Nonsize Criteria for Surgical Intervention on the Ascending Thoracic Aorta

For decades, aortic surgery has relied on size criteria for intervention on the ascending aorta. While diameter has served well, diameter alone falls short of an ideal criterion. Herein, we examine the potential application of other, nondiameter criteria in aortic decision-making. These findings are summarized in this review. We have conducted multiple investigations of specific alternate nonsize criteria by leveraging our extensive database, which includes complete, verified anatomic, clinical, and mortality data on 2,501 patients with thoracic aortic aneurysm (TAA) and dissections (198 Type A, 201 Type B, and 2102 TAAs). We examined 14 potential intervention criteria. Each substudy had its own specific methodology, reported individually in the literature. The overall findings of these studies are presented here, with a special emphasis on how the findings can be incorporated into enhanced aortic decision-making—above and beyond sheer diameter. The following nondiameter criteria have been found useful in decision-making regarding surgical intervention. (1) Pain: In the absence of other specific cause, substernal chest pain mandates surgery. Well-developed afferent neural pathways carry warning signals to the brain. (2) Aortic length/tortuosity: Length is emerging as a mildly better predictor of impending events than diameter. (3) Genes: Specific genetic aberrations provide a powerful predictor of aortic behavior; malignant genetic variants obligate earlier surgery. (4) Family history: Aortic events closely follow those in relatives with a threefold increase in likelihood of aortic dissection for other family members once an index family dissection has occurred. (5) Bicuspid aortic valve: Previously thought to increase aortic risk (as a “Marfan light” situation), current data show that bicuspid valve is not a predictor of higher risk. (6) Diabetes actually protects against aortic events, via mural thickening and fibrosis. (7) Biomarkers: A specialized “RNA signature test” identifies aneurysm-bearing patients in the general population and promises to predict impending dissection. (8) Aortic stress: Blood pressure (BP) elevation from anxiety/exertion precipitates dissection, especially with high-intensity weightlifting. (9) Root dilatation imposes higher dissection risk than supracoronary ascending aneurysm. (10) Inflammation on positron emission tomography (PET) imaging implies high rupture risk and merits surgical intervention. (11) A KIF6 p.Trp719Arg variant elevates aortic dissection risk nearly two-fold. (12) Female sex confers some increased risk, which can be largely accommodated by using body-size-based nomograms (especially height nomograms). (13) Fluoroquinolones predispose to catastrophic dissection events and should be avoided rigorously in aneurysm patients. (14) Advancing age makes the aorta more vulnerable, increasing likelihood of dissection. In conclusion, nondiameter criteria can beneficially be brought to bear on the decision to observe or operate on specific TAA.


Introduction
Although having previously touched on some concepts that may fall under the rubric of "nondiameter" criteria for intervention on the ascending thoracic aorta, 1 our team at the Aortic Institute at Yale had not previously organized intervention criteria under the concept of "nondiameter" factors. An invitation from Drs. Mark Field and Manoj Kuduvalli to deliver an invited lecture on this topic at the Liverpool Aortic Symposium 2022 stimulated our intensive look at nondimensional predictors of aortic behavior, which form the substance of the present study.
Nonsize factors that weigh in the decision to operate on the thoracic aorta include the following: • Pain.
• Root location of dilatation.
We will discuss each of these factors individually in the following sections.

Pain
As can be seen in ►Fig. 1, the aorta is invested by a network of afferent pain fibers, which transmit noxious sensations via to catastrophic dissection events and should be avoided rigorously in aneurysm patients. (14) Advancing age makes the aorta more vulnerable, increasing likelihood of dissection. In conclusion, nondiameter criteria can beneficially be brought to bear on the decision to observe or operate on specific TAA. AORTA  the sympathetic chain to the brain. Not all precordial chest pain originates in the myocardium. As emphasized by Boudoulas and Stafanadis, the aorta should always be kept in mind as a potential source of chest pain. 2 Pain is the only language in which the aorta can speak to us. Most importantly, if a patient with a known ascending aortic aneurysm develops chest pain, serious consideration should be given to resection of that aneurysm. Simply put, a painful aneurysm needs to be taken very seriously and likely merits resection, regardless of size. [3][4][5] ►Fig. 2 demonstrates the endoluminal aortic findings in patients who underwent surgery for pain rather than for aortic size. One can see dramatic disruptions of the aortic intima, with cavitary lesions penetrating into the media of the aorta. ►Fig. 2A-C shows aortic dissections that stopped their progress spontaneously but continued to cause ongoing pain. ►Fig. 2C was kindly contributed by Dr. Duke Cameron, to emphasize the importance with which pain is considered in his practice as well. ►Fig. 2D shows a virulent destructive process, with innumerable penetrations into the media, none of which had been recognized on preoperative imaging in this patient operated for pain.

Length/Tortuosity
In terms of aortic dimensions, attention has been nearly universally focused on diameter, largely ignoring other aortic size characteristics. We have referred to aortic length as "the neglected dimension." Recently, attention has been turned to aortic length and its correlate, aortic tortuosity. As the aorta lengthens, it needs to become tortuous (or "curved") to remain within its body confines. Although one could consider length a "size" measurement, we are including length in this manuscript because it extends analysis beyond diameter and, by recent evidence, appears to be of cardinal importance.
We and others have chosen to measure (or "define") the ascending aortic length as the distance from the aortic annulus to the base of the innominate artery (see ►Fig. 3). Both landmarks, the annulus and the base of the innominate artery, are clearly defined structures. While diameter is a good predictor of aortic adverse events, we have recently found that length is even a notch better prognostically. 6 As shown in ►Fig. 4, aortic lengthening to >13 cm signifies an extremely high risk of aortic dissection. Others have noted similar findings. [7][8][9] In fact, certain metrics show that ascending aortic length is an even better predictor than diameter. This is shown in ►Fig. 5, where we connote somewhat greater discrimination between patients free from or patients incurring adverse aortic events (AAEs). We found, favorably, that length (unlike diameter) does not change substantially at the moment of dissection, rendering length preferable in this regard. 6 We do not wish to denigrate diameter in any way; however, ascending aortic length is even a bit better predictor. We all need to incorporate ascending aortic length more fully in our day-to-day clinical assessments. We publish a nomogram using aortic length to predict expected risk of aortic events (►Fig. 6).

Genes
Once the familial nature of thoracic aortic aneurysm and dissection (TAAD) was documented in the late 1990s, 10,11 scientists have been on the hunt to identify the specific genes that cause this disease. Currently more than 60 genes have been implicated as being potentially causative of TAAD, 12 although only 24 of these have passed the ClinGen curation (with only 11 genes in the definitive or strong evidence category). 13 These genes can cause either syndromic or nonsyndromic TAAD, or both. 14 These genes encode regulatory molecules for the extracellular matrix (FBN1, FBN2, COL1A1, COL1A2, and COL3A1), the cytoskeleton in smooth muscle cells (ACTA2, MYH11, and MYLK), and the TGF-β signaling pathway (TGFβ2, TGFBR1, TGFBR2, SMAD3, and SLC2A10). 14 It is now well recognized that certain genes can alter the natural course of aortic disease by making patients more vulnerable to AAEs (such as rupture and dissection) at small aortic sizes, 15,16 well below the traditional intervention criteria, 17,18 and at a significantly younger age. This was first recognized for patients with mutations in the FBN1 gene (Marfan's syndrome), for whom the intervention criteria were lowered in the the American and European Aortic Disease guidelines. 18,19 Additional genes, for syndromic or nonsyndromic thoracic aortic aneurysm (TAA), also predispose to early clinical events. For example, mutations in the MYLK gene predispose patients exclusively to aortic dissection, 20 while mutations in in the ACTA2 gene cause aortic dissection at small aortic sizes. 21,22 So it is vital for the clinician to take these genetic variants into account in determining criteria for surgical intervention Note that lengthening of the ascending aorta to 13 cm is critical, incurring a very high risk of adverse aortic events (AAEs). Note also that there are two earlier "hinge points" at which risk rises incrementally, at 11.5 cm and at 12.5 cm. (Reproduced with permission from Wu et al. 6 ) Fig. 5 Herein are displayed the "density" curves of aortic size (i.e., the proportion of patients at different size ranges). (A) It indicates curves for diameter as a criterion, and (B) indicates length. Note that the aortic length better separates the curves of patients with versus without adverse aortic events (AAEs) than does diameter. (Reproduced with permission from Wu et al. 6 ) on the ascending aorta in specific patients in whom this detailed genetic information is available.
To aid clinicians with decision-making in the setting of a pathogenic variant in a particular gene, we publish and regularly update recommended intervention thresholds for each of the currently known TAAD genes (see ►Fig. 7). 23,24 Please note that many genes currently listed at the "Standard (5.0-5.5 cm)" may need to move "earlier" as more clinical data accumulate. Further guidance for surgical intervention is provided by Mariscalco et al, 25 who plot patient age versus diameter, both at the time of an aortic dissection event; they do this for several common genes that predispose to TAAD (see ►Fig. 8). This analysis can help the clinician considerably by incorporating the very important age parameter in decision-making.
In summary, knowing which gene is the culprit for TAAD in a particular patient is a critical criterion in the decision to operate.

Family History
Another important and often overlooked factor that should influence surgical intervention on the ascending aorta is family history. Once a patient has been identified as having thoracic aortic disease (aneurysm or dissection), it is critical to screen family members, who may be asymptomatic but still harbor an aneurysm. We screen very widely: parents, children, siblings, aunts, uncles, nieces and nephews, and grandchildren. An echocardiogram usually suffices as a screening tool, although it is usually unreliable in visualizing the aorta beyond the mid-ascending portion.
At our Aortic Institute at Yale, working together with Dr. Michael Coady, we determined that familial aneurysms are more malignant than sporadic aneurysms, presenting at a younger age (58 vs. 66 years), and growing faster (2.1 vs. 1.6 mm/y). 26 Thus, earlier surgical intervention may be necessary to prevent aortic dissection and rupture. However, if a patient has a family history of an aortic dissection event, this is an even more powerful predictor of malignant outcome, increasing the odds or rupture of dissection in family members by a massive three-to ninefold. [27][28][29] Furthermore, familial aortic dissection events tend to cluster in regard to the age of dissection onset: at least 50% of patients will develop an aortic dissection within 5 years (younger or older) of the age at which the initial dissection event in the family occurred. 30 So family history is vitally important as a criterion for surgical intervention. It has been our policy to offer surgical intervention to any patient with a "real" aneurysm (say 4.2 cm and above) who also has a family history of aortic dissection. Aortic height index (AHI) nomogram as a function of aortic size (diameter þ length; x-axis) and patient height (y-axis), with the AHI given within cells. The table has a four-tier, color-coded warning system, with red representing the most severe, followed by orange, yellow, and green. AAEs, aortic adverse events. (Reproduced with permission from Wu et al. 6

Bicuspid Aortic Valve
For decades, it was thought that aneurysms associated with bicuspid aortic valve were highly dangerous, almost as much as the aortas of patients with Marfan's disease. In fact, the term "Marfan's light" became popular to describe the increased susceptibility of the aorta in patients with bicuspid aortic valve. A recent analysis from our center has found that this is actually not the case. In fact, the aneurysmal ascending aorta in patients with bicuspid aortic valve is no more vulnerable than that of patients with a normal, trileaflet aortic valve (see ►Fig. 9). 31 The American Association for Thoracic Surgery consensus "Guidelines on bicuspid aortic valve-related aortopathy" agree with our findings, stating specifically that "An increasing amount of literature has recently shown that BAV (bicuspid aortic valve)-related aortopathy is less dangerous than previously described." 32 Therefore, ascending aortic aneurysm patients with bicuspid aortic valve do not need to be operated sooner or at smaller dimensions than other patients.

Diabetes
Diabetes has a long-standing, well-earned reputation as a profoundly negative health factor. That is why it is so frankly astounding that diabetes seems, along nearly every avenue of analysis, to play a protective role in patients with aneurysm disease. Our team recently reviewed the literature on this topic. 33 First, diabetes increases the thickness of the aortic wall. Surgeons know this from experience, but this has been shown scientifically as well 34 (see ►Fig. 10A). Specifically, wall thickness in the figure is seen to be substantially higher in diabetics than in nondiabetics. Because wall thickness appears in the denominator of Laplace's law ( where T is wall tension, P is intraluminal pressure, r is radius, and t is thickness, wall tension is decreased, as seen in ►Fig. 10B. Decreased wall tension is very beneficial in an aneurysmal wall. So, there are engineering benefits in the aortic wall. Do these translate to observable clinical benefit? We see in ►Fig. 11A that expansion rate was less in the diabetic aortas (circles). 35 In ►Fig. 11B, we see that far more nondiabetics than diabetics achieved an aortic growth rate of !5 cm during observation. 36 Studies have shown, and clinical experience in the operating room confirms, that the diabetic aorta is grossly thickened, dense, and fibrotic, and holds sutures well. The diabetic aorta is paradoxically protective against both abdominal aortic aneurysm (AAA) and TAA. Furthermore, the diabetic aorta is inimitable to aortic dissection.
Although most studies have been done in AAA and fewer in TAA, experiments have shown all the following effects of diabetes in aortic aneurysm disease: decreased incidence and prevalence, decreased aneurysm growth rate, lower matrix metalloproteinase (MMP) levels, decelerated matrix loss, decreased dissection rate and rupture rate, decreased mortality, and delayed age at rupture. 33,[37][38][39][40][41][42] This section has discussed the beneficial impact of diabetes itself. It is worth noting that the drug metformin, used so commonly in the treatment of diabetes, has in and of itself marked beneficial properties vis-à-vis aneurysm disease. 33,43 These salutary effects include protection against AAA and TAA and also decreased adverse factor levels (MMP2, MMP9, TNFa, and IL-6) in both mice and humans. Metformin also has cardioprotective and vascular protective effects-and it enhances weight loss as well.
So, paradoxically, diabetes, so detrimental in arteriosclerosis, has been found markedly beneficial from a purely aneurysm standpoint; the diabetic aorta, dense, thickened, and fibrous, has been found inimitable to aortic dissection. Also, the common medical treatment for diabetes, metformin, has been found to be additively beneficial as well.

Biomarkers
One of the most serious problems in aortic disease has to do with the identification of asymptomatic carriers of aneurysm in the general population. Once an aneurysm-bearing patient is identified, medical and surgical science can keep that patient safe. We know reasonably well when to operate (although this article aims to refine those abilities), and we know how to operate. So, we need to improve identification of asymptomatic carriers in the general population.
We have performed a biomarker study, in which we assessed the circulating levels of 33,000 ribonucleic acids (RNAs). It should be remembered that deoxyribonucleic acid (DNA) is the blueprint for how the body will be made, but RNAs are the worker moieties that carry out the actual process of building the body. RNAs can be upregulated or downregulated depending on the needed activity in different organs and tissues.
We were able to identify, from among those 33,000 RNAs, a panel of expression levels of 41 genes that was quite effective at discriminating which patients harbored an aneurysm and which did not, with an overall accuracy approaching 80% (►Fig. 12). 44 We are just now replicating this work in a different cohort of patients. This promises to enable detection of asymptomatic patients harboring TAAs.
Another clinical issue has to do with anticipation of impending aortic dissection. Although many "biomarkers" for aortic dissection have been developed, these detect dissection after it has occurred. So those tests are not very helpful, as the dissection has already occurred, and imaging studies have generally confirmed aortic dissection by the time those biomarker results become available. D-dimer (essentially always positive when there is an aortic dissection, reflecting clot in the false lumen of the dissected aorta) is of some utility in the emergency department in questionable cases prior to echocardiography or computed tomography (CT) scanning. However, what is really needed is a biomarker that can indicate that an aortic dissection is looming, about to occur.
Within the 33,000 genes tested in the study described earlier, we were able to identify a subpanel of a dozen up-or downregulated genes that were able to distinguish patients at the moment they suffered aortic dissection from those free of dissection (unpublished). We speculate that the abnormal RNA profile, reflecting disordered aortic biology, was likely abnormal for some time before dissection occurred-making this RNA profile a potential monitoring test. While still in the early stages, this molecularstyle approach holds promise for enhanced prediction of dissection based on specific activated or deactivated biological pathways. Such predictive markers of impending dissection could represent an important step forward in aortic care.

Aortic Stress
The level of mechanical stress imposed on the aorta has an important impact on the occurrence of aortic dissection. We have published the paradigm shown in ►Fig. 13 to represent the clinical path that eventuates in aortic dissection. This analysis is based on multiple studies we have performed and published over decades. [45][46][47][48] In brief, we feel that many, if not most, pathways toward dissection originate with a genetic mutation that predisposes to aortic disease. We currently find suspicious variants in about one-third of patients genetically sequenced. 49 This number increases as whole exome sequencing proliferates. We believe that the genetic abnormality, over decades, leads to destruction of the aortic wall elements (e.g., lamellar loss), thus weakening the wall and allowing its gradual enlargement over time-ultimately to aneurysmal dimensions. Then, we believe, an acute, severe hypertensive episode raises blood pressure and aortic wall tension beyond the tensile strength of the chronically weakened aortic wall-resulting in an acute aortic dissection event. Two-thirds of our patients recall a specific severe Fig. 12 Results of screening of ribonucleic acid (RNA) expression levels of 33,000 RNAs in patients with and without thoracic aortic aneurysm. Red indicates overexpression of our panel of select genes, and blue indicates under-expression. If all the reds were together and all the blues were together, the test would have been 100% accurate. The test was encouragingly accurate at predicting correctly which patients harbored a thoracic aortic aneurysm (TAA) and which did not. Sensitivity was 83% and specificity was 68%, and the overall accuracy was 78%. physical exertion or an especially troubling emotional event just before onset of their dissection pain.
So, aneurysm patients subject to severe physical exertion or severe emotional events should be considered dissection prone and triaged earlier (in time and in aortic size) to surgical therapy. This is an extremely important factor to take into account. (Of course, it is helpful to mitigate these physical and emotional stressors via advice and restrictions, but compliance is often marginal). We do recommend a Bblocker and an afterload reducing drug for most patients, to "take the edge off" hypertensive episodes.
Weightlifters are especially vulnerable, as blood pressure reaches astronomical levels (not seen in any other human setting or endeavor) during the effort cycle of the "lift" (see ►Fig. 14). We know of no other environment, in the outside world or within the hospital, where absolute magnitude blood pressures over 300 mm Hg are achieved. ►Table 1 presents dozens of cases made known to our team of exercise-induced aortic dissection-a very real and at least partially preventable cause for the loss of promising young (and older) athletes.

Root Location of Dilatation
In the past, we in cardiac surgery did not adequately differentiate risk levels for given diameters depending on whether they were located in the ascending aorta proper or on the aortic root. One of our team members has expertly sorted out these risks for the two zones: the ascending aorta above the sinotubular junction versus the aortic root itself.
Dr. Kalogerakos and colleagues performed a precise risk analysis for these two zones. Root dilatation was more malignant, producing more adverse events at a given level of enlargement. Earlier intervention for root dilatation than pure ascending dilatation was recommended 50 (see ►Fig. 15). Very recent data from the International Registry of Acute Aortic Dissection (IRAD) is supportive of this concept. 51

Inflammation (Positron Emission Tomography Imaging)
A career's worth of data have been accumulated by Sakalihasan et al 52 confirming that the inflammation that is known  Table 1 Synopses of cases sent to the senior author manifesting exercise-induced aortic dissection, often lethal to underlie aneurysmal degeneration of the aortic wall (driving the degeneration, in fact) can be visualized as "hot" uptake on PET of the chest or abdomen. Hot spots are seen more commonly in the abdominal aorta than in the ascending or descending thoracic aorta. Sakalihasan  estimate the likelihood of dissection events with greater precision than before.

Sex
Female sex raises risk of ascending aortic events mildly, but this is well accommodated by body size indexing, as in the height-based nomogram depicted in ►Fig. 6.

Fluoroquinolone Rx
Over recent years, through the precise work of LeMaire et al 57 and others, it has become abundantly clear that fluoroquinolone treatment disrupts the aorta and leads to aortic dissection in unfortunate patients. 58 This adverse relationship has been shown both in humans and in animal models. The Food and Drug Administration (FDA) has added a warning to the drug label. The dissection event may occur early or late after treatment and after small or large dosages of the drug. Eliciting a fluoroquinolone history has become an important part of aneurysm patient encounters. Fluoroquinolones should be fully avoided in patients with aortic disease.

Age
Finally, as can be seen in ►Fig. 17, the likelihood of aortic dissection occurring at a specific aortic diameter increases with increasing age, reflecting the "wear and tear" aging of the elements of the aortic wall. Serial studies of aortas of various ages have shown age-related deterioration-with thickening of the endothelium, thinning of the media, and loss of elastin. [59][60][61] Such changes make the aortic wall more susceptible to aortic dissection. This aging factor needs to be incorporated in our assessment of risks.

Conclusion
Diameter has served our profession well as a criterion for surgical intervention for ascending aortic aneurysms. However, diameter alone is far from perfect as a predictor. We have outlined here 14 additional criteria that clinicians can apply to fine-tune their risk strategy for ascending aneurysm patients (►Fig. 18). Most of these factors increase risk: pain, excess aortic length, genetic abnormalities, family history, excess aortic stress, root location, and inflammation. Bicuspid aortic valve, we have shown, is neutral. Biomarker analysis is still evolving, and the presence of diabetes substantially lowers risks. ►Table 2 summarizes these findings briefly, in tabular form, for easy reference. The clinician can apply added judgment along the axes discussed in this study to fine-tune the risk assessment for ascending aortic aneurysm patients-above and beyond that attainable solely by a diameter criterion. "Judgment" is the key word, as, despite decades of scientific investigation, the clinician's judgment is the key glue that assembles all the Fig. 17 The probability of aortic dissection at various diameters increases with increasing patient age, reflecting the adverse impact of age (wear and tear) on the aortic wall.   The aorta weakens as it ages, increasing the likelihood of aortic dissection at any given size

Judgment Key
Note: The criterion is named in the left column. The "snap" summary of the criterion is indicated in the right column. In the right column, text in red indicates a negative influence, text in green indicates a salutary influence, and text in blue indicates a neutral influence. Note (last row) our recommendation that despite scientific recommendations based on criteria 1-14, we recommend that clinical judgment be applied to put all component findings into a unified gestalt.