How to approach shared decision making when determining consolidation, maintenance therapy, and transplantation in acute myeloid leukemia

For those of us who care for patients with acute myeloid leukemia (AML), the approval of new therapies has had a significant impact not only on what treatments we have to offer, but also on how we discuss these options with patients and families. In addition to the complexity of explaining leukemogenesis and molecular mutations, we must also discuss differences in time to remission, length of treatment, need for hospitalization, and adverse effects. With experience we may refine our initial “AML talk” and be able to modify sections based on individual patient and disease characteristics. Although these conversations are well intentioned, they can be technically complex and one-sided as they tend toward delivering information and laying out options, risks, and benefits of therapy.¹  It is less likely that they begin by asking patients and families what information they want to know or how they prefer to receive this information. Similarly, an assessment of a patient’s values and how they wish to participate in the decision-making process with their family may not be discussed at the time of diagnosis.¹ 

For most patients, AML does not carry the familiarity of a more common solid tumor such as breast cancer, and it has come on with limited warning. This lack of understanding coupled with the suddenness of diagnosis has a significant impact on a patient’s ability to process initial conversations when key decisions are being made.¹  Research has confirmed that many patients are so stressed that they may not be able to participate in the decision-making process in a meaningful way and make decisions quickly.^2,3  Patients may also feel as though they are making a choice about choosing to live or die, not realizing that there is more than 1 option for treatment. In addition to underestimating the number of treatments available, patients also tend to overestimate the possibility of cure, as well as the risk of intensive therapy.¹  One of the more extreme examples of impaired information processing becomes apparent when patients present for the first outpatient visit after induction chemotherapy and are unaware that there is a need for additional chemotherapy and possibly bone marrow transplantation.

Unfortunately, we may not become aware of how our patients are processing information until new decision points arise such as making choices for consolidation, whether or not to consider maintenance therapy, or assessing the role of bone marrow transplantation. How might we help patients process information so they can make knowledgeable decisions about their therapy? Shared decision making (SDM) is an approach to medical decision making for those situations in which most clinicians would agree that there is more than 1 correct choice for a patient.⁴  During the SDM process, there is first a sharing of information between the physician and patient about existing options and the potential risks and benefits of each choice.⁴  Although this sounds quite similar to our usual patient conversations, SDM differs in that patients are encouraged to consider their personal preferences and to make a decision based on what matters most to them after understanding pertinent benefits and harms. With this type of approach, patients feel better informed about their decisions and express satisfaction with the plan and process.⁵  A systematic review of studies that measured SDM during a patient-clinician interaction and then evaluated the relationship of SDM with affective-cognitive, behavioral, and physiological and health outcomes confirmed this finding.⁶  In 66% of the included studies, affective-cognitive outcomes, which include patient satisfaction, decisional conflict, and general perceptions about the interaction with the clinician, were statistically associated with SDM in a positive manner.⁶  One of the initial models of SDM uses a framework with 3 sections: choice talk, option talk, and decision talk, which has evolved to team talk, option talk, and decision talk^4,7  (Figure 1; Tables 1-3). In both models, the physician notes that there are choices and describes them (choice/team talk), discusses these options in more detail (option talk), and then encourages patients to make a decision based on their preferences after understanding the risks and benefits (decision talk). Of note, while most physicians broadly support the inclusion of patient preference in decision making, commonly cited barriers to SDM include perceived time constraints, patient characteristics that suggest SDM cannot be successfully applied, and the notion that SDM does not have a role in certain clinical situations.⁸  Interestingly, data do not suggest a difference in the length of a clinic visit, but that the structure of the consultation did change after SDM was implemented, likely to accommodate the new interventions.⁹  Regardless of level of training or years of experience, evidence suggests that training programs for physicians are essential to effectively implement SDM into regular practice.¹⁰  In the following 3 cases we discuss how to incorporate SDM into decisions about consolidation, maintenance therapy, and transplantation for patients with AML.

Patient 1 is a 67-year-old woman with a history of hypertension who presented to her physician for evaluation of increasing shortness of breath and was found to have a white blood cell count of 120 000. She is an active former kindergarten teacher and was otherwise asymptomatic. Her bone marrow biopsy confirmed a diagnosis of AML with normal cytogenetics and mutations in DNMT3A and TET2. She underwent leukapheresis and was started on 7+3 chemotherapy, which she tolerated well aside from anticipated cytopenias and febrile neutropenia. Her end-of-treatment bone marrow biopsy confirmed a morphologic complete remission (CR) with persistence of the DNMT3A mutation. She has been doing well at home and presents to the clinic for further recommendations for treatment.

This is the first time you are meeting her, and after reviewing her hospital course and bone marrow results, you assess her understanding of next steps for treatment. Despite the intentions of your colleagues, she left the hospital unaware of the need for additional chemotherapy and the probability of relapse. She is at her appointment with a friend today because she lives 2.5 hours away and “doesn’t like to drive in the big city.” She thought that this would be her final and only visit with you.

Without additional chemotherapy after achieving a morphologic CR, patients with AML will experience relapse of their disease.¹¹  Although most physicians, if not all, would agree with this statement, there are varied opinions as to which consolidation chemotherapy is the best for preventing relapse for patients older than age 60 years who are not proceeding to allogeneic stem cell transplantation (allo-SCT). Furthermore, we have yet to identify a regimen that prolongs survival in this patient subgroup. Potential options include a regimen similar to cytarabine and daunorubicin induction or assorted doses and schedules of cytarabine (1 to 1.5 g/m² once per day to twice per day on days 1 to 5 or on days 1, 3, and 5), possibly with an anthracycline or in combination with gemtuzumab ozogamicin.^12,13  HOVON97 was a phase 3 trial of patients age 60 years or older with AML or myelodysplastic syndrome (MDS) refractory anemia with excess blasts who had achieved a CR or CR with incomplete count recovery (CRi) after at least 2 cycles of cytotoxic chemotherapy who were then assigned to either observation or 12 cycles of azacitidine 50 mg/m² once per day for 5 days once every 4 weeks.¹⁴  Despite gradual attrition in the azacitidine arm, therapy was well tolerated with few transfusion needs or adverse events. Importantly, disease-free survival was improved for those receiving azacitidine at the end of the 12 months (64% vs 42%; P = .04). There was no survival benefit in those receiving azacitidine: however, the trial was not powered to assess this. A more recent phase 3 trial, QUAZAR AML-001, is evaluating oral azacitidine (CC-486) vs placebo as maintenance therapy in patients age 55 years or older in first remission after intensive chemotherapy whether they received consolidation chemotherapy or not.¹⁵  Early results have reported a survival benefit of 24.7 vs 14.8 months in those randomly assigned to azacitidine, regardless of whether they had received consolidation chemotherapy. Furthermore, relapse-free survival was also prolonged in these patients (10.2 to 4.8 months). Results of this study are pending; however, if they are confirmed, this would be the first agent to improve overall survival and relapse-free survival when given as therapy after remission in AML.

Recognizing that there are knowledge gaps, you take a step back and discuss with patient 1 how she has approached decision making during the course of her treatment (Table 4) and find out that she has made decisions on her own and that she prefers to think about how her choices impact her family. You also ask her how she prefers to receive information (ie, in small increments vs all at once) and whether she would like to have your help to guide her decisions, given that she has made all of her other decisions on her own. During this conversation you learn that she plays a significant role in her granddaughter’s life because her own daughter is a single mother and that having to travel long distances for care will negatively impact her quality of life. This is not something that she is willing to compromise on. You then review the rationale for postremission therapy, emphasizing that there are several options for treatment, but you do so in the framework that she has set for you. She mentioned that she prefers to hear about “the long-term plan” and not just postremission therapy, so you discuss observation after treatment, detection of relapse, and salvage therapy. She asks for your help in deciding and you suggest proceeding with azacitidine on the basis of what you have understood her preferences to be. After considering her options she agrees, stating that this is the only choice that will allow her to receive care locally and hence spend more time with her family. She also agrees to telemedicine visits every 3 months so that you remain involved in her care and are able to monitor her progress.

Patient 2 is a 73-year-old man with a history of diabetes mellitus and chronic kidney disease after he received a kidney transplant from a living relative 7 years before his diagnosis of AML. He was found to have new onset pancytopenia during routine follow-up with his nephrologist. A bone marrow biopsy confirmed cytogenetically normal AML with 3 separate CEBPA mutations. Although he is active and otherwise in good health, his baseline creatinine ranges from 2.7 to 3.5 mg/dL. Induction chemotherapy was initiated with decitabine 20 mg/m² per day for 10 days and he achieved a CR with incomplete neutrophil recovery after 2 cycles. He then started maintenance therapy with decitabine 20 mg/m² per day for 5 days once every 28 days. He experienced neutrophil recovery and has continued with treatment for the past 25 months uneventfully without hospitalization or delays in treatment. Today he presents for cycle 26 and would like to discuss whether to continue with maintenance therapy or to stop treatment.

Treatment with hypomethylating agents (HMAs), alone or in combination with small molecular inhibitors such as venetoclax, has become a new standard of care in AML patients age 60 years or older who are deemed unfit for intensive chemotherapy. Generally, patients begin therapy and receive between 2 and 4 cycles of HMA-based treatment in the hopes of achieving a remission or hematologic improvement that would suggest benefit for ongoing therapy.¹²  For those patients for whom postremission therapy includes a bone marrow transplant, hypomethylating therapy is discontinued. For patients ineligible for transplantation, this becomes a chronic treatment until the time of relapse or the development of unacceptable toxicity. In a multicenter phase 1b study of patients age 65 years or older who are receiving venetoclax with either decitabine or azacitidine, the median duration of CR and CRi for those who received venetoclax 400 mg per day was 12.5 months.¹⁶  In the phase 2 trial of single-agent decitabine in patients age 60 years or older, the CR rate was 47%, with a median disease-free survival of 46 weeks.¹⁷  However, both studies include a range in duration of remission in which the upper limit was yet to be reached at the time of publication that may include patients such as the one mentioned here who has had a duration of response more than twice what was reported. In addition, many clinicians have “super responders” in their practice who have received chronic HMA therapy for several years.

There is little published data to guide cessation of HMA therapy in patients in a CR. Cabero et al¹⁸  retrospectively reported on 16 patients with AML or higher-risk MDS who were treated on HMA-based trials or who were in CR when treatment ended either per protocol or by personal request. Eleven patients (69%) relapsed with a median time to relapse of 4 months (range, 2-68 months). Although the sample size was small, it seemed as though patients without high-risk cytogenetics who had received more than 12 cycles of therapy had a longer time to disease relapse. Minimal residual disease (MRD) monitoring and its ability to identify molecular, cytogenetic, or flow cytometric recurrence of leukemia is being incorporated into clinical trials and to some extent clinical practice to identify the depth of the first remission, inform postremission treatment strategies, and identify impending relapse.¹⁹  Although published data about HMA “super responders” are limited, it is possible that evaluating MRD could be useful in cases such as this.

You discuss with patient 2 that data for guiding cessation of HMA therapy in patients who are in a CR are limited, and you mention and that he has 2 options for moving forward: continuing treatment without interruption or repeating a bone marrow biopsy to look for evidence of disease by morphology or MRD. If the bone marrow biopsy is negative, he could consider discontinuing treatment in favor of active surveillance alone. Because you have been observing patient 2 for more than 2 years, you know that he has made all of his treatment decisions with input from his sons, enjoys being active in his church, and values his independence, but has said that he does not like having to come to the clinic every month for treatment. You ask him a few more questions and learn that what he values most is being independent and not having to rely on his wife for help because she has developed more health issues over the past year. You discuss that given his kidney disease, he is not a candidate for a bone marrow transplant, may not tolerate cytotoxic chemotherapy well, and is unlikely to be eligible for any clinical trials. For these reasons, his chances of achieving a second CR are more limited, and he may develop more treatment related side effects than he has had thus far. A decision that would offer him the best chance at maintaining his quality of life would be continuing with treatment. He agrees with this decision and the plan for therapy.

Patient 3 is 52-year-old woman with a history of stage IIB breast cancer that was diagnosed and treated with chemotherapy 8 years before her diagnosis of AML. She was found to have new onset anemia after presenting for evaluation of shortness of breath. A bone marrow biopsy confirmed a therapy-related myeloid neoplasm, AML with MDS-related changes, and a complex karyotype with TP53, RUNX1, DNMT3A mutations. She received induction chemotherapy with liposomal daunorubicin-cytarabine. Her bone marrow biopsy on day 14 was consistent with persistent disease and she received a second induction with liposomal daunorubicin-cytarabine. She experienced complications in the hospital including a stay in the intensive care unit for neutropenic sepsis. Her counts slowly recovered and on day 55 of induction, she was discharged from the hospital. A bone marrow biopsy at the end of treatment confirmed both a morphologic and cytogenetic remission with disappearance of TP53 and RUNX1 mutations. She has been recovering at home and presents to the clinic with her partner to discuss further recommendations for treatment.

AML that arises after exposure to chemotherapy or radiation represents less than 10% of all AMLs but is more likely to be associated with high-risk features that portend a more aggressive clinical course with poor outcomes compared with de novo AML.²⁰  CPX-351 is a liposomal encapsulation of cytarabine and daunorubicin in a fixed 5:1 synergistic molar ratio that is now approved for patients with secondary AML or AML with MDS-related changes. In the randomized phase 3 study comparing CPX-351 to 7+3, patients enrolled on the experimental arm achieved higher remission rates (47.7% vs 33.3%; P = .016), and for those who underwent allo-SCT, there seemed to be a more favorable impact on survival.²¹  Although only 19.6% of patients enrolled had therapy-related AML, CPX-351 is used in this patient population, but for those who are eligible and willing, postremission consolidation with an allo-SCT is recommended.²¹ 

As you assess patient 3’s performance status and recovery from induction, you learn that she has 3 children aged 12, 14, and 16 years. She also shares with you that this treatment had many more serious adverse effects than her breast cancer treatment and that her wife has been overwhelmed to the point that she is not able to contribute to discussions about additional treatment like she normally does. Patient 3 also expresses a lot of fear about additional treatment and an allo-SCT. Despite this, she is clear that her goal is to live as long as possible to be able to parent her young children and that she is willing to do anything that will help her to achieve that. You discuss the adverse risk of therapy-related AML and treatment with CPX-351 consolidation and allo-SCT within this framework, including the risk for toxicity and that SCT does not completely remove the possibility of relapse, particularly for TP53-mutated AML. She expressed understanding of these risks and decides that she wants to proceed with SCT when possible.

The principles of shared decision making are based on the belief that a patient’s needs and desired outcomes should form the basis for all decisions. When using SDM, there is intentional engagement between the patient and his or her health care team as they navigate a diagnosis together.²²  SDM is not providing patients with a list of all possible treatments and asking them to decide. Patients want their health care providers involved in this process, and SDM allows us to integrate their values and preferences as we make decisions together. With the addition of new therapeutic options for patients with AML, consistent incorporation of SDM into our conversations may be needed now more than ever. To increase our understanding and confirm the impact of SDM on measureable outcomes in AML patients, additional studies of health care providers who use SDM techniques are needed, perhaps even in a randomized controlled study that includes usual practice as a comparator. One opportunity to consider might be the informed consent process for a clinical trial and a patient’s understanding of adverse events, expectations of treatment outcomes, and the decision to participate.²³  Regardless, current data suggest that when SDM is used, patients are more likely to feel informed about their disease and satisfied with their decision.⁵  Furthermore, health care teams are also more likely to feel as though they have done everything they can to honor a patient’s wishes, which at the end of the day is one of the most important things that we can do.