A novel use of CT attenuation value: increased bone marrow density in patients with acute myeloid leukemia

Abstract

Background

Acute myeloid leukemia (AML) is the most common type of acute leukemia in adults. Bone marrow computed tomography (CT) attenuation may increase in patients with myeloproliferative disorders; however, the actual threshold CT attenuation value predictive of myeloproliferative has not been reported.

Purpose

To determine whether the unenhanced CT attenuation value of the bone marrow may be useful for predicting AML.

Material and Methods

We retrospectively analyzed patients with AML (n = 56) who underwent unenhanced CT before treatment, and age- and sex-matched controls without any hematologic disease. For each patient, the CT attenuation value (HU) of the iliac bone was measured and compared between the two groups. Receiver operating characteristic (ROC) curve analysis was used to define the cutoff value for predicting AML on all patients, and only on late elderly patients (aged ≥75 years).

Results

Patients with AML showed higher bone marrow CT attenuation value (131.4 ± 58.3 vs. 53.9 ± 67.2 HU; P < 0.001), compared to the controls. The sensitivity and specificity for the diagnosis of AML in all patients were 78.6% and 80.4%, respectively, at a threshold value of 90 HU, whereas they were 83.3% and 91.7%, respectively, at 40 HU in late elderly patients.

Conclusion

The iliac bone CT attenuation value was elevated in patients with AML and may be useful for predicting AML.

Keywords

Acute myeloid leukemia computed tomography diagnostic radiology computed tomography attenuation value bone marrow

Introduction

Acute myeloid leukemia (AML) is a hematological malignancy that is characterized by infiltration of the hematopoietic organs (bone marrow, blood, spleen, and other tissues) by clonal, proliferative, abnormally differentiated, and non-functional hematopoietic blasts (1,2). It destroys the normal hematopoietic system and causes it to lose its functions by impeding the differentiation of hematopoietic stem cells, leading to life-threatening consequences (3). In the United States, an estimated 19,940 people are diagnosed with AML, and 11,180 patients died of the disease in 2020 (4).

AML is diagnosed by the presence of ≥20% blasts in the peripheral blood or in the bone marrow, or by the presence of specific genetic abnormalities in the bone marrow [t(8;21), inv(16), or t(15;17)] regardless of the blast count (5).

Although the diagnosis of AML is confirmed by pathological analysis, CT examination is occasionally performed due to the non-specific symptoms of AML like fatigue and fever. Anecdotally, we have encountered cases with increased bone marrow CT attenuation in patients with AML. While this finding is commonly seen in patients with myeloproliferative disorders in clinical practice, the actual threshold of the CT attenuation value predictive of AML has not been reported.

Therefore, the aim of the present study was to determine whether the unenhanced bone marrow CT attenuation value may be useful for predicting AML.

Material and Methods

Patients

We retrospectively analyzed all patients with AML treated at our hospital between January 2010 and December 2018. Inclusion criteria were as follows: (i) the final diagnosis of AML was made pathologically; (ii) unenhanced CT of the abdomen was performed before treatment; and (iii) age >20 years. Patients with relapsed AML were excluded from this study. As a control group, age- (within two years) and sex-matched controls without any hematologic disease were randomly selected from our hospital database. We recorded age, sex, body mass index (BMI), hemoglobin, white blood cell count, and platelet count for all patients. The chief complaint was also documented in patients with AML.

The appropriate institutional review board approved all aspects of this study. Considering the retrospective nature of the study, informed consent was waived, and the opt-out method was employed on the hospital website.

Computed tomography (CT)

All CT examinations were performed using one of the multidetector CT scanners (Aquilion ONE, Toshiba, Tokyo, Japan; Aquilion 64, Toshiba, Tokyo, Japan; Lightspeed VCT, GE Medical Systems, Waukesha, WI, USA; and SOMATOM Definition Flash, Siemens Healthcare, Forchheim, Germany) with automatic exposure control and a tube voltage of 120 kVp. Axial images at 5-mm thickness were used for the following image analysis.

Bone density was quantified by measuring the CT attenuation value (HU) by means of regions of interest (ROIs) with >100 mm² placed over both iliac bones at the level of posterior superior iliac spine (Fig. 1). Regions of cortical bone, beam-hardening, bone islands, and fractures were excluded from the ROIs. We also measured the craniocaudal length of the spleen to evaluate the presence of splenomegaly (6).

Two board-certified diagnostic radiologists with 10 and 9 years of experience in CT interpretation carried out the measurements, and the analysis of the data was made using the mean of the values they obtained.

Statistical analysis

We used Mann–Whitney U tests to compare age, BMI, hemoglobin, white blood cell count, platelet count, CT attenuation value of iliac bone, and the craniocaudal length of the spleen; we used the chi-square test to compare the sex proportions between groups. Linear regression analysis was performed to determine the correlation between age and bone marrow CT attenuation values. In addition, receiver operating characteristic (ROC) curve analysis was used to define the cutoff value of the attenuation value for predicting AML. These analyses were performed on all patients, and ROC curve analysis was also performed only on late elderly patients (aged ≥75 years). A P value <0.05 was considered significant. All statistical analyses were performed using SPSS version 25 software (IBM Japan, Tokyo, Japan).

Results

We found 188 patients with AML in the hospital database. Of these, 132 were excluded due to the absence of CT images before treatment (n = 116), only contrast-enhanced CT was performed (n = 10), or relapsed AML (n = 6). Therefore, 56 patients including 12 late elderly patients met the study criteria.

The chief complaints of patients with AML were blood test abnormality without any symptoms (n = 14), fatigue (n = 9), fever (n = 9), shortness of breath/dyspnea (n = 9), easy bleeding (n = 5), headache (n = 3), arthralgia (n = 3), sore throat (n = 3), stomatitis (n = 3), palpitations (n = 2), fainting (n = 2), and others (n = 7).

Patients with AML showed higher bone marrow CT attenuation values (131.4 ± 58.3 vs. 53.9 ± 67.2 HU; P < 0.001), compared to the control group (Figs. 2–4). Patients with AML also showed a lower hemoglobin (9.3 vs 12.9 mg/dL; P < 0.001), lower platelet count (54,000 vs 220,000/μL; P < 0.001), and a larger spleen (86.6 vs. 75.2 cm; P = 0.013). No significant differences in any other factors were apparent between the two groups (Table 1).

Table 1.

Summary of patients’ characteristics and laboratory /imaging data in the two groups.

	AML group (n = 56)	Control group (n = 56)	P value
Age (years)	61.8 ± 15.5	62.0 ± 15.2	0.940
Sex
Male	35	35	1.000
Female	21	21
BMI (kg/m²)	22.6 ± 4.0	22.1 ± 3.9	0.600
Hb (g/dL)	9.3 (4.2–14.5)	12.9 (7.1–16.6)	<0.001
WBC (/μL)	7250 (600–262,300)	5800 (2200–15,100)	0.330
Plt (/μL)	54,000 (7000–5,118,000)	220,000 (71,000–412,000)	<0.001
CT AV of iliac bone (HU)	131.4 ± 58.3	53.9 ± 67.2	<0.001
Spleen size (cm)	86.6 ± 25.1	75.2 ± 25.2	0.013

Values are given as n, mean ± SD, or median (range).

AV, attenuation value; BMI, body mass index; Hb, hemoglobin; Plt, platelet count; WBC, white blood cell count.

Linear regression analysis revealed a negative correlation in the study population between age and bone marrow CT attenuation values in both patients with AML and controls (Fig. 5). The results of the ROC analysis of all patients are shown in Fig. 6. The area under the ROC curve (AUC) was 0.824 for the bone marrow CT attenuation value. The threshold of iliac bone CT attenuation value was calculated for high sensitivity (approximately 90%), high specificity (approximately 90%), and balanced sensitivity-specificity (approximately 80%). The sensitivity and specificity for the diagnosis of AML were 91.1% and 51.8% at threshold values of 50 HU, 48.2% and 89.3% at a threshold value of 130 HU, and 78.6% and 80.4% at a threshold value of 90 HU.

The results of the ROC analysis of late elderly patients are shown in Fig. 7. The AUC was 0.958 for the bone marrow CT attenuation value. The sensitivity and specificity for the diagnosis of AML were 83.3% and 91.7% at threshold values of 40 HU.

Discussion

This study investigated the diagnostic value of bone marrow density with unenhanced CT in patients with AML. We showed that the iliac bone marrow CT attenuation value is significantly increased in patients with AML compared to controls. Our results suggest that CT images could be used as predictors of AML.

Although characteristic imaging findings of bone marrow infiltration by malignant cells have been extensively studied on magnetic resonance imaging (MRI) (7 –11), there have been few reports of CT findings of bone marrow infiltration in patients with AML. Although bone marrow CT attenuation has been known to sometimes increase in patients with myeloproliferative disorders, there has only been one case report that described increased bone marrow CT attenuation in a patient with AML (12). No previous reports have described the actual threshold of the CT attenuation value predictive of AML.

In the present study, we focused on AML, since this is the most common type of acute leukemia in adults with very poor overall survival rates (13). In the progression of AML, bone marrow infiltration of leukemia cells may increase the cellularity with more blasts and tumor angiogenesis (14,15). In leukemia, bone marrow infiltration usually occurred diffusely (7); however, it is difficult to detect bone marrow infiltration on CT images without lytic reaction in the axial skeleton, since the dense trabecular bone restricts the detection of medullary infiltration in the axial skeleton. Since the fatty marrow physiologically replaces the bone marrow of the appendicular skeleton in normal healthy adults, unenhanced CT easily visualizes an increase of cellularity in the medullary cavity of the appendicular skeleton as increased bone marrow density especially in late elderly patients (16 –18). Therefore, we used the iliac bone, which is often used for bone marrow aspiration, as the location to measure the CT attenuation value in the present study.

Peripheral blood and bone marrow blasts examination is the gold standard for the diagnosis of AML (2,19). However, we suspect that unenhanced CT is occasionally performed due to non-specific symptoms like fatigue and fever. Early suspicion of AML would lead to earlier diagnosis and appropriate management of associated complications, such as anemia or infection.

Of course, we would not suggest that CT itself may be useful for the diagnosis of AML. However, in the presence of abnormally high bone marrow density in unenhanced body CT, radiologists can suggest the possibility of AML. This may be particularly important because the elevation of bone marrow density may occur before the appearance of blood abnormalities and any symptoms related to AML (12).

The present study has some limitations. First, this study included a relatively small number of patients. However, we achieved high levels of statistical significance between the patients with AML and controls, suggesting that the results of the present study were reliable enough. Second, we did not evaluate the extramedullary disease of AML except splenomegaly; however, the incidence of extramedullary disease has been reported to be low and not specific by itself; lymph node involvement in 11.5% and liver involvement in 5.3% (20). Third, we did not evaluate the differences among subtypes of AML. Final diagnosis would be eventually made by pathology; therefore, we think that distinguishing these subtypes has no clinical importance at the time of pointing out the possibility of AML on CT images. Fourth, we used the four CT scanners in this study. Since CT attenuation values can vary depending on the CT scanners and its scan parameters (21,22), evidence on using a specific number of the bone marrow CT attenuation value on this study as a cutoff value for prediction of AML was limited. Finally, we have to add that other hematologic diseases involving the bone marrow may induce an increase in bone marrow density; therefore, this finding is not specific for AML.

In conclusion, the present study demonstrated that the iliac bone marrow CT attenuation value was elevated in patients with AML, which may be predictive of AML. We recommend that radiologists should routinely check the bone marrow density on unenhanced CT and report it if it is elevated even when AML is not clinically suspected.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Soma Kumasaka

References

Döhner

Weisdorf

Bloomfield

. Acute myeloid leukemia. N Engl J Med 2015;373:1136–1152.

Pelcovits

Niroula

. Acute myeloid leukemia: a review. R I Med J 2020;103:38–40.

Miraki-Moud

Anjos-Afonso

Hodby

, et al. Acute myeloid leukemia does not deplete normal hematopoietic stem cells but induces cytopenias by impeding their differentiation. Proc Natl Acad Sci U S A 2013;110:13576–13581.

Siegel

Miller

Jemal

. Cancer statistics, 2020. CA Cancer J Clin 2020;70:7–30.

Döhner

Estey

Grimwade

, et al. Diagnosis and management of AML in adults: 2017 ELN recommendations from an international expert panel. Blood 2017;129:424–447.

Robertson

Leander

Ekberg

. Radiology of the spleen. Eur Radiol 2001;11:80–95.

Silva

Hayashi

Yonenaga

, et al. MRI of bone marrow abnormalities in hematological malignancies. Diagn Interv Radiol 2013;19:393–399.

Nonomura

Yasumoto

Yoshimura

, et al. Relationship between bone marrow cellularity and apparent diffusion coefficient. J Magn Reson Imaging 2001;13:757–760.

Kan

Hernanz-Schulman

Frangoul

, et al. MRI diagnosis of bone marrow relapse in children with ALL. Pediatr Radiol 2008;38:76–81.

10.

Moulopoulos

Dimopoulos

Kastritis

, et al. Diffuse pattern of bone marrow involvement on magnetic resonance imaging is associated with high risk cytogenetics and poor outcome in newly diagnosed, symptomatic patients with multiple myeloma: a single center experience on 228 patients. Am J Hematol 2012;87:861–864.

11.

Grønningsæter

Ahmed

Vetti

, et al. Bone marrow abnormalities detected by magnetic resonance imaging as initial sign of hematologic malignancies. Clin Pract 2018;8:1061.

12.

El Rassi

Little

Holloway

, et al. Early diagnosis of acute myeloid leukemia by computed tomography scan. J Clin Oncol 2012;30:e207–e208.

13.

Pollyea

Bixby

Perl

, et al. NCCN guidelines insights: acute myeloid leukemia, version 2.2021, version 2.2021 featured updates to the NCCN guidelines. JNCCN J Natl Compr Cancer Netw 2021;19:16–27.

14.

Hussong

Rodgers

Shami

. Evidence of increased angiogenesis in patients with acute myeloid leukemia. Blood 2000;95:309–313.

15.

Testa

Riccioni

Militi

, et al. Elevated expression of IL-3Rα in acute myelogenous leukemia is associated with enhanced blast proliferation, increased cellularity, and poor prognosis. Blood 2002;100:2980–2988.

16.

Blebea

Houseni

Torigian

, et al. Structural and functional imaging of normal bone marrow and evaluation of its age-related changes. Semin Nucl Med 2007;37:185–194.

17.

Matsue

Kobayashi

Matsue

, et al. Prognostic significance of bone marrow abnormalities in the appendicular skeleton of patients with multiple myeloma. Blood Adv 2018;2:1032–1039.

18.

Abe

Ugai

Fujisawa

, et al. Prognostic implication of appendicular skeleton bone marrow abnormalities detected using low-dose multidetector computed tomography in patients with myelodysplastic syndrome. Leukemia 2018;32:2069–2073.

19.

Percival

Lai

Estey

, et al. Bone marrow evaluation for diagnosis and monitoring of acute myeloid leukemia. Blood Rev 2017;31:185–192.

20.

Ganzel

Manola

Douer

, et al. Extramedullary disease in adult acute myeloid leukemia is common but lacks independent significance: analysis of patients in ECOG-ACRIN cancer research group trials, 1980–2008. J Clin Oncol 2016;34:3544–3553.

21.

Birnbaum

Hindman

Lee

, et al. Multi-detector row CT attenuation measurements: assessment of intra- and interscanner variability with an anthropomorphic body CT phantom. Radiology 2007;242:109–119.

22.

Davis

Palmer

Pani

, et al. Assessment of the variation in CT scanner performance (image quality and Hounsfield units) with scan parameters, for image optimisation in radiotherapy treatment planning. Phys Med 2018;45:59–64.