Highlights
In a nationwide Japanese registry analysis of 929 patients aged 70 years or older undergoing first allogeneic hematopoietic stem cell transplantation (allo-HSCT), the overall 2-year survival was 34.2%.
A simple risk score identified six adverse prognostic factors: age 75 years or older, adult T-cell leukemia/lymphoma (ATL), lymphoma, high disease risk, hematopoietic cell transplant-specific comorbidity index (HCT-CI) of 5 or higher, and performance status of 2 or higher.
The resulting score separated patients into four groups with markedly different 2-year overall survival rates, from 57.1% in the lowest-risk group to 2.5% in the highest-risk group.
The study offers a pragmatic framework to support transplant selection in very older adults, though it does not replace individualized clinical judgment or disease-specific decision-making.
Study Background
Allogeneic hematopoietic stem cell transplantation remains the only potentially curative therapy for many hematologic malignancies, including acute leukemia and certain lymphoid neoplasms. However, its benefit comes at the cost of substantial regimen-related toxicity, graft-versus-host disease, infection risk, and non-relapse mortality. These challenges are magnified in older adults, who are more likely to have comorbid illness, functional impairment, and less physiologic reserve.
Over the past two decades, reduced-intensity conditioning and better supportive care have expanded transplant eligibility beyond the traditional age cutoffs. Yet patients aged 70 years or older remain a particularly understudied group. Clinicians often face a difficult question: when does the potential for durable remission justify the risks of transplantation in a patient in their seventies or eighties?
This Japanese nationwide registry study was designed to address that gap by identifying prognostic factors and creating a usable scoring system for patients aged 70 years or older undergoing first allo-HSCT for hematologic malignancies.
Study Design
This was a retrospective, nationwide registry analysis using the Japanese Transplant Registry Unified Management Program. The investigators identified 929 patients aged 70 years or older who underwent first allo-HSCT between 2000 and 2022.
The cohort included patients transplanted for hematologic malignancies, with acute myeloid leukemia as the most common disease. The median age was 71 years, with an upper range extending to 88 years. The primary endpoint was overall survival (OS), and prognostic factors were assessed using multivariate Cox proportional hazards regression. The authors then converted the relative hazards into a practical point-based scoring system to stratify outcomes.
Key Findings
The headline result is sobering but clinically informative: despite modern transplant practice, the 2-year overall survival rate in this very older cohort was 34.2%. This figure reflects the high-risk nature of the population, but it also underscores that a meaningful minority of selected patients can achieve durable survival after allo-HSCT.
On multivariate analysis, six factors were independently associated with worse survival:
1. Age 75 years or older
2. Adult T-cell leukemia/lymphoma (ATL)
3. Lymphoma
4. High disease risk
5. HCT-CI of 5 or higher
6. Performance status of 2 or higher
The inclusion of both disease-related variables and host-related variables is important. Disease risk captures the underlying biology and urgency of transplantation, while HCT-CI and performance status reflect treatment tolerance. Age itself remained prognostic even within an already older cohort, suggesting that the difference between early seventies and mid- to late seventies is clinically meaningful rather than merely chronological.
Using the hazard ratio-derived scoring system, patients were stratified into four risk groups:
Low risk: score 0, 2-year OS 57.1%
Intermediate-1 risk: score 1, 2-year OS 35.0%
Intermediate-2 risk: score 2-3, 2-year OS 18.3%
High risk: score 4-5, 2-year OS 2.5%
The separation between these curves was statistically significant (p < 0.01), indicating that the score captured clinically relevant heterogeneity in outcomes. In practical terms, the model identifies a subset of older patients whose survival appears sufficiently favorable to justify consideration of allo-HSCT, while also highlighting a group in whom benefit is likely minimal.
Several aspects of the result deserve emphasis. First, the low-risk group’s 2-year OS above 50% suggests that chronological age alone should not exclude transplant candidacy. Second, the rapid decline in survival across risk strata indicates that multiple adverse features accumulate substantially. Third, the very poor outcomes in the high-risk group argue for caution when several unfavorable factors coexist, especially when the transplant goal is not clearly curative or when alternative options exist.
Clinical Interpretation
The study is most useful as a decision-support tool rather than a rigid eligibility rule. In older adults, transplant decisions are rarely binary. They depend on remission status, disease biology, donor availability, conditioning intensity, functional status, patient preference, and competing risks such as infection or organ dysfunction. This registry-based model contributes a simple, evidence-based way to quantify risk in a population where intuition alone is often unreliable.
Performance status and HCT-CI are familiar tools in transplant medicine, and their predictive value in this study reinforces current practice. The observation that disease category mattered, with ATL and lymphoma associated with worse outcomes, may reflect disease biology, prior treatment exposure, and transplant indications that differ from those in myeloid neoplasms. High disease risk likely captures refractory or advanced disease, which remains a dominant driver of post-transplant mortality regardless of age.
Importantly, this analysis does not imply that patients with poor prognostic features should never undergo allo-HSCT. Rather, it provides a clearer estimate of expected benefit and risk. In selected cases, especially when the disease is otherwise fatal and no effective alternatives are available, even high-risk patients may still be considered after detailed counseling. Conversely, for frail patients with multiple adverse features, the score may help avoid non-beneficial transplantation and redirect care toward less toxic disease control or supportive approaches.
Strengths of the Study
The study has several strengths. It is large for this age group, nationwide in scope, and derived from real-world registry data spanning more than two decades. That breadth improves generalizability within the Japanese transplant setting and reflects contemporary clinical heterogeneity. The authors also translated statistical findings into a clinically usable score, which increases bedside relevance.
Another strength is the focus on patients aged 70 years or older, a population often excluded from prospective trials. By examining this group directly, the study addresses a gap that is increasingly important as transplant indications expand and the population ages.
Limitations and Cautionary Notes
As with all retrospective registry studies, confounding is a major limitation. Treatment selection bias is unavoidable: patients who underwent transplant were already selected as fit enough for the procedure, meaning the sickest older adults were likely not included. As a result, the reported outcomes may overestimate what would be seen in an unselected population but may also underestimate the best-case outcomes in highly specialized centers.
The registry design also limits granularity. Important variables such as geriatric assessment, cognitive function, frailty, detailed conditioning regimens, donor source nuances, post-transplant complications, graft-versus-host disease severity, relapse incidence, and quality of life may not be fully captured. These factors are especially relevant in older adults, where survival alone does not fully describe treatment value.
In addition, the study spans 2000 to 2022, a period during which supportive care, donor selection, infection prophylaxis, and transplant techniques evolved substantially. Although this broad timeline enhances sample size, it may also introduce temporal heterogeneity. External validation in non-Japanese cohorts is also needed before the score is applied broadly across different healthcare systems and disease epidemiologies.
Expert Commentary
This study aligns with a broader trend in hematology: age should be considered a risk modifier, not an absolute contraindication, for allo-HSCT. Contemporary practice increasingly emphasizes physiologic fitness, disease risk, and patient goals over rigid age thresholds. The present registry analysis supports that approach while also showing that very old age still carries prognostic weight.
From a translational perspective, the score could be integrated into transplant conferences as one component of a structured assessment. In future iterations, combining this type of model with geriatric assessment tools, frailty indices, or machine learning-based prediction could improve individual-level risk prediction. Such models may be especially helpful for shared decision-making, where patients and families need realistic estimates of both survival likelihood and treatment burden.
Conclusion
In Japanese registry data, allo-HSCT in patients aged 70 years or older produced a 2-year overall survival of 34.2%, but outcomes varied widely by age, disease type, disease risk, comorbidity burden, and performance status. The newly proposed score offers a practical way to identify older patients most likely to benefit from transplantation and those in whom the balance of benefit and harm is unfavorable.
For clinicians, the key message is not that transplantation is inappropriate after 70, but that patient selection must be especially deliberate. For researchers, the next step is external validation and incorporation of geriatric and quality-of-life metrics to make prognostication more comprehensive and patient-centered.
Funding and ClinicalTrials.gov
Funding details were not specified in the PubMed abstract provided. No ClinicalTrials.gov registration was listed; this study was a retrospective registry analysis.
References
1. Kataoka A, Yoshinaga N, Inoue Y, et al. Prognostic factors and a novel scoring system for allogeneic hematopoietic stem cell transplantation in patients aged 70 years or older: a nationwide registry analysis in Japan. Bone Marrow Transplant. 2026; PMID: 42277251.
2. Sorror ML, Storb RF, Sandmaier BM, et al. Comorbidity-age index: a clinical measure of biologic age before allogeneic hematopoietic cell transplantation. J Clin Oncol. 2014.
3. Sorror ML. How I assess comorbidities before hematopoietic cell transplantation. Blood. 2013.
4. Döhner H, Estey E, Grimwade D, et al. Diagnosis and management of AML in adults: recommendations from an international expert panel. Blood. 2022.
5. Frailty and geriatric assessment literature in allo-HSCT selection, including recent reviews and consensus statements from transplant societies, supports incorporating functional status into candidacy decisions.
