An Outcome Analysis of Childhood Acute Promyelocytic Leukemia Treated with Atra and Arsenic Trioxide, and Limited Dose Anthracycline

Abstract

The overall survival of Acute Promyelocytic Leukemia (APL), reported in recent studies, is approaching to 90% wherein, arsenic trioxide (ATO) and all-trans retinoic acid (ATRA) are used as the mainstay of treatment with either limited or no use of anthracycline and cytarabine. This study is aimed to ascertain the outcome of children with APL using similar approach. A total of 30 patients with APL, registered from January 2015 to December 2018, were reviewed. Diagnosis was established on bone marrow aspirate and confirmed by the presence of PML-RARA translocation. Treatment protocol was based on Australian APML 4 study performed by Australian Leukemia Lymphoma Group (ALLG). Lumbar puncture was not performed as it was not part of the protocol due to the risk of bleeding. The mean age in current cohort was 9 years with 53% males. Seven (23.3%) patients died and three (10%) abandoned treatment during induction. Twenty patients completed the intensive phase of chemotherapy and all (100%) of them attained molecular remission (MR). One patient dropped out after MR whereas, 19 remain on follow up with no evidence of disease, reflecting disease free survival (DFS) of 95%. With a median follow up of 2.5 years (range 2.1–4.8 years) the 5 years Kaplan–Meier estimate of OS was 63% and 73%, with and without abandonment, respectively. Analysis of outcome according to risk groups revealed inferior outcome of high risk (HR) group (38% and 50% with and without abandonment, respectively) in contrast to standard risk (SR) group which showed better outcome (82% and 88% with and without abandonment, respectively). The attainment of 100% molecular remission and absence of relapse supports the effectiveness of this regimen. Moreover, it is found to be less toxic and therefore, can be conveniently managed in day-care settings.

Introduction

Acute Promyelocytic Leukemia (APL) represents 4–8% of childhood acute myeloid leukemia (AML) and is classified as AML-M3 according to French American British (FAB) classification [1–3]. The hallmark of disease is reciprocal translocation between chromosome 15 and 17 which results in production of fusion protein PML-RARA that stops promyelocytes from differentiating into mature granulocytes.

All-trans retinoic acid (ATRA) and arsenic trioxide (ATO) act on PML-RARA protein to successfully mediate cell differentiation and treat APL [4–7]. Aggressive supportive measures and timely use of differentiating agents such as ATRA and ATO are of paramount importance in reducing the coagulopathy associated early mortality in APL [8, 9]. Since the introduction of these differentiating agents along with chemotherapy, the overall survival of APL has remarkably improved, approaching to 90% [10–12]. However, the morbidity associated with chemotherapy, such as anthracycline induced acute and chronic cardiotoxicity, remained a serious concern [13]. Data of recent clinical trials suggest that low risk APL can potentially be cured without chemotherapy and for high risk cases; comparable survival can be achieved with limited dose of anthracyclines [14–16]. Moreover, the role of cytarabine in APL is unclear and trials suggested that its omission did not demonstrate differences in terms of either complete remission (CR) or induction failure rates. Furthermore, chances of relapse in APL are noticeably low as compared to other AML [17–19]. Therefore, by using less intensive regimen, chemotherapy induced acute and long-term morbidity can be minimized without compromising the outcome.

Before 2015, all patients of APL in our institute were receiving United Kingdom (UK) Medical Research Council (MRC) based intensive chemotherapy along with ATRA. The unpublished data of that cohort reflects higher incidence of treatment related toxic morbidity and mortality resulting prolonged admissions and poor outcome. Therefore, we changed our approach according to the recent studies which suggest ATO as an important component of APL treatment [10, 15]. This regimen is primarily based on ATRA and ATO along with limited dose of anthracycline which was used in the first week of induction only. The rationale was to improve survival of this highly curable disease with less toxic and more cost-effective treatment regimens.

Material and Methods

Study Design

This is a retrospective study conducted in the Pediatric Hematology Oncology section of The Indus Hospital in Karachi, Pakistan after approval of the hospital ethic committee. Patient’s data was retrieved from the electronic medical record and documented on specifically designed case report forms. A total of 30 patients, aged 1 to 16 years, were diagnosed as APL from January 2015 to December 2018. The diagnosis was established on bone marrow aspirate (BMA) morphology by using World Health Organization (WHO) classification and confirmed by the presence of PML-RARA translocation on interphase fluorescent in situ hybridization (i-FISH) [3].

Treatment Protocol

Treatment regimen was based on an Australian APML4 study performed by Australian Leukemia Lymphoma Group (ALLG) [15]. All patients received the same protocol irrespective of their risk group. However, categorization to standard risk (SR) and high risk (HR) was done according to presenting white blood cell count (WBC) to see the difference in outcome. Patients with a WBC < 10,000/µL on the initial CBC at diagnosis are defined as SR and patients with a WBC ≥ 10,000/µL on the initial CBC at diagnosis are defined as HR in accordance with COG AAML0631 trial [20]. ATRA was started on morphological suspicion of promyelocytes whereas rest of the induction chemotherapy, including ATO and daunorubicin, was started after BMA and cytogenetic testing (Table 1). Since central nervous system (CNS) involvement in APL is extremely rare, lumbar puncture and CNS-directed treatment was not part of the protocol to avoid higher chances of procedure related bleeding. In contrast to APML4 study, re-assessment marrow was done once only at the end of consolidation II to document molecular remission (MR) by qualitative polymerase chain reaction (PCR) and those in remission, were followed with clinical and complete blood count parameters during the rest of the treatment course. Patients who failed to achieve remission at the end of consolidation II were to be considered off protocol. As our hospital is a philanthropy-based, non-governmental organization (NGO) which provides free-of-cost treatment, therefore, this deviation from the protocol was made in accordance with institutional policy to make treatment cost effective. Patients remained admitted during induction until they became afebrile, without any active bleeding or transfusion requirements and no signs of differentiation syndrome. Rest of the chemotherapy was administered in the day-care setting and admissions were reserved for febrile neutropenia or other toxicities.

Table 1.

Treatment protocol

Phase	Drugs	Dose and schedule
Induction	*Prednisolone	1 mg/kg/day q 8 h PO; days 1–10
	ATRA	45 mg/m2/day q 12 h PO; days 1–36
	ATO	0.15 mg/kg/day over 2 h IV; days 9–36
	#Daunorubicin	42 mg/m2/dose over 15 min IV; days 2, 4, 6, 8
Consolidation I	ATRA	45 mg/m2/day q 12 h PO; days 1–28
	ATO	0.15 mg/kg/day over 2 h IV; days 1–28
Consolidation II	ATRA	45 mg/m2/day q 12 h PO; days 1–7, 15–21, 29–35
	ATO	0.15 mg/kg/day over 2 h IV; days 1–5, 8–12, 15–19, 22–26, 29–33)
Maintenance (8 Cycles of 3 months each)	ATRA	45 mg/m2/day q 12 h PO; days 1–14
	6-Mercaptopurine	50 mg/m2 /day q 24 h PO; days 15–90
	Methotrexate	15 mg/m2 /day q weekly PO; days 15–90

ATRA all-trans retinoic acid, ATO arsenic trioxide, PO Per oral

^*Prednisolone was given as a prophylaxis of differentiation syndrome

^#Idarubicin was not available, therefore, replaced with Daunorubicin

Supportive care: As per APML 4 protocol oral prednisolone was given as a prophylaxis for differentiation syndrome to all patients irrespective of their presenting WBC at a dose of 1 mg/kg/day every 8 h for first 10 days of induction or until WBC counts fall below 1 × 10⁹/L. Dexamethasone was substituted for prednisolone for patients with manifestations of differentiation syndrome at a dose of 10 mg/m²/day. Routine monitoring of complete blood count (CBC) and prothrombin time (PT/INR) was performed daily for the initial two weeks of induction. Fibrinogen levels could not be performed for more than once in a week due to unavailability of in-house testing. Blood product transfusions from random donors were done for active bleeding and prophylactically to keep hemoglobin > 7 gm/dl, and platelets > 30 × 10⁹/L. Prophylactic antibiotics, antifungals and granulocyte colony stimulating factor (G-CSF) were not used. ECG once per week and serum electrolytes twice per week were done for monitoring of ATO associated toxicities. In those children who were unable to swallow ATRA capsules, doses were administered after softening the capsule in warm milk and then allowing child to chew and swallow. None of the children needed enteral tube administration of ATRA.

Data Analysis

Data was analyzed using the Statistical Package for Social Studies (SPSS; IBM, version 21.0 Armonk, New York, United States). Mean ± SD and median (Interquartile range; IQR) were computed for age and WBC. Frequency with percentages was computed for sex, cytogenetic status at presentation, risk group, molecular remission (MR) and treatment failure (abandonment, death, and relapse). A primary end point was MR, defined by the absence of detectable PML-RARA transcripts by PCR at the end of consolidation II of chemotherapy. Overall survival (OS) was defined as the time from the date of diagnosis till last follow-up or death from any cause. Event-free survival (EFS) was defined as time elapsed from date of diagnosis to occurrence of an event (relapse, death due to any cause and abandonment). Disease free survival (DFS) was defined as the time from documented MR to the earlier of either relapse or death or last contact if patients continued alive in remission. Early death was defined as death during the induction phase of chemotherapy. The Kaplan–Meier method was used to estimate the survival probability and log-rank test was used to determine statistical significance.

Results

There were 30 confirmed cases of APL in this cohort. The clinical variables of the entire cohort in accordance with the risk groups is displayed in Table 2.

Table 2.

Clinical variables

	Total cohort (n = 30)	Standard risk (n = 17)	High risk (n = 13)	p value
Sex, n (%)
Male	16 (53%)	9 (53%)	7 (54%)	0.961a
Female	14 (47%)	8 (47%)	6 (46%)
Age (years)
Mean ± SD	9.01 ± 3.8	9.4 ± 3.9	8.5 ± 3.9	0.540b
Median (range)	9.9 (1.6–16)	10 (3.0–16)	9 (1.6–14)
White cell count (× 109/L)
Mean ± SD	23.6 ± 7.2	4.0 ± 2.7	49.46 ± 45.59	0.000*b
Median (range)	7.5 (1.0–178)	3.0 (1.0–9.9)	37.0 (10.5–178)
Abandonment, n (%)	4 (13.3%)	1 (6%)	3 (23%)	0.175a
Mortality, n (%)	7 (23.3%)	2 (11.8%)	5 (38.5%)	0.876a

^aChi-square test, ^bIndependent t-test, *Significant value

One patient abandoned and two died within 24 h of presentation, therefore, induction was started in the remaining 27 cases. Five patients died and two abandoned therapy during the induction phase (first 35 days) of treatment. The remaining 20 patients completed the intensive phases of chemotherapy and were evaluable for MR by PCR on a BMA sample at the end of consolidation II. All (100%) of them were found to be in MR. One patient abandoned during maintenance phase (after achieving MR); hence 19 patients have remained on follow up till date. There were seven (23.3%) deaths, reported within four weeks of presentation and among them 5/7 (71%) were in the HR group. The details of adverse events and died patients are given in Table 3.

Table 3.

Comparison of alive and expired patients

	Alive N = 19	Expired N = 7	p-value
Age (years)
Mean ± SD	8.9 ± 3.8	9.3 ± 4.4	0.793b
Median (range)	9.8 (3.0–16)	10 (1.6 -15)
Sex
Male	11(48%)	5(71%)	0.399a
Female	12(52%)	2(29%)
WBC (× 109/L)
Mean ± SD	17.5 ± 36.7	43.8 ± 33.9	0.104b
Median (range)	7 (1–78)	50 (2.8–85)
Platelets (× 109/L)
Mean ± SD	28.3 ± 38.0	26.8 ± 24.4	0.926b
Median (range)	20 (3–191)	12 (5–65)
Bleeding	05 (26%)	04 (57%)	0.159a
CNS	–	2
Pulmonary	1	1
GI	1	1
Mucocutaneous	3	–
DS	01 (5%)	01 (14%)	0.474a
Septicemia	02 (10%)	02 (28%)	0.287a
Survival (days)
Median (range)	759 (363–1847)	12 (1–28)	0.000 *b

^aFisher Exact test, ^bIndependent t-test, *Significant value

WBC white blood cell, CNS central nervous system, GI gastrointestinal, DS differentiation syndrome

As there is no relapse to date, therefore, survival (OS and EFS) for the total cohort is the same. With a median follow up of 2.5 years (range 2.1–4.8 years) the 5 years Kaplan–Meier estimate of OS was 63% and 73%, with and without abandonment, respectively. The survival was found to be significantly different with respect to the risk groups (p-value 0.023); with abandonment included, it was found to be 82% for SR patients and 38% for HR cases. Besides one patient who abandoned therapy after achieving MR, no relapse or death was observed in the remaining 19 patients, which reflects DFS of 95% at 2.5 years of follow-up. These results are depicted in Fig. 1.

Fig. 1.

a Overall survival of total cohort, b Overall survival of standard risk (SR) and High risk (HR) with abandonment, c Overall survival of standard risk (SR) and high risk (HR) without abandonment and d Disease free survival (DFS)

Discussion

In addition to anti-leukemic activity and role of ATO in relapsed APL, its synergistic effect is well observed in small randomized trials when used in combination with ATRA [7]. We adopted a similar approach where combination of ATRA and ATO was used from induction phase with limited (4 doses) of anthracycline chemotherapy on days 2, 4, 6 and 8 of induction. When compared to historical (unpublished) data from our institute where intensive AML type of chemotherapy was used, treatment related morbidity and mortality was reduced in this cohort. Out of 22 patients in historical data who were treated on MRC based protocol, EFS was 45% (10/22 patients). In remaining 12 patients, 9 (41%) were died during treatment and 3 (14%) were relapsed. Moreover, treatment related morbidities including prolonged myelosuppression and mucositis was observed more frequently. In the current cohort, ATRA related toxicities including differentiation syndrome, headache and vomiting were reported in few patients. One patient died of differentiation syndrome whereas another patient required temporary discontinuation of ATRA along with therapeutic dose of dexamethasone. Similarly, transient dose reduction of ATRA was needed in a few patients for headache and vomiting. ATO was well tolerated in our patients and other than nausea and vomiting which responded to symptomatic treatment. No neurological or cardiac toxicities, including prolongation of QTc interval, were observed.

The median age of 9 years in our cohort is lower than that in various clinical trials conducted in western countries, including GIMEMA-AIDA 0493, PETHEMA group and European APL93, where the median age is reported 11.6, 12 and 15 years, respectively [12, 21, 22]. We observed 53% males and 47% females in our cohort. The Italian and Australian data comprised of equal male to female prevalence and studies from Spain and Korea showed higher female sex frequency [12, 15, 21, 22]. Our data showed 43% HR cases. The HR cases in clinical trials from developed countries is reported to be between 34 and 39% [12, 20, 21]. The difference may be attributed to delay in early recognition and referral of APL patients in developed countries but may not be reliably interpreted due to small number of cases in our cohort.

In the current cohort, re-assessment bone marrow aspirate was performed once only (post consolidation II) to document MR by using qualitative PCR. This contrasts with APML4 study where bone marrow morphology, cytogenetics and quantitative PCR were performed after induction and each cycle of consolidation. As decision of making patient off protocol was based on the status of post consolidation II MR, therefore, we followed this modified approach due to limited resources and unavailability of quantitative PCR facility in our setting. Consistent with the APML4 and COG AAML0631 studies, all 20 patients of our cohort that were evaluable for post consolidation PCR were found to be in MR [15, 21].

Despite having 100% MR, we observed 63% and 73% survival of total cohort with and without abandonment, respectively which is inferior to data available from developed countries where survival rate of 80–95% is reported in different clinical trials [11, 12, 15, 21]. However, our results were comparable to studies from countries with limited resources which show survival rates of 65–75% [23, 24]. On subsequent analysis of outcome according to risk groups, HR group revealed distinctly inferior outcome (38% and 50% with and without abandonment, respectively) in contrast to SR group which showed better outcome (82% and 88% with and without abandonment, respectively). The outcome of SR group of patients correlates well with results of major pediatric APL studies from developed countries [11, 12, 15, 20, 21].

Abandonment and early death are well known reasons of treatment failure in pediatric oncology in developing countries. Early death in APL is reported to be as high as 20%, a finding that contrasts with resource-rich countries where it ranges from 3 to 10% [20, 25–27]. Similar to developing countries, early death rate was found to be high (23.3%) in our cohort with majority (5/7–71%) of deaths were seen in HR cases. This finding indicates that high WBC remains the most significant predictor of early death in our cohort as has been reported previously [28, 29]. Our study demonstrated a pattern like these reports, wherein bleeding is the predominant cause of death (4/7 deaths—57%) followed by sepsis (2/7 deaths—29%) and differentiation syndrome (1/7 deaths—14%). This likely reflects a lack of comprehensive and standardized guidelines for the initial supportive care needed for the life-threatening bleeding, sepsis, and differentiation syndrome that is more prevalent in HR cases. Specific measures are needed to reduce early mortality in future including early recognition and referral of disease and prompt administration of ATRA and prophylactic steroids at the first suspicion of disease. Strict monitoring of platelet counts daily and fibrinogen levels at least twice weekly with liberal transfusion of blood products to keep platelet counts > 30 × 10⁹/L, fibrinogen level > 150 mg/dl, and prothrombin time and activated partial thromboplastin time within reference limits. Vigilant infection control practices and use of prophylactic antibiotics can be considered during induction and patient should not be discharged until there are no transfusion requirements, no risk for DS and ANC > 0.5 × 10⁹/L.

Higher rates of abandonment (13.3% in current cohort) are also an important contributor to treatment failure in developing countries which demands holistic approach to address the possible reasons, ranging from socioeconomic status to alternative medicine.

Conclusion

The attainment of 100% molecular remission and absence of relapse supports the effectiveness of this regimen. It is also found to be less toxic and therefore, can be conveniently managed in day-care settings. However, due to a relatively small sized, single center cohort, we suggest multicenter, larger scale study in our population for more reliable and meaningful interpretation.

Abbreviations

APL

Acute promyelocytic leukemia

AML

Acute myeloid leukemia

FAB

French American British

RARA

Retinoic acid receptor alpha

PML

Promyelocytic leukemia

ATRA

All-trans retinoic acid

CR

Complete remission

MRC

Medical Research Council

TIH

The Indus hospital

EMR

Electronic medical record

CRF

Case report form

BMA

Bone marrow aspirate

WHO

World Health Organization

FISH

Fluorescence in-situ hybridization

ALLG

Australian leukemia lymphoma group

WCC

White cell count

CNS

Central nervous system

MR

Molecular remission

PCR

Polymerase chain reaction

NGO

Non-government organization

CBC

Complete blood count

PT

Prothrombin time

G-CSF

Granulocyte colony stimulating factor

SPSS

Statistical package for social studies

IBM

International business machines

SD

Standard deviation

IQR

Interquartile range

OS

Overall survival

EFS

Event free survival

DFS

Disease free survival

ED

Early death

LR

Low risk

HR

High risk

Compliance with Ethical Standards

Conflict of interest

The authors report no conflict of interest.