Rare presentation of serous effusions in pediatric acute myeloid leukemia: insights from a retrospective case series

Abstract

Background

Serous effusions are rare extramedullary manifestations of pediatric acute myeloid leukemia (AML), for which limited data are available. This study aimed to systematically characterize the clinical features, biological profiles, and outcomes of pediatric AML patients who presented with serous effusions.

Methods

A retrospective case series was conducted at Beijing Children’s Hospital and Beijing Children’s Hospital Baoding Hospital from January 2015 to December 2024. Pediatric patients newly diagnosed with AML or isolated myeloid sarcoma presenting with serous effusion were included.

Results

Seven children (six males and one female) were included. The median time from symptom onset to diagnosis was 30 days (range: 10–150 days). Initial white blood cell counts ranged from 3.39 × 10⁹/L to 212.13 × 10⁹/L, with two patients exhibiting hyperleukocytosis. Genetic alterations included KMT2A rearrangements (4/7), RUNX1::RUNX1T1 fusions (2/7), and SFPQ::ZFP36L2 fusion (1/7). All patients had pleural effusions, and six had pericardial effusion. Four patients had additional extramedullary involvement. The median follow-up was 12.0 months (range: 1.0–68.0 months). Owing to critical illness, six of the seven patients received induction therapy in the intensive care unit. One patient died after induction therapy because of severe acute suppurative appendicitis and sepsis; the remaining six patients underwent hematopoietic stem cell transplantation. At the last follow-up, four patients were alive, while two died of early relapse following stem cell transplantation.

Conclusions

Serous effusion is a rare manifestation of pediatric AML, which is observed in patients with high-risk genetics and is often accompanied by a critical clinical course necessitating intensive care. Further studies are needed in this subgroup.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12887-026-06882-8.

Introduction

Acute myeloid leukemia (AML) is a heterogeneous hematologic malignancy with diverse clinical presentations and genetic profiles [1]. It is characterized by the rapid proliferation of abnormal myeloid precursor cells in the bone marrow and peripheral blood, leading to impaired normal hematopoiesis [2, 3]. In children, AML can present with a wide range of symptoms, including fever, fatigue, hemorrhage, and extramedullary infiltration. The presence of extramedullary involvement, such as isolated myeloid sarcoma (MS) or serous effusions, can complicate the diagnosis and treatment process [4]. Although serous effusions, including pericardial, pleural, and peritoneal effusions, have been reported in adult AML patients [5–7], with an incidence of 0.1%-0.4% for AML-associated serous effusions [8, 9], they are extremely rare in pediatric patients, with limited reports in the literature.

The rarity of AML with serous effusions in children highlights the need for a deeper understanding of its clinical manifestations, biological traits, and prognostic implications. Serous effusions in such cases might suggest a more aggressive disease course and extramedullary involvement, potentially resulting in delayed diagnosis and increased therapeutic difficulties [8, 10, 11]. Therefore, this study aimed to summarize the clinical features, biological properties, and outcomes of pediatric patients with AML who presented with serous effusions. We hope that the insights gained from this research will bolster diagnostic precision and enhance therapeutic effectiveness for this distinct group of patients.

Methods

This retrospective case series was conducted at Beijing Children’s Hospital and Baoding Hospital, Capital Medical University. We enrolled consecutive pediatric patients who were diagnosed with AML or isolated MS and who presented with serous effusion between January 2015 and December 2024. Informed consent was obtained from the parents or legal guardians of all the patients included in the study, and the study was conducted in accordance with the principles of the Declaration of Helsinki and approved by the Ethics Committee (2015-43, 209-k-343, [2023]-E-180-Y-01).

Inclusion and exclusion criteria

The inclusion criteria were as follows: (1) patients aged ≤ 18 years at the time of diagnosis and (2) children with a confirmed diagnosis of AML or MS on the basis of the uniform morphology–immunology–cytogenetics–molecular biology criteria. For cases of isolated MS, confirmation must be obtained through expert consultation in the pathology departments of at least two top-tier hospitals with consistent diagnostic opinions; (3) the presence of at least one type of serous effusion (including pericardial effusion, pleural effusion, peritoneal effusion, or pelvic effusion) prior to the initiation of antileukemic treatment, confirmed by radiography, computed tomography scan, or ultrasound examination. The exclusion criterion was patients whose clinical, laboratory, or follow-up data were incomplete.

Data collection

Data were collected from the medical records of eligible patients. The following information was extracted and analyzed. (1) Basic information, which consisted of age at onset, sex, and complete blood count at diagnosis, including white blood cell (WBC) count, hemoglobin (Hb) level, and platelet count. Detailed morphology–immunology–cytogenetics–molecular biology profiles, including bone marrow morphology, immunophenotyping, fusion gene, and mutation gene results, were also recorded. (2) Serous effusion-related information, such as the location of serous effusion (pericardial, pleural, peritoneal, or pelvic), imaging findings (such as radiography and computed tomography scans) associated with serous effusions, immunophenotyping of cells from serous effusions, and treatment methods for serous effusions, was recorded. (3) Treatment and prognosis: The specific treatment protocols used (CCLG-AML2015, CCLG-AML2019, or CCLG-AML2024), follow-up duration and outcomes, including relapse and survival status, were documented and evaluated.

Treatment protocols and evaluation

Patients were treated according to the CCLG-AML2015 [12], CCLG-AML2019 (ChiCTR1900027146), or CCLG-AML2024 (ChiCTR2400080307) protocols on the basis of their date of diagnosis. These protocols are standardized treatment guidelines for pediatric AML, incorporating induction chemotherapy, consolidation therapy, and hematopoietic stem cell transplantation (HSCT) for eligible patients. For isolated MS, a high-risk AML regimen was used. The specific regimens used were determined by the treating physicians on the basis of the patients’ clinical status and genetic profile. Treatment response was evaluated on the basis of the following criteria: complete remission (CR) was defined as < 5% blasts in the bone marrow. Flow cytometry minimal residual disease (FCM-MRD) testing was performed on bone marrow aspirates after each treatment cycle; FCM-MRD < 0.1% in bone marrow was considered negative. Regular tumor focus assessments were conducted for MS.

Statistical analysis

Statistical analysis was performed using SPSS 22.0 software. Given the small sample size, only descriptive statistics were applied without any comparative or inferential statistical analyses. Normally distributed continuous data are expressed as the mean ± standard deviation, whereas nonnormally distributed continuous data are expressed as the median (minimum, maximum).

Results

Patient characteristics

A total of 7 children, six males and one female, were included in the study. Their ages ranged from 2.0 to 17.0 years (median: 6.0 years) (Table 1). The median time from symptom onset to diagnosis was 30 days (range: 10–150 days). At diagnosis, the WBC counts had a wide range, from 3.39 × 10⁹/L to 212.13 × 10⁹/L, with two patients presenting with hyperleukocytosis. Anemia was observed in six patients, whereas thrombocytopenia occurred in four patients. Among the patients, three were diagnosed with AML alone, three with AML and MS, and one with isolated MS. According to the French–American–British (FAB) classification criteria, in addition to one case of isolated MS, three patients were classified as M5, two as M2, and one had an uncertain FAB classification. Genetic alterations were prominent: four patients harbored KMT2A rearrangements (KMT2A::AF9/AF10), two carried the RUNX1::RUNX1T1 fusion, and one had the SFPQ::ZFP36L2 fusion. The presence of serous effusions was closely associated with extramedullary involvement, which was observed in four of seven patients (Table 2). Notably, Patient 2 exhibited multisite leukemic infiltration involving the heart, central nervous system, testes, and abdominal organs. Moreover, Patients 3, 5, and 7 presented with mediastinal or pleural masses. Immunophenotypic analysis revealed consistent expression of CD34, CD117, and CD33 across all the patients, with CD56 positivity identified in five patients (Table 2).

Table 1.

Clinical features of pediatric acute myeloid leukemia with serous effusion

Patient	Sex	Age (years)	WBC (×109/L)	Hb (g/L)	PLT (×109/L)	Time from symptom onset to diagnosis (days)	Diagnosis	FAB classification	BM Blast	Fusion gene	Mutant gene	Karyotype
1	female	12	10.20	105	3	21	AML	M2	57.50%	RUNX1::RUNX1T1	c-kit mutation	46,XX, t(8;21)(q22;q22)[20]
2	male	3	4.44	83	141	150	AML with MS	M5	75%	KMT2A::AF9	-	46XY, t(9;11)(p22;23)[18]/47,idem,+8[1]/46,XY[1]
3	male	6	3.68	70	114	30	AML with MS	M2	39%	RUNX1::RUNX1T1	NRAS mutation	46,XY, t(8;21)(q22;q22)del(9)(q22)[19]/46,XY[1]
4	male	11	212.13	68	55	10	AML	M5	95%	-	FLT3-ITD, WT1 mutation	47,XY,+8[10]
5	male	17	3.39	144	305	60	MS	-	-	SFPQ::ZFP36L2	FLT3-ITD, JAK3 mutation	-
6	male	6	126.77	83	86	60	AML	indefinite	92.50%	KMT2A::AF10	KRAS, SETD2 mutation	46,XY, del(10)(p11.2), add(11)(q23)[18]/47, idem, + 21[2]
7	male	2	4.63	52	77	30	AML with MS	M5	96%	KMT2A::AF9	-	46,XY,1qh+,t(9;11)(p22;q23)[14]/46,idem, add(4)(q31.3)[6]

WBC white blood cell, Hb hemoglobin, PLT platelet, MS myeloid sarcoma, BM bone marrow, AML acute myeloid leukemia

Table 2.

Immunophenotype, extramedullary involvement and management of serous effusions in pediatric acute myeloid leukemia patients

Patient	Bone marrow immunophenotype	Extramedullary involvement	Pericardial effusion	Pleural effusion (unilateral or bilateral)	Ascites	Pelvic effusion	Percutaneous drainage	Induction therapy in ICU
1	Positive for CD34, CD117, CD13, HLA-DR, CD19, CD38, CD123, MPO	No	Yes	Yes, bilateral	No	No	No	Yes
2	Positive for CD33, CD11b, HLA-DR, CD56, CD117 (a subset of cells), CD15, CD64, CD38	heart, orbit, left lobe of the liver, pancreas, both kidneys, appendix, mesentery, omentum, testicular leukemia, central nervous system leukemia	Yes	Yes, bilateral	No	No	Yes, pericardiocentesis and drainage	Yes
3	Positive for CD34, CD117, CD33, HLA-DR, CD371, CD38, CD10, and CD123. Some cells also express CD13, CD56, and CD19	Mediastinum	Yes	Yes, unilateral	Yes	No	Yes, pleural drainage	No
4	Positive for CD117, CD13, CD33, and HLA-DR. A subset of these cells express CD34, CD123, CD7, CD38, CD64, and CD15dim	No	Yes	Yes, bilateral	No	Yes	No	Yes
5	Pericardial effusion: positive for CD7, CD56 (bright), CD34, CD38, CD117, CD33, CD13, and CD11b (dim). A subset of these cells express HLA-DR and CD64	Mediastinum, pericardial area, bilateral pleura, left hilar region, bilateral internal mammary areas	Yes	Yes, bilateral	Yes	No	Yes, pericardiocentesis and drainage	Yes
6	Positive for s CD34dim, CD117, CD13, CD33, CD64, CD56, and CD4dim. A subset of these cells express HLA-DR, CD7, CD19, CD14, CD38, and CD105	No	Yes	Yes, bilateral	Yes	No	No	Yes
7	Positive for HLA-DR (bright), CD33 (bright), CD38, CD56, CD64, and CD15. A subset of these cells express CD117, CD13, CD36, and CD4	Mediastinum, pleura	No	Yes, unilateral	No	No	Yes, pleural drainage	Yes

Serous effusions and clinical management

Among the seven patients, six had pericardial effusion, including two with massive pericardial effusion (Table 2). Pleural effusion was universal (7/7), with bilateral involvement in five patients and unilateral involvement in two patients. Peritoneal effusion and pelvic effusion were observed in three and one patients, respectively. Among the two patients with massive pericardial effusion, one patient with massive pericardial effusion developed cardiac tamponade. After percutaneous pericardial drainage and chemotherapy were performed, the patient’s symptoms improved, and the AML was controlled (Fig. 1A-C). The second patient, who also had a mediastinal mass, experienced massive pericardial effusion (Fig. 1D-F), resulting in a two-month delay in diagnosis because of the absence of bone marrow involvement. The diagnosis was eventually confirmed through pericardiocentesis and pleural biopsy (mediastinal biopsy was refused). Flow cytometry of the pericardial effusion revealed that 87.12% of the cells expressed CD7, CD56 (bright), CD34, CD38, CD117, CD33, CD13, and CD11b (dim), with partial expression of CD64 and HLA-DR, which are indicators of malignant myeloblasts. Multiple pathology consultations on the pleural lesion biopsy confirmed the diagnosis of MS. The patient, who carried the SFPQ::ZFP36L2 fusion gene, a FLT3-ITD mutation, and a JAK3 mutation, was treated with venetoclax in combination with homoharringtonine and cytarabine, followed by a regimen of venetoclax combined with decitabine and gilteritinib and subsequently administered mitoxantrone and cytarabine. The treatment achieved PET‒CT-confirmed remission, and the patient subsequently underwent haploidentical HSCT. However, the patient relapsed six months later and died because of concurrent circulatory failure. Two patients with massive pleural effusion presented with dyspnea and hypoxemia. Both underwent closed pleural drainage and chemotherapy, resulting in symptom relief and disease control.

Fig. 1.

Illustrative imaging of pericardial effusion before and after therapy in two pediatric patients with acute myeloid leukemia. A Pretreatment chest radiograph of Patient 2, highlighting an enlarged cardiac silhouette indicating pericardial effusion. B Pretreatment chest CT scan of Patient 2 clearly demonstrating substantial pericardial effusion. C Posttreatment chest radiograph of Patient 2 after one course of therapy and pericardiocentesis, revealing a marked reduction in cardiac silhouette size and a decrease in pericardial effusion. D Pretreatment chest radiograph of Patient 5, similarly showing an enlarged cardiac silhouette, suggesting the presence of pericardial effusion. E Pretreatment chest CT scan of Patient 5, revealing pronounced pericardial effusion. F Posttreatment chest radiograph of Patient 5, revealing a significant reduction in the cardiac silhouette size, which indicated successful management of pericardial effusion. These images are provided as illustrative examples and do not represent the findings in all patients

Biochemical and cytological profiles of serous effusions

Among the seven patients, four underwent diagnostic or therapeutic drainage of serous effusions (Table S1). Among these four patients, comprehensive immunophenotyping by flow cytometry was completed in two (Patients 3 and 5). For Patient 2 and Patient 7, immunophenotyping of the effusion was not performed because the AML had already been confirmed by bone marrow examination at the time of drainage. Patient 3 underwent pleural drainage at a local hospital, and routine and biochemical effusion results were unavailable from that facility.

Patient 2’s pericardial fluid exhibited markedly elevated LDH levels (6,616 U/L) and a low glucose concentration (0.26 mmol/L), which was consistent with an exudative effusion characterized by a yellow and turbid appearance with a positive Rivalta’s test and a mononuclear cell predominance (81%). Patient 5 exhibited bloody pericardial effusion with moderate LDH (265 U/L) and high glucose (7.9 mmol/L) levels. Mononuclear cells predominated (88%), and the immunophenotype was 87.12% of the cells expressing CD7, CD56 (bright), CD34, CD38, CD117, CD33, CD13, and CD11b (dim), with partial expression of CD64 and HLA-DR. Patient 7 had unilateral pleural fluid with a yellow and turbid appearance. The LDH concentration was extremely high at 5971 U/L, strongly suggesting leukemic infiltration. Neutrophils dominated the cell population, with polymorphonuclear cells constituting 57% of the total cells. Additionally, Staphylococcus epidermidis was cultured from the pleural effusion. Treatment with vancomycin for anti-infection therapy successfully alleviated the patient’s fever, indicating an improvement in clinical condition.

Primary disease treatment and follow-up

Among the seven patients, six underwent induction therapy in the intensive care unit (ICU) because of the severity of their condition: Patient 1 was admitted to the ICU because of severe pneumonia complicated by respiratory failure, and diffuse alveolar hemorrhage could not be excluded. Patient 2 was transferred to the ICU because of cardiac tamponade and presented with significant dyspnea, moaning, restlessness, and respiratory distress. Patient 4 was admitted to the ICU for acute suppurative appendicitis with perforation and diffuse peritonitis. Patient 5 required ICU admission for pericardiocentesis and catheter placement under ultrasound guidance because of the large volume of pericardial effusion. Patient 6 was transferred to the ICU because of septic shock and cardiac dysfunction during the induction therapy period. Patient 7 was admitted to the ICU because of a large left-sided pleural effusion with decreased oxygen saturation. In terms of treatment protocols, one was treated with the CCLG-AML2015 protocol, four with the CCLG-AML2019 protocol, and two with the CCLG-AML2024 protocol. In accordance with the criteria for risk stratification, all the patients were classified into the high-risk group. The follow-up duration for the seven patients ranged from 1.0 to 68.0 months (median: 12 months). With the exceptions of Patient 4 and Patient 5, all the remaining five pediatric patients achieved CR following Induction I. Patient 4 died after abandoning treatment during induction therapy because of concurrent acute suppurative appendicitis with perforation and diffuse peritonitis, while Patient 5 presented with isolated MS, and bone marrow evaluation was not feasible. Among the five patients who achieved a CR, three achieved MRD negativity after induction II, and another two achieved MRD negativity following consolidation I (Table 3). Six patients underwent HSCT: four were still alive after transplantation, one relapsed one month after transplantation and subsequently died of multiorgan failure, and one relapsed six months after transplantation and died due to circulatory failure.

Table 3.

Response to induction and outcome in pediatric acute myeloid leukemia patients with serous effusion

Patient	Response to Induction I		Response to Induction II		Outcome
	Morphology	FCM-MRD	Morphology	FCM-MRD
1	CR	0	CR	0	Alive
2	CR	1.21%	CR	0	Relapsed one month after HSCT and died of multiorgan failure
3	CR	0.139%	CR	0	Alive
4*	NA	NA	NA	NA	Died after abandoning treatment during induction therapy
5#	NA	NA	NA	NA	Relapsed six months after HSCT and died of circulatory failure
6	CR	0.15%	CR	0.13%	Alive
7	CR	1.35%	CR	0.39%	Alive

CR complete remission, FCM-MRD flow cytometry-detected minimal residual disease, NA not applicable, HSCT hematopoietic stem cell transplantation. *, Patient 4 died after abandoning induction therapy; no postinduction evaluation could be performed. ^#, Patient 5 presented with isolated myeloid sarcoma without bone marrow involvement; thus, bone marrow minimal residual disease assessment was not applicable. Patients 6 and 7 were FCM-MRD-negative after consolidation I therapy

Discussion

Serous effusions, while well documented in lymphoproliferative malignancies, are rare in pediatric patients with AML, with a reported incidence of 0.1%-0.4% in adult patients with AML [8, 9]. However, underdiagnosis due to nonspecific presentations may contribute to epidemiological underestimation. In our study, we provide a comprehensive analysis of the clinical characteristics, treatment approaches, and outcomes of seven pediatric patients with AML with serous effusions, highlighting the critical need for prompt diagnosis and tailored therapeutic strategies.

The diagnosis of AML-associated serous effusions, particularly in MS, remains challenging [13]. Postmortem studies have revealed that cardiac involvement is relatively common in leukemia patients, with an incidence rate of 30–44%. However, approximately 99% of these cases are asymptomatic [14]. This phenomenon underscores the risk of occult pathology, as illustrated by one of our cases in which massive pericardial effusion without bone marrow involvement resulted in a two-month delay in diagnosis. Flow cytometry of effusions and tumor biopsies proved indispensable for confirming leukemic infiltration, in agreement with prior studies [5, 6, 8, 15]. These findings support the use of a multimodal diagnostic approach when conventional methods fail, especially in patients with extramedullary disease.

In this small cohort, monocytic differentiation (FAB-M5) was observed in three of the six patients with AML. This observation was consistent with prior reports linking ​FAB-M4/M5 subtypes ​to extramedullary involvement [16, 17], as corroborated by Kanwalpreet Kaur DM et al., who reported AML-M4/M5 as the predominant subtype associated with effusions (6/9 cases) [8]. The frequent expression of adhesion molecules, notably CD56 (5/7 cases) and CD15 (2/7 cases), observed in our cases may be consistent with previous hypotheses suggesting a role of these surface molecules in extramedullary infiltration [13, 18]. Furthermore, in this small cohort, the mediastinum was the most frequently observed site of extramedullary involvement for AML patients with concurrent or isolated MS. This mediastinal MS, frequently characterized by infiltration into adjacent pleural tissues, may represent a potential pathological mechanism that contributes to pleural effusion formation. Although infections remain a common etiology for pleural effusion in patients with leukemia [19], the presentation of unilateral effusions in our cases—concurrent with mediastinal masses (necessitating drainage for symptomatic relief) and confirmed leukemic infiltration by flow cytometry in one case—emphasizes leukemic cell migration and serosal membrane infiltration as crucial causes of serous effusions.

The management of AML with serous effusions presents distinct therapeutic challenges because of its ​rapid onset and aggressive progression. Our study highlights this critical phenotype: six of seven patients required intensive care unit (ICU)-level management during induction therapy for life-threatening complications, including cardiac tamponade, respiratory failure from massive pleural effusions, septic shock, and surgical emergencies. The prognostic significance of the presence of serous effusion at diagnosis in patients with acute leukemia remains unclear. While some studies suggest that pleural effusion does not impact the rate of remission or survival, conflicting reports have associated this condition with poor prognosis [11, 20]. At the last follow-up, four of the seven patients were alive. All patients received HSCT as part of high-risk AML protocols. Nevertheless, two patients experienced early relapse (1–6 months post-HSCT), reflecting the high-risk biology of AML with serous effusions. However, this study is limited by its small sample size, necessitating validation in larger, multicenter investigations to confirm these findings.

Our study has several limitations that should be acknowledged. Its retrospective single-center design and small sample size limit the generalizability of the findings. Additional limitations include the inclusion of clinically distinct entities (AML, AML with MS, and isolated MS), an extended study period (2015–2024) with evolving diagnostic and therapeutic standards, and incomplete effusion analyses in some cases—all of which introduce heterogeneity and selection bias. Furthermore, the lack of a control or comparator cohort prevents definitive conclusions about the prognostic impact of serous effusions. Future prospective multicenter studies with standardized diagnostic and therapeutic protocols are needed to validate our observations.

Conclusions

Serous effusion is a rare manifestation of pediatric AML, observed in patients with high-risk genetics and often accompanied by a critical clinical course necessitating intensive care. Further studies are needed in this subgroup of patients.

Abbreviations

AML

Acute myeloid leukemia

HSCT

Hematopoietic stem cell transplantation

MS

Isolated myeloid sarcoma

WBC

White blood cell

Hb

Hemoglobin

CR

Complete remission

MRD

Minimal residual disease

FAB

French–American–British

ICU

Intensive care unit

Authors’ contributions

Jiaole Yu, Huyong Zheng, and Linya Wang designed the study. Linya Wang, Pengli Huang, Peijing Qi, Wei Lin, Qian Huang, Mi Liu, Dan Luo, Yan Liu, Ruidong Zhang, Huyong Zheng, and Jiaole Yu treated the patients and contributed data. Linya Wang and Qian Huang collected the data. Linya Wang, Jiaole Yu, Huyong Zheng and Pengli Huang wrote the paper. All the authors reviewed the paper and gave final approval.

Funding

This work was supported by the Beijing Municipal Administration of Hospitals Clinical Medicine Development of Special “Yang Fan” Grant (No. ZLRK202328), the Key Research and Development Health Innovation Project of the Hebei Provincial Department of Science and Technology (22377737D), the National Key Research and Development Program of China (2023YFC2706100) and the Beijing Hospital Authority Youth Program (QML20161202).

Data availability

The data are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The study was conducted in accordance with the principles of the Declaration of Helsinki and approved by the Ethics Committee of Beijing Children’s Hospital (2015-43, 209-k-343, [2023]-E-180-Y-01). Informed consent was obtained from all the patients or their families.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.