From armed-conflict to care: clinical outcomes, diagnostic challenges, and psychosocial burden in children evaluated for suspected cancer displaced from Gaza and treated at a regional center of excellence

Introduction

Childhood cancer is one of the most common causes of death from disease worldwide, with approximately 400,000 new cases diagnosed annually (1, 2). Nearly 80–90% of children with cancer live in LMICs, where inequities in access to timely diagnosis and treatment contribute to persistently poor survival outcomes (3), driven by delayed diagnosis, restricted access to treatment, and health system limitations (4, 5). These disparities are further exacerbated in conflict and forced displacement settings, where disruption of healthcare infrastructure, restricted mobility, and resource scarcity are associated with significant impact on cancer care delivery (6). Between 2020 and 2050, an estimated 13.7 million children worldwide will be diagnosed with cancer, most in resource-scarce settings, and without urgent systems strengthening and conflict reduction, 11.1 million are projected to die (7).

Children affected by armed conflict represent a particularly vulnerable population (8). Gaza exemplifies this vulnerability at an extreme scale: the escalation of conflict from October 2023 has severely degraded an already fragile healthcare system, limiting cancer services and disrupting routine immunization programs (9). Children with suspected or confirmed cancer are frequently evacuated to neighboring countries (9, 10), resulting in substantial diagnostic and treatment delays and advanced disease at first presentation (11, 12). Despite the growing number of conflict-displaced children with cancer, data on their clinical characteristics, diagnostic accuracy, and post-evacuation outcomes remain limited.

The King Hussein Cancer Center (KHCC) has served as a regional referral center for displaced children with cancer. It has provided care through coordinated humanitarian efforts with international partners, such as St. Jude Children’s Research Hospital (SJCRH) (13).

In addition to clinical challenges, these children carry a substantial psychosocial burden compounded by displacement, and conflict-related trauma (14, 15), while accompanying caregivers frequently present with unaddressed medical and psychological needs of their own. This study characterizes the demographic and clinical features of pediatric patients from Gaza referred to KHCC, with a focus on diagnostic concordance, disease stage at presentation, early outcomes, high-risk subgroup requiring intensive care and psychosocial and nutritional burden.

Methods

Study design and population

This retrospective cohort study was approved by the KHCC Institutional Review Board with a waiver of informed consent (IRB #24KHCC213); all data were de-identified. The cohort comprised all pediatric patients (aged ≤21 years) from Gaza evaluated for suspected malignancy at KHCC between November 2023 and March 2026 following medical evacuation to Jordan; the cohort does not include all evacuated pediatric patients broadly. Supplementary Figure 1 shows the patient flow from referral to final diagnostic outcome and the denominators used for each subsequent analysis. Non-malignant conditions were included in descriptive analyses but excluded from stage-at-presentation analyses; leukemia (systemic by definition) and retinoblastoma (staged by the International Intraocular Retinoblastoma Classification (IIRC) rather than Surveillance, Epidemiology, and End Results (SEER)) were also excluded from the pooled SEER staging denominator and are reported separately.

Data were extracted from electronic medical records and included demographics, clinical timelines (symptom onset, diagnosis, and referral), disease characteristics, treatment history, and outcomes. Cancer diagnoses were classified per the International Classification of Childhood Cancer, Third Edition (ICCC-3) (16), and staging used the SEER Program system (17). Diagnostic concordance was assessed between the referring and final KHCC diagnoses. Referring diagnoses from Gaza represented provisional working diagnoses based on the clinical, imaging, and — where available — pathology information accompanying the patient; most patients arrived with documentation (clinical/pathology reports) but without physical tissue. Where original pathology material (slides/blocks) accompanied the patient, it was reviewed by KHCC pathology; in its absence, the referring diagnosis was assessed against the diagnosis subsequently established at KHCC. The KHCC diagnosis — based on complete pathology review, immunophenotyping/molecular work-up, and imaging performed at KHCC — was adopted as the reference standard, reflecting the substantially greater diagnostic information available at the receiving center rather than an assumption that it was infallible. Discordant cases were adjudicated through multidisciplinary tumor board review and categorized as reclassification of cancer type, reclassification to a non-malignant condition, or non-confirmation of suspected relapse. Nutritional status was assessed via weight-for-age percentiles and serum markers (albumin, pre-albumin, vitamin D, and vitamin B12); testing rates varied by marker and are reported per-test rather than assuming universal ascertainment. Psychological distress was assessed using the Distress Thermometer scale (DTS) (18) by a trained psychologist or social worker following an interview with the child and caregiver. For younger children who were unable to provide a reliable self-report, assessment incorporated caregiver input and the clinical interview. A score ≥6 was used to identify clinically significant distress, consistent with the cited literature (19). Accompanying caregiver characteristics, administrative evacuation barriers, and funding source data were also collected. Patients with non-malignant conditions were included in descriptive analyses but excluded from staging and survival analyses. Funding source was extracted from institutional billing and financial records and categorized according to the organization supporting the patient’s care.

Statistical analysis

All statistical analyses were performed using R version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria), with the survival (version 3.5) and survminer (version 0.4.9) packages for Kaplan–Meier estimation and Cox proportional hazards modeling. For continuous variables, we calculated the median and the interquartile range (IQR). For categorical variables, we calculated the frequency and percentage. Categorical variables were compared between groups using Fisher’s exact test; continuous variables using the Mann–Whitney U test. Missing data were handled by exclusion from the denominator for the affected variable.

Overall survival (OS) was defined as time from KHCC presentation to death from any cause. Event-free survival (EFS) was defined as time from KHCC presentation to the first subsequent occurrence of relapse, disease progression, or death from any cause. For patients presenting with relapse or progressive disease, the presenting disease status was considered the baseline state rather than an EFS event. A subsequent event was recorded only if a new event occurred after KHCC presentation, such as death, a new relapse following achievement of remission, or subsequent disease progression. Patients who remained free of a subsequent event were censored at the date of last documented follow-up. Because patients entered the cohort at different points in their disease trajectories, including newly diagnosed, previously treated, and relapsed/progressive disease, survival time was anchored to KHCC presentation rather than original cancer diagnosis. Survival curves were estimated by the Kaplan–Meier method and compared with the log-rank test. A univariable Cox proportional hazards model was used to estimate the hazard ratio for pediatric intensive care unit (PICU) versus non-PICU admission; this model is unadjusted for disease stage, tumor type, and other clinical confounders. Survival is reported at 1 year and 2 years, consistent with the observed follow-up period (median 8.0 months; maximum approximately 28 months; data lock March 31, 2026). Early mortality was defined as death within 30 days of presentation.

Results

Patient characteristics

A total of 169 pediatric patients from Gaza were evaluated at KHCC between November 2023 and March 2026. The cohort was predominantly male (62%, n=105), with a median age of 6.7 years (range, 0.01–21.3); 40% were under 5 years, 27% aged 5–10 years, and 33% aged ≥10 years. Median follow-up was 8.0 months (IQR 5.2–22.0) (Table 1).

Table 1. Baseline characteristics of pediatric patients from Gaza evaluated at KHCC (N = 169).

¹Referral diagnoses reflect initial classification prior to KHCC evaluation. Disease status is assessed at time of arrival to KHCC (pre-treatment), distinct from Table 2’s final diagnosis category assigned after complete KHCC work-up; the non-cancer total here (n=13) includes 9 patients also captured in Table 2’s “Other/non-cancer” category (n=11 total; the remaining 2 arrived with disease status recorded as relapse/progression but were subsequently reclassified as non-malignant at KHCC), 2 patients with non-malignant hematologic disorders (aplastic anemia, transfusion-dependent thalassemia), and 2 patients in whom suspected relapse was ruled out at arrival and who are retained under their original cancer diagnosis category in Table 2, ²Percentage of subgroup. KHCC, King Hussein Cancer Center; PICU, pediatric intensive care unit.

At referral, 88% (n=148) were suspected new cancer diagnosis and 12% (n=21) suspected relapse. Half (50%, n=85) had received prior cancer-directed therapy (chemotherapy, radiotherapy, and/or surgery) in Gaza, though treatment details were frequently incomplete. At presentation, 61% (n=103) were managed as outpatients and 39% (n=66) required admission, including 14% (n=23) directly to the PICU. (Table 1) Median time from symptom onset to KHCC presentation was 3 months (range 1–62). This interval reflects the total elapsed time from symptom onset to KHCC assessment and encompasses any care received in Gaza prior to evacuation.

Disease characteristics and diagnostic concordance

The most common referral diagnoses were solid tumors (33%, n=56), acute leukemia (24%, n=40), lymphomas (18%, n=30), and CNS tumors (14%, n=24) (Table 1). Following KHCC evaluation, 57% (n=96) had active or progressive disease, 36% (n=60) were in remission, and 8% (n=13) were diagnosed with non-malignant conditions. (Figure 1A).

Overall diagnostic concordance was 80.5% (n=136), with discordance in 19.5% (n=33) (Figure 1B): reclassification of cancer type (54.5%, n=18), reclassification to a non-malignant condition (33.3%, n=11), and non-confirmation of suspected relapse (12.1%, n=4). (Figure 1C).

Disease stage at presentation

Because leukemia is systemic by definition and standard SEER staging does not apply to retinoblastoma, stage at presentation was assessed separately by diagnostic category among solid tumors, lymphomas, and CNS tumors (n=101; non-malignant conditions, leukemia, and retinoblastoma excluded from this denominator). Distant disease was present in 44.4% (n=24) of solid tumors, 65.2% (n=15) of lymphomas, and 12.5% (n=3) of central nervous system (CNS) tumors; localized disease in 31.5% (n=17), 4.3% (n=1), and 87.5% (n=21), respectively. All leukemia patients (n=40) presented with systemic disease at diagnosis. Retinoblastoma (n=6) was staged using the IIRC rather than SEER: 1 Group C, 1 Group D, and 4 Group E (67% high-risk) (Table 2).

Table 2. Final diagnosis, SEER stage, and diagnostic concordance (N = 169).

¹Stage classification (Localized/Regional/Distant) applied only to solid tumors, lymphomas, and CNS tumors; leukemia is not staged, as it is systemic by definition; retinoblastoma is staged separately using the International Intraocular Retinoblastoma Classification (IIRC) rather than SEER, as SEER’s localized/regional/distant framework does not capture intraocular disease extent. ²”Not applicable” (n=5 total) comprises: Pt 127 (bone sarcoma), lost to follow-up before staging or diagnostic confirmation could be completed; inflammatory myofibroblastic tumor and giant cell tumor, entities for which the SEER localized/regional/distant framework does not cleanly apply; and 2 patients in whom suspected relapse was ruled out, leaving no active tumor to stage. ³IIRC Group A–E reflects intraocular tumor extent (seeding, retinal detachment risk); Group D–E indicates high-risk disease. CNS, Central nervous system; IIRC, International Intraocular Retinoblastoma Classification; KHCC, King Hussein Cancer Center; RMS, Rhabdomyosarcoma; SEER, Surveillance, Epidemiology, and End Results.

Critically ill subgroup (PICU patients)

PICU-admitted patients (13.6%, n=23) had numerically higher rates of active or progressive disease (57% vs 36%, p=0.07) and significantly higher rates of distant disease (69% vs 37.5%, p=0.038) and infection at presentation (43% vs 8%, p<0.001). Rates of diagnostic discordance were numerically higher in the PICU group (30% vs 18%), though this did not reach statistical significance (p=0.12). Only 17% of PICU patients had received prior therapy in Gaza versus 55% of non-PICU patients (p<0.001) (Table 3).

Table 3. Clinical Characteristics of PICU (N = 23) vs. non-PICU patients (N = 146).

¹p-values calculated using Fisher’s exact test for categorical variables and the Mann–Whitney U test for continuous variables, ²Infection defined as documented bacterial, viral, or fungal infection at presentation. ³Represents 36% of deaths within the PICU subgroup. Distant disease is assessed only among solid tumors, lymphomas, and CNS tumors (SEER-eligible categories; PICU n=13, non-PICU n=88); leukemia, retinoblastoma, and non-malignant conditions are excluded from this denominator (see Table 2). Active/progressive disease reflects disease status at time of arrival to KHCC (columns AA/AB of source data), distinct from Table 1’s post-evaluation status. PICU, pediatric intensive care unit; SEER, Surveillance, Epidemiology, and End Results.

Mortality in the PICU subgroup was 48% (n=11), with early mortality (≤30 days) in 17% (n=4) — all four early deaths in the entire cohort. Median time to death was 61 days versus 182 days in non-PICU patients; the majority of deaths were attributable to progressive disease.

Infectious complications

Infectious complications affected 13% (n=22) at presentation, concentrated among PICU patients (43%). Pathogens included multi-drug resistant (MDR) organisms (Methicillin-resistant Staphylococcus aureus (MRSA), Vancomycin-resistant Enterococcus (VRE), Extended-spectrum beta-lactamase-producing Escherichia coli (E.coli) and carbapenem-resistant E. coli, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa) and viral infections including Coronavirus Disease of 2019 (COVID-19) and rhinovirus/enterovirus. Varicella was confirmed in 6% (n=10), occurring in immunocompromised patients in the context of documented disruption of routine immunization programs in Gaza during the conflict period. Individual vaccination histories were not available in this cohort; a direct causal link between immunization disruption and this specific outbreak cannot be established from the data presented.

Nutritional and psychosocial status

Nutritional deficiency was frequently observed among tested patients, though testing rates varied by marker: underweight status (weight-for-age <5th percentile) was assessed in 167/169 (98.8%) and present in 26.3% (n=44/167); vitamin D was measured in 149/169 (88.2%) and deficient (<20 ng/mL) in 50.3% (n=75/149); albumin was measured in 167/169 (98.8%) and low (<3.5 g/dL) in 9.6% (n=16/167). Pre-albumin was measured in only 39/169 patients (23.1%), reflecting selective ordering in patients with clinical concern for malnutrition; among those tested, 89.7% (n=35/39) had low values (<20 mg/dL), likely reflecting both this ascertainment bias and the acute-phase suppression of pre-albumin by active malignancy and inflammation, independent of chronic nutritional status. (Figure 2). Clinically significant psychological distress (DTS ≥6) was present in 41.4% (n=70).

Caregiver burden and system-level challenges

A total of 431 companions accompanied the 169 patients (median 2, range 1–10); (62%, n=268) were female. Siblings constituted the largest group (56%, n=239), followed by parents (40%, n=173), with other relatives accounting for 4% (n=19). The relationship structure of accompanying companions is reported descriptively; the study does not have data on primary caregiver identity or the specific reasons for companion composition. Administrative and logistical barriers — documentation requirements, movement restrictions, and evacuation delays — further complicated treatment initiation.

Survival outcomes

During follow-up, (Median 8.0; IQR 5.2–22.0; months) data lock March 31, 2026), 24 deaths were recorded (14.2%); early mortality was 2.4% (n=4), all in the PICU subgroup. Most deaths (83%, n=20) were due to progressive disease; 17% (n=4) to infectious or treatment-related complications. Given the short median follow-up, outcomes are reported as 1-year and 2-year landmark rather than 5-year projections. The 1-year OS was 81.3% (95% CI: 72.1%–87.7%) and 2-year OS was 75.7% (95% CI: 64.6%–83.7%); 1-year EFS was 76.5% (95% CI: 67.0%–83.6%) and 2-year EFS was 70.6% (95% CI: 59.4%–79.3%). (Figure 3A). These estimates are consistent with the observed follow-up period; the observation period spanned approximately 28 months and estimates beyond this point should be interpreted with caution.

Among PICU patients, 1-year OS was 52.7% versus 86.8% and 2-year OS was 42.2% versus 82.4% (HR = 5.56, 95% CI: 2.49–12.42, p<0.0001); 1-year EFS was 48.9% versus 81.8% and 2-year EFS was 39.1% versus 77.1% (HR = 4.27, 95% CI: 2.06–8.88, p<0.0001). (Figure 3B) This difference in survival outcomes between PICU and non-PICU patients reflects the substantially greater disease burden, higher rates of metastatic disease, and greater infectious morbidity present in PICU patients at presentation; the univariable Cox model is unadjusted for these and other confounders, and the observed association does not isolate delayed access to care as the sole contributing factor.

Discussion

This study describes the clinical characteristics and early outcomes of children evacuated from Gaza during active conflict and treated at KHCC between November 2023 and March 2026, and is, to our knowledge, one of the largest cohort studies of conflict-evacuated children with cancer at a regional oncology center. The observed clinical profile — advanced disease at presentation, high rates of diagnostic discordance, nutritional deficiency, and psychosocial distress — is consistent with the documented effects of healthcare system disruption in conflict-affected settings (1, 3, 5, 6). These findings should be interpreted as descriptive observations in a conflict-evacuated cohort; the observational study design without a pre-conflict or non-conflict comparator group does not permit causal attribution of individual clinical findings to the conflict itself.

A central observation is the substantial interval between symptom onset and KHCC presentation, with a median of three months. This interval encompasses the full care pathway in Gaza, including prior cancer-directed therapy received by half of patients; it therefore reflects the total elapsed time to specialist evaluation at KHCC rather than a period of continuous diagnostic absence, though it was not always accurate. The observed distribution of advanced disease at presentation is consistent with patterns documented in conflict-affected settings in the published literature (5, 7, 8, 11, 12), though the relative contributions of the symptom-onset interval, prior treatment in Gaza, and underlying disease biology to stage at presentation cannot be disaggregated in this study. The temporal pattern of referrals over the 28-month study period is reported descriptively in Figure 1; formal analysis of temporal trends was not performed and claims about progressive deterioration of referral capacity are not made. The 19.5% rate of diagnostic discordance between referring institution and KHCC diagnoses is a clinically significant finding in a cross-border referral context. Limited access to immunohistochemistry, molecular diagnostics, and advanced imaging is associated with diagnostic discordance in resource-limited settings (20, 21). Among the 33 discordant cases, reclassification of cancer type occurred in 54.5% (n=18), reclassification to a non-malignant condition in 33.3% (n=11), and non-confirmation of suspected relapse in 12.1% (n=4). For cases with documented prior therapy in Gaza, the appropriateness of that treatment in relation to the final KHCC diagnosis cannot be determined from the available data. These findings support the case for diagnostic quality assurance — pathological review and multidisciplinary tumor board evaluation — as a component of cross-border oncology referral programs (21–23).

Patients requiring PICU admission at presentation (13.6%) had markedly worse survival outcomes than non-PICU patients: 1-year OS 52.7% versus 86.8% and 1-year EFS 48.9% versus 81.8% (HR 5.56 and 4.27, respectively; P<0.0001). This survival difference reflects the substantially greater disease burden, higher rates of metastatic disease, and greater infectious morbidity in PICU patients at presentation. As the Cox model is univariable and unadjusted for these confounders, the association between PICU admission and worse survival outcomes is consistent with confounding by disease severity and indication; it does not demonstrate that delayed access to care was the primary driver of mortality in this group (3, 7).

Malnutrition — affecting 26.3% by weight-for-age and 50.3% by vitamin D levels among those assessed — is associated with reduced chemotherapy tolerance, increased infectious susceptibility, and worse survival in pediatric oncology (24, 25). These nutritional findings reflect a combination of underlying malignancy, prior treatment, and displacement-related factors that cannot be disaggregated in this study. Psychological distress was present in 41.4% of assessed patients. Contributing factors may include illness burden, displacement, family separation, and conflict exposure, although their relative contributions could not be disentangled in this study (14, 15). Nutritional assessment and psychosocial support must function as core clinical infrastructure, not supplementary services, at any center receiving conflict-displaced children with cancer. Infectious complications — including MDR organisms — were disproportionately concentrated among the most critically ill (26–28). Varicella was confirmed in 10 patients, including immunocompromised individuals, in the context of documented disruption of routine immunization programs in Gaza during the conflict period (27, 29). Individual vaccination histories were not available in this cohort; a direct causal relationship between immunization disruption and the observed varicella cases cannot be established. Nevertheless, these findings support systematic vaccination screening and urgent catch-up immunization at the earliest clinically safe opportunity upon arrival at receiving centers (9).

This study identifies caregiver instability and system-level administrative barriers as clinically relevant challenges in humanitarian oncology (25, 30, 31). Siblings constituted the largest group of accompanying companions (56%); companion data are reported descriptively and do not permit conclusions about primary caregiver capacity Administrative barriers — security clearances, documentation mandates, and movement restrictions — impose evacuation delays that have clinical implications (10). Structured repatriation planning, including emergency re-evacuation protocols and remote follow-up, should be standard in humanitarian oncology programs (10).

The 1-year OS of 81.3% and 2-year OS of 75.7% observed in this cohort are encouraging and compare favorably with data from equivalent humanitarian settings (12), demonstrating that meaningful survival is achievable even with late, malnourished, advanced-disease presentations — provided patients reach a capable receiving center. It is important to note that these outcomes describe only children who successfully navigated evacuation: those who died before evacuation, for whom evacuation was impossible, or in whom cancer was never diagnosed represent an unquantifiable additional burden not captured in this study. The observed survival reflects multiple contributing factors including the underlying disease distribution, prior treatment received in Gaza, the selection effect of the evacuation process, and the care delivered at KHCC; the independent contribution of each factor cannot be isolated in this observational study.

The KHCC–SJCRH collaboration illustrates the role of international partnerships in enabling continuity of pediatric oncology care for conflict-displaced children (13). In this cohort, 59% of patients were supported through King Hussein Cancer Foundation (KHCF) resources and 41% through SJCRH support. Although the sustainability of these funding arrangements was not formally evaluated in this study, the reliance on philanthropic and institutional support highlights the importance of predictable, long-term financing mechanisms for pediatric cancer care in humanitarian settings.

The following recommendations are informed by the findings of this cohort together with the broader humanitarian oncology literature; several of these areas, including evacuation logistics, telemedicine infrastructure, and financing mechanisms, were not directly evaluated in this study. Based on these findings and the broader literature, four priorities warrant consideration: First, protected, legally recognized medical evacuation corridors for children with cancer are required under International Humanitarian Law (10) — evacuation delays have direct clinical consequences for critically ill patients. Second, systematic diagnostic quality assurance must be integrated into all cross-border referral programs, including standardized pathological review, tumor board confirmation, and telemedicine-based support (21–23).Third, pediatric oncology must be explicitly named and resourced within global humanitarian emergency health frameworks, including the WHO Global Initiative for Childhood Cancer (6), as an acute life-threatening emergency — not a chronic post-conflict issue. Fourth, philanthropic funding dependence must be replaced by predictable multilateral mechanisms through WHO, UNRWA, or equivalent structures to ensure continuity independent of donor cycles (6, 13).

Without urgent global action, conflict-affected settings risk reversing decades of progress in childhood cancer survival. For every child in this cohort, an unknown number did not reach care. Ensuring equitable access to timely diagnosis and treatment for children with cancer in conflict settings is a central imperative for global oncology — and one this evidence supports.

Limitations

The most fundamental limitation is an unquantifiable referral and survival bias: this cohort includes only patients who successfully navigated evacuation from Gaza to Jordan. Children who died before evacuation was possible, those for whom physical access, administrative clearance, or family capacity made evacuation unachievable, and those in whom cancer was never suspected or diagnosed represent a wholly invisible denominator. The findings therefore characterize outcomes in a selected subpopulation and cannot be generalized to the full scope of pediatric oncologic burden in Gaza. This selection bias is inherent to any conflict-evacuated cohort and should be considered when interpreting all reported outcomes, particularly survival estimates.

The retrospective design and reliance on electronic medical records introduce information bias regarding pre-referral treatment details, prior diagnostic workup, and symptom onset documentation. No pre-conflict cohort and no contemporaneous non-conflict comparison group are available, precluding causal inference regarding the specific contribution of the conflict to the observed clinical findings. The reported 1-year and 2-year OS and EFS estimates are derived from the observed follow-up period; confidence intervals are correspondingly wide, particularly in the PICU subgroup, and should be interpreted with caution pending follow-up maturation. No multivariable Cox regression was performed due to sample size constraints and diagnostic heterogeneity; the reported hazard ratio for PICU admission is univariable and unadjusted for disease stage, tumor type, and other clinical confounders. Caregiver and administrative burden data were collected through routine clinical documentation rather than systematic prospective assessment, likely underestimating their true prevalence. EFS should be interpreted in the context of the study-specific time origin and heterogeneous baseline disease status; patients presenting with relapse or progression were not counted as having an EFS event at baseline, and therefore these estimates are not directly comparable with conventional diagnosis-to-event survival estimates. As a single-center experience at a well-resourced referral center supported by an international partnership, generalizability to humanitarian oncology programs with more limited resources may be constrained.

Conclusion

Children evacuated from Gaza during active conflict presented to KHCC with a high burden of advanced or active disease, diagnostic discordance, nutritional deficiency, and psychosocial distress — findings consistent with the documented effects of healthcare system disruption in conflict-affected settings, though the observational design without a comparator group precludes definitive causal attribution. The marked difference in survival between PICU-admitted and non-PICU patients reflects the concentration of mortality risk in the most clinically burdened subgroup, with confounding by disease severity acknowledged. At the same time, this study demonstrates that access to comprehensive, multidisciplinary care at a regional center of excellence — sustained by committed international partnerships — is associated with meaningful and encouraging survival outcomes even in the context of profound clinical vulnerability. These outcomes, however, describe only those children who successfully navigated evacuation: for every patient in this cohort, an unknown number did not reach care. The true oncologic burden of this conflict on children remains incompletely characterized and is almost certainly substantially greater than our data reflect.

Urgent, coordinated global action is required: to establish protected medical evacuation corridors; to integrate pediatric oncology into humanitarian emergency response frameworks; to invest sustainably and predictably in regional centers of excellence (13); to strengthen diagnostic and treatment capacity in conflict-affected settings (9, 10). Children with cancer in conflict-affected settings cannot wait for peace to receive equitable access to life-saving care. Ensuring that they receive it — now, in Gaza, and in every future conflict — is not a charitable aspiration: it is an obligation of global solidarity, and one this evidence demands we meet.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. The authors used an artificial intelligence language tool (Claude, Anthropic) to assist with language editing and manuscript preparation. The authors take full responsibility for the integrity and accuracy of all content.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fonc.2026.1934075/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Ibtihal Fadhil, NCD Alliance Eastern Mediterranean, Kuwait

Reviewed by: Qi Zhang, Zhejiang Provincial People’s Hospital, China

Joe El Khoury, Centre Hospitalier Public d’Hauteville, France