Response to Comments: Non-Invasive Fractional Flow Reserve (FFR) for Ischemic Heart Disease

Language clarification – stenosis range

AIM Specialty Health® (AIM)

• “Stable nonobstructive coronary artery disease with persistent symptoms requiring further test, and 40-90% stenosis on CCTA.” Nonobstructive CAD is defined as <50% stenosis (2021 AHA/ACC/ASE/ CHEST/SAEM/SCCT/SCMR Guideline for the Evaluation and Diagnosis of Chest Pain). Therefore, the above quoted text would apply ONLY to patients with stenosis in the 40-50% range. We doubt this was the intent of the authors.

Campbell Rogers, MD (HeartFlow)

• Revise Restriction 11 from “<50%” to “<40%” for consistency within the draft LCD

The 2021 AHA/ACC/ASE/ CHEST/SAEM/SCCT/SCMR Guideline for the Evaluation and Diagnosis of Chest Pain¹ state “the term “obstructive”, consistent with convention, will be used to indicate CAD with ≥50% stenosis and nonobstructive CAD will be used to indicate CAD <50% stenosis.” (Section 3 Cardiac Testing General Considerations and Figure 13). The language was intentional. FFRCT plays a role as a sequential test. If a patient has ≥50% stenosis the options within the algorithm do no included FFRCT. Therefore, in patients with non-obstructive CAD (<50% stenosis) who have persistent symptoms despite maximized medical management the pathway leads to option for further evaluation with CCTA or stress test. If CCTA is performed and ≥40-90% stenosis is found on this new study than FFRCT may be used to aid in further evaluation. The policy will remain consistent with the 2021 Guidelines.

The restriction from <50% to <40% was revised in draft for consistency.

Language clarification – persistent symptoms definition

AIM Specialty Health® (AIM)

• The term “persistent symptoms” also requires further definition. Does this mean that symptoms persist despite optimal medical therapy, or does it mean that the symptoms have been present for some period of time (whether or not optimal medical therapy was instituted)? If the latter, how much time? For example, if a patient with intermittent chest pain for a month undergoes CCTA and is found to have obstructive CAD (i.e., 70% RCA stenosis), should FFR be performed, or should the patient be treated medically per the ISCHEMIA trial protocol?

Persistent is defined by symptoms that persist despite optimal medical therapy. Optimal medical therapy is defined by ISCHEMIA trial protocol² as unacceptable ischemia-related symptoms despite maximally tolerated medical therapy. This includes at least two anti-anginal drugs from different drug classes added to beta-blocker therapy and titrated to maximally tolerated doses before medical therapy is considered to have failed. This has been added to the policy for clarification.

Stenosis location - proximal or middle segment of coronary artery

AIM Specialty Health® (AIM)

• In the indication that addresses FFR for stable CAD, there is no requirement that the stenosis be in the proximal or middle segment of the coronary artery, as is the case for acute CAD indications. Is this intentional or is it an oversight? We suspect it is an oversight because, in general, distal CAD is not managed with revascularization in which case FFR is not particularly helpful in clinical decision making.

The 2021 AHA/ACC/ASE/ CHEST/SAEM/SCCT/SCMR Guideline for the Evaluation and Diagnosis of Chest Pain¹ Section 5.1.3 Intermediate-High Risk Patients with Stable Chest Pain and No Known CAD states “For intermediate-high risk patients with stable chest pain and known coronary stenosis of 40% to 90% in a proximal or middle coronary segment on CCTA, FFR-CT can be useful for diagnosis of vessel-specific ischemia and to guide decision-making regarding the use of coronary revascularization”. This criterion is to identify patients who may benefit from further evaluation with FFRCT and does not restrict the vessels that can be evaluated with FFRCT or the clinical decision making from those results. The policy will remain consistent with the 2021 Guidelines.

Stress Testing

AIM Specialty Health® (AIM)

1. Patients who have undergone stress testing are excluded from consideration for FFRCT. We suggest that the authors consider the following:
   1. If stress testing is inconclusive, CCTA and FFR may have a role in patient management.
   2. If stress testing is performed without imaging, and is positive, and CCTA shows more than one obstructive lesion, FFR may be useful in determination of the flow limiting significance of each lesion thereby guiding PCI decisions.

1. If stress testing is inconclusive, CCTA and FFR may have a role in patient management. We agree that FFR-CT may be a reasonable alternative when stress test is inconclusive, and policy updated accordingly.
2. If stress testing is performed without imaging, and is positive, and CCTA shows more than one obstructive lesion, FFR may be useful in determination of the flow limiting significance of each lesion thereby guiding PCI decisions. This was not accompanied by supporting literature if such literature is published, please submit with reconsideration.

Stenosis range 40-90% support

Sadeer Al-Kindi, imaging cardiologist at University Hospitals in Cleveland, OH

• Expressed support of the expansion of the stenosis ranges to include 40-90% by CTA. The expanded stenosis ranges will extend patient access and improving patient outcomes using

Steven Raible, MD (Director, Noninvasive Cardiac Imaging, Norton Healthcare)

• Expressed support of the expansion of the stenosis ranges to include 40-90% by CTA. The expanded stenosis ranges will extend patient access and improving patient outcomes using FFRct.

Thank you for your support of these changes.

CTA image quality restriction

Sadeer Al-Kindi, imaging cardiologist at University Hospitals in Cleveland, OH

• Restrictions related to CTA image quality - prosthetic valves and prior pacemaker and defibrillator leads. These restrictions are unnecessary as HeartFlow has very strict image quality guidelines already in place. If there is artifact in the CTA image that does not allow HeartFlow to clearly see the lumen boundaries, the case is returned, and no analysis is completed.

Campbell Rogers, MD (HeartFlow)

• Remove restrictions related to image quality, valves (Restriction 1), known severe aortic stenosis (Restriction 2) and pacemaker or defibrillator leads (Restriction 8)

There are several reasons we will not remove these limitations at this time:

1. This policy is foundational so covers any technology that would provide equivalent service. If a new product that is equivalent to what is currently available emerges it will be held to the requirements within this policy. While current technology returns the images if quality is low, we cannot assume that will be the case for all future technologies.
2. To ensure safety and accuracy of results literature is needed to support the removal of these limitations. Pivotal trails consistently excluded patients with prior percutaneous coronary interventions.
3. Incomplete reporting of coronary artery anatomy defeats the purpose of this diagnostic test.

If literature or guidelines provide support for the safety and accuracy for CCTA and FFRCT in this population that can be evaluated with reconsideration.

The restriction for aortic stenosis is addressed in Comment #7.

 Extensive plaque & CCTA quality

Sadeer Al-Kindi, imaging cardiologist at University Hospitals in Cleveland, OH

• High plaque volume does not necessarily impair CCTA quality. There are many studies that support this including Syntax 2 and 3 attached. I would recommend the removal of this statement as to not limit access to Medicare patients that may have high plaque burden.

Steven Raible, MD (Director, Noninvasive Cardiac Imaging, Norton Healthcare)

• Request removal of
  • Prior placement of prosthetic valves
  • Known sever aortic stenosis
  • Prior pacemaker or defibrillator lead placement
  • Of note, CAST-FFR study validates use of FFR-CT in patients with known severe aortic stenosis

Society of Cardiovascular Computed Tomography

• The SCCT recommends removing the comment suggesting the low likelihood of high quality CCTA in the setting of extensive plaque.
• DL38771 proposes the following:
  • If extensive plaque is present a high quality CCTA is unlikely and stress testing is preferred.
• With the development of new technologies in CCTA, presence of extensive plaque or calcification does not necessarily exclude the use of FFRCT. We recommend the use of FFRCT if there is adequate image quality if determined by HeartFlow criteria (described above).
  • Suggestion: recommend removing “If extensive plaque is present a high quality CCTA is unlikely and stress testing is preferred.”

Campbell Rogers, MD (HeartFlow)

• We suggest removal of the last statement in the support paragraph. High plaque volume does not
• necessarily impair CCTA quality. Several studies have demonstrated utility in patients with severe
• multivessel coronary artery disease, and therefore with high plaque burden. [17, 18]
  • Remove “if extensive plaque is present a high quality CCTA is unlikely and stress testing is
  • preferred”.

Previous studies have demonstrated coronary CTA images in the presence of calcified plaques risk overestimation of stenosis severity and reduced agreement between the CTA and invasive angiography³. The Syntax 3 study⁴ suggest that the overestimation of CTA-SS was corrected by FFR-CT, resulting in higher correlation to invasive angiogram than CTA alone. This was reproduced in the FAME trial⁵. Therefore, the statement regarding excessive plaque formation is removed from the policy.

The 2021 CAST-FFR study supports FFR-CT in the setting of severe aortic stenosis so this limitation is removed from the policy⁶. 

 Include stable & acute chest pain to indications

Steven Raible, MD (Director, Noninvasive Cardiac Imaging, Norton Healthcare)

• Add stable and acute chest pain to indications to be consistent with 2021 ACC/AHA guidelines.

Campbell Rogers, MD (HeartFlow)

• Include coverage for both stable and acute chest pain patients in the indications, to conform to the ACC/AHA Chest Pain Guideline.
• Revise “stable” to “stable or acute” in the first sentence

Stable chest pain has been added to align with ACC/AHA Chest Pain Guidelines¹.

Request for additional restrictions for coronary stents

Steven Raible, MD (Director, Noninvasive Cardiac Imaging, Norton Healthcare)

• Request for additional restrictions for coronary stents including:
  • A stent is present in the left main coronary artery
  • There is a stenosis 30% in the left main coronary arty and 1 or more stent in the left system
  • There are stents present in 2 or more coronary systems (e.g., the left circumflex artery and left anterior descending artery)

The request to remove intracoronary metallic stents was accompanied by a suggested limitation which the manufacture provided as appropriate exclusion in the setting of stents, but without supporting literature. The challenges in reproducible image quality are a known limitation of the technology in the presence of intracoronary stenting. Moreover, the pivotal trials consistently excluded patients with prior percutaneous coronary intervention from participation. Upon review, we find no basis for coverage with the currently available evidence. If availability of evidence to support the efficacy of FFRCT analysis in the presence of intracoronary stents, we can reconsider coverage upon request.

Prior placement of prosthetic valves and pacemaker or defibrillator leads

Society of Cardiovascular Computed Tomography

Image quality plays a role in the performance of FFRCT. HeartFlow has established criteria for acceptable image quality which can be assessed in the presence of prosthetic valve, pacemaker, or defibrillator leads. The presence of prosthetic valves and pacemakers/defibrillators does not preclude performance of FFRCT. Prosthetic valves and pacemakers/defibrillators only preclude the performance of quality CCTA to the extent that these factors may affect adequate imaging of the lumen of the coronary arteries and prevent the reading physician from interpreting the presence of SIHD. However, these artifacts can be minimized with the use of certain scanning parameters and reconstruction algorithms. HeartFlow can process FFRCT results even for patients who have a prosthetic valve present in the heart if there are no significant artifacts involving coronary arteries. It is the opinion of the SCCT that the restriction on prosthetic valves and pacemakers/defibrillators be removed entirely.

• If CCTA image data quality is unacceptable as determined by HeartFlow, no FFRCT processing takes place, no claim is submitted, and no cost is incurred.

See comment #6.

Supporting literature submitted by Xu et al 2021 was a large retrospective study to evaluate influence of image quality on the diagnostic performance of FFRCT concluding support of FFRCT as a noninvasive test for evaluating lesion specific ischemia that can be influenced by image quality. However, the literature does not address prosthetic valve or pacemaker, or defibrillator leads therefore does not provide evidence for removal of these limitations from the policy.

Comment #11: Known severe aortic stenosis

Society of Cardiovascular Computed Tomography

Diagnostic quality images are routinely obtained in patients with severe valvular heart disease, even when vasoactive medications are not administrated prior to image acquisition. Transcatheter aortic valve replacement data demonstrates that coronary anatomy and coronary artery disease severity can be accurately evaluated with CCTA in patients with severe aortic stenosis undergoing cardiac CTA for pre-procedural planning. Lunardi et al. showed that FFR guidance was associated with favorable outcome in observational study in patients undergoing transcatheter aortic valve implantation. However, the left ventricular hypertrophy induced by aortic stenosis may alter the coronary flow reserve, potentially affecting the results obtained by invasive FFR. Since left ventricular mass is considered in the FFRCT analysis, severe aortic stenosis may not have a bearing on physiologic assessment of coronary stenosis. Recently, CAST-FFR study showed the safety and feasibility of FFRCT use in patients with severe aortic stenosis. The area under the receiver-operating characteristic curve for CT-FFR was 0.83 (0.72-0.93, P<0.0001), which was higher than that of CTA and quantitative coronary angiography (P=0.01 and P<0.001, respectively).

• Request: Aortic stenosis should be removed from the exclusion criteria

See comment #7.

Intracoronary metallic stents

Society of Cardiovascular Computed Tomography

HeartFlow can process FFR_CT results on patients who have had stents to one vessel. However, they can only interpret the vessels without stent. The IFU for HeartFlow state that there are 3 scenarios where FFR_CT cannot be utilized (and therefore would be appropriate to not cover):

1. Stent is present in the left main coronary artery
2. A stenosis > 30% in the left main coronary artery and 1 or more stent in the left system
3. Stents present in 2 or more coronary systems

• Suggestion: Intracoronary metallic should be removed from the exclusion criteria

Campbell Rogers, MD (HeartFlow)

• Revise language of the restriction related to intracoronary metallic stents (Restriction 5) to match verbiage in HeartFlow’s Instructions For Use (IFU) 11

See comment #9.

Non-obstructing stenosis (<50% of all major epicardial vessels) on CTA or catherization in the past twelve months, in the absence of a new symptom complex

Society of Cardiovascular Computed Tomography

The SCCT recommends clarification in the language related to item # 11.

• Suggestion: recommend revising item #11 to <40% of all major epicardial vessels to remain consistent with the 2021 ACC/AHA/ASE/CHEST/SAEM/SCCT/SCMR guidelines.

See comment #1.

Higher grade stenoses (greater than 90%)

Society of Cardiovascular Computed Tomography

• The SCCT recommends coverage for patients with multivessel disease where >90% stenosis coexists with 40-90% stenosis.
• DL38771 proposes the following:
  • If higher grade stenoses (i.e. greater than 90%) are present, this study is not medically necessary, as the patient should proceed to catheterization. Similarly, low-grade stenoses (less than 30%) do not require additional confirmatory data. This should be performed as an alternative to stress testing.
• In the case of 3 vessel disease, non-invasive functional Syntax Score (SS) provided by FFR_CT reclassified 30% of patients with multivessel CAD from the high and intermediate tertiles to the low risk tertile, where an alternative revascularization strategy (PCI over CABG) could be offered and reducing risk to the patient⁷.
  • Suggestion: recommend clarification in the coverage guidance to include studies with <90% stenosis co-existing in conjunction with >90% stenosis to allow hemodynamic assessment of the relevant vessel(s).

Campbell Rogers, MD (HeartFlow)

• Revise “If higher grade stenoses (i.e. > 90%) are present” to “If higher grade stenoses (i.e. greater
• than 90%) are present, and no other stenoses 40% - 90% are present.”
• Revise “less than 30%” to “less than 40%” for consistency within the draft LCD

The recommendations for changes related to high grade stenosis were not submitted with supporting literature. If literature is published to support the role of FFRCT in multi-vessel disease where one more vessel has >90% stenosis than it can be submitted with LCD reconsideration.

The revision from less than 30% to less than 40% has been made to ensure consistency with 2021 Guidelines¹.

Stress testing and FFR_CT

Society of Cardiovascular Computed Tomography

• DL38771 proposes the following:
  • Not in conjunction with stress testing (unless FFRCT was not high quality and alternative study needed)
• The performance of FFRCT is based on the image quality of the CCTA. If the CCTA image quality is not adequate for FFRCT analysis as determined by HeartFlow image quality criteria, then an alternative study such as stress testing would be a reasonable option. SCCT suggests the following language:
  • Not in conjunction with stress testing (unless CCTA was not of adequate quality for FFRCT and alternative study needed)

This has been revised in the policy to clarify.

Combine the two indications for “no known” and “known” CAD

Campbell Rogers, MD (HeartFlow)

• Request: Combine the two indications for “no known” and “known” CAD into one, as once a CCTA is performed and FFRCT ordered, CAD is always “known”.

In the 2021 AHA/ACC/ASE/ CHEST/SAEM/SCCT/SCMR Guideline for the Evaluation and Diagnosis of Chest Pain¹ the algorithm for suspected ACS at Intermediate Risk with No Known CAD and Known CAD are separate. Based on this algorithm FFRCT is a sequential test, therefore the evaluation and treatment plan may lead to CCTA in defined clinical scenarios. Based on the CCTA results the patient may benefit from FFR-CT or stress testing. While the FFR-CT would only be ordered in the setting of known CAD with 40-90% stenosis this if not analogous with what is known at presentation. Therefore, the policy language will remain consistent with the 2021 Guidelines.

Language change

Campbell Rogers, MD (HeartFlow)

• Request: Change “FFRCT” into “CCTA” in last indication of the draft LCD

Revision has been made.