Long-term efficacy and safety of arsenic trioxide for first-line treatment of elderly patients with newly diagnosed acute promyelocytic leukemia




The prognosis of acute promyelocytic leukemia (APL) in the elderly is poorer than that of younger patients after treatment with all-trans retinoic acid plus chemotherapy, which is the current standard therapy for APL. A significantly higher mortality during consolidation therapy was found, which is mainly due to deaths from sepsis following chemotherapy-induced myelosuppression.


A total of 33 patients aged 60 years or older with de novo APL were treated with single-agent arsenic trioxide (ATO) for remission induction and postremission therapy. The postremission therapy continued for up to 4 years.


Twenty-nine patients (87.9%) achieved a hematologic complete remission, and the most common adverse event during remission induction was leukocytosis (63.6%). Definite differentiation syndrome was observed in 5 patients. Nonhematologic adverse events were all manageable and reversible. Twenty-eight patients proceeded to postremission therapy. Adverse effects during postremission therapy were mild, transient, and no treatment was required. No patients died from ATO-related toxicities. With a median follow-up of 99 months, the 10-year cumulative incidence of relapse, overall survival, disease-free survival, and cause-specific survival were 10.3%, 69.3%, 64.8%, and 84.8%, respectively, which are comparable with those in the younger APL partners. No significant risks for development of chronic arsenicosis or second malignancy were observed during the follow-up period.


The results indicate that the single-agent ATO regimen is safe and effective with long-term durable remission, and could be used as first-line treatment for elderly patients with de novo APL. Cancer 2013. © 2012 American Cancer Society.

The combination of all-trans retinoic acid and anthracycline-based chemotherapy is currently the standard approach to newly diagnosed acute promyelocytic leukemia (APL) and has yielded encouraging clinical outcomes in adult patients. However, the treatment for elderly APL patients remains a challenge. APL is rarely seen in the elderly population aged 60 years or older. The proportion of elderly patients with APL is only 12% to 24% in most series.1-3 Compared with that in young adults, the prognosis of APL in the elderly is poorer after treatment by the current standard approach due to both a higher incidence of early deaths and a significantly higher mortality found during consolidation therapy, when about 15% of the patients died in hematologic complete remission (HCR), mainly from sepsis during the phase of myelosuppression induced by chemotherapy.4 Although a reduction of chemotherapy intensity has been proposed for patients in this age group to reduce treatment-related toxicities, the clinical implications remain unclear because of the potential increase of relapse risk.1, 5 In addition to higher mortality, elderly APL patients are often accompanied with a series of comorbidities, some of which represent an absolute contraindication to the current standard treatment.1, 2, 5 Effective treatment approaches that are particularly suitable for elderly APL patients are highly desired.

Arsenic trioxide (ATO), as a single agent, is most commonly used at present in cases of relapsed APL for induction of remission. It has been shown that it is equally effective in newly diagnosed cases of APL, with much less toxicity.6-8 Therefore, for the elderly APL patients, especially for those truly frail patients who are considered unfit for chemotherapy, ATO could be a reasonable alternative to the current standard approach.9 However, ATO is not commonly used as the first-line drug for remission induction in de novo APL cases, nor is it used for postremission therapy. The main concern is that long-term administration of ATO, which is best known as a poison, would render patients at high risks for developing chronic arsenicosis or a second malignancy.10 The results based on our 5-year follow-up study showed that a single-agent ATO regimen is highly efficacious and safe for the treatment of children with newly diagnosed APL.11 This report deals with the 10-year follow-up results of a single-agent ATO regimen for treatment of elderly patients with de novo APL.



The study was open to elderly patients diagnosed with de novo APL in the First Affiliated Hospital of Harbin Medical University, Harbin, China, from March 1996 to December 2002. During this period, a total of 38 consecutive patients 60 years or older were diagnosed with APL. Of these patients, 5 were excluded due to refusal of treatment. Thus, 33 patients were enrolled in this study, regardless of performance status or comorbid disease. The diagnosis of APL was based on clinical presentation, the characteristic French-American-British M3 bone marrow (BM) morphology, and was subsequently confirmed by cytogenetic assay for t(15;17) translocation. In 10 patients whose diagnosis was only made on a morphological basis without cytogenetic confirmation, review of initial marrow slides by an expert morphologist was mandatory. This study was started in 1996, when reverse transcription polymerase chain reaction was not widely available in China, and thus polymerase chain reaction information was not used for the diagnosis. Written informed consent was required, and the study protocol was reviewed and approved by the Provincial Medical Ethics Committee.

Treatment Protocol

For remission induction, ATO (0.16 mg/kg/day, for a maximum of 10 mg/day; Harbin Yida Pharmaceutical Company, Harbin, China) was prepared in 5% glucose or 0.9% sodium chloride (for patients with diabetes) solution. The total daily dose was infused intravenously over 3 to 4 hours/day for 28 days. After an intermission of 5 to 7 days, the drug was administered again until the presence of all of the following: no clinical evidence of APL, sustained platelet count exceeding 100 × 109/L for at least 5 to 7 days, no blasts or promyelocytes in blood, and less than 5% blasts plus promyelocytes in the BM, or to a maximum of 56 doses.

In patients with white blood cell (WBC) counts greater than 20×109/L at admission, adjusted-dose DA (daunorubicin at 40 mg for 3 days; Ara-c [cytosine arabinoside] at 50-100 mg for 5 days) was administered and a semi-dose of ATO was given simultaneously. For patients with leukocytosis (WBC count greater than 10 × 109/L) after ATO treatment was initiated, the ATO dose was reduced and hydroxyurea (1.0-2.0 g/day) was taken orally if the WBC counts exceeded 20 × 109/L; ATO was discontinued temporarily and hydroxyurea (3.0 g/day) was administered when the WBC count was higher than 40 × 109/L. For patients with cardiac, liver, or renal diseases at admission, the primary disease was treated aggressively. Depending on magnitude of organ dysfunction, full-dose or semi-dose of ATO was administered continuously or intermittently, and electrocardiograms (ECGs), liver, and renal functions were monitored closely. ATO was discontinued or reduced if a grade 3 to 4 toxic event or differentiation syndrome occurred during the induction phase. After resolution of the above events or recovery to baseline status, ATO was then increased to its full dose.

For patients with a leukocyte count greater than 20 × 109/L, oral or intravenous (according to cardiac and renal function) NaHCO3 and oral allopurinol were commonly administered, and no extra hydration was required. For patients with severe edema, frusemide was given and albumin was used to correct hypoalbuminemia. During the ATO therapy, serum potassium and magnesium concentrations were kept above 4 mEq/dL and 2 mg/dL, respectively.

Patients who achieved an HCR could receive up to 4 years of postremission therapy. The therapy was carried out on an outpatient basis. The schedule of ATO administration for postremission therapy is detailed in Table 1. For every ATO course during postremission therapy, a complete blood count was conducted before the administration of ATO and was repeated once per week thereafter. Postremission therapy would be intensified in some cases from our experience. In detail, the duration of this course would be extended to 14 days whenever blasts plus promyelocytes in BM increased to 4% to 5% or if before initiation of a course of ATO treatment, the WBC count was on the high side of the normal range (8-10 × 109/L) with a normal differential count. If the WBC count on the high side did not decrease markedly after 14 days of ATO administration, the duration of this course would be further prolonged to 21 days.

Table 1. Schedule of Arsenic Trioxide Administration for Postremission Therapy
TimeIntermission (wk)Course Duration (wk)FrequencyCumulative Time of Arsenic Trioxide Administration (wk)
First year21×211-13
 6 ×1 
Second year 1×17-8
 8 ×1 
Third year 2-3×110-12
Fourth year71×17-8

Safety Evaluation

During induction therapy, complete blood cell counts and liver and renal functions were closely monitored. ECGs were required at least twice per week. Throughout postremission therapy and for at least 3 years after completion of all therapy, complete physical examinations were performed every 3 months for liver function, renal function, electrolytes, ECGs, dermatological consultations, and detailed neurological examinations. BM morphology analysis was performed every 6 months until HCR had been sustained for 5 years.

To monitor the body retention of arsenic due to long-term administration of ATO, arsenic levels in urine, nails, and hair, which are good indicators of long-term exposure and in vivo accumulation of arsenic, were measured as described11 in 17 long-term survivors at the last follow-up visit and in 17 healthy controls matched for age, sex, and geographic location, who had never received therapeutic ATO.

Outcome Definitions

Early death was defined as death within the first week of treatment. HCR and hematologic relapse were defined as reported.11 Because only few molecular data were available after 2004, molecular remission and molecular relapse were not considered in the study.

Overall survival (OS) was calculated as the time from initiating treatment to death from any cause; cause-specific survival (CSS) was calculated as the time from the first day of the therapy to death attributed to APL disease or ATO treatment; disease-free survival (DFS) was calculated from time of achieving HCR to an event (relapse, other malignancy, or death from any cause). In the absence of these events, patients were censored at the last visit.

Statistical Considerations

This is a geriatric study with long-term follow-up where mortality unrelated to the study itself was substantial. Historically, survival following APL treatment in elderly patients has been reported, using all-cause survival analysis. CSS analysis, which focuses on the impact of a disease process and a therapeutic modality on survival and had been widely used in other long-term follow-up studies of elderly patients, was used here to reduce the impact of the unrelated deaths on survival rates.12

Univariate associations between dichotomous variables were evaluated with the Fisher exact test. The Student's t test was used to compare arsenic levels between the patient group and the healthy control group. The follow-up of the patients was initiated from the date of HCR and updated on December 2009. The survival analysis was performed with the Kaplan-Meier method using SPSS software, version 15.0. The survival estimates are reported ±1 standard error. The probability of relapse was estimated by the competing risk approach.13 All patients are included in intent-to-treat analysis, regardless of not receiving allocated postremission therapy on protocol.


Initial Patient Characteristics

The baseline characteristics of the 33 elderly patients are shown in Table 2. Of the 29 patients who had at least 1 comorbidity, 24 required specific treatment when admitted to the hospital. Five patients had chronic hepatitis B infection and 3 of them developed hepatic cirrhosis. Five patients had suffered from psoriasis and received bimolane treatment for 8 to 43 years (median, 22 years) before leukemia onset.

Table 2. Patient Demographics and Treatment Results
CharacteristicMedian (Range), No. of PatientsHCRP
  1. Abbreviations: APL, acute promyelocytic leukemia; DIC, disseminated intravascular coagulation; ECG, electrocardiogram; ECOG, Eastern Cooperative Oncology Group; FAB, French-American-British; HCR, hematologic complete remission; WBC, white blood cell.

Age, y65 (60-79)   
Body weight, kg69 (46-90)   
WBC count, ×109/L1.9 (0.49-33)   
Platelet count, ×109/L29×109/L (9-120)   
Bleeding or/and laboratory evidence of DIC    
ECOG performance status    
FAB subtype    
Relapse risk group14    
Tumor origin    
 Primary APL282485.70.60
 Therapy (bimolane)-related APL55100 
Mild or moderate increases in liver enzyme levels11981.80.59
Abnormal baseline ECG findings181688.91
 Coronary artery disease6466.70.14
 Chronic cor pulmonale221001
 Diabetes mellitus type 254800.50
 Hepatic cirrhosis3266.70.33

Response to Induction Therapy

Of the 33 patients, 29 (87.9%) achieved HCR. There were 3 early deaths at day 4, 6 and 7, respectively during the induction therapy resulted from cerebral hemorrhage. The other patient died on day 19 due to secondary uncontrolled sepsis.

The median time to BM remission (ie, elimination of all visible leukemic cells on BM aspirate review) was 32 days (range, 22-48 days). The median time to HCR was 57 days (range, 39-70 days) with a median dosage of 400 mg (range, 310-520 mg). The median times for recovery of the platelet count (100×109/L or higher) and absolute neutrophil count (1.5×109/L or higher) were 36 days (range, 0-49 days) and 55 days (range, 39-66 days), respectively. When HCR was achieved, the hemoglobin level, which was neither an index for withdrawing induction therapy nor that for HCR in APL, was 64-132 g/L (median, 89 g/L), and it generally returned to normal in 1 to 2 weeks.

Of the 29 patients who achieved HCR, all but 3 had a characteristic leukocytic response to ATO induction. First, the WBC counts increased gradually by a median of 8.6 times (range, 1.6-34.2 times) after initial treatment of ATO. The median peak WBC level was 19.4×109/L (range from 3 to 139.5×109/L), and reached at a median of 13 days (range, 3-21 days) following ATO administration and then the elevated WBC counts declined gradually to the lowest level ranged from 0.2×109/L to 3.4×109/L (median, 1.1×109/L) within median of 21 days (range, 11-33 days). Eventually, the low WBC counts went up again until back to normal. The WBC counts for the rest 3 patients had no significant change during induction therapy.

Univariate analysis demonstrated a significant relationship between baseline WBC count, Eastern Cooperative Oncology Group performance status and response to induction therapy (Table 2).

Adverse Events During Induction Treatment

Adverse events during ATO induction therapy are summarized in Table 3. The most common adverse events were leukocytosis. Of the 33 patients, 21 (63.6%) developed leukocytosis in response to the ATO treatment apart from 4 cases (12.1%) who had a high WBC count at presentation. Low-dose DA was administered in 1 patient and hydroxyurea was applied in 11 patients. Definite differentiation syndrome, which is often a severe or life-threatening complication, occurred in 5 patients with fever, edema, dyspnea, orthopnea, bilateral pleural effusion, and/or hydropericardium, and was evidently associated with severe leukocytosis (range of the peak WBC level from 28.6×109/L to 139.5×109/L). These patients were all treated successfully by controlling progressive leukocytosis, temporary discontinuation of ATO and administration of dexamethasone for 3 to 5 days.

Table 3. Adverse Events During Arsenic Trioxide (ATO) Induction (N = 33)
Adverse EventsPatients, n (%)
All EventsEvents Probably Related to ATOGrade 3/4 Eventsa Probably Related to ATO
  • a

    Grade 3/4 indicates grade 3 or 4 according to the National Cancer Institute Common Toxicity Criteria, version 2.0.

Any event33 (100)31 (93.9)6 (18.2)
Leukocytosis21 (63.6)21 (63.6)
Differentiation syndrome5 (15.2)5 (15.2)1 (3.0)
Electrocardiographic abnormalities20 (60.6)16 (48.5)2 (6.1)
 QTc prolongation (>0.48 s)12 (36.4)12 (36.4)2 (6.1)
 Ventricular premature contraction5 (15.2)5 (15.2)0 (0)
 Sinus tachycardia8 (24.2)4 (12.1)0 (0)
Gastrointestinal reactions25 (75.8)14 (42.4)1 (3.0)
Liver dysfunction7 (21.2)7 (21.2)0 (0)
Headache19 (57.6)8 (24.2)0 (0)
Edema16 (48.5)16 (48.5)2 (6.1)
Peripheral neuropathy11 (33.3)9 (27.3)0 (0)
Hyperglycemia13 (39.4)8 (24.2)0 (0)
Hyperpigmentation5 (15.2)5 (15.2)0 (0)

Grade 3 or 4 neutropenia was seen in 19 patients (57.6 %). With the exception of 1 patient who died of a sepsis, there were no severe or life-threatening infectious events. However, 1 patient without previous history of herpes zoster developed a herpes zoster infection during the induction therapy without relationship to dexamethasone administration.

Nonhematologic adverse events such as ECG abnormalities, gastrointestinal reactions, liver dysfunction, edema were all manageable and reversible. Other nonhematologic adverse events were all transient and mild, and no further management was required.

For patients with elevated levels of liver enzymes and abnormal ECGs at admission, management of primary hepatic disease and cardiac disease, symptomatic treatment such as protecting liver, reducing enzyme levels, anti-arrhythmia, and ATO induction were given simultaneously, after which liver enzyme levels or abnormal ECGs would usually return to normal or baseline when HCR was obtained. No treatment-related deaths were recorded.

It was found that for most comorbidities, such as hypertension, coronary artery disease, chronic hepatitis, and hepatic cirrhosis, as long as there was no organic function damage, administration of ATO seldom induced more toxicity, whereas in patients with ECG abnormalities, respiratory failure, heart failure, or abnormal liver or kidney function, adverse events of ATO therapy occurred more frequently. For examples, ECG changes were observed in 12 of 18 patients (66.7%) with abnormal baseline ECG which is significantly higher than 4 of 15 patients (26%) with normal baseline ECG (P = .037). Regarding patients with peptic ulcer or diabetes, ATO induction sometimes caused severe gastrointestinal reaction or abnormally high blood sugar levels.

In general, 14 patients had ATO decline or temporary discontinuation due to grade 3 or 4 adverse events (6 cases) or/and severe leukocytosis (12 cases). All of them were able to resume treatment and subsequently obtained a HCR.

Clinical Outcomes

Three patients underwent hematological relapse at 23, 28, and 31 months after achieving HCR. Of these, 2 patients had declined postremission therapy before relapse occurred due to personal request. One of the 2 patients who firmly refused postremission therapy after each HCR had 4 relapses (23, 33, 38, and 41 months, respectively) within 18 months. Following repeated ATO induction therapy, this patient achieved HCR 3 times, but eventually died of a cerebral hemorrhage in the fourth relapse. The other 2 patients both attained a second HCR after repeated ATO induction and have remained in the second HCR for 53 and 65 months, respectively. No ATO resistance has been found in this series. Eight patients died in HCR (see details in Table 4).

Table 4. Deaths in Hematologic Complete Remission
Patient No.Age, ySexTime of DeathCauses of Death
172MPostremission therapyLiver cirrhosis complication (bleeding in the upper digestive tract)
281FPostremission therapyChronic cor pulmonale with heart failure
366FOff-treatment for 2 moMyocardial infarction
474FOff-treatment for 8 moAccidental trauma
573MOff-treatment for 11 moHypertension complication (stroke)
671MOff-treatment for 69 moLiver cancer
780MOff-treatment for 71 moHypertension complication (heart attack)
883FOff-treatment for 107 moDiabetic complication

With a median follow-up of 99 months (range, 51-158 months) for all surviving patients, the 10-year cumulative incidence of relapse was 10.3% and the 10-year estimates of OS, DFS, and CSS were 69.3% ± 8.1%, 64.8% ± 10.2%, and 84.8% ± 6.3%, respectively (Fig. 1). In the CSS analysis, one ATO treatment–independent death from liver cancer was not considered as an event. The data of survival analysis indicated that in this series the 10-year mortality associated with APL or ATO treatment was only 15.2% while 15.5% deaths was APL or ATO treatment independent.

Figure 1.

Cumulative incidence curve and survival curves are shown, according to competing risks analysis for (A) cumulative incidence of relapse (CIR) (n = 29) and 10-year Kaplan-Meier product limit estimate of (B) overall survival (n = 33), (C) disease-free survival (n = 29), and (D) cause-specific survival (n = 33).

What should be noted here is the long-term outcomes of the 5 cases with secondary APL. Of these, 1 died in the fourth relapse, 1 was lost to follow-up after 109 months, and the other 3 were still alive at the last follow-up.

Tolerance of Postremission Therapy and Long-Term Toxicity

Of the 29 patients who achieved HCR, 28 were subjected to postremission therapy. The postremission therapy continued for up to 4 years in this study. Except for 2 patients who died during postremission therapy, the remaining 26 patients completed the programmed postremission therapy. Among these 26 patients, 21 cases had ever undergone intensified postremission therapy. The cumulative ATO doses in 24 cases without relapse were 2245-3380 mg (median, 3060 mg), whereas the doses in the 2 relapsed patients were 4670 mg and 5180 mg, respectively.

The periodic ATO infusion regimen was well tolerated for all 26 patients who completed the postremission therapy, including the 2 relapsed patients with higher cumulative ATO doses. Although headache, gastrointestinal reactions, and neutropenia occasionally occurred during the therapy, all these adverse effects were mild, transient, and needed no treatment.

The evaluation on the long-term toxicity demonstrates that until the final follow-up, neither obvious symptoms of chronic arsenicosis (such as liver and kidney damage, neuropathy, skin lesion, limb muscle atrophy, among others10) nor second malignancy had been observed, with the exception of 1 patient who had long-standing hepatitis B virus infection and hepatic cirrhosis, died of liver cancer 117 months after HCR, and the death was apparently independent of ATO treatment.

To monitor the toxicity of the ATO treatment, urine arsenic concentrations and arsenic contents in nails and hair were analyzed at the last follow-up visit in 17 long-term survivors, apart from 3 cases who were lost to follow-up. There had been no significant difference in arsenic levels in samples of nails, hair, and urine compared with the healthy control groups (Fig. 2).

Figure 2.

Arsenic levels are shown for (A) hair, (B) nails, and (C) urine samples of patients with acute promyelocytic leukemia. Arsenic levels in (A) hair, (B) nails, and (C) urine samples in group A (n = 15) and group B (n = 17) patients and the healthy controls (n = 17). Group A: All 17 long-term survivors were included regardless of the heterogeneity among patients. Group B: Only 15 long-term survivors, who had no hematological relapse after hematologic complete remission, with lower cumulative doses of arsenic trioxide (range, 2360-3150 mg; median, 2970 mg) and longer off-treatment duration (range, 36-110 months; median, 52 months) were included, whereas the other 2 patients, who had ever relapsed, with higher cumulative doses of arsenic trioxide (4670 mg and 5180 mg, respectively) and shorter off-treatment duration (5 and 17 months, respectively), were excluded. ▵ indicates the detection results of the 2 patients who had ever relapsed.


This is a long-term follow-up study of elderly APL patients who were treated with a single-agent ATO regimen. Due to the early initiation of the study, the molecular and cytogenetic data are insufficient. However, there is a characteristic leukocytic response to ATO induction including an early increase in WBC counts, a subsequent neutropenia, and then the recovery of WBC counts. Based on decades of our clinical observations, the treatment response can be found in more than 90% of patients with APL, whereas it never occurred in the cases with other diseases, suggesting that the characteristic leukocytic response to ATO induction is highly sensitive and specific and may serve as a further validation of the diagnosis of APL. In fact, this response was observed in all but 3 patients in this series.

In this study, ATO was administered in a low dose to patients with APL for a long period of time, in contrast to much shorter periods of treatment in all other single-agent studies.7, 8 It is in our hospital that ATO was initially attempted for clinical treatment of APL in the 1970s,15 and these treatment strategies are mainly derived from our previous clinical experiences: 1) We found that in adult patients, the toxic effects of ATO had much to do with its daily dose. A high daily dose of ATO for patients with excessive body weight often results in a notable increase in the incidence of severe ventricular arrhythmia and differentiation syndrome. Yet, as long as the maximum daily dose of ATO is limited to 10 mg, the increase can be avoided without affecting the therapeutic effect. 2) During postremission therapy, multiple short-course (7 days) regimens were used with the intermissions between courses prolonged gradually from 2 weeks to 6 weeks within the first year and sustained at 6 to 8 weeks during the next 3 years (Table 1). This treatment strategy can obviously reduce the occurrence of chronic arsenicosis such as numbness in extremities, skin pigmentation, and is especially suitable for senile patients who have a relatively slow drug metabolism rate. 3) The duration of postremission therapy was particularly long. Here, the duration of postremission therapy was mainly determined by the relapse risk. Originally, the duration of postremission therapy was much shorter. After a time, it was found that relapse seldom occurred during the treatment period but often happened at 1 to 3 years after treatment cessation.

In contrast to the tendency to develop all-trans retinoic acid resistance, patients could always obtain a second HCR after repeated ATO induction and seldom developed ATO resistance. Accordingly, it was inferred that the relapse was largely due to the too-short duration of postremission therapy, and thus the duration of postremission therapy was gradually lengthened. It has been found that the DFS curve appeared to reach a stable plateau after 4 years of postremission treatment in adult patients, and further ATO treatment after that time would have no apparent clinical benefit (data not shown); for this reason, the 4-year postremission therapy regimen formed. The postremission therapy was somewhat intensified during the third year, because it had been noticed that patients were often at high risks of relapse during that time.

APL is difficult to study in the elderly because of its rarity in this population. To our knowledge, only limited studies of APL in elderly patients have been published.1, 2, 5, 16, 17 Although these various studies may lack comparability due to differences in sample size, age distribution, severity of disease, medical conditions, and so forth, the main results from these studies were still summarized in Table 5 to make a rough and easy comparison. In 3 of the 6 series,1, 2, 5 selected patients fulfilling inclusion criteria were accrued, whereas unselected ones were included in the other 3 studies. Similar response rates of 84% to 92% were found in all but 1 of the 6 study series. In both the study in Spanish centers2 and ours, long-term outcomes were encouraged, which is better than those in the other 4 series. In addition, although reduced intensities of chemotherapy in older patients were used in 3 of the 5 series,1, 5, 16 the mortality during postremission therapy remains high, up to 10% to 18.6%, and the main cause of death was infection, whereas a high degree of compliance of postremission therapy was observed in our series, in which only 6.9% of patients died from noninfectious disease during postremission therapy. A similar cumulative incidence of relapse was also found between the Spanish study2 and the current study, which is comparable to that in the younger adult patients.1, 2 All these indicate a potential clinical significance of single-agent ATO treatment for elderly APL patients.

Table 5. Major Clinical Studies of APL in Elderly Patients
 Mandelli et al5Sanz et al2Ades et al1Disperati et al15Ferrara et al16Zhou et al (present study)
  • Abbreviations: 6-MP, 6-mercaptopurine; AMSA, Amsacrine; APL, acute promyelocytic leukemia; ATO, arsenic trioxide; ATRA, all-trans retinoic acid; CIR, cumulative incidence of relapse; CT, chemotherapy; DFS, disease-free survival; DNR, daunorubicin; GIMEMA, Gruppo Italiano Malattie Ematologiche Maligne dell'Adulto group; GO, gemtuzumab-izogamycyn; HCR, hematologic complete remission; Ida, idarubicin; MIT, mitoxantrone; MTX, methotrexate; NR, not reported; OS, overall survival; PETHEMA, Programa para el Tratamiento de Hemopatías Malignas group.

  • a

    AIDA, (IDA+Ara-C) ×1→ (MIT+ etoposide) ×1→ (IDA+Ara-C+6-thioquanine) ×1; aAIDA: (IDA+Ara-C) ×1.

GroupItalian GIMEMASpanish PETHEMAEuropeanCanadaItalianChinese
No. of patients134104129133433
Study periodJan 1993-Jun 2001Nov 1996-Dec 2003Apr 1993-Oct 19981999-2006Jan 2005-Dec 2007Mar 1996-Dec 2002
Median age, y65.8(60-75)68(60-83)66(62-70)78(71-87)70(61-84)65(60-79)
HCR, %86.68486926887.9
ConsolidationAIDA/aAIDAa(IDA→MIT→IDA)/ (reinforced single-agent CT+ATRA)(DNR+Ara-C)×1 or 2(ATRA+Ara-C +DNR/ AMSA) ×2aAIDAa / GOATO
MaintenanceAIDA: ATRA/ (6-MP+ MTX)/ (Alternating CT+ATRA)/no; aAIDA: ATRA6-MP +MTX+ATRAATRA/(6-MP+MTX)/ (ATRA+6-MP+ MTX)/noATRAATRA
Duration of maintenance2 y or no2 y2 y9 mo2 y4 y
OS, %6 y, 56NR4 y, 57.82 y, 76Median, 38 mo10 y, 69.3
DFS, %6 y, 596 y, 79NRNRNR10 y, 64.8
CIR, %NR6 y, 8.54 y, 15.6NRNR10 y, 10.3
Death during postremission, %

Adverse effects of ATO treatment had been a big concern. In this study, characterized by a low-dose method, almost all patients experienced 1 or more adverse events during remission induction; however, most side effects were manageable, reversible, and were usually less severe than those resulting from chemotherapy. No patients died from the toxicities caused by ATO treatment. Concerns have been raised about the potentially fatal cardiac and hepatic toxicity of ATO in a subset of patients with APL. However, most of these toxicities have been described in patients with relapsed APL who had received multiple courses of chemotherapy, which were often associated with cardiac and hepatic toxicities.18-20 In addition, the daily dose of ATO for these patients was generally calculated on the basis of body weight without maximum dosage limit, which was often too high for patients who had excessive body weight,18-20 whereas this was not the case in this series.

Another matter of great concern is whether long-term postremission therapy with ATO would result in chronic arsenicosis or secondary malignancy. The present study shows that although periodic ATO therapy continued for more than 4 years, no severe toxic effects or second malignancies could be attributed to administration of ATO even on long-term follow-up. Moreover, the analysis of arsenic retention from long-term survivors indicated that no significant ATO accumulation was observed in patients who had been off ATO therapy for more than 36 months. These results reinforce the long-term safety of the regime.

The single-agent ATO regimen is particularly suitable for elderly APL patients who are not tolerant to conventional chemotherapy. First, the side effects of ATO are mild and manageable. Even for elderly patients with antecedent cardiac, hepatic, and/or renal diseases, it also can be applied with discretion under carefully monitored conditions, since there is no better alternative. Second, as a result of the absence of myelosuppressive effects during postremission therapy, ATO treatment can therefore not only shorten the duration of hospitalization but also avoid deaths from infection in HCR, therefore prolonging the survival time of the elderly patients.

In conclusion, the data we show in this study demonstrate that the single-agent ATO regimen is highly effective and safe treatment for elderly APL patients without either myelosuppression or significant risks to develop chronic arsenicosis or secondary malignancy. Further studies with multiple centers, larger sample sizes, and complete cytogenetic and molecular data are highly desired to improve and optimize the treatment for elderly APL patients.


No specific funding was disclosed.


The authors made no disclosure.