Infection dynamics and response to chemotherapy in a rabbit model of tuberculosis using [¹⁸F]2-fluoro-deoxy-D-glucose positron emission tomography and computed tomography

Abstract

With a host of new antitubercular chemotherapeutics in development, methods to assess the activity of these agents beyond mouse efficacy are needed to prioritize combinations for clinical trials. Lesions in Mycobacterium tuberculosis-infected rabbits are hypoxic, with histopathologic features that closely resemble those of human tuberculous lesions. Using [(18)F]2-fluoro-deoxy-d-glucose ([(18)F]FDG) positron emission tomography-computed tomography (PET-CT) imaging, we studied the dynamics of tuberculosis infection in rabbits, revealing an initial inflammatory response followed by a consolidative chronic disease. Five weeks after infection, as much as 23% of total lung volume was abnormal, but this was contained and to some extent reversed naturally by 9 weeks. During development of this chronic state, individual lesions in the same animal had very different fates, ranging from complete resolution to significant progression. Lesions that remained through the initial stage showed an increase in volume and tissue density over time by CT. Initiation of chemotherapy using either isoniazid (INH) or rifampin (RIF) during chronic infection reduced bacterial load with quantitative changes in [(18)F]FDG uptake, lesion density and total lesion volume measured by CT. The [(18)F]FDG PET uptake in lesions was significantly reduced with as little as 1 week of treatment, while the volume and density of lesions changed more slowly. The results from this study suggest that rabbits may be a useful surrogate species for evaluating novel chemotherapies and understanding changes in both PET and CT scans in human clinical trials.

Fig 1

NZW rabbits infected with M. tuberculosis strain HN878 experience a transient replicative phase, followed by chronic disease with lesions slowly increasing in density. (A) Time course of M. tuberculosis HN878 infection in the NZW rabbit lung compiled from data for untreated animals, depicting the mean bacterial burden (error bars show SD). (B) Image of H&E stained section of rabbit lung showing solid cellular lesions typical of lesions observed 4-6 weeks p.i. including peribronchial lymphoid aggregates (arrow). (C) Necrotizing tuberculomas were the predominate lesion type after 9 weeks p.i. (D) Graphic representation of the changes in lung density (HU) during the course of infection in a representative rabbit (Fig. 2). Each bar represents the total volume of the lung divided into three density ranges, with the HD lesion percentage indicated on the top of each bar. (E) HD and MD tissue volumes (mean ± SD) for 4 or 5 rabbits during 20 weeks of infection. (F) Densities of individual lesions with increasing time PI (n = 5 rabbits). Data are means and 95% confidence intervals. As the time PI increased, the mean density of the lesions increased (P < 0.0001, ANOVA and posttest for linear trend).

Fig 2

FDG PET-CT scanning of M. tuberculosis-induced lung disease reveals an early inflammatory phase of disease that partially resolves, followed by slow progression. A group of 6 rabbits was infected and monitored by scanning 6 to 8 times over 20 weeks of infection. Images are presented with the sternum toward the top and vertebrae toward the bottom of each image and are composed of three merged axial slices (3.75-mm region of lung) from a single representative animal scanned 7 times (labeled [in weeks] on the left of the figure). A pronounced inflammatory response was observed from 4 to 6 weeks with many relatively homogeneous, 2- to 4-mm-diameter, density lesions observed (arrows in CT panel). Some lesions increased in density (as expressed in HU) and size slowly and cavitated (weeks 6 to 20, downward arrowhead) or lost FDG avidity and resolved (white arrowhead). The 3-mm-diameter lesion (left arrow in week 4 panel) shrank to 1.5 mm in diameter and decreased in FDG uptake at 14 weeks but remained detectable in the parenchyma.

Fig 3

Comparison of M. tuberculosis bacterial burden after 1 week or 8 weeks of INH or RIF treatment. (A) Mean bacterial burden (log CFU/g) (error bars show SD) of the right middle lobe of rabbits treated with either INH (30 mg/kg) or RIF (24 mg/kg) for 1 or 8 weeks in comparison with control rabbits (8 to 10 weeks p.i.) or rabbits treated with vehicle alone for 8 weeks (vehicle control). The ANOVA for the 6 groups did not show significant differences, but the difference in the means of the control and the INH 8 and RIF 8 CFU were each significant (*, P = 0.0061, nonpaired t test). (B) Mean bacterial burden of excised lung lesions (1- to 3-mm diameter) of rabbits treated with either INH or RIF as described for panel A (error bars show SD). Lesions that had become fibrotic but were still identifiable as abnormal lung were included but had few remaining cultivable bacilli, and many of these samples were at or below the limit of detection. The median CFU/g lesion of the control rabbits and those treated for 1 week were not significantly different, but values were significantly different when the control was compared with INH 8-week and RIF 8-week medians (* and **, P = 0.0002 and 0.0001, respectively, by ANOVA). (C) Box-and-whisker plot (whiskers represent minimum to maximum) of pathology scores of rabbits treated with INH or RIF for 1 or 8 weeks compared to untreated controls (9 to 10 weeks p.i.) or vehicle for 8 weeks (P = 0.025 ANOVA) with only a significant difference in control compared to INH for 8 weeks (*, P < 0.05 by Bonferroni's multiple comparison).

Fig 4

Histopathologic changes in response to 1 or 8 weeks of INH or RIF treatment. (A and B) Granulomas 1 to 3 mm in diameter (both necrotizing and cellular) were observed in animals euthanized just prior to experimental treatment initiation at 9 to 10 weeks p.i. (C to F) Similar lesions with central acellular necrotic regions often located adjacent to blood vessels or airways were present after 1 week of INH (C and D) or RIF (E and F) treatment. After 8 weeks of treatment with vehicle (G and H), INH (I and J), or RIF (K and L), the drug-treated lungs contained fewer lesions than lungs from rabbits given vehicle alone (P = 0.0306 by ANOVA). Small lesions proximal to the airways were less frequent in the drug-treated animals, but some fibrotic scars (I, arrow) and cavities (L, arrow) with mineralized necrotic interiors (L) remained. Bars, 1 mm. (M) Percent area of lesions was determined for each animal using custom image analysis on 3 to 5 lung sections per animal. The median lesion area was significantly reduced in the groups treated with either RIF or INH compared to those receiving vehicle alone (P = 0.016 and P = 0.016, respectively, by Mann-Whitney test with 95% confidence intervals). Bars show the medians.

Fig 5

INH or RIF treatment of infected rabbits results in a slow resolution of both FDG PET and CT findings. (A) Groups of 4 rabbits were treated orally with 30 mg/kg INH daily for 8 weeks; images from a representative rabbit are shown. The top panel in each column shows the CT scan of a lung region (3.75 mm thick) prior to infection (NI). The remaining panels depict the CT image and fused PET-CT image of the same lung region prior to treatment initiation (PT), after 4 weeks of treatment, and after 8 weeks of treatment. Three CT dense granulomas (black arrowheads) became smaller as treatment was administered. These same lesions showed a reduction in [¹⁸F]FDG PET activity. (B) During RIF treatment, one lesion (white arrowhead) cavitated and became more dense (as determined in HU) over 8 weeks; it also increased in FDG PET avidity from 4 to 8 weeks. For the histological appearance of this lesion, see Fig. 4L. Another lesion in this field was also reduced in size and increased in HU (black arrowhead). (C) Three-dimensional images of the same animal depicting the lungs (green) and the lesions (red) indicate that the lesions were mostly decreasing in size with treatment. The white arrow indicates the cavity shown in panel B.

Fig 6

Comparison of changes in individual lesion volume and [¹⁸F]FDG PET avidity in rabbits undergoing INH or RIF chemotherapy. The lesions from a representative animal in the vehicle control (A), the INH treatment (B), and the RIF treatment (C) groups were assessed for volume and FDG uptake in serial scans. For each lesion, the y axis shows the [¹⁸F]FDG uptake, the x axis shows the treatment time when the scan was captured, and the size of the sphere depicts the lesion volume in mm³. Lesions that resolve on CT are represented as an X at the background SUV_max for the region where the lesion was located in the PET-CT. Once the chronic stage of infection is reached, dynamic fluctuations in lesion size or SUV_max are observed (A), but the majority of lesions in animals receiving treatment (B and C) decrease in size and SUV_max. (D) The glycemic index (CT volume × SUV_max) for the lesions represented in panels A to C show that treatment with INH or RIF significantly reduce the lesion's glycemic index (P = 0.015 and P = 0.035 by repeated-measures ANOVA).

Fig 7

Assessment of lesion volume, lesion density, and lesion SUV_max as biomarkers for response to treatment. Individual small nodules from serial scans for each rabbit in the drug treatment groups were analyzed by either a paired t test (2 scans) or a repeated-measures ANOVA (3 scans) with a Dunnett's posttest for changes in lesion volume. The first groups of rabbits were imaged, received 1 week of either vehicle (control), INH, or RIF (A, C, and E), and imaged at 1 week, while the second groups of rabbits were treated for 8 weeks and imaged at 0, 4, and 8 weeks (B, D, and F). The P values for Dunnett's tests that were <0.05 are presented. SUV_max in PET changed more rapidly than density or volume in the CT scan and changed significantly after 1 week of treatment. All three measurements had significant changes after 4 weeks or more of treatment.