Patient Characteristics
Sixty-five patients were enrolled in this study in total, including 34 males and 31 females. The median age was 62 (32–79), and 34 patients had smoking history. Twenty-one patients had negative results of all lung tumor markers (NSE, CA199, CA125, SCC, CEA and CYFRA21-1) while 12 patients had single marker positive, and 14 patients with two or more markers positive. Number of patients in the three research groups are: nodules characterization (63 cases), N staging (35 cases) and M staging (9 cases) (Fig. 1).
Nodule characterization comparison of [ 68 Ga]Ga-DOTA-FAPI-04 and [ 18 F]FDG PET/CT in lung nodules
Sixty-eight lesions of 63 patients (two patients with three nodules and one patient with two nodules) were enrolled in the nodule characterization group. In reference to diagnosis, except for two inflammatory nodules diagnosed by follow-up CT imaging, pathological results of the other 66 lung nodules were obtained from either resection or aspiration biopsy. To analyze the tumor heterogeneity, 68 lung nodules were characterized as three types according to their pathological results: adenocarcinoma (AC, 51 lesions), squamous cell carcinoma (SCC, 9 lesions) and inflammation & granuloma (Inf & G, 8 lesions). Furthermore, all 68 lesions were also categorized by their density on CT images: 46 SN, 4 GGN and 18 PSN. Imaging characteristics and one-way ANOVA among groups are shown in Table 1. The density of SCC and Inf & G nodules were basically recognized as SN, while the density of AC nodules was different. Meanwhile, adenocarcinoma tended to have vacuoles, pleural indentation signs and lower uptake. Only lobular sign incidence exhibited differences between NSCLC and Inf & G, but the other imaging characteristics did not.
Table 1
Imaging characteristics of lung nodules
Characteristics | AC (%) | SCC (%) | Inf & G (%) | F | P |
Lesions | 51 | 9 | 8 | | |
Density | | | | | |
SN | 30 (58.8) | 9 (100) | 7 (87.5) | | |
PSN | 17 (33.3) | 0 | 1(12.5) | 3.666 | 0.031# |
GGN | 4 (7.8) | 0 | 0 | | |
Long diameter(cm) | 2.39 ± 1.17 | 2.47 ± 1.46 | 2.3 ± 0.73 | 0.038 | 0.963 |
Lobular sign | 41 (80.4) | 5 (55.6) | 1 (12.5) | 9.859 | 0.000* |
Spiculation | 45 (88.2) | 7 (77.8) | 7 (87.5) | 0.354 | 0.703 |
Vacuole | 28 (54.9) | 1 (11.1) | 2 (25) | 4.007 | 0.023# |
Pleural indentation | 41 (80.4) | 4 (44.4) | 4 (50) | 3.790 | 0.028# |
Calcification | 1 (2) | 3 (33.3) | 0 | 0.163 | 0.850 |
FDG SUVmax | 5.5 ± 4.8 | 11.4 ± 7.2 | 7.4 ± 4.9 | 5.128 | 0.009# |
FAPI SUVmax | 7.2 ± 4.5 | 9.7 ± 3.0 | 7.5 ± 3.7 | 1.330 | 0.272 |
FDG SUVmean | 3.9 ± 3.1 | 8.0 ± 4.0 | 5.1 ± 3.3 | 6.437 | 0.003# |
FAPI SUVmean | 5.0 ± 2.9 | 5.9 ± 1.9 | 5.0 ± 2.2 | 0.406 | 0.668 |
*: differences among groups caused by Inf & G. #: differences among groups caused by AC. |
Table 2
Diagnosis efficacy of NSCLC nodules in [68Ga]Ga-DOTA-FAPI-04 and [18F]FDG PET/CT
Diagnosis efficacy | FDG | FAPI |
SEN | 0.67 | 0.88 |
SPEC | 0.13 | 0.13 |
ACC | 0.60 | 0.79 |
PPV | 0.85 | 0.88 |
NPV | 0.05 | 0.14 |
SEN: sensitivity; SPEC: specificity; ACC: accuracy; PPV: positive predict value; NPV: negative predict value. |
High uptake of [68Ga]Ga-DOTA-FAPI-04 exhibited in AC, SCC and Inf & G nodule types, and the differences among them were not statistically significant. These three types showed high uptake of [18F]FDG as well, in which SCC was higher than that of AC, but the difference between NSCLC and inflammation was not statistically significant. Meanwhile, there was no differences in blood pool uptake of [68Ga]Ga-DOTA-FAPI-04 and [18F]FDG (Fig. 2A). Diagnostic efficacy of [68Ga]Ga-DOTA-FAPI-04 in NSCLC was better than [18F]FDG PET/CT, as the diagnostic accuracy was 0.79 and 0.6 respectively in this study with criteria as SUVmax ≥ 2.5 (Table. 2).
The AC nodules were further analyzed with respect to the densities. SN showed high uptake of both [68Ga]Ga-DOTA-FAPI-04 and [18F]FDG, while low uptake exhibited in GGN with no significant difference between two radiotracers. Though medium uptake exhibited in PSN, the uptake of [68Ga]Ga-DOTA-FAPI-04 was higher than [18F]FDG with significant differences (SUVmax were 4.8 ± 2.8 and 2.1 ± 1.1, p = 0.003) (Fig. 2B). A typical case where PSN of AC showed high uptake of [68Ga]Ga-DOTA-FAPI-04 but with low uptake of [18F]FDG is presented. The immunohistopathology (IHC) result verified high FAP expression in tumor interstitials (Fig. 3).
Diagnostic efficacy comparison of [ 68 Ga]Ga-DOTA-FAPI-04 and [ 18 F]FDG PET/CT in N staging of NSCLC patients
Totally, 100 LNs in 35 patients were enrolled, 17 of 35 (48.6%) patients companied with only non-metastatic LNs, while 18 of 35 (51.4%) patients have metastatic LNs. The diagnosis of non-metastatic LNs were obtained from post-surgery pathological results in 15 patients, and follow-up imaging confirmed such LNs in two patients. Eight patients with metastatic LNs were diagnosed by either resection or aspiration biopsy pathological results and ten patients were diagnosed by follow-up imaging. All these 100 LNs were assigned to metastatic (36 lesions) or non-metastatic (64 lesions) group depending on final diagnosis.
Imaging characteristics of LNs is shown in Table 3. The metastatic group tended to show lower density and larger short diameter, while no significant differences existed in risk categories and calcification between metastatic and non-metastatic groups. SUV of [68Ga]Ga-DOTA-FAPI-04 was effective in differentiating metastatic and non-metastatic groups, with a higher uptake in the metastatic group, while SUV of [18F]FDG did not show differences between the two groups. Moreover, SUV of [68Ga]Ga-DOTA-FAPI-04 was lower than [18F]FDG in the non-metastatic group (p < 0.01) and higher than [18F]FDG in the metastatic group (p < 0.01) (Fig. 4). This indicated the potential of 68Ga-DOTA-FAPI-04 for differential diagnosis of metastatic and non-metastatic LNs. One case (Fig. 5) depicted that several high uptake LNs which were suspected as metastatic LNs in [18F]FDG showed low uptake of [68Ga]Ga-DOTA-FAPI-04. The post-resection pathological result showed adenocarcinoma with no metastatic LNs (0 / 20),and the FAP expression was low as confirmed by IHC.
Table 3
Imaging characteristics of lymph nodes (by lesions)
Characteristics | Metastatic (%) | Non-metastatic (%) | χ2ort | p |
Lesions | 36 | 64 | | |
Risk categories | | | | |
High | 22 (61.1) | 31 (48.4) | | |
Intermediate | 6 (16.7) | 10 (15.6) | 2.112* | 0.348 |
Low | 8 (22.2) | 23 (35.9) | | |
Density | | | | |
High | 6 (16.7) | 37 (57.8) | | |
Iso-density | 22 (61.1) | 27 (42.2) | 24.977* | 0.000 |
Low | 8 (22.2) | 0 | | |
Calcification | 2 (5.6) | 9 (14.1) | 1.703* | 0.319 |
Short diameter (cm) | 1.2 ± 0.54 | 0.8 ± 0.16 | -5.314# | 0.000 |
FDG SUVmax | 6.5 ± 3.30 | 6.1 ± 2.32 | -0.754# | 0.453 |
FAPI SUVmax | 10.7 ± 4.72 | 3.1 ± 1.29 | -12.233# | 0.000 |
FDG SUVmean | 4.9 ± 2.10 | 4.5 ± 1.46 | -1.270# | 0.207 |
FAPI SUVmean | 7.4 ± 3.13 | 2.5 ± 0.96 | -11.576# | 0.000 |
FAPI SUVmax/ FDG SUVmax | 1.8 ± 0.93 | 0.5 ± 0.19 | -10.975# | 0.000 |
*, tested by chi-square test, χ2 values were documented; #, t tested by Mann–Whitney U test, t values were documented. |
Table 4
Multivariate logistic regression analysis of the relationship between PET parameters with metastatic LNs
Factor | Odds Ratio (95% CI) | p value |
FAPI SUVmax | 1.837 (1.189–2.838) | 0.006 |
FAPI SUVmax/FDG SUVmax | 11.438 (1.178-111.083) | 0.036 |
Table 5
The diagnosis efficacy for identifying metastatic LNs
Criteria | By lesions | | By cases |
SEN(%) | SPEC(%) | ACC(%) | | SEN(%) | SPEC(%) | ACC(%) |
FDG SUVmax≥2.5 | 91.70 | 1.60 | 34 | | 83.30 | 0 | 42.90 |
FAPI SUVmax≥6 | 86.10 | 100 | 95 | | 77.80 | 100 | 88.60 |
Ratio#≥1.1 | 80.60 | 100 | 93 | | 66.70 | 100 | 82.90 |
FAPI SUVmax≥6 or Ratio#≥1.1 | 89.20 | 100 | 97 | | 83.30 | 100 | 91.40 |
#: Ratio: FAPI SUVmax/ FDG SUVmax; Sensitivity, SEN; Specificity, SPEC; Accuracy, ACC. |
Diagnostic efficacy of [68Ga]Ga-DOTA-FAPI-04 was analyzed by taking parameters which could differentiate metastatic and non-metastatic groups into multivariate logistic regression, including density, short diameters, SUVmax of FAPI and FAPI SUVmax/ FDG SUVmax. The results indicated that FAPI SUVmax and FAPI SUVmax/FDG SUVmax could be used in differential diagnosis of metastatic and non-metastatic LNs while density and short diameters could not (Table.4). The area under receiver operating characteristic (ROC) curve of FAPI SUVmax and FAPI SUVmax/FDG SUVmax was 0.927 and 0.954, respectively (Fig. 6), which indicated a prominent diagnostic ability of these two parameters.
FAPI SUVmax ≥6 and the ratio of FAPI SUVmax/FDG SUVmax ≥1.1 were set as the cut-off value to analyze the diagnostic efficacy by the perspective of lesions and cases. The combination of these two conditions and [18F]FDG criteria were also analyzed and shown in Table. 5.
FAPI SUVmax and FAPI SUVmax/FDG SUVmax criteria identified more true negative cases than [18F]FDG, however it made two more false negative cases than [18F]FDG. Our result highlighted that LNs with low uptake of [68Ga]Ga-DOTA-FAPI-04 and low FAPI SUVmax/FDG SUVmax were more likely to be non-metastatic LNs.
Diagnostic efficacy comparison of [ 68 Ga]Ga-DOTA-FAPI-04 and [ 18 F]FDG PET/CT in M staging of NSCLC patients
Nine patients with multi-metastatic lesions were enrolled in the study, and the lesion detectability of [68Ga]Ga-DOTA-FAPI-04 PET/CT (206 lesions) was better than FDG PET/CT (106 lesions), but no patient’s staging was changed. Tumor to blood-pool ratio (TBR) of [68Ga]Ga-DOTA-FAPI-04 was higher than that of FDG with significant differences (6.5 ± 1.9 and 4.0 ± 2.0, p < 0.05). [68Ga]Ga-DOTA-FAPI-04 PET/CT presented more abnormal uptake lesions than [18F]FDG PET/CT in a multi-metastasis patient (Fig. 7).