Bone Marrow Involvement In Paediatric Non-Hodgkin Lymphoma: Comparing The Diagnostic Accuracy Of 18F-FDG PET/CT Versus That Of Bone Marrow Biopsy

Introduction: Non-Hodgkin lymphoma (NHL) is the fourth most common malignancy in paediatric population and is characterized by a high propensity to bone marrow involvement (BMI). Evaluation of BMI is an essential component of staging workup and accurate detection is critical since it indicates advanced stages and also affects the clinical management, treatment modalities and prognosis. It is an indicator of infusion-related reaction in patients with B cell NHL treated with rituximab. Bone marrow biopsy (BMB) although being the gold standard explores only a limited part of the bone marrow (unilateral or bilateral iliac crest) thereby having a small sample size and sampling error, thereby potentially missing focal bone marrow involvement. In addition to low sensitivity, BMB is invasive and causes pain to patients. Bleeding and infection are known complications although the risk is small. The clinical value of ¹⁸F-FDG PET/CT in assessing BMI in paediatric NHL is under debate still remains to be a timely question.

Aim: To evaluate the diagnostic accuracy of ¹⁸F-FDG PET/CT in detecting BMI compared to bone marrow biopsy (BMB) in newly diagnosed paediatric NHL patients.

The secondary aim of the study was to evaluate the prognostic value of BMI detected by PET/CT.

Methods: 55 consecutive, newly diagnosed and treatment naive paediatric NHL patients (68% boys) who underwent baseline staging ¹⁸F-FDG PET/CT and BMB were recruited in this study. Both the procedures were done within one week of one another. ¹⁸F-FDG uptake was evaluated qualitatively on a five point scale (0-4) where scores above and equal to 3 were taken positive for BMI. Histopathological confirmation for all PET/CT positive lesions was not carried out on ethical grounds. Results were compared by two blinded NM physicians and grouped as follows: True positive: 1) Focal uptake on scan matches the site of a positive BMB 2) No obvious CT changes in the bone indicating alternative underlying pathology 3) Follow up scan demonstrates disappearance of lesion post treatment suggestive of response to chemotherapy. False positive: Positive scan findings that do not fall under the above 3 criteria for TP. True negative: Negative scan for BMI matches negative BMB finding. False negative: Negative scan for BMI mismatched with positive BMB finding. The prognostic value of BMI demonstrated on ¹⁸F- FDG PET/CT was evaluated by overall survival (OS) and 3-year progression free survival (PFS).

Results: The sensitivity and specificity of ¹⁸F-FDG PET/CT in detecting BMI were 95%, and 100% respectively. BMB has a similar specificity however the sensitivity was 24%. PET/CT demonstrated a higher accuracy of 99% versus 81% using BMB. The negative predictive value of PET/CT scan was higher (98%) than BMB. BMI showed a strong correlation with PFS of 86% vs 97%; and OS of 83% vs 100% (p value<0.05 for both) as demonstrated for ¹⁸F-FDG PET/CT positive and negative BMI cases.

Conclusions: ¹⁸F-FDG PET-CT demonstrated higher diagnostic accuracy than BMB for evaluating BMI in paediatric NHL patients. BMB may be avoided in aggressive subtypes of NHL, wherein PET/CT scan have shown high sensitivity and specificity. In indolent NHL, however BMB has a higher diagnostic value owing to low sensitivity of PET/CT. We suggest that ¹⁸F-FDG PET/CT be used as a complementary tool to BMB for BMI evaluation in paediatric NHL population for accurate staging of the disease.