Lymphoma Therapy Response Assessment with Low-Dose [¹⁸F]FDG Total-Body PET/CT

Visual Abstract

Abstract

The improved sensitivity of total-body (TB) PET/CT offers the possibility of reducing injected activities. The aim of our study was to define a lower limit of reduced injected activities in [¹⁸F]FDG TB PET/CT for interim and end-of-treatment assessment of patients with lymphoma at 2 acquisition times. Methods: Twenty-four consecutive patients with lymphoma who were undergoing interim and end-of-treatment TB PET/CT were prospectively enrolled in this study. An [¹⁸F]FDG activity of 3.0 MBq/kg served as the reference standard (RS). Images simulating low doses of 1.0, 0.5, 0.25, and 0.125 MBq/kg were reconstructed at 1 and 2 h after injection. The coefficient of variation of the liver was assessed. Lymphoma lesions were segmented and semiquantitatively compared with the RS using the SUV. Additionally, metabolic tumor volume (MTV) for each lesion, patient-based total MTV, and total-lesion glycolysis (TLG) were analyzed. Semiquantitative parameters were normalized to the liver and blood pool by tumor-to-background ratios (TBRs) and contrast-to-noise ratios. Therapy response was assessed using Deauville criteria. Results: Overall, 191 lymphoma lesions were analyzed. SUV_max demonstrated a trend toward a statistically significant increase in scans with reduced activity at 1 h after injection (6.28 ± 5.87 for RS vs. 7.76 ± 6.69 for 0.125 MBq/kg; P = 0.07) and 2 h after injection (7.14 ± 7.16 for RS vs. 8.67 ± 7.62 for 0.125 MBq/kg; P = 0.13). SUV_peak, SUV_mean, MTV, and TLG did not significantly differ between the reduced injected activities and the RS. The coefficient of variation for the liver increased significantly with decreasing injected activities (P < 0.01). The TBR for the liver did not differ significantly, whereas the TBR for the blood pool was significantly higher only for the lowest injected activity (P < 0.01) at 2 h after injection. The contrast-to-noise ratio significantly decreased with reduced activities. Deauville scores did not differ significantly, up to a dose of 0.25 MBq/kg at 1 h after injection and a dose of 1.0 MBq/kg at 2 h after injection. Below this limit, we noted significantly lower Deauville scores for reduced injected activities (P < 0.01). Conclusion: Reduction of injected activities with [¹⁸F]FDG TB PET/CT for therapy response assessment in patients with lymphoma may be possible and does not result in significant differences in MTV, TBR, or TLG. SUV_max and Deauville scores were comparable to the RS to a lower limit of 0.25 MBq/kg at 1 h after injection and 1.0 MBq/kg at 2 h after injection.

PET/CT has become an established imaging method to assess therapy response in Hodgkin lymphoma (HL) and non-HL (1,2). International guidelines recommend the use of [¹⁸F]FDG PET/CT for interim and end-of-treatment assessment for HL and diffuse large B-cell lymphoma (3–6). HL most commonly occurs during the second and third decades of life. Patients with advanced-stage disease are potentially curable with state-of-the-art therapeutic regimens (7). In addition, young patients (<20 y) with HL have a greater than 25% risk of developing a secondary malignant neoplasm within 30 y after diagnosis (8). Therefore, reducing radiation exposure is of particular interest, as this patient population is relatively young and undergoes multiple consecutive scans during their lifetime. Low-dose imaging protocols would also benefit children and pregnant women (9).

Next-generation PET technology has resulted in the development of long–axial-field-of-view (LAFOV), including total-body (TB) scanners (10,11). Along with their ability to image an entire patient in a single bed position (12), LAFOV and TB PET/CT have increased signal-collection efficiency, which allows them to provide high-quality images with low-dose protocols in children and adults (13–15). In a study of pediatric patients with lymphoma, low-dose [¹⁸F]FDG TB PET/CT yielded high-quality images, with a lower limit of 0.37 MBq/kg (16). Moreover, case reports have shown that [¹⁸F]FDG injections of 0.3 MBq/kg were possible during pregnancy for staging and restaging lymphoma, resulting in a reduction of PET-derived fetal doses (17,18).

Although initial studies and case reports have indicated that TB PET/CT can provide good image quality with low-dose protocols, some concerns remain unaddressed. Specifically, background noise levels increase substantially with low-dose protocols, resulting in abnormal lesion quantification (19,20). A previous study reported an SUV_max of up to 20% higher and the occurrence of false-positive lesions due to background noise on short–axial-field-of-view (SAFOV) PET systems (21). Other studies have found that the conspicuity of small, PET-positive lesions was impaired with low-dose protocols (20,22). It remains unclear if therapy response assessment by Deauville score in low-dose PET (i.e., reduced injected activities) is comparable to the reference standard (RS; 3.0 MBq/kg) proposed by the European Association of Nuclear Medicine guidelines (23,24).

The aim of this study was to examine differences in lower injected activities and determine a potential reduced activity level for lymphoma therapy assessment with [¹⁸F]FDG TB PET/CT at 2 different acquisition times. Furthermore, we characterized [¹⁸F]FDG lesion uptake with regard to background noise and assessed additional PET biomarkers at low-dose reconstructions.

MATERIALS AND METHODS

Patient Population, Radiopharmaceutical, and Imaging Study Design

This study was approved by the institutional review board at the University of California, Davis (IRB1470016) and performed in accordance with the Declaration of Helsinki. All patients gave written informed consent for study inclusion. In this study of prospectively collected registry-type data, 24 patients with biopsy-confirmed lymphoma were consecutively enrolled for interim (n = 22) or end-of-treatment (n = 2) [¹⁸F]FDG TB PET/CT (uEXPLORER; United Imaging Healthcare) to assess their response to therapy. Patients older than 18 y underwent a dual-time-point protocol after receiving a median dose of 3.8 ± 0.8 MBq/kg of [¹⁸F]FDG. Patients fasted for at least 6 h before injection and had a median blood glucose level of 101 ± 20 mg/dL. Standard clinical imaging was performed in list mode for 20 min, beginning 1 h after injection, as previously described (25). A second 20-min list-mode acquisition on the TB PET/CT system was performed 2 h after [¹⁸F]FDG injection, using a standard clinical protocol (26).

Low-Dose Reconstruction Parameters and Scanner Characteristics

The uEXPLORER TB PET/CT scanner has a LAFOV of 194 cm, allowing imaging of the entire body of most patients with a single bed position (12). The LAFOV scanner has high spatial resolution (∼3.0 mm) and high sensitivity (approximately 15–68 times higher than conventional PET scanners) (11).

All PET data were acquired in list mode, which enabled the creation of simulated lower-dose datasets by randomly removing detected events, while preserving the original motion and physiological uptake, as previously described by our group (13). The list-mode data from the original acquisition was divided into 5 dose levels: the RS (3.0 MBq/kg) and 4 reduced injected activities (1.0, 0.5, 0.25, and 0.125 MBq/kg). For each patient, 5 list-mode datasets were generated (RS and 4 reduced activities) and reconstructed following our standard clinical protocol: time-of-flight–based 3-dimensional ordered-subset expectation maximization with 4 iterations, 20 subsets, 2.344-mm isotropic voxel size, and 600 × 600 × 1,940 mm³ field of view, with all standard data corrections applied (without point-spread function modeling). In accordance with the clinical protocol, no postreconstruction smoothing was applied to the RS or reconstructed images with reduced activities.

Image Interpretation and Semiquantitative Analysis

Target lymphoma lesions were delineated using a semiautomated artificial intelligence (AI)–guided approach. The AI algorithm highlighted all lesions from the RS image, and a dual-certified radiologist and nuclear medicine physician edited the AI segmentation using open-source software (3D Slicer) (27,28). Additionally, a spheric, 10-mm volume of interest for mediastinal blood pool and 30-mm volume of interest for liver background were segmented in the ascending aorta and the right liver lobe, respectively. Individual lesions were then separated by connected component labeling (29). Further, 40% isocontour lesion volumes of interest were generated automatically by identifying voxels with SUV greater than 40% of SUV_max determined per target lesion, identified by the AI-guided approach (13). Lesion segmentations were generated independently for each RS image at 1 and 2 h after injection.

SUV_max, SUV_mean, and SUV_peak were recorded on both the patient and lesion levels. Lesions were also quantified by metabolic tumor volume (MTV) and total lesion glycolysis (TLG). Additionally, global lymphoma burden was characterized by the total MTV per patient. Background noise (coefficient of variation [CoV]) was characterized by the liver SUV_mean and standard deviation (Eq. 1), as previously published (19):

$$\mathrm{CoV}=\left(\frac{\mathrm{\sigma }}{\mathrm{\mu }}\right),$$ Eq. 1

where σ is the standard deviation of the background volume of interest and μ is the background SUV_mean. Tumor-to-background ratio (TBR) (Eq. 2) and contrast-to-noise ratio (CNR; Eq. 3) were defined as previously published (13,30):

$$\mathrm{TBR}=\frac{{\mathrm{SUV}}_{\max }\mathrm{\hspace{0.17em}}\left(\mathrm{lesion}\right)}{{\mathrm{SUV}}_{\mathrm{mean}}\mathrm{\hspace{0.17em}}\left(\mathrm{background}\right)}\mathrm{ },$$ Eq. 2

$$\mathrm{CNR}=\frac{{\mathrm{SUV}}_{\mathrm{mean}}\hspace{0.17em}\left(\mathrm{lesion}\right)-{\mathrm{SUV}}_{\mathrm{mean}}\mathrm{\hspace{0.17em}}\left(\mathrm{background}\right)}{{\mathrm{SUV}}_{\mathrm{mean}}\mathrm{\hspace{0.17em}}\left(\mathrm{background}\right)}.$$ Eq. 3

OsiriX MD version 13.0 (Pixmeo SARL) was used for semiquantitative image analysis.

Therapy Response Assessment

Lymphoma treatment response was assessed visually by 2 dual-certified radiologists and nuclear medicine physicians in consensus using the 5-point Deauville criteria and Lugano classification of PET/CT treatment response (progressive disease, stable disease, partial metabolic response, or complete metabolic response) (23,31). Visual assessment and quantification were possible for all scans; reconstruction was performed using an appropriate workstation and software, as previously described (32). Baseline [¹⁸F]FDG PET/CT scans were available for all patients. [¹⁸F]FDG uptake was compared with the mediastinal blood pool and liver activity in each reconstruction for both time points.

Statistical Analysis

Statistical analysis was performed using Prism version 10 (GraphPad). Semiquantitative data are presented as mean ± SD or median and range. After checking for normality, paired t tests were used to compare low-dose reconstructions with the full-dose clinical RS. Comparisons yielding P values of less than 0.05 were considered statistically significant. Fleiss κ was used to compare low-dose Lugano response groups to the RS. A κ greater than 0.6 was considered comparable between the reduced injected activities.

RESULTS

In total, 24 subjects were included in this study (2 with diffuse large B-cell lymphoma, 4 with follicular lymphoma, 14 with HL, 3 with marginal zone lymphoma, and 1 with unspecified non-HL). Overall, 191 lymphoma lesions were identified, segmented, and analyzed in all reconstructions for both acquisition times (1 and 2 h after injection). Patient characteristics are shown in Table 1.

TABLE 1.

Characteristics of Study Patients (n = 24)

Characteristic	Value
Sex
Male	13 (54)
Female	11 (46)
Age (y)	46 (19–73)
Weight (kg)	79 (53–127)
Height (cm)	169 (157–201)
BMI (kg/m2)	27 (19–45)
Lymphoma subtype
DLBCL	2 (8)
FL	4 (17)
HL	14 (58)
MZL	3 (13)
Other	1 (4)

BMI = body mass index; DLBCL = diffuse large B-cell lymphoma; FL = follicular lymphoma; MZL = marginal zone lymphoma.

Qualitative data are number and percentage; continuous data are median and range.

Target Lesion Quantification in Low-Dose TB PET/CT

SUV_max demonstrated a trend toward a statistically significant increase in scans obtained with reduced activity at 1 h after injection (6.28 ± 5.87 for RS vs. 7.76 ± 6.69 for 0.125 MBq/kg; P = 0.07) and 2 h after injection (7.14 ± 7.16 for RS vs. 8.67 ± 7.62 for 0.125 MBq/kg; P = 0.13). SUV_peak and SUV_mean were comparable between low-dose reconstructions and the RS for both acquisition times (P = 0.99). MTV did not significantly differ between low-dose reconstructions and the RS (Fig. 1).

FIGURE 1.

Violin plots of SUV_max (A), SUV_peak (B), SUV_mean (C), and MTV (D) per lesion for 1 and 2 h after injection of [¹⁸F]FDG for TB PET/CT in low-dose reconstructions. ns = not significant.

Patient-Based TB PET/CT Evaluation

Patient-based analysis confirmed the results of the lesion-based evaluation. Total MTV was comparable between low-dose reconstructions and the RS for both acquisition times (P = 1.00). TLG also did not differ significantly between low-dose reconstructions and the RS at 1 h after injection (14.77 ± 94.03 for RS vs. 15.01 ± 95.98 for 0.125 MBq/kg; P = 0.99) and 2 h after injection (16.59 ± 119.37 for RS vs. 16.75 ± 120.91 for 0.125 MBq/kg; P = 0.99). However, the CoV significantly increased with reduced injected activity at 1 h after injection (0.09 ± 0.01 for RS vs. 0.43 ± 0.06 for 0.125 MBq/kg; P < 0.01) and 2 h after injection (0.12 ± 0.02 for RS vs. 0.54 ± 0.08 for 0.125 MBq/kg; P < 0.01) (Fig. 2).

FIGURE 2.

Mean ± SD of total MTV (TMTV; A), TLG (B), and liver CoV (C) for 1 and 2 h after injection of [¹⁸F]FDG for TB PET/CT. *P < 0.05; ns = not significant.

Normalized Lesion Characteristics in Low-Dose and Delayed TB PET/CT

Lesion uptake normalization to liver background did not reveal statistically significant differences between low-dose reconstructions and the RS, despite absolute increases associated with the administration of reduced activity 1 h after injection (2.74 ± 2.55 for RS vs. 3.32 ± 2.84 for 0.125 MBq/kg; P = 0.12) and 2 h after injection (3.85 ± 3.86 for RS vs. 4.43 ± 3.91 for 0.125 MBq/kg; P = 0.42). However, the CNR decreased significantly with reduced injected activity at 1 h after injection (7.95 ± 17.67 for RS vs. 1.62 ± 3.66 for 0.125 MBq/kg; P < 0.01) and 2 h after injection (11.83 ± 19.83 for RS vs. 2.37 ± 4.18 125 MBq/kg; P < 0.01).

The TBR for the blood pool (TBR_BP) demonstrated a trend toward a statistically significant difference in images acquired 1 h after injection (3.53 ± 3.26 for RS vs. 4.35 ± 3.67 for 0.125 MBq/kg; P = 0.07). At 2 h after injection, the TBR_BP was significantly higher for the lowest-dose reconstruction (6.47 ± 6.56 for RS vs. 10.05 ± 10.45 for 0.125 MBq/kg; P < 0.01). The other low-dose reconstructions did not reveal statistically significant differences in TBR_BP at either acquisition time. The CNR for the blood pool was significantly lower in the low-dose reconstructions compared with the RS at 1 and 2 h after injection (P < 0.01), except for the 1.0-MBq/kg dose at 1 h after injection (12.50 ± 20.65 for RS vs. 9.53 ± 15.85 for 1.0 MBq/kg; P = 0.11) (Fig. 3).

FIGURE 3.

Mean ± SD of TBR (A and C) and CNR (B and D) with respect to liver or blood-pool activity of low-dose reconstruction for 1 and 2 h after injection of [¹⁸F]FDG for TB PET/CT. *P < 0.05; ns = not statistically significant.

Therapy Response Assessment in Low-Dose and Delayed TB PET/CT

There was no significant difference in Deauville scores at 1 h after injection with 3.0, 1.0, 0.5, or 0.25 MBq/kg of [¹⁸F]FDG. However, the Deauville score for the 0.125 MBq/kg reconstruction was significantly lower than that for the full-dose RS at 1 h after injection (P < 0.01) (Fig. 4). This was also reflected in the Lugano response groups. Up to a lower limit of 0.25 MBq/kg, low-dose reconstructions yielded good to perfect agreement with the RS. However, for the 0.125 MBq/kg reconstruction, the agreement was only fair (κ, 0.29; Supplemental Table 1; supplemental materials are available at http://jnm.snmjournals.org). The proportion of patients with a complete metabolic response increased in the lowest-dose reconstruction, from 50% with the RS to 79% with 0.125 MBq/kg (Fig. 5).

FIGURE 4.

Mean ± SD Deauville score measuring therapy response in low-dose reconstructions for imaging performed 1 and 2 h after injection. *P < 0.05; ns = not significant.

FIGURE 5.

Treatment response groups according to Lugano criteria at 1 h after injection (A) and 2 h after injection (B). CMR = complete metabolic response; PD = progressive disease; PMR = partial metabolic response.

At 2 h after injection, only the low-dose reconstruction simulating an injected activity of 1.0 MBq/kg yielded a Deauville score that was not significantly lower than that associated with the full-dose RS (P = 0.33). All other low-dose reconstructions revealed significantly lower Deauville scores compared with those of the RS (P = 0.02 for 0.5 MBq/kg vs. RS; P = 0.01 for 0.25 MBq/kg vs. RS; P < 0.01 for 0.125 MBq/kg vs RS) (Fig. 4). We noted good Lugano group agreement between 3.0 and 1.0 MBq/kg. At 0.5 MBq/kg, agreement decreased to fair and was moderate or poor for the 2 lowest-dose reconstructions (Supplemental Table 1). The percentage of patients with complete metabolic response increased from 29% with the RS to 69% with the lowest-dose reconstruction (Fig. 5).

Subgroup analysis for the largest cohort (patients with HL) yielded Deauville scores similar to those obtained in the overall response assessment (Supplemental Fig. 2).

DISCUSSION

This study aimed to assess the lower limit of reduced injected activities in [¹⁸F]FDG TB PET/CT for lymphoma treatment response analysis at 2 acquisition times. Although the guidelines use 3.0 MBq/kg as the RS, we demonstrated that injected activities can be reduced to 8% of the RS with TB PET/CT in patients with lymphoma without loss of quantification and therapy response assessment. This is of particular interest for young adults who undergo multiple consecutive PET studies and for children and pregnant women (8,9).

In our cohort, images simulating lower administered activities of [¹⁸F]FDG for TB PET/CT were reconstructed, as previously described (13). This allowed us to segment target lymphoma lesions using the clinical RS (3.0 MBq/kg) and transfer them to the low-dose images for the same patient and imaging time point (24). Although previous studies investigated whether delayed imaging (e.g., at 2 h after injection) increased the diagnostic accuracy of PET in patients with lymphoma, this study emphasized the use of lower administered activities (33,34). Moreover, previous work indicated that a reduction of injected activities to 0.5 MBq/kg in children was possible with TB PET/CT performed 2 h after injection. (13). Therefore, we aimed to apply a simulated low-dose protocol to imaging performed at 1 and 2 h after injection in patients with lymphoma.

In our study, semiquantitative lymphoma lesion parameters, such as SUV_max, SUV_mean, SUV_peak, and MTV did not significantly differ in low-dose reconstructions compared with the full-dose RS. Contrarily, another study on SAFOV PET systems reported an increase in SUV_max when positron emission was low (14). Previous work showed that TB PET’s high signal-collection efficiency can better reduce the noise-derived bias of the SUV_max than can SAFOV PET, making further dose reductions possible (10,11,13,35). This explains why the SUV_max did not increase significantly in our simulated low-dose PET images at either acquisition time. However, we found that the CoV increased significantly with reduced activities (Fig. 2), which confirmed a significant increase in noise with images obtained with low doses.

Additionally, patient-based biomarkers (i.e., total MTV and TLG) did not differ significantly in low-dose reconstructions compared with the RS. The TBR of the liver did not reveal any statistically significant differences. TBR_BP values were similar at 1 h after injection. At 2 h after injection, the TBR_BP increased with decreasing injected activities, resulting in a significantly higher TBR_BP only for 0.125 MBq/kg compared with the RS (P < 0.01). This is consistent with findings on the use of lower doses with other LAFOV PET/CT scanners. In a study of 26 patients, Calderón et al. (36) reported no statistical difference in TBR for the same low-dose reconstructions that were applied in our study. However, in our study, TBR_BP increased at 2 h after injection, which was the result of [¹⁸F]FDG washout in late acquisition times (37). Blood-pool [¹⁸F]FDG washout is a well-known phenomenon and may have an impact on visual and quantitative interpretation (38).

Increased noise in low-dose reconstructions was confirmed by significantly reduced CNRs for both the liver and blood pool (Fig. 3). This increase in noise resulted in significantly lower Deauville scores for the lowest injected activity 1 h after injection and for 0.5–0.125 MBq/kg at 2 h after injection when compared with the RS. Figure 6 displays images of a 24-y-old patient with HL who underwent interim [¹⁸F]FDG TB PET/CT. The red arrows show remaining bilateral neck lymphoma lesions. Lesion uptake was higher than the liver background with the full-dose RS at both acquisition times, resulting in a Deauville score of 4. At 1 h after injection, lesion uptake was lower than the liver background because of increased noise in images obtained using 0.125 MBq/kg, whereas lesion uptake was already lower than the liver background in images acquired using 0.5 MBq/kg at 2 h after injection, resulting in downstaging the Deauville score to 3. This could have affected the treatment response assessment (as measured using Lugano group classification) and impacted clinical decisions (23). Therefore, we suggest lower limits for dose reduction: 0.25 MBq/kg for images taken 1 h after injection and 1.0 MBq/kg for images obtained 2 h after injection imaging.

FIGURE 6.

[¹⁸F]FDG TB interim PET/CT images of 24-y-old patient with HL. Maximum-intensity projection and [¹⁸F]FDG-positive cervical lymphoma lesions (red arrowheads) are shown at 1 and 2 h after injection at all activities evaluated.

Our finding of significant downstaging in Deauville score in low-dose scans, indicating an artificial improvement in treatment response (measured using Lugano group classification), is in agreement with previous work by Hornnes et al. (39), who found that acquisition times could be reduced by 50% using a digital SAFOV system. Thus, TB PET/CT demonstrated that Deauville score assessment was maintained to 8.3% of the RS, indicating that TB PET/CT systems may be less affected by noise than are conventional SAFOV systems (11,40).

Nevertheless, some limitations need to be addressed. First, only 24 patients met the inclusion criteria and were enrolled in our study. Moreover, our analysis included multiple lymphoma subtypes. However, 191 lymphoma lesions were quantifiable, more than the number reported in comparable studies (13,39–41). Second, patients’ median body mass index was above the normal range, which may have negatively affected lesion quantification. In a cohort with a normal body mass index, we expect that further reductions of [¹⁸F]FDG activities may be possible. Third, despite high signal-collection efficacy, image noise, especially for the lowest injected activities at 2 h after injection, was high, reflected by a significant increase of the CoV and decrease of the CNR, indicating that reconstructed images may benefit from smoothing or parameter optimization. Additionally, at 2 h after injection, increased washout resulted in low blood-pool activity, limiting its utility for TBR normalization and Deauville score assessment in low-dose reconstructions. The reduction of injected activities with a delayed-imaging protocol may be challenging; therefore, further consideration is needed before implementation.

If low-dose protocols were implemented in standard clinical care, then the reconstruction parameters would likely need to be optimized for these modified protocols. We did not perform reconstruction parameter optimization in this study; however, in our prior work, we found that readers preferred only minor modifications to low-dose protocols and that these modifications were limited to very low dose levels, corresponding to 0.25 and 0.125 MBq/kg in the pediatric population (13). Reconstruction parameter optimization for low-dose acquisitions would most likely incorporate postreconstruction smoothing into the images. This optimization would improve image noise and CoV metrics; however, this would not improve lesion contrast and therefore may not allow a fair comparison of lesion metrics (e.g., TBR, Deauville score) (42).

Lastly, our study investigated the potential of lower injected activities in patients with lymphoma by simulating low-dose images from the list-mode data. Results in a real clinical setting may differ because of lower random rates and less dead time (11,13). Additionally, the number of patients included was low due to a complex study protocol requiring PET scans 1 and 2 h after injection, and the patients assessed for interim and end-of-treatment response were imbalanced. Moreover, long-term follow-up data would provide more-detailed information about the patients’ remission status and long-term prognosis. A follow-up, prospective head-to-head comparison with different injection protocols and scanners can be used to assess detectability and quantification capability in low-dose [¹⁸F]FDG TB PET/CT. However, the design of this study allowed us to perform lesion segmentations using the RS images, which avoided the bias of imaging at 2 different time points, using different patient cohorts, or accounting for patient movement (43). Future prospective studies with homogenous lymphoma subgroups and different injection protocols are necessary to address these limitations.

CONCLUSION

Reduction of the injected activity for therapy response assessment with [¹⁸F]FDG TB PET/CT in patients with lymphoma may be possible. Semiquantitative values were comparable to the RS up to a certain limit. We identified lower limits of 0.25 and 1.0 MBq/kg for [¹⁸F]FDG TB PET/CT performed 1 and 2 h after injection, respectively. Lower injected activities showed significantly reduced Deauville scores in interim and end-of-treatment [¹⁸F]FDG TB PET/CT, which may have overestimated patients’ response to therapy. TB PET/CT has the potential to reduce PET/CT-mediated radiation exposure and possibly reduce the risk of developing secondary malignancy for vulnerable patient populations, especially young patients with lymphoma, who undergo multiple consecutive PET/CT scans during their lifetime.

DISCLOSURE

Research reported in this publication was supported by the National Institutes of Health (R01CA249422), In Vivo Translational Imaging Shared Resources (NCI P30CA093373), and the Fred and Julia Rusch Foundation for Nuclear Medicine Research and Education. Hande Nalbant receives funding from a United Imaging Health Fellowship Gift. Axel Rominger has received research support and speaker honoraria from Siemens. Lorenzo Nardo is principal investigator of a service agreement with United Imaging Healthcare, site principal investigator of clinical trials supported by Novartis AG, and principal investigator of clinical trials supported by Telix Pharmaceuticals, Lantheus, and GE HealthCare. Lorenzo Nardo and Ramsey Badawi are investigators of a clinical trial supported by Lilly. Ramsey Badawi received research support from United Imaging Healthcare during the course of this work. UC Davis has a revenue-sharing agreement with United Imaging Healthcare. No other potential conflict of interest relevant to this article was reported.

KEY POINTS

QUESTION: Can we reduce the injected activity of [¹⁸F]FDG in TB PET/CT for therapy response assessment in patients with lymphoma?

PERTINENT FINDINGS: Injected activity can be reduced with [¹⁸F]FDG TB PET/CT in lymphoma treatment response assessment to 0.25 MBq/kg for imaging performed 1 h after injection and to 1.0 MBq/kg for imaging performed 2 h after injection.

IMPLICATIONS FOR PATIENT CARE: Dose reduction of injected activity in patients with lymphoma is possible with TB PET/CT, reducing the radiation burden for young or pregnant patients, who undergo multiple consecutive PET scans during their treatment for response assessment.