Throughout the evidence review (18)F-fluorodeoxyglucose positron emission tomography-computed tomography will be referred to as 18F-PET or 18F-PET/CT.
Contractor Advisory Meeting (CAC)
A CAC meeting on “Pet Scans for Inflammation and Infection” was held on 11/09/22 and subject matter experts (SMEs) reviewed pertinent literature and provided input for the development of this policy. The transcripts and recording are available in the CGS Administrators website: https://www.cgsmedicare.com/partb/medicalpolicy/lcd_discussion_recordings.html. Reference to the SME input will be included throughout the policy.
Fever of Unknown Origin (FUO)
Fever of unknown origin (FUO) in adults is defined as a temperature higher than 38.3 degrees C (100.9 degrees F) lasting greater than three weeks with no obvious source despite proper investigation.2,4 There are four categories of potential etiology of FUO which are classic, nosocomial, immune deficient, and human immunodeficiency virus-related. There are four subgroups of the differential diagnosis of FUO are infections, malignancies, autoimmune conditions, and miscellaneous.3 Physical examination, and standard laboratory testing continue to be the clinical basis of the initial evaluation of the patient with FUO.6,12
Chen et al. (2022) conducted a prospective study in a Chinese population to identify the most suitable model for diagnosing patients with FUO based on 18F-PET/CT imaging.13 Patients were included if they were 14 years or older and had a diagnosis of classical FUO who were hospitalized from January 2016 to July 2021. All patients underwent a standard diagnostic work up which included a thorough patient history questionnaire, physical examination, routine clinical investigation, and second-stage examinations (e.g., chest/abdominal CT with contrast, magnetic resonance imaging, nuclear medicine techniques, and biopsy) and 18F-PET/CT scans when there were no diagnostic clues. Final diagnosis was determined by the study team and causes were divided into five (5) categories which included infection, noninfectious inflammatory disease (NIID), malignancy, miscellaneous cause, and unknown cause. 18F-PET/CT scans were attained using following the guidelines issued by the Chinese Society of Nuclear Medicine. The median fever duration was 37 days (21-732, range). The testing cohort resulted in AUCs of the infection prediction model, the malignancy diagnostic model, and the noninfectious inflammatory disease (NIID) prediction model were 0.89 (0.86-0.92), 0.94 (0.92-0.97), and 0.95 (0.93-0.97), respectively. The validation cohort resulted in ACUs of 0.88 (0.82-0.93), 0.93 (0.89-0.98), and 0.95 (0.92-0.99), respectively. Infectious diseases, tumors, NIIDs, and miscellaneous causes were reported in 223, 121, 109, and 22 patients, respectively. At 6 month follow up, 49 patients remained undiagnosed. Authors concluded 18F-PET/CT diagnostic models displayed good performance, deeming it a reliable tool to help discriminate the cause of FUO. They further acknowledge more robust studies are needed.
Minamimoto 4 (2022) conducted a comprehensive review to help determine the optimal use of 18F-PET/CT as a diagnostic tool in patients with FUO. Results of 26 clinical studies (3 prospective and 23 retrospective) and meta-analysis were reviewed for the diagnosis of FUO, sensitivity and diagnostic yield. It was concluded that final diagnosis by 18F-PET/CT in patients with FUO was dependent on the underlying disease. Earlier diagnosis has led to reduced mortality rates in cases associated with infection such as staphylococcus aureus bacteremia, gram-positive bacteremia, and pacemaker or defibrillator infection have been reported. Authors report the current status of 18F-PET/CT applied to patients with FUO which may affect the future utilization to improve FUO patient outcomes. This meta-analysis is limited by lack of high-quality studies, combining observational and retrospective studies and absence of RCTs.
Kubota et al. (2021) performed a multicenter prospective study to compare the sensitivity of 18F-PET/CT with that of 67 Ga single photon emission computed tomography (SPECT) for the identification of the site of greatest importance for the final diagnosis of the cause of FUO. Ninety-one (91) 18F-PET/CT results were assessed. The sensitivity of 18F-PET/CT (45%, 95% CI 33.1-58.2%) was significantly higher than that for 67 Ga-SPECT (25%, 95% CI 15.5-37.5%) (P = 0.0029) according to the patient-based assessments. Clinical impact of 18F-PET/CT (91%) was significantly higher than 67 Ga-SPECT (57%, P < 0.001). Authors concluded 18F-PET/CT yielded superior sensitivity to 67 Ga-SPECT for the identification of the site of greatest importance for the final diagnosis of the cause of FUO.5
Letertre et al. (2021) performed a retrospective study to explore the diagnostic contribution of the 18F-PET/CT in a population of patients with classical FUO, to pinpoint its place in the diagnostic decision tree in a real-life setting, and to identify those factors associated with a diagnostic 18F-PET/CT.14 Forty-four patients with FUO were included. Diagnoses were obtained in 31 patients (70.5%), 17 had non-infectious inflammatory diseases, 9 had infections (20.5%), and 3 had malignancies (6.8%). 18F-PET/CT assisted with making a final diagnosis (true positive) in 43.6% of all patients. Sensitivity and specificity were 85% and 37%, respectively. A total of 135 investigations were performed prior to 18F-PET/CT including CT scans (93.2%) and echocardiography (59.1%), and 108 after 18F-PET/CT mostly biopsies (including the biopsy of a temporal artery) (25%) and MRIs (34%). Authors concluded that 18F-PET/CT may be proposed as a routine initial non-invasive procedure in the diagnostic workup of FUO, especially in anemic patients who could be more likely to benefit from 18F-PET/CT.14
Buchrits et al. (2021) executed a single center retrospective cohort study to assess the yield of 18F-PET/CT versus contrast enhanced CT (alone) for the diagnosis of classical FUO.15 The 8-year study included 303 patients that underwent 18F-PET/CT for FUO between 1/2012-1/2020. The final diagnosis was based on clinical, microbiological, radiological, and pathological data available in the patient’s last follow-up (at least six months after discharge). Each case was reviewed to determine if 18F-PET/CT was necessary, or if diagnosis could have been attained by CT scan alone. 18F-PET/CT and CT scan sensitivity and specificity results were reviewed. Final diagnosis was achieved as follows: infectious disease 36.5% (111/303), malignancies 18.4% (56/303), and non-infectious inflammatory conditions 17.1% (52/303). 18F-PET/CT scans had an overall sensitivity of 88.7% and a specificity of 80.9% whereas CT scans were 75.2% and 90.2%, respectively.
Georga et al. (2020) performed a retrospective study at a tertiary academic general hospital in Northern Greece in order to assess the diagnostic value of 18F-PET/CT in patients with FUO. Authors retrospectively reviewed 18F-PET/CT scans performed on 50 consecutive adult patients referred for further investigation of classic FUO. Histopathological and microbiological findings, clinical criteria, or clinical follow-up were used to make final diagnosis in 39/50 (78%) of the patients. The cause of FUO was infection in 20/50 (40%), noninfectious inflammatory diseases in 11/50 (22%), and malignancy in 8/50 (16%) patients. Fever remained unexplained in 11/50 (22%) patients. 18F-PET/CT scan contributed to the diagnosis in 70% of the patients, by aiding in identifying the underlying cause of FUO or by directing to the most appropriate site for biopsy. Sensitivity, specificity, accuracy, positive predictive value (PPV) and negative predictive value (NPV) of 18F-PET/CT FUO for active disease detection in patients were 94.7%, 50.0%, 84.0%, 85.7%, and 75.0%, respectively. Authors concluded 18F-PET/CT was a sensitive method for the detection of the underlying cause of FUO or directing to site for biopsy. 16
Kan et al. (2019) performed a meta-analysis to evaluate the diagnostic performance of 18F-PET/CT in the diagnosis of Fever of unknown origin (FUO) and inflammation of unknown origin (IUO). 17 A comprehensive literature search was conducted and returned 23 studies comprised of 1927 patients. A per-patient based analysis was calculated and a pooled diagnosis performance yielded: sensitivity=0.84 (95% confidence interval [CI]=0.79– 0.89), specificity =0.63 (95% CI=0.49–0.75), positive likelihood ratio=2.3 (95% CI=1.5–3.4), negative likelihood ratio=0.25 (95% CI=0.16–0.38), diagnostic odds ratio=9 (95% CI=4.0–20), and AUC=0.84 (95% CI=0.81–0.87). Authors concluded 18F-PET/CT assists with identifying and excluding diseases, guiding future diagnostic decisions, and circumventing unnecessary invasive examinations in patients with non-specific symptoms. Limitations included heterogeneity among included studies, limited data available for analysis and high risk of selection bias.
Lawal et al. (2019) retrospectively reviewed the scans of 46 patients to evaluate the role of 18F-PET/CT in identifying the cause of FUO in patients on renal replacement therapy (RRT) for end-stage renal disease (ESRD).18 18F-PET/CT assisted in identifying the cause of FUO in 22/46 patients (47.83%). Infection resulted in fever in all 22 patients. According to multiple regression analysis, the odds of a helpful 18F-PET/CT increased with every unit increase in CRP level (OR: 1.009; 95% CI: 1.003-1.016; P=0.005). Authors concluded 18F-PET/CT scans (22/46) identified the cause of FUO in patients on RRT for ESRD with a higher CRP level being predictive of a positive 18F-PET/CT outcome.
Kouijzer and colleagues (2018) conducted a review of current literature. Pooled diagnostic yield was 56% (95% CI 50-61%), I2=61%, 18 studies and 905 patients. Only 5 studies reported results of previous imaging, and sub-group analysis estimated diagnostic yield beyond conventional CT at 32% (95% CI 22-44%), I2=66%. Consensus was established that 18F-PET/CT is increasingly available with an emerging role, but there is prevailing variability in practice. Authors concluded that there was insufficient evidence to support the value of 18F-PET/CT in investigative algorithms of FUO. There is a need for research, involving prospective studies recruiting at diagnosis of FUO, with updated case definitions and hard outcome measures. These studies are vital for balancing both radiation exposure and costs against possible benefits of utilizing 18F-PET/CT.19
Schönau et al. (2018) prospectively enrolled 240 patients of which 72 were diagnosed with FUO, 142 with inflammation of unknown origin (IUO) and 26 had FUO or IUO previously (exFUO/ IUO) to assess the diagnostic utility of 18F-PET/CT in patients with FUO or IUO and to define parameters that increase the likelihood of diagnostic 18F-PET/CT.20 Final diagnosis was made in 190 patients (79.2%) in which 132 had NIID (69.5%), 27 had infections (14.2%), 20 malignancies (10.5%) and 11 other diseases (5.6%). 18F-PET/CT was helpful in determining final diagnosis in 136 FUO patients. Chief diagnoses were adult-onset Still’s disease at 15.3% in the FUO group, large vessel vasculitis at 21.1% and polymyalgia rheumatica 18.3% in the IUO group then IgG4-related disease (15.4%) in the exFUO/IUO group. Authors conclude that a more accurate diagnosis was achieved in more than 50% of FUO and IUO cases with utilization of 18F-PET/CT.
Takeuchi et al. (2018) performed a systematic review and meta-analysis to assess the association of 18F-PET or PET/CT results with spontaneous remission in classic fever of unknown origin.21 The search yielded nine studies of PET/CT results (418 patients) and four studies of standalone PET results (128 patients). Negative PET/CT results were significantly more probable to present with spontaneous regression than positive results (summary RR=5.6; 95% CI: 3.4–9.2; P<.001; I2=0%). Standalone PET results and spontaneous remission yielded no significant association. As result of random-effects study-level meta-regression, PET/CT results [relative RR (rRR)= 7.4; 95% CI: 2.5–21.3; P=.002], compared with standalone PET results, and publication year (rRR=1.2 per 1 year; 95% CI: 1.0–1.3; P=.013) were significantly associated with spontaneous remission. Authors concluded FUO patients with negative PET/CT results yielded a high likelihood of spontaneous remission, but further studies should be conducted to validate their findings. Limitations included small sample size, limited data, inclusion of studies with high risk of bias, retrospective study design, short follow-up period and variability in methods across studies.
Bharucha et al. (2017) conducted a systematic review and meta-analysis to assess the diagnostic yield of 18F-PET/CT in patients with fever of unknown origin.22 After the quality assessment was performed, 18 studies with a total of 905 patients were selected for meta-analysis. Eighteen studies (905 patients) resulted in a pooled diagnostic yield of 56% (95% confidence interval [CI]: 50-61%, I2=61%). Previous imaging was only reported in five studies, and diagnostic yield beyond conventional CT at 32% (95% CI: 22-44%; I2=66%) was estimated from subgroup analysis. Authors conclude although consensus was established that 18F-PET is becoming more available and has an emerging role, there is high variability in practice and resolve there is insufficient evidence to support 18F-PET/CT in investigative algorithms of FUO. Limitations included studies with a high risk of bias, significant heterogeneity, and overall low quality.
Hung et al. (2017) prospectively evaluated 68 patients with a diagnosis of FUO to compare the efficacy of 18F-PET and 67Ga SPECT/CT in diagnosing FUO on a head-to-head basis. First-step examinations included clinical history, physical examinations, routine laboratory tests, chest radiographs (including CT in nine patients), and abdominal ultrasound were assessed, but yielded no potentially diagnostic clues. The patients were entered prospectively into the study. As a second step investigation technique, 18F-PET and 67Ga SPECT/CT, was performed within a 7-day interval. All studies were read and interpreted independently by two board-certified nuclear medicine physicians. Ten of the 68 patients were excluded. In 23 patients an underlying infectious disease was found, 10 patients had a malignant disorder, 11 patients had non-infectious inflammatory disease, two patients had adrenal insufficiency. The cause of FUO was not found for 12 patients. A high false-positive rate of 44% (7/16) was observed for 18F-PET/CT, while a high false-negative rate of 55% (23/42) was observed for 67Ga SPECT/CT. 18F-PET/CT studies depicted all 67Ga-avid lesions. The sensitivity (79% vs. 45%) and clinical contribution (72% vs. 55%) of 18F PET/CT in diagnosing FUO were significantly higher than those of 67Ga SPECT/CT (p < 0.05). Authors concluded the diagnostic performance of 18F-PET/CT was superior to 67Ga SPECT/CT in the work-up of patients with FUO. With its rapid results and superior sensitivity, 18F-PET/CT may replace 67Ga SPECT/CT where this technique is available. 23
Takeuchi et al. (2016) performed a meta-analysis to review the test performance, diagnostic yield, and management decision impact of nuclear imaging tests in patients with classic FUO.24 The literature search included 42 studies comprised of 2,058 patients. Studies with higher prevalence of neoplasms and infections resulted in higher diagnostic yields. Nonneoplastic causes were found to be less successfully localized. Among the four imaging modalities, 18F-PET/CT resulted in the best test performance and diagnostic yield with summary sensitivity was 0.86 (95% confidence interval [CI], 0.81–0.90), specificity 0.52 (95%CI, 0.36–0.67), and diagnostic yield 0.58 (95% CI, 0.51–0.64). Authors concluded nuclear imaging tests, predominantly 18F-PET/CT, may be helpful in identifying sources of fever in patients with classic FUO. They acknowledge studies with standardized diagnostic algorithms are needed to establish optimal timing for testing and further assessments are needed to fully understand management decision impacts and outcomes. Limitations include heterogeneity, small sample size, methodologic limitations, and a lack of standardization of diagnostic algorithms.
Besson et al. (2016) conducted a stratification based meta-analysis to quantify the contribution of 18F-PET/CT to the diagnostic assessment of fever of unknown origin (FUO). PubMed/MEDLINE was searched from 2000 to September 2015 for articles with inclusion criteria of: (1) FUO as the initial diagnosis, (2) no immunosuppressed or nosocomial condition, (3) final diagnosis not based on 18F-PET/CT, (4) a follow-up period specified, (5) adult population, and (6) availability of adapted data for calculation of odds ratios (ORs). Ultimately, 14 studies were included in the meta-analysis. Statistical analysis and sensitivity analyses were performed. Authors concluded that “normal 18F-PET/CT findings led to an increase in the absolute final diagnostic rate of 36% abnormal 18F-PET/CT findings to an increase of 83%, corresponding to a pooled OR of 8.94 (95% CI 4.18–19.12, Z=5.65; p<0.00001).” They further state that the design of the studies also influenced results as follows: (OR 2.92, 95% CI 1.00–8.53 for prospective studies; OR 18.57, 95% CI 7.57–45.59 for retrospective studies; p= 0.01). Authors concluded that geographic area and length of follow up did not influence study results and increased final diagnostic rate in FUO was associated with abnormal 18F-PET/CT findings. Further randomized prospective studies with standardized 18F-PET/CT procedures are needed. Limitations include heterogeneity among included studies, lack of diagnostic protocol and possible selection bias.25
Pereira et al. (2016) conducted a retrospective study on 76 consecutive patients to determine the clinical usefulness of 18F-PET/CT for FUO and to define determinants of its diagnostic utility. 18F-PET/CT exhibited characteristically increased 18F activity in 56 patients (74%), leading to confirmation of, or change in, the suspected cause of FUO in 57 and 17%, respectively. Final diagnosis in 18F-PET/CT patients included infection (21%), malignancy (22%), noninfectious inflammatory disease (12%), others (5%), or an unknown cause (40%). The success rate, sensitivity, and specificity of 18F-PET/CT were 60, 77, and 31%, respectively. Patients with suspected malignancy (100%, 95% confidence interval 79-100%) produced the highest sensitivity. Diagnostic performance was independent of the investigated variables apart from suspected infection as a cause of FUO (odds ratio 0.1, 95% confidence interval 0.01-0.8, P=0.033). The authors recognize most studies investigating 18F-PET/CT in patients with FUO are difficult to compare based on heterogeneity of the patient population, variation among 18F-PET/CT techniques, and the stages at which it was performed. This study’s findings yielded a success rate of 61% and a sensitivity of 77% in FUO patients. However, they go further and state a final diagnosis remained unclear in 39% of the FUO patients despite performing an 18F-PET/CT.26
Gafter-Gvili (2015) conducted a retrospective single center study over 4 years on 112 patients that were hospitalized in the years 2008–2012 for FUO evaluation and underwent 18F-PET/CT.27 Final diagnosis was based on clinical, microbiological, radiological, and pathological data available at the final follow-up. A total of 83 patients (74%) resulted in a final diagnosis which included: infectious disease 43% (n=49), non-infectious inflammatory disease 16% (n=17), malignancies 14% (n=15) and other diagnosis in 1.7% (n=2). 18F-PET/CT were considered helpful in 66% (74 studies) and contributed to 46% of positive diagnosis and 20.5% exclusion of diagnosis. 18F-PET/CT demonstrated a sensitivity of 72.2%, a specificity of 57.5%, a positive predictive value (PPV) of 74.2% and a negative predictive value (NPV) of 53.5%. FUO spontaneously resolved in 20% of patients without a diagnosis. Authors concluded support for 18F-PET/CT as the most important imaging modality based on their study findings reporting 66% of studies were contributory to diagnosis in patients with FUO.
Hao and associates (2013) performed a meta-analysis to assess the diagnostic performance of 18F-PET/CT in patients with fever of unknown origin. 28 Fifteen studies comprised of 595 patients with FUO were included. On a per-patient-based analysis, a pooled sensitivity of 18F-PET/CT in detecting the cause of FUO was 85% (95% confidence interval 81–88%). The area under the receiver-operating characteristic curve was 0.88. Authors concluded that 18F-PET/CT demonstrated high sensitivity and value in diagnosis of patients with FUO. They further stated 18F-PET/CT is an accurate technique in this setting, but the possibility of false-positive results should be considered outside of this setting. Limitations resulted from inclusion of studies of low to moderate quality, bias, limited sample size.
Dong et al. (2011) conducted a meta-analysis to evaluate the value of 18F-PET/CT in patients with fever of unknown origin.29 Five 18F-PET studies and four 18F-PET/CT studies were evaluated. For the detection of FUO, 18F-PET resulted in a pooled sensitivity and specificity were 0.826 (95% CI; 0.729–0.899) and 0.578 (95% CI; 0.488–0.665), respectively, and the AUC was 0.810. Among the results of 18F-PET studies severe heterogeneity was present (QSE=12.40, I2=67.7%; QSp=35.98, I2 =88.9%). Pooled sensitivity and specificity of 18F-PET/CT were 0.982 (95% CI; 0.936–0.998) and 0.859 (95% CI; 0.750–0.934), respectively, and the AUC was 0.947. No statistical differences were observed in the AUC and Q index between FDG-PET and 18F-PET/CT (Z = 0.566, p > 0.05). Limitations of this meta-analysis include incorporation of studies with poor design, small number of studies with limited sample sizes, and risk of bias.
Keidar and associated (2008) conducted a study to prospectively assess the role of PET/CT using 18F in the investigation of FUO. A total of 48 consecutive patients (25 men, 23 women; age range, 24–82 yo) with FUO underwent 18F-PET/CT scans. Final diagnosis was based on histopathology, microbiologic assays, or clinical and imaging follow-up. 18F-PET/CT detected suggestive foci of increased 18F uptake in 27 patients. In 22/27 positive studies (81%), 18F-PET/CT identified the underlying disease and diagnosed infection in 9 patients, an inflammatory process in 10 patients, and malignancy in 3 patients. 18F-PET/CT was negative in 21 patients. All these patients were diagnosed as having systemic non focal infection or drug-induced fever or showed spontaneous resolution of the febrile state with no evidence of a localized inflammatory, infectious, or malignant process for a clinical follow-up period of 12–36 mo. Authors concluded-18F-PET/CT identified the underlying cause of the fever in 46% of the present study population and contributed to the diagnosis or exclusion of a focal pathologic etiology of the febrile state in 90% of patients. 18F-PET/CT has a high negative predictive value (100%) for assessment of FUO.30
Bleeker-Rovers (2007) conducted a prospective multicenter study to integrate the use of 18F-PET as part of a diagnostic protocol in the general FUO patient population.31 Patients with FUO were recruited from a university hospital and five community hospitals. A total of 70 patients were enrolled at a university hospital (n=38) and five community hospitals (n=32). A diagnostic protocol utilizing 18F-PET was employed. 33% of scans contributed to the final diagnosis and 70% of abnormal 18F-PET scans were clinically helpful, however, the probability of reaching a diagnosis was only 50%. 18F-PET had a sensitivity of 88% and a specificity of 77%, with a positive predictive value of 70% and negative predictive value of 92%. There was no significant difference between patents diagnosed in the university hospital versus the community hospitals and the patients with normal erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) 18F-PET proved not to be helpful. Authors concluded that 18F-PET, when included in a diagnostic protocol, could be a useful imagining modality in patients with FUO and raised ESR or CRP.
Jaruskova and Belohlavek (2006) retrospectively reviewed 124 patients with fever of unknown origin (76%) or prolonged fever due to a suspected post-surgical infection of a joint or vascular prosthesis (24%). Fifty-seven (46%) patients had positive 18F-PET findings. In six patients, no further clinical information was available. Positive 18F-PET findings in 51 patients and 118 patients in total were subsequently evaluated. Systemic connective tissue disease was confirmed in 17 patients, lymphoma in three patients, inflammatory bowel disease in two patients, vascular prosthesis infection in seven patients, infection of a hip or knee replacement in seven patients, mycotic aneurysm in two patients, abscess in four patients and AIDS in one patient. Eight (16%) patients the finding was falsely positive. Authors concluded 18F-PET or 18F-PET/CT contributed to establishing a final diagnosis in 84% of the 51 patients with positive 18F-PET findings and in 36% of all 118 evaluated patients with prolonged fever.32
Mourad and colleagues (2003) performed a systematic review to develop evidence-based recommendations for the diagnostic workup of FUO. The quality of articles was rated as “good,” “fair,” or “poor”. Sensitivity, specificity, and diagnostic yield of tests were calculated. Eleven studies indicate that diseases include “no diagnosis” (19%), infections (28%), inflammatory diseases (21%), and malignancies (17%). Deep vein thrombosis (3%) and temporal arteritis in the elderly (16%-17%) were considered. Four good studies indicate that patients with undiagnosed FUO recover spontaneously (51%-100%). A single fair-quality study yielded a high specificity (99%) for the diagnosis of endocarditis in FUO by applying the Duke criteria. One fair-quality study showed that computed tomographic scanning of the abdomen had a diagnostic yield of 19%. Ten studies of nuclear imaging revealed that technetium was the most promising isotope, showing a high specificity (94%), albeit low sensitivity (40%-75%) (2 fair-quality studies). Two fair quality studies showed liver biopsy to have a high diagnostic yield (14%-17%), but with risk of harm (0.009% 0.12% death). Empiric bone marrow cultures showed a low diagnostic yield of 0% to 2% in two articles rated as fair quality. Authors concluded that diagnosis of FUO may be assisted by the Duke criteria for endocarditis, computed tomographic scan of the abdomen, nuclear scanning with a technetium based isotope, and liver biopsy (fair to good evidence). Routine bone marrow cultures are not recommended.33 Of FUO patients, 18F-PET/CT was deemed as necessary in 26% (79/303). Endovascular infection, hematological malignancy, and large vessel vasculitis were the only factors associated with 18F-PET/CT necessity on multivariable analysis. Authors concluded they were able to successfully define the patient and conditions where referral for 18F-PET/CT would be the first in line imaging modality in cases such as hematological malignancies, LVV and endovascular infections. They acknowledge the need for more robust studies comparing 18F-PET/CT and CT scan as a first line diagnostic tool.
The role of 18F-PET/CT in immunocompromised patients has not been established. A retrospective study of 10 patients with HIV-associated FUO were studied with 18F-PET/CT and the authors reported the study was “helpful for diagnosis”.36 Another retrospective report compared identifying etiology of FUO in 10 HIV-infected patients with FUO to 10 viremic patients without FUO.37 The authors felt the results of the PET study helped to identify the etiology of the fever in 80% of those with FUO. Both studies were limited by retrospective design, small sample size and lack of controls.
A multi centered, open labeled RCT randomized patients undergoing chemotherapy for hematopoietic stem cell transplantation or chemotherapy for acute leukemia with neutropenic fever defined as greater than 72 hours to 18F-PET/CT (n=73) or CT (n=74). Antimicrobial rationalisation occurred in 53 (82%) of 65 patients in the 18F-PET-CT group and 45 (65%) of 69 patients in the CT group (OR 2·36, 95% CI 1·06–5·24; p=0·033). The authors concluded that the 18F-PET/CT was helpful in the management of high-risk neutropenic fever after chemotherapy or transplant conditions.38 While this study shows a potential role in the management of neutropenic fever in this population it does not provide data regarding if PET imaging had improved outcomes compared to the CT group nor addresses how many patients management would have been changed if they initially underwent a CT and reserved PET as second line study in unclear cases.
Additional Guidance
UpToDate states that F-fluorodeoxyglucose positron emission tomography (18F-PET) may be sensitive in identifying anatomic sites of inflammation and malignancy. 18F-PET may have a valuable place in the evaluation of FUO, but additional data are needed to determine its added value beyond repeated clinical evaluation over time and routine CT.2
A 2022 review paper published in the New England Journal of Medicine provides a summary of fever of unknown origin and includes a diagnostic and management algorithm for FUO.6 This algorithm includes 18F-PET/CT after history and physical and minimal and advance FUO evaluation fails to determine a diagnosis. The author reports the range of performance for 18F-PET/CT with sensitivities ranging from 52-85% and diagnostic yield ≥ 50% which is 30% higher than conventional CT. The yield seems to be higher in patients with infection or neoplasia as compared to autoimmune sources. They also state PET scan seems to be superior to other nuclear imaging methods such as PET without CT, or gallium or leukocyte scintigram.23
The SMEs felt there is a role for PET scan for FUO citing it can be useful in 20-30% of cases to identify the source of FUO in cases that meet the definition above (3 weeks of unexplained fever). The addition of CT may increase yield by 30%. They report pooled sensitivity between 75-86% based on current literature. They also discussed the value of a negative study in which literature shows a higher rate of spontaneous resolution of fever allowing providers to feel more confident in a watchful waiting approach. The discussed limitations such as recent post-operative changes, antibiotic and steroid use, and importance of standardized protocols. Overall, they felt there is role for this study in FUO as they consider this equivalent to superior to other studies used in these diagnostic conundrums with faster results.
Cardiac
Preparation is critically important for the success of cardiac PET/CT. Our SMEs explained there are standardized protocols that must be followed to get useful results from the study. These protocols are well established and readily available.8,9
Infections: Cardiac Endocarditis and Infected Cardiac Implantable Devices
Infective endocarditis (IE) is an infection of the endocardial surface of the heart. Frequently including the heart valves, predominantly prosthetic heart valves. Infective endocarditis has an annual incidence of 3-9 cases per 100,000 with the highest levels reported in patients with prosthetic valves, cardiac devices, unrepaired cyanotic heart disease and those with a family history.39
A report by the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines released in 2020 recommends in patients with suspected endocarditis, it is recommended that the Modified Duke Criteria are to be utilized as the current standard for diagnosis.40 This criterion incorporates the use of clinical, imaging, and bacteriological criteria. Recommendation 12.2 states “In patients classified by Modified Duke Criteria as having ‘possible IE’, 18F-PET/CT is reasonable as adjunct diagnostic imaging.” They go further to state “18F-PET/CT may also be considered a complementary diagnostic tool for some patients with suspected native valve endocarditis.” This recommendation is supported at the CCF/AHA Task Force on Practice Guidelines. Methodology Manual and Policies From the ACCF/AHA Task Force on Practice Guidelines at level of 2A which is a moderate strength of recommendation and considered reasonable.
UpToDate published a topic on the clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis.41 It recognizes echocardiography as the standard imaging modality for evaluation of cardiac valves however, 18F-PET/CT has emerged as a diagnostic tool for IE that can identify infection of native valves, paravalvular areas and extracardiac sites of infection.
A 2015 ESC Guidelines for the management of infective endocarditis: The Task Force for the Management of Infective Endocarditis of the European Society of Cardiology (ESC) Endorsed by: European Association for Cardio-Thoracic Surgery (EACTS), the European Association of Nuclear Medicine (EANM)10 recommends “in the setting of the suspicion of endocarditis on a prosthetic valve, abnormal activity around the site of implantation detected by 18F-PET/CT (only if the prosthesis was implanted for >3 months) or radiolabeled leucocyte SPECT/CT should be considered a major criterion.”
The Royal College of Physicians published a contemporary update infective endocarditis in 2020. 18F-PET/CT or radiolabeled leucocyte single photon may assist as an adjunctive investigation to determine presence of inflammation or infection of the prosthetic heart valve that would result in an IE diagnosis. Whereas in cases of perivalvular pathology especially when root abscess or aortic valve endocarditis is suspected, CT was determined to be of value.42
The National Heart Association of Malaysia (NHAM) and the Academy of Medicine Malaysia produced a clinical practice guideline for prevention, diagnosis and management of infective endocarditis in 2017 (reviewed in 2021). Guidelines suggest when used as an adjunct modality, WBC SPECT/CT and 18F-PET/CT may have a role in detecting peripheral embolic, metastatic infectious lesions, cases with high clinical suspicion of IE, prosthetic valve endocarditis and in other difficult cases.
Albano et al. (2021) conducted a systematic review and meta-analysis to assess the diagnostic role of 18F-PET/CT imaging in native valve endocarditis (NVE).43 Twelve papers were included in the analysis comprised of 600 patients with NVE. 18F-PET/CT imaging resulted in a pooled sensitivity of 31% (95% confidence interval [CI]: 26% to 36%) with a heterogeneity (I2=73.1%; p = 0.001) and a pooled specificity was 82% (95% CI: 77% to 86%) with a heterogeneity (I2 = 59%; p = 0.012). Authors conclude 18F-PET/CT imaging in NVE remains suboptimal. This is attributed to the low accuracy heterogeneity among papers, despite good specificity (82%). They further state specific protocols may improve diagnostic performance. Limitations include selection bias, small sample size, low study quality and design, and high heterogeneity.
De Camargo (2020) published a prospective study of 303 patients with left sided infective endocarditis (IC) who underwent 18F-PET/CT. Of the 188 in this population with prosthetic valves the sensitivity, specificity, and positive and negative predictive values of 18F-PET/CT focal uptake were 93%, 90%, 89%, and 94%, respectively. Among 115 patients with native valve endocarditis, the corresponding values were 22%, 100%, 100%, and 66%. The authors concluded the use of 18F-PET/CT increased the diagnostic capability of the modified due criteria in the setting of prosthetic valves. in the setting of native valves, it was found to be less accurate and could only be considered a complementary diagnostic tool for those with persistent suspicion of IE but inconclusive results.44
Mahmood et al. (2019) conducted a meta-analysis of 13 studies including 537 patients performed to evaluate the role of 18F-PET/CT as an adjunctive diagnostic test for prosthetic valve endocarditis and culture negative endocarditis. The pooled sensitivity of 18F-PET/CT for diagnosis of IE was 76.8% (95% CI 71.8–81.4%; Q 5 39.9, P < 0.01; I2 5 69.9%) and the pooled specificity was 77.9% (95% CI 71.9–83.2%; Q 5 44.42, P < 0.01; I2 5 73.0%). Diagnostic accuracy was improved with prosthetic valves with a sensitivity of 80.5% and specificity of 73.1%. Additional extracardiac foci of infection were found in 17% of patients on whole body 18F-PET/CT. The authors concluded that 18F-PET/CT is a useful adjunctive diagnostic tool in the evaluation of diagnostically challenging cases if IE especially those with prosthetic valves. Five of the 13 studies involving 203 patients included both native and prosthetic value endocarditis, however the number with native values was not included nor was this group analyzed separately. Limitations of this study is moderate heterogeneity including variation in imaging protocols, data acquisition and methodological concerns in included studies including lack of blinding. The authors also address the uncertainty of the impact of prior antibiotic treatment on the sensitivity of 18F-PET/CT and call for additional prospective studies.45
Juneau et al. (2017) performed a systematic review and meta-analysis to systematically assess the diagnostic accuracy of 18F-PET/CT, labeled leukocyte scintigraphy (LS), and Gallium-67 citrate scintigraphy for the diagnosis of CIED infection.46 Literature was systematically reviewed on the three modalities in CIED infection identifying 2493 articles in which 13 articles (11 studies for 18F-PET/CT and 2 for LS), were ultimately selected for analysis. The meta-analysis on the use of 18F-PET/CT resulted in a sensitivity of 87% (95% CI, 82%–91%), a specificity of 94% (95% CI, 88%–98%) and an AUC of 0.952. There was insufficient data for LS therefore meta-analysis was not performed. Authors concluded that 18F-PET/CT and LS may be useful for the diagnosis of CIED infection and suggest that 18F-PET/CT may be preferred when available. Limitations include lack of gold standard for diagnosis of CIED, lack of blinding, possible selection bias and heterogenous methods data. A similar 2018 review and meta-analysis to evaluate the potential utility of 18F-PET/CT, (67) Ga citrate and radiolabeled white blood cell (WBC) scintigraphy in the diagnosis of IE.47 reported on 14 studies with sensitivity 81% for 18F-PET/CT and 86% for LS again concluding both tests are accurate for diagnosis of IE.47
Yan et al. (2016) performed a systematic review and meta-analysis to assess the role of 18F-PET/CT in patients with infectious endocarditis (IE).48 The literature search returned 6 studies (2 retrospective and 4 prospective studies) comprised of 246 patients. The meta-analysis yielded a pooled sensitivity of 61% (95% confidence interval (CI), 52 – 88%), and a pooled specificity of 88% (95% CI, 80 – 93%). Authors concluded due to its low sensitivity, 18F-PET/CT is currently not sufficient for the diagnosis of IE, however, it might be a useful in the diagnosis of prosthetic valve endocarditis (PVE) and skin and pocket cardiovascular implantable electronic device (CIED) infections. Limitations include bias, selection of studies with low to moderate quality with small sample sizes, lack of testing for heterogeneity due to limited number of studies and use of suboptimal reference tests in selected studies.
Pizzi et al. (2015) performed a prospective study at a cardiac surgery referral center to evaluate the clinical value of 18F-PET/CT and 18F-PET/CTA for the diagnosis of IE in patients with prosthetic valves or intracardiac devices.49 The cardiac surgery referral center was comprised of a multidisciplinary IE unit including cardiologists, infectious diseases physicians, cardiac imaging specialists, heart surgeons, neurologists, and microbiologists. A total of 92 patients with prosthetic valves or cardiac devices consecutively admitted for suspected IE were evaluated in this study. Patients underwent echocardiography and 18F-PET/CT and 76 patients had cardiac CTA. Findings were compared and yielded concordant results in 54% of cases (K=0.23). Comparisons were evaluated between final diagnostic consensus and initial diagnosis with DC at admission, 18F-PET/CT, and Duke Criteria (DC)+18F-PET/CT. Reclassification of 90% of cases resulted from DC+18F-PET/CT that were initially classified as possible IE. “Sensitivity, specificity, and positive and negative predictive values were 52%, 94.7%, 92.9%, and 59.7% for DC; 87%, 92.1%, 93.6%, and 84.3% for 18F-PET/CT; and 90.7%, 89.5%, 92%, and 87.9% for DC+18F-PET/CT. Use of 18F-PET/CTA yielded better diagnostic values than PET/nonenhanced CT (91%, 90.6%, 92.8%, and 88.3% versus 86.4%, 87.5%, 90.2%, and 82.9%) and reduced the rate of doubtful cases from 20% to 8% (P<0.001). DC+18F-PET/CTA reclassified an additional 20% of cases classified as possible IE with DC+PET/nonenhanced CT. In addition, 18F-PET/CTA enabled detection of a significantly larger number of anatomic lesions associated with active endocarditis than PET/nonenhanced CT (P=0.006) or echocardiography (P<0.001).” Authors concluded “18F-PET/CT improves the diagnostic accuracy of the modified DC in patients with suspected IE and prosthetic valves or cardiac devices. 18F-PET/CTA yielded the highest diagnostic performance and provided additional diagnostic benefits.” Limitations include lack of gold standard, and authors bring up concern regarding radiation exposure resulting from the imaging examination.
Saby et al. (2013) conducted a prospective study to determine the value of 18F-PET/CT for diagnosing prosthetic valve endocarditis (PVE).50 A total of 72 consecutive patients suspected of having PVE were included. All subjects underwent clinical, microbiological, and echocardiographic evaluation and cardiac 18F-PET/CT at admission. At a 3 month follow up the final diagnosis was defined according to the clinical and/or pathological modified Duke criteria. Abnormal 18F uptake around the prosthetic valve was seen in 36 (50%) subjects. Sensitivity, specificity, positive predictive value, negative predictive value, and global accuracy were (95% confidence interval): 73% (54% to 87%), 80% (56% to 93%), 85% (64% to 95%), 67% (45% to 84%), and 76% (63% to 86%), respectively. A significant increase in the sensitivity of the modified Duke criteria at admission (70% [52% to 83%] vs. 97% [83% to 99%], p = 0.008) was observed when abnormal FDG uptake around the prosthetic valve as a new major criterion was added. Yielding a significant reduction in the number of possible PVE cases from 40 (56%) to 23 (32%). Authors concluded 18F-PET/CT was a useful imaging modality in diagnosing PVE and study results support adding abnormal FDG uptake as a major criterion for PVE. Limitations include presence of false positive and false negative tests, defining optimal timing for performing 18F-PET/CT, use of modified Duke criteria at the end of the follow-up was used as the gold standard in this study, and short follow up period.
Diagnosis of cardiac implantable electronic device (CIED) can be challenging. Echocardiogram is currently the first line imaging technique in patients with suspected CIED but is of limited value unless there is vegetation visualized along the lead. This has led to further investigation of the role of PET scan because this is an infective process. 46. Alternative studies are somewhat limited in this population because MRI is often contraindicated in patients with CIED.
Juneau et al. (2017) conducted a systematic review and meta-analysis which included 11 studies on use of 18F-PET/CT for cardiac implantable electronic device (CIED) infections with 260 patients. This analysis also included studies for labeled leukocyte scintigraphy (LS), and Gallium-67 citrate scintigraphy. They report pooled sensitivity of 18F-PET/CT was 87% (95% CI, 82-91%) and specificity of 94% (95% CI, 88-98%). The authors conclude that 18F-PET/CT yields a high sensitivity specificity and accuracy and is useful in the diagnosis of CIED infections. 46
Cardiac Sarcoidosis (CS)
Cardiac sarcoidosis is rare, affecting 10 out of 100,000 people a year. CS is estimated to affect 2-5% of those with systemic disease although autopsy studies suggest it may be as high as 20-27%. In a joint SNMMO-ASNC expert consensus document on the role of 18F-PET/CT in cardiac sarcoid detection and therapy monitoring. Cardiac sarcoidosis may be present in as many as 20% of sarcoid cases, may be asymptomatic, and is a major cause of death from this condition. CS remains an underdiagnosed condition and prognosis may be improved with treatment. Diagnosis is challenging given the low yield of extracardiac tissue biopsy and no useful gold standards. Cardiac 18F-PET Has been increasingly added to diagnostic algorithms for CS to detect cardiac involvement as well as assessing the presence and severity of myocardial inflammation when done in patients without contraindication and with proper preparation.11
A Joint Procedural Position Statement on Imaging in Cardiac Sarcoidosis. From the Cardiovascular and Inflammation & Infection Committees of the European Association of Nuclear Medicine (EANM), the European Association of Cardiovascular Imaging (EACVI), and the American Society of Nuclear Cardiology (ASNC)(2017) reports “FDG-PET has emerged as a powerful and most commonly used technique not only to assess the extent of systemic sarcoidosis but also to assess extent and activity of myocardial involvement”. The panel recommends 18F-PET in conjunction with radionuclide myocardial perfusion imaging (MPI) is therefore the currently recommended radionuclide method for evaluation of cardiac sarcoidosis. The sensitivity for detecting CS in the reviewed studies was 85-100% (39-100% range). This is consistent with other SR/MA which report similar range.11
Aiken et al. (2022) conducted a systematic review and meta-analysis on the diagnostic accuracy of cardiac MRI versus 18F-PET/CT for CS. Thirty-three studies including 678 patients with CS were included. Six studies offered a direct comparison between cardiac MRI and 18F-PET/CT in the same 303 patients. Cardiac MRI had a higher sensitivity than 18F-PET/CT (95% vs 84%; P = .002) with no difference in specificity. This study was limited by high risk of bias in 30 of the studies included and few studies with direct comparison.51
Kim et al. (2020) 52 conducted a systematic review and meta-analysis which includes 17 studies with 891 patients with CS. They a report pooled sensitivity of 0.84 [95% confidence interval (95% CI) 0.71-0.91] with heterogeneity (I2 = 77.5) and a pooled specificity of 0.83 (95% CI 0.74-0.89) with heterogeneity (I2 = 80.0). Meta-regression suggests the heterogeneity was caused by studies which combined 18F-PET/CT with myocardial perfusion studies. The authors acknowledge the limited studies on use of 18F-PET/CT for CS and need for further large multicentered studies. The suggested addition of myocardial perfusion imaging may improve diagnostic accuracy.
A 2020 review describes the importance of appropriate preparation and protocol for accurate 18F-PET/CT results, they describe 18F-PET/CT as the “primary tool” to assess active inflammation and a powerful complementary modality in cases where diagnosis remains uncertain despite initial investigation including cardiac MRI. They describe the ability of 18F-PET/CT to assess the degree of CS involvement and aid in guiding the course of therapy as well as identification of extracardiac disease that can be potential targets for biopsy. The authors also reference studies in which reclassification occurred after the study improving diagnostic accuracy.53 The Joint SNMMI-ASNC Consensus Document on the role of 18F-PET/CT in cardiac sarcoid states that this can be a useful tool in the investigation of suspected cardiac sarcoidosis but diagnosis should not be based on PET study alone. Their guidelines state a role for PET imaging for patients with histological evidence of extracardiac sarcoidosis and abnormal screening for cardiac sarcoidosis, patients 60 years or younger with unexplained, new onset, significant conduction system (such as sustained second or third degree AV block), and patients with idiopathic sustained ventricular tachycardia unexplained by other causes.11,54,55
Vita et al (2018) conducted a retrospective review of 107 consecutively enrolled subjects referred for cardiac sarcoidosis (CS) evaluation.56 Evaluations of CS were conducted by cardiac magnetic resonance (CMR) and 18F-PET/CT to determine the value of combining them in assessing the likelihood of CS and guiding patient management. Two blinded experienced readers assessed the likelihood of CS after reviewing CMR and 18F-PET images. Outlined by predefined criteria, readers were asked to categorize the likelihood of CS as no (<10%), possible (10%–50%), probable (50%–90%), or highly probable (>90%). A total of 107 patients (age, 55±11 years; left ventricular ejection fraction, 43±16%), 91 (85%) had late gadolinium enhancement, whereas 82 (76%) had abnormal F18-fluorodeoxyglucose uptake on 18F-PET, which was suggestive of active inflammation. A total of 91 patients with positive late gadolinium enhancement, 60 (66%) had abnormal F18-fluorodeoxyglucose uptake. When 18F-PET and CMR were combined, 48 (45%) patients were reclassified as having a higher or lower likelihood of CS, most of them (80%) being correctly reclassified when compared with the final diagnosis. In patients with highly probable CS, immunosuppressive therapies were significantly more likely to change. Combining CMR and 18F-PET resulted in the reclassification to highly probable (>90%) likelihood in 12 new patients. Authors concluded results show that including data from both CMR and 18F-PET may increase the diagnostic certainty, chiefly following inconclusive CMR or 18F-PET imaging.
Osbourne et al. (2014) aimed to determine if monitoring CS with 18F-PET/CT improved outcomes, specifically ejection fraction (EF). This was a small prospective study with 23 patients with cardiac sarcoidosis followed for a total of 90 serial 18F-PET/CT scans averaging 4 scans per patient. In this study they found that a reduction in the intensity and extent of myocardial inflammation seen on the 18F-PET/CT was associated with improvements in the EJ. Therefore, the authors concluded that serial 18F-PET/CT may be helpful in titration of immunosuppressive therapy for improvement or prevention of heart failure in CS.57
In another small observational study with 12 patients followed with 18F-PET/CT before and after treatment with infliximab demonstrated that the detection of changes on the 18F-PET/CT correlated with the clinical improvement in 92% of these patients supporting the role of 18F-PET/CT for evaluation of response to treatment. This is limited by the small sample size and lack of comparative group.58
Vascular Graft infection
Pelletier-Galarneau et al. (2020) 59 published an editorial discussing improving diagnosis with imaging in regard to vascular graft infections. Historically, clinical manifestations of VGI are challenging. To add to the complexity, there are various vascular prostheses and techniques that may impact treatment outcomes of VGI. While CTA and MRI can be utilized in these cases, they have their limitations. The combination of these challenges led to the Management of Aortic Graft Infection (MAGIC)60 producing a criterion for the diagnosis of aortic graft infection in 2016. Several small studies and three meta-analyses on the use of 18F-PET in the diagnosis of VGI have been investigated. 18F-PET for the detection of VGI yielded similar results including a pooled sensitivity of 93%, 95%, and 96% and a pooled specificity of 74%, 78%, and 80%. The gold standard for diagnosis may not always be readily available therefore, it is suggested 18F-PET/CT coupled with the MAGIC criteria may serve as a desirable technique to detect vascular graft infection. Authors acknowledge the need for more robust studies to establish the diagnostic accuracy and impact of 18F-PET/CT in the evaluation and management of VGI patients.
Bowles et al. (2020) 61 performed a prospective cohort study to determine the diagnostic yield of 18F-PET/CT in patients with suspected PVGI. A total of 49 patients with suspected prosthetic vascular graft infection (PVGI) were included. Uptake patterns were defined as: (i) focal, (ii) patched (PVGI criteria), and (iii) diffuse (no PVGI criterion). 18F-PET/CT values were sensitivity (88%), specificity (79%), positive predictive values (67%), and negative predictive value (93%). 18F-PET/CT successfully identified 14/16 PVGI cases resulting in a focal (n = 10) or patched pattern (n = 4), with a true negative in 26/33 cases with either a diffuse pattern (n = 16) or without uptake (n = 10). A patched pattern occurred in 71% (5/7) false positive cases, all of which were associated with all applications of adhesives for PVG placement. Authors concluded results have established 18F-PET/CT as a useful technique for the diagnosis of PVGI. They further recognized that adhesives that were applied for PVG placement resulting in a patched pattern on 18F-PET/CT does not indicate infection.
Rojoa and associates (2019) 62 A systemic review and meta-analysis were performed to evaluate the diagnostic accuracy of 18F-PET/CT in vascular prosthetic graft infection (VPGI) cases. A total of 433 prostheses were identified in 12 studies. Of those 433, 202 were infected. 18F-PET and 18F-PET/CT were used to analyzed five different methods including: graded uptake, focal uptake, maximum standardized uptake value (SUVmax), tissue to background ratio (TBR), and dual time point (DTP). Pooled estimates were graded based on sensitivity and specificity uptake. Authors concluded 18F-PET/CT demonstrated a high sensitivity when used as a diagnostic tool for VPGI. They note accuracy could be increased by the addition of CT with PET.
Husmann et al. (2018) 63 produced a prospective study that included subjects from the prospective Vascular Graft Infection Cohort Study (VASGRA) if they had microbiologically proven AGI to evaluate the role of 18F-PET/CT in the long-term monitoring of patients. Metabolic activity in 18F-PET/CT was quantified by using maximum standardized uptake value (SUVmax) and was further classified by documenting it as being focal or diffuse. A total of 68 subjects with AGI were evaluated by 266 PET/CTs. A total of 36 examinations were performed after the stop of antimicrobial therapy. Results include “higher C-reactive protein (CRP) (adjusted coefficient per log10 mg/L 0.05 [95% C.I. 0.02–0.08]) was associated with higher SUVmax. CRP, metabolic and clinical findings informed the decision to either start (medians of SUVmax 7.1 and CRP 31.5 mg/L; 100% focal uptake), escalate (SUVmax 9.5; CRP 31.5; 100% focal uptake), continue (SUVmax 6.0; CRP 9.95 mg/L; 90% focal uptake), or stop (SUVmax 4.3; CRP 3.5 mg/L; 61% focal uptake) antibiotic treatment. Of note, decisions to escalate or continue antibiotic treatment were taken despite normal CRP values in 12.5 and 35.7% of PET/CTs, respectively.” Authors conclude PET/CTs taken consecutively may influence decision-making patterns in AGI patients.
Folmer et al. (2018) 64 performed a systematic review and meta-analysis to determine diagnostic value of imaging techniques to diagnosis VGI. A total of 14 articles (eight prospective and six retrospective) were included. WBC scintigraphy ± SPECT/CT (I2 78.6%) demonstrated substantial heterogeneity. CTA (I2 7.4%), 18F-PET (I2 36.5%), and 18F-PET/CT (I2 36.6%) yielded negligible to moderate heterogeneity. Pooled sensitivity yielded 0.67 (95% CI 0.57–0.75) for CTA, 0.94 (95% CI 0.88–0.98) 18F-PET of, 0.95 (95% CI 0.87–0.99) 18F-PET/CT, 0.90 (95% CI 0.85–0.94) WBC scintigraphy, and WBC scintigraphy with SPECT/CT was 0.99 (95% CI 0.92–1.00). Pooled specificities were 0.63 (95% CI 0.48–0.76) for CTA, 0.70 (95% CI 0.59–0.79) for 18F-PET, 0.80 (95% CI 0.69–0.89) for 18F-PET/CT, 0.88 (95% CI 0.81–1.94) for WBC scintigraphy, and WBC scintigraphy SPECT/CT 0.82 (95% CI 0.57–0.96). Results demonstrated that WBC SPECT/CT favors 18F-PET/CT when considering pre- and post-test results, yielding a positive post-test probability of 96% versus 83%. Authors concluded that the greatest method in diagnosing VGI results when diagnostic performance of WBC scintigraphy is joined with SPECT/CT.
Ratliff et al. (2017) 65 performed a retrospective chart review study to identify factors in patients that may increase the risk of vascular graft infections (VGI) following abdominal or lower extremity revascularization surgery. Electronic medical records were reviewed in patients who underwent VGI or lower extremity revascularization surgery which included 28 preoperative, intraoperative, and post-operative factors. A total of 33 cases of VGIs yielded an incidence rate of 15%. Preoperative factors were identified with statistical significance for the development of VGI included sequential procedures (P = .003), diabetes mellitus (P = .002), hemoglobin A1c greater than 7.0 (P = .0002), blood glucose greater than 180 mg/dL (P = .0006), and lack of mobility (0.0097). Associated VGI Intraoperative factors consisted of hemostatic agents when applied to the surgical field intraoperatively (P = .003) and perioperative hypoxemia (P = .027). Associated VGI postoperative influences at discharge from the hospital to skilled nursing facility or acute rehabilitation facility (P = .005) and unscheduled clinic visits (P = .008). Authors concluded the results show vigilance is required to prevent VGI and the importance of identifying associated risk factors.
Keidar and associates (2014) 66 retrospectively mined a database for cancer patients with prosthetic vascular grafts. Data points included graft location, material, and time from surgery. Two nuclear medicine physicians reviewed imagines for consensus considering presence and patterns of increased 18F-FDG uptake. A total of 107 prostheses (43 patients) were identified. Prostheses materials consisted of Dacron (67), Gore-Tex (33) and 7 were native veins. Diffuse homogeneous uptake was noted in 68 and inhomogeneous uptake in 30 grafts. A total of 9 grafts demonstrated no increased 18F-FDG uptake. Diffuse 18F-FDG uptake was found in 92% of noninfected vascular prostheses, with Dacron grafts demonstrating more than with other materials. The degree of 18F uptake remained the same over time. Authors concluded the information of 18F uptake in noninfected vascular prostheses including presence, patterns and persistence could increase the quality of interpretation of 18F-PET/CT studies in suspected prosthetic infection patients.
The European Society for Vascular Surgery (ESVS) published clinical practice guidelines on the management of Vascular graft and endograft infections in 2020.67 ESVS recommends “for patients with a clinical suspicion of vascular graft/endograft infection and with non-convincing findings on CTA, the use of 18F- PET combined with low dose CT is recommended as an additional imaging modality to improve diagnostic accuracy” offering a Class B recommendation “Data derived from a single randomized clinical trials or large non-randomized studies.”
Large-vessel Vasculitis (Aortitis and Systemic Vasculitis)
Angelotti et al.68 (2021) retrospectively assessed 47 patients affected by large-vessel vasculitis (LVV), evaluating clinics, blood chemistry and 18F-PET results to evaluate the utility of 18-FDG, compared with traditional assessments, in the short- (average 8 months after diagnosis) and long-term (average 29 months) follow-up. Patterns of vascular involvement in patients with Takayasu’s arteritis (TAK) and giant cell arteritis (GCA) were compared. Disease activity was assessed by clinical condition and inflammation markers were evaluated at two time points. A significant decrease in the mean value of SUV max in 18F-PET uptake was reported during follow-up in all patients. When comparing 18F-PET uptake and target-to-background ratio (TBR) between short and longer term follow up, the radiologist determined 38 positive 18F-PETs at short-term evaluation. Fifteen remained positive, 23 turned out to be negative and one that was previously negative turned positive. The mean value of SUV max at short term follow-up was 4.09 (SD±2.13) with long-term follow-up of 2.23 (SD±1.13) in all patients. In all patients 18F-PET/CT uptake decreased significantly (p<0.001) during follow-up. All TBR districts analyzed changed significantly (p<0.001) when comparing short- and long-term durations. Authors concluded that results establish 18F-PET as a useful tool in the evaluation of LVV, at diagnosis and monitoring phases of disease. They further state study data confirms that GCA and TAK are part of the same disease range.
Quinn et al (2020) conducted a prospective, observational cohort study to determine the impact of imaging acquisition time on interpretation of disease activity on 18F-PET in large-vessel vasculitis (LVV). A total of 79 patients (GCA=44, TAK=35) were assessed and provided 168 paired one- and two-hour PET studies resulting in 56% (94/168) scans were interpreted as active at the one-hour time point, and 77% (129 scans) were interpreted as active at the two-hour time point (p< 0.01). Results show that patients in the Dual active group (Odds Ratio 1.71, 95%CI 1.06-2.93; p =0.04) and the delayed group were significantly more likely to have clinically active disease (Odds Ratio 1.94, 95%CI 1.13-3.53; p=0.02) when compared to the to the dual inactive group. Authors concluded a large sum of LVV patients were detected as active when delayed imaging was executed.
Lee et al. (2019)69 conducted a systematic review and meta-analysis to evaluate the performance of 18F-PET or 18F-PET/CT in the assessment of disease activity in patients with LVV. Nine studies comprised of 439 18F-PET images from 298 patients resulted in a sensitivity of 0.88 [95% confidence interval (CI) 0.79-0.93] without heterogeneity (χ2 = 14.42, P = .07) and a pooled specificity of 0.81 (95% CI 0.64-0.91) with heterogeneity (χ2= 63.72, P = .00) for the detection of active LVV. The pooled DOR was 30 (95% CI 8-107). Hierarchical sROC curve indicates that the area under the curve was 0.91 (95% CI 0.89-0.94) suggesting good performance for the detection of active disease. Authors concluded the detection of active disease status in patients with LVV was achieved by performing 18F-PET or 18F-PET/CT. Further assessments should be conducted to determine the impact on outcomes of 18F-PET imaging in LVV patients.
A 2018 joint procedural recommendation by European Association of Nuclear Medicine (EANM), Society of Nuclear Medicine and Molecular Imaging (SNMMI), and the PET Interest Group (PIG), and endorsed by the American Society of Nuclear Cardiology (ASNC) 70 regarding 18F-PET/CT(A) imaging in large vessel vasculitis and polymyalgia rheumatica was published. Regarding management of patients with suspected LVV and 18F-PET, a consensus was reached that states “Based on the available evidence, 18F-PET imaging exhibits high diagnostic performance for the detection of LVV and PMR (evidence level II, grade B).” They go on to state “further studies are needed to select the most clinically relevant and reproducible criteria for defining the presence of LVV with 18F-PET, as well as to test the clinical impact of 18F-PET imaging on the management of patients with suspected LVV.” The paper explains that the interpretation of FDG-PET images for LVV can be challenging, and there is currently no consensus on how to interpret the images in this setting established. Additional areas that lack consensus is the role of glucocorticoids and guidelines for image acquisition are lacking. The sensitivity and specificity cited is based on a several meta-analyses of pooled studies, however these meta-analyses were based on case series, retrospective chart reviews and small prospective studies, therefore the quality of literature used was low. Different PET technologies were used across the studies with about 50% using nonhybrid technologies, lack of clear diagnostic criteria for LVV, various conditions of controls and preparation for the studies result in significant heterogeneity and inability to determine if the meta-analysis results are reliable.71
Grayson et al.72 (2018) performed a prospective, longitudinal study with a cohort of patients with large vessel vasculitis to evaluate the clinical value of 18F-PET in a prospective cohort of patients with large-vessel vasculitis (LVV) and disease comparators. A total of 115 subjects (LVV=56; comparators=59) provided 170 18F-PET scans. Clinically active LVV and disease comparators were differentiated utilizing 18F-PET with a sensitivity=85% (95%CI: 69–94%) and specificity=83% (95% CI: 71–91%). A result of 58% (41/71) were classified as active vasculitis in LVV patients in clinical remission. In clinical remission patients who underwent 18F-PET it was determined that future relapse was more common in elevated PETVAS scores (45% versus 11%, p=0.03) over 15 months. Authors concluded that 18F-PET was beneficial in the assessment of LVV aside from clinical assessment alone and future clinical relapse was associated with 18F-PET scan activity during clinical remission.
Barra and associates73 (2018) performed a systematic review and meta-analysis to determine the effectiveness of imaging modalities for the management of Takayasu's Arteritis (TAK). A total of 57 studies were included with most lacking robust design, quality and adequate sample size. In regard to TAK diagnosis, ultrasound (US) demonstrated a lower pooled sensitivity of 81% (95% CI: 69-89%) when compared to Magnetic Resonance Angiography (MRA) with sensitivity (SN) 92% (95% CI: 88-95%), and US and MRA yielded specificities (SP) >90%. Computed tomography angiography (CTA) resulted in a SN and a SP for TAK diagnosis of >90%. Disease prediction varied across studies when considering the utility of vessel wall thickening and enhancement by MRA and CTA. Pooled results for disease activity utilizing 18F-PET were SN of 81% (95% CI: 69-89%) and SP of 74% (95% CI: 55-86%). Authors concluded although US, CTA and MRA are effective in the diagnosis of TAK, the utility remains uncertain when considering assessing disease activity.
Puppo et al.74 (2014) executed a systematic review to assess the various qualitative and semiquantitative methods for diagnosis and grading of vascular inflammation in giant cell arteritis (GCA) patients (with or without associated polymyalgia rheumatica (PMR)) utilizing 18F-PET. TAK was excluded from analysis. A total of 19 articles (13 prospective and 6 retrospective) were included resulting in a total of 442 patients (with or without PMR symptoms) and 535 controls. Vascular inflammation was assessed in GCA in 10 studies that used qualitative 18F uptake criteria, 6 used semiquantitative criteria, and 3 utilized qualitative and semiquantitative criteria. Diagnostic GCA-related vascular inflammation utilizing 18F-PET scans to determine severity based on diagnostic performance and examining their respective advantages and limitations. Authors concluded most studies reviewed used the qualitative methods of 18F-PET image assessment. Whereas the aortic to blood pool uptake ration for the aortic arch proved to be more accurate among semiquantitative approaches.
The role of PET imaging for monitoring disease progression is under investigation. There is limited data based six retrospective studies and one prospective study included in the Slart review.71 The timeline for the PET imaging to return negative is not established therefore the length of time before performing a follow-up study or if a follow-up PET scan offers any benefit in clinical management. These consensus guidelines conclude that FDG-PET/CT may be of value for evaluating response to treatment based on level III evidence with grade C recommendation and state additional studies are warranted. 71
Cardiac Amyloidosis
The ASNC/AHA/ASE/EANM/HFSA/ISA/SCMR/SNMMI published a 2-part expert consensus recommendation for multimodality imaging in cardiac amyloidosis: evidence base and standardized methods of imaging in which states while molecular imaging techniques show promise, limited data is available therefore further studies are needed for F-labelled PET tracers to detect early amyloidosis. 75,76
Device Infections
The Journal of the American College of Cardiology published an expert panel statement on best practices for cardiac device-related infections and endocarditis.77 They stated that transthoracic (TEE) and transesophageal (TEE) echocardiogram are the first line imaging test for patients with suspected cardiac implantable electronic devices (CIEDs) including pacemakers and implantable cardioverter-defibrillators, prosthetic valves and left ventricular assist devices (LVADs). Echocardiography has been assigned a Class I recommendation by expert guideline documents for any suspicion of endocarditis. However, TTE alone has been shown to have inadequate sensitivity for detecting prosthetic valve vegetations (around 50%), prosthetic valve abscess (30%-50%), and CIED lead infections (25%-40%) with TEE resulting in improved sensitivity of 90% or greater. In suspected prosthetic valve or CIED infections a negative echocardiography does not completely exclude the possibility of infections in patients with a high clinical suspicion and sometimes other processes can be misdiagnosed as infection therefore false negative and false positive interpretations are possible. Cardiac CT plays a role but has notable limitations in detection of CIEDs infections. 18F-PET/CT has been found to be sensitive during the early phase of the infectious process with a high sensitivity for CIEDs infections. For CIED infection evaluation, a meta-analysis showed that 18F-PET/CT had a pooled sensitivity and specificity of 87% and 94%, respectively and for detecting PVE, 18F-PET/CT had a sensitivity in the range of 73% to 100% and a specificity of 71% to 100%.46,77,78 18F-PET/CT has been shown in several studies and a systematic review to accurately detect infection in LVAD in equivocal cases and has been proposed to be superior to CT for predicting clinical outcomes.77,79,80 It also has been shown to diagnose earlier than TEE before anatomical damage ensures and by adding 18F-PET/CT findings as an additional major criterion to the modified Duke Criteria increases sensitivity from 52-70% to 91-97%.49, 50, 81-83
Joint Arthroplasty
Periprosthetic joint infection (PJI) occurs in around 0.5 to 1.0 percent of hip arthroplasties. This percentage is expected to rise due to an aging population. The occurrence of PJI can lead to significant quality of life impairment, including chronic infection until loss of limb or amputation.84
UpToDate states “positron emission tomography (PET) scans, computed tomography (CT) scans, magnetic resonance imaging (MRI) scans, or bone scans are not useful for routine diagnostic evaluation in most cases of suspected prosthetic joint infection (PJI).” False positives have also been presented when 18F-PET/CT was conducted in the first year of arthroplasty. When PJI was predetermined, the accuracy of 18F-PET was higher in THA infection.84
The American College of Radiology (ACR)85 constructed an appropriateness criterion for imaging following total hip arthroplasty. In-111 WBC and Tc-99m sulfur colloid scan hip was often considered the best imaging test for evaluating patients with a painful primary total hip arthroplasty where infection was not excluded. The 2023 updated guidelines state for “symptomatic hip arthroplasty, infection not excluded” additional imaging following radiographs that PET/CT imaging is “Usually not appropriate” and has high levels of radiation exposure. The guidelines state reported results for diagnosis have been inconsistent wide variations in reported sensitivity and specificity. The study also cannot distinguish infection from aseptic inflammation and another study cited reported false-positive rate of FDG PET/CT compared with culture alone of 77%. There was no advantage found for Fluoride PET/CT over FDG. 86
The SNMMI Appropriate Use Criteria for use of Nuclear Medicine in Musculoskeletal Infection Imaging scores the diagnosis of prosthetic joint injection of the hip with 18F-PET/CT as AUC of 7 (Appropriate). This recommendation was based on 3 systematic reviews and meta-analysis which were rated as fair for level of evidence. These 3 papers all reported sensitivities and specificities above 83%.87
The American Academy of Orthopaedic Surgeons (AAOS) produced Evidence-Based Clinical Practice Guideline on the Diagnosis and Prevention of Periprosthetic Joint Infections.88 Limited strength evidence supports the use of 18F-PET/CT, 18F-NaF and CT to aid in the diagnosis of PJI. Strength of Recommendation: Limited – “Description: Evidence from two or more “Low” quality studies with consistent findings or evidence from a single “Moderate” quality study recommending for or against the intervention or diagnostic test or the evidence is insufficient or conflicting and does not allow a recommendation for or against the intervention.” 88
A consensus document was produced by the European Association of Nuclear Medicine (EANM), European Society of Radiology (ESR), European Bone and Joint Infection Society (EBJIS), and with European Society of Clinical Microbiology and Infectious Diseases (ESCMID) endorsement.89 A consensus was reached that 18F-PET in patients with suspected prosthetic joint infection demonstrated a high sensitivity but lower specificity when compared to white blood cell scintigraphy or anti-granulocyte antibody scintigraphy with a level of evidence: 2.
Jin and colleagues90 (2014) conducted a systematic review and meta-analysis of published literature on the diagnosis of prosthetic infection after arthroplasty. Fourteen studies with 838 prostheses with suspicious of prosthetic infection after arthroplasty were included in the meta-analysis. “A pooled sensitivity of 18F-PET or 18F-PET/CT in detecting prosthetic infection was 86% (95% confidence interval [CI] 82-90%) on a per prosthesis-based analysis. The pooled specificity of 18F-PET or 18F-PET/CT in detecting prosthetic infection was 86% (95% CI 83-89%) on a per prosthesis-based analysis. The area under the ROC curve on a per prosthesis-based analysis was 0.93.” Authors concluded that patients suspicious of prosthetic infection, 18F-PET or 18F-PET/CT demonstrated high sensitivity and specificity but report possible sources of false positive results. Limitations included statistically heterogeneous studies, inclusion of studies with limited power, moderate quality, and small sample sizes. The authors report no conflicts of interest.
Authors van der Bruggen et al.91 (2010) conducted a systematic review. PubMed/MEDLINE and Embase were searched for publications on SPECT and 18F-PET on osteomyelitis and prosthetic bone and joint infections. The search produced 44 articles (15 for SPECT and 29 for 18F-PET comprised of 1634 patients (580 patients SPECT, 1054 patients 18F-PET). In 18F-PET performance in orthopedic implant infections, sensitivity varies widely from 28% to 91% and specificity from 9% to 97%. The highest diagnostic accuracy of for diagnosis of bone and joint infections is achieved with combined 111In-WBC and 99mTc-sulfur colloid for SPECT was reported as 95%. Authors concluded SPECT/computed tomography (CT) with 111In-WBC combined with 99mTc-MDP or 99mTc-sulfur colloid is the ideal imaging technique for diagnosis of bone and joint infections. While 18F-PET is useful for diagnosis of osteomyelitis which enables more accurate localization with improved spatial resolution over SPECT imaging. Adding CT further improves localization. Limitations include small sample sizes, low and moderate levels of evidence in included studies, with a wide range inclusion criterion and outcomes.
Mushtaq92 and associates (2019) performed a review on radiological imaging modalities in the evaluation of the failing total hip arthroplasty. Despite 18F-PET enabling visualization of inflammatory cells during infection, authors state published results to date are inconclusive and have contradictory findings.
Kumar and colleagues (2016)93 conducted a prospective study to evaluate 57 hip components in 45 patients (bilateral prostheses in 12 patients, 45 painful and 12 asymptomatic contralateral hip components) with dual phase 18F-PET/CT (DPFP) and three-phase bone scan (TPBS). The aim was to establish the clinical utility of dual phase 18F-PET/CT in preoperative diagnosis of implant loosening, differentiation between septic and aseptic loosening and to compare the diagnostic accuracy of DPFP and TPBS. Two expert nuclear medicine physicians were blinded (clinical findings and final diagnosis) evaluated skeletal scintigraphy and 18F-PET/CT findings. Authors concluded that 18F-PET/CT is a useful modality for diagnosing septic loosening of hip prostheses. Diagnostic DPRP imaging (early and delayed) demonstrated potential to differentiate septic and aseptic loosening and over TPBS. Limitations include small sample size, and lack of histopathology and microbiological confirmation on asymptomatic stable joints. One patient experienced a high tracer uptake resulting in the 18F-PET/CT being unable to differentiate metastatic deposit from septic loosening. Authors report no conflicts of interest.
Basu et al. (2014) performed a prospective study to assess and compare the value of 18F-PET to combined 111In-labeled leukocyte/99mTc-sulfur colloid bone marrow (WBC/BM) imaging for diagnosing infection in hip and knee prostheses.94 Suspected infected or noninfectious loosening of 221 prostheses (134 hip and 87 knee) were evaluated. WBC/BM imaging and 18F-PET was performed in 88 prostheses and 18F-PET was performed on all 221 prostheses. The evaluation of hip arthroplasty using 18F-PET yielded a sensitivity, specificity, PPV, NPV, and AUC were 81.8%, 93.1%, 79.4%, 94.0%, and 0.874, respectively. The evaluation of hip arthroplasty using WBC/BM imaging yielded a sensitivity, specificity, PPV, NPV, and AUC of 38.5%, 95.7%, 71.4%, 84.6%, and 0.671, respectively. There was no significant difference in sensitivity (P=0.0625) or specificity (P=1.000) in those cases that underwent both 18F-PET and WBC/BM imaging modalities. The evaluation of knee arthroplasty using 18F-PET yielded a sensitivity, specificity, PPV, NPV, and AUC were 94.7%, 88.2%, 69.2%, 98.4%, and 0.915, respectively. The evaluation of knee arthroplasty using WBC/BM imaging yielded a sensitivity, specificity, PPV, NPV, and AUC were 33.3%, 88.5%, 25.0%, 92.0%, and 0.609, respectively. Cases that underwent both 18F-PET and WBC/BM imaging yielded no significant differences in sensitivity and specificity (P=0.500 and P=1.000, respectively). Authors conclude use of 18F-PET scan in detecting infection in painful hip and knee prostheses is optimal based on diagnostic performance in this study and when considering the cost and safety issues associated with WBD/BM imaging. Limitations include a limited number of patients had both studies, not all patients with knee arthroplasty were assessed with both imaging modalities. This study made use of standalone 18F-PET rather than 18F-PET/CT which is more commonly used.
Koob et al. (2019) conducted a retrospective analysis to assess the sensitivity and specificity in diagnosing periprosthetic loosening in total hip and knee arthroplasty utilizing 18F-Fluoride 18F-PET/CT in 26 patients.95 A sensitivity of 95.00%, a specificity of 87.04% and an accuracy of 89.19% for 18F- PET/CT. Authors conclude 18F-PET/CT is helpful in diagnosing periprosthetic loosening of total hip and knee arthroplasties, further investigation is needed with larger sample sizes and more focus on uptake kinetics to assess 18F-PET/CT in periprosthetic joint infection to establish its value in distinguishing aseptic from septic loosening. Authors report no conflicts of interest.
Verberne et al (2016) conducted a systematic review and meta-analysis and the accuracy of imaging techniques and the assessment of periprosthetic hip infection. This report was given Level of Evidence rating of IV. They reported a pooled sensitivity and specificity of 86% (95% CI, 80% to 90%) and 93% (95% CI, 90% to 95%) for 18F-PET/Ct They acknowledge that greater than 50% of the studies had insufficient information and risk of bias was a concern. They conclude “FDG PET has an appropriate accuracy in confirming or excluding periprosthetic hip infection, but may not yet be the preferred imaging modality because of limited availability and relatively higher cost”.96
Kwee et al. (2008) conducted a systematic review and meta-analysis on published data on the diagnostic performance of 18F-PET/CT in detecting prosthetic hip or knee joint infections. Eleven studies were included with a total sample size of 635 prosthesis reporting a pooled sensitivity and specificity of 82.1% (95%CI = 68.0–90.8%) and 86.6% (95%CI = 79.7–91.4%), respectively for prosthetic hip or knee joint infections. High heterogeneity among the individual studies was present (I 2 = 68.8%). Subgroup analysis revealed the overall specificity was higher in hip than knee. The authors reported an overall moderate to high diagnostic performance of 18F-PET, however due to the high heterogeneity caution is warranted and they state further studies should be conducted to explore the potential causes of this heterogeneity and validate the use of-PET for diagnosis of prosthetic joint infections.97
SMEs discussed there is a limited role in evaluation of arthroplasty and that laboratory analysis and aspiration of joint with cultures is standard evaluation. This indication is challenged by lower quality and greater variability in the evidence but feels there is a role in complicated cases. They also discussed the potential to replace the current 3-phase bone scan which is time-consuming, labor-intensive approach that also delays results and could potentially simplify management. There is also lack of clarity within the literature on the duration of time after surgery to obtain the most accurate results with 1-2 years being discussed from literature sources.
Chronic Osteomyelitis
Osteomyelitis is an infection involving the bone and may be classified based on mechanism of infection (hematogenous versus nonhematogenous) and illness duration. Diagnosis of osteomyelitis is established via culture obtained from biopsy of the involved bone and should be performed before antibiotic therapy is initiated.98 Radiographic imaging can be helpful in the diagnosis, delineating the extent of disease, and planning therapy in patients with osteomyelitis. Magnetic resonance imaging (MRI) is the imaging modality with greatest sensitivity and a negative predictive value for diagnosis of osteomyelitis. An alternative test may include 18F-PET, computed tomography (CT), tagged white blood cell (WBC) or three-phase bone scan plus/minus, or single photon emission CT (SPECT-CT) in cases where MRI is not feasible.99
SMEs discussed there are multiple entities to consider including acute vs. chronic osteomyelitis, diabetic foot and other locations to consider, but there is overall a lack of high-quality evidence for the role of 18F-PET scan. Current standard for diagnosis of osteomyelitis is MRI and some studies suggest 18F-PET may have similar sensitivity. For diabetic foot consideration of blood sugar control must be made as this is important to get an accurate 18F-PET scan. Also, diagnosis typically requires biopsy and while imaging may guide determining the location for biopsy it will not typically replace biopsy which is necessary to confirm the diagnosis and for cultures. Once advantage to 18F-PET scan is the whole-body scan to understand extent of disease. This may also impact staging of disease and decision making.
A consensus document100 for the diagnosis of peripheral bone infection in adults by European Bone and Joint Infection Society, European Society of Microbiology and Infectious Diseases, European Society or Radiology, and European Association of Nuclear Medicine state the gold standard for diagnosis of bone infection is bone biopsy and culture of affected bone. This should be performed under image guidance typically CT guidance however X-ray of fluoroscopy may be utilized. MRI has a high diagnostic performed radiation with minimal radiation exposure and wide availability. 18F-PET/CT has high sensitivity and specificity implants or recent surgery or fractures, but also has higher radiation exposure. The panel states Level of Evidence 2 for 18F-PET/CT as having high diagnostic accuracy in peripheral bone infection without fracture and osteosynthesis. The panel did agree there is not sufficient evidence for use of this modality as the current reference standard and is typically performed in chronic osteomyelitis to the role in acute phase is still unknown. Comparative studies to WBS scintigraphy are too small to be definitive.
Wang et al. conducted a meta-analysis including 23 studies on 851 patients evaluating the role of 18F-PET, three-phase bone scintigraphy, leukocyte scintigraphy, and MOAB scintigraphy in the assessment of suspected osteomyelitis. They report that 18F-PET had a pooled sensitivity of 0.923, specificity of 0.920, and AUC of 0.9666, whereas for bone scintigraphy, the corresponding values were 0.827, 0.446, and 0.6514. The authors conclude 18F-PET is superior to other radionucleotide imaging modalities for the detection of chronic osteomyelitis.101
Kulkarni and colleagues (2020) conducted a retrospective analysis to evaluate the diagnostic performance of 18F-PET/CT in patients with skull base osteomyelitis (SBO) compared to magnetic resonance imaging (MRI) whenever available. Seventy-seven patients with suspected skull base osteomyelitis (SBO) who underwent regional 18F-PET/CT were included in the review. Images were analyzed for 18F uptake. Diagnostic performance of 18F-PET/CT was analyzed based on histopathology, culture, and clinical/imaging follow-up. Fifty-six patients underwent both 18F-PET/CT and MRI. Agreement analysis between the modalities yielded agreement for delineation of soft tissue and bony involvement with K values of 0.82 and 0.81, respectively. Authors concluded 18F-PET/CT was a sensitive tool in evaluation of patients with SBO. 18F-PET/CT shows a good agreement with the MRI. This study is limited by the retrospective study design, potential selection bias and a lack of histopathology across the study population.102
Recommendations for the treatment of osteomyelitis (2014) were published after a multi-professional team conducted a literature review. The aim was to produce a guided medical approach to different types of osteomyelitis, aiming to obtain better clinical outcomes and reducing the social costs of this disease. A multidisciplinary approach while considering palliative treatment and a curative approach is important in the assessment of each case. Treatment for chronic osteomyelitis include “correct microbiological diagnosis; improvement of the host’s defenses; stabilization of underlying diseases; correct anatomical localization of bone involvement; adequate antimicrobial therapy; surgical debridement of all devitalized tissue; repair of soft tissues; and bone reconstruction and rehabilitation.” 103
Termaat et al. (2005) conducted a systematic review and meta-analysis to determine the accuracy of current imaging modalities in the diagnosis of chronic osteomyelitis. Each imaging technique was evaluated by determining its sensitivity and specificity compared with the results of histological analysis, findings on culture, and clinical follow-up of more than six months.104 Twenty-three clinical studies included accuracy information for radiography (2 studies), magnetic resonance imaging (5), computed tomography (1), bone scintigraphy (7), leukocyte scintigraphy (13), gallium scintigraphy (1e), combined bone and leukocyte scintigraphy (6), combined bone and gallium scintigraphy (3), and fluorodeoxyglucose positron emission tomography (4) and were included in the review. Pooled sensitivity established that 18F-PET was the most sensitive technique, with a sensitivity of 96% (95% confidence interval, 88% to 99%) compared with 82% (95% confidence interval, 70% to 89%) for bone scintigraphy, 61% (95% confidence interval, 43% to 76%) for leukocyte scintigraphy, 78% (95% confidence interval, 72% to 83%) for combined bone and leukocyte scintigraphy, and 84% (95% confidence interval, 69% to 92%) for magnetic resonance imaging. Pooled specificity established that bone scintigraphy had the lowest specificity, with a specificity of 25% (95% confidence interval, 16% to 36%) compared with 60% (95% confidence interval, 38% to 78%) for magnetic resonance imaging, 77% (95% confidence interval, 63% to 87%) for leukocyte scintigraphy, 84% (95% confidence interval, 75% to 90%) for combined bone and leukocyte scintigraphy, and 91% (95% confidence interval, 81% to 95%) for PET scan. The sensitivity of leukocyte scintigraphy in detecting chronic osteomyelitis in the peripheral skeleton was 84% (95% confidence interval, 72% to 91%) compared with 21% (95% confidence interval, 11% to 38%) for its detection of chronic osteomyelitis in the axial skeleton. The specificity of leukocyte scintigraphy in the axial skeleton was 60% (95% confidence interval, 39% to 78%) compared with 80% (95% confidence interval, 61% to 91%) for the peripheral skeleton. Authors concluded 18F-PET exhibited the highest diagnostic accuracy for confirming or excluding the diagnosis of chronic osteomyelitis. While leukocyte scintigraphy demonstrated an appropriate diagnostic accuracy in the peripheral skeleton, 18F-PET was superior for detecting chronic osteomyelitis in the axial skeleton. This analysis was limited by the overall low quality of the evidence with the author rating the level of evidence as III due to multiple factors including non-consecutive studies, studies without consistently applied reference standards, lack of blinding, small sample sizes and other factors.
The SNMMI report for the diagnosis of uncomplicated peripheral bone osteomyelitis 18F-PET/CT was given an Appropriate Use Criteria (AUC) of nine. The report cited two systematic reviews101,105 with fair level of evidence.87 For complicated peripheral bone osteomyelitis, including orthopedic hardware infection, the AUC was eight. This was based on one systematic review104. The overall level of evidence was rated fair. They conclude 18F-PET/CT as the nuclear medicine imaging test of choice for the diagnosis of uncomplicated and complicated peripheral bone osteomyelitis.
The ACR reports the following: suspected osteomyelitis 18F-PET AUC rating of 2 and suspected osteomyelitis at sight of previous nonarthoplasty hardware AUC rating 2. They report that PET imaging has shown a high accuracy in the detection of osteomyelitis in patients with prior surgery, trauma and orthopedic hardware and a recent meta-analysis confirmed the superiority of PET to other radionucleotide examinations with pooled sensitivity of 92% and specificity of 92% and another meta-analysis reporting 96% and 91%, respectively. However, they acknowledge the challenge of differentiating between infection and inflammation on 18F-PET/CT in the acute postoperative or post traumatic setting as tracers may not normalize until 3 to 4 months after surgery or trauma, and a high false positive rate from fractures. They state the diagnostic role for 18F-PET for combined evaluation of acute and chronic bone and soft tissue infections continue to evolve and the capability for diagnosis of chronic osteomyelitis and spinal infections offer great promise, but data is limited.106
Diabetic Foot
Llewellyn and associates (2020) published a systematic review and meta-analysis to review the evidence on the diagnostic accuracy of imaging tests to diagnose osteomyelitis in people with diabetic foot ulcers.107 36 studies were evaluated and reported the sensitivities of MRI (96.4%), PET scans (84.3%), SPECT (95.6% but only based on 3 studies), scintigraphy (84.2%), and X- rays (61.9%). The authors conclude that MRI and PET scan both reliably diagnosed osteomyelitis and diabetic foot ulcer patients with Maria being the current preferable test given its wider availability and lack of potential harmful radiation.
Lauri et al.108 (2020) conducted a retrospective multicenter study to compare the diagnostic accuracy of white blood cell scintigraphy (WBC), 18F-PET/CT, and Magnetic Resonance Imaging (MRI) in patients with suspected diabetic food infection (DFI). A total of 251 patients from five centers were enrolled and clinical and imaging data was reviewed. Sensitivity, specificity and accuracy of WBC, FDG, and MRI for diagnosing osteomyelitis (OM), soft-tissue infection (STI), and Charcot osteoarthropathy. Median values of CRP and ESR were significantly higher in OM patients when compared to patients without infection (p = 0.017 and p = 0.027 respectively) in a post hoc analysis. In patients with STI and normal subjects and between patients with OM and STI, no significant difference was observed. When information on the pathogen casing the infection could not be obtained, biopsy was used as the gold standard for final diagnosis in those 36 patients. A total of 67 subjects underwent skin cultures and 14 of 50 subjects underwent pre- or intra-operative biopsies resulting in causative pathogens being recorded. Clinical follow up was utilized in the remaining 121 subjects to establish a final diagnosis. Results included OM (n=93), STI (n=76), and Charcot (n=10). No pathology was required in 72 subjects according to reference standard. Imagining modalities resulted in WBC scintigraphy (n=119), 18F-PET/CT (n=46), and MRI (n=59). WBC and 18F-PET/CT were performed in 10 subjects, WBC scintigraphy and MRI were performed in 15 subjects, and all three imaging modalities were performed in 2 subjects. When considering STI, FDG and WBC reached a significantly higher specificity than MRI (97.9% and 95.7% vs. 83.6%, p = 0.04 and p = 0.018, respectively). In Charcot, WBC was more specific than MRI (89.3% vs. 88.2% p < 0.0001) and MRI and WBC were significantly higher specificity and accuracy than FDG (88.2% and 89.3% vs. 62.5%, p = 0.0009; 80.3% and 87.9% vs. 62.1%, p < 0.02, respectively). When considering WBC in patients with suspected DFI, the utilization of EANM guidelines resulted in the most reliable imaging modality to differentiate between OM, STI, and Charcot. Authors concluded the study results confirm WBC scintigraphy is the superior imaging modality when delineating pedal OM, STI, and Charcot. Examinations results include high sensitivity, specificity, and accuracy when European Society of Nuclear Medicine (EANM) guidelines are applied. Authors recognize the need for additional research with more robust study designs.
The same author conducted a 2017 systematic review and meta-analysis on 27 articles evaluating compared the diagnostic performances of MRI, radiolabeled white blood cell (WBC) scintigraphy (either with 99mTc-hexamethylpropyleneamineoxime [HMPAO] or 111In-oxine), and 18F-PET in diagnosis of diabetic foot infections.109 The authors report the sensitivity of 18F-PET/CT as 89%, 111In-oxine as 92%, WBC scan as 91% and MRI as 93%. Specificities were 92%, 75%, 92% and 75% respectively. They conclude that the 18F-PET/CT and WBC scan offered highest specificity but larger prospective studies with direct comparison among the imaging techniques are required. This study was limited by the number of published articles, variability in technique, small number of patients in most studies, risk of biased and variability and diagnostic criteria of osteomyelitis.
Diez et al. (2020) performed a prospective study to compare the diagnostic accuracy of diffusion-weighted imaging (DWI) and dynamic contrast-enhanced-magnetic resonance imaging (DCE-MRI) involving two region of interest (ROI) sizes with 18F-PET/CT to differentiate diabetic foot osteomyelitis (DFO) from Charcot neuro-osteoarthropathy (CN). Thirty-one diabetic patients were included. Two readers independently evaluated DWI and DCE-MRI parameters using two different ROI sizes, and 18F-PET/CT parameters (visual assessment, SUVmax, delayed SUVmax, and percentage changes between SUVmax and delayed SUVmax). Authors concluded DWIr, Ktrans and iAUC60 allowed reliable differentiation of DFO and CN. The 18F-PET/CT was the most accurate technique for differentiation in visual assessment. 110
Kagna and associates (2012) conducted a prospective study assess the value of 18F-PET/CT in diabetic patients with clinically suspected osteomyelitis. A total of 39 consecutive diabetic patients were enrolled. 18F-PET/CT was evaluated, and final diagnosis was founded on histopathology and bacteriology of surgical samples, or clinical and imaging follow-up yielding in 18 confirmed cases of osteomyelitis and exclusion of 21 sites. Twenty lesions with focal bone FDG uptake, yielded 2 that were false-positive with no further evidence of osteomyelitis. Diffuse FDG uptake involving more than one bone on CT were correctly diagnosed as diabetic osteoarthropathy in 5 sites. Results of 18F-PET/CT yielded a sensitivity, specificity, and accuracy of 100%, 92% and 95% in a patient-based analysis and 100%, 93% and 96% in a lesion-based analysis, respectively, for the diagnosis of osteomyelitis in the diabetic foot. Authors concluded “18F-PET/CT was found to have high performance indices for evaluation of the diabetic foot. The PET component identified FDG-avid foci in sites of acute infection which were precisely localized on fused 18F-PET/CT images allowing correct differentiation between osteomyelitis and soft-tissue infection.” 111
Schwegler et al. (2008) performed a prospective study in 20 diabetic patients with a chronic foot ulcer (>8 weeks) without antibiotic pretreatment and without clinical signs for osteomyelitis. With the purpose of assessing the prevalence of clinically unsuspected osteomyelitis and to compare the value of magnetic resonance imaging (MRI), 18F-PET and 99mTc-labelled monoclonal anti granulocyte antibody scintigraphy (99mTc-MOAB). Bone biopsy confirmed presence of osteomyelitis in seven of the 20 clinically unsuspected foot ulcers. MRI was positive in six of the seven patients with proven osteomyelitis, whereas 18F-PET and 99mTc-MOAB were positive only in (the same) two patients. Authors conclude that MRI appears to be superior to 18F-PET and 99mTc-MOAB in detecting foot ulcer-associated osteomyelitis. 112
Spondylodiscitis (SD)
Guidance was published by the Society of Nuclear Medicine and Molecular Imaging and the American College of Nuclear Medicine (SNMMI) to describe appropriate use of nuclear medicine imaging in patients suspected of having various musculoskeletal infections.87 Appropriate use was based on expert panel consensus and evidence review from existing systematic reviews. In all the scenarios reviewed below evidence was rated fair unless there was no evidence to review.
- Diagnosis of spondylodiscitis in patients without spinal hardware, 18F-PET/PET-CT (Score 9 – Appropriate) based on fair evidence and expert opinion.
- Diagnosis of spondylodiscitis in patients with spinal hardware, 18F-PET/PET-CT (Score 8 – Appropriate)- based on fair evidence and there were higher false positive rates in the present of spinal hardware than without (12.8% vs. 7%) based on fair evidence and expert opinion.
- Diagnosis of uncomplicated peripheral bone osteomyelitis, 18F-PET/PET-CT (Score 9 – Appropriate) based on fair evidence (from 2 SRs) and expert opinion.
- Diagnosis of complicated peripheral bone osteomyelitis, including orthopedic hardware infection, 18F-PET/PETCT (Score 8 – Appropriate).
- Diagnosis of foot osteomyelitis in diabetic patients, 18F-PET/PET-CT (Score 8 – Appropriate)
- Diagnosis of PJI of the hip, 18F-PET/PET-CT (Score 7 – Appropriate)
- Diagnosis of PJI of the knee, 18F-PET/PET-CT (Score 6 – May be Appropriate)
- Diagnosis of PJI of the shoulder, 18F-PET/PET-CT (Score 5 – May be Appropriate)- no evidence to review.
- Diagnosis of septic arthritis, 18F-PET/PET-CT (Score 6 – May be Appropriate)- no evidence to review.
A joint consensus document was published by EANM/ESNR and ESCMID-endorsed for the diagnosis of spine infection (spondylodiscitis) in adults. 113
- In primary and post-surgical SD, if MRI is contraindicated, the imaging modality of choice is 18F-PET/CT Level of evidence: 2.
- In post-surgical SD, with or without spinal hardware, 18F-PET/CT can detect both spine infection and soft tissue infection Level of evidence: 2.
- In patients with suspected spine infection and elevated ESR and/or CRP and doubtful MRI, 18F-PET/CT should be performed Level of evidence: 1.
- In patients with suspected spine infection, elevated ESR and/or CRP, doubtful or unperformable MRI, and doubtful or unperformable 18F-PET/CT, a CT scan should be performed with an image-guided biopsy. Level of evidence: 2
- The role of hybrid PET/MRI, although promising, needs to be evaluated. Level of evidence: 5
- In case of negative MRI or negative 18F-PET/CT, the diagnosis of SD should be excluded. Level of evidence: 5
- In patients with SD diagnosed by 18F-PET/CT, a second 18F-PET/CT scan can be performed to evaluate the response to antibiotic therapy. Level of evidence: 4
ACR 2021 Appropriateness Criteria® Suspected Spine Infection states FDG-PET is “Usually not Appropriate” for suspected spine infections as in initial imaging or after recent intervention. they state it “May be Appropriate” for suspected spine infections with abnormal radiographs or CT findings as a next imaging study.114
A 2014 meta-analysis reported on patient based meta-analysis of diagnostic data for 18F-PET/CT against clinical, laboratory and or radiological evidence of disease. 12 studies (n=224) and combined sensitivity across studies was 0.97 [95% confidence interval (CI), 0.83–1.00], the specificity was 0.88 (95% CI, 0.74–0.95), and the area under the curve was 0.98 (95% CI, 0.96–0.99). The authors conclude it is an excellent diagnostic test for spondylodiscitis.115 A systematic review was not included as part of this analysis and there was no evaluation of quality of literature therefore the limitations of the included study is uncertain and the subsequent meta-analysis results are uncertain.
A 2020 systematic review and meta-analysis reported on the role of 18F-PET/CT in patients with spinal infection (SI). Twenty-six articles(n=833) were included, and 12 studies (n=396) selected for meta-analysis. The bivariate meta-analysis on 18F-PET/CT in patients with suspected SI was reported as: sensitivity 94.8% (95% CI 88.9– 97.6%) and specificity 91.4% (95% CI 78.2–96.9%). Heterogeneity was reported as low (I2 <50%). The authors conclude that 18F-PET/CT offers good diagnostic performance in patients with SI and can be used in patients where MRI cannot be performed or is non diagnostic or inconclusive.116 Limitations of the meta-analysis include that the studies were all retrospective or small prospective studies and there were no randomized controlled trials, therefore the included studies had lower quality of evidence intrinsic to their study design and weaken the results of the meta-analysis.
Corticosteroid, antiepileptics, catecholamines and chemotherapy (within 1 month of end of therapy) can alter the qualitative analysis of 18F-PET and must be considered for patient selection.113 While sensitivity and negative predictive value is high for 18F-PET/CT it may be reduced by presence of tumor, degenerative spinal disease and/or spinal implants. In one small prospective study MRI was superior for abscess detection while 18-PET/CT was better for metastatic infection.117,118
