Keywords
Head Pain - neck pain - aortic dissection
Introduction
A 53-year-old Caucasian man was admitted to the hospital because of intense frontal
headache and neck pain. In the past, he suffered from occasional migraine with aura,
with predominant parietotemporal and occipital headache. The present episode started
36 hours before admission and was ushered in with a stabbing anterior chest pain lasting
10 minutes. A few minutes after the onset of chest pain, severe bifrontal headache
evolved within 2 minutes, followed by additional neck pain. Moreover, the patient
complained of nausea, dizziness, and flickering eyes. The headache persisted despite
analgesics. Computed tomography scan showed no subarachnoid hemorrhage or aneurysm.
Chest X-ray showed no abnormalities. However, transthoracic echocardiography showed
a widened ascending aorta (54 mm) with an intimal flap and a severely regurgitant
tricuspid aortic valve.[1]
Acute aortic dissection is a highly lethal condition, which has an estimated incidence
of 2.9 per 100,000 in western Hungary to 3.5 per 100,000 person-years in Olmsted County,
Minnesota.[2]
[3] Though this condition is rare compared to myocardial infarction that has an incidence
of 208 cases per 100,000 person-years, aortic dissection is still the most common
life-threatening disorder affecting the aorta.[4]
[5]
[6] This acute syndrome is rapidly fatal if undetected with a 20% prehospital lethality
rate and up to 30% in-hospital mortality.[7]
[8]
[9] The mortality reaches 1 to 3% per hour in the first 48 hours and without treatment,
more than 75% of patients with a Type A dissection will die in 2 weeks.[10] One of the difficulties with aortic dissection is the myriad ways in which it can
present. Work done by the International Registry of Acute Aortic Dissection (IRAD)
has shown that though many patients present with sudden severe chest pain, pulse deficits,
and pain radiation, 10 to 15% of patients do not present with pain and a subset of
other patients present with nonclassic symptoms such as neurologic deficits and syncope.[4]
[8]
[11]
These atypically presenting patients have a potential for delayed diagnosis, treatment,
and worse outcomes compared with classically presenting patients. One of these atypical
presentation symptoms is head and neck pain. Clinical teaching states that proximal
dissections are more often associated with head and neck pain and neurologic symptoms
due to arch involvement; however, there are little data in the published literature
to support this claim.[12] We sought to analyze the truth of this clinical anecdote while also looking at the
demographics, presentation, treatment, and outcomes for acute Type A dissection patients
with head and neck pain.
Materials and Methods
The patient's information regarding acute aortic dissection was collected from 43
centers in North America, Europe, Asia, and Australia on patients who presented between
January 1996 and March 2015. These patients were identified via hospital discharge
diagnosis records and the imaging databases of surgical and echocardiography laboratories.
A 290-question survey, which included questions on demographics, history, physical
findings, management, imaging, and outcomes, was then filled out. These data were
then entered into an online database and reviewed by the IRAD coordinating center
at the University of Michigan for face validity and completeness.[4]
The 3,027 acute Type A dissection patients were first separated into two cohorts:
those with head and neck pain (791 patients) and those without head and neck pain
(2,236 patients) for our main analysis. These groups were then further analyzed regarding
demographic characteristics, presentation signs and symptoms, treatment types, and
outcomes. Given our results, we performed a subgroup study in patients with head and
neck pain, separating that group into those who had chest and back pain versus those
without chest and back pain (therefore only head and neck pain). We also performed
a logistic regression of mortality in this subset controlling for head and neck pain
status, preoperative hypotension, age > 65 years, site of intimal tear, and signs
of intramural hematoma.
Categorical variables were analyzed using Pearson's chi-square analysis. Continuous
variables without skewed means were analyzed using two-sided t-tests. Skewed continuous variables were analyzed using the Whitney-Mann U asymptotic
test. Logistic regressions models were analyzed using Hosmer-Lemeshow tests. Wald
chi-square tests were used to find significance of the predictors.
Results
Our main analysis found several differences in the demographics between patients with
head and neck pain versus those without head and neck pain as seen in [Table 1]. There were significantly more females with head and neck pain compared with those
without (39.8% vs. 31.3% p < 0.001). There were significantly more white patients with head and neck pain (90.2%)
compared with those without (84%, p < 0.001). There were fewer African American patients with head and neck pain (4.9%
vs. 8.1% p = 0.004) and fewer Asian patients with head and neck pain (2.3% vs. 5.3% p = 0.001). Finally, atherosclerosis and a family history of aortic disease were found
more often in those with head and neck pain than those without (23.7% vs. 18.9% p = 0.005, and 14.5% vs. 8.8% p = 0.002, respectively).
Table 1
Comparison of baseline demographic risk factors in patients with head and neck pain
compared with those without head and neck pain
|
Demographics
|
Head and neck pain group
|
No head and neck pain group
|
p-Value
|
|
Age (overall mean)
|
61.49
|
61.45
|
0.939
|
|
< 40 y
|
8.6
|
7.4
|
0.27
|
|
> 70 y
|
30
|
30.4
|
0.831
|
|
Female sex
|
39.8
|
31.3
|
< 0.001
|
|
White
|
90.2
|
84
|
< 0.001
|
|
African American
|
4.9
|
8.1
|
0.004
|
|
Asian
|
2.3
|
5.3
|
0.001
|
|
Family history of aortic disease
|
14.5
|
8.8
|
0.002
|
|
Risk factors:
|
|
|
|
|
Marfan's syndrome
|
4.6
|
3.7
|
0.281
|
|
Smoking
|
53.7
|
51.9
|
0.549
|
|
Hypertension
|
72.6
|
72.5
|
0.965
|
|
Atherosclerosis
|
23.7
|
18.9
|
0.005
|
|
Cocaine usage
|
1.5
|
1.9
|
0.534
|
|
Bicuspid aortic valve
|
5.3
|
3.9
|
0.13
|
|
Diabetes
|
8.8
|
7.2
|
0.134
|
|
Prior surgery:
|
|
|
|
|
Coronary artery bypass grafting
|
3.5
|
5.3
|
0.053
|
|
Aortic valve replacement
|
3.1
|
4.8
|
0.061
|
|
Aortic aneurysm/dissection repair
|
7
|
6.1
|
0.407
|
Note: Bold values indicate significance at p < 0.05.
Regarding presentation symptoms ([Table 2]), there appeared to be no significant difference in time from symptom onset to diagnosis
or treatment between the two groups. There were interesting patterns in the quality
of pain, however. Those with head and neck pain described significantly more sharp
pain (49.6% vs. 43.6% p = 0.024), pressure-type pain (50.4% vs. 38% p < 0.001), abrupt onset of pain (89.8% vs. 78.5% p < 0.001), back pain (43.3% vs. 37.5% p = 0.005), chest pain (87.6% vs. 79.3% p < 0.001), and migrating pain (20.1% vs. 11% p < 0.001). Patients with head and neck pain also tended to present with cardiovascular
accident (CVA) more often (7% vs. 4.6% p = 0.01), arch vessel involvement (44.3% vs 38% p = 0.01), and intramural hematoma (11.7% vs. 8.1% p = 0.003). However, patients with head and neck pain had significantly lower rates
of congestive heart failure (CHF) (4.8% vs. 7% p = 0.03) and shock and tamponade (8.3% vs. 11.1% p = 0.03).
Table 2
Comparison of presenting symptoms in patients with head and neck pain compared with
those without head and neck pain
|
Variables
|
Head/neck pain group
|
No head/neck pain group
|
p-Value
|
|
Hours from symptom onset to presentation at initial hospital
|
0.75–3.33
|
0.833–3.5
|
0.189
|
|
Presenting within 6 h of symptom onset
|
87.4
|
84
|
0.076
|
|
Hours from symptom onset to presentation at initial hospital
|
0.75–3.33
|
0.833–3.5
|
0.189
|
|
Hours from symptom onset to diagnosis (avg)
|
2.87–12.75
|
2.66–12.19
|
0.534
|
|
Hours from symptom onset to diagnosis:
|
|
|
|
|
0–4 h
|
40.6
|
41.9
|
|
|
4–24 h
|
45.2
|
43.1
|
|
|
24 h
|
14.2
|
15.1
|
|
|
Quality of pain:
|
|
|
|
|
Tearing/ripping
|
29.9
|
30.2
|
0.926
|
|
Sharp
|
49.6
|
43.6
|
0.024
|
|
Pressure
|
50.4
|
38
|
< 0.001
|
|
Burning
|
13.8
|
12.6
|
0.534
|
|
Abrupt onset of pain
|
89.8
|
78.5
|
< 0.001
|
|
Back pain
|
43.3
|
37.5
|
0.005
|
|
Chest pain
|
87.6
|
79.3
|
< 0.001
|
|
Migrating pain
|
20.1
|
11
|
< 0.001
|
|
Coma/altered consciousness
|
10
|
10.7
|
0.598
|
|
Syncope
|
16.7
|
16.5
|
0.927
|
|
Cardiovascular accident
|
7
|
4.6
|
0.01
|
|
Congestive heart failure
|
4.8
|
7
|
0.03
|
|
Hypotension/shock/tamponade
|
24.9
|
27.6
|
0.168
|
|
Shock/tamponade
|
8.3
|
11.1
|
0.03
|
|
Shock
|
5.7
|
7.8
|
0.07
|
|
Tamponade
|
2.5
|
3.4
|
0.266
|
|
Murmur of aortic insufficiency
|
35.8
|
33.7
|
0.337
|
|
Any pulse deficit
|
33
|
34.8
|
0.451
|
|
Diagnostic imaging findings:
|
|
|
|
|
Arch vessel involvement
|
44.3
|
38
|
0.01
|
|
Intramural hematoma (def)
|
11.7
|
8.1
|
0.003
|
|
Periaortic hematoma
|
21
|
19.6
|
0.463
|
|
False lumen thrombosis, complete
|
9.3
|
9
|
0.855
|
|
False lumen thrombosis, partial
|
18.9
|
20.3
|
0.517
|
|
Coronary artery compromise
|
12.9
|
12.9
|
0.978
|
|
Pericardial effusion
|
41.9
|
41.9
|
0.993
|
|
Chest X-ray findings:
|
|
|
|
|
% abnormal
|
72.8
|
73.3
|
0.82
|
|
Wide mediastinum
|
53
|
52.7
|
0.931
|
|
Abnormal aortic contour
|
41.6
|
41.8
|
0.922
|
|
Abnormal cardiac contour
|
22.6
|
24.8
|
0.362
|
|
Displacement/calcification of the aorta
|
5.4
|
6.5
|
0.423
|
|
Pleural effusion
|
9.3
|
12.8
|
0.051
|
Note: Bold values indicate significance at p < 0.05.
When treating patients in cohort 1, those with head and neck pain were more likely
to receive surgical treatment (89.8% vs. 85.2% p = 0.001) and correspondingly less likely to undergo medical treatment alone (7.8%
vs. 11.4% p = 0.005). Significantly more patients with head and neck pain had partial arch replacement
compared with those without head and neck pain (47.3% vs. 41.8% p = 0.017) ([Table 3]).
Table 3
Treatment modalities for the head and neck pain and no head and neck pain
|
Variables
|
Head/neck pain group
|
No head/neck pain group
|
p-Value
|
|
Type of management
|
|
|
|
|
Type A surgical
|
89.8
|
85.2
|
0.001
|
|
Type A medical
|
7.8
|
11.4
|
0.005
|
|
Endovascular
|
1.3
|
1.8
|
0.284
|
|
Surgery
|
|
|
|
|
Surgery after 24 h
|
15.5
|
15.4
|
0.938
|
|
Hours from symptom onset to surgery
|
6–21
|
6–21
|
0.244
|
|
Root replacement
|
34.7
|
37.9
|
0.171
|
|
Ascending aortic replacement
|
95.1
|
93.7
|
0.202
|
|
Complete arch replacement
|
16.4
|
17.1
|
0.704
|
|
Partial arch replacement
|
47.3
|
41.8
|
0.017
|
|
Initial management of β-blockers
|
56.5
|
56
|
0.807
|
Note: Bold values indicate significance at p < 0.05.
While still in the hospital, patients with head and neck pain had higher rates of
CVA as a complication (13% vs. 9.9% p = 0.016). Regarding overall outcomes ([Table 4]), mortality among patients with head and neck was 19.5% compared with 23% in those
without head and neck pain (p = 0.038). Surgical mortality was slightly lower among patients with head and neck
(14.8% and 18.8% p = 0.017). Conversely, those patients treated medically showed higher mortality rates
among patients with head and neck pain compared with those without (67.7% vs. 51.6%
p = 0.022).
Table 4
Hospital complications and overall outcomes for patients with head and neck pain compared
with those without head and neck pain
|
Variables
|
Head/neck pain group
|
No head/neck pain group
|
p-Value
|
|
In-hospital complications:
|
|
|
|
|
New neurologic deficit
|
26.5
|
24.8
|
0.343
|
|
Cerebral vascular injury
|
13
|
9.9
|
0.016
|
|
Transient neurodeficit
|
67.9
|
60.1
|
0.085
|
|
Coma
|
4.5
|
5.5
|
0.314
|
|
Myocardial ischemia
|
85.1
|
88.2
|
0.063
|
|
Hypotension
|
29.4
|
29.3
|
0.939
|
|
Cardiac tamponade
|
18.2
|
18.6
|
0.799
|
|
Mortality:
|
|
|
|
|
Type A overall
|
19.5
|
23
|
0.038
|
|
Type A surgical mortality
|
14.8
|
18.8
|
0.017
|
|
Type A medical mortality
|
67.7
|
51.6
|
0.022
|
|
Cause of mortality:
|
|
|
|
|
Neurologic
|
10.4
|
8.2
|
0.387
|
|
Tamponade
|
4.5
|
6.8
|
0.312
|
|
Visceral Ischemia
|
8.4
|
8.5
|
0.968
|
|
Bleeding
|
5.2
|
2.5
|
0.095
|
|
multiorgan failure
|
9.7
|
9.7
|
0.991
|
|
Cardiac
|
7.8
|
9.1
|
0.609
|
|
Rupture
|
18.2
|
18.3
|
0.984
|
|
Unknown
|
1.9
|
4.1
|
0.557
|
Note: Bold values indicate significance at p < 0.05.
In our subgroup analysis of patients with head and neck pain only versus those who
had head and neck pain with chest or back, we found that those with head and neck
pain only were significantly older (66.4 vs. 61.1 p = 0.001). Furthermore, those with head and neck pain only had higher rates of diabetes
(19.6% vs. 7.4% p = 0.003), atherosclerosis (39.2% vs. 24.1% p = 0.018), and previous coronary bypass surgery (12% vs. 1.9% p < 0.001) compared with those who also had chest or back pain ([Table 5]).
Table 5
Comparison of baseline demographic risk factors in patients with head and neck pain
only compared with those with head and neck pain with chest or back pain
|
Demographics
|
Head/neck pain only group
|
Head/neck pain + chest/back pain group
|
p-Value
|
|
Age (overall mean)
|
66.4 ± 10.6
|
61.1 ± 14.4
|
0.001
|
|
< 40 y
|
1 (1.9%)
|
45 (8.3%)
|
0.167
|
|
> 70 y
|
19 (36.5%)
|
160 (29.5%)
|
0.292
|
|
Female sex
|
14 (26.9%)
|
207 (38.2%)
|
0.108
|
|
White
|
45 (91.8%)
|
481 (93.6%)
|
0.552
|
|
African American
|
1 (2%)
|
16 (3.1%)
|
1.000
|
|
Asian
|
1 (2%)
|
9 (1.8%)
|
0.601
|
|
Family history of aortic disease
|
0 (0%)
|
30 (14.2%)
|
0.084
|
|
Risk factors:
|
|
|
|
|
Marfan's syndrome
|
1 (2%)
|
25 (4.7%)
|
0.717
|
|
Smoking (current)
|
5 (26.3%)
|
68 (33%)
|
0.551
|
|
Hypertension
|
39 (78%)
|
378 (71.5%)
|
0.324
|
|
Atherosclerosis
|
20 (39.2%)
|
126 (24.1%)
|
0.018
|
|
Cocaine usage
|
0 (0%)
|
8 (1.6%)
|
1.000
|
|
Bicuspid aortic valve
|
2 (4.8%)
|
29 (6.1%)
|
1.000
|
|
Diabetes
|
10 (19.6%)
|
39 (7.4%)
|
0.003
|
|
Prior surgery:
|
|
|
|
|
Coronary artery bypass grafting
|
6 (12%)
|
10 (1.9%)
|
< 0.001
|
|
Aortic valve replacement
|
1 (2%)
|
13 (2.5%)
|
1.000
|
|
Aortic aneurysm/dissection repair
|
4 (8.2%)
|
36 (7%)
|
0.769
|
Note: Bold values indicate significance at p < 0.05.
At presentation, there were higher rates of coma and altered mental status (32.7%
vs. 9.1% p < 0.001), syncope (32.7% vs. 14.4% p = 0.001), and CVA (18.4% vs. 6.6% p = 0.003) in the head and neck pain only group as seen in [Table 6]. However, we found no difference in time to presentation or time to diagnosis between
the two groups in cohort 2. Furthermore, there was no difference found in management
type or extent of surgery between the two groups ([Table 7]).
Table 6
Comparison of presenting symptoms in patients with head and neck pain only compared
with those with head and neck pain with chest or back pain
|
Variables
|
Head/neck pain only group
|
Head/neck pain + chest/back pain group
|
p-Value
|
|
Presenting within 6 h of symptom onset
|
28 (90.3%)
|
324 (86.9%)
|
0.782
|
|
Hours from symptom onset to presentation at initial hospital
|
1.25 (0.85–4.42)
|
1.25 (0.75–3.21)
|
0.600
|
|
Hours from symptom onset to diagnosis (avg)
|
5.62 (3.13–10.13)
|
5.17 (2.88–14.00)
|
0.889
|
|
Hours from symptom onset to diagnosis:
|
|
|
0.770
|
|
0–4 h
|
10 (34.5%)
|
135 (40.1%)
|
|
|
4–24 h
|
15 (51.7%)
|
151 (44.8%)
|
|
|
24 h
|
4 (13.8%)
|
51 (15.1%)
|
|
|
Quality of pain:
|
|
|
|
|
Tearing/ripping
|
3 (12.5%)
|
116 (35.3%)
|
0.024
|
|
Sharp
|
11 (44%)
|
169 (50.8%)
|
0.515
|
|
Pressure
|
11 (39.3%)
|
172 (55.3%)
|
0.103
|
|
Burning
|
5 (20.8%)
|
42 (15.3%)
|
0.478
|
|
Abrupt onset of pain
|
43 (84.3%)
|
477 (90.2%)
|
0.19
|
|
Back pain
|
0 (0%)
|
256 (47.2%)
|
< 0.001
|
|
Chest pain
|
0 (0%)
|
521 (96.1%)
|
< 0.001
|
|
Migrating pain
|
2 (4%)
|
112 (21.5%)
|
0.001
|
|
Coma/altered consciousness
|
16 (32.7%)
|
47 (9.1%)
|
< 0.001
|
|
Syncope
|
17 (32.7%)
|
77 (14.4%)
|
0.001
|
|
Cardiovascular accident
|
9 (18.4%)
|
34 (6.6%)
|
0.003
|
|
Congestive heart failure
|
3 (6.4%)
|
23 (4.5%)
|
0.471
|
|
Hypotension/shock/tamponade
|
14 (28.6%)
|
126 (24.3%)
|
0.505
|
|
Shock/tamponade
|
2 (4.1%)
|
43 (8.3%)
|
0.412
|
|
Shock
|
1 (2%)
|
31 (6%)
|
0.510
|
|
Tamponade
|
1 (2%)
|
12 (2.3%)
|
1.000
|
|
Murmur of aortic insufficiency
|
17 (36.2%)
|
178 (38.5%)
|
0.751
|
|
Any pulse deficit
|
11 (33.3%)
|
111 (27.8%)
|
0.499
|
|
Diagnostic imaging findings:
|
|
|
|
|
Arch vessel involvement
|
23 (56.1%)
|
169 (41.8%)
|
0.079
|
|
Intramural hematoma (def)
|
5 (9.6%)
|
62 (11.6%)
|
0.663
|
|
Periaortic hematoma
|
7 (14.9%)
|
103 (23.8%)
|
0.168
|
|
False lumen thrombosis, complete
|
0 (0%)
|
34 (10.1%)
|
0.093
|
|
False lumen thrombosis, partial
|
6 (21.4%)
|
62 (18.4%)
|
0.692
|
|
Coronary artery compromise
|
3 (7%)
|
52 (13.6%)
|
0.336
|
|
Pericardial effusion
|
23 (45.1%)
|
203 (42.3%)
|
0.700
|
|
Chest X-ray findings:
|
|
|
|
|
% abnormal
|
30 (78.9%)
|
301 (75.3%)
|
0.612
|
|
Wide mediastinum
|
18 (48.6%)
|
199 (52.5%)
|
0.654
|
|
Abnormal aortic contour
|
15 (40.5%)
|
153 (41.2%)
|
0.934
|
|
Abnormal cardiac contour
|
9 (24.3%)
|
76 (20.7%)
|
0.601
|
|
Displacement/calcification of the aorta
|
4 (10.8%)
|
20 (5.4%)
|
0.258
|
|
Pleural effusion
|
1 (2.7%)
|
35 (9.4%)
|
0.231
|
Note: Bold values indicate significance at p < 0.05.
Table 7
Treatment modality used for those with head and neck pain only compared with those
with head and neck pain and chest or back pain
|
Variables
|
Head/neck pain only group
|
Head/neck pain + chest/back pain group
|
p-Value
|
|
Type of management:
|
|
|
|
|
Type A surgical
|
44 (84.6%)
|
491 (90.6%)
|
0.169
|
|
Type A medical
|
7 (13.5%)
|
42 (7.7%)
|
0.153
|
|
Endovascular
|
0
|
3 (0.6%)
|
1
|
|
Surgery:
|
|
|
|
|
Surgery after 24 h
|
9 (20%)
|
79 (16.2%)
|
0.51
|
|
Hours from symptom onset to surgery
|
11 (7.75–25.17)
|
10.64 (6.73–23.63)
|
0.481
|
|
Root replacement
|
14 (34.1%)
|
1
|
0.793
|
|
Ascending aortic replacement
|
42 (93.3%)
|
443 (94.5%)
|
0.733
|
|
Complete arch replacement
|
6 (14%)
|
64 (14.4%)
|
0.93
|
|
Partial arch replacement
|
20 (44.4%)
|
20 (46.4%)
|
0.798
|
|
Initial management of β-blockers
|
21 (47.7%)
|
268 (53.5%)
|
0.463
|
Regarding outcomes, there was a significantly higher percentage of patients with head
and neck pain only who had cardiac tamponade while in the hospital (30% vs. 16.6%
p = 0.018). Regarding mortality rates, patients with head and neck pain only had higher
overall mortality (34.6% vs. 19.9% p = 0.013) and higher surgical mortality (29.5% vs. 15.3% p = 0.014) ([Table 8]). A logistic regression of mortality was conducted with predictive variables of
head and pain, preoperative hypotension, presence of ascending intimal tear, or intramural
hematomas, which found that head and neck pain status was not associated with higher
mortality after controlling for the other factors (p = 0.36) ([Tables 9–11]).
Table 8
Hospital complications and overall outcomes for patients with head and neck pain only
compared with those with head and neck pain with chest or back pain
|
Variables
|
Head/neck pain only group
|
Head/neck pain + chest/back pain group
|
p-Value
|
|
In-hospital complications:
|
|
|
|
|
New neurologic deficit
|
19 (36.5%)
|
144 (27.5%)
|
0.167
|
|
Cerebral vascular injury
|
14 (26.9%)
|
71 (13.8%)
|
0.011
|
|
Transient neurodeficit
|
|
|
|
|
Coma
|
4 (7.7%)
|
24 (4.7%)
|
0.312
|
|
Myocardial ischemia
|
4 (8.3%)
|
63 (12.6%)
|
0.494
|
|
Hypotension
|
16 (31.4%)
|
163 (31.5%)
|
0.989
|
|
Cardiac tamponade
|
15 (30%)
|
86 (16.6%)
|
0.018
|
|
Mortality:
|
|
|
|
|
Type A overall
|
18 (34.6%)
|
108 (19.9%)
|
0.013
|
|
Type A surgical mortality
|
13 (29.5%)
|
27 (15.3%)
|
0.014
|
|
Type A medical mortality
|
5 (71.4%)
|
28 (66.7%)
|
1
|
|
Cause of mortality:
|
|
|
|
|
Neurologic
|
3 (16.7%)
|
13 (12%)
|
0.701
|
|
Tamponade
|
0
|
6 (5.6%)
|
0.593
|
|
Visceral ischemia
|
3 (16.7%)
|
9 (8.3%)
|
0.377
|
|
Bleeding
|
1 (5.6%)
|
6 (5.6%)
|
1
|
|
Multiorgan failure
|
1 (5.6%)
|
6 (5.6%)
|
1
|
|
Cardiac
|
0 (5.6%)
|
9 (8.3%)
|
1
|
|
Rupture
|
1 (5.6%)
|
26 (24.1%)
|
0.118
|
|
Unknown
|
6 (33.3%)
|
31 (28.7%)
|
0.69
|
Note: Bold values indicate significance at p < 0.05.
Table 9
Logistic regression model of mortality with independent variables of head and neck
pain status, age > 65 years, preoperative hypotension, ascending intimal tear, intramural
hematoma
|
Variables
|
β-Coefficient
|
p-Value
|
Risk ratio
|
95% confidence interval: lower
|
95% confidence interval: upper
|
|
Head and neck and chest or back pain
|
−0.396
|
0.359
|
0.673
|
0.289
|
1.568
|
|
Age > 65 y
|
0.806
|
0.003
|
2.238
|
1.318
|
3.802
|
|
Preoperative hypotension
|
1.395
|
0.000
|
4.037
|
2.348
|
6.942
|
|
Ascending intimal tear
|
−0.837
|
0.002
|
0.433
|
0.252
|
0.745
|
|
Intramural hematoma
|
−0.705
|
0.05
|
0.494
|
0.244
|
1.001
|
|
Constant
|
−1.277
|
0.004
|
0.279
|
|
|
Note: Bold values indicate significance at p < 0.05.
Table 10
Hosmer-Lemeshow test for model fit
|
Chi-square
|
Degrees of freedom
|
Significance
|
|
10.495
|
8
|
0.232
|
Note: Bold values indicate significance at p < 0.05.
Table 11
Area under the curve for the logistic regression model
|
Area under the curve
|
|
0.736
|
Discussion
The acute and lethal nature of acute aortic dissection creates strong impetus for
discovering risk factors, symptoms, and other patterns that may speed its diagnosis
and decrease its mortality. Complicating this effort are the multiple presenting patterns
that occur clinically.[9]
[13]
[14]
[15]
[16] Head and neck pain is one of these aberrant presentation patterns, which is present
in 26.1% of all Type A patients in our database to date.
The pathophysiology of this pain location is not well defined in the literature. Whereas
classic angina pectoris is thought to be caused by stimulation of sympathetic afferent
nerves around the heart, several studies suggest that the afferent pathway of the
vagus nerve may be a cause of cardiac-related head and neck pain. Stimulation of vagal
afferents stimulates nerve endings in C1–C3, which corresponds to the receptive field
for the neck, jaw, upper arm, and ear.[17]
[18]
[19]
Another possible mechanism of head pain might be partially explained by our findings
that there were significantly more patients with arch vessel involvement and CVA in
the head and neck cohort. Because the anterior and posterior circulations of the brain
derive from the carotids and the vertebral arteries, respectively, which ultimately
obtain blood flow from the aorta, arch vessel involvement has the potential to directly
alter flow to the brain. Thus, this would represent a primary cause of the pain rather
than referred pain from the heart. Studies have shown that patients with carotid or
vertebral dissection can present with headache and neck pain.[20]
[21]
Regarding the higher proportion of head and neck pain found in women in our first
analysis ([Table 1]), it has been shown in two IRAD analyses that women on average present differently
than men do with aortic dissection. Women were found to present more frequently with
coma and/or altered mental status compared with men.[13]
[22] Our current data suggest that one reason for higher rates of head and neck pain
among women might be more frequent CVA. A 2004 IRAD paper found that 7.9% of women
with acute aortic dissection presented with CVA compared with 5.2% men, though this
difference was not significant.[13] When analyzing the subgroups in cohort 2, this gender preponderance of head and
neck pain in women did not remain. This suggests that though women are more likely
to experience head and neck pain overall, those women who have head and neck pain
are equally likely to present with chest pain and back pain or with head and neck
pain only.
An unexpected finding in our first analysis was the higher rates of chest pain and
back pain in the head and neck group along with higher rates of all of the various
permutations of pain types such as tearing, sharp, and migrating ([Table 2]). This may be because “worst-ever pain” in the chest, which radiates to the back,
is still the predominant way that patients experience acute aortic dissection, it
will remain the most prevalent presentation despite the presence of head and neck
pain.[16]
[23]
[24] On the other hand, a reason that the non–head and neck group had fewer types of
pain may be related to the fact that 15% of patients present with painless dissection,
suggesting that there may be aortic dissection patients who are biologically predisposed
to feel less pain.[8]
[25] In fact, in our second analysis, those who experienced isolated head and neck pain
were more likely to be older, have more atherosclerosis, and had a higher prevalence
of diabetes, all of which were shown to be higher in patients who presented with painless
cardiac syndromes.[25]
[26]
[27]
[28] In addition, patients with head and neck pain had higher rates of syncope, coma/altered
mental status, syncope, and CVA, which are neurologic phenomena, which may affect
the ability to perceive pain, and which may explain why these patients did not report
more types of pain.[29]
Despite the potential for delayed diagnosis, we found no delay to diagnosis or treatment
in either cohort. Furthermore, in cohort 1, those with head or neck pain had lower overall mortality and lower surgical mortality. This was unexpected given the atypical presentation. However, given the fact that
those with head and neck pain had a higher rate of chest pain and back pain than the
group without head and neck pain, we wondered whether the presence of chest and back
pain obscured the impact of having head and neck pain. Thus, we conducted a second
analysis (cohort 2) with head and neck pain only versus those with head and neck pain
with chest or back pain to see the effect of having only head and neck pain (without
the influence of chest pain or back pain). We found in the second analysis that there
was significantly higher overall and surgical mortality in the head and neck pain
only group.
Reasons for the difference in mortality between cohorts 1 and 2 are not clear. Despite
those in cohort 1 with head and neck pain having higher rates of the classic symptoms
such as chest pain and back pain, we did not see any improvement in time to diagnosis
in that group, which would explain their lower mortality. On the other hand, in cohort
2, those who experienced only head and neck pain were older, had higher rates of atherosclerosis,
diabetes, and had more coronary artery bypass grafting operations, which represent
several important risk factors for mortality.[29]
[30]
[31] To analyze this further, we created a logistic regression model for mortality with
head and neck pain status, age > 65 years, preoperative hypotension, site of intimal
tear and evidence of intramural hematoma as the dependent variables. Ultimately, when
controlled for those variables, head and pain neck pain status was no longer significant
associated with increased mortality. Given that age > 65 and preoperative hypotension
had significant, positive risk ratios in the model, this suggests that these two variables
explained some of the increase in mortality.
In conclusion, we have shown that patients with head and neck pain have higher rates
of proximal arch involvement and stroke, which adds credence to classic clinical teaching
about the association of head and neck pain with arch involvement. Furthermore, patients
with head and neck overall tend to have classic dissection symptoms and thus have
no delay in diagnosis and also lower mortality compared with those without head and
neck pain. However, in our subgroup analysis (cohort 2), we showed that those with
isolated head and neck pain tend to have worse comorbidities, more neurologic symptoms
on presentation, and worse overall and surgical mortality. This higher mortality may
be partially explained by the older age and more preoperative hypotension in the head
and neck pain only group. These patients, though a small fraction of our Type A dissection
population, represent an area for added vigilance and improvement in our pursuit of
better care of acute aortic dissection patients.
Our study has several limitations. IRAD is composed of patients in academic referral
centers and may not represent the overall population of aortic dissection patients.
In addition, patients with abnormal presentations may pass away prior to diagnosis
or transfer to tertiary aortic centers and are thus underrepresented in our population.
Furthermore, given our international scope, there is possible variation in treatment
patterns, which may complicate outcomes measurements.
