Outcomes of endovascular occlusion and stenting in the treatment of carotid blowout

Abstract

Background and purpose

Carotid blowout is a life threatening complication of invasive head and neck cancers and their treatments. This is commonly treated with endovascular embolization and carotid stenting. Using the Nationwide Inpatient Sample, we report the immediate clinical results of patients receiving embolization and/or stenting for treatment of carotid blowout associated with head and neck cancer.

Materials and methods

Using the Nationwide Inpatient Sample from the period 2003–2011, we defined carotid blowout patients as those with head and neck malignancies receiving carotid stenting and/or endovascular embolization without open surgery. Outcomes studied included mortality, acute ischemic stroke, hemiplegia/paresis, and other post-operative neurologic complications. Outcomes for the endovascular embolization and carotid stenting group were compared.

Results

A total of 1218 patients underwent endovascular treatment for carotid blowout. Of these, 1080 patients (88.6%) underwent embolization procedures and 138 patients (11.4%) underwent carotid stenting. The mortality rate of endovascular embolization patients was similar to that of carotid stenting patients (8.0%, 95% confidence interval (CI) = 6.5%–9.7% versus 10.2%, 95% CI=6.0%–16.4%, p = 0.36). Stroke rate was similar between embolization patients and stenting patients (2.3%, 95% CI=1.6%–3.4% vs. 3.4%, 95% CI=1.3%–8.4%, p = 0.43). Hemiplegia rates were significantly higher rate in stenting patients compared with endovascular occlusion patients (3.8%, 95% CI=1.3%–8.4% vs. 1.4%, 95% CI=1.4%–2.4%, p = 0.05). The rate of post-operative neurologic complications was higher in stenting patients compared with embolization patients (6.5%, 95% CI=3.3%–12.1% vs. 1.4%, 95% CI=0.9%–2.4%, p < 0.0001).

Conclusions

Given the natural history of carotid blowout, carotid stenting and endovascular embolization are acceptable means of treating this disease. Endovascular embolization remains the most common treatment among patients with head and neck cancers with lower overall rates of post-operative neurologic complications, including hemiplegia/paresis and stroke.

Introduction

Carotid blowout is an uncommon but life threatening complication of invasive head and neck cancers and occurs following tumoral invasion of the internal carotid artery (ICA), common carotid artery (CCA) or external carotid artery (ECA) branches. This complication has also been associated with surgical neck dissections and radiation therapies.^1–5 When untreated, carotid blowout syndrome is associated with mortality rates of 60% and morbidity rates of 40%.^1–3 The emergence of endovascular treatments such as parent artery occlusion and covered stents has resulted in markedly improved outcomes.^3,5–14 However, since this condition is relatively uncommon, available information about outcomes and treatment modalities is based only on small single center series. Using the Nationwide Inpatient Sample (NIS), we report the immediate clinical results of patients receiving embolization and/or stenting for treatment of carotid blowout associated with head and neck cancer.

Methods

Patient population, demographics and comorbidities

We purchased the NIS hospital discharge database for the period 2003–2010 from the Healthcare Cost and Utilization Project of the Agency for Healthcare Research and Quality, Rockville, MD, USA. The NIS is a hospital discharge database that represents 20% of all inpatient admissions to nonfederal hospitals in the United States. Patient procedures and diagnoses were identified using both International Classification of Diseases 9th Revision (ICD-9) and Clinical Classification Software (CCS) codes. CCS codes are specific to the NIS and are a means of coding multiple diagnoses into one broader code (i.e. multiple different head and neck malignancies can be categorized as one CCS code: “11”). Further explanation is provided in the following link: http://www.hcup-us.ahrq.gov/toolssoftware/ccs/ccsfactsheet.jsp.

Patients who had head and neck malignancies other than paragangliomas and underwent endovascular embolization or carotid stenting were identified ICD-9 codes. These patients were considered to have carotid blowout as generally the primary reason to embolize patients with head and neck malignancies (excluding paraganglioma) and not pursue further surgery is because the tumors have invaded the carotid artery branches and have started to bleed. Patients with carotid artery stenosis were excluded so as to exclude any patients who underwent stenting for post-radiation carotid artery stenosis. For determining treatment outcomes, two groups of patients were defined: 1) patients undergoing endovascular embolization (ICD-9 Procedure Code 397.2, 397.9), and 2) patients undergoing carotid stenting (ICD-9 Procedure Code 0061, 0063, 0064). Demographic variables for each group included age, gender and race (White, non-White). Charlson co-morbidity index was calculated for each patient. The distribution of tumor location for each group was also compared. Tumor location was categorized as: 1) laryngeal cancer (ICD-9 161.0, 161.1–161.9, 146.4, 146.5); 2) hypopharyngeal cancer (ICD-9 148.0–148.9; 3) oral cavity cancer (ICD-9 140.0–140.9, 141.1, 141.2, 141.3, 141.4, 141.9, 143.0, 143.1, 143.8, 143.9, 144.0, 144.4, 144.8, 144.9, 145.0–145.9, 170.1); 4) oropharyngeal/nasopharyngeal cancer (ICD-9 141.0, 141.5, 141.6, 141.8, 145.3, 145.4, 146.0–146.9,, 147.0–147.9, 149.0, 149.1),; 5) sinus cancer (160.2–160.9); and 6) other/non-specified (142.0–142.9, 148.0–148.2, 170.0. 173.3, 173.4, 193, 194.1, 195.0, 196.0).

Outcomes

Outcomes studied included mortality, acute ischemic stroke, hemiplegia/paresis, other post-operative neurologic complications (i.e. post-operative neuropathy, post-operative headache), post-operative hemorrhage, acute renal failure, post-operative infection, sepsis, skin/soft tissue infection, pneumonia, pulmonary embolism, and length of stay. Outcomes for the endovascular embolization and carotid stenting group were compared.

Statistical analysis

For comparison of categorical variables, chi-squared testing was performed. For comparison of continuous variables, Student's t-test was performed. Means are presented with their associated standard deviations. Proportions are presented with their 95% confidence intervals (CIs). A multivariate logistic regression analysis was performed to determine predictors of mortality and stroke. Variables included in this model were age, gender, Charlson co-morbidity index, tumor location, and treatment modality. All statistical analyses were performed using the SAS-based statistical software package JMP 9.0 (www.jmp.com). To obtain national estimates, discharge weights were applied.

Results

Patient demographics and trends in treatment

From 2003 to 2011, a total of 1218 patients underwent endovascular treatment for carotid blowout. Of these, 1080 patients (88.6%) underwent embolization procedures and 138 patients (11.4%) underwent carotid stenting. Overall, the most common location for a primary malignancy was oro/nasopharyngeal (486 patients, 39.9%). Patients undergoing carotid stenting were generally older than patients undergoing embolization (63.1 ± 36.9 years vs. 51.6 ± 44.4 years, p < 0.0009). Patients undergoing stenting had a similar mean Charlson co-morbidity index compared with endovascular embolization patients (0.9 ± 3.1 vs. 0.6 ± 2.5, p = 0.18). Tumor location varied significantly between the stenting and endovascular embolization groups. There was a higher proportion of laryngeal cancer patients in the stenting group (39 patients, 28.3%, 95% CI=21.4%–36.3%) when compared with the endovascular embolization group (10.7%, 95% CI=9.0%–12.7%, p < 0.0001). There was a lower proportion of patients with oro/nasopharyngeal cancer in the stenting group (34 patients, 24.6%, 95% CI=18.2%–32.5%) when compared with the endovascular embolization (451 patients, 41.8%, 95% CI=38.9%–44.7%, p < 0.0001). These data are summarized in Table 1.

Table 1.

Patient characteristics.

	Endovascular blowout repair	Embolization	Stenting	p
N	1218	1080 (88.6)	138 (11.4)	–
Demographics
Mean age, years (SD)	52.9 (44.3)	51.6 (44.4)	63.1 (36.9)	0.0009
N (%, 95% CI) female	362 (29.8, 27.2–32.4)	296 (27.5, 24.8–30.2)	66 (47.5, 39.7–56.1)	<0.0001
Mean (SD) CCI	0.6 (2.6)	0.6 (2.5)	0.9 (3.1)	0.18
Tumor location
N (%, 95% CI) laryngeal cancer	155 (12.7, 11.0–14.7)	116 (10.7, 9.0–12.7)	39 (28.3, 21.4–36.3)	<0.0001
N (%, 95% CI) hypopharyngeal cancer	34 (2.8, 2.0–3.9)	30 (2.7, 1.9–4.0)	<10 (3.6, 1.3–8.4)
N (%, 95% CI) oral cavity cancer	102 (8.4, 6.9–10.1)	88 (8.1, 6.7–9.9)	15 (10.8, 6.6–17.3)
N (%, 95% CI) Oropharyngeal/nasopharyngeal cancer	486 (39.9, 37.2–42.7)	451 (41.8, 38.9–44.7)	34 (24.6, 18.2–32.5)
N (%, 95% CI) sinus	55 (4.5, 3.5–5.8)	55 (5.1, 3.9–6.6)	0 (0.0, 0.0–3.3)
N (%, 95% CI) other/non-specified	386 (31.7, 29.1–34.4)	340 (31.5, 28.8–34.3)	45 (32.6, 25.3–40.8)

Exact numbers in cases where outcomes have <10 patients cannot be reported per the Healthcare Cost and Utilization Project of the Agency for Healthcare Research and Quality-Nationwide Inpatient Sample user agreement.

CI: confidence interval; CCI: Charlson comorbidity index

Outcomes

Overall in-hospital mortality rate for patients receiving endovascular treatment for carotid blowout was 8.2% (95% CI=6.8%–9.9%; 100/1218). Mortality rate for patients undergoing endovascular embolization (86/1080, 8.0%, 95% CI=6.5%–9.7%) was similar to that of patients undergoing carotid stenting (14/138, 10.1%, 95% CI=6.0%–16.4%, p = 0.36). Overall mean length of stay was 7.5 ± 21.4 days and was similar between the endovascular embolization (7.5 ± 21.6 days) and stenting (7.4 ± 20.3 days) groups (p = 0.96).

Endovascular repair for carotid blowout was associated with a 2.3% (30/1218; 95% CI=1.6%–3.4%) risk of acute ischemic stroke with no significant difference between the two groups (p = 0.43). The overall rate of hemiplegia/paresis was 1.7% (95% CI=1.1%–2.6%) with a significantly higher rate in the stenting group (3.8%, 95% CI=1.3%–8.4%, < 10/138) when compared with the endovascular occlusion group (1.4%, 95% CI=0.9%–2.4%, 16/1080, p = 0.05). The overall rate of other post-operative neurologic complications was higher in the stenting group (6.5%, 95% CI=3.3%–12.1% < 10/138) when compared with the embolization group (1.4%, 95% CI=0.9%–2.4%, 16/1080, p < 0.0001). The overall rate of infectious complications, acute renal failure and post-operative hemorrhage was similar between the two groups. These data are summarized in Table 2.

Table 2.

Outcomes of endovascular repair for carotid blowout.

	Endovascular blowout repair	Embolization	Stenting	p
N	1218	1080 (88.6)	138 (11.4)	–
Outcomes
In-hospital mortality	100 (8.2)	86 (8.0)	14 (10.2)	0.36
Mean (SD) length of stay in days	7.5 (21.4)	7.5 (21.6)	7.4 (20.3)	0.96
Mean (SD) charges in US dollars	116,640 (259,005)	122,692 (265,969)	67,249 (158,747)	0.0008
Complications, N (%, 95% CI)
Acute ischemic stroke	30 (2.4, 1.7–3.5)	25 (2.3, 1.6–3.4)	<10 (3.6, 1.3–8.4)	0.43
Hemiplegia/paresis	21 (1.7, 1.1–2.6)	16 (1.4, 0.9–2.4)	<10 (6.5, 3.3–12.1)	0.05
Other post-operative neurologic complications	25 (2.1, 1.4–3.0)	16 (1.4, 0.9–2.4)	<10 (6.5, 3.3–12.1)	<0.0001
Other post-operative complications	15 (1.2, 0.7–2.0)	15 (1.4, 0.9–2.4)	0 (0.0, 0.0–3.3)	0.17
Acute renal failure	38 (3.2, 2.3–4.3)	33 (3.1, 2.2–4.3)	<10 (3.8, 1.3–8.4)	0.67
Post-operative hemorrhage	25 (2.0, 1.4–3.0)	20 (1.9, 1.2–2.9)	<10 (3.6, 1.3–8.4)	0.37
Pulmonary embolism	27 (2.2, 1.5–3.2)	27 (2.5, 1.7–3.6)	0 (0.0, 0.0–3.3)	0.06
Pneumonia	82 (6.8, 5.5–8.3)	71 (6.6, 5.2–8.2)	11 (7.9, 4.4–13.8)	0.81
Sepsis	20 (1.6, 1.1–2.5)	20 (1.8, 1.2–2.9)	0 (0.0, 0.0–3.3)	0.11
Skin/soft tissue infection	14 (1.1, 0.7–1.9)	14 (1.3, 0.8–2.2)	0 (0.0, 0.0–3.3)	0.19

Exact numbers in cases where outcomes have <10 patients cannot be reported per the Healthcare Cost and Utilization Project of the Agency for Healthcare Research and Quality-Nationwide Inpatient Sample user agreement.

CI: confidence interval

Multivariate analysis

On multivariate analysis, variables associated with a significantly higher odds of mortality included age (p < 0.0001), female gender (p = 0.03), and laryngeal cancer (p = 0.004), hypopharyngeal cancer (p = 0.0001) and oral cavity cancer (p = 0.007). Variables associated with a significantly higher odds of stroke included age (p = 0.002) and increasing Charlson co-morbidity index (p = 0.04). These data are summarized in Table 3.

Table 3.

Multivariate predictors of stroke and mortality.

	Mortality		Stroke
	OR (95% CI)	p	OR (95% CI)	p
Age, by year	1.03 (1.02–1.05)	<0.0001	1.04 (1.01–1.07)	0.002
Female gender	1.73 (1.07–2.79)	0.03	0.00 (0.00–0.11)	0.004
CCI	1.00 (0.85–1.24)	0.97	1.11 (1.01–1.59)	0.04
Tumor location
Laryngeal	2.79 (1.41–5.54)	0.004	0.52 (0.11–2.14)	0.37
Hypopharyngeal	7.26 (2.74–18.52)	0.0001	0.00 (0.00–1.67)	0.11
Oral cavity	2.86 (1.34–5.96)	0.007	3.25 (0.90–11.76)	0.07
Oropharyngeal	1.68 (0.94–3.11)	0.08	1.25 (0.47–3.78)	0.67
Sinus	0.00 (0.00–0.17)	0.008	0.00 (0.00–2.65)	0.20
Treatment
Embolization	Ref.	Ref.	Ref.	Ref.
Stenting	0.79 (0.41–1.43)	0.45	1.82 (0.57–4.83)	0.29

OR: odds ratio; CI: confidence interval; CCI: Charlson comorbidity index; Ref.: reference

Discussion

This study demonstrated that patients receiving endovascular embolization or stenting for carotid blowout have acceptable morbidity and mortality rates given the natural history of carotid blowout. Recently, some studies have highlighted the safety and efficacy of parent artery preservation through the use of carotid stents to treat carotid blowout or near-carotid blowout syndrome.^3,5–14 However, our study demonstrated that endovascular embolization is still a more common treatment modality when compared with carotid stenting. Patients receiving carotid stenting had significantly higher rates of hemiplegia/paresis, cerebral artery occlusion and post-operative neurologic complications; however, on multivariate analysis the difference in stroke rate was not statistically significant. It is important to mention that our study is not aimed at identifying which type of treatment is preferable for treating carotid blowout syndrome. Because each case is unique regarding goals of treatment, severity of disease and location of disease, cases should be managed on a case-by-case basis. Rather, these data can serve to inform providers and patients on expected in-hospital outcomes following endovascular treatment of carotid blowout syndrome.

A number of previously published case series have studied outcomes with both embolization and stenting for treatment of carotid blowout. These studies are limited by small size and thus perioperative complication rates following treatment are highly variable between studies. Studies of endovascular embolization (deconstructive technique) for treatment of carotid blowout have demonstrated stroke rates ranging from 0% to 25%.^6,7,15,16 Reported rebleeding rates also vary considerably, from 0% to nearly 30%.^6,7,15,16 Early mortality rates are similarly variable ranging from 0% to 20%.^6,7,15,16 These prior studies included only small numbers of patients, so wide variation in outcomes can occur simply from small sample size. Our series of 1080 patients receiving endovascular embolization demonstrates a post-operative stroke rate of approximately 2% and an in-hospital mortality rate of 8%. Patients with carotid blowout tend to have very advanced cancer and be approaching the end of life, so it is not surprising that there is significant in-hospital mortality associated with this condition.

While embolization (deconstructive technique) is considered to be quite efficacious in treatment of carotid blowout, several smaller case series have examined the efficacy and safety of ICA preservation with stenting.¹⁶ This is the preferred treatment modality in cases where patients cannot tolerate sacrifice of the ICA.^3,17 Several series have demonstrated that stents have good immediate results in controlling carotid blowout; however, some studies suggest that the long-term efficacy of these stents in preventing bleeding is poor.^3,7,17 In a study of 11 patients receiving stenting and 13 patients receiving endovascular vessel occlusion, Chang et al. found a rebleeding rate of 46% for the stenting group compared with 23% for the occlusion group.⁷ We did not find a statistically significant difference in rebleeding rates between embolization and stenting, but the database used only identifies rebleeding during the initial hospitalization. Reports of early mortality and stroke rates among patients undergoing stenting for carotid blowout range from 0% to 20%.^{3,6,8,9,11,13,14,18} Our series of 138 patients receiving stenting for carotid blowout demonstrates a stroke rate of 3% and an in-hospital mortality rate of 10%.

Limitations

There are a number of limitations to the use of the NIS database, largely stemming from the lack of important clinical data that could provide a more detailed explanation of outcomes. Information on the severity of disease is not provided. Using the NIS, we are unable to determine how successful endovascular embolization and stenting were in treating carotid blowout and no long-term data are available on re-bleeding rates and long-term mortality rates. Data on the treated artery as well as the embolization and stent materials used are not provided in this database. It is possible that patients receiving stenting may have had more serious pathologies compared with the endovascular embolization group as ICA/CCA rupture is more likely to be treated with stenting, while ECA branch vessel bleeding is more likely to be treated with embolization. Data on embolization materials used are not provided due to limitations in ICD-9 codes. For this reason, we have no data as to whether particles, balloons or coils were used in endovascular embolization procedures. Likewise, we have no data as to what types of stents were used as ICD-9 codes are not that specific. As mentioned previously, for patients suffering in-hospital mortality, it is difficult to determine the cause of death and whether it was related to the surgery itself or the patients' underlying comorbidities. Coding errors are also a potential limitation of this study as they are with any study of a large administrative database.

Conclusions

Our study of a large nationwide administrative database demonstrates that given the natural history of carotid blowout, carotid stenting and endovascular embolization are acceptable means of treating this disease. Endovascular embolization remains the most common treatment among patients with head and neck cancers with lower overall rates of post-operative neurologic complications including hemiplegia/paresis and cerebral artery occlusion. Acute ischemic stroke, one of the more common complications associated with treatment of carotid blowout, occurred in less than 3% of patients.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest

The authors declare no conflict of interest.