Response to Comments: Positron Emission Tomography (PET) for Inflammation and Infection

The American Society of Nuclear Cardiology and the American College of Cardiology submitted multiple comments addressed in comments #1-9.

“ASNC and ACC reviewers recommend adding “device infections (pacemaker, defibrillators, LVAD, metallic implants suspected” to the list of conditions for which a patient should be undergoing evaluation for PET imaging coverage. Though transthoracic and transesophageal echocardiography are generally first line tests for suspected endocarditis and for assessing hemodynamic complications, recent literature suggests “that cardiac computed tomography (CT) or CT angiography and functional imaging with 18F-fluoro-2-deoxyglucose (FDG) positron emission tomography (PET) with CT (FDG PET/CT) may have an incremental role in technically limited or inconclusive cases on echocardiography.” Most notably, FDG PET/CT can detect inflammatory cells early in the infection process before morphological damage takes place.” Supporting references were provided Dilsizian 2022, Ferro 2023, Ten Hove 2022, Roy SG 2023.

Thank you for your comments. The literature submitted has been reviewed and added to the LCD. Based on this review the policy has been expanded to include PET imaging coverage for suspected device infections in equivocal cases.

ASNC and ACC suggest “adding aortitis and systemic vasculitis to the list of conditions for which a patient should be undergoing evaluation for infection and inflammation PET coverage. In a multi societal joint procedural recommendation by the EANM, SNMMI, the PET Interest Group (PIG), and endorsed by the ASNC, FDG-PET imaging was noted to exhibit high diagnostic performance for the detection of large vessel vasculitis. FDG-PET and CTA have complementary roles in the diagnosis of large vessel vasculitis, but most importantly FDG-PET/CT(A) may be of value for evaluating response to treatment by monitoring functional metabolic information and detecting structural vascular changes.” Supporting reference Slart 2018 was provided. 

Slart 2018 was included in the evidence review of the LCD. These joint procedural recommendations state that based on the available evidence PET imaging exhibits high diagnostic performance for LVV (evidence level II, grade B) and goes on to explain “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 sensitivity of 18F-PET is reported in meta-analysis (Lee 2019 & Barra 2018) of 81-88%. However, the included studies were case series, retrospective chart reviews and small prospective studies, therefore the quality of literature in the meta-analysis was low reducing the reliability of its conclusions. However, based on the high sensitivity it will be added to the LCD in limited coverage capacity for inconclusive cases when additional imaging is necessary in lieu if other nuclear imaging modalities.

The role of PET imaging for monitoring disease progression has been added to the LCD. There is limited data based on mixed and small patient cohorts to determine if there is a role. The Slart 2018 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 additional studies are warranted. Therefore, due to insufficient evidence and lack of standard of care guidelines use of PET imaging for monitoring the disease course of LVV remains investigational.

“ASNC and ACC reviewers recommend adding ventricular arrythmia without another explanatory cause to the list of conditions that would indicate PET imaging for suspected sarcoidosis is appropriate. The Joint SNMMI–ASNC Expert Consensus Document on the Role of 18F-FDG PET/CT in Cardiac Sarcoid Detection and Therapy Monitoring confirms that postmortem studies have shown that sarcoidosis “may involve any part of the heart but most commonly involves the myocardium. Depending on the type and extent of involvement, CS can present as conduction abnormalities, ventricular arrhythmias, sudden cardiac death, systolic and diastolic heart failure or valvular disorders…” Supporting reference Chareonthaitawee 2017 was submitted.

The joint SNMMI-ASNC Expert Consensus guidelines have been added to the LCD. Based on the available literature and guidelines the policy has been updated with “young patient” replaced by <60 years old and to include idiopathic sustained ventricular tachycardia unexplained by other causes consistent with the guidelines. The request to add ‘History of systemic autoimmune disease and evidence of non-coronary late gadolinium enhancement on cardiac magnetic resonance imaging’ was not included in the guidelines or other evidence so this was not added.

Additional comment regarding cardiac sarcoidosis includes “There may often be a role to do both forms of imaging as they have complementary value in addition to the perfusion exam conducted around the time of FDG PET. Thus, it should be permissible to obtain both CMR and PET where clinically indicated. Reference Vita 2018 provided. 

The language prohibiting PET imaging with other cardiac imaging this was reworded to clarify the intentions that if 18F-PET/CT is the primary imaging modalities for surveillance it is not to be combined with other cardiac imaging modalities for this same purpose. This does not prohibit the evaluation for other conditions if the medical necessity for the additional studies is clearly documented in the medical record.

“ASNC and ACC clinical reviewers recommend deletion of the requirement that the “patient does not have any conditions that would limit the ability to interpret the PET scan including cardiac/vascular surgery within the past three months.” The data to support this condition is limited and it is possible to make the diagnosis for the diseases deemed appropriate for PET imaging in this policy within three months after cardiac/vascular surgery.” No reference submitted.

No supporting evidence was submitted for this request. The data on the impact of surgery, glucocorticoid, and other factors that can potentially affect the interpretation of the study remain to be determined. In the absence of evidence to contrary the limitation will remain, however the ‘within the past 3 months’ timeframe has been removed as we agree the data to support that is limited.

The ASNC and ACC clinical reviewers recommend the deletion of patient is unable to undergo an MRI because of their device (the first point under infection of cardiovascular implantable devices) “given that it is not related to the diagnosis of device infection with FDG PET. In addition, the reviewers recommend adding “or support prolonged antibiotic therapy or not” as a frequently occurring example of where additional diagnostic studies would impact clinical care.” No reference submitted

These suggestions have been considered and changes are reflected in the LCD. 

Clinical reviewers suggest deletion of no conditions that would make PET scan study less diagnosis (such as surgical adhesives/recent surgery (subsection v.) The reviewers argue that conditions such as surgical adhesives or recent surgery are not sufficient to mean that the imaging should not be done. Rather, the reviewers think that physicians should exercise caution with interpretation of the imaging.” No reference submitted.

The limitation remains but the example of surgical adhesion or recent surgery has been removed. Any condition which could impact the ability to accurately interpret the PET scan still is considered a limitation as it is necessary to obtain reliable and accurate interpretation to justify the use of the image modality.

Comments were received from the Society of Nuclear Medicine and Molecular Imaging (SNMMI). They express concern that the policy is too restrictive. The state “SNMMI strongly believes that PET scans will not be overused in cases of suspected infection/inflammation but rather will be applied as a problem-solving tool when other diagnostic methods have come up negative or equivocal, often in critically ill patients. They submitted Best Practices for Imaging Cardiac Device-Related Infections and Endocarditis; A JACC Cardiovascular Imaging Expert Panel Statement, which is attached to this comment letter and specifically to suspected prosthetic heart valve infection, suspected cardiac implantable electronic device infection, and suspected left ventricular assist device infection. 

Thank you for your comments. Please see comment #1.

While SNMMI takes the position that PET scans will not be overused in cases of suspected inflammation/infection this assumption has already been proven inaccurate by internal data analytics and record reviews. 

“SNMMI is a strong supporter of appropriate use criteria (AUC) and has developed AUC for many imaging modalities. CGS should ensure that any LCD is aligned with the available AUC…At a minimum, CGS must cover all indications that the SNMMI AUC rated with a score of 7 or higher. We have attached the SNMMI AUC for the use of nuclear medicine in musculoskeletal infection imaging and in FUO for your reference.”

Thank you for providing this document which has been reviewed and incorporated into the LCD. AUC from other societies have also been reviewed or updated throughout the document.

“Under the proposed LCD, CGS would cover PET scans for certain patients with fever of unknown origin (FUO) and cardiac sources of infection and inflammation. Other uses of PET for inflammation and infection would be non-covered. SNMMI strongly recommends that substantial revisions be made to the proposed LCD to align it with the clinical evidence. SNMMI understands that for a number of inflammation/infection indications, use of PET and PET/CT may not be first-line diagnostic tools. In those cases, CGS should cover PET and PET/CT as second-line tools when the primary tools (e.g., echocardiography for suspected native cardiac valve infection) are inconclusive. As a general matter, that is a much better approach to coverage than not covering PET and PET/CT at all.”

For a service to be covered by Medicare it must be reasonable and medically necessary. Per §1862(a) (1) (A) of the Act this requires a service is not experimental or investigational; and meets accepted standards of medical practice for the diagnosis or treatment. CGS has provided an evidence review of the conditions where evidence was available to help determine the role of PET imaging in these conditions. Conditions in which evidence is not available or insufficient to meet the above criteria do not qualify for coverage. If evidence is developed that supports expansion in the future from the current policy, it can be considered via the LCD reconsideration process.

SNMMI request removal of the non-coverage position for FUO in immunocompromised patients. They do not feel the literature cited in the LCD support this position.

“SNMMI is not aware of any literature supporting non-coverage of PET for FUO in this population and there are data that clearly support its use.”

Submitted references include: Castaigne 2009, Douglas 2022, Haidar 2022, Jasper 2010, Yang 2015, Wang 2018, Chang 2016.

The submitted literature has been reviewed and added to the LCD. The Haidar 2022 review paper which was already in the LCD does not address immunocompromised patients and therefore was not included this section. In regard to FUO evaluation in children (Jasper 2010, Yang 2015, Wang 2018, Chang 2016 ) the FUO coverage criteria was modified to include ≥18 years as there is a separate body of literature to address this age group and is beyond the scope of this LCD. The definition of FUO used in most studies excluded immunocompromised patients therefore there is a paucity of literature to understand the role in this population. The literature consists of 2 retrospective reviews of 10 patients and 20 patients resulting in very low-quality evidence. We do not have sufficient evidence to understand a role in the immunocompromised population therefore no further changes will be made to the LCD and if new literature is available can be considered via the reconsideration process. In highly unique situations such as those in these retrospective reports further consideration can be made on appeal with medical necessity documented in the medical record. 

“SNMMI agrees with the comments submitted by the American Society of Nuclear Cardiology (ASNC). Specifically, SNMMI agrees with the ASNC comments for cardiac conditions.” The advocate for coverage for native cardiac valves where native valve endocarditis is suspected but not proven by echocardiography and blood cultures. They state “While PET and PET/CT may not be quite as sensitive and specific in infective endocarditis in native valves as it is in artificial valves, that is not a sufficient justification for non-coverage.” They cite a 2020 report from American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines recognizes the value of PET/CT in this population.

See Comments #1-9.

The 2020 report (which was included in the LCD) in Section 12.2 states “In patients classified by Modified Duke Criteria as having “possible IE” 18F-PET/CT is reasonable adjunct diagnostic imaging” with 3 citations included. This recommendation does not distinguish between prosthetic and native valves. Of the citations used to support this recommendation one study was specific to prosthetic valves. The Mahmood (2019) study which was also included in the LCD does include native valves however they did not separate out the valve types beyond mentioning that 5/13 studies included patients with suspected NVE and these represented a small population of the total number evaluated. The third citation De Camargo (2020) specifically addresses suspected NVE and in this retrospective report the authors concluded that the diagnostic accuracy in the setting of native valves was less accurate. They conclude it could be a complementary tool, but do not offer criteria of when it would play a role. We agree that there is evidence to support the role of 18F-PET/CT as an adjunctive test in inconclusive cases of suspected endocarditis in prosthetic valve patients, however there is not sufficient evidence to support this role in suspected NVE where the sensitivity and specificity is significantly lower and there is a lack of evidence to understand if/what role it may play in this population. If additional studies become available that provide this evidence, it can be submitted through the LCD reconsideration process. 

SNMMI opposes non-coverage of PET and PET/CT for the diagnosis of infection in artificial joints. They state the articles reviewed in the LCD, professional society guidelines and SMEs supports the clinical utility of PET and PET/CT for these indications.

The SNMMI submitted Appropriate Use Criteria for use of Nuclear Medicine in Musculoskeletal Infection Imaging in which the diagnosis of prosthetic joint injection of the hip with 18F-PET/CT was given AUC of 7 (Appropriate). This recommendation was supported by three systematic reviews and was given level evidence rating of fair. Jin (2014) was included in the LCD and the other 2 and SNMMI AUC have been added to the LCD. While these three reports had sensitivities 80% or higher the report does not include other systematic review and meta-analysis where lower sensitivities were reported. Within these reports the authors acknowledge pertinent limitations including the potential high risk for false positives, the high heterogeneity of the included literature making the conclusions less certain and the call for the need for additional studies to clarify the role of PET imaging for prosthetic infections.

The 2023 updated ARC Guidelines conclude that PET imaging for symptomatic hip arthroplasty, infection not excluded was “usually not appropriate”. Additionally, several other guidelines and consensus statements do not conclude a recommendation for PET imaging in this setting. The timing after surgery in which the study can be considered valid is not clearly established.

This illustrates the current challenge as there is variations among the societal recommendations and inconsistencies within the literature. There is not a clear standard for the role of PET imaging for artificial joint infections established which is required to meet the definition of reasonable and medically necessary therefore the non-coverage position remains. If additional literature clarifies the role in this area, it can be submitted through the reconsideration process.

SNMMI advocates for coverage for chronic osteomyelitis. They cite several reviewed papers that conclude high sensitivity and specificity of PET and PET/CT for chronic osteomyelitis and provide SNMMI AUC recommendations for osteomyelitis. The cite Wang, European Consensus Statement, Kulkarni publication and Termaat review for support.

The SNMMI AUC recommendations provide a level 9 recommendation for the role of PET/CT for osteomyelitis. The ACR AUC is a level 2 recommendation, and they state that the technology appears very promising, but data is limited. The supporting studies were included in the LCD. It is clear there is a high sensitivity for the detection of osteomyelitis with PET scan. What is unclear is how PET images compares to existing diagnostics (esp. gold standard MRI) which also have high sensitivity without the additional radiation exposure. Additional questions such as addressing the potentially high false positive rate, timing with post operative and post-traumatic time periods and the role of blood sugars has not been established. Additional investigation is necessary to fully establish the role of PET/CT imaging for osteomyelitis. However due to the consistent reports of high sensitivity in the rare circumstances when the gold standard of MRI is not possible or inconclusive and nuclear medicine studies are clinically indicated it is reasonable to replace the alternative nuclear medicine studies with 18F-PET/CT which has been demonstrated to be equivalent or superior to these modalities.

SNMMI supports coverage for spondylodiscitis and provide SNMMI AUC for this condition. Supporting paper is Palestro 2021. They state “As CGS recognizes in its analysis of the evidence, the results of PET and PET/CT can be affected by taking certain medicines or the presence of certain conditions. However, those are not reasons to deny all coverage nor is the fact that the clinical trials do not show superiority of PET to MRI.”

SNMMI AUC for spondylodiscitis is 9. This is based on a systematic review with fair evidence and a meta-analysis. AUC the recommendation for diagnosis of spondylodiscitis for patients with spinal hardware is an 8 based on expert opinion of the work group. They state is the nuclear medicine of choice for the diagnosis of spondylodiscitis with and without hardware. ACR rates FDG-PET/CT as “May be Appropriate’ in suspected spine infections with abnormal radiographs or CT findings as next imaging study. Two meta-analyses were identified and added to the LCD. Additional studies are necessary to delineate role of this modality as well as the potential interference of antibiotics and other factors, however due to the high sensitivity it will be included as an alternative nuclear medicine study in cases of suspected spine infections with abnormal radiographs or CT findings when MRI is non diagnostic or unable to be performed.