Fitness assessment in acute myeloid leukemia: recommendations from an expert panel on behalf of the European LeukemiaNet Open Access

Numerous factors predict short- and long-term outcomes in patients with acute myeloid leukemia (AML).¹ With many new therapeutic options becoming available, these factors could serve to assess the risks and benefits of a specific therapeutic intervention for each individual patient. However, considering the diversity of patients, including variations in disease-related characteristics, biological factors, personal circumstances, and preferences, no single clinical or biological parameter provides sufficient information for this purpose.² Thus, several models integrating various parameter combinations have been developed for daily clinical practice.^3-6 The application of these models is referred to as a fitness assessment, because it aims to select which of the currently available treatments would better “fit” each patient’s unique clinical context.⁷ Although a precise, comprehensive fitness assessment is widely recognized as a critical step of care, there is still no consensus on its exact definition or which specific parameters should be included.⁸ To help address this, an international panel of experts convened on behalf of the European LeukemiaNet (ELN), focused on relevant aspects of fitness in patients with AML, aiming to (1) provide tools for a general definition of fitness/unfitness, and (2) identify and categorize individual factors that contribute to this definition.

Thirty-one hematologists were convened based on their qualifications and track record in the clinical management of AML, in geriatrics and quality-of-life (QoL) assessment. Furthermore, several members had previously contributed to guidelines on fitness definition and geriatric assessment in AML. Moreover, patients’ representatives from the Acute Leukemia Advocated Network (ALAN) provided additional insights and contribution based on their evidence-based advocacy work. After a comprehensive review of recent literature covering the last 25 years, a set of 21 statements was developed. Literature research was primarily conducted by the steering committee (A.V., R.P., L.M., D.d.L., G.O., and F.E.), although all authors contributed to this task. The statements covered 5 main aspects: general definitions (n = 5); fitness components (n = 4); influence of fitness and disease genetics on treatment choice (n = 4); timing for fitness/genetic determination (n = 4); expected impact on QoL; and available social support (n = 4). Based on the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) approach, each statement was assigned an evidence level (level of evidence [LoE]) and a recommendation grade (grade of recommendation [GoR]).⁹ Subsequently, a 2-rounds Delphi poll was conducted to determine the level of agreement on the statements.^10,11 Participants rated their level of agreement on a Likert scale of 1 (completely disagree) to 5 (completely agree). Consensus on each statement was defined as ≥70% agreement (ratings of 4 or 5) or disagreement (ratings of 1 or 2). After the first round of voting, 3 statements (5, 12, and 20) did not reach a consensus and were adjusted after discussion by the steering committee. The resulting statistics were presented in the final discussion with the participants. Despite adjustments after round 1, 2 statements (12 and 21) still did not reach a consensus (supplemental Materials). Statistics and figures were computed using R version 4.2.1.¹² 

Age as a single parameter is a significant prognostic factor¹³; however, it also serves as a surrogate for comorbidities, functional limitations, and frailty (Table 1). Age significantly affected outcomes in patients with poor performance status (PS) but had only a minor effect on outcomes of older patients with excellent PS.¹⁴ Analysis of the Swedish Acute Leukemia Registry showed that older patients with good PS had low early death rates, whereas those with poor PS had increased early mortality across all ages.¹⁵ Poor PS has been associated with increased 30-day and 8-week mortality, lower rates of complete remission (CR), and shorter overall survival (OS).^16,17 Comorbidities should be assessed in patients with AML to help define overall fitness for a specific treatment. They are also an independent predictor of CR.^6,18,19 Comprehensive comorbidity indices, such as the Charlson Comorbidity Index and the Hematopoietic Cell Transplantation (HCT)–specific comorbidity index (HCT-CI), determine organ dysfunction and predict early death rates after induction chemotherapy or transplant in AML.^5,19,20 The Società Italiana di Ematologia/Società Italiana di Ematologia Sperimentale/Gruppo Italiano Trapianto Midollo Osseo consensus criteria defined fitness for intensive chemotherapy (IC) or non-IC by associating the presence of a set of geriatric and comorbidity parameters with a specific treatment choice.³ A large retrospective analysis of patients treated with IC confirmed that these criteria had excellent accuracy in predicting 28- and 100-day mortality.²¹ The AML composite model (which incorporates HCT-CI, cytogenetic risk, and age) also accurately predict early and late mortality in patients with AML treated with IC.²² Because aging and age-induced frailty are not solely the result of comorbidities, a comprehensive geriatric assessment (CGA) that examines multiple factors (comorbidities, physical function and cognition, depression or geriatric syndromes, malnutrition, polypharmacy, and social isolation) has been shown to predict IC side effects and OS.^6,23,24 AML CGA can detect physical and cognitive abnormalities even in patients considered fit by standard oncological evaluation.^6,25 The feasibility of CGA has also been demonstrated in multicenter trials of patients treated with hypomethylating agents (HMAs).^25,26 

PANEL RECOMMENDATION: Several parameters, including age, PS, comorbidities, and a CGA, concur to influence a patient’s ability to tolerate AML-directed therapies. Therefore, a comprehensive, patient-centered evaluation of these factors should be conducted before initiating any therapeutic program.

Current treatments for patients with AML are divided into intensive and low intensity. IC options include cytarabine and anthracycline (7+3) with or without the combination of other agents, (gemtuzumab ozogamicin, midostaurin, or targeted agents) or CPX-351.²⁷ Low-intensity therapies include HMAs, venetoclax (VEN) combined with HMAs, or low-dose cytarabine (LDAC), glasdegib combined with LDAC, ivosidenib or enasidenib single agents or combined with azacytidine, and single agents gentuzumab ozogamicin or gilteritinib.^27,28 Pretreatment cytogenetics and mutational profiles are important independent predictors of treatment effectiveness. Poor-risk karyotypic abnormalities and mutational profiles, as defined by the ELN genetic risk stratification guidelines,²⁹ are more prevalent in older patients and may confer resistance to chemotherapy. In high-risk patients, IC regimens such as 7+3 are associated with low CR rates and poor OS.³⁰ Other disease features, such as AML evolving from an antecedent myeloid neoplasm or therapy-related AML, also significantly influence outcomes.³¹ Thus, besides clinical fitness, biological characteristics should be considered in the decision-making process (Table 1).³² 

PANEL RECOMMENDATION: The diverse intensities of available treatments emphasize the necessity of recognizing additional fitness levels extending beyond the conventional categories of “fit,” “unfit,” and “frail.” Therefore, patients should be considered “eligible” for a specific treatment strategy based on clinical fitness and biological factors, with a clear explanation of the treatment approach and the provision of alternative options.

Allogeneic stem cell transplant (allo-SCT) is an important step in achieving a curative outcome for patients with AML, yet its potential benefits must be weighed against the risk of transplant-related mortality (TRM). Age and, more importantly, PS and CGA predict outcomes in patients undergoing allo-SCT³³ and are commonly assessed to determine their suitability for the procedure.^34,35 Because no absolute upper age limit exists to define allo-SCT eligibility/ineligibility,²⁹ dedicated tools including HCT-CI, are currently available to predict outcomes.^35,36 For patients aged ≥70 years, a geriatric assessment before allo-SCT may support a judicious selection of those who are most likely to benefit from the procedure.³⁷ However, patients considered fit for both IC and allo-SCT at diagnosis may experience health deterioration during the treatment and become ineligible for allo-SCT. Conversely, some patients may improve during treatment after achieving a response, becoming eligible for allo-SCT (Table 1).

PANEL RECOMMENDATION: Because a patient’s fitness status can change during treatment, it should be reassessed before each treatment phase, including before allo-SCT.

Allo-SCT significantly reduces relapse risk compared with chemotherapy alone and represents an important option for fit patients with AML.³⁸ Currently, allo-SCT indication is decided based on disease risk.^29,39 However, the high morbidity and mortality rates associated with allo-SCT remain a major drawback.⁴⁰ Accordingly, patients in good general condition and with an acceptable estimated TRM are considered suitable candidates for allo-SCT.⁴¹ Based on this, only patients receiving standard induction and consolidation chemotherapy are usually considered for transplant. However, allo-SCT has shown promising results in patients achieving CR after less-intensive therapies, with 40% to 55% of them remaining in CR and alive 1 year after transplant.^42-44 Although nonintensive approaches cannot yet be considered the standard of care, they may serve as a “bridge to transplant” for selected subgroups of patients, reducing treatment-related toxicities (Table 1).⁴⁴ 

PANEL RECOMMENDATION: Ineligibility for IC does not necessarily exclude eligibility for allo-SCT in patients achieving CR, after nonintensive treatment.

Antileukemic treatment is the straightforward approach to reducing symptoms, prolonging life, and maintaining patients’ QoL, particularly compared with supportive care only. In the AML-001 study, azacitidine (AZA) improved survival over best supportive care (BSC) alone, even among patients who were preselected for BSC. The median OS was 5.8 with AZA vs 3.7 months with BSC. Moreover, no QoL deterioration was observed in either groups during treatment.⁴⁵ The addition of VEN to HMA or LDAC significantly prolonged survival and preserved QoL, thereby further underlining the correlation between active treatment, response, prognosis, and improved QoL.⁴⁶ Data from the Swedish registry and the US American SEER database showed that any cytoreductive treatment was associated with lower early death rates and longer survival.^15,47 These results implicate that any antileukemic treatment is preferable to no treatment, with the potential side effects outweighed by the benefits of extended survival and maintained QoL (Table 1).

PANEL RECOMMENDATION: Antileukemic therapy should be considered for mitigating disease-related symptoms, even if a curative intent is unattainable. Decisions regarding symptom management should be informed by patient preferences and QoL objectives, ensuring they align with individual values and expectations. Real-life data collection and activation of dedicated registries should be implemented to better investigate the impact of this strategy on clinical practice and patients' lives.

Advanced age is a predominant determinant in assessing patients’ fitness and suitability for IC^48,49 and is linked to an increased likelihood of comorbidities, resulting in low CR rates and shorter OS.^50-53 However, as life expectancy increases within the general population, many older individuals reach advanced age in a better overall health.⁵⁴ Consequently, chronological age does not inherently reflect biological age.⁴ Indeed, early TRM rates in physically fit older patients are comparable to younger ones and are potentially even better than those observed in younger patients with multiple comorbidities (Table 2).¹³ 

PANEL RECOMMENDATION: Although fitness for IC declines with aging, age as a single parameter should not be considered as the sole criterion to define patients’ fitness, Overall health status, physiological reserve, QoL, and life expectancy are additional components to examine.

Concerns regarding administration of IC to very old yet fit patients have been reinforced by recent changes in AML treatment.^55,56 Novel less-intensive therapies have significantly improved response rates and OS in patients ineligible for IC.^42,57-61 Particularly, less-intensive VEN-based approaches have demonstrated promising efficacy and better tolerability in older patients, and those considered at high risk for TRM.^62-64 A trial by the European Organization for Research and Treatment of Cancer and Gruppo Italiano Malattie Maligne dell’Adulto (EORTC-GIMEMA 1301) randomized fit patients aged >60 years between 10-day decitabine and IC and demonstrated comparable survival rates and improved QoL in decitabine-treated patients.⁴⁴ Randomized trials are needed to determine whether the addition of VEN will increase responses without elevating toxicity, thereby enhancing the benefit of VEN-based regimens over IC (Table 2).^65-69 

PANEL RECOMMENDATION: Given that only a small proportion of patients aged >75 years can benefit from IC and considering the lack of clinical trials focusing exclusively on intensively treated patients in this age group, careful evaluation of eligibility for IC is needed for patients aged >75 years. It is crucial to also consider the patient’s QoL expectations, psychosocial circumstances, and willingness to undergo IC.

A correlation was observed between PS and the risk of complications in adults undergoing IC.^{4,15,17,18,70} Patients with a good Eastern Cooperative Oncology Group PS have comparable early mortality rates across all age groups when treated with IC.¹³ In contrast, in patients with PS score of ≥3, the probability of early death substantially increases with age.¹³ Older patients with poor PS have a high likelihood of early mortality, irrespective of treatment intensity. Nevertheless, a subset of these patients achieves long-term survival, indicating that selected individuals may potentially benefit from IC.^15,71 Conversely, physiologic reserve capacity varies considerably within the geriatric population. Even patients with a favorable PS may show significant variations and impediments in cognitive, emotional, and physical well-being, potentially affecting outcomes (Table 2).^23,72 

PANEL RECOMMENDATION: PS should be determined during treatment planning. Nevertheless, considering its subjective nature and limited sensitivity, a geriatric assessment is particularly useful for patients with a PS score of >2 to inform decision about treatment intensity. For patients with a PS score of >3, the decision between active treatment and supportive therapy alone should be made on a case-by-case basis.

Several studies showed that patient-reported physical functioning is significantly associated with OS.^73-76 This prognostic value remains regardless of whether physical functioning is assessed by patients or physicians.⁷⁷ Combining physician-assessed and patient-reported scores may represent a valid approach to enhancing prognostic accuracy, as physical functioning scales alone do not fully capture the spectrum of functional abilities in older patients. More detailed self-reported questionnaires may assess a patient’s ability to perform basic activities of daily living needed for independence at home or instrumental activities of daily living required for community independence.^23,77,78 Findings from ALAN underscore key factors, such as age, sex, and socioeconomic status, that clinicians should consider when developing support strategies.⁷⁹ Additionally, the role of social functioning in relation to patient QoL must be incorporated into these considerations (Table 2).

PANEL RECOMMENDATION: Although retrospective self-evaluation may provide biased results, patient-reported physical and social functioning before AML onset should be considered appropriately.

Patients with unfavorable genetic risk have lower remission rates and shorter OS after standard IC compared with those with more favorable profiles.^80-85 Retrospective analyses indicated that less-intensive, HMA-based treatments are at least equally effective and better tolerated than IC in patients with adverse-risk AML.^86-89 In a relatively small, propensity-matched cohort of patients with AML treated with IC or VEN-AZA, age of >65 years, ELN adverse risk, and a RUNX1 mutation favored VEN-AZA for OS, whereas IC was superior in patients with ELN intermediate-risk disease.⁶³ In real-world data, early mortality and OS were not significantly different between patients treated with CPX-351 or VEN-AZA. Similarly, no survival benefit with any treatment was observed in patients with adverse-risk AML. Conversely, CPX-351 treatment was associated with higher rates of documented infections, neutropenic fever, and longer inpatient hospital stays compared with VEN-AZA (Table 3).⁹⁰ 

PANEL RECOMMENDATION: Treatment decisions should be always individualized and made in partnership with patients after careful consideration of all patient- and disease-related characteristics that may affect efficacy or tolerability of any therapeutic intervention.

Survival in patients unable to tolerate any active therapy is very poor.^47,91,92 Ivosidenib and enasidenib are paradigmatic examples of effective targeted therapies that can offer advantages over supportive therapy.^93,94 Results from a phase 1 study indicate that ivosidenib monotherapy showed CR/CR with partial hematological recovery rates of 42.5% in patients with newly diagnosed isocitrate dehydrogenase-1 (IDH1)-mutated AML who were considered too frail to tolerate HMA-based treatment. Responses were durable with a median OS of 12.6 months.⁹⁵ In 39 patients who are frail with newly diagnosed IDH2-mutated AML who received enasidenib monotherapy, the CR/CR with incomplete count recovery rate, ORR, and median OS were 21%, 31%, and 11.3 months, respectively.⁹⁴ These outcomes are superior to those reported for BSC alone in noncomparative studies.⁹⁶ Other oral, targeted, small-molecule inhibitors (eg, menin inhibitors, SYK inhibitors, etc) are currently available in clinical trials and may offer future options for patients with AML who are frail (Table 3).

PANEL RECOMMENDATION: Patients for whom supportive therapy seems the only option should be tested for actionable mutations.

The negative impact of adverse genetic/cytogenetic profiles on chemosensitivity has been extensively reported.⁹⁷ Considering disease-related factors can potentially alter the traditional paradigms and clinical fitness algorithms.⁷ The EORTC-GIMEMA randomized trial investigating 10-day decitabine vs 7+ 3 regimen demonstrated a CR/CR with incomplete count recovery rate of 44% and 43% with decitabine and 7+3, respectively. However, in patients within the ELN 2017 adverse-risk category, the OS was poor, particularly in those with Tp5 mutations, consistent with other studies.⁴⁴ Randomized studies are investigating whether standard cytotoxic chemotherapy, such as 7 + 3 or CPX-351–based regimens, vs lower-intensity VEN combinations should be preferred for older, fit patients who are candidates for allo-SCT. This approach, followed by a prompt bridging to allo-SCT, may lead to durable response with improved OS/disease-free survival.^65,68 In this highly heterogeneous group of patients with adverse genetic and cytogenetic profile,²⁹ the differential impact of IC and nonintensive strategies remains controversial. The impact of new agents and combinations, within adverse subgroups, exemplifies the current complexity of treatment decision-making (Table 3).⁹⁸ 

PANEL RECOMMENDATION: Although an agreement was not reached, 67% of the panelists concurred that no clear contraindications exist to treating adverse risk, older, fit patients with adverse-risk profiles using IC. The lack of consensus on this issue stems from unsatisfactory results when treating patients with adverse risk with either intensive or nonintensive approaches. The authors agree on the necessity to implement treatment strategies for this patient population.

Mental disorders and overall cognitive status may significantly affect therapy compliance and lead to prolonged hospitalizations. Along with physical function and comorbidities, geriatric scales also explore domains such as cognition, depression, and overall psychological function.^6,99 Cognitive function evaluation through CGA may uncover undiagnosed cognitive impairments that could increase the risk of complications during and after intensive treatment.⁷⁸ However, few studies have addressed the impact of mental disorders, including cognitive deterioration, on clinical outcomes in patients treated non-intensively (Table 3).

PANEL RECOMMENDATION: Psychiatric comorbidities and cognitive disorders should be carefully assessed to inform therapeutic decisions, as they may affect patient compliance with therapy and increase the risk of complications. Involving specialists, such as psychiatrists or psychologists, for targeted support may be beneficial.

Whether the time from AML diagnosis to treatment affect the short- and long-term outcome has been addressed in recent years.^100-102 Such an impact appears negligible in patients with stable disease. Therefore, waiting a short period to better characterize leukemia, manage concurrent and transient medical conditions, and then design adapted treatments seem feasible with low risk for the patient (Table 4).

PANEL RECOMMENDATION: A comprehensive approach that considers both patient- and disease-related factors is recommended. Because diagnostic workup may require time, patients with clinically stable disease may wait up to 3 weeks^100-102 from diagnosis to treatment. Patients should be informed that the time spent on comprehensive assessments is critical for tailoring the most effective treatment plan.

AML presentation often involves significant deterioration in the patient’s general condition.¹⁰⁰ Therefore, it should be carefully evaluated whether the general condition is affected by the underlying AML or reflects the patient’s usual health status (Table 4).¹⁰³ Supportive therapy should be initiated at AML presentation to revert disease-induced conditions that hamper accurate fitness assessment,¹⁰² potentially leading to overtreatment or undertreatment.⁷ A more objective measure of fitness status could be provided by CGA, because such methodology is only moderately affected by disease burden (Table 5).

PANEL RECOMMENDATION: If poor general condition may be attributed to disease burden rather than preexisting conditions, fitness level should be reassessed after adequate supportive therapy.

Just as a patient’s clinical fitness may improve after supportive therapy at disease presentation, additional changes may also occur during or after treatment.¹⁰⁴ In some cases, IC can worsen the patient’s general condition, requiring deintensification of the subsequent therapeutic strategy.^105-108 Conversely, patients initially considered ineligible for IC may improve after nonintensive approaches.¹⁰⁹ Therefore, dynamically reevaluating fitness status is critical to recalibrate treatment intensity according to the patient’s actual clinical conditions (Table 4).⁴³ 

PANEL RECOMMENDATION: Fitness modifications should be assessed dynamically throughout therapy to modulate treatment intensity as appropriate. Patients and caregivers should be kept informed about how fitness changes may affect treatment adjustments, ensuring they understand the rationale behind modifications in therapy intensity.

Genetic/cytogenetic profiling is essential at baseline to define disease risk and is also recommended in relapsing/refractory AML to intercept clonal transformations.^110-112 Identifying new, emerging molecular mutations is crucial to personalize salvage therapy.¹¹³ This is the case of FLT3 mutations, that can be acquired in up to 25% of relapsing cases.¹¹⁰ Indeed, novel targeted agents, such as FLT3-inhibitors, have been effective as salvage therapy, offering lower toxicity than IC.¹⁰⁶ This approach may benefit patients who, once in remission, become candidates for allo-SCT.^114,115 From this perspective, molecular retesting represents a fundamental step, providing therapeutic alternatives for patients who were historically considered only for BSC (Table 4).

PANEL RECOMMENDATION: Biological reassessment should be performed at disease relapse to assist the choice of salvage therapy.

A meta-analysis found a relationship between perceived social support, social network, and marital status with OS in patients with cancer, including those with leukemias/lymphoma.¹¹⁶ Perceived social support had a stronger association with reduced mortality in patients with leukemia/lymphoma than in those with other cancers.¹¹⁶ Patients with AML living alone were less likely to receive IC or undergo postremission allo-SCT, potentially leading to worse outcomes.¹¹⁷ An ALAN study demonstrates that living alone, younger age, and being female are associated with higher rates of nonadherence to treatment. Proactive strategies should be implemented before treatment initiation to address all these factors.⁷⁹ In a study of hospitalized patients with hematologic malignancies, limited social support was independently associated with a shorter OS and a higher likelihood of death or readmission within 90 days of hospital discharge (Table 6).¹¹⁸ 

PANEL RECOMMENDATION: Social support provided through a multidisciplinary approach involving professionals beyond physicians should be offered to all patients from the beginning of therapy.

Most patients with AML are older and often lack caregivers.¹¹⁹ When caregivers are present, they may experience high levels of burden.¹²⁰ Therefore, alleviating distress and enhancing their skills and service capabilities should be considered.¹²¹ Accordingly, increasing social support may be a decisive component in managing patients with AML (Table 6).

PANEL RECOMMENDATION: Although social support is not a component of fitness, it may affect the success of a therapeutic program. Accordingly, patients’ social support should be intensified when considering any therapeutic interventions, especially in the absence of caregivers.

QoL is an end point nowadays incorporated in AML trials to determine the value of novel drugs.⁴⁶ Systematic QoL data collection in cancer care is associated with several benefits, including improved symptom control, better patient-physician communication, and higher satisfaction with care.^122,123 Although not specifically in the AML setting, previous trials showed that monitoring patient-reported symptoms is associated with longer survival than traditional follow-up care.^124-126 Investigating QoL indicators provides independent prognostic information beyond PS or other clinical and genetic tests, a finding replicated across several cancer populations, including those with hematologic malignancies.^127-130 The independent prognostic value of QoL data was also observed in patients with high-risk myelodysplastic syndromes and AML,^131,132 further emphasizing its importance in these patients. Findings from one ALAN study revealed that demographic factors most strongly associated with poorer QoL include younger age, female sex, and lower income (Table 6).⁷⁹ 

PANEL RECOMMENDATION: Although there is insufficient evidence to consider QoL as a measure of fitness, any antileukemic treatment significantly affects patients' QoL. Therefore, the extent of such an impact should be considered and discussed with patients when making therapeutic decisions and monitored throughout treatment.

EPC refers to the integration of palliative care into active disease treatment within 8 weeks of cancer diagnosis.¹³³ EPC includes communication with patients, symptom management, social coping support, regular follow-up visits addressing physical and psychological symptoms, as well as setting realistic treatment goals and facilitating patient choices about care planning.¹³⁴ In patients with advanced non–small-cell lung cancer,¹³⁵ EPC resulted in longer OS than standard care, along with significant improvements in QoL, mood, and survival advantages.¹³⁶ 

A Cochrane systematic review and meta-analysis published in 2017 confirmed the value of EPC in reducing symptom intensity and improving QoL vs standard cancer care.^134,137-141 Similar benefits were observed in patients with AML. EPC is particularly valuable for patients with AML because they are burdened by several physical and psychological symptoms and have several unmet palliative care needs.¹⁴² In a trial enrolling 160 patients with high-risk AML receiving IC, those receiving EPC reported significantly better QoL, along with reduced depression, anxiety, and posttraumatic stress disorder symptoms at both early (week 2) and late (week 24) time points compared with those receiving standard care. The study concluded that EPC should become standard of care for patients with AML.¹⁴³ In a retrospective study, EPC was associated with high frequency of quality palliative care indicators and low rates of treatment aggressiveness at the end of life.¹⁴⁴ However, consensus was not reached because of factors including the novelty of the topic in AML setting and the experts’ focus on prioritizing life-saving interventions. Future studies should explore this crucial aspect further (Table 6).

PANEL RECOMMENDATION: Although not reaching a consensus, 67% of the panelists agreed that EPC should be standard of care, even for patients receiving IC, to improve QoL, and treatment outcome. EPC should be integrated at diagnosis, tailored to the patient’s goals, preferences, and expectations, with a focus on symptom management and psychosocial support.

Given the current therapeutic options, fitness assessment in patients with AML is becoming increasingly complex yet essential. The growing knowledge of AML genetics, alongside the development of novel targeted agents, has complicated the process. The continuous discovery of new gene mutations, which can serve as therapeutic targets, has challenged old-fashioned, dualistic, “yes or no” approach (fit or unfit for IC) and introduced a spectrum of intermediate situations in which almost any patient with AML could be considered for active treatment. Moreover, in recent years, QoL assessments and patient-reported outcomes have gained prominence, providing a more comprehensive understanding of the patients' experience and perspective. The present recommendations represent an effort to collate and harmonize all these evolving factors. Similar studies have recently dissected this topic¹⁴⁵ but they mainly focused on “clinical- and biological-oriented” therapeutic programs. This work also contemplates a broad range of “patient-centered” aspects, thus extending the concept of fitness to include the patients’ reported perception of their well-being and preferences. Clear communication of these fitness levels to patients, using nontechnical language, is essential to support informed decision-making. Future research should deepen the involvement of patients’ and caregivers’ representatives. By doing so, we may gain valuable insights into the definition of fitness in the AML setting. Physicians’ awareness of patients’ and caregivers’ expectations and goals can contribute to a more informed and accurate treatment selection, ensuring that it is a shared decision.

The panel is very grateful to the European LeukemiaNet, on behalf of which this task was conducted.

Contribution: A.V. and G.O. conceptualized and designed the study; R.P. and L.M. performed literature research; L.L.N. and J.T. organized the meetings and analyzed the Delphi polls; A.V., R.P., L.M., C. Röllig, A.W., D.d.L., F.E., and A.C. were responsible for section assembly and manuscript writing; G.O., S.N., and C.G.M. revised the manuscript; and all authors were responsible for final approval of the manuscript.

Conflict-of-interest disclosure: A.V. reports research funding from Jazz Pharmaceuticals, and consultancy for Amgen, Servier, AstraZeneca, Pfizer, Kite-Gilead, AbbVie, Janssen, Astellas, Astex, Otzuka, Stemline Menarini, Bristol Myers Squibb, Glycostem, Novartis, and Delbert (all are not related to this manuscript). R.P. reports speakers bureau participation with AbbVie, Janssen, Jazz, Pfizer, Astellas, Novartis, and Sobi; and served as consultant for Janssen and AbbVie (all are not related to this manuscript). C. Röllig reports institutional research funding from AbbVie, Novartis, and Pfizer; and advisory honoraria from AbbVie, Astellas, Bristol Myers Squibb, Jazz, Janssen, Novartis, Otsuka, Pfizer, and Servier (all are not related to this manuscript). D.d.L. reports consultancy for Servier, Roche, and AbbVie; and scientific advisory board participation for Takeda and Servier (all are not related to this manuscript). F.E. reports consultancy or advisory role for AbbVie, Incyte, Novartis, and Jazz Pharmaceuticals; and research funding (institution) from Daiichi Sankyo (all are not related to this manuscript). A.C. reports advisory board role for AbbVie, Novartis, and Menarini-Stemline; and research support from AbbVie (all are not related to this manuscript). A.A. reports honoraria from AbbVie, Bristol Myers Squibb, and Novartis (all are not related to this manuscript). L.B. reports honoraria from AbbVie, Amgen, Astellas, Bristol Myers Squibb, Celgene, Daiichi Sankyo, Gilead, Hexal, Janssen, Jazz Pharmaceuticals, Menarini, Novartis, Pfizer, Roche, and Sanofi; and research support from Bayer and Jazz Pharmaceuticals (all are not related to this manuscript). J.E. reports participation in advisory boards of AbbVie, Novartis, Astellas, Jazz Pharmaceuticals, Bristol Myers Squibb–Celgene, Pfizer, and Amgen; and research grants from Novartis, Jazz Pharmaceuticals, and Pfizer (all are not related to this manuscript). M.H. reports research funding to institutions from AbbVie, Servier, Astellas, BerGenBio, Glycostem, Jazz Pharmaceuticals, Karyopharm, Loxo Oncology, Novartis, and PinotBio (all are not related to this manuscript). P. Mehta reports speaker/advisory fees from AbbVie, Astellas, Daiichi Sankyo, Jazz Pharmaceuticals, Menarini, and Pfizer. P. Montesinos reports receiving research support, consultancy fees, speakers bureau fees, and advisory board membership from Agios, Astellas, AbbVie, Pfizer, Daiichi-Sankyo, Tolero Pharmaceuticals, Incyte, Novartis, Forma Therapeutics, GlycoMimetics, Bristol Myers Squibb/Celgene, Janssen, Karyopharm, Sanofi, Servier, Syndax, and Teva (all are not related to this manuscript). C. Récher reports consulting or advisory role with AbbVie, Amgen, Astellas, Bristol Myers Squibb, Boehringer, Jazz Pharmaceuticals, and Servier; received research funding from AbbVie, Amgen, Astellas, Bristol Myers Squibb, Iqvia, and Jazz Pharmaceuticals; and received support for attending meetings and/or travel from AbbVie, Novartis, and Servier. N.V. received honoraria from AbbVie, Amgen, Bristol Myers Squibb/Celgene, Daiichi, Novartis, and Roche; reports consultancy with Novartis, Janssen, Biokinesis, and Servier; and reports research grants from Bristol Myers Squibb, Jazz, and Innate Pharma. R.B.W. received laboratory research grants and/or clinical trial support from Aptevo, Celgene, ImmunoGen, Janssen, Jazz, Kite, Kura, and Pfizer; has ownership interests in Amphivena; and is (or has been) a consultant to AbbVie, Adicet, Amphivena, BerGenBio, Bristol Myers Squibb, GlaxoSmithKline, ImmunoGen, Kura, Orum, and Wugen (all are not related to this manuscript). E.W. reports advisory/consulting role with AbbVie, Bristol Myers Squibb, CTI Biopharma, Blueprint, Immunogen, Kite, Kura, Johnson & Johnson, Qiagen, Rigel, Schrodinger, Daiichi Sankyo, Gilead, GlaxoSmithKline, Stemline, Syndax, Novartis, NuProbe, Rigel, Schrodinger, and Sellas; reports speaking for Pfizer, Astellas, and Dava; and served on the data safety and monitoring committee of AbbVie and Gilead (also research scholar award committee; all are not related to this manuscript). The remaining authors declare no competing financial interests.

Correspondence: Adriano Venditti, Hematology, Department of Onco-Hematology, Fondazione Policlinico Tor Vergata, Viale Oxford 81, Rome 00133, Italy; email: [email protected].