Survival and Prognostic Factors in Patients with Relapsed/Refractory Acute Lymphoblastic Leukemia Receiving Supportive Care

Abstract

Background/Objectives: Adult acute lymphoblastic leukemia (ALL) often has poor outcomes, especially after relapse or treatment failure. Many patients eventually become ineligible for curative treatment and require only supportive care or low-intensity chemotherapy. However, data on prognosis and predictive factors in this context are limited. The study aim was to evaluate survival and identify prognostic factors in patients with relapsed/refractory ALL receiving supportive care. Methods: We conducted a retrospective observational study of 59 patients at two tertiary hospitals in Mexico. All patients had exhausted curative treatment options. Clinical variables at diagnosis and relapse were analyzed, including age, leukocyte counts, relapse timing, prior treatment lines, transfusion needs, and use of prognostic scores. Kaplan–Meier analysis was used to estimate survival, and multivariate models were applied to identify predictors of overall survival. Results: Fifty-nine patients were included (median age 31 years, balanced gender). Most received two prior high-intensity chemotherapy lines. Median overall survival was 137 days, with transfusion requirements being the only significant prognostic factor; neither the Palliative Prognostic Index nor the Charlson Comorbidity Index demonstrated predictive value. Conclusions: In patients with relapsed/refractory ALL managed with supportive care, survival remains limited. Transfusion dependence is a strong adverse prognostic factor, likely reflecting disease burden and logistical barriers to outpatient care. These findings highlight the need for earlier integration of palliative care and the development of tailored prognostic tools for this population.

1. Introduction

Acute lymphoblastic leukemia (ALL) is a heterogeneous hematologic malignancy characterized by the proliferation of immature lymphoid cells in the bone marrow, blood, and extramedullary sites [1]. While pediatric patients have seen significant survival gains over the past decades, outcomes for adults remain suboptimal, particularly in the relapsed or refractory setting [2]. Advances in targeted therapy and immunotherapy—such as tyrosine kinase inhibitors, bispecific antibodies, and CAR-T cell therapy—have provided new options, but a considerable proportion of patients still progress after multiple lines of treatment. In this context, understanding the clinical course, prognostic factors, and supportive care needs of patients beyond curative options is essential for guiding patient-centered care. Additionally, in low- and middle-income countries, access to these advanced therapies is limited, increasing the relevance of studies focused on supportive management strategies [3].

Acute lymphoblastic leukemia prognosis is influenced by biological features, patients’ functional status, treatment tolerance, and access to novel therapies such as tyrosine kinase inhibitors (e.g., BCR::ABL1 inhibitors), immunotherapies (monoclonal or bispecific antibodies), and hematopoietic stem cell transplantation [4,5]. Treatment resistance may arise from both tumor-related mechanisms (e.g., drug-resistance genes, antigen loss) and host factors such as poor adherence or pharmacogenetic variability [6,7,8,9].

Although curative approaches, including transplantation and CAR-T cell therapy, have improved outcomes, relapses remain common. Consequently, many patients undergo multiple lines of therapy, often accumulating toxicities and experiencing diminished quality of life [10]. Molecular profiling using next-generation sequencing has identified key mutations (e.g., TP53, MLL rearrangements) associated with treatment failure and poor survival [11,12].

Despite these advances, the prognosis for relapsed/refractory patients beyond curative options remains poorly defined. Identifying individuals unlikely to benefit from further aggressive therapies is critical to avoid overtreatment and preserve quality of life [10,11]. Overtreatment is increasingly recognized as a challenge, particularly in settings with expanding access to new therapies, which may paradoxically delay the integration of palliative care [13].

Early palliative care in leukemia has been associated with improved symptom control, fewer ICU admissions, and better end-of-life outcomes. However, its implementation remains limited and often occurs too late to have a meaningful impact on quality of life or enable home-based care [14].

These challenges are further compounded in Latin America, where healthcare systems face structural limitations, including fragmented public policies, insufficient funding, geographic disparities, and inadequate clinician training. As a result, timely access to multidisciplinary palliative services is limited, especially in rural areas [15].

In Mexico, leukemia care is highly centralized in urban tertiary hospitals, with minimal availability of outpatient or home-based supportive care. This presents a particular challenge for patients with acute lymphoblastic leukemia (ALL), who are typically young adults with preserved functional status despite treatment failure. Understanding the clinical trajectory and prognostic factors in this group is essential to improve care planning and ensure appropriate palliative integration [16,17].

Therefore, this study aimed to characterize survival and identify prognostic factors among relapsed/refractory ALL patients receiving supportive care after exhausting curative treatment options.

2. Materials and Methods

Patient selection was based on a review of electronic and paper medical records from two tertiary care hospitals in Mexico City between January 2015 and December 2023. Inclusion criteria were as follows: age ≥ 18 years, confirmed diagnosis of ALL according to WHO criteria, and discontinuation of curative-intent therapy due to refractory disease or patient preference. Exclusion criteria included Philadelphia chromosome-positive disease still eligible for tyrosine kinase inhibitors, incomplete clinical records, or concurrent enrollment in a clinical trial. Variables collected included demographics, disease characteristics at diagnosis, prior lines of therapy, transfusion requirements, comorbidities (Charlson Comorbidity Index), and prognostic scores (Palliative Prognostic Index).

For prognostic classification, we evaluated two widely used tools: the Charlson Comorbidity Index (CCI) to quantify comorbidity burden, and the Palliative Prognostic Index (PPI) to estimate short-term survival in patients receiving palliative care. Both scores were calculated retrospectively from available clinical data at the time patients were deemed ineligible for curative therapy. The CCI was derived from recorded comorbid conditions, while the PPI was based on clinical parameters such as performance status, oral intake, edema, dyspnea, and delirium. These scores were incorporated into survival analyses to explore their potential prognostic value in this specific population.

Immunophenotypic classification at diagnosis (B-ALL vs. T-ALL) was determined by multiparametric flow cytometry performed on bone marrow aspirates at initial presentation.

Leukocyte counts were measured using automated hematology analyzers, with hyperleukocytosis defined as >100 × 10³/μL. Transfusion requirement was quantified as the number of red blood cell or platelet units administered per week. Demographic and clinical variables at diagnosis included age, sex, initial leukocyte count, and relapse risk stratification. Disease-related factors at relapse were also recorded, including time to relapse (early <1 year vs. late ≥1 year), leukocyte count at relapse, number of prior treatment lines, and transfusion requirements. The presence of newly acquired mutations (e.g., BCR::ABL1) was documented when available. Relapse timing was classified as early (<1 year from diagnosis) or late (≥1 year), consistent with prior studies in adult ALL where most relapses occur within the first year [18].

2.1. Statistical Analysis

The demographic characteristics of the study were presented using median (range) for quantitative variables and cases (n, %) for categorical variables. Non-parametric analysis was used for every variable since they did not follow a normal distribution according to the Shapiro–Wilk test. The difference between the groups was calculated using a Mann–Whitney U test for the quantitative variables and chi-squared tests for the categorical variables. To determine the relationship between the main variables and the main outcomes, we calculated the Odds Ratio. Regardless of the median value, differences between groups for clinical outcomes were analyzed by a Log-Rank test. We established a p-value of <0.05 as a statistical difference. We conducted all statistical analyses using SPSS version 25 (SPSS Statistics for Windows, Version 25.0. Armonk, NY, USA: IBM Corp) software and generated figures using GraphPad Prisma version 7.

2.2. Ethical Considerations

This study was approved by the Research and Ethics Committees of the Hospital General de México “Dr. Eduardo Liceaga” (approval number: HGM-DG-024/2023). Due to the retrospective nature of the study, informed consent was waived. All data were anonymized to ensure patient confidentiality and in accordance with the Declaration of Helsinki and national research regulations.

3. Results

A total of 59 patients with a diagnosis of ALL who were no longer candidates for curative treatment and were receiving supportive care were included in this study. The gender distribution was balanced, and the median age was 31 years (range: 18–71 years). Among these, 33 patients were under 30 years old and exhibited no significant functional limitations. Only three patients had Down syndrome at the time they were deemed ineligible for intensive treatment. These patients received a maximum of one line of salvage chemotherapy.

The median leukocyte count at initial diagnosis was 14 × 10³/μL (range: 0.03–490 × 10³/μL), and 33.9% (n = 20) of patients were classified as hyperleukocytic at the time of treatment discontinuation. When patients were deemed ineligible for further curative therapy, the median total leukocyte count was 0.4 × 10³/μL (range: 0–1.0 × 10³/μL), consistent with profound leukopenia. This finding likely reflects severe bone marrow suppression secondary to prior high-intensity chemotherapy and, in some cases, may indicate extensive leukemic infiltration leading to pancytopenia. It does not necessarily represent peripheral disease activity, as circulating blast counts were minimal or absent in most patients at that stage.

Regarding risk stratification, 79.7% of patients were considered high-risk at diagnosis, primarily due to age and leukocyte count. High-risk cytogenetic abnormalities such as BCR::ABL1 and MLL rearrangements were infrequent. When analyzing relapse timing, 64.4% of patients experienced early relapse (<1 year from diagnosis), while 35.6% (n = 21) experienced late relapse (≥1 year). Demographic and clinical characteristics are shown in

Table 1. Demographic and clinical characteristics of study population.

Variables	Values
Age (years)	31.0 (18.0–74.0)
WBC (×103/μL)	14.0 (0.0–490.0)
Survivorship (days)	137.0 (11.0–1244.0)
Gender     Male     Female	29 (49.2%) 30 (50.8%)
Lineage at diagnosis     B cell     T cell	57 (96.6%) 02 (03.4%)
Risk at diagnosis     Standard     High	12 (20.3%) 47 (79.7%)
Relapse     Absence     Presence	38 (64.4%) 21 (35.6%)
Transfusion requirement     Without     High	25 (42.4%) 34 (57.6%)
Hyperleukocytosis     Absence     Presence	39 (66.1%) 20 (33.9%)
Age Risk     <35 years     >35 years	32 (54.7%) 27 (45.8%)
Overall survival     Live     Death	04 (06.8%) 55 (93.2%)

WBC: White blood cell count. Quantitative variables are presented as median (range); categorical variables are shown as absolute frequency and percentage.

.

3.1. Treatment

In our cohort, 6 patients (10.2%) received two lines of therapy, 36 patients (61.0%) underwent three lines, and 13 patients (22.0%) were treated with four lines. In the first-line setting, most patients were managed with high-intensity chemotherapy protocols adapted from pediatric-inspired or adult regimens, typically HyperCVAD or similar combinations. For second-line therapy, the most frequently used regimens were augmented HyperCVAD, FLAG (fludarabine, cytarabine, G-CSF), and mitoxantrone–etoposide–cytarabine. A small subset of patients proceeded to third-line or later therapies, which occasionally incorporated newer agents such as blinatumomab (4 patients, 6.8%), while none received inotuzumab. Hematopoietic stem cell transplantation was infrequent due to limited center availability, referral constraints, and pandemic-related barriers.

3.2. Supportive Care and Survival Outcomes

More than half of the patients required weekly transfusion support, mainly with packed red blood cells. An additional 25 patients required transfusions only occasionally.

At the time of analysis, the estimated median overall survival was 137 days (range: 11–1244 days). No significant differences in survival were observed based on relapse timing (early vs. late; log-rank, p = 0.677), age (p = 0.120), or initial risk stratification (p = 0.204). In contrast, transfusion requirement was the only variable significantly associated with prognosis (log-rank, p = 0.050). (Figure 1)

Only nine patients (15.3%) received multidisciplinary end-of-life care. These cases were evaluated using the Palliative Prognostic Index (PPI) or the Charlson Comorbidity Index. However, neither score demonstrated a statistically significant impact on survival (log-rank, p = 0.701). When these variables were incorporated into a multivariate model, none showed independent prognostic value.

4. Discussion

Our findings reinforce that transfusion dependence is a key adverse prognostic factor in relapsed/refractory ALL patients managed with supportive care. This aligns with prior reports indicating that high transfusion needs may limit the feasibility of home-based palliative care, particularly in resource-constrained settings where transfusion services are hospital-based [19].

The contemporary treatment of ALL in adults includes high-intensity chemotherapy combined with immunotherapy agents such as blinatumomab or Inotuzumab, and tyrosine kinase inhibitors for Philadelphia chromosome–positive cases. Allogeneic hematopoietic stem cell transplantation remains a cornerstone for long-term survival, particularly in settings where CAR-T cell therapy is not widely available [20].

Accurate risk stratification at diagnosis plays a crucial role in optimizing outcomes. In current practice, measurable residual disease (MRD)–oriented therapy has become the cornerstone for tailoring treatment intensity, guiding decisions on consolidation and transplantation. However, its implementation remains challenging in many centers due to limited availability of standardized assays, variability in sensitivity across techniques, and restricted access in low- and middle-income countries [21]. Despite therapeutic progress, prognosis remains poor for patients with refractory or relapsed disease unresponsive to salvage regimens [18]. For these individuals, treatment options are often limited to low-dose chemotherapy or best supportive care.

In Latin America, treatment success is influenced not only by disease biology but also by structural healthcare disparities. Access to novel agents, supportive therapies (e.g., transfusions, antibiotics), and proximity to referral centers significantly affect outcomes [22]. These challenges are even more pronounced in patients managed with palliative intent, for whom home-based care and outpatient transfusion support are rarely available.

In our cohort, the median survival was 137 days, aligning with previously reported estimates in similar populations. Notably, high transfusion requirements emerged as the only variable significantly associated with worse prognosis. This likely reflects greater disease burden and bone marrow failure, which in turn necessitate frequent transfusions. These transfusion requirements present a major barrier to home-based palliative care, as they often require recurrent hospital visits—especially in resource-limited settings where outpatient transfusion infrastructure is lacking [23].

Interestingly, we did not observe a statistically significant difference in overall survival between early and late relapse subsets, which contrasts with prior reports in adult ALL where early relapse is typically associated with worse prognosis [18]. This finding may reflect the impact of inadequate supportive care in our cohort. Limited access to timely transfusions, outpatient palliative programs, and novel agents likely minimized the potential survival advantage usually seen in late relapses. Thus, disparities in supportive care may have attenuated the prognostic effect of relapse timing in our setting.

Fatigue is the most frequently reported symptom among palliative patients with acute leukemia, significantly affecting both physical and emotional well-being [24]. These findings support the development of early integration models—such as the EASE (Emotion and Symptom-focused Engagement) program—which aims to monitor functional decline and improve quality of life from diagnosis onward [25]. We explored the use of prognostic tools such as the Charlson Comorbidity Index and the PPI to identify patients with poorer survival. However, only a small number of patients in our study were assessed using these tools, and neither index demonstrated significant discriminatory power in our cohort. While the Charlson Index has been validated in older AML patients to predict early mortality [26], and the PPI has shown utility in solid tumors [27,28], their applicability in relapsed/refractory ALL remains limited.

Transfusion dependence has been previously associated with increased healthcare utilization and poorer quality of life. High transfusion requirements frequently necessitate inpatient care, limiting access to hospice services and home-based palliation [29]. Integrating palliative care has been linked to reduced hospitalizations and a lower use of aggressive interventions at the end of life [30,31], Patients with hematologic malignancies, unlike those with solid tumors, are more likely to receive aggressive treatments, prolonged hospitalizations, and ICU admissions near the end of life [32,33].

In our experience, few patients accessed multidisciplinary end-of-life care, largely due to geographic, institutional, and socioeconomic barriers, as well as the impact of the COVID-19 pandemic. The inability to provide transfusions outside hospital settings remains a major limitation—even in high-income countries—and contributes to frequent readmissions and loss of autonomy [34].

A key strength of this study is that it provides real-world data from a population that is often underrepresented in clinical trials—patients with relapsed/refractory ALL receiving exclusively supportive care. The multicenter nature of the cohort and the detailed clinical variables analyzed add to the robustness of the findings.

However, the study has limitations. Its retrospective design may introduce selection bias and incomplete data capture, particularly for variables related to symptom burden and quality of life. The relatively small sample size limits the statistical power of multivariate analyses and may have contributed to the lack of discriminatory performance observed with the Charlson Comorbidity Index and Palliative Prognostic Index. Moreover, the study did not systematically assess patient-reported outcomes, which are critical in evaluating the true impact of supportive interventions.

Future research should explore prognostic models specifically validated in this population, incorporating both hematologic and patient-reported outcome measures. Moreover, qualitative studies could provide insight into patient and caregiver perspectives, guiding interventions aimed at improving quality of life in advanced ALL. Expanding access to community-based palliative programs and developing low-resource transfusion strategies could help bridge the care gap. Finally, multi-center prospective studies are warranted to validate these findings and assess interventions tailored to this high-risk group.

5. Conclusions

Efforts to incorporate palliative care earlier in the disease trajectory and to develop more tailored prognostic tools for this population are needed. Ultimately, improving access to outpatient transfusion services and integrating supportive care from the time of diagnosis may help optimize quality of life for patients with relapsed or refractory ALL who are no longer candidates for intensive therapy.

Future research should focus on developing prognostic models tailored to patients with relapsed/refractory ALL who are beyond curative treatment. Prospective studies incorporating quality-of-life metrics, transfusion needs, and symptom trajectories could help refine patient selection for palliative interventions. Additionally, exploring the integration of outpatient transfusion programs and early palliative care pathways in hematologic malignancies may improve both survival-related outcomes and the quality of the end-of-life experience.