Identification of Malignancy-Associated Haemophagocytic Lymphohistiocytosis By 18F-FDG PET/CT Based Multivariate Models in Clinical Practice

doi:10.21203/rs.3.rs-482576/v1

Download PDF

Research Article

Identification of Malignancy-Associated Haemophagocytic Lymphohistiocytosis By 18F-FDG PET/CT Based Multivariate Models in Clinical Practice

https://doi.org/10.21203/rs.3.rs-482576/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Purpose: ¹⁸F-FDG PET/CT is often used for detection of malignancy in newly diagnosed haemophagocytic lymphohistiocytosis (HLH), with acceptable sensitivity but relatively low specificity. The aim of this study was to improve the diagnostic ability of ¹⁸F-FDG PET/CT to identify malignancy in HLH by a combination of ¹⁸F-FDG PET/CT and clinical parameters.

Methods: A cohort of patients with secondary HLH (n=71, age≥14 years) was retrospectively reviewed and pre-therapeutic variables of ¹⁸F-FDG PET/CT and laboratory were collected. Patients were divided into two groups: malignancy-associated HLH (M-HLH) (n=22) and non-malignancy-associated HLH (NM-HLH) (n=49). The variables were analyzed by Mann-Whitney U or Pearson Chi-Square test and the nomograms were constructed by multivariate binary logistic regression. The predictors were also ranked by decision-tree analysis.

Results: Asymmetry distribution of hypermetabolic lymph nodes was more commonly observed in M-HLH (p=0.002). SUVmax-bone lesions, SUVmax-lymph nodes and C-reactive protein were selected to construct the model. This model displayed a good predictive performance with the area under receiver-operating-characteristic (ROC) curve of 0.884 (95%CI: 0.791-0.978). At an appropriate cut-off value, the sensitivity and specificity of identifying M-HLH reached 86.4% (65.1%-97.1%) and 87.2% (74.3%-95.2%), respectively. The decision tree integrating SUVmax-bone lesions and SUVmax-lymph nodes achieved a high sensitivity (86.4%) and specificity (87.8%) for identification of malignancy.

Conclusion: ¹⁸F-FDG PET/CT can be a practical technique for identifying M-HLH and the models constructed by ¹⁸F-FDG PET/CT features alone or combined with clinical parameters are able to detect malignancy with better accuracy than visual analysis of ¹⁸F-FDG PET/CT images.

Nuclear Medicine & Medical Imaging

hemophagocytic lymphohistiocytosis

malignancy

PET/CT

multivariate prediction model

Hemophagocytic lymphohistiocytosis (HLH) is a rare but life-threatening hyperinflammatory syndrome, characterized by persistent activation of cytotoxic T cells, natural killer (NK) cells and macrophages [1]. This disease can be classified into primary (genetic) and secondary (reactive) HLH. Primary HLH generally has genetic mutation or familial inheritance with an incidence of 1/50,000 to 1/100,000 live-born children [2]. In contrast, secondary HLH occurs in older children and adults but it is more frequently presented in adults. Secondary HLH consists of two main types: malignancy-associated HLH (M-HLH) and non-malignancy-associated HLH (NM-HLH). Currently, the precise incidence of HLH in adults is still unknown, with an estimated annual incidence of approximately 1/800,000 [3]. The triggers linked with secondary HLH are extensive, such as malignancy, infection, autoimmune disease, metabolic disease and acquired immune deficiency. In particular, malignancy and infection are the most popular triggers for secondary HLH. Emerging data showed that M-HLH accounts for about 40-70% of secondary HLH in adults and has the worst prognosis with a mortality rate of greater than 80% and a median survival of less than 2.0 months [2, 4]. The vast majority of malignancy triggers are hematological malignancies including lymphomas, T/NK-cell disorders, acute leukemias, lymphoproliferative diseases and myelodysplastic syndrome. It was noted that the prevalence of M-HLH is 0.9% in adults with hematological malignancies, but could be 20% in patients with specific rare types of lymphoma [2]. Patients with HLH have non-specific manifestations, including intermittent fever, hemocytopenia, hepatosplenomegaly, lymphadenopathy, liver injury and a sepsis-like syndrome that may rapidly progress to terminal multiple organ failure [5]. The diagnosis of M-HLH is considered as a big challenge to clinicians as a result of variable overlaps of symptoms with other types of HLH, sepsis and multiorgan failure, thus leading to greater misdiagnosis and poorer outcomes. Treatment must be tailored to the founding trigger, patient performance, organ functions and concomitant therapies, since HLH is a heterogeneous disease. Traditional therapy includes suppression of the overactive immune system and induction therapy followed by allogeneic stem cell transplant. Particularly, M-HLH requires specific treatment and intensive care. Promptly identifying the underlying trigger is essential to initiate appropriate therapy and reduce relapse in patients with M-HLH.

As addressed above, accurate identification of malignancy in the context of HLH is extremely difficult. Bone marrow assessment is performed in all patients with HLH, but negative does not exclude malignancy. Additionally, immune system is unable to adequately restrict stimulatory effects of various triggers, so levels of laboratory parameters may be influenced in M-HLH [4, 6-9]. For example, the increases of LDH (lactate dehydrogenase), beta-2-microglobulin and sCD25/ferritin ratio have been reported in M-HLH [6-8] while the levels of fibrinogen and platelet count might be lower [4, 7]. Although laboratory parameters are pathognomonic, none of them is sufficient for malignancy diagnosis. ¹⁸F-FDG PET/CT is a noninvasive whole-body functional imaging approach and has been advised to reveal malignancy [10]. A few studies have indicated that ¹⁸F-FDG PET/CT helps detecting malignancy and guiding biopsy [10-13], but the imaging findings are unspecific for HLH. In general, it shows hepatosplenomegaly and diffusely increased FDG uptake in liver, spleen, axial skeleton, bone marrow and multiple lymph nodes. These findings are partially due to hyperinflammatory state of HLH and have a high rate of false-positive in detecting malignancy. Our previous study highlighted that ¹⁸F-FDG PET/CT alone is insufficient to correctly make a differential diagnosis, though it is helpful to identify the potential trigger for secondary HLH [10]. Therefore, efforts focusing on developing novel strategy in order to improve our ability to rapidly and accurately identify M-HLH are underway. This study was designed to evaluate the diagnostic performance of ¹⁸F-FDG-PET/CT in secondary HLH with suspected malignancy and to develop novel predictive models for identifying M-HLH by incorporation of PET/CT with clinical variables.

Patients

The records of patients with HLH were reviewed retrospectively from August 2016 to June 2020 in our department of Beijing Friendship Hospital affiliated to Capital Medical University. The diagnostic criteria of HLH and hematological malignancy were in accordance with HLH-2004 protocol [14] and 2016 World Health Organization criteria [15]. Patients aged <14 years or treated with chemotherapy or granulocyte colony-stimulating factor before ¹⁸F-FDG PET/CT scan were excluded. Consequently, a total of 71 patients were included in this study. All cases had pathology tests and 22 of them were pathologically diagnosed as malignancy-associated HLH. Patients were divided into two groups: malignancy-associated HLH (M-HLH) (n=22) and non-malignancy associated HLH (NM-HLH) (n=49) (Fig. 1). This retrospective study was approved by Institutional Review Board of Beijing Friendship Hospital of Capital Medical University and the requirement of written informed consent was waived.

Clinical Data Collection

A standard form was used to record clinical parameters, including age, gender and clinical laboratory features. All data were collected before initial HLH-specific therapy and within 7 days before or after ¹⁸F-FDG PET/CT scan.

¹⁸F-FDG PET/CT Image Acquisition and analysis

All ¹⁸F-FDG PET/CT images were acquired from basis cranii to upper thigh on Siemens biograph mCT PET/CT scanner (Siemens Healthineers). Patients were instructed to fast for at least 6 h and blood glucose levels were confirmed to be less than 11.1mmol/L. Approximately ¹⁸F-FDG (4.4MBq/kg) was injected intravenously before PET/CT scan. After a 60-minute uptake time, low-dose CT scan was performed first to visualize anatomic structures and attenuation correction with 140keV, automatic mAs and a slice thickness of 3mm. The whole-body PET scan was carried out with 2.5 min per bed position using three-dimensional mode immediately after a whole-body CT scan. Images were reconstructed with an iterative reconstruction algorithm.

Two experienced nuclear medicine physicians reviewed PET/CT images carefully on a dedicated workstation (syngo Multimodality Workstation, Siemens Healthineers). Hypermetabolic lymph nodes and their distributions were recorded. Maximal standardized uptake value (SUVmax) and the short and long axis of the largest node were also documented. In bone marrow, focal increased FDG uptake and SUVmax of bone lesions were recorded. It was excluded that hypermetabolic foci were caused by other factors (such as bone marrow puncture or degeneration) in bone marrow. Meanwhile, it was ruled out if lymph nodes distributed symmetrically in hilar which were considered as reactive lymphoid hyperplasia. An ellipsoidal region of interest [16] was drawn in lumbar vertebrae 4 (L4) to measure SUVmax of bone marrow background (SUVmax-bone background). If SUVmax-bone background could not be measured because of inhomogeneously and diffusely increased FDG uptake in bone marrow, these cases would be excluded. When the focal bone lesion located in L4, the adjacent lumbar vertebrae L3 was measured. In liver and spleen, focal increased FDG uptake, SUVmax, volume and total lesion glycolysis (TLG) were documented. The SUVmax, volume and TLG were measured using 3D SlicerTM software (14) (version 4.10.0, http://www.slicer.org). The TLG of liver and spleen was calculated by multiplying SUVmean by total volume.

Statistical Analysis

Statistical analyses were performed by SPSS Statistical Software version 26.0 (IBM Corp., Armonk, NY), MedCalc Statistical Software version 19.0.7 (MedCalc Software bvba, Ostend, Belgium) and R version 3.6.1 (http://www.r-project.org). Continuous variables are presented as medians with interquartile ranges. Categorical variables are presented as frequencies and percentage. Skewed continuous variables were tested by Mann-Whitney U test and categorical variables were tested using Pearson Chi-Square (χ²) test, in order to compare clinical and PET/CT parameters between two groups. All P values were two-sided, with a significance level of less than 0.05. Multivariate logistic regression with stepwise method was served to select independent predictors. The predictive models were constructed by the logistic equations and were simplified into nomogram. The selected predictors were utilized for decision-tree analysis. The diagnostic ability of variables and predictive models were determined by receiver-operating-characteristic curve analysis and cutoff values were estimated. Decision-curve analysis (DCA) was used to assess the clinical usefulness of the models.

Characteristics of patients

The characteristics of 71 patients with secondary HLH are shown in Table 1. Their median age was 38 years (range, 14-79 years) and 37 (52.1%) were male. Among them, 22 (31.0%) were diagnosed as M-HLH. Eleven of M-HLH were detected with B cell neoplasms including 6 diffuse large B-cell lymphoma, 2 classical Hodgkin lymphoma, 1 follicular lymphoma, 1 non-Hodgkin B cell lymphoma and 1 B-cell acute lymphocytic leukemia and 4 of them had EBV infection simultaneously. The rest of M-HLH were identified with T/NK cell neoplasms including 6 NK/T-cell lymphoma, 2 peripheral T-cell lymphoma, 2 unspecified T-cell lymphoma and 1 anaplastic large cell lymphoma and 7 of them had EBV infection concurrently. In 49 (69.0%) of NM-HLH, there were 24 EBV infection, 12 other infections, 5 adult Still’s disease, 1 systemic lupus erythematosus, 1 undifferentiated systemic rheumatic disease and 6 with unspecific underlying causes.

Comparison of baseline ¹⁸F-FDG PET/CT findings and laboratory data between M-HLH and NM-HLH

The comparisons of ¹⁸F-FDG PET/CT and laboratory features were listed in Table 2 and 3. Regarding gender and age, it seemed that M-HLH occurred more commonly in males and older adults. The percentage of males and the median age in the M-HLH type were larger than those in the NM-HLH type [72.7% vs. 42.9%, p=0.020; 49 (32-58) vs. 33 (25-52), p=0.041]. Most of PET/CT variables had a significant difference between the two groups. For instance, the asymmetrical distribution of hypermetabolic lymph nodes (p=0.002), bone lesion (p<0.001) and focal liver lesion (p=0.015) were more frequently presented in the M-HLH group. Moreover, the patients with M-HLH tended to have a greater value in the SUVmax of lymph nodes and the SUVmax, volume, TLG of spleen and liver compared with the NM-HLH (p<0.05). Similarly, the SUVmax of bone lesions in the M-HLH was significantly higher than in the NM-HLH at a level of p<0.1. There were no differences between the groups in terms of hypermetabolic lymph nodes, the long and short diameter of the lymph node with the highest FDG uptake and the SUVmax of bone marrow background. As for the laboratory parameters in the cases with M-HLH, white blood cell, absolute neutrophil and platelet counts were lower, whilst the levels of C-reactive protein, lactate dehydrogenase, soluble CD25, beta-2-microglobulin, IL-8 and IL-10 were higher in comparison with the NM-HLH category (p<0.05).

Diagnostic performance of visual analysis of ¹⁸F-FDG PET/CT images

Patients were considered “positive” if ¹⁸F-FDG PET/CT images showed hypermetabolic lymph nodes and/or focal increased FDG uptake in bone, liver and/or spleen. This approach was applied to diagnose M-HLH with great sensitivity (90.9%, 20/22) but poor specificity (36.7%, 18/49) and its accuracy was only 53.5%. The false-negative PET/CT images of two M-HLH were shown in Fig. 2a and b. It was noteworthy that multifocal of increased FDG uptake in bone marrow was usual in patients with EBV infection (positive predictive value (PPV) for malignancy: 53.3%, 8/15) (Fig. 2c and d), but was a malignant sign in cases without EBV infection (PPV: 87.5%, 7/8). Only one patient without malignancy and EBV infection had focal hypermetabolic bone lesions (Figure 2e). She had cytomegalovirus (CMV) infection, and the SUVmax of bone lesions was lowest at 3.9.

In this study, focal spleen lesions were observed in 13 patients and the potential causes were malignancy (7/13), EBV infection (5/13) and unknown (1/13). Notably, there were 6 cases with both malignancy and EBV (2 classical Hodgkin lymphoma, 2 NK/T-cell lymphoma, 1 diffuse large B-cell lymphoma and 1 follicular lymphoma) and 1 case with malignancy alone (peripheral T-cell lymphoma). Focal liver lesions were discovered in 6 patients and the possible causes were malignancy (5/6) and EBV infection (1/6). There were 2 cases with both malignancy and EBV (1 classical Hodgkin lymphoma and 1 NK/T-cell lymphoma) and 3 cases with malignancy alone (2 diffuse large B-cell lymphoma and 1 peripheral T-cell lymphoma).

Multivariate logistic regression models

¹⁸F-FDG PET/CT and laboratory parameters were analyzed separately and conjointly by multivariate logistic regression. Lastly, two PET/CT (SUVmax of hypermetabolic lymph nodes and SUVmax of bone lesions) and three laboratory variables (lactate dehydrogenase, C-reactive protein and sCD25) were selected to build the predictive models. The Odd ratios and p values of these variables were shown in Table 4. Two best logistic models were picked up to construct nomograms. One nomogram was only based on two PET/CT parameters (SUVmax of hypermetabolic lymph nodes and SUVmax of bone lesions) and the other included both types of features (SUVmax of hypermetabolic lymph nodes, SUVmax of bone lesions and C-reactive protein) (Fig. 3).

The diagnostic performance of each parameter was shown in Supplemental Table 1 (p<0.05). Although none of these parameters was individually suited to predict malignancy, there were interesting data shown in Table 5. The multivariate logistic model constructed by SUVmax-bone lesions and SUVmax-lymph nodes displayed acceptable diagnostic accuracy with an area under the ROC curve (AUC) of 0.867 (0.766-0.936) (Fig. 4a and c). At a cut-off of 27.0%, its sensitivity and specificity were 90.9% (70.8-98.9) and 81.6% (68.0-91.2), whilst the positive and negative predictive value were 69.0% (54.8-80.3) and 95.2% (84.1-98.7), respectively. Remarkably, the multivariate logistic model built by SUVmax-bone lesions, SUVmax-lymph nodes and C-reactive protein had the largest AUC (0.884, 95%CI: 0.791-0.978) (Fig. 4b and c). At a cut-off of 32.4%, its sensitivity and specificity were 86.4 (95%CI: 65.1-97.1) and 87.2% (74.3-95.2), while the positive and negative predictive value were 76.0% (59.6-87.2) and 93.2% (82.6-97.5) (Fig. 4b and c). The decision-curve analysis showed that those two models had greater clinical utility for prediction of malignancy than models established by laboratory variables alone (Fig. 4d).

Decision tree for diagnosing M-HLH

The decision tree showed that the model constructed by two PET/CT features alone (SUVmax-bone lesions and SUVmax-lymph nodes) had a predictive ability to identify malignancy, with a sensitivity of 86.4% (19/22), a specificity of 87.8% (43/49) and an accuracy of 87.3% (62/71) (Fig. 5 and Table 5). Eighty percent (12/15) of the patients with SUVmax-bone lesions≥7.0 and 70% (7/10) of the patients with SUVmax-bone lesions<7.0 and SUVmax-lymph nodes≥12.7 were M-HLH, while those with SUVmax-bone lesions<7.0 and SUVmax-lymph nodes<12.7 were almost NM-HLH (43/46, 93.5%).

In recent years, ¹⁸F-FDG PET/CT has been suggested to detect a potential trigger for secondary HLH [5, 17]. Because of cytokine storm, reticuloendothelial system activation and myeloproliferation, ¹⁸F-FDG PET/CT images nonspecifically show diffusely increased FDG uptake in liver, spleen and axial skeleton [18-20]. Hence, it is necessary to improve diagnostic specificity of ¹⁸F-FDG PET/CT for prompt and precise detection of M-HLH. In the present study, ¹⁸F-FDG PET/CT had a power to assist identifying malignancy in secondary HLH and the multivariate model was established to predict malignancy by integration of ¹⁸F-FDG PET/CT and laboratory variables (SUVmax-bone lesions, SUVmax-lymph nodes and C-reactive protein). The predictive models and nomograms built by these parameters were able to distinguish M-HLH and NM-HLH before treatment, with the better performance than visual analysis of images. The decision tree depending on ¹⁸F-FDG PET/CT features alone (SUVmax-bone lesions and SUVmax-lymph nodes) can predict M-HLH with the greater diagnostic accuracy.

The diagnostic performance of visual analysis of ¹⁸F-FDG PET/CT images for detection of malignancy in HLH was shown with acceptable sensitivity but relatively low specificity. One study reported that the sensitivity, specificity and diagnostic accuracy of ¹⁸F-FDG PET/CT for detecting M-HLH were 83.0, 62.5 and 71.4%, sequentially, in a small sample size (n=14) [13]. In addition, several studies suggested that focal FDG uptake was a predictive sign for an underlying malignancy [11, 12, 21]. However, it was found that extranodal NK/T cell lymphoma nasal type, aggressive NK cell leukemia, B-cell lymphoma and T-cell lymphoma occasionally lack focal FDG uptake lesions [11, 13, 20]. In our study, if considering focal FDG uptake in liver, spleen, bone marrow and lymph nodes as positive, the accuracy of M-HLH diagnosis was only around 50%, with high sensitivity (90.9%) but poor specificity (36.7%). The two false-negative patients without focal hypermetabolic lesion were pathologically diagnosed as non-Hodgkin B cell lymphoma and T-cell lymphoma using bone marrow biopsy. As is well known, EBV infection is the most likely cause of focal FDG uptake lesions in bone marrow. Multiple hypermetabolic lymph nodes and/or patchy bone lesions were seen in several cases of EBV associated HLH [13, 22-24]. If without EBV infection, focal FDG uptake lesions in bone marrow are a reliable sign for diagnosing malignancy in secondary HLH. In agreement with these, our results showed that the PPV of multifocal FDG uptake in bone marrow was 53.3% in the cases with EBV infection, and was 87.5% in the cases without EBV infection. Only one NM-HLH case without EBV had focal FDG uptake bone lesions, and the SUVmax-bone lesions was low which might be caused by heterogeneous active proliferation of bone marrow cells. Although multiple lymphadenopathies with increased FDG uptake occur more frequently in lymphoma associated HLH than non-lymphoma associated HLH [12], infections and autoimmune diseases often present hypermetabolic lymph nodes. Our data suggested that there was no significant difference in the incidence of hypermetabolic lymph nodes between the two groups, but the asymmetry distribution of hypermetabolic lymph nodes could aid to indicate malignancy. Focal splenic FDG uptake is correlated with EBV infection, lymphoma, angiosarcoma, hamartoma[25], visceral leishmaniasis[26], spleen tuberculosis[27], abdominal sarcoidosis[28] and reactive lymphoid hyperplasia[29]. Our results indicated that focal splenic FDG uptake was closely linked with EBV infection, and focal hepatic FDG uptake was relatively rare and also associated with malignancy and EBV infection in HLH. Therefore, both focal splenic FDG uptake and focal hepatic FDG uptake cannot indicate malignancy. Additionally, M-HLH had higher SUVmax of bone lesions, lymph nodes, spleen and liver than NM-HLH in consistence with previous studies [12, 30], but there is no report about the volume and TLG of liver and spleen so far. The present study showed that the volume and TLG of liver and spleen in M-HLH were larger than NM-HLH.

Apart from PET/CT, laboratory variables could contribute to differentiate M-HLH from NM-HLH. It has been found that patients with lymphoma-associated HLH have lower levels of fibrinogen and platelet count and/or higher levels of lactate dehydrogenase and beta 2 microglobulin than NM-HLH [4, 7, 8]. Cytokine profiles vary in HLH with different underlying etiologies [31, 32]. INF-γ and IL-10 are markedly elevated in HLH in comparison with sepsis [33], while IL-18 is starkly raised in macrophage activation syndrome (autoimmune disease associated HLH) [34]. Although PET/CT and laboratory parameters are helpful in the detection of malignancy in HLH, none of these variables has adequate efficiency to identify M-HLH individually. In this study, two predictive models for malignancy were developed to improve PET/CT diagnostic ability. One model only included ¹⁸F-FDG PET/CT parameters and the other integrated PET/CT and laboratory variables. Using the nomogram and appropriate cutoff, both of them achieved better diagnostic accuracy with 84.5% and 87.0%, respectively. Decision-curve analysis found that both models could be practical to predict malignant risk ranging from 0.1 to 0.8. SUVmax-bone lesions, SUVmax-lymph nodes and C-reactive protein were three critical predictors for M-HLH, since FDG accumulation should be higher in malignant cells. It is reasonable to infer that highly FDG uptake of focal bone lesions are likely skeletal involvement of malignancies, and highly FDG uptake of lymph nodes are likely due to lymphoma rather than reactive lymphoid hyperplasia. C-reactive protein plays both pro- and anti- inflammatory role as a prototypical acute phase serum protein[35], and is elevated in many malignancies[36]. Importantly, C-reactive protein is served as a good indicator for the usefulness of ¹⁸F-FDG PET/CT in HLH [11]. The nomogram built by these three parameters had the best performance for predicting M-HLH with high sensitivity and specificity.

Decision tree was executed to make it easy to use in clinical setting. First of all, if focal FDG-avid bone lesions and the SUVmax ≥7.0 were observed, it was very likely that the patient had M-HLH. If not, SUVmax-lymph nodes were utilized. When its value was ≥12.7, the possibility of M-HLH was very high. The rest of patients would be considered as non-malignancy. This model was simple and intuitive, with the predictive accuracy of 87.3%. The false positives might occur in patients with EBV infection, Kikuchi disease or adult Still’s disease, because Kikuchi disease and adult Still’s disease sometimes had highly FDG-avid lymphadenopathy. As a result of relatively low metabolic activity in a few types of lymphoma including B-cell and T-cell lymphoma[37] and diffusely elevated FDG uptake in bone marrow caused by cytokine storm[20], the false negatives cannot be avoided. Hence, the biopsy of bone marrow and lymph nodes was definitely needed to confirm the diagnosis.

Our study had limitations, because small sample size from single center and sample selection bias may influence the generalized ability, sensitivity and specificity of the predictive models and nomograms. Occult malignancies may be difficult to detect and require repetitive tissue sampling. Even after a meticulous search, there may be still undetectable malignancies in HLH. There was heterogeneity of the patients in terms of age, physical condition and economic status. The predictive models and nomograms should be validated by prospective larger cohort from multi-center.

¹⁸F-FDG PET/CT can be a practical technique for identifying M-HLH and the models and nomograms constructed by ¹⁸F-FDG PET/CT features alone or combined with clinical parameters are able to predict malignancy with better accuracy than visual analysis of PET/CT images. Future efforts should be made to validate its diagnostic value with prospective larger cohort from multi-center.

Funding: Not applicable

Conflict of interest: The author declares that there are no conflicts of interest.

Availability of data and material: The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Code availability: Not applicable

Author contributions: Xu Yang and Xia Lu made substantial contributions to study design, image acquisition, data analysis and interpretation and new software creation in this work.

Shuxin Zhang, Wei Wang, Ying Kan, Lei Liu, and Jixia Li contributed writing and/or revising the manuscript.

Jigang Yang approved all versions to be published and were responsible for all aspects of this study.

Ethics approval: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional ethics committee and with the 1964 Declaration of Helsinki and its later amendments.

Acknowledgement:

Jigang YANG was supported by National Natural Science Foundation of China (No: 81971642,81771860), Beijing Natural Science Foundation (No:7192041), National Key Research and Development Plan (No: 2020YFC0122000).

Al-Samkari H, Berliner N. Hemophagocytic lymphohistiocytosis. Annu Rev Pathol. 2018;13:27–49.
Daver N, McClain K, Allen CE, Parikh SA, Otrock Z, Rojas-Hernandez C, et al. A consensus review on malignancy-associated hemophagocytic lymphohistiocytosis in adults. Cancer. 2017;123:3229–40. doi:10.1002/cncr.30826.
Ramos-Casals M, Brito-Zeron P, Lopez-Guillermo A, Khamashta MA, Bosch X. Adult haemophagocytic syndrome. Lancet. 2014;383:1503–16. doi:10.1016/s0140-6736(13)61048-x.
Parikh SA, Kapoor P, Letendre L, Kumar S, Wolanskyj AP. Prognostic factors and outcomes of adults with hemophagocytic lymphohistiocytosis. Mayo Clin Proc. 2014;89:484 – 92. doi:10.1016/j.mayocp.2013.12.012.
La Rosée P, Horne A, Hines M, von Bahr Greenwood T, Machowicz R, Berliner N, et al. Recommendations for the management of hemophagocytic lymphohistiocytosis in adults. Blood. 2019;133:2465–77. doi:10.1182/blood.2018894618.
Tabata C, Tabata R. Possible prediction of underlying lymphoma by high sIL-2R/ferritin ratio in hemophagocytic syndrome. Ann Hematol. 2012;91:63–71. doi:10.1007/s00277-011-1239-7.
Li F, Li P, Zhang R, Yang G, Ji D, Huang X, et al. Identification of clinical features of lymphoma-associated hemophagocytic syndrome (LAHS): an analysis of 69 patients with hemophagocytic syndrome from a single-center in central region of China. Med Oncol. 2014;31:902. doi:10.1007/s12032-014-0902-y.
Jiang T, Ding X, Lu W. The Prognostic Significance of Beta2 Microglobulin in Patients with Hemophagocytic Lymphohistiocytosis. Dis Markers. 2016;2016:1523959. doi:10.1155/2016/1523959.
Yoon SE, Eun Y, Huh K, Chung CR, Yoo IY, Cho J, et al. A comprehensive analysis of adult patients with secondary hemophagocytic lymphohistiocytosis: a prospective cohort study. Ann Hematol. 2020;99:2095–104. doi:10.1007/s00277-020-04083-6.
Yuan L, Kan Y, Meeks JK, Ma D, Yang J. 18F-FDG PET/CT for identifying the potential causes and extent of secondary hemophagocytic lymphohistiocytosis. Diagn Interv Radiol. 2016;22:471–5. doi:10.5152/dir.2016.15226.
Zheng Y, Hu G, Liu Y, Ma Y, Dang Y, Li F, et al. The role of (18)F-FDG PET/CT in the management of patients with secondary haemophagocytic lymphohistiocytosis. Clin Radiol. 2016;71:1248–54. doi:10.1016/j.crad.2016.05.011.
Wang J, Wang D, Zhang Q, Duan L, Tian T, Zhang X, et al. The significance of pre-therapeutic F-18-FDG PET-CT in lymphoma-associated hemophagocytic lymphohistiocytosis when pathological evidence is unavailable. J Cancer Res Clin Oncol. 2016;142:859–71. doi:10.1007/s00432-015-2094-z.
Kim J, Yoo SW, Kang SR, Bom HS, Song HC, Min JJ. Clinical implication of F-18 FDG PET/CT in patients with secondary hemophagocytic lymphohistiocytosis. Ann Hematol. 2014;93:661–7. doi:10.1007/s00277-013-1906-y.
Henter JI, Horne A, Arico M, Egeler RM, Filipovich AH, Imashuku S, et al. HLH-2004: Diagnostic and therapeutic guidelines for hemophagocytic lymphohistiocytosis. Pediatr Blood Cancer. 2007;48:124–31. doi:10.1002/pbc.21039.
Cazzola M. Introduction to a review series: the 2016 revision of the WHO classification of tumors of hematopoietic and lymphoid tissues. Blood. 2016;127:2361–4. doi:10.1182/blood-2016-03-657379.
Yamamoto S, Shimizu S, Kiyonaka S, Takahashi N, Wajima T, Hara Y, et al. TRPM2-mediated Ca2 + influx induces chemokine production in monocytes that aggravates inflammatory neutrophil infiltration. Nat Med. 2008;14:738–47. doi:10.1038/nm1758.
Canna SW, Marsh RA. Pediatric hemophagocytic lymphohistiocytosis. Blood. 2020;135:1332–43. doi:10.1182/blood.2019000936.
Boddu P, Oviedo SP, Rausch CR, Yam C, Daver N, Kantarjian H, et al. PET-CT in AML-related hemophagocytic lymphohistiocytosis. Leuk Lymphoma. 2018;59:1486–9. doi:10.1080/10428194.2017.1379079.
Shieh AC, Guler E, Smith DA, Tirumani SH, Beck RC, Ramaiya NH. Hemophagocytic Lymphohistiocytosis: A Primer for Radiologists. AJR Am J Roentgenol. 2020;214:W11-W9. doi:10.2214/AJR.19.21788.
Yang YQ, Ding CY, Xu J, Fan L, Wang L, Tian T, et al. Exploring the role of bone marrow increased FDG uptake on PET/CT in patients with lymphoma-associated hemophagocytic lymphohistiocytosis: a reflection of bone marrow involvement or cytokine storm? Leuk Lymphoma. 2016;57:291–8. doi:10.3109/10428194.2015.1048442.
Merrill SA, Naik R, Streiff MB, Shanbhag S, Lanzkron S, Braunstein EM, et al. A prospective quality improvement initiative in adult hemophagocytic lymphohistiocytosis to improve testing and a framework to facilitate trigger identification and mitigate hemorrhage from retrospective analysis. Med (Baltim). 2018;97:e11579. doi:10.1097/MD.0000000000011579.
Hao R, Yang X, Liu Z, Yang J. EBV-Associated T-Cell. Lymphoproliferative Disorders Demonstrated on FDG PET/CT in a Patient With Hemophagocytic Lymphohistiocytosis. Clin Nucl Med. 2019;44:829–30. doi:10.1097/rlu.0000000000002649.
Pan Q, Luo Y, Wu H, Ma Y, Li F. Epstein-Barr Virus-Associated Hemophagocytic Lymphohistiocytosis Mimicking Lymphoma on FDG PET/CT. Clin Nucl Med. 2018;43:125–7. doi:10.1097/rlu.0000000000001923.
Lu J, Fang Q, Ma C, Su F, Chen G, Huang M, et al. Atypical Epstein-Barr virus-associated hemophagocytic lymphohistiocytosis simulating lymphadenitis on (18)F-FDG PET/CT and its differential diagnosis. Hell J Nucl Med. 2017;20:254–7. doi:10.1967/s002449910612.
Dong A, Wang Y, Lu J, Zuo C. Enhanced CT. and FDG PET/CT findings of splenic hamartoma. Clin Nucl Med. 2014;39:968–71. doi:10.1097/RLU.0b013e3182a75ccc.
Rinaldi F, Giachè S, Spinicci M, Corsi P, Ambu S, Gianfaldoni G, et al. Focal spleen lesions in visceral leishmaniasis, a neglected manifestation of a neglected disease: report of three cases and systematic review of literature. Infection. 2019;47:507–18. doi:10.1007/s15010-019-01279-5.
Cegla P, Spychala A, Marszalek A, Wierzchoslawska E, Cholewinski W. Atypical spleen tuberculosis in a melanoma patient accidentally detected during a (18)F-FDG PET/CT study: Case report. Molecular clinical oncology. 2018;8:89–92. doi:10.3892/mco.2017.1498.
Milojevic IG, Sobic-Saranovic D, Petrovic N, Beatovic S, Tadic M, Artiko VM. Hybrid Imaging in Evaluation of Abdominal Sarcoidosis. Curr Med Imaging Rev. 2019;15:26–31. doi:10.2174/1573405614666180531111616.
Ito Y, Nakahara T, Nishihara Y, Shiraishi J, Jinzaki M. FDG PET/CT Findings in a Localized Reactive Lymphoid Hyperplasia of the Spleen. Clin Nucl Med. 2018;43:354–5. doi:10.1097/rlu.0000000000002020.
Zhang LJ, Xu J, Liu P, Ding CY, Li JY, Qiu HX, et al. The significance of 18F-FDG PET/CT in secondary hemophagocytic lymphohistiocytosis. J Hematol Oncol. 2012;5:40. doi:10.1186/1756-8722-5-40.
Chen Y, Wang Z, Luo Z, Zhao N, Yang S, Tang Y. Comparison of Th1/Th2 cytokine profiles between primary and secondary haemophagocytic lymphohistiocytosis. Ital J Pediatr. 2016;42:50. doi:10.1186/s13052-016-0262-7.
Ohno T, Ueda Y, Nagai K, Takahashi T, Konaka Y, Takamatsu T, et al. The serum cytokine profiles of lymphoma-associated hemophagocytic syndrome: a comparative analysis of B-cell and T-cell/natural killer cell lymphomas. Int J Hematol. 2003;77:286–94. doi:10.1007/bf02983788.
Xu XJ, Tang YM, Song H, Yang SL, Xu WQ, Zhao N, et al. Diagnostic accuracy of a specific cytokine pattern in hemophagocytic lymphohistiocytosis in children. J Pediatr. 2012;160:984 – 90.e1. doi:10.1016/j.jpeds.2011.11.046.
Weiss ES, Girard-Guyonvarc'h C, Holzinger D, de Jesus AA, Tariq Z, Picarsic J, et al. Interleukin-18 diagnostically distinguishes and pathogenically promotes human and murine macrophage activation syndrome. Blood. 2018;131:1442–55. doi:10.1182/blood-2017-12-820852.
Marnell L, Mold C, Du Clos TW. C-reactive protein: ligands, receptors and role in inflammation. Clin Immunol. 2005;117:104–11. doi:10.1016/j.clim.2005.08.004.
Deme D, Telekes A. [Prognostic importance of plasma C-reactive protein (CRP) in oncology]. Orv Hetil. 2017;158:243–56. doi:10.1556/650.2017.30646.
Cronin CG, Swords R, Truong MT, Viswanathan C, Rohren E, Giles FJ, et al. Clinical utility of PET/CT in lymphoma. AJR Am J Roentgenol. 2010;194:W91-w103. doi:10.2214/ajr.09.2637.

Table 1 Patients characteristics(n=71)
Characteristic	Number (%) of patients
Gender
Female	34(47.9)
Male	37(52.1)
Age at the time of HLH diagnosis（years）
Mean±SD	38±16.8
Median	38
Range	14-79
Malignancy associated HLH	22(31.0)
B cell neoplasms	11(15.5)
Diffuse large B-cell lymphoma	6(8.5)
Classical Hodgkin lymphoma	2(2.8)
Non-Hodgkin B cell lymphoma	1(1.4)
Follicular lymphoma	1(1.4)
B-cell acute lymphocytic leukemia	1(1.4)
T/NK cell neoplasms	11(15.5)
NK/T-cell lymphoma	6(8.5)
T-cell lymphoma	2(2.8)
Peripheral T-cell lymphoma	2(2.8)
Anaplastic large cell lymphoma	1(1.4)
Non-malignancy associated HLH	49(69.0)
EBV infection	24(33.8)
Other infections	12(16.9)
Adult Still's disease	5(7.0)
Systemic lupus erythematosus	1(1.4)
Undifferentiated systemic rheumatic disease	1(1.4)
Unknown	6(8.5)

Table 2 The comparison of Baseline 18F-FDG PET/CT findings in M-HLH and NM-HLH
Variables			M-HLH (n=22)	NM-HLH(n=49)	p values
Clinical variables
	sex(male)		16(72.7%)	21(42.9%)	0.020^*
	age		49(32-58)	33(25-52)	0.041^*
lymph node features
	high FDG uptake		17(77.3%)	28(57.1%)	0.103
		long diameter	1.8(1.1-2.4)	1.6(1.3-2.0)	0.851
		short diameter	1.3(0.8-1.7)	1.0(0.7-1.3)	0.313
		SUVmax-Lymph nodes	12.9(3.7-19.7)	4.6(3.5-9.7)	0.022^*
		asymmetry	14(82.4%)	10(35.7%)	0.002^*
bone features
	bone lesion		15(68.2%)	8(16.3%)	<0.001^*
		SUVmax-bone lesions	11.4(7.1-18.9)(n=15)	6.3(4.6-13.5)(n=8)	0.076
		SUVmax-bone background	4.2(3.7-5.8)(n=14）	4.0(3.6-4.9)(n=44)	0.349
Spleen features^a
	focal spleen lesion		7(33.3%)	6(12.2%)	0.081
	SUVmax-Spleen		4.2(3.1-9.6)	3.2(2.4-4.4)	0.009^*
	Volume-Spleen		650(516-1038)	432(281-878)	0.008^*
	TLG-Spleen		1791(1212-2598)	961(463-1688)	0.003^*
Liver features
	focal liver lesion		5(22.7%)	1(2.0%)	0.015^*
	SUVmax-liver		3.9(3.1-8.6)	2.9(2.5-3.4)	0.001^*
	Volume-Liver		1980(1525-2479)	1549(1212-1966)	0.020^*
	TLG-Liver		4171(3042-5628)	3089(2179-3896)	0.001^*
^aOne HLH patient with hematologic malignancy had been removed spleen, so the number of M-HLH spleen features is 21. *Significance at P<0.05

Table 3 The comparison of baseline laboratory parameters in M-HLH and NM-HLH
Variables			M-HLH (n=22)		NM-HLH(n=49)^a		p values
Variables			n^b	summary measure	n^b	summary measure	p values
Laboratory parameters			22		47
	WBC(×10⁹/L)			2.56(1.40-5.44)		4.90(2.04-7.18)	0.076
	ANC(×10⁹/L)			1.420.76-3.01)		2.65(1.20-4.97)	0.083
	HGB(g/L)			77.5(66.8-110.0)		93.0(84.0-104.0)	0.166
	PLT(×10⁹/L)			67.5(33-94)		107.0(57-191)	0.035^*
	CRP(mg/L)			47.4(10.8-94.2)		15.0(3.0-42.5)	0.011^*
	ALT(U/L)			46.5(19.5-105.5)		56.0(32.0-100.0)	0.507
	AST(U/L)			68.5(39.7-137.8)		62.9(29.3-113.2)	0.421
	TG(mmol/L)			2.04(1.5-2.8)		1.89(1.4-2.6)	0.532
	SF(ng/ml)		22	1957.2(1226.8-5692.5)	46	1390.0(440.3-3910.8)	0.132
	FBG(g/L)			2.84(1.71-3.93)		2.01(1.42-3.02)	0.155
	ESR(mm/h)		22	25.5(10.0-57.5)	46	20(10.0-36.5)	0.273
	LDH(U/L)			719(464-1374)		455(284-737)	0.028^*
	sCD25(pg/ml)		20	39516(9561-44000)	42	10038.5(3589-29099)	0.004^*
		sCD25/SF	20	10.2(2.1-23.9)	42	5.9(1.7-23.0)	0.598
	β2-MG(mg/L)		20	4.18(3.20-4.96)	35	3.17(2.30-4.82)	0.090
inflammatory cytokines^c(pg/ml)			17		33
	IL-1α			0.40(0.31-1.36)		0.40(0.30-0.60)	0.732
	IL-1β			1.00(0.70-1.95)		1.20(0.76-1.85)	0.735
	IL-2			3.40(2.50-4.10)		3.40(2.50-5.70)	0.538
	IL-4			5.60(3.00-20.77)		5.31(4.25-7.10)	0.984
	IL-6			8.90(4.45-42.85)		5.20(4.05-22.80)	0.282
	IL-8			12.00(3.64-25.35)		1.40(1.00-19.10)	0.025^*
	IL-10			45.70(8.57-194.85)		1.40(1.00-41.00)	0.007^*
	IL-12			3.40(2.75-5.50)		3.40(2.60-6.15)	0.878
	IL-17			1.10(0.70-1.40)		0.90(0.60-1.20)	0.340
	IL-18			168.10(75.39-390.98)		262.80(49.61-533.75)	0.862
	IL-23			8.10(4.81-11.70)		6.70(4.45-8.63)	0.384
	IFN-α			0.20(0.15-0.28)		0.20(0.19-0.30)	0.272
	IFN-γ(			190.40(67.25-265.42)		125.70(28.35-676.90)	0.975
	GM-CSF			7.00(5.95-7.15)		6.90(5.14-8.10)	0.984
	TNF-α			10.70(4.60-25.75)		13.40(5.40-31.85)	0.798
EBV infection			22	11(50%)	49	24(49%)	0.937
Hemophagocytosis			22	16(72.7%)	48	32(66.7%)	0.612
Data are median(interquartile range), mean ± SD or number(percentage)^aThe number of NH-HLH group is 47 because 2 HLH patients lacked available laboratory data. ^bData are number of patients with data available, blank indicates no missing data. ^cThe number of patients whose cytokines is available is 17 and 33 with M-HLH and NM-HLH respectively. ^Significance at P<0.05. ^*Significance at P<0.01. WBC=White blood cell, ANC=Absolute neutrophil count, HGB=Hemoglobin, PLT=Platelet count, CRP=C-reactive protein, ALT=alanine aminotransferase, AST=aspartate aminotransferase, TG=Triglycerides, SF=serum ferritin, FBG=fibrinogen, ESR=erythrocyte sedimentation rate, LDH=lactate dehydrogenase, sCD25=soluble interleukin-2 receptor (sIL-2R), β2-MG=Beta-2-microglobulin, IL=interleukin, IFN=interferon, GM-CSF=granulocyte-macrophage colony stimulating factor, TNF=tumor necrosis factor

Table 4 Multivariate logistic regression models for prediction of malignancy in patients with HLH
variables		Odds ratio(95%CI)	p values
Laboratory parameters(n=62)
	sCD25	1.36(1.08-1.70)^b	0.009
Laboratory parameters(n=69)^a
	LDH	1.90(1.10-3.29)^c	0.022
	CRP	1.13(1.02-1.24)^d	0.016
Combined parameters(n=69)
	SUVmax-Lymph nodes	1.15(1.04-1.27)	0.006
	SUVmax-bone lesions	1.22(1.08-1.38)	0.002
	CRP	1.12(1.00-1.25)^d	0.046
¹⁸F-FDG PET/CT parameters(n=71)
	SUVmax-Lymph nodes	1.13(1.03-1.24)	0.008
	SUVmax-bone lesions	1.24(1.10-1.40)	0.001
^aThe regression result when variables with missing values were removed such as sCD25. ^bThe OR is for every 6400pg/ml sCD25 increase. ^cThe OR is for every 500U/L LDH increase. ^dThe OR is for every 8mg/L CRP increase. sCD25=soluble interleukin-2 receptor (sIL-2R), LDH=lactate dehydrogenase, CRP=C-reactive protein,

Table 5 Comparison of ROC Curve of models in predicting malignancy in patients with HLH
	Cutoff	Accuracy(%)(95%CI)	Sensitivity(%)(95%CI)	Specificity(%)(95%CI)	PPV(%)(95%CI)	NPV(%)(95%CI)	Area under curve(95%CI)
sCD25(pg/ml)	38836	79.0(62.6-90.7)	55.0(31.5-76.9)	90.5(77.4-97.3)	73.3(50.0-88.3)	80.9(72.0-87.4)	0.730(0.603-0.835)
LDH＆CRP	24.6%	69.6(51.9-83.3)	77.3(54.6-92.2)	66.0(50.7-79.1)	51.5(40.2-62.7)	86.1(73.6-93.2)	0.741(0.621-0.839)
SUVmax-bone lesions＆SUVmax-lymph nodes	27.0%	84.5(68.9-93.6)	90.9(70.8-98.9)	81.6(68.0-91.2)	69.0(54.8-80.3)	95.2(84.1-98.7)	0.867(0.766-0.936)
SUVmax-bone lesions＆SUVmax-lymph nodes＆CRP	32.4%	87.0(71.4-95.8)	86.4(65.1-97.1)	87.2(74.3-95.2)	76.0(59.6-87.2)	93.2(82.6-97.5)	0.884(0.791-0.978)
tree(SUVmax-bone lesion＆SUVmax-lymph nodes)	+	87.3(72.1-95.9)	86.4(65.1-97.1)	87.8(75.2-95.4)	76.0(59.5-87.2)	93.5(83.3-97.6)	0.871(0.770-0.939)
PPV= positive predictive value, NPV= negative predictive value, LDH=lactate dehydrogenase, sCD25=soluble interleukin-2 receptor (sIL-2R), IL=interleukin,

YangSupplementalTable.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Identification of Malignancy-Associated Haemophagocytic Lymphohistiocytosis By 18F-FDG PET/CT Based Multivariate Models in Clinical Practice

Status:

Version 1

Abstract

Figures

Introduction

Methods

Results

Discussion

Conclusions

Declarations

References

Tables

Supplementary Files

Status:

Version 1