Acute promyelocytic leukemia: preventing early complications and late toxicities

Acute promyelocytic leukemia (APL) is a subtype of acute myeloid leukemia (AML), characterized by a balanced reciprocal translocation, resulting in the fusion of the PML gene on chromosome 15 with the RARA gene on chromosome 17. Patients with APL frequently present with a distinct hemorrhagic syndrome caused by hyperfibrinolysis, disseminated intravascular coagulation (DIC), and thrombocytopenia. Historically, APL was among the deadliest subtypes of AML, but through therapeutic advances and with a thorough understanding of the genetics driving this disease, APL now represents the most curable form of AML.¹  Steady improvements beginning with the incorporation of anthracyclines in the 1970s, followed by the introduction of all-trans retinoic acid (ATRA) in the 1980s, to the frontline use of ATRA in combination with chemotherapy, have pushed the complete response (CR) rate in excess of 90% for most patients, with a durable cure in >80% of patients with APL.^2-6  More recently, excellent outcomes have been seen using an approach with no or minimal chemotherapy with an ATRA and arsenic trioxide (ATO)-based induction.^7-10 

Although APL has seen incredible success over the last decades, there are still issues remaining in this disease that must be addressed. The initial diagnosis of APL must continue to be managed as a medical emergency, because early mortality from hemorrhage, differentiation syndrome, and infection persist at a significant rate.^11,12  Further, clinicians must recognize that, although using ATRA and chemotherapy has successfully reduced the relapse rate in APL and improved long-term survival, it comes with a cost of short- and long-term toxicities. These toxicities can be fatal, and this is particularly true among older patients or patients with significant comorbidities.^13-17  This review will focus on early management considerations to minimize early death and relevant considerations to prevent long-term complications in patients with APL. Specifically, we will address the necessary and important steps that need to be taken to reduce early hemorrhagic death and differentiation syndrome. In the context of the remarkable results seen using up-front ATRA and ATO over the past 5 years, we will then focus on the potential impact this will have on older adults with APL. Finally, we will discuss the immediate and long-term toxicities from APL-directed therapy, and discuss surveillance considerations.

Early death (ED) in APL is defined as death within the first 30 days of diagnosis. As a result of selection bias, clinical trial data have underestimated the impact of ED, but a series of epidemiologic studies in the mid-2000s revealed that a significant proportion of patients continue to suffer early death.^18-20  However, it is encouraging that newer epidemiologic studies now suggest ED rates may be improving.^21-24  Recently, results reported from 17 French centers from 2006 through 2011 revealed that 51% of patients with suspected APL received ATRA on the day of admission, subsequently resulting in an ED rate of only 9.6%.²⁴  Another recent study, reporting outcomes from the California Cancer Registry, reported a significant decrease in 30-day mortality, from 26% between the years 1988 and 1995 to 14% between the years 2004 and 2011.²¹  Catastrophic hemorrhage, differentiation syndrome, and infection remain the main causes of early death.^21-24  Although these studies show that early deaths persist, they also indicate that current early measures are effective in reducing the ED rate. In the next few sections, we will review best practices to minimize early death.

Fatal early hemorrhagic complications account for the largest proportion of early deaths in patients with APL. When fatal bleeding occurs, it is most commonly the result of intracranial hemorrhage. Patients at greatest risk for fatal bleeding are those presenting with high-risk disease, defined by an initial white blood cell (WBC) count >10 × 10⁹/L, and patients older than 60 years of age.^11,12 

Early suspicion and recognition of APL, and prompt ATRA administration, are critical steps to mitigate early death in APL. Nearly 90% of patients with APL demonstrate some evidence of bleeding at presentation.¹¹  Accordingly, clinicians should inquire about bruising, gum bleeding, epistaxis, and gastrointestinal bleeding. Further, clinicians should ask about headaches, visual symptoms, or neurologic symptoms given risk of intracranial hemorrhage with APL. Careful mucocutaneous examination for ecchymosis, petechiae, or excessive bruising at catheter insertion or venipuncture sites should follow. Laboratory studies should be ordered to assess hemostatic parameters, with careful review for thrombocytopenia, abnormal coagulation studies including a prolonged prothrombin time (PT) and/or partial thromboplastin time (PTT), a decreased fibrinogen, and an elevated d-dimer. Peripheral smear review should promptly be performed to identify circulating promyelocytes. The morphology of the abnormal promyelocytes in APL can present as 1 of 2 distinct subgroups. The more common subgroup is a hypergranular form, characterized by abundant azurophilic granules and potentially with multiple Auer rods. However, there is also a microgranular form, without apparent granules. Both subtypes are characterized by a bilobed or reniform nuclear contour, and it is of critical importance to recognize this, especially in the microgranular variant. Table 1 summarizes these findings.

If APL is suspected after initial assessment, ATRA should be initiated immediately at 45 mg/m² per day in 2 divided doses. Initiation of ATRA should not wait for confirmation of the disease, but instead should be started at first suspicion of APL. Genetic testing (cytogenetics, FISH for t(15;17) or polymerase chain reaction for PML/RARA) should be ordered and expedited to confirm the diagnosis. Delay in ATRA administration, especially for high-risk patients, can be costly. In a retrospective study assessing time to ATRA administration, we found that in high-risk patients for whom ATRA administration was delayed over 2 days, ED occurred in 80% (4/5) of patients.²⁵  Three of these deaths were attributed to hemorrhage (75%).²⁵  In comparison, among high-risk patients with APL who received ATRA within 2 days, the ED rate was lower at 20% (8/40 patients died), and only 3 of the 8 (37.5%) patients died as a result of hemorrhage.²⁵  Although an increase in early death rate was in this data set was only able to be shown in high-risk patients in whom ATRA was delayed by 2 days; this may be secondary to small patient numbers and we recommend initiation of ATRA at the very first suspicion of APL.

In addition to administration of ATRA, thrombocytopenia, DIC, and hypofibrinogenemia must also be monitored and promptly corrected. Meticulous attention to hemostatic support is as critical as initiation of ATRA, especially because correction of coagulopathy after ATRA initiation may take several days. Key measures for supportive care are obtaining blood work every 6 hours and aggressively transfusing blood products such that platelets are maintained at 50 × 10⁹/L or higher, fibrinogen is kept >150 mg/dL, and PT and PTT are sustained near normal levels.^26,27  Of note, as a result of coagulopathy, multiple platelet or cryoprecipitate transfusions may be required to meet these recommended parameters.¹¹  Finally, any invasive procedures, including the insertion of central access catheters, should be delayed until coagulopathy has resolved.²⁷ 

Beyond these measures, the incorporation of ATO into induction therapy, at least in theory, may have the potential to reduce hemorrhagic complications. Arsenic trioxide promotes promyelocytic differentiation and apoptosis and in combination with ATRA may reverse coagulopathy even more promptly than ATRA alone.²⁸  Although clinical data are limited, some information exists to allow one to hypothesize about the concurrent benefit of ATRA and ATO. The APL 0406 trial was a phase 3 trial, which randomized adults with standard-risk APL to up-front ATRA plus ATO compared with ATRA plus chemotherapy, reported no hemorrhagic deaths in patients receiving ATRA plus ATO.⁹  Of note, there were no hemorrhagic deaths reported in the ATRA/chemotherapy arm. The AML17 trial included high-risk and standard-risk patients and in this phase 3 trial, patients were randomized to receive ATRA plus ATO (and gemtuzumab ozogamicin in high-risk patients) or ATRA plus chemotherapy. This trial reported 3 deaths resulting from hemorrhage in the chemotherapy arm, but none in the ATO arm.⁷  Further data are required to determine whether this trend is significant, but prompt ATO along with ATRA may have the potential to minimize hemorrhage in patients with APL, particularly in those with high-risk disease who are at highest risk for early hemorrhagic death.

Differentiation syndrome (DS) is a complication during induction caused by the effects of differentiating agents on leukemic blasts.²⁹  Hyperleukocytosis frequently but not always accompanies DS and often precedes the clinical manifestations of DS. DS leads to a systemic inflammatory response syndrome–like syndrome. The most common problem seen with DS is acute respiratory distress caused by diffuse interstitial pulmonary infiltrates, which appear as a pleural effusion and pulmonary infiltration on chest imaging.²⁹  Other features that may indicate DS include fevers, weight gain, pericardial effusions, acute renal insufficiency, hyperbilirubinemia, or peripheral edema. Severe DS can be fatal, and patients with a WBC count >5 × 10⁹/L at diagnosis are at an increased risk for early mortality.²⁹ 

Given the risk of fatal DS, up-front steroid prophylaxis should be considered. The evidence for the use of steroids as a prophylactic approach to prevent DS, however, is limited. Prophylactic dexamethasone was used in high-risk patients enrolled in the APL2000 trial, which randomized patients to receive either ATRA, daunorubicin (DNR), and ara-C or ATRA and DNR alone during induction.²⁹  Compared with the earlier APL93 trial, which randomized patients to receive either ATRA followed by DNR and ara-C or ATRA and early DNR and ara-C during induction, and did not use prophylactic dexamethasone, the DS-related death rate decreased from 5.7% to 3.9%.⁵  The LPA99 trial, using ATRA and idarubicin during induction, broadened the use of prophylactic prednisone to all patients during the first 15 days of induction therapy.³⁰  Compared with the earlier LPA96 trial that used ATRA and idarubicin during induction, and selectively provided steroid prophylaxis for patients with a WBC count >5 × 10⁹/L, there was a trend toward a further reduction in severe DS (16.6% vs 11.3%).³⁰  Overall, in the setting of treatment with ATRA and chemotherapy, there may be a benefit using prophylactic steroids. Further, with the recent incorporation of ATO and ATRA for frontline therapy, there is concern that the use of 2 differentiating agents may result in an even greater risk of differentiation syndrome.

Therefore, both the APL 0406 trial and the APML4 trial, which was a phase 2 single-arm study using ATRA plus ATO plus idarubicin, used prophylactic steroids. In both trials, there were no reported deaths from DS. Given these findings, weighed against the minimal complications anticipated with short-term steroid use, we advocate for the use of prophylactic steroids in and ATO and ATRA–based induction approach.

Further, in patients presenting with leukocytosis, or patients with progressive leukocytosis, additional agents to prevent DS may be required. APL cells are exquisitely sensitive to anthracyclines, and anthracycline use can allow for rapid APL clearance. This is of particular importance in high-risk patients. The APML4 protocol used up-front idarubicin, in part to prevent hyperleukocytosis and DS.⁸  In this trial, no deaths from DS were reported, including among high-risk patients.⁸  Further, gemtuzumab ozogamicin has been used in clinical trials in high-risk patients to prevent DS and appeared to be effective, but is not currently accessible in the United States for use outside clinical trials.⁷  Leukopheresis typically has no up-front role in high-risk patients with leukocytosis. Furthermore, hydrea is generally not used in adults with high-risk disease. However, in patients with a rising WBC count after treatment initiation, hydroxyurea has a role. Hydroxyurea was used in the APL0406 trial to reduce the peripheral WBC count if hyperleukocytosis occurred during treatment, and appeared to be effective.⁹  No deaths occurred from DS. In general, we recommend adjusting induction therapy based on WBC count, with an anthracycline planned up front for patients with high-risk disease. Hyperleukocytosis that results during the treatment of standard-risk APL should be managed with hydroxyurea, reserving anthracycline use for resistant cases. Table 2 summarizes these recommendations and suggested supportive measures to prevent hemorrhagic complications and differentiation syndrome.

The use of ATO and ATRA in the up-front setting has changed the landscape of APL therapy. Recently updated data indicate that up-front therapy with ATO and ATRA may result in both fewer relapses and deaths. With a median follow-up of 53 months, patients enrolled in the APL0406 trial and randomized to ATRA plus ATO had a significantly higher event-free survival (EFS) compared with patients treated with conventional ATRA plus chemotherapy (EFS: 96% vs 81%, P = .003).³¹  Comparable results were seen in patients treated on the APML4 trial using ATRA plus idarubicin and ATO up front, followed by ATRA and ATO in consolidation. With a median follow-up of 4.2 years, patients treated in the APML4 trial had an EFS of 90%, compared with an EFS of 72% in patients previously treated on the APML3 trial, which used ATRA and chemotherapy.³²  Equally as remarkable is that these potentially superior outcomes have occurred with potentially less immediate toxicity, as outlined in Table 3.^32,33 

An unmet necessity in the treatment of APL is the management of disease in older adults. Chemotherapy-based regimens are especially toxic for older patients, with particular concern for death in remission. Older patients appear to see a CR rate of 80% to 90%, comparable with younger adults.^15-17  Overall survival (OS), however, is significantly worse than that in younger adults, with rates between 41% and 65%.^16,17 

Several factors contribute to lower OS in older adults. One is a higher rate of ED with advanced age. Several studies have reported a higher ED rate in older adults ranging between 24% and 29%, and in a Swedish population–based study, a very high ED rate of 50% was reported in patients older than 60 years.^18-20  In addition to death from hemorrhage, infection and multi-organ failure also significantly affected older adults. In fact, the German AML cooperative group studied outcomes in older patients and found that 70% of early deaths were attributed to infection and multi-organ failure, whereas hemorrhage accounted for 30% of early deaths.¹⁷  In comparison, the same German group reported that only 9% of early deaths in younger adults occurred as a result of severe sepsis. In addition, older adults appear to have a higher death rate in CR. Various studies have reported a significantly higher mortality rate at 18% to 24% among older adults in CR compared with younger patients.^15-17  Most of these deaths are attributed to either complications during consolidation chemotherapy or later toxicities from chemotherapy including secondary malignancies.

Given the increased mortality seen as a result of chemotherapy, no group of patients stands to benefit from chemotherapy, sparing induction and consolidation, more than older adults. The AML17 study and APML0406 study both provide evidence for the safety of up-front ATRA plus ATO.^7,9  The results from the UK-led AML17 trial demonstrated that patients who randomly received ATRA plus ATO required significantly less supportive care.⁷  This study demonstrated that compared with conventional therapy with ATRA and chemotherapy, patients treated with ATRA plus ATO required almost 4 fewer units of both blood and platelets, almost 10 fewer days of antibiotics, and were hospitalized an average of 5 fewer days.⁷  A reduction in any of these measures significantly favors frail patients. Further, chemotherapy-induced neutropenia is thought to be a major contributor to infectious deaths. In the APL0406 trial, treatment with ATRA plus ATO compared with ATRA plus chemotherapy resulted in a significantly lower percentage of patients with neutropenia (62% vs 35%, P = .0001).³²  This reduction in neutropenia appeared to translate into a lower rate of infectious deaths, particularly in older adults. The APL0406 trial enrolled 35 patients between 60 and 71 years of age.³⁰  Nineteen patients were randomized to receive AIDA, and 16 were randomized to ATRA plus ATO. Among older patients who received chemotherapy, there were 3 deaths as a result of infection.³¹  There were no reported deaths from infection among the older patients that received ATRA and ATO.³²  The results from these trials indicate that an ATRA-plus-ATO approach has reduced toxicities and has the potential to improve survival, particularly in older adults.

Over the last several years, cooperative groups have published very-long-term outcomes of patients with APL after treatment with ATRA and chemotherapy. Using ATRA and anthracycline for induction, consolidation, and low-dose chemotherapy maintenance in the LPA99 trial, the PETHEMA group reported a 5-year OS of 82%.³⁴  With a longer follow-up interval of 10 years, the European APL group reported a 10-year survival rate of 77% in patients treated in the APL93 trial, using ATRA and chemotherapy followed by maintenance therapy.³⁵  Although these outcomes reflect a significant achievement, these trials also observed a 3% to 5% death rate in CR.^34,35  Chemotherapy-related events including infection, heart failure, and secondary cancer were the major reasons for these deaths. A significant number of surviving patients with APL today were exposed to chemotherapy, and thus understanding the potential toxicities is essential to provide sufficient long-term care.

Anthracycline chemotherapy has been an integral part of APL therapy since the 1970s. Secondary malignancies and cardiotoxicity are both well-characterized long-term complications attributed to anthracycline use, and contribute to late nonrelapse deaths and morbidity in patients with APL. The greatest numbers of secondary cancers are therapy-related myeloid neoplasms (t-MNs). According to long-term follow-up data regarding patients enrolled in the PETHEMA group LPA96, LPA99, and LPA2005 protocols, the 6-year incidence of t-MN was 2.2%.³⁶  Consistent with a therapy effect, patients in whom t-MN developed commonly had abnormalities of chromosomes 5 and 7, or 11q23 rearrangements identified on cytogenetic analysis.³⁶  Beyond t-MNs, a higher than expected rate of solid malignancies has been reported in patients with APL treated with anthracyclines. One study, which followed a cohort of patients with APL treated with ATRA and idarubicin induction (consolidation and maintenance approach was not described) for a median of 136 months, reported that a secondary cancer developed in 17% of these patients, including breast cancer, prostate cancer, colon cancer, melanoma, and soft-tissue sarcoma.³⁷  These findings underscore the importance that routine blood work during follow-up visits post remission is important not only to detect recurrent APL but also to assess for t-MN. Furthermore, the medical team must stress the importance of age-appropriate cancer screening in APL survivors.

Cardiotoxicity is a well-known late complication of anthracycline use. Chronic cardiomyopathy is the most commonly described cardiac complication from anthracyclines, and is typically thought to progress from asymptomatic diastolic or systolic dysfunction ultimately to symptomatic disease.³⁸  Cardiomyopathy has been specifically described in patients with APL. In a study examining 34 patients at a median of 7 years after APL diagnosis, despite a normal ejection fraction, segmental wall motion abnormalities were detected in nearly one third of patients.³⁸  In addition, compared with healthy volunteers, diastolic dysfunction was significantly more common in patients with APL exposed to anthracyclines.³⁸  Because anthracycline-induced cardiomyopathy is generally considered to be irreversible, early detection is important. No formal guidelines in adults exist for post-therapy screening. The children’s oncology group has proposed electrocardiogram and echocardiogram monitoring after the completion of therapy, and every 2 years thereafter.³⁹  It is reasonable to consider similar monitoring, particularly in older patients or patients with a history of heart disease, because these patients may be more vulnerable to the occurrence of anthracycline-related cardiotoxicity.

The up-front use of ATO plus ATRA is anticipated to reduce the long-term morbidity currently seen as a result of anthracycline therapy. However, studies indicate that potential long-term complications exist with ATO. In one long-term follow-up study, among 217 newly diagnosed patients with APL treated with up-front ATRA plus ATO, with a median follow-up interval of 72 months, a higher rate of liver dysfunction and hepatic steatosis was seen compared with healthy controls.⁴⁰  Further, in a retrospective study of patients treated for APL with ATRA plus ATO at MD Anderson Cancer Center, after a minimum of 3 year follow-up, hypertension developed in 43% of patients, diabetes mellitus developed in 26%, and cardiac arrhythmia developed in 11%.⁴¹  These medical conditions were all more frequent compared with patients who received ATRA plus chemotherapy. Finally, several secondary malignancies were reported in patients who received ATRA plus ATO; 15 of 104 patients treated with ATRA and ATO developed a second cancer on average 4 years after diagnosis; this was not significantly different compared with the rate in patients treated with ATRA plus chemotherapy.⁴¹  Overall, these findings highlight the need for collection of long-term data to determine the long-term risk profile when using up-front and consolidative ATO. Based on this limited data, in patients treated with ATO, liver function testing should likely continue at periodic intervals after therapy has been completed. This should especially be continued in patients who had liver toxicity during therapy. Further, close primary care follow-up should be established to monitor for and manage hypertension and diabetes. Finally, age-appropriate cancer screening should be emphasized in all patients after completion of APL therapy. Table 4 lists the secondary malignancies observed in APL survivors, along with appropriate screening practices.

A significant number of patients with APL will be cured with current therapy, but early death and long-term toxicities remain major concerns. For early death, timely ATRA administration and aggressive supportive transfusions have begun to minimize the early hemorrhagic death rate and are critical in the initial treatment of patients with APL. However, especially in high-risk patients with APL, a need to minimize the duration of coagulopathy persists as hemorrhagic deaths continue to occur. One measure that has promise is the incorporation of ATO therapy up front for APL. As a potent differentiation agent in APL, up-front ATO in addition to ATRA may more quickly resolve the clinical signs of coagulopathy, and thus potentially reduce mortality. Further, preventive measures for differentiation syndrome should be incorporated into the early treatment plan. There is now evidence to suggest prophylactic steroids for DS reduces the incidence of severe DS; therefore, given the consequences of this complication, all patients starting ATRA and ATO therapy for APL should receive steroids during induction. Hyperleukocytosis commonly occurs, either at initial presentation or during therapy in patients treated with ATRA and ATO induction. Thus, in standard-risk APL, hydroxyurea should be initiated if the WBC count rises to >10 × 10⁹/L, and in high-risk disease, anthracycline chemotherapy during induction should be considered. Finally, with increasing evidence confirming the efficacy of therapy using ATO in the up-front setting and in consolidation, a chemotherapy-sparing approach with ATO should be planned for all standard-risk patients with APL. An approach that minimizes chemotherapy exposure is also generally an appropriate therapeutic consideration for adults with high-risk APL. Older patients with APL who have inferior survival outcomes compared with younger patients stand to significantly benefit from ATRA and ATO–based induction.

Strategies to identify potential long-term effects of APL therapy should be included in the care plan for patients with APL. In patients who received anthracycline chemotherapy during treatment, clinicians must recognize the risks of cardiotoxicity and secondary malignancies using this treatment. Noninvasive cardiac imaging including echocardiography should be considered after completion of therapy and at defined intervals, particularly in patients with risk factors for heart disease, to look for subclinical signs of heart failure, recognizing that the data to support this are limited. Also, clinicians should emphasize the importance of age-appropriate cancer screening including skin examinations, mammography, and colonoscopy, because these may occur at a higher rate in treated patients. Clinicians should maintain a high level of suspicion for the development of t-MNs if cytopenias or macrocytosis occurs. Extended data regarding the use of ATRA and ATO induction will be important to assess not only long-term efficacy but also late toxicities. Limited data suggest that liver toxicity may persist after treatment; therefore, we suggest that liver function monitoring be included as part of routine monitoring post therapy. Higher rates of hypertension, diabetes, and arrhythmias were also seen post ATO therapy at a single institution; therefore, primary care follow-up continues to be important for patients with APL. The outcomes for adults with APL are excellent, thus underscoring the importance of the prompt recognition of APL and administration of supportive care and the need for vigilant monitoring and management of toxicities during and after the completion of therapy.

Jessica K. Altman, Robert H. Lurie Comprehensive Cancer Center of Northwestern University, 303 E. Superior St, Lurie 3-119, Chicago, IL 60611; e-mail: j-altman@northwestern.edu.