Try a new search

Format these results:

Searched for:



Total Results:


Common iliac vein stenting for May-Thurner syndrome and subsequent pregnancy

Speranza, Giancarlo; Sadek, Mikel; Jacobowitz, Glenn
BACKGROUND:For women with left common iliac vein compression (ie, May-Thurner syndrome) who undergo venous stenting and subsequently become pregnant, concerns have been raised regarding a possible compromise of stent patency due to compression from the gravid uterus and the hypercoagulability induced by pregnancy. Only a small body of literature exists on this subject, and limited management guidelines are available. The present study was designed to evaluate the safety of iliac vein stenting for May-Thurner syndrome (MTS) with subsequent pregnancy. METHODS:Female patients who had undergone common iliac vein stenting at our center who were aged 18 to 45 years and had subsequently become pregnant were identified. A retrospective medical record review of eight eligible patients was conducted, recording the demographics, procedural characteristics, and anticoagulation strategies. The primary outcome evaluated was stent patency. RESULTS:All eight patients had undergone left common iliac vein stenting for MTS. A total of eight stents were placed, and all demonstrated duplex ultrasound patency throughout pregnancy and postpartum. Seven patients delivered healthy pregnancies, and one experienced a stillbirth. The clinical CEAP (clinical, etiologic, anatomic, pathophysiologic) class remained unchanged or improved from pregnancy to postpartum for all patients. The average age at stent placement was 31 ± 5 years, and the average interval from stent placement to pregnancy was 28 ± 19 months. One patient developed nonobstructive deep vein thrombosis (DVT) of the left femoral vein during pregnancy and was treated with therapeutic enoxaparin. The nonobstructive DVT did not compromise the iliac vein stent. Two patients received low-dose aspirin and prophylactic doses of enoxaparin, one for a history of DVT and factor V Leiden and one for a recent history of fertility treatment. The five remaining patients received no anticoagulation, three received low-dose aspirin, and two received no antiplatelet therapy. CONCLUSIONS:Common iliac vein stent patency was not compromised by subsequent pregnancy in our eight patients with MTS. Furthermore, the stents remained patent throughout pregnancy in patients receiving a wide range of anticoagulation and antiplatelet treatments, suggesting that no uniform therapeutic threshold exists and treatment should be individualized. For most patients, low-dose aspirin alone or no treatment was adequate. This could have implications for counseling women who require intervention for MTS and are of child-bearing age.
PMID: 34438090
ISSN: 2213-3348
CID: 5011602

Comparison of Outcomes for Open Popliteal Artery Aneurysm Repair Using Vein and Prosthetic Conduits

Chang, Heepeel; Veith, Frank J; Rockman, Caron B; Siracuse, Jeffrey J; Jacobowitz, Glenn R; Cayne, Neal S; Patel, Virendra I; Garg, Karan
BACKGROUND:Autologous vein is considered the preferred conduit for lower extremity bypass. There is, however, limited literature regarding conduit choice for open popliteal artery aneurysm (PAA) repair. We sought to compare outcomes of PAA repair using vein versus prosthetic conduits. METHODS:The Vascular Quality Initiative database (2003-2019) was queried for patients with PAAs undergoing elective conventional revascularization originating from the superficial femoral and popliteal arteries. Conduits were categorized as vein or prosthetic. Primary outcomes were primary graft patency, freedom from major adverse limb event (MALE) and MALE-free survival at 2-years. Kaplan-Meier method with log-rank tests was used for estimation and comparison of patency. RESULTS:A total of 1,146 limbs in 1,065 patients underwent elective open revascularization for popliteal artery aneurysm. Vein was used in 921 limbs (80%), and prosthetic in 225 (20%). Patients in the prosthetic cohort had a shorter procedure time, were older, and had a higher prevalence of COPD. Postoperatively, prosthetic patients were more likely to be started and maintained on anticoagulation without increased incidence of hematoma. There was no significant difference in the rate of surgical site infection (2% vs. 2%; P = .946). There was an increased tendency toward more symptomatic patients in the vein cohort although not statistically significant (49% vein vs. 41% prosthetic; P = .096). On a mean follow-up of 13 ± 5 months, the incidence of MALE and MALE-free survival were comparable between the two groups. The 2-year primary and secondary patency rates were similar, 87% and 96% in the vein, and 91% and 95% in the prosthetic groups, respectively. At multivariable analysis, outflow bypass targets to the infrapopliteal arteries (HR 2.05; 95% confidence interval (CI), 1.16-3.65; P = .014) and symptomatic aneurysm (HR 1.81; 95% CI, 1.04-3.15; P = .037) were independently associated with loss of primary patency. Conduit type did not make a difference in MALE-free survival, or primary graft patency at 2-years. CONCLUSIONS:Our study demonstrates that conventional open PAA repair with prosthetic conduit yields results comparable to those with vein conduit with regard to primary and secondary patency and major adverse limb events at 2-years for targets to the popliteal artery. However, when the distal target was infrapopliteal, worse outcomes were observed with prosthetic conduit. Our results suggest that vein conduit should be preferentially used for infrapopliteal targets, while prosthetic conduit confers comparable outcomes in a subset of patients who do not have suitable autologous vein conduits.
PMID: 33819593
ISSN: 1615-5947
CID: 4839012

The Presence of a Pathologic Perforator May be Predictive of Central Venous Pathology and Multilevel Disease in Severe Chronic Venous Insufficiency

Li, Chong; Nwachukwu, Chukwuma; Jacobowitz, Glenn R; Kabnick, Lowell S; Maldonado, Thomas S; Rockman, Caron B; Berland, Todd L; Sadek, Mikel
INTRODUCTION/BACKGROUND:The treatment of a refluxing perforator is indicated in the setting of severe chronic venous insufficiency, but there are limited data on the presence of multilevel disease in these patients. This study sought to evaluate whether the presence of a pathologic perforator is predictive of the presence of central venous pathology. METHODS:This study was a retrospective review of the institutional vascular quality initiative (VQI) database. Consecutive patient-limbs were identified who underwent intervention of refluxing perforators. The patients who underwent imaging, including MRI or CT (Group A) were compared to those who did not undergo imaging (Group B). The treated limbs in Group A were also compared to the contralateral limbs as an internal control. Anatomical findings on imaging were analyzed by two independent investigators. The primary outcome was the presence and degree of central venous stenosis as measured by an orthogonal diameter reduction of > 50% by axial imaging. Secondary outcomes included demographic and clinical differences between the two groups, frequency of central venous intervention, and duration of ulcer healing. Standard statistical analysis was performed. RESULTS:Ninety-three patient-limbs underwent treatment of a pathologic perforator, with 30 in Group A and 63 in Group B. The following demographic and clinical variables were higher in Group A compared to Group B: Male gender, BMI, deep venous thrombosis history, recent or active anticoagulation use, perforator diameter, Clinical Etiology Anatomy Pathophysiology class 4, 5 or 6, and Venous Clinical Severity Score. Radiographic analysis of Group A revealed concordance of a treated pathological perforator with an ipsilateral central venous stenosis in 53.3% of patients, and a higher frequency of common iliac vein stenosis (50% vs 21.4%, P = 0.024) and external iliac vein stenosis (20% vs 0%, P = 0.012) compared to the contralateral limbs. When separated by left or right limb, the left limbs exhibited a greater degree of common iliac vein stenosis as compared to the contralateral limbs (50.7±20.9% vs 16.3±16.5%, P < 0.001) as well as a greater frequency of >50% common iliac vein stenosis (46.7% vs 13.3%, P = 0.046). The right limbs exhibited a greater frequency of > 50% external iliac vein stenosis as compared to contralateral limbs (33.3% vs 0%, P = 0.022). CONCLUSIONS:This study suggests that patients with severe chronic venous insufficiency who undergo treatment for a pathologic perforator may have additional ipsilateral central venous pathology, supporting the presence of multilevel disease. Additional axial imaging might unmask central venous pathology and provide another option for treatment.
PMID: 34252577
ISSN: 2213-3348
CID: 4938302

Thoracic Endovascular Aortic Repair for Symptomatic Penetrating Aortic Ulcers and Intramural Hematomas is Associated with Poor Outcomes

Rokosh, Rae S; Rockman, Caron B; Patel, Virendra I; Milner, Ross; Osborne, Nicholas H; Cayne, Neal S; Jacobowitz, Glenn R; Garg, Karan
INTRODUCTION/BACKGROUND:The natural history of penetrating aortic ulcers (PAU) and intramural hematomas (IMH) of the aorta is not well described. While repair is warranted for rupture, unremitting chest pain or growth, there is no established threshold for treating incidental findings. Thoracic endovascular aortic repair (TEVAR) offers an attractive approach in treating these pathologies, however, peri-procedural and post-operative outcomes are not well defined. METHODS:Patients 18 or older identified in the VQI database who underwent TEVAR for PAU and/or IMH between 1/2011-2/2020 were included. We identified 1042 patients, of whom 809 had available follow-up data. Patient demographics and comorbidities were analyzed to identify risk factors for major adverse events (MAE), as well as postoperative and late mortality. RESULTS:The cohort was 54.8% female and 69.9% former smokers with a mean age of 71.1 years. Comorbidities were prevalent with 57.8% classified ASA IV; 89.8% had hypertension, 28.3% chronic obstructive pulmonary disease (COPD), 17.9% coronary artery disease, and 12.2% congestive heart failure (CHF). Patients were predominately symptomatic (74%) and 44.5% underwent non-elective repair. MAE incidence was 17%. Independent predictors of MAE were history of CAD, non-Caucasian race, emergent procedural indication, ruptured presentation, and deployment of two or more endografts. In-hospital mortality was 4.3%. Seventy-three percent of index hospitalization mortalities were treatment-related. Of 809 patients with follow-up (mean 25.1 months±19 months), all-cause mortality was 10.6%. Predictors of late mortality in follow-up included age greater than 70 years, ruptured presentation, and history of COPD and ESRD. Subset analysis comparing symptomatic (74%) vs. asymptomatic (26%) patients demonstrated the former were frequently female (58.2% vs. 45.3%, p<.001) with a higher incidence of MAE (20.6% vs. 6.9%, p<.001), notably higher in-hospital reintervention rates (5.9% vs. 1.5%, p=.002) and mortality (5.6% vs. 0.7%, log-rank p=.015), and prolonged length of stay (6.9 vs. 3.7 days, p<.0001) despite similar procedural risks. In follow-up, late mortality was higher in the symptomatic cohort (12.2% vs. 6.5%, log-rank p=.025), with all treatment-related mortalities limited to the symptomatic group. CONCLUSIONS:We demonstrate significantly higher morbidity and mortality in symptomatic patients undergoing repair compared to asymptomatic patients, despite similar baseline characteristics. Asymptomatic patients treated with TEVAR had no treatment-related mortality in follow-up, with overall prognosis largely dependent on pre-existing comorbidities. These findings, in conjunction with growing evidence highlighting the risk of disease progression and attendant morbidity associated with these aortic entities, suggest a need for natural history studies and definitive guidelines on the elective repair of IMH and PAU.
PMID: 33340703
ISSN: 1097-6809
CID: 4725982

Outcomes Of Translumbar Embolization Of Type II Endoleaks Following Endovascular Abdominal Aortic Aneurysm Repair

Charitable, John F; Patalano, Peter I; Garg, Karan; Maldonado, Thomas S; Jacobowitz, Glenn R; Rockman, Caron B; Veith, Frank J; Cayne, Neal S
OBJECTIVES/OBJECTIVE:Presence of an endoleak can compromise aneurysm exclusion after endovascular abdominal aortic aneurysm repair(EVAR). Type II endoleaks(T2Es) are most common and may cause sac expansion. We report outcomes of translumbar embolization(TLE) of T2Es following EVAR. METHODS:We conducted a retrospective chart review of patients with T2E after EVAR treated with TLE from 2011-2018 at a single academic institution. Treatment indications were the presence of persistent T2E and aneurysm growth ≥5mm. Sac stabilization was defined as growth ≤5mm throughout the follow-up period. RESULTS:Thirty consecutive patients were identified. The majority were men (n=24) with a mean age of 74.3 years (70.9-77.6, 95% CI). The most common comorbidities were hypertension (83.3%) and coronary artery disease (54.0%). The mean maximal sac diameter at T2E discovery was 5.8 cm (5.4-6.2, 95% CI). The mean time to intervention from endoleak discovery was 33.7±28 months with a mean growth of 0.84 cm (0.48-1.2, 95% CI) during that time period. The mean follow-up time after TLE was 19.1 months (11.1-27.2, 95% CI). Twenty-eight patients were treated with cyanoacrylate glue(CyG) alone, and 2 were treated with CyG plus coil embolization(CE). There was immediate complete endoleak resolution as assessed intraoperatively, and sac stabilization in 15 cases (50.0%). Eleven (36.7%) patients had evidence of persistent T2E on initial imaging after the embolization procedure; additional follow-up revealed eventual sac stabilization at a mean of 21.3±7.2 months and therefore these patients did not require further intervention. In the remaining four cases (13.3%) there was persistent T2E after the initial TLE requiring a second intervention. Repeat TLE stabilized growth in three of these four patients after a mean of 17.6±12.9 months. One patient required open sacotomy and ligation of lumbar vessels due to continued persistence of the T2E and continued aneurysm growth. There were no ischemic complications related to the embolization procedures. Factors associated with persistent endoleak after initial embolization were: larger aneurysm diameter at the time of initial endoleak identification (p<0.001), and the use of antiplatelet agents (p<0.02). The use of anticoagulation was not a significant risk factor for endoleak recurrence or aneurysm growth after TLE. CONCLUSIONS:TLE of T2E is a safe and effective treatment option for T2E with aneurysm growth following EVAR. Patients taking antiplatelet medication and those with larger aneurysms at the time of endoleak identification appear to be at increased risk for persistent endoleak and need for subsequent procedures following initial TLE. These patients may require more intensive monitoring and follow-up.
PMID: 34197948
ISSN: 1097-6809
CID: 4926902

Anticoagulation and Antiplatelet Medications Do Not Affect Aortic Remodeling after Thoracic Endovascular Aortic Repair for Type B Aortic Dissection

Chang, Heepeel; Rockman, Caron B; Cayne, Neal S; Veith, Frank J; Jacobowitz, Glenn R; Siracuse, Jeffrey J; Patel, Virendra I; Garg, Karan
OBJECTIVE:There is a lack of evidence regarding the effect of anticoagulation and antiplatelet medications on aortic remodeling for aortic dissection after endovascular repair. We investigated whether anticoagulation and antiplatelet medications affect aortic remodeling after thoracic endovascular aortic repair (TEVAR) for Type B aortic dissection (TBAD). METHODS:Records of the Vascular Quality Initiative TEVAR registry (2012-2020) were reviewed. Procedures performed for TBAD were included. Aortic reintervention, false lumen thrombosis of the treated aorta and all-cause mortality at follow-up were compared between patients treated with and without anticoagulation medications. A secondary analysis was performed to assess the effect of antiplatelet therapy in patients not on anticoagulation. Cox proportional hazards models were used to estimate the effect of anticoagulation and antiplatelet therapies on outcomes. RESULTS:1,210 patients (mean age, 60.7±12.2 years; 825 (68%) males) were identified with a mean follow-up of 21.2±15.7 months (range 1-94 months). 166 (14%) patients were on anticoagulation medications at discharge and at follow-up. Patients on anticoagulation were more likely to be older (mean age, 65.5 vs 60 years; P<.001) and Caucasian (69% vs 55%; P=.003), with higher proportions of coronary artery disease (10% vs 3%; P<.001), congestive heart failure (10% vs 2%; P<.001) and chronic obstructive pulmonary disease (15% vs 9%; P=.017). There were no differences in the mean preoperative thoracic aortic diameter or the number of endografts used. At 18-month, the rates of aortic reinterventions (8% vs 9% log-rank P=.873), complete false lumen thrombosis (52% vs 45%; P=.175) and mortality (2.5% vs 2.7%; P=.209) were similar in patients with and without anticoagulation, respectively. Controlling for covariates with the Cox regression method, anticoagulation use was not independently associated with a decreased rates of complete false lumen thrombosis (hazard ratio (HR) 0.74; 95% confidence interval (CI), 0.5-1.1; P=.132), increased need for aortic reinterventions (HR 1.02; 95% CI, 0.62-1.68; P=.934), and mortality (HR 1.25; 95% CI, 0.64-2.47; P=.514). On a secondary analysis, antiplatelet medications did not affect the rates of aortic reintervention, complete false lumen thrombosis and mortality. CONCLUSIONS:Anticoagulation and antiplatelet medications do not appear to negatively influence the midterm endpoints of aortic reintervention or death in patients undergoing TEVAR for TBAD. Moreover, it did not impair complete false lumen thrombosis. Anticoagulation and antiplatelet medications do not adversely affect aortic remodeling and survival in this population at midterm.
PMID: 34182028
ISSN: 1097-6809
CID: 4926292

An algorithm combining VVSYmQ® and VCSS scores may help to predict disease severity in C2 patients

Sadek, Mikel; Pergamo, Matthew; Almeida, Jose I; Jacobowitz, Glenn R; Kabnick, Lowell S
OBJECTIVES/OBJECTIVE:The purpose was to assess whether combining patient reported scores (VVSymQ®) and physician reported scores (VCSS) stratifies disease severity in C2 patients. METHODS:Consecutive patients were pooled from the VANISH-1 and VANISH-2 cohorts. VCSS and VVSymQ® were calculated for each patient. The relationship between scoring systems was evaluated using Pearson's correlation and frequency distribution analysis. RESULTS:Two-hundred and ten C2 limbs were included. Scoring systems demonstrated: VVSymQ®: mean = 8.72; VCSS: mean = 6.32; correlation (r = 0.22, p = 0.05). Frequency distribution analysis demonstrated 61.4% of patients had low VVSymQ® and low VCSS; 31.3% had elevated VVSymQ® and increased VCSS; 7.3% were inconsistent with C2 disease. Strict concordance analysis revealed 40.5% had VVSymQ® (< 9)/VCSS (0-6), 18.6% had VVSymQ® (≥ 9)/VCSS (7-9), and 2.9% had VVSymQ® (≥9)/VCSS (≥10). CONCLUSIONS:For combined elevated VVSymQ® and VCSS, moderate/severe disease is corroborated, and intervention may be indicated. For combined lower scores, the disease severity is mild and conservative therapy is more appropriate.
PMID: 34121506
ISSN: 1758-1125
CID: 4907182

High incidence of patients lost to follow-up after venous thromboembolism diagnosis - Identifying an unmet need for targeted transition of care

Rokosh, Rae S; Grazi, Jack H; Ruohoniemi, David; Yuriditsky, Eugene; Horowitz, James; Sista, Akhilesh K; Jacobowitz, Glenn R; Rockman, Caron; Maldonado, Thomas S
OBJECTIVES/OBJECTIVE:Venous thromboembolism, including deep venous thrombosis and pulmonary embolism, is a major source of morbidity, mortality, and healthcare utilization. Given the prevalence of venous thromboembolism and its associated mortality, our study sought to identify factors associated with loss to follow-up in venous thromboembolism patients. METHODS:-value of <0.05 set for significance. RESULTS: = 0.03) as opposed to home. CONCLUSIONS:Our study demonstrates that over one-third of patients diagnosed with venous thromboembolism at our institution are lost to venous thromboembolism-specific follow-up, particularly those discharged to a facility. Our work suggests that significant improvement could be achieved by establishing a pathway for the targeted transition of care to a venous thromboembolism-specific follow-up clinic.
PMID: 34080914
ISSN: 1708-539x
CID: 4891752

Endovascular Treatment of Popliteal Artery Aneurysms Has Comparable Long-Term Outcomes to Open Repair with Shorter Length of Stay

Shah, Noor G; Rokosh, Rae S; Garg, Karan; Safran, Brent; Rockman, Caron B; Maldonado, Thomas S; Sadek, Mikel; Lamparello, Patrick; Jacobowitz, Glenn R; Barfield, Michael E; Veith, Frank; Cayne, Neal S
OBJECTIVE:Over the past two decades, the treatment of popliteal artery aneurysms (PAAs) has undergone a transformation. While open surgical repair (OR) remains the gold standard for treatment, endovascular repair (ER) has become an attractive alternative in select patient populations. The objective of this study was to compare the outcomes of open versus endovascular repair of PAAs at a single institution. METHODS:We performed a retrospective chart review of all patients between 1998 and 2017 who underwent repair for PAA. Patient baseline, anatomic, and operative characteristics as well as outcomes were compared between the open and endovascular cohorts. The intervention and treatment were at the discretion of the surgeon. RESULTS:Between 1998 and 2017, a total of 64 patients underwent repair of 73 PAAs at our tertiary care center. Twenty-nine patients with 33 PAAs underwent OR, and 35 patients with 40 PAAs underwent ER. When comparing the 2 cohorts, there were no statistically significant differences in demographic characteristics such as age, gender, or number of run-off vessels. There were significantly more patients in the ER group (21/53%) than the OR group (7/21%) with hyperlipidemia (p=.008) and a prior carotid intervention (6% vs. 0%, p=.029). Overall, the presence of symptoms was similar amongst the two groups; however, patients in the OR group had a significantly higher number of patients presenting with acute ischemia (p=.01). Length of stay (LOS) was significantly shorter in the ER cohort (mean 1.8 days [1-11]) compared to the OR group (5.4 days [2-13]) (p<.0001). There was no significant difference in primary or secondary patency rates between the two groups. In the ER group, good runoff (≥2 vessels) was a positive predictor for primary patency at 1 year (3.36 [1.0-11.25]), however, it was not in the OR group. Post-operative single and/or dual anti-platelet therapy did not affect primary patency in either cohort. CONCLUSIONS:The results of our study demonstrate that ER of PAAs is a safe and durable option with comparable patency rates to OR and a decreased LOS, with good run-off being a positive predictor for primary patency in the ER cohort.
PMID: 33957229
ISSN: 1097-6809
CID: 4866682

Contemporary Outcomes of Endovascular Abdominal Aortic Aneurysm Repair in Patients Deemed Unfit for Open Surgical Repair

Chang, Heepeel; Rockman, Caron B; Jacobowitz, Glenn R; Ramkhelawon, Bhama; Cayne, Neal S; Veith, Frank J; Patel, Virenda I; Garg, Karan
OBJECTIVE:Endovascular abdominal aortic aneurysm repair (EVAR) is preferred to open surgical repair (OSR) for the treatment of abdominal aortic aneurysm (AAA) in high-risk patients. We sought to compare perioperative and long-term outcomes for EVAR in patients designated as unfit for OSR using a large national dataset. METHODS:The Vascular Quality Initiative database collected from 2013 to 2019 was queried for patients undergoing elective EVARs for AAA > 5cm. The patients were stratified into two cohorts based on the suitability for OSR (FIT vs. UNFIT). Primary outcomes included perioperative (in-hospital) major adverse events, perioperative mortality, and mortality at 1 and 5 years. Patient demographics and postoperative outcomes were analyzed to identify predictors of perioperative and long-term mortality. RESULTS:Of 16,183 EVARs, 1,782 patients were deemed unfit for OSR. The UNFIT cohort was more likely to be older and female, with higher proportions of HTN, CAD, CHF, COPD, and larger aneurysm diameter. Postoperatively, the UNFIT cohort was more likely to have cardiopulmonary complications (6.5% vs. 3%; P<.001), with higher perioperative mortality (1.7% vs. 0.6%; P<.001) and 1 and 5-year mortality (13% and 29% UNFIT vs. 5% and 14% FIT; P<.001). Subgroup analysis within the UNFIT cohort revealed those deemed unfit due to hostile abdomen had significantly lower 1 and 5-year mortality (6% and 20%) compared to those unfit due to cardiopulmonary compromise and frailty (14% and 30%; P=.451). Reintervention-free survival at 1 and 5-years was significantly higher in the FIT cohort (93% and 82%) as compared to the UNFIT cohort (85% and 68%; P<.001). Designation as unfit for OSR was an independent predictor of both perioperative (OR 1.59; 95% CI, 1.03-2.46; P=.038) and long-term mortality (HR 1.92; 95% CI, 1.69-2.17; P<.001). Advanced age (OR 2.91; 95% CI, 1.28-6.66; P=.011) was the strongest determinant of perioperative mortality while ESRD (HR 2.51; 95% CI, 1.78-3.55; P<.001) was the strongest predictor of long-term mortality. Statin (HR 0.77; 95% CI, 0.69-0.87; P<.001) and ACE inhibitor (HR 0.83; 95% CI, 0.75-0.93; P<.001) were protective of long-term mortality. CONCLUSION/CONCLUSIONS:Despite low perioperative mortality, long-term mortality of those designated by the operating surgeons as unfit for OSR was rather high in patients undergoing elective EVARs, likely due to the competing risk of death from their medical frailty. Unfit designation due to hostile abdomen did not confer any additional risks after EVAR. Judicious estimation of the patient's life expectancy is essential when considering treatment options in this subset of patients deemed unfit for OSR.
PMID: 33035595
ISSN: 1097-6809
CID: 4627322