Discussion
This study highlights a high prevalence of musculoskeletal misdiagnoses among Danish children who were diagnosed with spinal tumors. In two-thirds of children with spinal tumors, musculoskeletal misdiagnoses were identified, resulting in a longer diagnostic interval compared to children without musculoskeletal misdiagnoses. This subgroup of patients frequently exhibited less aggressive tumors, rarely developed metastases or required additional therapy beyond surgery and tended to have a higher 5-year survival. These findings suggest that less aggressive tumors can gradually manifest as skeletal abnormalities and musculoskeletal symptoms without concurrent neurological or general symptoms, thereby leading to misdiagnosis.
Physicians who evaluate children with musculoskeletal pain must always consider the potential presence of an underlying tumor. Although it is widely acknowledged that pediatric cancers are frequently misdiagnosed, with rates ranging from 52-60%, musculoskeletal misdiagnoses have previously mainly been studied in children with hematological cancers.
Patients who were misdiagnosed with musculoskeletal conditions often presented with localized pain in the lower limb, back, or neck, accompanied by gait abnormalities and nocturnal pain. The literature states pain as the most prevalent symptom of spinal tumors, followed by motor weakness, sciatica, and sensory deficits. It is imperative to approach any occurrence of new-onset, persistent, localized, and severe pain in a previously asymptomatic child with utmost seriousness, warranting thorough consideration of an underlying pathological condition.
Despite medical advances, the diagnostic intervals are still very long for spinal tumors. Multiple studies have revealed a median diagnostic interval ranging from 2-8 months. In the presence of a misdiagnosis the diagnostic interval for pediatric cancers is twice as long. To the best of our knowledge, the present study is the first to investigate the impact of a musculoskeletal misdiagnosis on the time intervals of pediatric spinal tumors. Our results indicate that patients with a musculoskeletal misdiagnosis have a significantly longer median total interval of 5.5 months compared to 3 months for patients without a misdiagnosis, primarily due to a longer parental and primary care interval. In addition, the subgroup with a specific rheumatic misdiagnosis had an even longer first hospital doctor interval, often due to a delay after a visit to the emergency room. The median duration was four months for this subgroup compared to one week for the children without a musculoskeletal misdiagnosis (p=0.01).
Multiple studies have previously revealed associations that extend the diagnostic interval (age, symptoms, and tumor grade, among others). Koshimizu et al., conducted a study evaluating factors contributing to a delay in the diagnosis of pediatric spinal tumors. They discovered neurological symptoms as the presenting symptom was associated with early diagnosis, and the symptom of pain in the lower limb or back was associated with a longer duration of symptoms until diagnosis. They recommended that children experiencing pain lasting more than one month should undergo an MRI to exclude serious spinal disorders.
The present study found no association between longer time intervals and decreased survival. Patients with a musculoskeletal misdiagnosis had the highest 5-year survival rate. This is in accordance with most literature. Bouffet et al., found that patients with a shorter symptom duration (<2 months) had a poorer outcome than those with longer symptom duration (>2 months).
Previous investigations of pediatric CNS tumors have identified tumor type and grade as key factors influencing survival, rather than the diagnostic interval. Crawford et al., found that patients with high-grade malignant spinal tumors had a shorter duration of symptoms and significantly poorer survival compared to patients with low-grade tumors. In the present study, patients with a musculoskeletal misdiagnosis frequently had less aggressive tumors, rarely developed metastases or required additional therapy beyond surgery. Furthermore, these patients demonstrated a tendency towards higher 5-year survival rates. This may indicate that less aggressive tumors can provoke slowly developing skeletal abnormalities and musculoskeletal symptoms without neurological or general symptoms, causing misdiagnoses and diagnostic delays.
Survivors of spinal tumors experience a considerable disease burden due to long-term complications from both the tumor and it’s treatment, which markedly reduce their quality of life. In this study, sequelae were highly prevalent in all patients, occurring in two-thirds of the cases, consistent with literature, underscoring the significant impact spinal tumors implicate on patients. Notably, the patients with a musculoskeletal misdiagnosis did not demonstrate a higher prevalence or severity of sequelae compared to those without a misdiagnosis. This is consistent with some studies, while others have suggested that delayed diagnosis leads to decreased long-term quality of life.
The strengths of this study include the population-based setting including a non-selected cohort of all pediatric spinal tumor cases diagnosed in Denmark over a long period of 23 years. The study’s broad representation of tumor morphologies reflects the general population. The comprehensive data collection enabled a detailed analysis of the clinical course of pediatric spinal tumors, providing valuable insights into potential areas for improving the diagnostic process. Further, the evaluation of the diagnostic intervals was strengthened by using a standardized model with several subintervals, which increases the generalizability of the results.
This study has some limitations that require cautious interpretation. Firstly, pediatric spinal tumors are rare, leading to few cases and small subgroups, which may limit the statistical precision. Secondly, retrospective evaluation of the clinical presentation based on doctors’ notes from medical charts may introduce recall bias, particularly for time intervals and symptoms. Thirdly, a potential methodological limitation is the fact that the data does not include primary care and presumably underestimates the number of preliminary musculoskeletal misdiagnoses. Fourthly, the study did not record where in the health system misdiagnoses occurred, which precludes a definitive determination of a correlation between misdiagnosis and diagnostic delay. Lastly, 12 out of 70 (17%) patients were excluded due to insufficient or missing medical charts, which could affect our findings.