We collected data from electronic patient files. Two different researchers (E.A.M.Z. and L.A.W.) collected data independently. Uncertainties were discussed together and, if necessary, with the other researchers (A.B., R.J.N.) to reach a consensus. Due to the retrospective design, no standard formats were used for the description of the investigated variables. For example, involvement of the body regions was scored based on the radiology reports. Sometimes, one of the investigators was not certain how to interpret the radiology reports. This was discussed with the colleagues, and if necessary, the radiology investigations were re-analyzed by the involved radiologist of our study.
Data were processed anonymously and encrypted.
We identified predicting factors for lymphoma based on an extensive search of the literature using PubMed, Medline, and Embase. We searched for studies using Medical Subject Heading terms including “lymphadenopathy”, “child”, “adolescent”, and “lymphoma”. An overview of potential predicting factors based on this search of the literature and their results are given inTable S1 [8 (link),31 (link),32 (link),33 (link),34 (link),35 (link),36 (link),37 (link),38 (link),46 ,47 (link),48 (link),49 (link),50 (link),51 (link),52 (link)]. We identified 39 potential predictors and included these in our univariate analyses: age, gender, presence of B-symptoms, 11 laboratory parameters including TARC, and several imaging findings. These variables and their definitions are listed in Table S2 .
The body regions of the involved areas were scored individually. An overview of the separately scored anatomical body regions and an explanation is provided inTable S3 .
We used pathology reports primarily for defining the diagnosis; 158 out of 182 patients underwent biopsy, including all cases of lymphoma. Twenty-four patients were diagnosed without a biopsy, but based on clinical, radiological, microbiological, and laboratory results (twenty infectious/reactive lymphadenopathy, one venous malformation, one lymphangioma, one branchiogenic cyst, and one dermoid cyst).
We categorized the patients into 12 groups according to their diagnosis. The malignant diagnoses in the study population included: cHL, NLPHL, ALCL, primary mediastinal large B-cell lymphoma (PMBCL), diffuse large B-cell lymphoma (DLBCL), Burkitt lymphoma (BL), T-LBL, B-cell lymphoblastic lymphoma (B-LBL), and other malignancies (Langerhans cell histiocytosis (LCH)). Furthermore, there were three groups with benign causes of lymphadenopathy: reactive or infectious lymphadenopathy, progressive transformation of germinal centers (PTGC), and other non-malignant causes.
For the identification of predictive factors, we divided the outcome into the benign group and the malignant group for univariate analysis. However, the malignant group contained nine different diagnoses, which differ significantly in incidence and clinical presentation. Therefore, we subdivided the group into five categories for multivariate analysis: cHL, NLPHL, NHL, other malignancies, and the benign group. In brief, we collected data from electronic patient files. We identified 39 potential predictors based on an extensive search of the literature. We used pathology reports for defining the diagnosis.
Data were processed anonymously and encrypted.
We identified predicting factors for lymphoma based on an extensive search of the literature using PubMed, Medline, and Embase. We searched for studies using Medical Subject Heading terms including “lymphadenopathy”, “child”, “adolescent”, and “lymphoma”. An overview of potential predicting factors based on this search of the literature and their results are given in
The body regions of the involved areas were scored individually. An overview of the separately scored anatomical body regions and an explanation is provided in
We used pathology reports primarily for defining the diagnosis; 158 out of 182 patients underwent biopsy, including all cases of lymphoma. Twenty-four patients were diagnosed without a biopsy, but based on clinical, radiological, microbiological, and laboratory results (twenty infectious/reactive lymphadenopathy, one venous malformation, one lymphangioma, one branchiogenic cyst, and one dermoid cyst).
We categorized the patients into 12 groups according to their diagnosis. The malignant diagnoses in the study population included: cHL, NLPHL, ALCL, primary mediastinal large B-cell lymphoma (PMBCL), diffuse large B-cell lymphoma (DLBCL), Burkitt lymphoma (BL), T-LBL, B-cell lymphoblastic lymphoma (B-LBL), and other malignancies (Langerhans cell histiocytosis (LCH)). Furthermore, there were three groups with benign causes of lymphadenopathy: reactive or infectious lymphadenopathy, progressive transformation of germinal centers (PTGC), and other non-malignant causes.
For the identification of predictive factors, we divided the outcome into the benign group and the malignant group for univariate analysis. However, the malignant group contained nine different diagnoses, which differ significantly in incidence and clinical presentation. Therefore, we subdivided the group into five categories for multivariate analysis: cHL, NLPHL, NHL, other malignancies, and the benign group. In brief, we collected data from electronic patient files. We identified 39 potential predictors based on an extensive search of the literature. We used pathology reports for defining the diagnosis.
Free full text: Click here
