Epidemiology of childhood malignant mediastinal masses and clinical factors associated with intensive care unit admission: A Singapore experience.

AIM: Majority of mediastinal masses in children are malignant. These masses are complex to manage as they have a risk of compression to surrounding structures. Many of these children have to be managed in the intensive care unit (ICU). Hence we sought to evaluate the local epidemiology of malignant mediastinal masses in children and their clinical presentation, and identified factors associated with ICU admission so that at-risk patients may be identified early. METHODS: This study is a retrospective review of institutional case records of 56 children below 18 years of age from 2000 to 2015 with a malignant mediastinal mass. We collected data on their presenting symptoms, clinical signs, radiological investigations, treatment and correlated these factors with admission to our ICU. RESULTS: Lymphoma was most common diagnosis, comprising 37 children (66.0%). There were 6 patients with neuroblastoma (10.7%), 3 patients with germ-cell tumour (5.4%) and 10 patients with T-cell acute lymphoblastic leukaemia (17.9%). Overall, 21 patients (37.5%) had to be admitted to the ICU. Almost all patients (98.2%) were symptomatic on presentation, of which lymphadenopathy was the most common (69.6%). Factors that are significantly associated with ICU admission are stridor, pericardial effusion and need for pleural drainage. CONCLUSIONS: Malignant mediastinal masses in children in our institution range from leukaemias and lymphomas to germ cell tumours and neuroblastomas, of which almost all are symptomatic. These children have a risk of cardiorespiratory collapse and many of them require intensive care. We identified factors that are associated with ICU admission, with the aim of early intervention of at-risk cases.

Successful toxicity reduction during delayed intensification in the non-high-risk arm of Malaysia-Singapore Acute Lymphoblastic Leukaemia 2010 study.

In non-high-risk (non-HR) patients, the Malaysia-Singapore Acute Lymphoblastic Leukaemia 2003 (MS2003) study achieved good outcomes. However, its delayed-intensification (DI) phase, comprising repeated blocks of protocol III (2003-PIII), was toxic and caused significant treatment delays. The successor MS2010 study attempted to lower DI toxicity by replacing myelosuppressive drugs (doxorubicin, cytarabine) with vincristine and asparaginase. PATIENTS AND METHODS: We analysed 1748 admissions for fever in 315 Singapore children with non-HR acute lymphoblastic leukaemia (ALL) (MS2003, n = 183; MS2010, n = 132), comprising 76% of the total cohort (n = 413), to study the impact of these changes. RESULTS: The new 2010-PVa which has no doxorubicin, was associated with significantly fewer hospitalisations due to fever (0.08 versus 0.30 admissions per block [A/blk], p < 0.001), as than 2003-PIIIa. Similarly in 2010-PIIIb and PVb, where one block of cytarabine was replaced by two doses of vincristine, admissions for fever were also fewer (0.47 versus 0.74 A/blk, p < 0.001) than in 2003-PIIIb. However, the addition of single doses of vincristine and asparaginase in 2010-PIIIa, even with a mandatory seven-day rest, led to more hospitalisations (0.45 A/blk, p < 0.001), increased risk of bacteraemia (relative-risk (RR) = 7.66, p = 0.005) and critical-care admissions (RR = 4.31, p = 0.13). Despite this, overall treatment-related mortality decreased from 2.7% to 0.8%. Taken together, the reduced phase delays allowed earlier completion of the intensive phase of treatment (standard risk: 38.1 versus 49.4 weeks, p < 0.001; intermediate risk: 50.9 versus 58.8 weeks, p < 0.001), while maintaining excellent 10-year event-free survival of 95.4% and overall survival of 96.2%. CONCLUSIONS: In non-HR ALL, replacing doxorubicin/cytarabine with vincristine/asparaginase during some DI blocks is effective in reducing toxicity without compromising outcomes. CLINICAL TRIAL INFORMATION: NCT0289464.