Digital ischemia: Etiologies and long-term follow up. A cohort study of 323 patients

Abstract
Aim: Upper Extremity Digital ischemia (UEDI) is a rare heterogeneous condition with digital upper limb ischemia occurring forty times less frequently than toe ischemia. Using the largest cohort ever published, the aim of this study was to evaluate etiologies, treatments, prognosis and long-term clinical outcomes of DI.
Method: Patients included had UEDI with or without cutaneous necrosis. They were followed in a university hospital setting between January 2000 to December 2016. Etiologies, recurrence of UEDI, digital amputation and overall survival were analyzed retrospectively.
Results: DI due to cardio-embolic disease (DICE) was the highest occurring etiology with 59 patients (18.3%), followed by UEDI due to Systemic Sclerosis (SSc) (16.1%), idiopathic causes (11.7%), Thromoboangitis obliterans (TAO) (9.3%), iatrogenic causes (9.3%) and cancer (6.2%). DICE patients tended to fall into an older age bracket and featured a greater numbers of cases with hypertension whereas TAO patients smoked more tobacco and cannabis. DICE presented a greater number of cases with paleness and neuropathic and nociceptive pain whereas SSc patients presented more cases with necrotic pathophysiology. Long-term follow-up of this study showed recurrences were significantly more frequent in SSc than in all others tested groups (p <0.0001 vs idiopathic and DICE, p=0.003 vs TAO) and that TAO patients had significantly more recurrences than DICE (p = 0.005). Global survival was significantly lesser in DICE than in SSc (p<0.0001), idiopathic (p=0.0093) and TAO (p =0.0007). 
Conclusion: The main etiologies of UEDI were CE, SSc and idiopathic. This study highlighted an increase of iatrogenic UEDI cases compared to previous studies. UEDI associated with SSc had a poor local prognosis and DICE presented poor general prognosis. UEDI with SSc and TAO presented with greater recurrence.


Introduction
Upper extremity digital ischemia (UEDI) is a persistent peripheral vascular disease.
Diagnosis remains a real challenge to identify and consequently evaluation of etiologies and determination of therapeutic strategies has proved difficult. The incidence of UEDI is about 2 cases per 100,000 persons per year (1,2) with symptomatic ischemia of the upper extremity being encountered less frequently than lower extremity vascular insufficiency.
The main causes generally identified in literature are cardiac or arterial embolism, local thrombosis, systemic autoimmune connective tissue diseases (especially systemic sclerosis), vasculitis or traumatic injury (3).
Past literature shows that only a few studies have focused on the prevalence of etiologies, in patients diagnosed with digital ischemia, with long-term clinical outcomes thus being rarely discussed. Only two retrospective studies about UEDI, with a cohort of over one hundred patients, were reported (2,4,5).
Moreover, there are only few studies published about long-term follow up of UEDI. Three of them were among specific patients : cancer-induced UEDI (4), hypothenar hammer syndrome (6), systemic sclerosis (7)). Only two studies were about general follow up of UEDI, both with a limited number of cases (8,9).
The purpose of this retrospective study was to evaluate etiologies in a large cohort and to assess long-term clinical outcomes of patients diagnosed with digital ischemia. [Texte]

Material and Methods
In this retrospective cohort study, we included patients hospitalized for UEDI and outpatients visiting for UEDI in a university hospital between January 1, 2000 and December 31, 2016. We identified patients, based on IDC-10 (I742, T345 and I744 codes) through the medical data processing department. This study was approved by the Local Ethics Review committee. UEDI inclusion criterion was the first episode of UEDI diagnosed, defined by: painful and digital blanching, cyanosis, ulceration of the fingers, gangrene or change of cutaneous temperature, due to vascular pathology. We include all UEDI cases, not only those limited to fingers.
Clinical assessment included demographic data (age, gender, occupation), medical history, cardiovascular risk factors (CVRF), substance abuse, number of recurrence, long-term clinical outcomes and cardiovascular events (stroke, myocardial ischemia, lower limbs ischemia).
Each etiology was determined as follow: Systemic Sclerosis cases (SSc) were diagnosed through ACR/EULAR 2013's criteria (10); thromboangiitis obliterans cases (TAO) by Mills 2003's criteria (11); cardio-embolic (CE) diseases were diagnosed through ultrasound criteria and Holter monitoring; vasculitis cases were diagnosed by histology or by cryoglobulinemia positivity or ANCA positivity and further with giant cell arteritis and Takayasu's Arteritis by ACR or Ishikawa's criteria; thrombophilias were determined through specific tests; cancer were diagnosed by histology and molecular biology; iatrogenic etiology was defined by intrinsic and extrinsic probability for drug-induced iatrogenic ischemia and by corresponding temporality, for technical-induced iatrogenic ischemia ( such as artery cannulation) and finally hypothenar hammer syndrome was identified with the help of occupational health service. When criteria were not found, diagnosis was made following intrinsic and extrinsic causalities. Idiopathic ischemia was defined as such after negative clinical, laboratory and imaging examination. When several etiologies could be incriminated, we chose to include the UEDI case in the etiology most susceptible to be responsible for.
Long-term follow-up was performed by chronological analysis of medical records and by regimenting telephone interviews with patient's general practitioners.

Statistical analysis
Results were expressed as mean ± standard deviation (SD) or as median, range. Categorical variable for the groups were compared using chi-square tests or Fisher's exact tests when any of the expected cell counts of a 2x2 table was less than 5. Comparisons of quantitative variables were performed using Student's t-test. ANOVA test was performed with a Bonferroni correction after test to individualize group differences. A p-value < 0.05 was considered statistically significant. Survival [Texte] curves were developed to compare cardio-embolic, SSc, idiopathic and TAO groups. Only new events were considered. Survival free of digital ischemia, cardiovascular event (stroke, acute myocardial infarction or acute lower limb arteritis), fingers amputation and global survey were compared. Kaplan-Meier curves were then produced and a log-rank test was performed to analyze the event-free survival. All calculations were done using Graph Pad Prism 6.

Results
Patients' characteristics 323 patients were analyzed. The etiologies of digital ischemia are presented in Table 1 Etiology of digital    Successful surgical revascularizations were significantly more often than all other groups (p< 0.0001).
Idiopathic group had also significantly more surgical revascularization than SSc group (p=0.002) and TAO group (p =0.02).  Frequencies of digital amputations and auto-amputations varied, depending on etiologies ( Figure   2B). The cumulated rate of digital amputation and auto-amputation was higher in SSc group (n=18) (p=0.02) and TAO (n=7) (p=0.03) than in DICE group (n=2). In terms of survival (free of amputation) features, there were no significant differences between the other groups. There were only 4 amputations and auto-amputations in the idiopathic group.

Discussion
This study on digital ischemia is the largest cohort ever reported with 323 patients encountered in tertiary center recruitment in vascular pathology. [Texte] At the end of the follow-up DICE was the highest-occurring etiology in this cohort, the second highest occurring was SSc. Idiopathic UEDI were the third highest occurring, TAO   Cancer-associated UEDI was found in 2 to 15% of cases. In this study, as in Le Besnerais et al (4), the most common histological type was adenocarcinoma. Myeloproliferative diseases were also frequent here, accounting for 30% of digital ischemia associated with cancer. The time between UEDI and cancer diagnosis varied, with it taking 2 months in a study with intensive initial investigation (4) and 14.3 months, in this study, after a clinical examination, standard blood test and clinical follow-up.

Limits
In this study, patients were included retrospectively in a large time-lapse of sixteen years and so there was data loss. We presented episodes of UEDI from tertiary center recruitment in vascular pathology, where prevalence of CTD and multiple comorbidities are higher. Because diagnosis of occupational pathologies is carried out upstream of our center, they are under-represented in our study while they represent up to 33% in some UEDI studies. [Texte]

Conclusion
This study describes the largest cohort of UEDI ever reported. For the first time, this study showed that cardio-embolic diseases are the highest occurring of the UEDI etiologies studied. This study also found that iatrogenic cause of UEDI was frequent and increased with the use of vasopressor drugs and multiplication of endoluminal arterial procedures. The long-term follow-up showed: UEDI associated with SSc had a poor local prognosis and DICE presented poor general prognosis. UEDI with SSc and TAO presented with more recurrence. When idiopathic UEDI shows recurrence, those etiologies have to be researched as a priority. This research is even more important because of the higher risk of clinical course towards amputations in those etiologies.