information regarding diagnosis, treatment, transplant, and follow-up were collected. Data on donor types, cell sources, disease prog

ression or relapse, and mortality were also recorded. The HCT database is housed by the CTTC registry at CancerCare Manitoba. Regular central auditing of the data is performed to ensure consistency and quality.

2.3. Outcomes:

Trends in the patient characteristics, transplant type, and indication of HCTs in Canada were assessed as primary outcomes. Clinically relevant outcomes including overall survival (OS) and non-relapse mortality (NRM) were also examined as primary outcomes for HCT recipients over this 20-year period. OS was defined as the time from HCT until death and NRM was defined as death from any cause other than relapse or disease progression after HCT. All data were censored at the date of last follow-up, date of last contact, date of death or date of second transplant.

2.4. Statistical Analysis:

Baseline patient, disease indication, and HCT characteristics for first transplants were summarized using descriptive statistics and reported as absolute numbers and proportions for each of the two eras. Statistical significance for comparisons were performed using Fisher's exact test for categorical data. OS and TRM were analyzed by HCT type (allogeneic or autologous), disease type (malignant or non-malignant), age groups (adult, age >18 years or pediatric, age <18 years) and by era (early or later) for first transplants. Survival probabilities were estimated using the Kaplan-Meier method and survival curves compared using the log-rank test. NRM was analyzed using cumulative incidences and Fine and Gray’s tests to accommodate competing risks of death from relapse/progression. Survival probabilities and cumulative incidences were reported with 95% confidence intervals (CI) and p-values <0.05 were considered statistically significant.

2.5. Ethics Approval:

All patients included in this study gave written consent to participate in the research database and to have their data included in observational research. This study was approved by the institutional review board of the University of Manitoba.

3. Results

A total of 18,046 first HCTs were reported from January 1, 2000, to December 31, 2019, from 15 transplant centers in Canada. Autologous transplants accounted for 10,475 (58%) of first HCTs performed. Patient characteristics for first allogeneic and autologous transplants are presented in Tables 1 and 2 respectively. Median age at HCT increased between the early (2000-2009) and later (2010-2019) eras: for allogeneic HCT, median age increased from 42 years to 50 years( p<0.0001), while for autologous transplants median age increased from 53 years to 58 years (p<0.0001). The proportion of older adults (>64 years old) receiving allogeneic HCT increased 5-fold between early and later eras (1.7% vs. 9.4%, p<0.0001). The proportion of older adults receiving autologous transplants doubled (10.4% vs. 21.5%, p<0.0001).

Donor types for allogeneic HCT are summarized in Table 1. Matched related donors (MRDs) accounted for the largest proportion of donors in the early era (54.6%), while matched unrelated donors (MUDs) accounted for the largest proportion in the later era (56.3%). The decrease in MRDs between the two eras (54.6% early vs. 36.6% later, p<0.0001) was proportionate to the increase in MUDs (35.1% early vs. 56.3% later, p<0.0001). Use of human leukocyte antigen (HLA) mismatched related (haploidentical) donors doubled (3.4% in the early era vs. 6.1% in the later era, p<0.0001).

Regarding cell source, peripheral blood stem cells (PBSCs) were the most common source in both allogeneic and autologous transplants (Table 1 and Table 2). PBSC use in allogeneic transplants increased from 60.3% to 74.0% (p<0.0001) over the two eras. In contrast, bone marrow for allogeneic transplants decreased from 30.0% to 15.9% (p<0.0001). Placental cord blood represented a small proportion of allogeneic transplants over these two time periods (6.3% in the early era vs. 5.2% in later era, p=0.044).

Indications for first transplants performed are presented in Table 1 and Table 2. For allogeneic HCT, acute myeloid leukemia (AML) was the most common indication across both eras (36.7%), followed by ALL (14.2%) and myelodysplastic or myeloproliferative disorders (MDS/MPN) (12%). Allogeneic HCT for AML increased between early and later eras (32.1% vs. 40.5%, p<0.0001), while HCTs for chronic myeloid leukemia (CML) dropped by approximately half over this time (7.3% early vs. 3.4%, p<0.0001). The most common non-malignant disease for which HCT was undertaken was severe aplastic anemia (4.5%). The proportion of allogeneic HCT for aplastic anemia decreased from 5.4% in the early era to 3.8% in the later era (p<0.0001). In contrast, use of allogeneic HCT for other non-malignant diseases, including inherited bone marrow diseases and immune system disorders, increased from 5.5% to 8.0% (p<0.0001).

Plasma cell disorders (including multiple myeloma) were the most common indication for autologous transplants over both eras (49.4%), followed by non-Hodgkin (30.8%), and Hodgkin lymphoma (9.7%). Relative contributions for most disease indications for autologous HCT remained similar between both eras (Table 2). Breast cancer was a rare indication for autologous HCT in the earlier era (N =31), which was no longer observed in the later era (N=0).

Outcomes were assessed through 5-year OS and 100-day NRM. For adults, 5-year OS after allogeneic HCT was similar between early and later eras (49% [95% C.I. 47-52%], vs. 49% [95% C.I. 47-51%], p=0.83). Autologous HCT recipients, experienced improved 5-year OS in the later era (55% [95% C.I. 54-57%] vs. 65% [95% C.I. 62-67%], p<0.0001). (Figure 1A, Table 3). For pediatric allogeneic HCT, there was significant improvement in 5-year OS in the later era (64% [95% C.I. 60-67%], vs. 76% [95% C.I. 71-80%], p<0.0001) while 5-year OS after pediatric autologous HCT showed a numerical increment, albeit with statistical non-significance (51% [95% C.I. 45-57%], vs. 57% [95% C.I. 49-65%], p=0.45). (Figure 1B, Table 3).

100-day NRM for allogeneic HCT improved for adults in the later era (8.1% [95% C.I. 7.1-9.0%], vs. 10.9% [95% C.I. 9.7-12.2%], p=0.0002). Improvement in 100-day NRM was also noted in pediatric allogeneic HCT patients (3.7% [95% C.I. 2.5-5.3%], vs. 7.5% [95% C.I 5.7-9.6%], p=0.0025) (Figure 2A, Table 4). For autologous HCT, 100-day NRM improved in the later era for adult patients (1.4 % [95% C.I. 1.0-1.7%], vs. 2.1% [95% C.I. 1.7-2.6%], p=0.0067), but for pediatric patients there was no difference in 100-day NRM between the two eras (1.4 % [95% C.I. 0.5-3.4%], vs. 1.9% [95% C.I. 0.8-3.8%], p=0.534) (Figure 2B, Table 4). As can be seen in these comparisons, 100-day NRM was consistently higher amongst adults compared to pediatric patients for all transplant modalities.

To better evaluate changes in clinically relevant outcomes, OS and NRM were separately assessed for non-malignant and select malignant disease indications. For non-malignant disease indications, 5-year OS was numerically improved in the later era for adult patients (85 % [95% C.I. 75-90%] vs. 70% [95% C.I. 59-79%], p=0.1529) with a statistically significant improvement for pediatric patients (85 % [5% C.I. 78-89%] vs. 78% [95% C.I. 70-84%] , p=0.0393) (Figure 3A, Table 5A). 100-day NRM for non-malignant diseases decreased numerically in the later era for adult patients (p=0.1686). For pediatric patients, 100-day NRM similarly showed a downward trend (p=0.2425) between early and later eras (Figure 3B, Table 5B).

A subgroup analysis of allogeneic HCTs was performed on adults with indications of acute leukemias, myelodysplastic syndromes, and myeloproliferative neoplasms (AML, ALL, AL of ambiguous lineage and MDS or MPN), excluding chronic myeloid leukemia. Within this subgroup, 5-year OS was stable: 44.8% [95% C.I 42.1 - 47.5 %] in the early era vs. 45.7 [95% C.I. 43.1-48.2%], in the later era, p=0.1693 (Figure 4A, Table 6A). For these patients, 100-day NRM decreased significantly in the later era (7.85% [95% C.I 6.8-8.9] vs. 11.4% [95% C.I 9.9-13.1], p=0.0002) (Figure 4B, Table 6B).

Table 6A. Overall Survival by Era following Allogeneic HCT indicated for Adult Acute Leukemias. Malignant disease indications included were AML, ALL, AL of ambiguous lineage, and MDS/MPN. Number of individuals at risk and overall survival with 95% confidence intervals are indicated at 1, 2 ,5 and 10 years post transplant.

	2000-2009		2010-2019		P-value	Overall P-value
	No. at risk	OS, % (95% CI)	No. at risk	OS, % (95% CI)
1-yr	831	62.4 (59.8 - 64.9)	1222	65.6 (63.6 – 67.6)	0.014	0.1513
2-yr	672	53.6 (50.9 – 56.2)	762	55.0 (52.8 - 57.3)	0.0911
5-yr	423	44.8 (42.1 - 47.5)	307	45.7 (43.1 - 48.2)	0.1693
10-yr	268	40.1 (37.3 - 42.9)	24	41.3 (37.6 – 44.9)	0.1522

Table 6A. Overall Survival by Era following Allogeneic HCT indicated for Adult Acute Leukemias. Malignant disease indications included were AML, ALL, AL of ambiguous lineage, and MDS/MPN. Number of individuals at risk and overall survival with 95% confidence intervals are indicated at 1, 2 ,5 and 10 years post transplant.

	2000-2009		2010-2019		P-value	Overall P-value
	No. at risk	OS, % (95% CI)	No. at risk	OS, % (95% CI)
1-yr	831	62.4 (59.8 - 64.9)	1222	65.6 (63.6 – 67.6)	0.014	0.1513
2-yr	672	53.6 (50.9 – 56.2)	762	55.0 (52.8 - 57.3)	0.0911
5-yr	423	44.8 (42.1 - 47.5)	307	45.7 (43.1 - 48.2)	0.1693
10-yr	268	40.1 (37.3 - 42.9)	24	41.3 (37.6 – 44.9)	0.1522

Table 6B. Non-Relapse Mortality by Era following Allogeneic HCT indicated for Adult Acute Leukemias. Malignant disease indications included were AML, ALL, AL of ambiguous lineage, and MDS/MPN. Number of individuals at risk and cumulative incidence of non-relapse mortality with 95% confidence intervals are indicated at 100 days, 1 year, 2 years, and 5 years post transplant.

	2000-2009		2010-2019		P-value	Overall P-value
	No. at risk	Cumulative Incidence, % (95% CI)	No. at risk	Cumulative Incidence, % (95% CI)
100 days	1216	11.4 (9.9 - 13.1)	1925	7.85 (6.8 - 8.9)	0.0002	0.0337
1-yr	781	20.1 (18.1 - 22.2)	1157	18.1 (16.5 – 19.7)	0.0642
2-yr	630	23.0 (20.8 - 25.3)	724	21.1 (19.3 - 22.9)	0.0815
5-yr	413	26.3 (23.9 - 28.7)	293	23.0 (21.1 – 25.0)	0.0349

Table 6B. Non-Relapse Mortality by Era following Allogeneic HCT indicated for Adult Acute Leukemias. Malignant disease indications included were AML, ALL, AL of ambiguous lineage, and MDS/MPN. Number of individuals at risk and cumulative incidence of non-relapse mortality with 95% confidence intervals are indicated at 100 days, 1 year, 2 years, and 5 years post transplant.

	2000-2009		2010-2019		P-value	Overall P-value
	No. at risk	Cumulative Incidence, % (95% CI)	No. at risk	Cumulative Incidence, % (95% CI)
100 days	1216	11.4 (9.9 - 13.1)	1925	7.85 (6.8 - 8.9)	0.0002	0.0337
1-yr	781	20.1 (18.1 - 22.2)	1157	18.1 (16.5 – 19.7)	0.0642
2-yr	630	23.0 (20.8 - 25.3)	724	21.1 (19.3 - 22.9)	0.0815
5-yr	413	26.3 (23.9 - 28.7)	293	23.0 (21.1 – 25.0)	0.0349

4. Interpretation

This national level study provides insights into the trends and outcomes of HCT in Canada over a 20-year period. We observed an increasing number of transplants being offered to Canadians, which is likely to be reflective improvements in health care access, greater awareness of the potential benefits of HCT, and higher number of medically eligible patients as HCT can be delivered more safely. We noted aa shift towards older adults receiving both allogeneic and autologous HCTs, with a 5-fold increase in older adults (>64 years old) accessing allogeneic transplants. We also observed that MUDs displaced MRDs as the dominant donor type in allogeneic HCT, suggesting increased reliance on national and international stem cell donor registries to match Canadian patients.

Regarding key clinical outcomes, there were significant temporal improvements in OS that benefited both pediatric and adult allogeneic HCT recipients. We highlight the ongoing challenge of higher NRM in allogeneic HCT compared to autologous HCT, likely due to deeper and more prolonged immune deficiency after allogeneic HCT, often compounded by GVHD. Age-related differences in NRM after allogeneic HCT were apparent, with higher NRM in adults compared to children; suggesting that adult patients are more vulnerable to the adverse effects of allogeneic HCT. However, there were temporal improvements in NRM in the 2010-2019 (later) era for both adult autologous and allogeneic HCT recipients as well as for children receiving allogeneic HCT.

The rise in both allogeneic and autologous HCT among older adults in the later era is consistent with trends in other jurisdictions ¹⁰ and poses unique challenges. While older adults may benefit from allogeneic HCT with appropriate selection¹⁵, they experience inferior outcomes in OS and NRM compared to younger adults¹⁶. This is due to the associated increase incidence in frailty, medical co-morbidities, and inherently more treatment resistant hematological diseases in older adults^17,18. We observed a 5-fold increase in allogeneic HCT among older adults during the later era, which potentially impacted OS in these patients. However, there was a measurable reduction in 100-day NRM of allogeneic HCT patients treated for acute leukemia and related myeloid malignancies, which attests to temporal improvements in both patient selection and hospital-based care of transplant recipients despite their older age. Advances in pre- and post-transplant care, such as the of reduced intensity conditioning, more effective graft-vs-host disease prophylaxis, and optimal use of anti-microbial agents also likely contributed to these positive changes¹⁹.

In the later era , we observed that MUDs became the dominant donor type for allogeneic HCT with a proportionate reduction in the use of matched related donors. We also noted a rise in haploidentical (HLA mismatched) related donors. While similar trends have been reported in other registries^10,11, the proportion of haploidentical related transplants in Canada remains relatively low compared to other registries^10,11, while cord blood transplant usage remains stable. The shift towards matched unrelated donors may be attributed to decreasing family sizes in Canada^20,21, stem cell registry expansion, and greater acceptance of HLA mismatching associated with improved GVHD prophylaxis and treatment^22,23. Additionally, the preference for PBSCs over bone marrow as the donor cell source in the later era highlights improvements in procedural logistics, relative ease of PBSC collection, and improved donor experiences associated with this apheresis-based collection technique²⁴.

Our study highlights the persistent challenge of higher NRM in allogeneic HCT compared to autologous HCT. This difference is likely attributable to greater immune suppression associated with allogeneic HCT, as well as the inherent risk of GVHD and its potentially severe complications²⁵. We also observed a higher NRM in adult allogeneic recipients compared to pediatric recipients. NRM rises for patients aged 21-40 compared to those under 20²⁶. Similar findings from a national study in South Korea reported increased early NRM following allogeneic HCT in AML patients over 20 years of age compared to younger patients²⁷. Moreover, a recent review highlighted studies reporting a lower incidence of severe acute GVHD among pediatric recipients compared to adults²⁸, with GVHD incidence increasing with age within pediatric cohorts²⁹. Taken together, these insights demonstrate age-related differences in HCT outcomes and emphasize the importance of tailored approaches for distinct patient groups.

Amongst all age groups and time periods that we studied, outcomes after allogeneic HCT for non-malignant diseases were better than those for patients with malignancies. Amongst non-malignant hematological diseases, the principal indication in Canada was severe aplastic anemia, a disease that is expected to be cured after allogeneic HCT in most patients. As the ancestry of the Canadian population becomes enriched for individuals of Asian and African ancestry, we expect that sickle cell disease and thalassemia may be become more frequent reasons for allogeneic HCT, especially in children and young adults. Future outcomes studies should assess trends and outcomes for these emerging conditions.

This study has several limitations. Firstly, this analysis is based on transplants voluntarily reported to the CTTC registry, and thus, does not account for all HCTs performed in Canada during that time. While CTTC registry data underrepresents the entirety of transplant activity in Canada, the potential for geographical bias is low given that individual HCT centers do not specialize in specific patient populations or treatment modalities Thus, potential under-reporting of HCT data from some centres does not represent a source of differential bias in transplant type or outcome. Regarding outcomes reported in this study, it is important to consider the decreasing reliability of the data with time due to follow-up loss, especially beyond 10 years post-transplant, and the increased likelihood of other, unrelated factors influencing OS and NRM. However, transplant centres are generally highly committed to the long-term care of their patients, especially in the allogeneic HCT setting, where follow-up with recipients is often indefinite in duration. This close follow-up of patients ensures higher quality of long-term outcome data. Additionally, we were unable to report on trends in the use of reduced intensity or non-myeloablative conditioning regimens for allogeneic HCT, which gained popularity in the recent era. Despite this limitation, we acknowledge the increased use of these novel conditioning regimens likely contributed to the observed improvements in outcomes in allogeneic HCT during the later era.

5. Conclusion:

In conclusion, over the 20-year study period, transplant activity has increased, and key clinical outcomes have generally improved for adult and pediatric populations receiving HCT in Canada. The landscape of HCT in Canada has also evolved to serve older patients, with increased reliance on volunteer unrelated donors. Despite offering allogeneic HCT to older patients, OS rates in adults remained stable. These data serve as a benchmark for quality management in HCT centers across the country and should be helpful for resource planning by health providers and funding authorities.