The role of routing in the diagnosis of acute leukemia in children: an observational retrospective non-randomized study

Full Text

BACKGROUND

The treatment outcomes of children with acute leukemia (AL) have improved significantly over the years, owing to several clinical trials [1, 2]. The use of risk-adapted specific therapy and improvements in accompanying therapy protocols have significantly reduced the relapse rate and treatment-related mortality in patients with AL [3]. Despite improvements in the treatment outcomes of children and young adults with AL, certain socioeconomic and healthcare organization issues and the type of healthcare facility providing this care can affect treatment outcomes [4–6].

In 2017, the World Health Organization published “Guidelines for the Early Detection of Cancer” to strengthen early detection programs worldwide. In this article, “delay” is classified into patient delay (from symptom onset to first visit to a physician, presentation interval) and healthcare delay which is further divided into diagnosis establishment delay (from initial visit to a physician to confirmed diagnosis, diagnostic interval) and treatment delay (from confirmed diagnosis to treatment initiation, therapeutic interval) [7]. The duration of each of the three stages of early diagnostics is ideally no more than 30 days for all types of malignancies in children and adults. Most studies revealed that the interval from symptom onset to chemotherapy initiation for children with AL is less than 60 days [8–13]. However, with a reduction in the total time to early diagnosis of AL to 30 days or less, relapse-free and overall survival of patients significantly increases [5, 13].

A factor influencing the delay in diagnostics at stage 2 is the distance from the patient’s place of residence to the healthcare facility with pediatric oncologists and hematologists who can provide specialized care [4, 13, 14]. The current study provides basis for cancer control programs that will systematically eliminate challenges in the timely provision of pediatric oncological care at each stage.

AIM

This study aimed to identify the correlation between the distance from the place of residence of children diagnosed with AL to the Children’s Regional Clinical Hospital (CRCH) of Tver Region (level 3 institution) and delay in establishing the diagnosis of the disease.

MATERIALS AND METHODS

Study design

An observational retrospective non-randomized study was conducted.

Eligibility criteria

The criteria for inclusion of patients in the study were age <18, confirmed AL diagnosis, and voluntary informed consent of parents/legal representatives for participation of patients in the study and publication of results.

The exclusion criteria was refusal to participate in the study.

Conditions and duration of the study

The analysis included 35 patients aged <18 years available for analysis and hospitalized in the oncohematology department of the CRCH between 2018 and 2023 with a confirmed diagnosis of AL.

Description of medical intervention

Acute lymphoblastic leukemia (ALL, C91.0) was diagnosed in 30 cases (86%), acute myeloid leukemia (C92) in 3 cases (9%), and leukemia of undetermined cell type (C95.0) in 2 cases (5%).

The initial blood analysis showed the following results:

• Thrombocytopenia and anemia at diagnosis in 76% of patients in group 1 and 78% in group 2.
• Platelet concentration: thrombocytopenia; 50–99×10⁹/L in 50% of patients, 20–49×10⁹/L in 37.5%, and <20×10⁹/L in 12.5%.
• Severe anemia (<70 g/L) in 47% of patients.
• Leukocyte concentration: leukocytosis (>20×10⁹/L) in 58% of patients, leukopenia (<3.5×10⁹/L) in 15%, and hyperleukocytosis (>100×10⁹/L) in 14%. The average concentration of leukocytes was 261.6×10⁹/L (126–470). In 27% of cases, the concentration of leukocytes in the peripheral blood was within the normal reference values, and in 97% of cases (n=34), blasts were detected in the peripheral blood of 2–95% of cases.

All patients were transferred to one of the Federal Children’s Oncology Centers immediately after the provisional diagnosis was made because performing a full diagnostics on site was impossible. The average transfer period was 4.2 (1–9) days. Bone marrow puncture was performed in 27 (77%) patients. Myelography was conducted if no blasts in the periphery in the presence of cytopenia were found in the clinical blood test and if the patient did not exhibit characteristics determining high risk and therapeutic urgency (e.g., bleeding grade III and higher, hyperleukocytosis, organ dysfunction). In the presence of risk factors, patients were urgently transferred to a specialized federal center.

The clinical presentation mainly showed hepatosplenomegaly (58%) and lymphadenopathy (54%). Infectious syndrome with fever necessitating antibacterial therapy was noted in 23% of patients. Hemorrhagic syndrome of grade III and higher, requiring transfusion of blood-derived products, was observed in 12% (n=4) of patients.

According to the distance from the place of residence to the CRCH, patients were distributed into the following groups:

• Group 1: Tver: “0” kilometers
• Group 2: Tver Region: <50 km, 51–99 km, and >100 km.

By the delay in diagnosis from the onset of the first symptoms, four groups were identified:

• Group 1: <2 weeks
• Group 2: 2–4 weeks
• Group 3: 4–8 weeks
• Group 4: >8 weeks

Subgroups analysis

The average age of patients was 61 months (5.1 years). The gender ratio was 17 boys (46%) and 18 girls (54%). In Tver (group 1) and Tver Region (group 2), 16 (46%) and 19 (54%) patients were identified, respectively. The average age of patients was 28.6 months in group 1 and 72.3 months in group 2 (p=0.1).

Ethics committee

The study was approved by the local ethics committee of the Children’s Regional Clinical Hospital of Tver Region on February 20, 2024 (no. 125284).

Statistical processing

Statistical data processing was performed using Microsoft Excel 2019 (Microsoft Inc., USA). In the variation series with the distribution, the Student’s t-test was utilized to identify the significant differences between the two groups. The differences were considered significant at p <0.05.

RESULTS

In groups 1 and 2, the following distribution of AL forms was revealed:

• Acute lymphoblastic leukemia in 14 (88%) and 16 (84%) patients (p=0.6)
• Acute myeloid leukemia in 1 (6%) and 2 (11%) patients (p=0.7)
• Leukemia of undetermined cell type in 1 (6%) and 1 (5%) patients (p=0.95)

The delay in diagnosis averaged 21 (6–62) and 18 (5–64) days in groups 1 and 2, respectively. In the overall group (n=35), a delay in diagnosis of less than 2 weeks was noted in 21 (60%) cases, of 2–4 weeks in 7 (20%) cases, of 4–8 weeks in 4 (11%) cases, and of more than 8 weeks in 3 (9%) cases.

When comparing the delay in diagnostics indicator between groups 1 and 2, the following results were obtained:

• Less than 2 weeks: 7 (44%) versus 13 (68%) cases (p=0.15)
• 2–4 weeks: 6 (38%) versus 3 (17%) cases (p=0.16)
• 4–8 weeks: 1 (6%) versus 1 (5%) case
• More than 8 weeks: 2 (12%) versus 2 (10%) cases

The difference between AL patients in groups 1 and 2 was not significant (p=0.37).

Among patients in group 2, the distance of the place of residence from the children’s institution level 3 providing specialized care (CRCH) had no significant effect on the time of diagnosis. When patients were located <50 km from the CRCH, the delay in diagnostics was the following:

• <2 weeks (36% of patients)
• 2–4 weeks (36% of patients)
• 4–8 weeks (21% of patients)
• >8 weeks (7% of patients)

With a distance of 50–100 km, diagnosis was made within 2–4 weeks in 100% of cases.

The delay in diagnostics when patients lived >100 km from the CRCH was as follows:

• Less than 2 weeks in 30%
• 2–4 weeks in 30%
• 4–8 weeks in 20%
• More than 8 weeks in 20% (p=0.78)

DISCUSSION

Delay in diagnosis and treatment has been described in various studies with different results depending on the type of cancer studied. AL is diagnosed quickly compared with other types of cancer in children [5, 15]. However, the results may vary significantly depending on the country income level and healthcare organization system.

Most studies have noted a negative impact of delay from the onset of the first symptoms to the time of diagnosis on overall and relapse-free survival and early mortality during therapy [4–6, 12, 13]. In a group of 166 pediatric patients with acute lymphoblastic leukemia, a significant improvement in survival was observed in children whose overall treatment delay was <30 days, compared with those whose overall treatment delay was 30–120 days (86.4 versus 60%; p=0.02). Moreover, the average time interval from the onset of symptoms to the start of chemotherapy (overall delay) was 53.5 (38–93.5) days. Only 13.3% of patients started treatment within 30 days after the onset of symptoms [13]. Gardie et al. reported that acute lymphoblastic leukemia was diagnosed within 30 days in only 66% of children studied, which had a negative impact on the therapy results [15]. A similar negative impact was revealed in a Chinese study [6].

In the current study, the interval from the onset of the first symptoms to the time of diagnosis was 19 (5–64) days, and an excess of the diagnostic interval by more than 30 days was found in only 20% of patients. Thus, our data positively correlate with studies conducted in Brazil, Mexico, and Nicaragua, where the median of the overall delay is 30–35 days, and with data from Japan and China (median: 20 and 21 days, respectively) [6, 9, 11, 16, 17].

In most cases, an increase in the distance from the place of residence to a medical institution capable of making a diagnosis negatively affects the results of therapy and is directly related to an extended diagnostic interval. In a study in Greece, higher mortality rate was reported among children living 50 km or more from a treatment center compared with those living within 50 km (mortality ratio: 1.77; 95% confidence interval (CI): 0.93–3.37; n=293) [18]. Youlden et al. revealed higher mortality rates among children with leukemia living in remote areas compared with those in major cities in Australia (relative risk: 1.52; 95% CI: 1.11–2.08; n=6,289) [19]. In a recent analysis of pediatric and young adult (up to 39 years) patients in the North American National Cancer Database, Rotz et al. showed higher survival among those living within 50 miles (80 km) of the cancer center compared with those living more than 50 miles away from it (83 versus 77%; p <0.001) [4]. In a univariate analysis of 288 children with AL, late diagnostics was noted in 45% of cases when living 100 km or less away from the hospital and in 55% of cases when the distance from the hospital was more than 100 km. The relative risk of a delay in diagnosis of more than 30 days was 1.62 (95% CI: 1.01–2.58; p=0.04) [9].

In contrast, not all studies noted the association between the distance of residence and outcome of AL therapy. Janitz et al. did not observe differences between distance-to-care categories and relapse-free survival in children and adolescents aged ≤20 years with acute lymphoblastic leukemia (n=275). However, in this study, an increased risk of relapse was noted when the patient lived more than 121 km (75 miles) from the treatment center [20].

The present study did not detect an effect of the distance from the patient’s place of residence to the medical facility level 3 on the delay in diagnosis of AL in children. Among children living 100 km or more from the CRCH, the average delay in diagnosis was 20 [6–62] days compared to 19 [5–64] days in those living less than 100 km from it (p=0.97). Interestingly, among patients living in Tver, diagnosis was made within 30 days in 72% of patients, whereas among those in Tver Region, the diagnosis was made within this period in 85% of patients. The average delay in diagnosis in Tver and Tver region was 21 and 18 days, respectively (p=0.58). The higher frequency of early (<4 weeks) detection of AL in the districts of Tver Region compared to that in the city can be elucidated by the fact that doctors in the districts, when faced with abnormal blood test results or an unexplained deterioration in the patient’s health, attempt to immediately transfer the patient to the CRCH, where there is a pediatric oncologist. In contrast, city patients are first examined at the city clinic; then, they are often hospitalized in level 2 institutions and only then transferred to the CRCH.

Study limitations

First, considering the child population of Tver Region, the average incidence of all types of AL was 3 per 100 thousand individuals, which is slightly lower than the average for children aged <18 years [21]. Thus, it can be assumed that some patients living in the border of the Moscow and Leningrad regions visit directly the federal pediatric oncology institutions when pathological symptoms appear, bypassing the CRCH. Second, the distance to the level 3 medical facility is geocoded from the postcode level of the patient’s actual residential address and does not take into account how long it took the patient to reach the primary care medical facility or whether the patient was elsewhere at the time of disease onset.

ADDITIONAL INFORMATION

Funding source. This study was not supported by any external sources of funding.

Competing interests. The authors declare that they have no competing interests.

Authors’ contribution. All authors made a substantial contribution to the conception of the work, acquisition, analysis, interpretation of data for the work, drafting and revising the work, final approval of the version to be published and agree to be accountable for all aspects of the work. Dolgopolov IS — development of research design, data collection, analysis of literary sources, text editing; Rykov MYu — development of research design, data collection, analysis of literary sources, editing and writing of text.

About the authors

Igor S. Dolgopolov

Regional Children’s Clinical Hospital; Tver State Medical University

Email: irdolgopolov@gmail.com
ORCID iD: 0000-0001-9777-1220
SPIN-code: 4312-9786

MD, Dr. Sci. (Medicine)

Russian Federation, Tver, 170100; Sovetskaya st. 4, Tver 170100

Maksim Yu. Rykov

Russian State Social University

Author for correspondence.
Email: wordex2006@rambler.ru
ORCID iD: 0000-0002-8398-7001
SPIN-code: 7652-0122

MD, Dr. Sci. (Medicine), Associate Professor

Russian Federation, 4 build. 1 Wilhelm Peak st., Moscow 129226

References

Supplementary files