Patients’ clinical characteristics
In this study, 23 patients with NTM infection were included, and their demographics and clinical features are presented in Table 1. They were 19 men (82.6%) and four women (17.4%), with age ranging between 14 and 84 years (median: 60 years). The most common clinical manifestations were fever (12 cases, 52.2%), coughing and expectoration (8 cases, 39.1%), ostealgia (5 cases, 21.7%), fatigue (2 cases, 8.7%), chest tightness (1 case, 4.3%), and weight loss (1 case, 4.3%).
In terms of immunological status and risk factors, six patients (26.1%) had high risk factors due to immunological dysfunction, such as chemotherapy (3 cases, 13.0%), acquired immune deficiency syndrome (AIDS) (2 cases, 8.7%), and monocytopenia and mycobacterial infection syndrome (MonoMAc syndrome) (1 case, 4.3%). In addition, 8 patients (37.8%) had mild risk factors, such as a history of chemotherapy (4 cases, 17.4%), splenectomy (1 case, 4.3%), hepatitis B cirrhosis (1 case, 4.3%), tuberculosis (1 case, 4.3%) and use of small-dose glucocorticoids (1 case, 4.3%). Furthermore, six patients (26.1%) had mild chronic obstructive pulmonary disease as a result of long-term smoking, while three patients (13.0%) had no risk factors or underlying diseases.
NTM infection was diagnosed in the included cases using mNGS (14 cases, 60.9%) and specimen culture (9 cases, 39.1%). Specimens were collected from blood or alveolar lavage fluid, pus, surgical sites, and other sites. Regarding the NTM species, the following were identified: M. intracellulare in 12 cases (52.2%); M. avium (4 cases, 17.4%); M. abscessus (3 cases, 13.0%); M. paraintracellulare (1 case, 4.3%); and M. chimaera (1 case, 4.3%).
The laboratory characteristics of the 23 patients with NTM infection are presented in Table 2. The median white cell count was 7.6 × 109/L [interquartile range (IQR): 2.3–22.8 × 109/L], with nearly half of the patients having leukocytosis (10 cases, 43.5%). Of these patients, the neutrophil count was elevated in nine patients, and the lymphocytic count was elevated in one. The median platelet count was 226 × 109/L [interquartile range (IQR): 85–492 × 109/L], with more than one-third of the patients having polyplastocytosis (8 cases, 34.8%).However, red blood cell counts and hemoglobin levels were reduced in eight patients (34.8%) and 11 patients (47.8%), respectively, with eight patients having both decreased. Although the majority of patients had variable levels of elevation in C-reactive protein (CRP) (16 cases, 69.6%) and erythrocyte sedimentation rate (ESR) (15 cases, 65.2%), the procalcitonin (PCT) levels which is commonly used to determine bacterial infection were all within the normal range. The serum ferritin and CA125 levels were elevated in 13 (56.5%) and 9 (39.1%) patients, respectively. In addition, the median serum ferritin and CA125 levels were 326.6 (IQR: 43.2–2286.6 g/dL) and 34.6 (IQR: 2.1–290.8 U/L), respectively.
In terms of serum cytokine levels, the median interleukin 6 (IL-6) level was 13.9 pg/ml (IQR: 1.1–184.3 pg/ml), with 20 patients (87%) having significant elevations. In addition, tumor necrosis factor-alpha (TNF-α) and interferon-gamma (IFN-γ) levels were both elevated in two patients. Furthermore, serum IL-10 level was elevated in only one patient, whereas serum IL-2 and IL-4 levels were all within the normal range.
The functional immune cell percentage in peripheral blood was within the reference range in the majority of patients, with only two patients having elevated total T lymphocytes, two patients having decreased assisted/induced T lymphocytes and inhibited/cytotoxic T lymphocyte, and one patient having decreased B lymphocytes.
18F-FDG PET/CT scan imaging characteristics
The 18F-FDG PET/CT scan imaging characteristics in patients with NTM infection are summarized in Table 3
The 18F-FDG PET/CT scan imaging revealed that 21 patients (91.3%) had an intensive or moderate uptake of 18F-FDG in the pulmonary lesions, with a single lobe (5 cases) and double lobe involvement (16 cases). Of these patients, 18 (85.7%) had patchy infiltration on breath-holding lung CT. In addition, bronchiectasis (8 cases), cavitation (5 cases), pleural effusion (4 cases), nodular infiltration (3 cases), calcification (3 cases) and atelectasis (2 cases) were detected. Furthermore, high FDG-uptake (SUVI-lung﹥2.5) lesions were found in 14 patients (Fig. 2A, blue arrows) and low FDG-uptake (SUVI-lung ≤ 2.5) lesions were found in seven patients (Fig. 2B, white arrows).
The 18F-FDG PET/CT scan imaging revealed abnormally high 18F-FDG FDG uptake in extra-pulmonary lesions in 17 patients (73.9%), with increased 18F-FDG accumulation in lymph nodes (14 cases), bone (6 cases), subcutaneous tissues (1 case), spleen (1 case), and adrenal glands (1 case).
Distribution of lesions
The 18F-FDG PET/CT scan imaging revealed that the lesion involved multiple systems in 16 patients, including the lung and lymph nodes (8 cases), lung and bone (3 cases) (Fig. 3), lung, lymph nodes, bone, and subcutaneous tissues (one case), lung, lymph nodes, and bone (one case), lung, lymph nodes, and spleen (one case), lung, lymph nodes, and adrenal gland (one case), and lymph nodes and bone (one case). Furthermore, NTM infection was found in multiple sites within a single-system in six patients (five patients with pulmonary involvement and one patient with lymph node involvement), whereas NTM infection was found in a single site within a single-system in one case (nodular infiltrate in the left upper lobe of the lung).
The median SUVMax in 23 patients with pulmonary involvement was 10.0 (range: 1.2–28.1).The median SUVI-lung in 21 patients with pulmonary involvement was 5.5 (range: 1.2–13.9). The median SUVE-lung in 23 patients was 5.4 (range: 1.8–28.1). The median SUVLiver, SUVMarrow and SUVSpleen were 2.7 (range: 1.7–3.8), 3.1 (range: 1.7–6.7), and 2.1 (range: 1.7–6.0), respectively. The median SURLiver and SURBlood were 3.2 (IQR: 0.5–7.6) and 4.1 (IQR: 0.7–10.3), respectively.
18F-FDG PET/CT scan imaging characteristics in different immunological states, NTM lesion distribution and severity
The clinical and FDG PET/CT scan imaging characteristics of different immunological states, as well as NTM lesion distribution and severity, are presented in Table 4. In terms of the NTM lesion distribution, there were significant differences in the SURLiver, SURBlood and SUVMax of the 18F-FDG-avid lesions between the localized and disseminated groups (p = 0.009, 0.003, and 0.005, respectively) (Fig. 4A). In terms of the NTM lesion severity, SUVI-Lung, SUVMarrow, and platelet count significantly differed between the severe and non-severe groups (p = 0.036, 0.006, and 0.007, respectively) (Fig. 4C). Furthermore, one-way analysis of variance revealed significant differences in metabolic parameters of 18F-FDG-avid lesions, including SURLiver, SURBlood and SUVMax among the three groups with different immunological states (p = 0.007, 0.012, and 0.010, respectively) (Fig. 4F).
Receiver operating characteristic (ROC) curves were created to assess the consistency of 18F-FDG PET/CT metabolic parameters and clinical features (Fig. 4). In the localized and disseminated groups, the areas under the curve (AUCs) for SUVMax, SURLiver, and SURBlood were 0.796 (p = 0.027), 0.853 (p = 0.008), and 0.826 (p = 0.015), respectively (Fig. 4B). When the cut-off values of SUVMax, SURLiver, and SURBlood were set at 7.7, 2.9, and 3.3, respectively, the sensitivity of 18F-FDG PET/CT for lesion distribution was 68.8%, 75%, and 87.5%, respectively, while the specificity was 71.4%, 71.4%, and 71.4%, respectively. In the severe and non-severe groups, the AUCs for SUVMarrow, SUVI-lung, and platelet count were 0.897 (p = 0.001), 0.688 (p = 0.158), and 0.794 (p = 0.019), respectively (Fig. 4D). When the cut-off values of SUVMarrow, SUVI-lung, and platelet count were set at 3.1, 4.6, and 202.5 × 109/L, the sensitivity of 18F-FDG PET/CT for severity detection was 76.9%, 76.9%, and 92.3%, respectively, while the specificity was 88.9%, 55.9%, and 55.6%, respectively. In the immunocompromised and immunocompetent groups, the AUCs for SUVMax, SURLiver, and SURBlood were 0.759 (p = 0.05), 0.768 (p = 0.04), and 0.696 (p = 0.14), respectively (Fig. 4E). The cut-off values for SUVMax, SURLiver, and SURBlood were set at 5.3, 3.8 and 3.0, respectively, and the sensitivity of 18F-FDG PET/CT for compromised immune function in patients with NTM infection was 88.2%, 52.9%, and 76.5%, respectively, while the specificity was 50.0%, 83.3% and 50.0%, respectively.
Correlations between metabolic parameters of 18F-FDG PET/CT scan imaging and clinical laboratory results
The Spearman’s correlation analysis (Fig. 5) revealed that PCT and IL-6 correlated positively with SUVMarrow of the 18F-FDG PET/CT scan imaging (r = 0.507 and 0.530, respectively, both p < 0.05). Furthermore, PCT and ferritin correlated positively with SUVSpleen (r = 0.577 and 0.558, respectively, both p < 0.05).