Cigarette smoking as a significant risk factor for digital vascular disease in patients with systemic sclerosis

Abstract

Objective

Patients with systemic sclerosis (SSc) are at high risk for digital vascular complications, including amputation and gangrene. Cigarette smoking is an important risk factor for vascular disease in the general population. We investigated the influence of cigarette smoking on digital ischemia in patients with SSc.

Methods

We studied 101 patients with SSc (87 women and 14 men, median age 53 years, median disease duration 13 years). Smoking history was defined in terms of current smoking status and total number of pack-years. Digital ischemic events were classified as debridement, hospital admission for intravenous (IV) administration of vasodilators, and digital amputation. The influence of smoking on digital ischemic events was examined using logistic regression, adjusting for age, sex, and disease duration. Results are expressed as the odds ratio (OR) and 95% confidence interval (95% CI).

Results

Of the 101 patients, 21 (21%) were current smokers, 37 (37%) were ex-smokers, and 43 (43%) had never smoked. After adjusting for age, sex, and disease duration, current smokers were significantly more likely than never-smokers to have had debridement (OR 4.5, 95% CI 1.1–18.3) or admission for IV vasodilators (OR 3.8, 95% CI 1.1–12.9). Patients smoking at higher intensity were more likely to require admission for IV vasodilators.

Conclusion

Among patients with SSc, current smokers are 3–4 times more likely than never-smokers to incur digital vascular complications. Resources should be directed to supporting smoking cessation in patients with SSc.

Systemic sclerosis (SSc) is a generalized disorder characterized principally by skin thickening, vascular disease, and immune dysfunction (1). Microvascular abnormalities are well recognized, manifesting as digital ischemia that may require surgical amputation. More recently, it has been recognized that the prevalence of macrovascular disease is increased in patients with SSc (2, 3).

Several factors have been implicated in the pathogenesis of vascular disease in SSc, including damage to the vascular endothelium, hypercoagulability, and reduced fibrinolysis (4–8). Cigarette smoking is a well-known risk factor for other vascular diseases, including ischemic heart disease, atherosclerosis, and stroke (9, 10). Only one previous study has investigated the influence of cigarette smoking on digital vascular disease in patients with SSc (11). That study included 98 patients with SSc referred to the Johns Hopkins Medical Center, Baltimore. The investigators concluded that smoking was not a risk factor for digital ulceration or amputation. However, the demographic characteristics of that study cohort were different from those normally observed in the UK: the proportion of nonwhite patients was relatively high (22%), and the proportion of current smokers was low (13%). We therefore attempted to determine whether cigarette smoking was associated with digital vascular disease in a UK-based cohort of patients with SSc.

PATIENTS AND METHODS

Study population.

Patients were recruited from the SSc clinic at Hope Hospital. The study was approved by the local research ethics committee. All patients attending this clinic are assessed annually by one of us (ALH), using a standardized protocol, and the results are entered onto a clinical database. The database was begun in 1994. At the time that each patient's data were first entered into the database, information from previous hospital visits and procedures was recorded retrospectively. By February 2000, the database included details of 118 patients with SSc who were currently attending the clinic. A short questionnaire about smoking habits was mailed to each of these 118 patients, and 101 (86%) responded. These 101 patients comprise the study population for this analysis. There were no differences between the 101 responders and 17 nonresponders with regard to age, sex, disease duration, disease subtype, or smoking status.

Ascertainment of demographic and clinical data.

Patients were classified as having limited cutaneous SSc if skin involvement was confined to the face and/or was distal to the elbows/knees. Patients with more extensive skin involvement were classified as having diffuse cutaneous SSc. The date of disease onset was based on the patient's first recollection of the onset of symptoms of Raynaud's phenomenon (or of the first non-Raynaud's feature for patients without Raynaud's phenomenon). The level of anticentromere antibodies (ACA) was determined by an indirect immunofluorescence technique that measured IgG antibodies directed against HEp-2 cells, at a serum dilution of 1:40.

Assessment of digital ischemia.

For the purposes of this study, a patient was classified as having or having had digital ischemia if 1 or more of the following were reported: hospital admission for intravenous (IV) administration of vasodilator therapy (either prostacyclin or iloprost), surgical debridement of ≥1 digits, or surgical amputation of ≥1 digits. These data were obtained from the clinical database. In addition, the case notes of all patients were reviewed to verify this information.

Assessment of smoking history.

Smoking history was obtained using a short, self-completed questionnaire. The questionnaire asked whether patients were smoking currently, had smoked in the past, or had never smoked. Smoking was classified as regular smoking of at least 1 cigarette per day. Current smokers and ex-smokers were asked at what age they started smoking and, for ex-smokers, at what age they stopped smoking. Patients were also asked to record the average number of cigarettes they smoked per day. For current smokers and ex-smokers, the total number of pack-years was calculated. A pack normally contains 20 cigarettes. Therefore, if a patient reported smoking an average of 10 cigarettes per day for 6 years, this would be equivalent to 3 pack-years.

Statistical analysis.

The characteristics of patients with limited and diffuse skin disease were compared using the Mann-Whitney U test for continuous data, and the percentage of difference in risk (95% confidence interval [95% CI]) for categorical data. The association of smoking history with digital ischemia was determined using logistic regression, after adjusting for the potential confounders of age, sex, and disease duration. A further analysis was performed, adjusting for ACA. Results are reported as the odds ratio (OR) and 95% CI.

Smoking was entered as an independent variable using 2 separate methods. According to the first method (current smoking status), current smoking, ex-smoking, and never-smoking were entered as independent variables. The ORs represent the association of digital ischemia with current smoking and ex-smoking compared with never smoking (referent group). For the second method (smoking intensity), the total number of pack-years reported for current smokers and ex-smokers was divided by tertiles (1–10, 11–22, and 23–51 pack-years). The ORs represent the association of smoking with digital ischemia in each tertile compared with never-smokers (referent group).

RESULTS

Clinical and demographic characteristics of patients.

Of the 101 patients with SSc, 73 (72%) had limited skin involvement, and 28 (28%) had diffuse disease. The characteristics of patients with limited and diffuse disease are shown in Table 1. There was no difference between patients with limited or diffuse disease with regard to age, sex, or disease duration. Patients with limited skin disease were significantly more likely than those with diffuse disease to have undergone digital debridement or amputation. The prevalence of ACA was nonsignificantly higher in patients with limited skin disease (44%) than in those with diffuse skin disease (11%). Of the 46 patients who had at least 1 digital ischemic event, 23 had received IV vasodilators only, 6 had received IV vasodilators and undergone amputation, 9 had received IV vasodilators and debridement, and 8 had undergone all of these treatments.

Table 1. Clinical and demographic characteristics of patients with systemic sclerosis

	Whole cohort (n = 101)	Limited disease (n = 73)	Diffuse disease (n = 28)	% difference (95% confidence interval)*
* Limited disease vs. diffuse disease. † P not significant, versus limited disease.
Age, years, median (range)	53 (26–80)	53 (26–80)	53 (30–74)†	–
Female, no. (%)	87 (86)	65 (89)	22 (79)	10 (−6, 27)
Disease duration, years, median (range)	13 (1–58)	15 (1–58)	11 (2–56)†	–
Anticentromere positive, no. (%)	35 (35)	32 (44)	3 (11)†	33 (−7, 49)
Admission for intravenous vasodilators, no. (%)	46 (46)	36 (49)	10 (36)	13 (−8, 35)
Digital amputation, no. (%)	15 (15)	14 (19)	1 (4)	15 (4, 27)
Digital debridement, no. (%)	17 (17)	15 (21)	2 (7)	14 (1, 27)

Smoking history.

A total of 21 patients (21%) were current smokers at the time that they completed the questionnaire. Thirty-seven patients (37%) were ex-smokers, and 43 (43%) had never smoked. Among both current smokers and ex-smokers, the median age at which they started smoking was 16 years (range 11–37). For current smokers, the median duration of smoking was 33 years (range 1–55), and the median number of pack-years was 23 (range 1–51). For ex-smokers, the median duration of smoking was 21 years (range 1–44), and the median number of pack-years was 11 (range 1–44). There was no difference in age, disease duration, disease subtype, or ACA with respect to smoking status. However, current smokers were more likely to be male (29%) compared with ex-smokers (16%) and never-smokers (5%).

The influence of cigarette smoking on digital ischemia.

Table 2 illustrates the influence of smoking on digital ischemia in the 101 patients with SSc. Compared with never-smokers, current smokers were approximately 4 times as likely to have been admitted for IV vasodilator therapy (OR 3.8, 95% CI 1.1–12.9) and to have had 1 or more digital debridements (OR 4.5, 95% CI 1.1–18.3). The data for digital amputation showed a similar, but not significantly different, trend (OR 3.4, 95% CI 0.8–15.1). Ex-smokers had an increased (but nonsignificant) risk of debridement or amputation compared with never-smokers.

Table 2. Influence of cigarette smoking on digital ischemia*

	Admission for IV vasodilators	Digital debridement	Digital amputation
* Values are the odds ratio, adjusted for age, sex, and disease duration (95% confidence interval), versus never-smokers. IV = intravenous.
Current smokers	3.8 (1.1–12.9)	4.5 (1.1–18.3)	3.4 (0.8–15.1)
Ex-smokers	0.9 (0.3–2.3)	1.5 (0.4–6.1)	2.0 (0.5–7.9)
Smoking intensity
23–51 pack-years	3.2 (1.0–10.6)	3.0 (0.6–14.4)	2.0 (0.4–10.4)
11–22 pack-years	1.2 (0.4–3.8)	3.0 (0.6–14.7)	2.9 (0.6–14.4)
1–10 pack-years	0.9 (0.3–2.7)	1.9 (0.4–8.5)	2.5 (0.5–11.1)

Patients with digital vascular disease were more likely to be positive for ACA. For example, 60% of patients with a history of digital amputation were ACA positive, compared with 30% of those who had not undergone amputation. However, ACA status was no different in current smokers, ex-smokers, and never-smokers. We analyzed the effect of adjusting for ACA (in addition to age, sex, and disease duration). This adjustment had no influence on the observed association between smoking status and digital outcome. For example, the ORs for current smoking were as follows: IV vasodilators 3.8 (95% CI 0.3–2.2), debridement 4.9 (95% CI 0.4–6.6), and amputation 3.4 (95% CI 0.7–16.2). This means that once the effect of smoking has been taken into account, there is no additional influence of ACA on digital vascular outcome.

The effect of smoking intensity was examined by calculating the total number of pack-years for current smokers and ex-smokers. Patients in the highest tertile (23–51 pack-years) had a 2–3-fold increased risk of digital ischemia compared with never-smokers. This was significant with respect to admission for IV vasodilators (OR 3.2, 95% CI 1.0–10.6). For all of these analyses, similar results were obtained when the data were examined separately for patients with limited and diffuse disease.

DISCUSSION

In this study of 101 UK-based patients with SSc, we determined that current smokers are 3–4 times more likely than never-smokers to require treatment for digital ischemia (e.g., admission for IV vasodilators, digital debridement, or amputation). Digital ischemia was also related to the intensity of smoking, as measured by pack-years. To our knowledge, this is the first study to report an association between smoking and digital vascular disease in patients with SSc. We may have underestimated the burden of digital ischemia in this cohort, because many patients may have had ischemic ulcers that did not require any of the above-mentioned interventions.

SSc is a rare disease, with a prevalence in the UK of only 30.8 per million population (12). It is therefore difficult to recruit large numbers of patients for investigation. In this study, we recruited 101 patients from a center with a special interest in SSc, which attracts tertiary referrals. Although such patients are likely to have more severe disease, this is unlikely to influence the relationship between cigarette smoking and disease severity. In addition, because we performed a cross-sectional study, it is important to consider whether recall bias is an explanation for our results. However, we thought that it was unlikely that patients with more severe digital ischemia would differentially recall that they were current smokers. It is also important to consider whether patients with more severe disease started smoking after they had a digital vascular event. In this study, most of the smokers started smoking long before they developed symptoms consistent with SSc (median age of starting smoking 16 years [range 11–37], median age of disease onset 35 years [range 6–67]). This is, therefore, an unlikely explanation for our results.

The pathogenesis of the microvascular changes in SSc is not fully understood. Abnormalities in the endothelium, in the neuroendothelial control of vascular tone, and in platelet activation are all thought to play key roles (13). Only 1 previous study has investigated the influence of smoking on digital vascular disease in SSc (11). In a cross-sectional study of 98 patients with SSc attending the rheumatology clinic at the Johns Hopkins Medical Center, Wigley et al reported that smoking status (assessed at the last clinic visit) was not associated with ischemic digital ulceration or amputation. The study cohort was similar to ours in terms of demographics and prevalence of digital amputation. Current smoking was slightly less common in their cohort compared with ours (13% versus 21%). However, the main difference was that the North American study included a high proportion of nonwhite patients (22%), compared with only 3% in our study. This difference suggests that genetic factors may underlie the association of digital vascular disease with environmental insults.

Cigarette smoke contains >4,000 components, including reactive oxygen and nitrogen species causing oxidative and nitrosative stress (14, 15). Oxidative stress has been implicated in the pathogenesis of SSc and Raynaud's phenomenon (16). A study of antioxidant levels revealed significantly lower levels of ascorbic acid and selenium in patients with SSc compared with those in controls, and this effect was more marked in current smokers (17). Trials of antioxidant therapy in patients with SSc have yielded conflicting results (18, 19).

Smoking also has direct effects on the vascular endothelium. For example, cigarette smoke increases the adhesion of leukocytes and platelets to endothelial cells (20, 21), and endothelial cell–dependent vasodilation is impaired in chronic smokers (22). These effects may be mediated by disturbances in nitric oxide biosynthesis and reactivity. In addition, smokers have higher plasma viscosity, with elevated hematocrit, red cell volume, and plasma fibrinogen and thrombin levels (23, 24). Together with impaired fibrinolysis (25), these factors cause reduced blood flow to peripheral tissues (26).

Finally, smoking has been shown to increase the risk of vascular complications in other autoimmune diseases, such as systemic vasculitis in patients with rheumatoid arthritis (27), and progression to end-stage renal disease in those with lupus nephritis (28). Among patients with SSc, smokers have more severe pulmonary involvement (29). Although one might expect that smoking is a risk factor for primary Raynaud's phenomenon, the association is not clear-cut (30–32). Of interest, a large American–French collaborative study found no association between smoking and severity of Raynaud's attacks (32). In a study of male US veterans, smoking was no more common in patients with SSc than in controls (33).

Interestingly, smokers undergoing joint replacement surgery require longer operating times and incur higher costs (34). Of particular relevance to our study, microvascular surgeons have used animal models to show that smoking delays wound healing and increases the risk of skin flap necrosis (35, 36). Although we did not study the effects of smoking on surgical complications, this would be an interesting area for future study.

In conclusion, we observed that cigarette smokers with SSc are significantly more likely than nonsmokers to require treatment for digital vascular disease. Previous studies also suggest that surgical complications may be increased in smokers. Current smokers had more severe vascular disease than did ex-smokers, which suggests that at least some of the effects are reversible. This result is compatible with data reported by Cherniack et al, who observed that the effects of smoking on vibration white finger (occupational Raynaud's phenomenon) were eliminated by cessation of smoking (37).

It is our clinical experience that many patients with severe SSc, despite advice to the contrary, continue to smoke. In this UK-based study, 21% of patients, including 33% of those with a previous amputation or debridement, are current smokers. We propose that further research in SSc pathogenesis should take into account the similarities with smoking-related vascular disease, and that resources be directed to supporting smoking cessation in this small but important group of patients.

Acknowledgements

We are grateful to Dr. Ronan O'Driscoll for his help with the smoking questionnaire.