The authors have declared that no competing interests exist.
Patients with multiple myeloma (MM) are at increased risk of arterial thrombosis. Our aim was to determine the risk factors, mechanisms and outcome of strokes in these patients.
We conducted a retrospective matched case–control study from our database of MM patients enrolled in Total Therapy (TT) 2, TT3a and TT3b protocols who developed a vascular event (transient ischemic attack, ischemic stroke, or intracerebral hemorrhage) from October 1998 to January 2014. Cases were matched for age-matched selected controls. Baseline demographics, risk factors, MM characteristics, laboratory values, and mortality of cases were compared to those of controls. Multivariate logistic regression analysis identified risk factors associated with stroke. Ischemic strokes (IS) were classified with modified Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria.
Of 1,148 patients, 46 developed a vascular event (ischemic stroke, 33; transient ischemic attack, 11; hypertensive intracerebral hemorrhage, 2). Multivariate logistic regression analysis determined renal insufficiency (odds Ratio, 3.528; 95% CI, 1.36–9.14; P = 0.0094) and MM Stages I and II (odds Ratio, 2.770, 95% CI, 1.31–5.81; p = 0.0073) were independent predictors of stroke. In our study, strokes attributable to hypercoagulability, atrial fibrillation and small-vessel occlusion were common mechanisms. After a stroke, 78% of patients were discharged to home or a rehabilitation facility and 4% to a long-term nursing facility; in-hospital mortality was 15%. Despite suffering a stroke no significant differences in survival were observed.
In our cohort of multiple myeloma patients, renal failure and MM Stages I and II had increased risk of stroke.
Multiple myeloma (MM) is the second most common hematological malignancy among adults in the United States [
In a large population-based study (18,627 MM patients, 70,991 matched controls) the hazard ratios of VTE at 1, 5, and 10 years were 7.5, 4.6, and 4.1, and the risk of cerebrovascular disease was 1.5, 1.2, and 1.2 respectively [
Our study was retrospective review of our database of MM patients enrolled in Total Therapy (TT) 2, TT3a, and TT3b protocols and managed at our academic tertiary care medical center from October 1998 to January 2014; the institutional review board of University of Arkansas for Medical Sciences approved the study. Details of the protocols and patient outcomes were previously reported [
Briefly, TT2 consisted of two arms of newly diagnosed MM patients who were ≤ 75 years old. At enrollment, patients were randomly assigned either to a control group (no thalidomide) or experimental group (thalidomide). Both arms consisted of multi-agent induction chemotherapy along with tandem autologous transplants, with the only difference in the inclusion or exclusion of thalidomide. Thalidomide doses were 400 mg daily during induction chemotherapy, 100 mg daily between transplantations, 200 mg daily with consolidation therapy, 100 mg daily during the first year of maintenance therapy, and then 50 mg on alternating days thereafter; the drug was given until relapse or adverse events occurred. Low-molecular-weight heparin was given prophylactically to all patients in the thalidomide group starting in July 2001.
Total Therapy 3 protocols were characterized by incorporation of bortezomib in addition multi-agent chemotherapy that consisted of thalidomide at induction and maintenance phases. TT3a and TT3b differed only in the maintenance phase; TT3a applied bortezomib, thalidomide and dexamethasone only in the first year and only thalidomide and dexamethasone thereafter, but TT3b applied bortezomib, lenalidomide and dexamethasone for all 3 years. Thus only the maintenance phase of TT3a and TT3b deferred with TT3a having thalidomide and TT3b with lenalidomide. Thalidomide dosage did not exceed 200 mg in any phase of the protocols, and titration of dosage to a minimum of 50 mg was permitted to ensure maximum compliance to the treatment. For lenalidomide, the maximum and minimum permitted dosages at the time were 15 mg and 5 mg, respectively. No routine prophylaxis for deep vein thrombosis was administered in the TT3 protocols. Of note, renal insufficiency was not an exclusion criterion for TT2 as long as it was of recent onset and due to Bence-Jones proteinuria or hypercalcemia, but patients were not eligible for TT3 protocols if creatinine levels were >3 mg/dl, regardless of the reason. We specifically selected patients enrolled in TT2 and TT3 protocols because wanted to study the effects of thalidomide and lenalinomide exclusively. This cut of makes it feasible to include patients with mild to moderate renal failure. Thus varying degrees of renal insufficiency was prevalent in all treatment protocols.
All cases (i.e., MM patients who had strokes) were compared with age-matched controls (i.e., MM patients who did not have strokes). Baseline demographics (age, gender, race), risk factors (hypertension, hyperlipidemia, diabetes mellitus, coronary artery disease, congestive heart failure, atrial fibrillation, obstructive sleep apnea, smoking, alcohol use, other malignancies, renal insufficiency, need for hemodialysis, prior VTE), MM characteristics (protocol, stage, isotype, risk category), laboratory values (complete blood count, coagulation profile, fasting blood glucose, β2 microglobulin, blood urea nitrogen, creatinine, albumin), and mortality were compared between both groups. Additionally, among the cohort of stroke cases, baseline laboratory values at the initial visit were compared with those during the vascular insult.
Transient ischemic attack was defined as rapidly developing symptoms and signs of loss of cerebral function of vascular origin with a negative diffusion-weighted imaging (DWI) sequence on magnetic resonance imaging (MRI) [
The disposition at discharge and recurrence of vascular event upon follow up were assessed in the cohort of stroke cases (i.e., MM patients who had strokes). In addition, all-cause mortality and survival upon follow up was compared between both groups.
Baseline characteristics, risk factors, laboratory parameters, and outcomes of the two groups were compared with Pearson’s Chi-square test or Fisher’s exact test for categorical variables and student
During this study period, 1,148 patients with MM were enrolled; TT2 without thalidomide (TT2−Thal;
IS was the most common type of stroke (72%; 33 patients) followed by transient ischemic attack (24%; 11 patients), and hypertensive basal ganglia hemorrhage (4%; 2 patients). All patients diagnosed as TIA in our cohort had a negative DWI on MRI. The most common mechanisms of stroke in our cohort of MM patients was thrombotic in nature from hypercoaguable state (15%; 7 patients), followed by cardioembolism (13%; 6 patients), small vessel disease (13%; 6 patients), multiple other known mechanisms (watershed patterns of strokes from hypotension, sepsis) (15%; 7 patients), large-vessel disease (7%; 3 patients), and cryptogenic in (9%; 4 patients). Inter current systemic infections were observed in 28% (13 patients) of stroke patients but none of them had evidence of nonbacterial thrombotic endocarditis. Vascular events of ischemic etiology occurred despite prior use of antiplatelet and anticoagulation therapy (two patients in each group). After a vascular event, antiplatelet agents were administered to 16 patients and anticoagulants to 7 patients; however 23 patients were ineligible for both antiplatelet and anticoagulant agents due to thrombocytopenia.
Demographics, risk factors, and baseline laboratory values are described in Tables
Stroke (N = 46) |
No Stroke (N = 138) |
||
---|---|---|---|
Age, years | 60.6 (7.7) | 60.7 (7.8) | 0.8960 |
Women | 23 (50.0) | 57 (41.6) | 0.3023 |
Race, Caucasian | 44 (95.7) | 125 (90.6) | 0.3635 |
Hypertension | 25 (54.4) | 60 (43.5) | 0.2003 |
Hyperlipidemia | 15 (33.3) | 36 (26.1) | 0.3464 |
Diabetes mellitus | 8 (17.4) | 13 (9.4) | 0.1409 |
Coronary artery disease | 5 (10.9) | 15 (10.9) | >0.99 |
Congestive heart failure | 2 (4.4) | 12 (8.7) | 0.5237 |
Atrial fibrillation | 8 (17.4) | 13 (9.4) | 0.1409 |
Transient ischemic attack | 2 (4.4) | 0 | 0.0615 |
Obstructive sleep apnea | 0 | 2.2 (3) | 0.5743 |
Smoking | 12 (26.1) | 18 (13.0) | 0.0381 |
Alcohol abuse | 1 (2.2) | 2 (1.5) | >0.99 |
Other malignancy | 4 (8.7) | 22 (16.1) | 0.2159 |
Renal insufficiency | 11 (23.9) | 11 (8.0) | 0.0039 |
Hemodialysis | 5 (10.9) | 1 (0.7) | 0.0040 |
Prior deep vein thrombosis | 13 (28.9) | 27 (19.6) | 0.1888 |
Protocol | |||
T2 TT2 with Thalidomide TT2 without Thalidomide TT3a TT3b |
25 (54.4) 17 (37.4) 8 (17.4) 15 (32.6) 6 (13.0) |
78 (55.6) 55 (39.9) 23 (16.7) 30 (21.7) 30 (21.7) |
0.3787 |
MM stage | |||
I II III |
11 (23.9) 8 (17.4) 27 (58.7) |
13 (9.4) 17 (12.3) 108 (78.3) |
0.0182 |
MM isotype | |||
IgG IgA FLC-κ FLC-λ Other |
27 (58.7) 10 (21.7) 4 (8.7) 5 (10.9) 0 |
68 (49.3) 34 (24.6) 14 (10.1) 15 (10.9) 7 (5.1) |
0.6128 |
MM risk | –0.13 (0.61) | 0.09 (0.67) | 0.0941 |
Values are expressed as n (%), mean ± standard deviation as appropriate
*Hyperlipidemia and for Prior deep vein thrombosis, N = 45; for MM risk, N = 37
†Other malignancy, N = 137; for MM risk, N = 88
‡Fisher’s exact test
Stroke (N = 46) |
No Stroke (N = 138) |
||
---|---|---|---|
Hemoglobin, g/dl | 11.4 (1.8) | 11.4 (1.9) | 0.9135 |
Hematocrit, % | 34.3 (5.3) | 34.1 (5.7) | 0.8519 |
Platelets, 109/L | 258.7 (99.2) | 247.7 (101.7) | 0.5230 |
Prothrombin time, sec | 13.4 (1.5) | 13.4 (1.8) | 0.9240 |
APTT, sec | 30.5 (6.5) | 28.7 (4.5) | 0.0881 |
International normalized ratio | 1.07 (0.12) | 1.09 (0.15) | 0.5321 |
Serum glucose, mg/dl | 105.5 (25.8) | 107.2 (30.9) | 0.7359 |
Beta 2-microglobulin, mg/L | 5.2 (7.0) | 4.6 (3.8) | 0.5783 |
Fibrinogen, mg/dl | 428 (159) | 406 (136) | 0.4534 |
Creatinine, mg/dl | 1.3 (1.0) | 1.2 (1.0) | 0.5855 |
Lactate dehydrogenase, U/L | 166.8 (47.8) | 170.0 (75.9) | 0.7627 |
Serum albumin g/dl | 3.93 (0.52) | 3.95 (0.61) | 0.7797 |
White blood cell count, 103/dl | 5.9 (2.4) | 6.1 (3.1) | 0.7138 |
Values represented as n (%), mean ± standard deviation;
APTT, activated partial thromboplastin time
*Fibrinogen, N = 36
†Beta 2-microglobulin, N = 137; Fibrinogen, N = 63
Unadjusted | Adjusted | ||||
---|---|---|---|---|---|
Odds Ratio | 95% CI | Odds Ratio | 95% CI | ||
Smoking | 2.35 | 1.03–5.36 | 2.324 | 0.98–5.54 | 0.0572 |
Nephropathy | 3.63 | 1.45–9.07 | 3.528 | 1.36–9.14 | 0.0094 |
MM Stage I and II | 2.53 | 1.24–5.17 | 2.77 | 1.31–5.81 | 0.0073 |
In our cohort of 46 cases, 78% (36 patients) were discharged home or to a rehabilitation facility and 4% (2 patients) to a long-term nursing facility. In-hospital mortality was 15% (7 patients), with stroke-related mortality in 13% (6 patients). The two groups did not differ significantly in median overall survival (121 vs 137 months,
Patients who were still alive were censored at the date of last contact. Survival curves were estimated using the method of Kaplan-Meier.
Although baseline laboratory values (
Variable | N | Baseline | Stroke | |
---|---|---|---|---|
Hemoglobin, g/dl | 35 | 11.4 (1.9) | 11.3(1.8) | 0.7906 |
Hematocrit, % | 35 | 34.5 (5.6) | 34.2 (5.5) | 0.8138 |
Platelets, 109/L | 35 | 255.2 (104.2) | 112.6 (78.6) | <0.0001 |
Prothrombin time, sec | 30 | 13.5 (1.5) | 14.9 (2.8) | 0.0145 |
APTT, sec | 28 | 31.1 (5.9) | 32.2 (8.4) | 0.4959 |
International normalized ratio | 30 | 1.08 (0.12) | 1.25 (0.33) | 0.0096 |
Serum glucose, mg/dl | 30 | 108.4 (28.4) | 113.9 (30.5) | 0.4394 |
Values represented as n (%), mean ± standard deviations as appropriate
APTT, activated partial thromboplastin time
Observations with missing data were dropped in Table 4. This was done on a variable-by-variable basis, which is why the sample size changes depending on the variable being analyzed.
This large single-center retrospective, comparative case−control study of MM patients enrolled in TT2, TT3a, and TT3b protocols over 10 years, identified 4% of patients who developed an acute stroke. Only 4% of the strokes were of the hemorrhagic type, which is different from the general population (10%) and patients with cancer (14–18%) [
The major risk factors associated with stroke in our study were renal insufficiency and Stage I or II MM, with a trend for smoking, which is a novel finding when compared to prior results, indicating that hypertension and smoking were significant risk factors [
Increased risk of stroke was observed in patients with Stage I and II MM, but no such relationship was observed for various treatment protocols or thalidomide use. However, the relationship with thalidomide may be underpowered given the case-control nature of our study. No study has yet evaluated the risk of VTE and arterial thrombosis for patients with Stage 1 or II MM; differences have only been noted only for patients with newly diagnosed MM versus those with relapsed MM [
Management of strokes in patients with MM depends upon the etiology. Common mechanisms of stroke in patients with underlying malignancy include hypercoagulability and cardioembolism, among other mechanisms [
Some major drawbacks of our study include a retrospective review, referral bias rather than true population patterns, inclusion of a subset of MM patients enrolled in specific treatment protocols, aggressiveness of the myeloma during the index event, a small sample size, and a case−control analysis. Because of a small sample size, the results of our study may need to be interpreted with caution until these are validated by large studies. Although the stroke subtype was arrived based on the TOAST criteria, the lack of long-term cardiac monitoring evaluation may have possibly underestimated the percentage of cardioembolic strokes. Besides, certain patients had multiple factors in which case the imaging pattern and triggering factor were the deciding factors. Although nephropathy preceded the occurrence of stroke, the exact time frame and the degree of renal failure was unavailable because definitions varied over the entire decade. Finally, the etiology of renal disease may either be due to myeloma or non-myeloma related origin and could not be evaluated because of the retrospective nature of our analysis.
To conclude, in our cohort of multiple myeloma patients, renal failure, MM Stage I and II were significant risk factors associated with stroke, which is different when compared to the general population. Despite a high rate of in-hospital mortality, patients with MM who develop a vascular event have good outcomes at discharge and overall long-term survival. Besides myeloma- related thrombosis, cardioembolic causes and small-vessel occlusions were other possible stroke mechanisms. Unlike other malignancies, patients with MM had greater risk of stroke if they had renal disease; thus renal disease may serve as a marker to identify high-risk patients, but will need confirmation in large population-based studies. Adoption of various preventive measures, especially for patients with renal insufficiency, may reduce the risk of stroke in future.
(DOCX)
(DOCX)
The manuscript was edited by Peggy Brenner, PhD, ELS, in the Grants and Scientific Publications at the University of Arkansas for Medical Sciences.