Journal of the Neurological Sciences 347 (2014) 205–209
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Early infarct growth predicts long-term clinical outcome in ischemic stroke Seung Min Kim a,b, Sun U. Kwon a, Jong S. Kim a, Dong-Wha Kang a,⁎ a b
Department of Neurology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea Department of Neurology, Veterans Health Service Medical Center, Seoul, South Korea
a r t i c l e
i n f o
Article history: Received 24 June 2014 Received in revised form 27 August 2014 Accepted 26 September 2014 Available online 2 October 2014 Keywords: Infarct volume Infarct growth Outcome Ischemic stroke
a b s t r a c t Background: Ischemic lesions dynamically evolve during the acute phase of stroke. Although the ischemic lesion volume has been considered as a predictor of clinical outcome, it is still controversial whether early changes in ischemic lesion have prognostic information in addition to clinical variables. We hypothesized that early infarct growth on diffusion-weighted imaging (DWI) might be independently associated with long-term outcome in acute ischemic stroke patients. Methods: This was a prospective study for acute ischemic stroke patients admitted to the Stroke Unit of Asan Medical Center. The patients underwent DWI at baseline (within 24 h) and subsequently at 5 days after stroke onset. Early infarct growth was defined as the absolute difference between follow-up and baseline infarct volumes. Poor outcome was a modified Rankin Scale (mRS) at 3 months of 2–6 or 3–6. The association between infarct growth on DWI and clinical outcome was explored using multivariate analysis adjusting for demographics, risk factors for stroke, and other clinical variables. The cut-off values of early infarct growth predicting long-term outcomes were estimated using receiver operating characteristic analysis. Results: Of 409 patients enrolled, 345 (84.4%) showed any infarct growth (median, 0.63 cm3; interquartile range [IQR], 0.11–6.33 cm3; mean ± standard deviation, 9.55 ± 25.54 cm3). At the 3-month follow-up, the good outcomes were observed in 217 patients (53.1%) for mRS 0–1 and 303 patients (74.1%) for mRS 0–2. The larger infarct growth was associated with poor clinical outcome (for mRS 2–6, 0.29 cm3 [IQR 0.04–2.19] vs. 2.16 cm3 [IQR 0.26–17.68], p b 0.001; and for mRS 3–6, 0.39 cm3 [IQR 0.05–3.25] vs. 7.36 cm3 [IQR 0.57–26.48], p b 0.001). After adjusting age, diabetes, baseline National Institutes of Health Stroke Scale, and baseline infarct volume by multivariate logistic regression analysis, infarct growth was an independent predictor of poor clinical outcomes (for mRS 2–6, odds ratio [OR], 1.03, 95% confidence interval [CI], 1.004–1.06, p = 0.03; and for mRS 3– 6, OR, 1.03, 95% CI, 1.01–1.05, p = 0.01). The cut-off values of infarct growth discriminating between good and poor outcomes were 0.99 cm3 for mRS 0–1 vs. 2–6 (area under curve, 0.685; P b 0.001) and 8.86 cm3 for mRS 0–2 vs. 3–6 (area under curve, 0.736; P b 0.001). Conclusions: Our present study findings show that infarct growth within a week of onset independently predicts 3-month clinical outcomes. This suggests that short-term changes in infarct volume may serve as a surrogate marker of long-term clinical outcomes after ischemic stroke. © 2014 Elsevier B.V. All rights reserved.
1. Introduction An accurate prognosis of acute ischemic stroke is important for establishing a long-term treatment plan. Moreover, prognostic information can be used to evaluate the efficacy of thrombolysis or neuroprotective drugs. Clinical variables such as age, severity, and stroke subtype have consistently been associated with clinical outcomes [1–3]. A simple, reliable and inexpensive imaging predictor may be useful if it can ⁎ Corresponding author at: Department of Neurology, Asan Medical Center, University of Ulsan College of Medicine, 88 Olympic-ro 43-gil, Songpa-gu, Seoul 138-736, South Korea. Tel.: +82 2 3010 3440; fax: +82 2 474 4691. E-mail address: [email protected] (D.-W. Kang).
http://dx.doi.org/10.1016/j.jns.2014.09.048 0022-510X/© 2014 Elsevier B.V. All rights reserved.
demonstrate long-term outcome or early treatment effect. In previous reports, recanalization of vessels and lesion volume were suggested as imaging predictors [4]. However, it is controversial whether the lesion volume at a single time point can predict the clinical outcome in a stroke patient [5–7]. Ischemic lesion size increases dynamically and substantially during the first few days after onset and slowly decreases during the subsequent 3 to 4 weeks [8–11]. Thus, a hyperacute lesion on initial diffusion-weighted imaging (DWI) does not indicate the final lesion size, and it is assumed that assessment of infarct growth is a more reliable predictor of clinical outcome than ischemic lesion size at a single time point. In this regard, previous studies have tried to investigate the association between infarct growth and various clinical outcomes,
206
S.M. Kim et al. / Journal of the Neurological Sciences 347 (2014) 205–209
but have been limited to a retrospective design, specific stroke subtype (subcortical infarction) or treatment (thrombolysis) group, or a relatively small number of patients [12–16]. It remains unclear whether early infarct growth has predictive value for long-term prognosis in a large number of patients in routine clinical practice. In our prospective study, we hypothesized that early infarct growth might be independently associated with long-term outcome in acute ischemic stroke patients and attempted to estimate a cut-off value of infarct growth for predicting a poor outcome. 2. Methods 2.1. Patients Our current study was a prospective investigation of acute ischemic stroke patients admitted to the Stroke Unit of the Asan Medical Center between December 2004 and December 2006. We screened consecutive patients who had (1) acute ischemic stroke confirmed by initial MRI including DWI and MR angiography within 24 h after onset, (2) a followup DWI scan performed at 5 (± 1) days after symptom onset and (3) clinical follow-up for 3 months. Patients who had contraindications to MRI were excluded. The onset time was defined as the time patients were last known to be without neurological symptoms. This study was approved by the institutional review board of Asan Medical Center, and each patient or legal guardian provided written informed consent to participate in the study. 2.2. Imaging protocol and analysis MRI examinations were performed using a 1.5-T MR imaging unit (Signa; GE Medical Systems, Milwaukee, WI) with echo-planar capabilities. The initial MRI protocol included DWI, fluid-attenuated inversion recovery imaging (FLAIR), gradient echo T2-weighted imaging (GRE), perfusion-weighted imaging, 3D time-of-flight MRA, and 3D contrastenhanced MRA. The follow-up MRI protocol at 5 (± 1) days after onset included DWI, GRE, and FLAIR in all patients. The detailed MRI protocol has been previously described [17]. The infarct volumes on DWI were measured by an investigator (S.M.K.) who was blind to the clinical data. The infarct volume was measured on DWI as the sum of the infarct area in each slice multiplied by the slice thickness with the use of the Picture Archiving and Communication System at the Asan Medical Center. Early infarct growth was defined as the absolute difference between the follow-up infarct volume and the baseline infarct volume.
by using a Student's t-test or Mann–Whitney U test. Categorical variables were analyzed by a chi-square test or Fisher's exact test. A multivariate logistic regression analysis considering all variables was then conducted to assess the independent association of the infarct growth with clinical outcome. Infarct growth was considered as a continuous variable in univariate and multivariate analyses. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. Finally, we estimated the cut-off value of infarct growth for predicting poor clinical outcome at 3 months using receiver operating characteristic (ROC) analysis. The cut-off value of infarct growth was regarded as the value that had a maximum sum of sensitivity and specificity. In addition, we further evaluated the association of early infarct growth and clinical outcome after excluding the patients with small-vessel occlusion. All statistical analyses were performed with SPSS 20.0 for Windows (IBM Corp., Armonk, NY) and P b 0.05 was considered statistically significant. 3. Results 3.1. Baseline characteristics and infarct growth During the study period, we screened 1351 patients with acute ischemic stroke. Of these patients, 942 were excluded (570 patients underwent initial DWI after 24 h, 96 patients did not provide study consent, 184 patients did not undergo follow-up DWI, 76 patients had a contraindication for the MRI, and 16 patients were excluded due to loss to follow-up at 3 months). Thus, 409 patients (253 men) were included in the final analysis, with a mean age (standard deviation) of 63.6 (12.0) years. The median NIHSS score was 5 (IQR, 2–8). Because lesion volumes were not normally distributed, we used median value for analysis and expressed both median and mean values in the tables. The median infarct volume was 1.00 cm3 (IQR, 0.32–7.40) at baseline and 2.29 cm3 (IQR, 0.61–17.07) at follow-up, resulting in a median infarct growth of 0.63 cm3 (IQR, 0.11–6.33). Some degree of infarct growth was observed in 345 patients (84.4%). In 61 patients (14.9%), the infarct was smaller at follow-up than at baseline. In trend analysis, the infarct growth became larger as the mRS score increased (Fig. 1; P b 0.001). 3.2. Predictors of poor clinical outcomes At the 3-month follow-up, 53.1% and 74.1% of patients had excellent and independent outcomes, respectively. Univariate analysis demonstrated that a larger baseline infarct volume and larger infarct growth
2.3. Clinical assessment Stroke severity at admission was measured by the National Institutes of Health Stroke Scale (NIHSS) score. Stroke subtypes were determined according to the classification of the Trial of the Org 10172 in Acute Stroke Treatment (TOAST) [18]. Clinical outcomes were rated according to the modified Rankin Scale (mRS) at 3 months and were assessed by a certified research coordinator who was blind to the imaging and other clinical data. Clinical outcomes were dichotomized as (1) excellent (mRS 0–1) or poor (mRS 2–6) and (2) independent (mRS 0–2) or poor (mRS 3–6). 2.4. Statistical analysis We analyzed the relationship between early infarct growth and clinical outcome at 3 months. We also compared demographics, risk factors for stroke, initial NIHSS scores, stroke subtypes, initial glucose levels, reperfusion therapy, and baseline infarct volumes between patients with good and poor outcomes according to the clinical outcome criteria. Continuous or numerical variables were expressed as the mean (standard deviation) or median (interquartile range [IQR]) and were compared
Fig. 1. Association between infarct growth and mRS at 3 months. mRS, modified Rankin Scale. Infarct growth is presented as mean (standard deviation).
S.M. Kim et al. / Journal of the Neurological Sciences 347 (2014) 205–209
207
Table 1 Baseline characteristics associated with clinical outcomes at 3 month.
Age (years) Male Hypertension Diabetes Hyperlipidemia Smoking Previous stroke Initial glucose Onset to initial DWI (hours) Time interval between two DWIs (days) Initial NIHSS Reperfusion therapy Baseline infarct volume (cm3) Median Mean Infarct growth (cm3) Median Mean Stroke subtype Small-vessel occlusion Large artery atherosclerosis Cardioembolism Others
Total (n = 409)
Excellent outcome (mRS 0–1) Yes (n = 217)
No (n = 192)
P
Independent outcome (mRS 0–2) Yes (n = 303)
No (n = 106)
P
63.6 ± 12.0 253 (61.9%) 312 (76.3%) 126 (30.8%) 182 (44.5%) 121 (29.6%) 101 (24.7%) 148.1 ± 59.1 10.8 ± 6.8 4.7 ± 1.0 5 (2–8) 49 (12.0%)
61.4 ± 11.9 141 (65.0%) 158 (72.8%) 53 (24.4%) 94 (43.3%) 68 (31.3%) 49 (22.6%) 142.4 ± 53.1 10.3 ± 6.7 4.5 ± 0.9 3 (1–5) 14 (6.5%)
65.4 ± 11.8 112 (58.3%) 154 (80.2%) 73 (38.0%) 88 (45.8%) 53 (27.6%) 52 (27.1%) 154.8 ± 65.5 11.3 ± 6.8 4.9 ± 1.1 7 (4–12.25) 35 (18.2%)
0.001 0.167 0.079 0.003 0.609 0.409 0.292 0.044 0.124 b0.001 b0.001 b0.001
61.4 ± 11.4 199 (65.7%) 225 (74.3%) 82 (27.1%) 126 (41.6%) 96 (31.7%) 72 (23.8%) 145.7 ± 59.8 10.6 ± 6.8 4.5 ± 0.9 3 (2–6) 28 (9.0%)
68.7 ± 12.2 54 (51.0%) 87 (82.0%) 44 (41.5%) 56 (52.8%) 25 (23.6%) 29 (27.4%) 155.6 ± 58.0 11.1 ± 6.6 4.9 ± 1.2 10 (6–14.75) 21 (19.8%)
b0.001 0.007 0.103 0.006 0.045 0.116 0.460 0.155 0.502 0.002 b0.001 0.004
1.00 (0.32–7.40) 9.29 ± 21.13
0.66 (0.24–3.45) 4.04 ± 9.33
2.16 (0.46–16.24) 15.20 ± 28.06
b0.001 b0.001
0.73 (0.27–4.35) 6.12 ± 15.74
3.23 (0.64–21.38) 18.27 ± 30.11
b0.001 b0.001
0.63 (0.11–6.33) 9.53 ± 25.63
0.29 (0.04–2.19) 2.55 ± 6.50
2.16 (0.26–17.68) 17.46 ± 35.03
b0.001 b0.001 0.042
0.39 (0.05–3.25) 3.91 ± 11.03
7.36 (0.57–26.48) 25.67 ± 42.79
b0.001 b0.001 0.027
110 122 101 76
69 (31.8%) 64 (29.5%) 45 (20.7%) 39 (18.0%)
41 (21.4%) 58 (30.2) 56 (29.2%) 37 (19.3%)
94 (31%) 88 (29%) 63 (21%) 58 (19%)
16 (15%) 34 (32%) 38 (36%) 18 (17%)
Data are expressed as a number (percent), mean ± standard deviation, or median (interquartile range). NIHSS, National Institutes of Health Stroke Scale.
were associated with poor outcome at 3 months. In addition, poor outcomes were associated with an older age (for mRS 2–6 and for mRS 3–6), higher initial NIHSS score (for mRS 2–6 and for mRS 3–6), diabetes (for mRS 2–6 and for mRS 3–6), reperfusion therapy (for mRS 2–6 and for mRS 3–6), time interval between the two DWI scans (for mRS 2–6 and for mRS 3–6), initial glucose level (for mRS 2–6), hyperlipidemia (for mRS 3–6), and stroke subtype (for mRS 2–6 and for mRS 3–6) (Table 1). Multivariate logistic regression analysis showed that a larger infarct growth, older age, higher initial NIHSS score (for mRS 2–6 and for mRS 3–6), diabetes and time interval between the two DWI scans (for mRS 2–6) were independent predictors of poor clinical outcomes. In contrast, baseline infarct volume was no longer significant upon multivariate analysis (Table 2). Excluding patients with small-vessel occlusion, larger infarct growth, older age and higher initial NIHSS were still found to be associated with poor clinical outcome (Supplemental Tables 1 and 2). 3.3. Cut-off values of early infarct growth predicting long-term clinical outcomes The cut-off values for infarct growth that discriminate between poor and good clinical outcomes were 0.99 cm3 for mRS 2–6 (area under curve [AUC], 0.685; P b 0.001) and 8.86 cm3 for mRS 3–6 (AUC, 0.736;
P b 0.001) (Fig. 2). Excluding patients with small-vessel occlusion, cutoff values were 8.85 cm3 for mRS 2–6 (AUC, 0.694; P b 0.001) and 10.10 cm3 for mRS 3–6 (AUC, 0.741; P b 0.001) (Supplemental Fig. 1). 4. Discussion In the present study, we demonstrated that early infarct growth in acute ischemic stroke is independently associated with long-term clinical outcome. In addition, the association was still significant even after exclusion of patients with small-vessel occlusion. Among clinical variables, age, initial NIHSS, diabetes and time interval between the two DWI scans were also identified as predictors of clinical outcome. We also estimated a cut-off value of early infarct growth for predicting poor outcome, which may better predict long-term prognosis. This study has some advantages over previous ones, by being the largest prospective study to date, by adjusting for most of the clinical variables, and by considering all treatment groups seen in routine practice to reflect the real clinical situation. Early infarct growth on MRI likely reflects various pathophysiological processes. It can reflect expansion of the cerebral infarction from the ischemic penumbra, shown by the increase in cytotoxic edema on DWI that occurs in the first few hours after stroke onset. In addition, infarct growth can reflect the formation of vasogenic edema, which plays
Table 2 Predictors of poor clinical outcomes by multivariate logistic regression analysis. Poor outcome (mRS 2–6) OR (95% CI) Male sex Age Diabetes Hyperlipidemia Time interval between two DWIs (per 1 day) NIHSS (per 1) Reperfusion therapy Baseline infarct volume (per 1 cm3) Infarct growth (per 1 cm3) Stroke subtype
Poor outcome (mRS 3–6) P
1.025 (1.004–1.047) 1.708 (1.028–2.836)
0.019 0.039
1.327 (1.031–1.709) 1.228 (1.140–1.323) 0.967 (0.398–2.345) 1.004 (0.983–1.026) 1.033 (1.003–1.064)
0.028 b0.001 0.940 0.691 0.029 0.539
OR, odds ratio; CI, confidence interval; NIHSS, National Institutes of Health Stroke Scale.
OR (95% CI)
P
0.684 (0.382–1.225) 1.059 (1.030–1.089) 1.677 (0.914–3.074) 1.603 (0.906–2.837) 1.318 (0.971–1.788) 1.223 (1.137–1.317) 0.560 (0.210–1.491) 0.989 (0.973–1.006) 1.031 (1.010–1.052)
0.201 b0.001 0.095 0.105 0.076 b0.001 0.245 0.213 0.003 0.887
208
S.M. Kim et al. / Journal of the Neurological Sciences 347 (2014) 205–209
Fig. 2. Cut-off values of early infarct growth from ROC curves. Poor outcomes were defined as mRS 2–6 (A) and mRS 3–6 (B).
an important role in the subsequent 2 to 3 days. These mechanisms can be affected by various clinical factors, such as vessel occlusion, collateral flow patterns, reperfusion, systemic blood pressure, and stroke etiology. Despite the concern that estimation of infarct growth on MRI might contain a measurement error due to the sensitivity of MRI, intra- and inter-rater concordances were excellent in a previous study [19]. Although infarct volume is considered to be a variable predicting clinical outcome, infarct volume measurement at a single time point does not reflect the true biology and dynamic nature of progressive ischemic injury because early cellular changes following ischemia can potentially be reversed and infarct growth can be influenced by reperfusion therapy or other neuroprotective drugs [20,21]. Hence, an image taken at a single time point, such as that of the final infarct volume or the initial infarct volume, may not be sufficient for evaluating the long-term prognosis or treatment effect. In this regard, infarct growth was used as the outcome variable in several stroke trials. However, infarct growth was previously defined as the difference between the initial DWI lesion and 90-day T2 or FLAIR lesions [13,22,23], which is inadequate for the evaluation of an early treatment effect. In addition, evaluation may not be completed due to follow-up loss or death. Because we evaluated the early infarct growth in the initial 5 days after stroke onset, our findings provide evidence that this growth can be used as a marker of early treatment effects in acute ischemic stroke. Moreover, because most patients are hospitalized for the first week, follow-up MRI at 5 days may be more practical than at 90 days after onset [8]. In addition to early infarct growth, our study showed that age, initial NIHSS, and diabetes were significantly correlated with long-term clinical outcome. These results are in agreement with those of previous studies [2,3,24–27]. Our present study thus provides clear evidence that early infarct growth has a predictive value for prognosis, even after adjustment for these clinical variables. Although reperfusion therapy was associated with poor outcome in univariate analysis, it did not reach significance in multivariate analysis. This result may be due to the baseline characteristics of patients who received reperfusion therapy. They had higher initial NIHSS scores, larger infarct growth, and a high proportion of cardioembolism compared to patients who did not receive reperfusion therapy (data not shown). More severe stroke patients tended to receive reperfusion therapy which resulted in poor clinical outcome, but the association between reperfusion therapy and poor outcome disappeared after adjusting for other clinical variables. We also investigated the cut-off values of early infarct growth for predicting poor clinical outcomes—defined as mRS 2–6 and mRS 3–6—
and measured them at 0.99 cm3 and 8.86 cm3, respectively. The cutoff value of infarct growth might be clinically useful when evaluating the effect of thrombolysis or neuroprotective treatment. The cut-off volumes and infarct growth are smaller than that of a previous study (approximately 15 cm3) that studied acute stroke patients with anterior circulation infarct (baseline lesion volume N 2 cm3) or treated with thrombolysis [11,15]. A plausible explanation for this disparity may be the difference in study population: this study included patients regardless of thrombolytic treatment, which resulted in the enrollment of a large number of patients with small-vessel occlusion. Our study had several limitations of note. First, we did not consider the perfusion status and recanalization of occluded vessels. Because infarct progression from the ischemic penumbra is one of the most important causes of infarct growth [9,22,28,29], this growth might reflect the size of the penumbra. However, PWI lesions also include the benign oligemic area and would thereby overestimate hypoperfused tissues [30]. Thus, a PWI–DWI mismatch may not accurately reflect the potentially salvageable ischemic tissue [31]. Second, because we assessed infarct growth using the absolute difference in volumes between initial and follow-up infarcts, there was a tendency for a larger infarct growth with a larger initial infarction. However, if infarct growth was calculated as a percentage change, a smaller infarct at baseline would give a greater increase in percentage change. Nevertheless, this limitation did not significantly affect our results because we included patients with all stroke subtypes and the result was significant after adjusting for initial lesion volume. Hence, the evaluation of infarct growth via the absolute difference in volume might be a more appropriate method. Despite these limitations, our study findings have significant clinical implications. Although the most important end point in acute stroke trials is clinical outcome, an ancillary outcome measure may be valuable, especially in phase 2 clinical trials. Because long-term clinical outcome can be affected by various clinical factors, including long-term treatment and rehabilitation, early infarct growth can be used as a good surrogate marker for clinical outcome focused on the acute treatment effect. In conclusion, our results demonstrate that early infarct growth independently predicts poor clinical outcomes in acute ischemic stroke patients. These findings support the belief that estimation of early infarct growth may be a good surrogate marker for clinical outcome in acute treatment trials. Conflict of interest statement The authors report no disclosures.
S.M. Kim et al. / Journal of the Neurological Sciences 347 (2014) 205–209
Acknowledgements This study was supported by grants from the Korea Health Technology R&D Project, Ministry for Health & Welfare, Republic of Korea (HI12C1847 and HI11C1531). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.jns.2014.09.048. References [1] Johnston KC, Connors Jr AF, Wagner DP, Knaus WA, Wang X, Haley Jr EC. A predictive risk model for outcomes of ischemic stroke. Stroke 2000;31:448–55. [2] Weimar C, Konig IR, Kraywinkel K, Ziegler A, Diener HC. German Stroke Study C. Age and National Institutes of Health Stroke Scale Score within 6 hours after onset are accurate predictors of outcome after cerebral ischemia: development and external validation of prognostic models. Stroke 2004;35:158–62. [3] Saposnik G, Guzik AK, Reeves M, Ovbiagele B, Johnston SC. Stroke Prognostication using Age and NIH Stroke Scale: SPAN-100. Neurology 2013;80:21–8. [4] Vogt G, Laage R, Shuaib A, Schneider A, Collaboration V. Initial lesion volume is an independent predictor of clinical stroke outcome at day 90: an analysis of the Virtual International Stroke Trials Archive (VISTA) database. Stroke 2012;43:1266–72. [5] Wardlaw JM, Keir SL, Bastin ME, Armitage PA, Rana AK. Is diffusion imaging appearance an independent predictor of outcome after ischemic stroke? Neurology 2002; 59:1381–7. [6] Hand PJ, Wardlaw JM, Rivers CS, Armitage PA, Bastin ME, Lindley RI, et al. MR diffusion-weighted imaging and outcome prediction after ischemic stroke. Neurology 2006;66:1159–63. [7] Johnston KC, Wagner DP, Wang XQ, Newman GC, Thijs V, Sen S, et al. Validation of an acute ischemic stroke model: does diffusion-weighted imaging lesion volume offer a clinically significant improvement in prediction of outcome? Stroke 2007;38:1820–5. [8] Lansberg MG, O'Brien MW, Tong DC, Moseley ME, Albers GW. Evolution of cerebral infarct volume assessed by diffusion-weighted magnetic resonance imaging. Arch Neurol 2001;58:613–7. [9] Schwamm LH, Koroshetz WJ, Sorensen AG, Wang B, Copen WA, Budzik R, et al. Time course of lesion development in patients with acute stroke: serial diffusion- and hemodynamic-weighted magnetic resonance imaging. Stroke 1998;29:2268–76. [10] Beaulieu C, de Crespigny A, Tong DC, Moseley ME, Albers GW, Marks MP. Longitudinal magnetic resonance imaging study of perfusion and diffusion in stroke: evolution of lesion volume and correlation with clinical outcome. Ann Neurol 1999;46:568–78. [11] Gaudinski MR, Henning EC, Miracle A, Luby M, Warach S, Latour LL. Establishing final infarct volume: stroke lesion evolution past 30 days is insignificant. Stroke 2008;39: 2765–8. [12] Cho KH, Kang DW, Kwon SU, Kim JS. Lesion volume increase is related to neurologic progression in patients with subcortical infarction. J Neurol Sci 2009;284:163–7. [13] Warach S, Kaufman D, Chiu D, Devlin T, Luby M, Rashid A, et al. Effect of the Glycine Antagonist Gavestinel on cerebral infarcts in acute stroke patients, a randomized placebo-controlled trial: The GAIN MRI Substudy. Cerebrovasc Dis 2006;21:106–11.
209
[14] Olivot JM, Mlynash M, Thijs VN, Kemp S, Lansberg MG, Wechsler L, et al. Relationships between infarct growth, clinical outcome, and early recanalization in diffusion and perfusion imaging for understanding stroke evolution (DEFUSE). Stroke 2008; 39:2257–63. [15] Cho KH, Kwon SU, Lee DH, Shim W, Choi C, Kim SJ, et al. Early infarct growth predicts long-term clinical outcome after thrombolysis. J Neurol Sci 2012;316:99–103. [16] Barrett KM, Ding YH, Wagner DP, Kallmes DF, Johnston KC, Investigators A. Change in diffusion-weighted imaging infarct volume predicts neurologic outcome at 90 days: results of the Acute Stroke Accurate Prediction (ASAP) trial serial imaging substudy. Stroke 2009;40:2422–7. [17] Cho KH, Kim JS, Kwon SU, Cho AH, Kang DW. Significance of susceptibility vessel sign on T2*-weighted gradient echo imaging for identification of stroke subtypes. Stroke 2005;36:2379–83. [18] Adams Jr HP, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993;24:35–41. [19] Luby M, Bykowski JL, Schellinger PD, Merino JG, Warach S. Intra- and interrater reliability of ischemic lesion volume measurements on diffusion-weighted, mean transit time and fluid-attenuated inversion recovery MRI. Stroke 2006;37:2951–6. [20] Minematsu K, Li L, Sotak CH, Davis MA, Fisher M. Reversible focal ischemic injury demonstrated by diffusion-weighted magnetic resonance imaging in rats. Stroke 1992;23:1304–10 [discussion 10–1]. [21] Takano K, Carano RA, Tatlisumak T, Meiler M, Sotak CH, Kleinert HD, et al. Efficacy of intra-arterial and intravenous prourokinase in an embolic stroke model evaluated by diffusion–perfusion magnetic resonance imaging. Neurology 1998;50:870–5. [22] Warach S, Pettigrew LC, Dashe JF, Pullicino P, Lefkowitz DM, Sabounjian L, et al. Effect of citicoline on ischemic lesions as measured by diffusion-weighted magnetic resonance imaging. Citicoline 010 Investigators. Ann Neurol 2000;48:713–22. [23] Merino JG, Latour LL, Todd JW, Luby M, Schellinger PD, Kang DW, et al. Lesion volume change after treatment with tissue plasminogen activator can discriminate clinical responders from nonresponders. Stroke 2007;38:2919–23. [24] Rabinstein A, Rundek T. Prediction of outcome after ischemic stroke: the value of clinical scores. Neurology 2013;80:15–6. [25] Megherbi SE, Milan C, Minier D, Couvreur G, Osseby GV, Tilling K, et al. Association between diabetes and stroke subtype on survival and functional outcome 3 months after stroke: data from the European BIOMED Stroke Project. Stroke 2003;34: 688–94. [26] Grau AJ, Weimar C, Buggle F, Heinrich A, Goertler M, Neumaier S, et al. Risk factors, outcome, and treatment in subtypes of ischemic stroke: the German stroke data bank. Stroke 2001;32:2559–66. [27] De Silva DA, Ebinger M, Christensen S, Parsons MW, Levi C, Butcher K, et al. Baseline diabetic status and admission blood glucose were poor prognostic factors in the EPITHET trial. Cerebrovasc Dis 2010;29:14–21. [28] Barber PA, Darby DG, Desmond PM, Yang Q, Gerraty RP, Jolley D, et al. Prediction of stroke outcome with echoplanar perfusion- and diffusion-weighted MRI. Neurology 1998;51:418–26. [29] Schlaug G, Benfield A, Baird AE, Siewert B, Lovblad KO, Parker RA, et al. The ischemic penumbra: operationally defined by diffusion and perfusion MRI. Neurology 1999; 53:1528–37. [30] Hossmann KA. Viability thresholds and the penumbra of focal ischemia. Ann Neurol 1994;36:557–65. [31] Sobesky J, Zaro Weber O, Lehnhardt FG, Hesselmann V, Neveling M, Jacobs A, et al. Does the mismatch match the penumbra? Magnetic resonance imaging and positron emission tomography in early ischemic stroke. Stroke 2005;36:980–5.
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Early infarct growth predicts long-term clinical outcome in ischemic stroke Seung Min Kim a,b, Sun U. Kwon a, Jong S. Kim a, Dong-Wha Kang a,⁎ a b
Department of Neurology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea Department of Neurology, Veterans Health Service Medical Center, Seoul, South Korea
a r t i c l e
i n f o
Article history: Received 24 June 2014 Received in revised form 27 August 2014 Accepted 26 September 2014 Available online 2 October 2014 Keywords: Infarct volume Infarct growth Outcome Ischemic stroke
a b s t r a c t Background: Ischemic lesions dynamically evolve during the acute phase of stroke. Although the ischemic lesion volume has been considered as a predictor of clinical outcome, it is still controversial whether early changes in ischemic lesion have prognostic information in addition to clinical variables. We hypothesized that early infarct growth on diffusion-weighted imaging (DWI) might be independently associated with long-term outcome in acute ischemic stroke patients. Methods: This was a prospective study for acute ischemic stroke patients admitted to the Stroke Unit of Asan Medical Center. The patients underwent DWI at baseline (within 24 h) and subsequently at 5 days after stroke onset. Early infarct growth was defined as the absolute difference between follow-up and baseline infarct volumes. Poor outcome was a modified Rankin Scale (mRS) at 3 months of 2–6 or 3–6. The association between infarct growth on DWI and clinical outcome was explored using multivariate analysis adjusting for demographics, risk factors for stroke, and other clinical variables. The cut-off values of early infarct growth predicting long-term outcomes were estimated using receiver operating characteristic analysis. Results: Of 409 patients enrolled, 345 (84.4%) showed any infarct growth (median, 0.63 cm3; interquartile range [IQR], 0.11–6.33 cm3; mean ± standard deviation, 9.55 ± 25.54 cm3). At the 3-month follow-up, the good outcomes were observed in 217 patients (53.1%) for mRS 0–1 and 303 patients (74.1%) for mRS 0–2. The larger infarct growth was associated with poor clinical outcome (for mRS 2–6, 0.29 cm3 [IQR 0.04–2.19] vs. 2.16 cm3 [IQR 0.26–17.68], p b 0.001; and for mRS 3–6, 0.39 cm3 [IQR 0.05–3.25] vs. 7.36 cm3 [IQR 0.57–26.48], p b 0.001). After adjusting age, diabetes, baseline National Institutes of Health Stroke Scale, and baseline infarct volume by multivariate logistic regression analysis, infarct growth was an independent predictor of poor clinical outcomes (for mRS 2–6, odds ratio [OR], 1.03, 95% confidence interval [CI], 1.004–1.06, p = 0.03; and for mRS 3– 6, OR, 1.03, 95% CI, 1.01–1.05, p = 0.01). The cut-off values of infarct growth discriminating between good and poor outcomes were 0.99 cm3 for mRS 0–1 vs. 2–6 (area under curve, 0.685; P b 0.001) and 8.86 cm3 for mRS 0–2 vs. 3–6 (area under curve, 0.736; P b 0.001). Conclusions: Our present study findings show that infarct growth within a week of onset independently predicts 3-month clinical outcomes. This suggests that short-term changes in infarct volume may serve as a surrogate marker of long-term clinical outcomes after ischemic stroke. © 2014 Elsevier B.V. All rights reserved.
1. Introduction An accurate prognosis of acute ischemic stroke is important for establishing a long-term treatment plan. Moreover, prognostic information can be used to evaluate the efficacy of thrombolysis or neuroprotective drugs. Clinical variables such as age, severity, and stroke subtype have consistently been associated with clinical outcomes [1–3]. A simple, reliable and inexpensive imaging predictor may be useful if it can ⁎ Corresponding author at: Department of Neurology, Asan Medical Center, University of Ulsan College of Medicine, 88 Olympic-ro 43-gil, Songpa-gu, Seoul 138-736, South Korea. Tel.: +82 2 3010 3440; fax: +82 2 474 4691. E-mail address: [email protected] (D.-W. Kang).
http://dx.doi.org/10.1016/j.jns.2014.09.048 0022-510X/© 2014 Elsevier B.V. All rights reserved.
demonstrate long-term outcome or early treatment effect. In previous reports, recanalization of vessels and lesion volume were suggested as imaging predictors [4]. However, it is controversial whether the lesion volume at a single time point can predict the clinical outcome in a stroke patient [5–7]. Ischemic lesion size increases dynamically and substantially during the first few days after onset and slowly decreases during the subsequent 3 to 4 weeks [8–11]. Thus, a hyperacute lesion on initial diffusion-weighted imaging (DWI) does not indicate the final lesion size, and it is assumed that assessment of infarct growth is a more reliable predictor of clinical outcome than ischemic lesion size at a single time point. In this regard, previous studies have tried to investigate the association between infarct growth and various clinical outcomes,
206
S.M. Kim et al. / Journal of the Neurological Sciences 347 (2014) 205–209
but have been limited to a retrospective design, specific stroke subtype (subcortical infarction) or treatment (thrombolysis) group, or a relatively small number of patients [12–16]. It remains unclear whether early infarct growth has predictive value for long-term prognosis in a large number of patients in routine clinical practice. In our prospective study, we hypothesized that early infarct growth might be independently associated with long-term outcome in acute ischemic stroke patients and attempted to estimate a cut-off value of infarct growth for predicting a poor outcome. 2. Methods 2.1. Patients Our current study was a prospective investigation of acute ischemic stroke patients admitted to the Stroke Unit of the Asan Medical Center between December 2004 and December 2006. We screened consecutive patients who had (1) acute ischemic stroke confirmed by initial MRI including DWI and MR angiography within 24 h after onset, (2) a followup DWI scan performed at 5 (± 1) days after symptom onset and (3) clinical follow-up for 3 months. Patients who had contraindications to MRI were excluded. The onset time was defined as the time patients were last known to be without neurological symptoms. This study was approved by the institutional review board of Asan Medical Center, and each patient or legal guardian provided written informed consent to participate in the study. 2.2. Imaging protocol and analysis MRI examinations were performed using a 1.5-T MR imaging unit (Signa; GE Medical Systems, Milwaukee, WI) with echo-planar capabilities. The initial MRI protocol included DWI, fluid-attenuated inversion recovery imaging (FLAIR), gradient echo T2-weighted imaging (GRE), perfusion-weighted imaging, 3D time-of-flight MRA, and 3D contrastenhanced MRA. The follow-up MRI protocol at 5 (± 1) days after onset included DWI, GRE, and FLAIR in all patients. The detailed MRI protocol has been previously described [17]. The infarct volumes on DWI were measured by an investigator (S.M.K.) who was blind to the clinical data. The infarct volume was measured on DWI as the sum of the infarct area in each slice multiplied by the slice thickness with the use of the Picture Archiving and Communication System at the Asan Medical Center. Early infarct growth was defined as the absolute difference between the follow-up infarct volume and the baseline infarct volume.
by using a Student's t-test or Mann–Whitney U test. Categorical variables were analyzed by a chi-square test or Fisher's exact test. A multivariate logistic regression analysis considering all variables was then conducted to assess the independent association of the infarct growth with clinical outcome. Infarct growth was considered as a continuous variable in univariate and multivariate analyses. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. Finally, we estimated the cut-off value of infarct growth for predicting poor clinical outcome at 3 months using receiver operating characteristic (ROC) analysis. The cut-off value of infarct growth was regarded as the value that had a maximum sum of sensitivity and specificity. In addition, we further evaluated the association of early infarct growth and clinical outcome after excluding the patients with small-vessel occlusion. All statistical analyses were performed with SPSS 20.0 for Windows (IBM Corp., Armonk, NY) and P b 0.05 was considered statistically significant. 3. Results 3.1. Baseline characteristics and infarct growth During the study period, we screened 1351 patients with acute ischemic stroke. Of these patients, 942 were excluded (570 patients underwent initial DWI after 24 h, 96 patients did not provide study consent, 184 patients did not undergo follow-up DWI, 76 patients had a contraindication for the MRI, and 16 patients were excluded due to loss to follow-up at 3 months). Thus, 409 patients (253 men) were included in the final analysis, with a mean age (standard deviation) of 63.6 (12.0) years. The median NIHSS score was 5 (IQR, 2–8). Because lesion volumes were not normally distributed, we used median value for analysis and expressed both median and mean values in the tables. The median infarct volume was 1.00 cm3 (IQR, 0.32–7.40) at baseline and 2.29 cm3 (IQR, 0.61–17.07) at follow-up, resulting in a median infarct growth of 0.63 cm3 (IQR, 0.11–6.33). Some degree of infarct growth was observed in 345 patients (84.4%). In 61 patients (14.9%), the infarct was smaller at follow-up than at baseline. In trend analysis, the infarct growth became larger as the mRS score increased (Fig. 1; P b 0.001). 3.2. Predictors of poor clinical outcomes At the 3-month follow-up, 53.1% and 74.1% of patients had excellent and independent outcomes, respectively. Univariate analysis demonstrated that a larger baseline infarct volume and larger infarct growth
2.3. Clinical assessment Stroke severity at admission was measured by the National Institutes of Health Stroke Scale (NIHSS) score. Stroke subtypes were determined according to the classification of the Trial of the Org 10172 in Acute Stroke Treatment (TOAST) [18]. Clinical outcomes were rated according to the modified Rankin Scale (mRS) at 3 months and were assessed by a certified research coordinator who was blind to the imaging and other clinical data. Clinical outcomes were dichotomized as (1) excellent (mRS 0–1) or poor (mRS 2–6) and (2) independent (mRS 0–2) or poor (mRS 3–6). 2.4. Statistical analysis We analyzed the relationship between early infarct growth and clinical outcome at 3 months. We also compared demographics, risk factors for stroke, initial NIHSS scores, stroke subtypes, initial glucose levels, reperfusion therapy, and baseline infarct volumes between patients with good and poor outcomes according to the clinical outcome criteria. Continuous or numerical variables were expressed as the mean (standard deviation) or median (interquartile range [IQR]) and were compared
Fig. 1. Association between infarct growth and mRS at 3 months. mRS, modified Rankin Scale. Infarct growth is presented as mean (standard deviation).
S.M. Kim et al. / Journal of the Neurological Sciences 347 (2014) 205–209
207
Table 1 Baseline characteristics associated with clinical outcomes at 3 month.
Age (years) Male Hypertension Diabetes Hyperlipidemia Smoking Previous stroke Initial glucose Onset to initial DWI (hours) Time interval between two DWIs (days) Initial NIHSS Reperfusion therapy Baseline infarct volume (cm3) Median Mean Infarct growth (cm3) Median Mean Stroke subtype Small-vessel occlusion Large artery atherosclerosis Cardioembolism Others
Total (n = 409)
Excellent outcome (mRS 0–1) Yes (n = 217)
No (n = 192)
P
Independent outcome (mRS 0–2) Yes (n = 303)
No (n = 106)
P
63.6 ± 12.0 253 (61.9%) 312 (76.3%) 126 (30.8%) 182 (44.5%) 121 (29.6%) 101 (24.7%) 148.1 ± 59.1 10.8 ± 6.8 4.7 ± 1.0 5 (2–8) 49 (12.0%)
61.4 ± 11.9 141 (65.0%) 158 (72.8%) 53 (24.4%) 94 (43.3%) 68 (31.3%) 49 (22.6%) 142.4 ± 53.1 10.3 ± 6.7 4.5 ± 0.9 3 (1–5) 14 (6.5%)
65.4 ± 11.8 112 (58.3%) 154 (80.2%) 73 (38.0%) 88 (45.8%) 53 (27.6%) 52 (27.1%) 154.8 ± 65.5 11.3 ± 6.8 4.9 ± 1.1 7 (4–12.25) 35 (18.2%)
0.001 0.167 0.079 0.003 0.609 0.409 0.292 0.044 0.124 b0.001 b0.001 b0.001
61.4 ± 11.4 199 (65.7%) 225 (74.3%) 82 (27.1%) 126 (41.6%) 96 (31.7%) 72 (23.8%) 145.7 ± 59.8 10.6 ± 6.8 4.5 ± 0.9 3 (2–6) 28 (9.0%)
68.7 ± 12.2 54 (51.0%) 87 (82.0%) 44 (41.5%) 56 (52.8%) 25 (23.6%) 29 (27.4%) 155.6 ± 58.0 11.1 ± 6.6 4.9 ± 1.2 10 (6–14.75) 21 (19.8%)
b0.001 0.007 0.103 0.006 0.045 0.116 0.460 0.155 0.502 0.002 b0.001 0.004
1.00 (0.32–7.40) 9.29 ± 21.13
0.66 (0.24–3.45) 4.04 ± 9.33
2.16 (0.46–16.24) 15.20 ± 28.06
b0.001 b0.001
0.73 (0.27–4.35) 6.12 ± 15.74
3.23 (0.64–21.38) 18.27 ± 30.11
b0.001 b0.001
0.63 (0.11–6.33) 9.53 ± 25.63
0.29 (0.04–2.19) 2.55 ± 6.50
2.16 (0.26–17.68) 17.46 ± 35.03
b0.001 b0.001 0.042
0.39 (0.05–3.25) 3.91 ± 11.03
7.36 (0.57–26.48) 25.67 ± 42.79
b0.001 b0.001 0.027
110 122 101 76
69 (31.8%) 64 (29.5%) 45 (20.7%) 39 (18.0%)
41 (21.4%) 58 (30.2) 56 (29.2%) 37 (19.3%)
94 (31%) 88 (29%) 63 (21%) 58 (19%)
16 (15%) 34 (32%) 38 (36%) 18 (17%)
Data are expressed as a number (percent), mean ± standard deviation, or median (interquartile range). NIHSS, National Institutes of Health Stroke Scale.
were associated with poor outcome at 3 months. In addition, poor outcomes were associated with an older age (for mRS 2–6 and for mRS 3–6), higher initial NIHSS score (for mRS 2–6 and for mRS 3–6), diabetes (for mRS 2–6 and for mRS 3–6), reperfusion therapy (for mRS 2–6 and for mRS 3–6), time interval between the two DWI scans (for mRS 2–6 and for mRS 3–6), initial glucose level (for mRS 2–6), hyperlipidemia (for mRS 3–6), and stroke subtype (for mRS 2–6 and for mRS 3–6) (Table 1). Multivariate logistic regression analysis showed that a larger infarct growth, older age, higher initial NIHSS score (for mRS 2–6 and for mRS 3–6), diabetes and time interval between the two DWI scans (for mRS 2–6) were independent predictors of poor clinical outcomes. In contrast, baseline infarct volume was no longer significant upon multivariate analysis (Table 2). Excluding patients with small-vessel occlusion, larger infarct growth, older age and higher initial NIHSS were still found to be associated with poor clinical outcome (Supplemental Tables 1 and 2). 3.3. Cut-off values of early infarct growth predicting long-term clinical outcomes The cut-off values for infarct growth that discriminate between poor and good clinical outcomes were 0.99 cm3 for mRS 2–6 (area under curve [AUC], 0.685; P b 0.001) and 8.86 cm3 for mRS 3–6 (AUC, 0.736;
P b 0.001) (Fig. 2). Excluding patients with small-vessel occlusion, cutoff values were 8.85 cm3 for mRS 2–6 (AUC, 0.694; P b 0.001) and 10.10 cm3 for mRS 3–6 (AUC, 0.741; P b 0.001) (Supplemental Fig. 1). 4. Discussion In the present study, we demonstrated that early infarct growth in acute ischemic stroke is independently associated with long-term clinical outcome. In addition, the association was still significant even after exclusion of patients with small-vessel occlusion. Among clinical variables, age, initial NIHSS, diabetes and time interval between the two DWI scans were also identified as predictors of clinical outcome. We also estimated a cut-off value of early infarct growth for predicting poor outcome, which may better predict long-term prognosis. This study has some advantages over previous ones, by being the largest prospective study to date, by adjusting for most of the clinical variables, and by considering all treatment groups seen in routine practice to reflect the real clinical situation. Early infarct growth on MRI likely reflects various pathophysiological processes. It can reflect expansion of the cerebral infarction from the ischemic penumbra, shown by the increase in cytotoxic edema on DWI that occurs in the first few hours after stroke onset. In addition, infarct growth can reflect the formation of vasogenic edema, which plays
Table 2 Predictors of poor clinical outcomes by multivariate logistic regression analysis. Poor outcome (mRS 2–6) OR (95% CI) Male sex Age Diabetes Hyperlipidemia Time interval between two DWIs (per 1 day) NIHSS (per 1) Reperfusion therapy Baseline infarct volume (per 1 cm3) Infarct growth (per 1 cm3) Stroke subtype
Poor outcome (mRS 3–6) P
1.025 (1.004–1.047) 1.708 (1.028–2.836)
0.019 0.039
1.327 (1.031–1.709) 1.228 (1.140–1.323) 0.967 (0.398–2.345) 1.004 (0.983–1.026) 1.033 (1.003–1.064)
0.028 b0.001 0.940 0.691 0.029 0.539
OR, odds ratio; CI, confidence interval; NIHSS, National Institutes of Health Stroke Scale.
OR (95% CI)
P
0.684 (0.382–1.225) 1.059 (1.030–1.089) 1.677 (0.914–3.074) 1.603 (0.906–2.837) 1.318 (0.971–1.788) 1.223 (1.137–1.317) 0.560 (0.210–1.491) 0.989 (0.973–1.006) 1.031 (1.010–1.052)
0.201 b0.001 0.095 0.105 0.076 b0.001 0.245 0.213 0.003 0.887
208
S.M. Kim et al. / Journal of the Neurological Sciences 347 (2014) 205–209
Fig. 2. Cut-off values of early infarct growth from ROC curves. Poor outcomes were defined as mRS 2–6 (A) and mRS 3–6 (B).
an important role in the subsequent 2 to 3 days. These mechanisms can be affected by various clinical factors, such as vessel occlusion, collateral flow patterns, reperfusion, systemic blood pressure, and stroke etiology. Despite the concern that estimation of infarct growth on MRI might contain a measurement error due to the sensitivity of MRI, intra- and inter-rater concordances were excellent in a previous study [19]. Although infarct volume is considered to be a variable predicting clinical outcome, infarct volume measurement at a single time point does not reflect the true biology and dynamic nature of progressive ischemic injury because early cellular changes following ischemia can potentially be reversed and infarct growth can be influenced by reperfusion therapy or other neuroprotective drugs [20,21]. Hence, an image taken at a single time point, such as that of the final infarct volume or the initial infarct volume, may not be sufficient for evaluating the long-term prognosis or treatment effect. In this regard, infarct growth was used as the outcome variable in several stroke trials. However, infarct growth was previously defined as the difference between the initial DWI lesion and 90-day T2 or FLAIR lesions [13,22,23], which is inadequate for the evaluation of an early treatment effect. In addition, evaluation may not be completed due to follow-up loss or death. Because we evaluated the early infarct growth in the initial 5 days after stroke onset, our findings provide evidence that this growth can be used as a marker of early treatment effects in acute ischemic stroke. Moreover, because most patients are hospitalized for the first week, follow-up MRI at 5 days may be more practical than at 90 days after onset [8]. In addition to early infarct growth, our study showed that age, initial NIHSS, and diabetes were significantly correlated with long-term clinical outcome. These results are in agreement with those of previous studies [2,3,24–27]. Our present study thus provides clear evidence that early infarct growth has a predictive value for prognosis, even after adjustment for these clinical variables. Although reperfusion therapy was associated with poor outcome in univariate analysis, it did not reach significance in multivariate analysis. This result may be due to the baseline characteristics of patients who received reperfusion therapy. They had higher initial NIHSS scores, larger infarct growth, and a high proportion of cardioembolism compared to patients who did not receive reperfusion therapy (data not shown). More severe stroke patients tended to receive reperfusion therapy which resulted in poor clinical outcome, but the association between reperfusion therapy and poor outcome disappeared after adjusting for other clinical variables. We also investigated the cut-off values of early infarct growth for predicting poor clinical outcomes—defined as mRS 2–6 and mRS 3–6—
and measured them at 0.99 cm3 and 8.86 cm3, respectively. The cutoff value of infarct growth might be clinically useful when evaluating the effect of thrombolysis or neuroprotective treatment. The cut-off volumes and infarct growth are smaller than that of a previous study (approximately 15 cm3) that studied acute stroke patients with anterior circulation infarct (baseline lesion volume N 2 cm3) or treated with thrombolysis [11,15]. A plausible explanation for this disparity may be the difference in study population: this study included patients regardless of thrombolytic treatment, which resulted in the enrollment of a large number of patients with small-vessel occlusion. Our study had several limitations of note. First, we did not consider the perfusion status and recanalization of occluded vessels. Because infarct progression from the ischemic penumbra is one of the most important causes of infarct growth [9,22,28,29], this growth might reflect the size of the penumbra. However, PWI lesions also include the benign oligemic area and would thereby overestimate hypoperfused tissues [30]. Thus, a PWI–DWI mismatch may not accurately reflect the potentially salvageable ischemic tissue [31]. Second, because we assessed infarct growth using the absolute difference in volumes between initial and follow-up infarcts, there was a tendency for a larger infarct growth with a larger initial infarction. However, if infarct growth was calculated as a percentage change, a smaller infarct at baseline would give a greater increase in percentage change. Nevertheless, this limitation did not significantly affect our results because we included patients with all stroke subtypes and the result was significant after adjusting for initial lesion volume. Hence, the evaluation of infarct growth via the absolute difference in volume might be a more appropriate method. Despite these limitations, our study findings have significant clinical implications. Although the most important end point in acute stroke trials is clinical outcome, an ancillary outcome measure may be valuable, especially in phase 2 clinical trials. Because long-term clinical outcome can be affected by various clinical factors, including long-term treatment and rehabilitation, early infarct growth can be used as a good surrogate marker for clinical outcome focused on the acute treatment effect. In conclusion, our results demonstrate that early infarct growth independently predicts poor clinical outcomes in acute ischemic stroke patients. These findings support the belief that estimation of early infarct growth may be a good surrogate marker for clinical outcome in acute treatment trials. Conflict of interest statement The authors report no disclosures.
S.M. Kim et al. / Journal of the Neurological Sciences 347 (2014) 205–209
Acknowledgements This study was supported by grants from the Korea Health Technology R&D Project, Ministry for Health & Welfare, Republic of Korea (HI12C1847 and HI11C1531). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.jns.2014.09.048. References [1] Johnston KC, Connors Jr AF, Wagner DP, Knaus WA, Wang X, Haley Jr EC. A predictive risk model for outcomes of ischemic stroke. Stroke 2000;31:448–55. [2] Weimar C, Konig IR, Kraywinkel K, Ziegler A, Diener HC. German Stroke Study C. Age and National Institutes of Health Stroke Scale Score within 6 hours after onset are accurate predictors of outcome after cerebral ischemia: development and external validation of prognostic models. Stroke 2004;35:158–62. [3] Saposnik G, Guzik AK, Reeves M, Ovbiagele B, Johnston SC. Stroke Prognostication using Age and NIH Stroke Scale: SPAN-100. Neurology 2013;80:21–8. [4] Vogt G, Laage R, Shuaib A, Schneider A, Collaboration V. Initial lesion volume is an independent predictor of clinical stroke outcome at day 90: an analysis of the Virtual International Stroke Trials Archive (VISTA) database. Stroke 2012;43:1266–72. [5] Wardlaw JM, Keir SL, Bastin ME, Armitage PA, Rana AK. Is diffusion imaging appearance an independent predictor of outcome after ischemic stroke? Neurology 2002; 59:1381–7. [6] Hand PJ, Wardlaw JM, Rivers CS, Armitage PA, Bastin ME, Lindley RI, et al. MR diffusion-weighted imaging and outcome prediction after ischemic stroke. Neurology 2006;66:1159–63. [7] Johnston KC, Wagner DP, Wang XQ, Newman GC, Thijs V, Sen S, et al. Validation of an acute ischemic stroke model: does diffusion-weighted imaging lesion volume offer a clinically significant improvement in prediction of outcome? Stroke 2007;38:1820–5. [8] Lansberg MG, O'Brien MW, Tong DC, Moseley ME, Albers GW. Evolution of cerebral infarct volume assessed by diffusion-weighted magnetic resonance imaging. Arch Neurol 2001;58:613–7. [9] Schwamm LH, Koroshetz WJ, Sorensen AG, Wang B, Copen WA, Budzik R, et al. Time course of lesion development in patients with acute stroke: serial diffusion- and hemodynamic-weighted magnetic resonance imaging. Stroke 1998;29:2268–76. [10] Beaulieu C, de Crespigny A, Tong DC, Moseley ME, Albers GW, Marks MP. Longitudinal magnetic resonance imaging study of perfusion and diffusion in stroke: evolution of lesion volume and correlation with clinical outcome. Ann Neurol 1999;46:568–78. [11] Gaudinski MR, Henning EC, Miracle A, Luby M, Warach S, Latour LL. Establishing final infarct volume: stroke lesion evolution past 30 days is insignificant. Stroke 2008;39: 2765–8. [12] Cho KH, Kang DW, Kwon SU, Kim JS. Lesion volume increase is related to neurologic progression in patients with subcortical infarction. J Neurol Sci 2009;284:163–7. [13] Warach S, Kaufman D, Chiu D, Devlin T, Luby M, Rashid A, et al. Effect of the Glycine Antagonist Gavestinel on cerebral infarcts in acute stroke patients, a randomized placebo-controlled trial: The GAIN MRI Substudy. Cerebrovasc Dis 2006;21:106–11.
209
[14] Olivot JM, Mlynash M, Thijs VN, Kemp S, Lansberg MG, Wechsler L, et al. Relationships between infarct growth, clinical outcome, and early recanalization in diffusion and perfusion imaging for understanding stroke evolution (DEFUSE). Stroke 2008; 39:2257–63. [15] Cho KH, Kwon SU, Lee DH, Shim W, Choi C, Kim SJ, et al. Early infarct growth predicts long-term clinical outcome after thrombolysis. J Neurol Sci 2012;316:99–103. [16] Barrett KM, Ding YH, Wagner DP, Kallmes DF, Johnston KC, Investigators A. Change in diffusion-weighted imaging infarct volume predicts neurologic outcome at 90 days: results of the Acute Stroke Accurate Prediction (ASAP) trial serial imaging substudy. Stroke 2009;40:2422–7. [17] Cho KH, Kim JS, Kwon SU, Cho AH, Kang DW. Significance of susceptibility vessel sign on T2*-weighted gradient echo imaging for identification of stroke subtypes. Stroke 2005;36:2379–83. [18] Adams Jr HP, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993;24:35–41. [19] Luby M, Bykowski JL, Schellinger PD, Merino JG, Warach S. Intra- and interrater reliability of ischemic lesion volume measurements on diffusion-weighted, mean transit time and fluid-attenuated inversion recovery MRI. Stroke 2006;37:2951–6. [20] Minematsu K, Li L, Sotak CH, Davis MA, Fisher M. Reversible focal ischemic injury demonstrated by diffusion-weighted magnetic resonance imaging in rats. Stroke 1992;23:1304–10 [discussion 10–1]. [21] Takano K, Carano RA, Tatlisumak T, Meiler M, Sotak CH, Kleinert HD, et al. Efficacy of intra-arterial and intravenous prourokinase in an embolic stroke model evaluated by diffusion–perfusion magnetic resonance imaging. Neurology 1998;50:870–5. [22] Warach S, Pettigrew LC, Dashe JF, Pullicino P, Lefkowitz DM, Sabounjian L, et al. Effect of citicoline on ischemic lesions as measured by diffusion-weighted magnetic resonance imaging. Citicoline 010 Investigators. Ann Neurol 2000;48:713–22. [23] Merino JG, Latour LL, Todd JW, Luby M, Schellinger PD, Kang DW, et al. Lesion volume change after treatment with tissue plasminogen activator can discriminate clinical responders from nonresponders. Stroke 2007;38:2919–23. [24] Rabinstein A, Rundek T. Prediction of outcome after ischemic stroke: the value of clinical scores. Neurology 2013;80:15–6. [25] Megherbi SE, Milan C, Minier D, Couvreur G, Osseby GV, Tilling K, et al. Association between diabetes and stroke subtype on survival and functional outcome 3 months after stroke: data from the European BIOMED Stroke Project. Stroke 2003;34: 688–94. [26] Grau AJ, Weimar C, Buggle F, Heinrich A, Goertler M, Neumaier S, et al. Risk factors, outcome, and treatment in subtypes of ischemic stroke: the German stroke data bank. Stroke 2001;32:2559–66. [27] De Silva DA, Ebinger M, Christensen S, Parsons MW, Levi C, Butcher K, et al. Baseline diabetic status and admission blood glucose were poor prognostic factors in the EPITHET trial. Cerebrovasc Dis 2010;29:14–21. [28] Barber PA, Darby DG, Desmond PM, Yang Q, Gerraty RP, Jolley D, et al. Prediction of stroke outcome with echoplanar perfusion- and diffusion-weighted MRI. Neurology 1998;51:418–26. [29] Schlaug G, Benfield A, Baird AE, Siewert B, Lovblad KO, Parker RA, et al. The ischemic penumbra: operationally defined by diffusion and perfusion MRI. Neurology 1999; 53:1528–37. [30] Hossmann KA. Viability thresholds and the penumbra of focal ischemia. Ann Neurol 1994;36:557–65. [31] Sobesky J, Zaro Weber O, Lehnhardt FG, Hesselmann V, Neveling M, Jacobs A, et al. Does the mismatch match the penumbra? Magnetic resonance imaging and positron emission tomography in early ischemic stroke. Stroke 2005;36:980–5.
