PET/CT in therapy control of infective native aortic aneurysms

Infective native aortic aneurysms (INAA) are aneurysms arising from infection of the aortic wall. Treatment is demanding with 5-year survival rates between 53 and 55%. The aim of our study was to evaluate the usefulness of 18F-fluorodeoxyglucose positron emission tomography/computed tomography (PET/CT) in the long-term monitoring of patients with proven INAA. Fifty-three PET/CT were performed in 15 patients with INAA in this single-center retrospective cohort study and retrospective analysis of prospectively collected Vascular Graft Cohort Study (VASGRA) data. Median metabolic activity (as measured by maximum standardized uptake value, SUVmax) of the aneurysms at the initial PET/CT was high (6.8 (IQR 5.7–21.8)), and lower at the last PET/CT prior to the end of antimicrobial therapy (3.9 (IQR 2.7–6.8); n = 11) as well as in the first PET/CT after the end of the treatment (3.9 (IQR 3.0–4.4);n = 6). Compared to the course of C-reactive protein alone, PET/CT provided different (> 20% difference in trend) or altering (opposed trend) information on the course of disease in at least 14 comparisons (56%) in 11 patients (73%). The one-year and five-year freedom from all-cause lethality was 92% (95% confidence interval 57%-99%). As compared to the course of C-reactive protein, PET/CT provides different and occasionally altering information in therapy control of INAA.


Scientific Reports
| (2021) 11:5065 | https://doi.org/10.1038/s41598-021-84658-z www.nature.com/scientificreports/ number 2018-01904). All procedures were performed in accordance with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Diagnosis of INAA was made in an overall appraisal of clinical presentation (pain, fever, sepsis), laboratory (positive microbiological culture of aortic/aneurysmatic wall, presence of bacteria in thrombus or blood culture; elevation of inflammatory markers such as C-reactive protein and leucocytes) and imaging 5 . Diagnosis of secondary VGI due to placement of stentgrafts relied on the MAGIC criteria 12 .
In all prospectively enrolled patients, criteria for the termination of antimicrobial therapy were a combination of the absence of clinical features of infection, normal CRP, and reduced metabolic activity in PET/CT. Cure of INAA was defined as a combination of absence of clinical features, normal CRP, and reduced metabolic activity in PETCT at least three months after termination of antibiotic therapy (the latter only in the prospectively enrolled patients).
For retrospective inclusion of patients, we performed a retrospective chart review in all patients with suspected INAA, who were examined with PET/CT between the years 2005 and 2018, if the term "mycotic aneurysm", "infective aneurysm", or "infected aneurysm" was mentioned in the written PET/CT report, and was found to refer to an aneurysm of the thoracic, abdominal or pelvic arteries.
PET/CT examinations and patient follow-up. The study design for all consecutively and prospectively enrolled patients included consecutive PET/CT scans, at baseline, during follow-up on antimicrobial therapy, and at the end of antimicrobial treatment. If feasible, a control PET/CT three months after the end of antimicrobial therapy was performed, to document continuous cure or possible signs for recurrence of infection. For all retrospectively enrolled patients, PET/CT was performed depending on the clinical situation of the patient; the reasons for patient's referral are given in the results section. The exact time points of follow-up examinations with regard to the baseline are demonstrated in Fig. 1. For all patients, baseline was defined as first PET/CT scan at diagnosis of INAA.
We performed clinical follow-up of all patients by reviewing electronic patient charts. Patient data were recorded at the time of imaging, and at the last recorded clinical visit (recorded until March 2020).
Recorded data at baseline and at follow-up included patient demographics, clinical information, laboratory data (e.g. level of CRP and leucocyte count), results of microbiology, results from other diagnostic procedures, and information about treatment.
PET/CT data acquisition and image analysis. Five different types of PET/CT scanners were used within the study period between 2005 and 2018, i.e. a Discovery ST16, a Discovery VCT, two Discovery MI a Discovery 690 and a Discovery 710 (all GE Healthcare, Waukesha, WI). Body weight, height, and blood glucose level were measured prior to imaging (blood glucose levels < 12 mmol/l were accepted 13 ). All PET/CT examinations followed basic study protocols: patients fasted for at least four hours, body-weight adjusted intravenous injection of FDG 14 , standardized uptake time of 60 min in supine position, non-enhanced CT scans for attenuation correction, data acquisition with arms overhead whenever possible.
All PET/CT examinations were independently analysed by two double board certified radiologists and nuclear medicine physicians, blinded to all clinical patient data. Readers determined the FDG uptake pattern in the aneurysm or graft to be focal or diffuse 15 . A consensus reading was performed if results differed. Furthermore, both readers quantified the FDG-uptake in the aneurysm or in the vascular graft by measuring the maximum standardized uptake value (SUVmax) in the aneurysm or graft as well as in the liver and mediastinal bloodpool (the latter as background for reference). In case SUVmax measurements were not identical among both readers concerning the measurements in the aneurysm or graft, a consensus reading was performed; in case of discordant SUVmax values for the background measurements, the mean was calculated. To compensate for differences in the sensitivity of the different PET/CT scanner generations between the years 2005 and 2018, and resulting differences in quantitative PET parameters (i.e. SUVmax), we calculated relatively scanner-independent FDG uptake ratios (i.e. SUVratio was defined as uptake in the aneurysm of graft in relation to the mediastinal blood pool). To determine differences between the course of C-reactive protein and SUVratio we calculated trends between consecutive examinations, and defined an arbitrary cut-off of > 20% between differences in trends as relevant.
Furthermore, for all baseline and follow-up PET/CT examinations, readers determined whether a INAA or a VGI was present or not, using a visual 4-point grading score. Score 1 (no signs for infected aortic aneurysm/VGI) and score 2 (most likely a non-infected aneurysm/VGI) were considered negative for infected aortic aneurysm, while score 3 (suspicion of infected aortic aneurysm/VGI) and score 4 (clear signs of infected aortic aneurysm/ VGI) were considered positive for infection.

Statistical analyses.
Variables were expressed as median and interquartile range (25th, 75th percentiles) or percentages. Kaplan-Meier estimates were used to describe survival at 1 and 5 years. We compared sizes and visual grading scores of INAA from initial and last CT scans using nonparametric pairwise Wilcoxon signedrank test. Mixed-effects multilevel linear regression were applied to analyze the individual changes of SUVmax, SUVratio and CRP. Statistical analysis were performed using commercially available software (Stata/SE, Version 15.1, StataCorp, College Station, Texas).

Results
Patient population. Seven out of 21 (33%) patients with confirmed INAA were prospectively enrolled, while 14 (66%) patients were retrospectively included. Five out of 14 retrospective patients were excluded due to lack of written consent and one out of the seven prospective patients was excluded due to lack of clinical followup after the first PET/CT. At the time of the initial PET/CT examination, patients had a median age of 61 years (IQR 54-85), one patient (7%) was female, six (40%) were smokers or had a history of smoking, two (13%) patients were diabetic, and six (40%) had renal insufficiency. The diagnoses of INAA were confirmed by blood culture (n = 11; 73%), culture or PCR from tissue obtained during surgical revision (n = 2; 13%) and serology (n = 2, 13%) ( Table 2).
The first follow-up PET/CT (112 (IQR 95-721) days after the first PET/CT) was performed in all patients, and 13 (87%) patients received antimicrobial therapy at the time (in one patient antimicrobial therapy was started after the second PET/CT in another therapy was already terminated). The median visual 4-point grading score www.nature.com/scientificreports/ decreased to 3 (3 × score 4, 8 × score 3, 2 × score 2, 2 × score 1). The first follow-up showed an overall decrease in SUVmax (5.5 (IQR 3.0-8.2) and SUVmax ratio (2.2 (IQR 1.6-2.9)); SUVmax decreased in 13 and increased in two patients, SUVmax ratio decreased in 10 and increased in five patients (Figs. 1, 2, 3, Table 3). In 11 patients, PET/CT was performed prior to the termination of the antimicrobial therapy. Compared to background activity, PET/CT showed focally increased FDG uptake in nine patients (82%) and diffuse increased FDG uptake in two patients (18%) with SUVmax 3.9 (IQR 2.7-6.8) and SUV ratio 1.7 (IQR 1.2-2.9). The FDG uptake in the aneurysm/graft was higher than blood pool background in eight (73%) patients ( Fig. 2) and higher than liver background in seven (64%) patients. The median CRP was 12.0 mg/L (IQR 3-218). Of note, the median time difference between the last PET/CT prior to the termination of antimicrobial therapy and the actual end of antimicrobial therapy was a median 30 days (IQR 18-1059), and six examinations (55%) were considered to be directly linked to the decision to end treatment.
The comparison of the last PET/CT prior to the termination of antimicrobial therapy to the first PET/CT after termination of antimicrobial therapy (feasible in six patients) showed less focal FDG-uptake (from five to four) www.nature.com/scientificreports/ and decreasing FDG uptake in three patients for SUVmax and in two patients for SUVratio, while the uptake increased in three patients for SUVmax and in four patients for SUVratio (SUVmax from 3.8 to 3.9 (IQR 2.5-4.4 and 3.0-4.4) and SUV ratio from 1.6 to 1.7 (IQR 0.9-1.9 and 1.5-4.1)). Notably, the two patients with a significantly rising SUV ratio (from 0.7 to 2.0 and from 0.7 to 1.2; patient 9 and 13 in Table 3) had no signs of infection at the last clinical follow-up, which was more than five and more than eight years after the last PET/CT and without antimicrobial therapy. None of these patients had clinical signs of new VGI at the last clinical follow-up.
Sizes of INAA, measured on CT images, decreased significantly by 1.45 cm on average from the initial to the last examination (p = 0.016, Table 3). In parallel, visual grading scores for INAA decreased by 1.2 on average (p < 0.01, Table 3).

Discrepancies between trends of SUVratio and CRP.
With 53 PET/CT performed in 15 patients we were able to calculate 38 trends of SUVratio and 25 trends of CRP (immediate data of CRP missing for 13 PET/ CT).
The decrease or increase of SUVratio was paralleled by decreasing or increasing CRP values in 16 of 25 (64%) trends, while it was opposed in nine (36%) examinations. Nineteen comparisons (76%) showed either opposed trends (as mentioned above, n = 9) and/or differences between trends of CRP and SUVratio > 20%. For these trends, PET/CT provided a conclusive explanation for the difference in five comparisons (26%) in three patients (i.e. additional foci of infection with different response to therapy: patient 05, 11, and 14 in Table 3 and Fig. 2). No apparent reason for high differences in trend could be determined for four comparisons (21%) in four patients (i.e. three opposed trends between first and second PET/CT in patient 04, 07, and 11 (Table 3) and a high 46% increase in CRP paralleled by a low 11% increase of SUVratio in patient 09 in Table 3). In five comparisons of five patients, CRP values showed very good or complete responses to therapy at the time of the first follow-up PET/CT, which were paralleled by less pronounced partial responses in SUVratio (patient 02, 03, 06, 08, and 15); further clinical follow-up could not determine which methods better described the actual clinical response to therapy (as all other clinical parameters defining cure of INAA were normal/unchanged). The final five comparisons of differences between trends of CRP and SUVratio which were > 20%, were rated to be errors as the values were very low and partly normal, i.e. between the fourth and fifth PET/CT of patient 02 (Table 3), CRP values decreased by 80% from 2.1 to 1.3, which was paralleled by a 20% increase in SUV ratio from 1.65 to 1.84.
Hence, PET/CT provided additional or altering information on the course of disease in at least 14 comparisons (56%) in 11 patients (73%), as compared to CRP alone. Patient outcome. Patients were clinically followed for a median of 1496 days (IQR 961-3187 days) after their initial PET/CT examination and for a median of 1005 days (IQR 683-3109 days) after their last PET/CT Table 2. Data on microbiology and antimicrobial of the final study population in patients with proven infective native aortic aneurysms. Pat: Patient number; na, not applicable. *Microbiology resulted from blood culture (n = 11), culture or PCR from tissue obtained during surgical revision (n = 2) and serology (n = 2).  The two patients (13%) on continuous antibiotic therapy, showed higher metabolic activity in the INAA (SUVmax 7.9 and 4.7, SUVratio 2.4 and 1.6) than the mean metabolic activity of those patients that were scanned prior to the termination of antimicrobial treatment (as mentioned above: SUVmax 3.9 (IQR 2.7-6.8) and SUV  Tables 1, 2 and 3) presented with chest pain and signs for infection (C-reactive protein 122 mg/L). The initial PET/CT examination showed strongly increased FDGuptake in the wall of an aortic arch aneurysm and no other infectious foci. Both PET/CT readers suspected a infective native aneurysm; which was clinically confirmed (Streptococcus agalactiae in blood cultures). At first PET/CT follow-up (ongoing antimicrobial therapy, after thoracic endovascular repair with a Conformable GORE TAG Thoracic Endoprosthesis and debranching) a strong residual focal FDG-uptake was detected adjacent to the graft, in line with a secondary vascular graft infection. The second PET/CT follow-up (ongoing antimicrobial therapy), showed only very faintly increased FDG-uptake adjacent to the graft. However, a new pneumonia in the right lower lobe was incidentally detected. The patient died 60 days after the last PET/CT due to a gastrointestinal bleeding not related to the thoracic infective native aortic aneurysm.   2-2.9)). The conditions of these two patients were determined to be stable at the last clinical visit 261, and 729 days after the last performed PET/CT. One patient (7%) showed recurrent signs of infection at the last clinical visit, which were attributed to an infected pancreatic cyst. All other patients (n = 10, 67%) did not show any signs of infection at the last clinical visit and all were without antimicrobial therapy (Table 1).

Discussion
To the best of out knowledge, this is the first study investigating the role of PET/CT in therapy control of INAA. Our study results show: (i) PET/CT adds additional information in therapy control of INAA. (ii) Metabolic activity in the aneurysms remains slightly elevated after the end of antimicrobial therapy and therefore should not be mistaken for persistent infection.
All patients of the present study were treated with open and/or endovascular repair, and as this treatment in INAA frequently leads to secondary graft infections, we may compare the present study results to the findings of three previous publications 9-11 on therapy control with PET/CT in VGI. In line with these studies 9-11 , we found that in addition to laboratory and clinical information, consecutive PET/CT examinations were a valuable source of information for treatment monitoring, often displaying additional and sometimes even opposed results to clinical and laboratory parameters.
Previous studies have documented a correlation between the course of CRP and SUVmax in therapy control of patients with VGI 10 . In the present study, only a minority of comparisons (i.e. 24%) of CRP and the metabolic activity in the aneurysm showed a similar course in therapy control of patients with INAA. The majority of cases displayed large discrepancies between trends of CRP and metabolic activity or even opposed courses. The latter, may partly be explained by additional foci of infection with different response to therapy, which was in line with results of a previous study, showing correlation between CRP and SUVmax only in a subpopulation of patients with VGI without additional infectious foci 9 .
However, we could identify another group of comparisons with no apparent reason for the large discrepancies between trends of CRP and metabolic activity, which may indicate, that CRP and metabolic activity in the aneurysm are two unrelated identities, which may have to be evaluated independently, when treating patients with INAA. Because of such discrepancies and the overall severity of INAA in general, we believe that in all patients with INAA, decisions for further therapy should always be based also on other parameters such as clinical aspects (general health status, fever, clinical signs of INAA and/or graft infection), and other laboratory parameters (CRP, ESR, leucocytes).
Furthermore, we documented in line with previous reports [16][17][18] , that the metabolic activity in INAA is initially high or very high, and comparable to the initial presentation of VGI in PET/CT [9][10][11]15,[19][20][21][22][23][24][25] . Under antimicrobial therapy the metabolic activity generally decreases [9][10][11] . However, the metabolic activity remains above background level before and even after the end of medical treatment; despite the fact that we did not observe any cases of recurrent infection. High rates of residual metabolic activity were also observed in patient populations with VGI at the end of medical treatment. Husmann et al. 10 observed a complete metabolic response to treatment in only 33% of patients with VGI, while we could not observe a complete response in any of our patients. The reason for this difference remains unclear, and may possibly be due to differences in the course of two similar but yet different diseases, or it may be due to bias as our study cohort was considerably smaller in number. Notably, residual metabolic activity (i.e. SUVmax values slightly above background level) in the aneurysms after the end of medical treatment should not be mistaken for persistence of infection in INAA, but may possibly represent a sterile ongoing inflammatory reaction, as none of our patients showed signs of recurrence during the follow-up of this series and an excellent patient outcome was observed.
Two patients died due to reasons not considered unrelated to INAA. The one-year and five-year freedom from all-cause mortality was 93%, and was much higher than in previously published studies, which found five-year survival rates between 53% 26 and 55% 7 . Whether disease monitoring with PET/CT contributed to the excellent clinical outcome in the present study still remains unanswered. However, we could not find any other obvious systematic difference, which may account for the improved clinical outcome.
Limitations of the study. The present study population is heterogeneous, with varying numbers of follow-up PET/CT examinations, as well as prospectively and retrospectively included patients, which precludes standardized PET/CT intervals. Furthermore, due to the lack of comparable data in the literature, we defined an arbitrary cut-off of > 20% between differences in trends of SUVratio and CRP as relevant.
Despite these limitations, the results of the present study appear consequential and should be confirmed in further studies.

Conclusion
As compared to the course of C-reactive protein, PET/CT provides different and occasionally altering information in therapy control of INAA. Of note, metabolic activity in the aneurysms remains slightly elevated even after the end of antimicrobial therapy and should not be mistaken for persistent infection. www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.