You are on page 1of 11

Clinical Section: Research Article

Gerontology 2023;69:571–580 Received: August 2, 2022


Accepted: December 27, 2022
DOI: 10.1159/000528951 Published online: January 5, 2023

Lactate Dehydrogenase Predicts


Hemorrhagic Transformation in Patients
with Acute Ischemic Stroke

Downloaded from http://karger.com/ger/article-pdf/69/5/571/3949659/000528951.pdf by Univ. of California Davis user on 01 April 2024


Lingli Chen a Minjie Xu b Qiqi Huang c Caiyun Wen d Wenwei Ren a
aDepartment
of Neurology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, China; bSchool of
Mental Health, Wenzhou Medical University, Wenzhou, China; cDepartment of Cardiac Care Unit, The First Affiliated
Hospital of Wenzhou Medical University, Wenzhou, China; dDepartment of Radiology, The First Affiliated Hospital of
Wenzhou Medical University, Wenzhou, China

Keywords infarction subgroup (281.0 [230.0–340.0] U/L versus 258.0


Lactate dehydrogenase · Hemorrhagic transformation · [209.0–311.0] U/L, p < 0.001). The area under the ROC curve
Acute ischemic stroke · Risk · Association of LDH was 0.890 (95% confidence level [CI] 0.874–0.905,
p < 0.001). Besides, logistic regression revealed that high
LDH levels (LDH >215 U/L) showed a higher risk of HT (odds
Abstract ratio = 10.958, 95% CI 7.964–15.078, p < 0.001). Conclusion:
Introduction: Hemorrhagic transformation (HT) is a severe High LDH levels were linked with an increased risk of HT in
but frequent complication of acute ischemic stroke (AIS). AIS patients. Practical measures should be considered in pa-
This study aimed to evaluate the relationship between se- tients with increased LDH levels (LDH >215 U/L).
rum lactate dehydrogenase (LDH) levels and HT. Methods: © 2023 S. Karger AG, Basel
We retrospectively included 542 AIS patients with HT and
1,091 age- and gender-matched patients without HT. De-
mographic and clinical data were obtained from medical Introduction
records, and blood samples were obtained within 24 h after
admission. The characteristics of the groups were compared. Hemorrhagic transformation (HT) is a common com-
With the receiver operating characteristic (ROC) curve anal- plication of cerebral infarction, often leading to the dete-
ysis, we assessed the discriminating capacity of LDH levels rioration of clinical symptoms and increased death and
in predicting HT in patients with AIS. The logistic regression disability rates [1–4]. Exploring the risk factors for HT
model was used to determine the connection between LDH may assist physicians in providing appropriate therapy
and HT. Results: The HT group had considerably higher LDH for patients with high HT risk after acute ischemic stroke
levels than the non-HT group (263.0 [216.0–323.3] U/L versus (AIS).
178.0 [162.0–195.0] U/L, p < 0.001). We also observed that
the levels of LDH in the parenchymal hemorrhage sub- Lingli Chen, Minjie Xu, and Qiqi Huang have contributed equally to
group were significantly higher than those in the hemorrhagic this work.

Karger@karger.com © 2023 S. Karger AG, Basel Correspondence to:


www.karger.com/ger Caiyun Wen, wencaiyunwenzhou @ qq.com
Wenwei Ren, renwenwei @ wmu.edu.cn
Downloaded from http://karger.com/ger/article-pdf/69/5/571/3949659/000528951.pdf by Univ. of California Davis user on 01 April 2024
Fig. 1. The flow of participants. AIS, acute
ischemic stroke; HT, hemorrhagic trans-
formation.

Fig. 2. CT scan illustrations of HI and PH, as


defined by the European Cooperative Acute
Stroke Study II criteria. HI, hemorrhagic in-
farction; PH, parenchymal hemorrhage.

Lactate dehydrogenase (LDH), an end product of ischemia, myocarditis, hepatitis, liver cirrhosis, tumors,
glycolysis, is a hydrogen transfer enzyme in many cells pneumonia, and sepsis [9–13]. For the central nervous
and organs, including the heart, lung, liver, and central system, patients with hypoxic-ischemic encephalopathy,
nervous system [5–8]. The irregular extracellular occur- cerebral hemorrhage, posterior reversible encephalopathy
rence of LDH, which can be detectable in serum to discov- syndrome, and subarachnoid hemorrhage with delayed
er cell or organ injury, has been noted as a warning sign cerebral ischemia had higher LDH levels than control
in a wide range of clinical situations, including myocardial groups [10, 14, 15]. Previous studies have demonstrated

572 Gerontology 2023;69:571–580 Chen/Xu/Huang/Wen/Ren


DOI: 10.1159/000528951
that high LDH levels could predict elevated all-cause
mortality and poor functional outcomes in patients with
AIS [16]. Besides, a retrospective observational study
revealed a link between higher serum LDH levels and
all-cause mortality, including stroke, in patients with
metabolic syndrome [17]. Although LDH has long been
considered a predictor of various adverse health out-
comes and the LDH/pyruvate dehydrogenase activity
ratio increased following ischemia in rat brain slices,
the role of LDH in predicting HT in patients with AIS
is not well explored [18]. Cell death is usually accom-
panied by cytoplasmic destruction, which releases
LDH into the peripheral circulation [19]. HT patients

Downloaded from http://karger.com/ger/article-pdf/69/5/571/3949659/000528951.pdf by Univ. of California Davis user on 01 April 2024


often have more severe pathological destruction than Fig. 3. a The levels of LDH in HT and non-HT groups. b The lev-
els of LDH in the HI and PH subtypes. LDH, lactate dehydroge-
non-HT patients, with more cell necrosis and theoreti- nase; HT, hemorrhagic transformation; HI, hemorrhagic infarc-
cally higher peripheral LDH release than non-HT pa- tion; PH, parenchymal hemorrhage. **p < 0.01, ***p < 0.001.
tients [20, 21]. A recent study reported that elevated
serum LDH levels could predict unfavorable outcomes
following recombinant tissue plasminogen activator
thrombolysis in patients with AIS [22]. Therefore, we
speculate whether HT patients have higher serum LDH
levels than non-HT patients.
Based on these findings, this study aimed to determine
if LDH had a role in developing HT in patients with AIS.
We hypothesized that a high LDH was related to an in-
creased risk of HT in AIS patients.

Methods

Participants
We retrospectively analyzed consecutive patients with AIS be-
tween December 2012 and December 2021 from the First Affiliated
Hospital of Wenzhou Medical University. The following criteria
were used to determine inclusion: (1) age 18 years or older; (2) di-
agnosed with AIS and verified using MRI or CT; (3) onset of stroke
within 7 days; and (4) had undergone the follow-up CT or MRI
scans. Patients with incomplete data were excluded (Fig. 1).

Data Collection
The electronic medical record system was used to gather base- Fig. 4. The ROC curve of LDH predicts the risk of HT in individu-
line characteristics. General demographic factors included age, als with AIS. ROC, receiver operating characteristic; LDH, lactate
gender, body mass index (BMI), smoking, and drinking history. dehydrogenase; HT, hemorrhagic transformation; AIS, acute is-
Hypertension, diabetes, and atrial fibrillation (AF) were all identi- chemic stroke.
fied as comorbidities. The National Institutes of Health Stroke
Scale (NIHSS) score, systolic blood pressure, diastolic blood pres-
sure, the trial of ORG 10172 in acute stroke treatment (TOAST),
and infarct location were all collected clinically. Blood parameters, NIHSS score was used to assess the severity of the stroke. All
including platelet count, creatinine, LDH, low-density lipoprotein individuals were categorized into large artery atherosclerosis,
(LDL), prothrombin time (PT), the international normalized ratio cardioembolism, small vessel occlusion, and other subtypes (in-
(INR), and the prothrombin time ratio (PTR), were measured in cluding stroke of other determined etiology and stroke of undeter-
the hospital biochemistry department after an overnight fast (at mined etiology) using the TOAST criteria [23]. Infarct locations
least 8 h) within 24 h after admission. Therapy was divided into were categorized as follows based on brain radiological examination:
conventional therapy and thrombolysis or thrombectomy therapy. lobar (frontal, insular, temporal, occipital, parietal), subcortical

Prediction of Hemorrhagic Gerontology 2023;69:571–580 573


Transformation DOI: 10.1159/000528951
Table 1. Comparison of clinical features between HT and non-HT groups

Variables Non-HT (n = 1,091) HT (n = 542) Statistic p value

Age, years (median, IQR) 72.0 (62.0–78.0) 72.0 (62.0–78.3) −0.018 0.986
Gender, n (%)
Male 743 (68.1) 371 (68.5) 0.020 0.887
Female 348 (31.9) 171 (31.5)
BMI, kg/m2 (median, IQR) 23.5 (22.2–24.8) 23.8 (22.8–24.8) −2.749 0.006
Smoking, n (%) 442 (40.5) 223 (41.1) 0.060 0.807
Drinking, n (%) 373 (34.2) 190 (35.1) 0.120 0.729
Hypertension, n (%) 755 (69.2) 342 (63.1) 6.117 0.013
Diabetes, n (%) 331 (30.3) 157 (29.0) 0.325 0.568
Atrial fibrillation, n (%) 107 (9.8) 231 (42.6) 237.520 <0.001
NIHSS (median, IQR) 2.0 (1.0–4.0) 3.0 (1.0–8.0) −9.154 <0.001

Downloaded from http://karger.com/ger/article-pdf/69/5/571/3949659/000528951.pdf by Univ. of California Davis user on 01 April 2024


SBP, mm Hg (median, IQR) 153.0 (138.0–169.3) 155.0 (138.0–168.0) −0.337 0.736
DBP, mm Hg (median, IQR) 84.0 (76.0–92.0) 86.0 (75.0–93.0) −0.985 0.325
TOAST, n (%)
Large artery atherosclerosis 543 (49.8) 169 (31.2) 246.547 <0.001
Cardioembolism 102 (9.3) 221 (40.8)
Small vessel occlusion 75 (6.9) 2 (0.4)
Other subtypes 371 (34.0) 150 (27.7)
Therapy, n (%)
Conventional therapy 1,090 (99.9) 499 (92.1) 84.933 <0.001
Thrombolysis or thrombectomy 1 (0.1) 43 (7.9)
Infarct location, n (%)
Lobar 149 (13.7) 113 (20.8) 530.666 <0.001
Subcortical 543 (49.8) 37 (6.8)
Brainstem 178 (16.3) 5 (0.9)
Cerebellum 20 (1.8) 35 (6.5)
Mixed type 201 (18.4) 352 (64.9)
HT, n (%)
Non-HT 1,091 (100) 0 (0) 1,633.000 <0.001
HI 0 (0) 351 (64.8)
PH 0 (0) 191 (35.2)
Platelet count, 109/L (median, IQR) 209.0 (174.0–249.0) 196.0 (162.0–234.4) −4.124 <0.001
Creatinine, μmol/L (median, IQR) 72.0 (61.0–86.0) 74.0 (62.0–88.0) −1.100 0.271
LDH, U/L (median, IQR) 178.0 (162.0–195.0) 263.0 (216.0–323.3) −25.687 <0.001
LDH >215 U/L, n (%) 124 (11.4) 408 (75.3) 673.384 <0.001
LDL, mmol/L (median, IQR) 2.55 (2.04–3.11) 2.64 (2.17–3.22) −2.505 0.012
PT, s (median, IQR) 13.6 (13.1–14.2) 13.9 (13.3–14.5) −5.393 <0.001
INR (median, IQR) 1.05 (1.01–1.11) 1.07 (1.02–1.14) −4.490 <0.001
PTR (median, IQR) 0.98 (0.93–1.03) 0.97 (0.92–1.02) −2.614 0.009

IQR, interquartile range; HT, hemorrhagic transformation; HI, hemorrhagic infarction; PH, parenchymal
hemorrhage; BMI, body mass index; NIHSS, Institutes of Health Stroke Scale; SBP, systolic blood pressure; DBP,
diastolic blood pressure; TOAST, the trial of ORG 10172 in acute stroke treatment; LDH, lactate dehydrogenase; LDL,
low-density lipoprotein; PT, prothrombin time; INR, international normalized ratio; PTR, prothrombin time ratio.

(corona radiata, corpus callosum, thalamus, basal ganglia, internal Statistical Analyses
capsule), brainstem (medulla, midbrain, pons), cerebellum, and the Continuous data were presented as mean ± standard devia-
mixed type (comprises at least two of the types mentioned above). tion or median and interquartile range, depending on the data
Two neuroimaging physicians retrospectively evaluated MRI or distribution. The normally distributed continuous variables
CT scans. The European cooperative acute stroke study classifica- were compared with the Student’s t test and one-way analysis of
tion criteria were used to classify HT, and examples of HT subtypes variance. For non-normally continuous variables, the Mann-
(hemorrhagic infarction [HI] and parenchymal hemorrhage [PH]) Whitney U test was applied to assess the differences between two
were given in Figure 2 [24]. When a dispute arose, the third neu- groups, while the Kruskal-Wallis test was for three or more
roimaging physician made the decision. groups. Categorical variables were represented as frequencies or

574 Gerontology 2023;69:571–580 Chen/Xu/Huang/Wen/Ren


DOI: 10.1159/000528951
Table 2. Comparison of subtypes

Variables Non-HT (n = 1,091) HI (n = 351) PH (n = 191) Statistic p value

Age, years (median, IQR) 72.0 (62.0–78.0) 71.0 (61.0–78.0) 72.0 (64.0–79.0) 1.870 0.393
Gender, n (%)
Male 743 (68.1) 232 (66.1) 139 (72.8) 2.564 0.277
Female 348 (31.9) 119 (33.9) 52 (27.2)
BMI, kg/m2 (median, IQR) 23.5 (22.2–24.8) 23.8 (22.6–24.7) 24.0 (22.9–24.9) 9.197 0.010
Smoking, n (%) 442 (40.5) 138 (39.3) 85 (44.5) 1.438 0.487
Drinking, n (%) 373 (34.2) 121 (34.5) 69 (36.1) 0.270 0.874
Hypertension, n (%) 755 (69.2) 223 (63.5) 119 (62.3) 6.201 0.045
Diabetes, n (%) 331 (30.3) 119 (33.9) 38 (19.9) 11.909 0.003
AF, n (%) 107 (9.8) 140 (39.9) 91 (47.6) 242.055 <0.001
NIHSS (median, IQR) 2.0 (1.0–4.0) 3.0 (1.0–8.0) 6.0 (2.0–9.0) 93.430 <0.001

Downloaded from http://karger.com/ger/article-pdf/69/5/571/3949659/000528951.pdf by Univ. of California Davis user on 01 April 2024


SBP, mm Hg (median, IQR) 153.0 (138.0–169.3) 155.0 (139.0–168.0) 154.0 (134.0–167.0) 2.212 0.331
DBP, mm Hg (median, IQR) 84.0 (76.0–92.0) 87.0 (76.0–92.0) 83.0 (74.0–95.0) 1.440 0.487
TOAST, n (%)
Large artery atherosclerosis 543 (49.8) 117 (33.3) 52 (27.2) Fisher <0.001
Cardioembolism 102 (9.3) 125 (35.6) 96 (50.3)
Small vessel occlusion 75 (6.9) 2 (0.6) 0 (0)
Other subtypes 371 (34.0) 107 (30.5) 43 (22.5)
Therapy, n (%)
Conventional therapy 1,090 (99.9) 329 (93.7) 170 (89.0) 95.475 <0.001
Thrombolysis or thrombectomy 1 (0.1) 22 (6.3) 21 (11.0)
Infarct location, n (%)
Lobar 149 (13.7) 80 (22.8) 33 (17.3) 535.246 <0.001
Subcortical 543 (49.8) 27 (7.7) 10 (5.2)
Brainstem 178 (16.3) 3 (0.9) 2 (1.0)
Cerebellum 20 (1.8) 22 (6.3) 13 (6.8)
Mixed type 201 (18.4) 219 (62.4) 133 (69.6)
Platelet count 109/L (median, IQR) 209.0 (174.0–249.0) 201.0 (167.0–235.0) 188.0 (156.0–230.0) 20.348 <0.001
Creatinine, μmol/L (median, IQR) 72.0 (61.0–86.0) 71.0 (61.0–86.0) 76.0 (64.0–88.0) 5.355 0.069
LDH, U/L (median, IQR) 178.0 (162.0–195.0) 258.0 (209.0–311.0) 281.0 (230.0–340.0) 664.270 <0.001
LDH >215 U/L, n (%) 124 (11.4) 255 (72.6) 153 (80.1) 676.514 <0.001
LDL, mmol/L (median, IQR) 2.55 (2.04–3.11) 2.64 (2.14–3.21) 2.6 (2.20–3.25) 6.666 0.036
PT, s (median, IQR) 13.6 (13.1–14.2) 13.8 (13.2–14.4) 14.0 (13.4–14.6) 32.441 <0.001
INR (median, IQR) 1.05 (1.01–1.11) 1.07 (1.02–1.13) 1.08 (1.02–1.15) 22.870 <0.001
PTR (median, IQR) 0.98 (0.93–1.03) 0.96 (0.92–1.01) 0.98 (0.93–1.03) 15.207 <0.001

IQR, interquartile range; HT, hemorrhagic transformation; HI, hemorrhagic infarction; PH, parenchymal hemorrhage; BMI, body mass
index; NIHSS, Institutes of Health Stroke Scale; SBP, systolic blood pressure; DBP, diastolic blood pressure; TOAST, the trial of ORG 10172 in
acute stroke treatment; LDH, lactate dehydrogenase; LDL, low-density lipoprotein; PT, prothrombin time; INR, international normalized
ratio; PTR, prothrombin time ratio.

percentages, and their significance was determined using the χ2 regression analysis due to their clinical importance in HT [2, 25,
test or Fisher’s exact test. The receiver operating characteristic 26]. SPSS version 21.0 and R (version 4.0.3) were used for all
(ROC) curve was used to assess the discrimination ability of statistical analyses.
LDH level in predicting HT in total patients, patients receiving
conventional therapy, and patients undergoing thrombolysis or
thrombectomy therapy. Multivariate logistic regression analysis Results
was used to determine the impact of LDH levels contributing to
HT, with the confounders being controlled. The multivariate lo- Comparisons between HT and Non-HT Patients
gistic regression analysis adjusted confounders with the p < 0.05
in Table 1, including BMI, AF, NIHSS score, TOAST criteria, As shown in Figure 1, 4,389 patients were screened.
infarct location, platelet count, LDL, PT, and INR. Additionally, Nine hundred sixty patients were eliminated, and 3,429
age, gender, and hypertension (all p > 0.05) were included in the patients were initially enrolled in this research. We used

Prediction of Hemorrhagic Gerontology 2023;69:571–580 575


Transformation DOI: 10.1159/000528951
Table 3. Multivariate logistic model of the
clinical determinants of HT Variable OR (95% CI) p value

BMI 1.030 (0.967–1.097) 0.354


Hypertension 0.824 (0.594–1.145) 0.249
AF 1.656 (0.829–3.309) 0.153
NIHSS 1.069 (1.021–1.119) 0.004
TOAST
Large artery atherosclerosis 0.987 (0.683–1.427) 0.944
Cardioembolism 1.109 (0.529–2.324) 0.784
Small vessel occlusion 0.530 (0.119–2.371) 0.406
Other subtypes Reference
Infarct location
Lobar 0.640 (0.434–0.944) 0.024
Subcortical 0.087 (0.056–0.135) <0.001
Brainstem 0.051 (0.019–0.132) <0.001

Downloaded from http://karger.com/ger/article-pdf/69/5/571/3949659/000528951.pdf by Univ. of California Davis user on 01 April 2024


Cerebellum 1.806 (0.873–3.736) 0.111
Mixed type Reference
Therapy
Conventional therapy 0.027 (0.003–0.223) 0.001
Thrombolysis or thrombectomy Reference
Platelet count 0.999 (0.996–1.001) 0.302
LDL 1.212 (1.007–1.459) 0.042
PT 2.056 (0.806–5.244) 0.131
INR 0.001 (0.000–7.205) 0.126
PTR 0.739 (0.463–1.178) 0.203
LDH
LDH >215 10.958 (7.964–15.078) <0.001
LDH ≤215 Reference

OR, odds ratio; CI, confidence level; HT, hemorrhagic transformation; BMI, body mass
index; NIHSS, Institutes of Health Stroke Scale; TOAST, the trial of ORG 10172 in acute stroke
treatment; LDH, lactate dehydrogenase; LDL, low-density lipoprotein; PT, prothrombin
time; INR, international normalized ratio; PTR, prothrombin time ratio.

the propensity score matching approach at a ratio of 1:2 Comparative Analysis of Subtypes
to match 1,091 age- and gender-matched AIS patients The subgroup analysis of patients with HT is shown
without HT to 542 consecutive patients with HT. in Table 2. There were 351 patients with HI (351/542,
Differences in the demographic, clinical, and laboratory 64.76%) and 191 patients with PH (191/542, 35.24%). A
features between HT and non-HT patients are shown in significant difference (p < 0.001) was found in LDH lev-
Table 1. LDH levels were significantly higher in the HT els among the non-HT, HI, and PH groups with the
group than in the non-HT group (263.0 [216.0–323.3] U/L Kruskal-Wallis test (Table 2), and the LDH level was sig-
versus 178.0 [162.0–195.0] U/L, p < 0.001), as shown in nificantly higher in the PH group than in the HI group
Table 1 and Figure 3a. Patients with a higher BMI, a history after the Bonferroni modification (281.0 [230.0–340.0]
of AF, no history of hypertension, higher NIHSS scores, a U/L versus 258.0 [209.0–311.0] U/L, p < 0.001) (Fig. 3b).
higher percentage of cardioembolism, treatment with In addition, there were also differences in BMI, hyper-
thrombolysis or thrombectomy therapy, a lower platelet tension, AF, NIHSS score, TOAST criteria, therapy, in-
count, a higher LDL, more prolonged PT, a higher INR, and farct location, platelet count, LDL, PT, INR, and PTR
a lower PTR were more likely to develop HT (all p < 0.05). (all p < 0.05).
In addition, there were differences in infarct location be-
tween the HT and non-HT groups (p < 0.001). However, Associations between LDH and HT
these groups had no statistical differences regarding age, As seen in Figure 4, the area under the ROC curve
gender, smoking, drinking, diabetes, systolic blood pressure, (AUC) of LDH in predicting the occurrence of HT in to-
diastolic blood pressure, and creatinine level (all p > 0.05). tal AIS patients was 0.890 (95% CI: 0.874–0.905, p <

576 Gerontology 2023;69:571–580 Chen/Xu/Huang/Wen/Ren


DOI: 10.1159/000528951
0.001). With a cut-off value of 215 U/L in total patients, A recent study reported that elevated serum LDH lev-
the prediction of HT was the best: 63.91% accuracy, els could predict unfavorable outcomes following recom-
88.63% specificity, and 75.28% sensitivity in total pa- binant tissue plasminogen activator thrombolysis in pa-
tients. Patients were then classified into high-LDH tients with AIS [22]. While no relationship between LDH
(LDH >215 U/L) and low-LDH (≤215 U/L) groups and symptomatic HT was found in that study, this dispar-
based on the best ROC cut-off value. As seen in Table 3, ity may be attributable to their limited sample size of pa-
multivariable logistic regression analysis revealed that tients with symptomatic intracranial hemorrhage (n =
those with a high LDH level (LDH >215 U/L) were more 28). This study preliminarily explored and compared the
prone to developing HT after adjusting the confounding predictive value of LDH for HT in patients with and
variables with p < 0.05 (BMI, hypertension, AF, NIHSS without thrombolytic or thrombectomy therapy. The re-
score, TOAST criteria, therapy, infarct location, platelet sults showed that the predictive value of conventional
count, LDL, PT, INR, and PTR) (odds ratio [OR] = treatment was much higher than that of thrombolysis or
10.958, 95% CI: 7.964–15.078, p < 0.001). Furthermore, thrombectomy therapy. However, the sample size of pa-

Downloaded from http://karger.com/ger/article-pdf/69/5/571/3949659/000528951.pdf by Univ. of California Davis user on 01 April 2024


the findings remained significant when LDH was added tients receiving thrombolytic or thrombectomy therapy
as a continuous variable in the multivariable logistic re- was relatively limited (n = 43), which may limit the cred-
gression model (p < 0.001). ibility of the results. In the future, we will expand the
Further analysis was conducted on the patients receiv- sample size to further explore the predictive value of
ing conventional therapy or thrombolytic and thrombec- LDH on HT in patients with thrombolytic or thrombec-
tomy therapy. The results showed that the AUC of LDH tomy therapy.
in patients with conventional therapy was 0.888 (95% CI: LDH, mainly found in the cytoplasm and mitochon-
0.872–0.904, p < 0.001), and the results of the multivariate dria of numerous organs, plays a vital role in different
logistic regression analysis were consistent with that in diseases [8, 12, 27, 28]. LDH releases into the extracellular
the total patients (OR = 17.268, 95% CI: 12.054–24.739, p space upon tissue injury, resulting in an elevated serum
< 0.001). In patients with thrombolytic or thrombectomy LDH concentration. Since it reflected cardiomyocyte ne-
therapy, there were 43 patients (43/224, 19.20%) with HT crosis, LDH was historically used to diagnose myocardial
and 181 patients (181/224, 80.80%) without HT. Further infarction [12, 29]. Additionally, LDH is a generic sign of
analysis showed that the AUC of LDH in patients with acute or chronic tissue injury and is considered an in-
thrombolysis or thrombectomy therapy was 0.714 (95% flammatory marker [30]. Infection with COVID-19,
CI: 0.650–0.773, p < 0.001). As seen in online Supple- chronic obstructive pulmonary disease, acute lung injury,
mentary Figure 1(for all online suppl. material, see www. and pulmonary embolism have been reported to be asso-
karger.com/doi/10.1159/000528951), the predictive value ciated with elevated blood LDH levels [30–32]. Previous
of LDH in conventional treatment was much higher than research has shown that LDH was not only associated
that in patients with thrombolysis or thrombectomy ther- with activating certain proto-oncogenes but also played a
apy (AUC: 0.888 vs. 0.714, p < 0.001). While after adjust- critical role in the maintenance of tumor invasiveness,
ing for confounding factors in the multivariate logistic metastatic potential, chemoresistance, and radioresis-
regression analysis, no significant correlation was found tance [6, 33–35]. Besides, LDH is involved in immuno-
between LDH and HT in patients receiving thrombolysis suppression [34, 35]. LDH may be used as a prognostic
or thrombectomy therapy (OR = 2.152, 95% CI: 0.806– marker for immune surveillance, and the aberrant eleva-
5.745, p = 0.126), which may be owing to the small sample tion of LDH was predictive of poor outcomes in cancer
size of patients with HT (n = 43). patients [35].
The above research results suggested that serum LDH
has substantial predictive value in various disorders of
Discussion other systems. Previous studies revealed that LDH also
played a role in central nervous system diseases [9, 10,
To the best of our knowledge, this is one of the few 16, 18]. According to a survey based on proteomics anal-
studies comprehensively evaluating the effects of LDH ysis, serum LDH was found to independently predict the
levels on HT after AIS. In the current study, we found that formation of early hematomas in individuals with spon-
LDH was significantly linked with the occurrence of HT taneous intracerebral hemorrhage [10]. Another study
in patients with AIS, and patients with LDH levels >215 revealed that increased LDH level was statistically asso-
U/L had a higher risk of HT. ciated with hematoma expansion [9]. Data from the

Prediction of Hemorrhagic Gerontology 2023;69:571–580 577


Transformation DOI: 10.1159/000528951
third China National Stroke Registry (CNSR-III), a na- Conclusion
tionwide prospective registry with 9,796 patients who
submitted to 201 medical institutions in 22 provinces LDH may be a biomarker of HT in individuals with
and four municipalities with AIS or TIA, demonstrated AIS. HT risk should be noted when AIS individuals show
that an increased LDH level is independently related to a high level of LDH (LDH >215 U/L).
adverse outcomes, including all-cause death and high
modified Rankin Scale scores at 3 months and 1 year in
patients with AIS or TIA [16]. Furthermore, higher se- Acknowledgments
rum LDH levels were found in electrocoagulation-in-
We express our gratitude to all the patients who participated in
duced thrombotic focal ischemic rats receiving an intra- our study. Key Laboratory of Intelligent Treatment and Life Sup-
venous infusion of tissue plasminogen activator 6 h after port for Critical Diseases of Zhejiang Province, Wenzhou 325000,
ischemia. By modulating LDH and endogenous metabo- Zhejiang, People’s Republic of China.
lites, pinocembrin prevents HT after delayed tissue plas-

Downloaded from http://karger.com/ger/article-pdf/69/5/571/3949659/000528951.pdf by Univ. of California Davis user on 01 April 2024


minogen activator therapy in thromboembolic stroke
rats [18]. Statement of Ethics
As shown in online supplementary Figure 2, there
were various reasonable explanations for the relation- This retrospective review of patient data did not require written
informed consent from participants in accordance with local/na-
ship between LDH and HT. First, structurally, research tional guidelines. The Medical Ethics Committee of the First Affili-
showed that the breakdown of the blood-brain barrier is ated Hospital of Wenzhou Medical University authorized this re-
the fundamental mechanism resulting in blood extrava- search with the registration number KY2021-R077. The study com-
sation [36, 37]. The release of LDH by cell disruption plied with the guidelines for human studies and was conducted
may lead to the destruction of the blood-brain barrier ethically in accordance with the World Medical Association Decla-
ration of Helsinki.
and the triggering of HT through inflammatory re-
sponse, vasodilation, and oxidative stress. For inflam-
matory response, LDH has been touted as a potential Conflict of Interest Statement
biomarker for inflammatory loads. Its inhibitors have
anti-inflammatory properties and play a role in oxida- The authors declare that the research was conducted without
tive stress cascade, associated with the destruction of the any commercial or financial relationships construed as a potential
blood-brain barrier, leading to HT [37–40]. For vasodi- conflict of interest.
lation, LDH catalyzes the conversion of glucose to lac-
tate acid, which has been shown to cause vasodilation
Funding Sources
via various pathways [41–43]. For oxidative stress, LDH
weakens the integrity of the basal layer and tight endo- The present study was supported by the Zhejiang Provincial
thelial connections through oxidative stress cascades Natural Science Foundation of China (Grant No. LQ22H020003;
and may be the trigger of HT [25, 37, 44]. Secondly, after to LL.C.) and Wenzhou’s Science and Technology Bureau Project
the occurrence of HT, a series of cascade reactions fur- (Grant No. Y2020934; to LL.C.).
ther promote cell necrosis and release more LDH [19,
21, 22]. The above reasons may be the cause of elevated
LDH in HT patients. However, in vitro and in vivo ex- Author Contributions
periments are needed to further clarify the relationship Wenwei Ren engaged in the conception and design of the re-
between LDH and HT in the future. search. Lingli Chen, Minjie Xu, and Qiqi Huang gathered the clinical
This research had several limitations. First, this was data. Lingli Chen was in charge of data analysis and paper writing.
a retrospective investigation, and a causal link between Caiyun Wen reviewed and modified the manuscript. All authors
LDH and HT could not be established. Second, since contributed to the article and approved the submitted version.
the patients were from a single institution in China, ad-
ditional confirmation of the findings in other areas is
required. Third, this study only measured the LDH lev- Data Availability Statement
el once at admission. As LDH levels may fluctuate over The raw data supporting the conclusions of this article will be
time, it is preferable to test LDH more times during made available by the authors without undue reservation. Further
hospitalization. inquiries can be directed to the corresponding author.

578 Gerontology 2023;69:571–580 Chen/Xu/Huang/Wen/Ren


DOI: 10.1159/000528951
References
1 Okada Y, Yamaguchi T, Minematsu K, Miyashita dehydrogenase complex and lactate dehy- acute ischaemic stroke (ECASS II). Second
T, Sawada T, Sadoshima S, et al. Hemorrhagic drogenase are targets for therapy of acute European-Australasian Acute Stroke Study
transformation in cerebral embolism. Stroke. liver failure. J Hepatol. 2018; 69(2): 325– Investigators. Lancet. 1998; 352(9136): 1245–
1989;20(5):598–603. 35. 51.
2 Alvarez-Sabin J, Maisterra O, Santamarina 14 Anan M, Nagai Y, Fudaba H, Fujiki M. Lac- 25 Hong JM, Kim DS, Kim M. Hemorrhagic
E, Kase CS. Factors influencing haemor- tate and lactate dehydrogenase in cistern as transformation after ischemic stroke: mech-
rhagic transformation in ischaemic stroke. biomarkers of early brain injury and delayed anisms and management. Front Neurol.
Lancet Neurol. 2013; 12(7): 689–705. cerebral ischemia of subarachnoid hemor- 2021; 12: 703258.
3 Yaghi S, Willey JZ, Cucchiara B, Goldstein rhage. J Stroke Cerebrovasc Dis. 2020; 29(5): 26 Hong L, Hsu TM, Zhang Y, Cheng X. Neu-
JN, Gonzales NR, Khatri P, et al. Treatment 104765. roimaging prediction of hemorrhagic trans-
and outcome of hemorrhagic transforma- 15 Park JS, You Y, Ahn HJ, Min JH, Jeong W, formation for acute ischemic stroke. Cere-
tion after intravenous alteplase in acute is- Yoo I, et al. Cerebrospinal fluid lactate dehy- brovasc Dis. 2022; 51(4): 542–52.
chemic stroke: a scientific statement for drogenase as a potential predictor of neuro- 27 Passarella S, de Bari L, Valenti D, Pizzuto R,
healthcare professionals from the American logic outcomes in cardiac arrest survivors Paventi G, Atlante A. Mitochondria and L-
heart association/American stroke associa- who underwent target temperature manage- lactate metabolism. FEBS Lett. 2008; 582(25–

Downloaded from http://karger.com/ger/article-pdf/69/5/571/3949659/000528951.pdf by Univ. of California Davis user on 01 April 2024


tion. Stroke. 2017; 48(12): e343–61. ment. J Crit Care. 2020; 57: 49–54. 26): 3569–76.
4 Paciaroni M, Agnelli G, Corea F, Ageno W, 16 Wang A, Tian X, Zuo Y, Wang X, Xu Q, 28 Kline ES, Brandt RB, Laux JE, Spainhour SE,
Alberti A, Lanari A, et al. Early hemorrhagic Meng X, et al. High lactate dehydrogenase Higgins ES, Rogers KS, et al. Localization of
transformation of brain infarction: rate, pre- was associated with adverse outcomes in pa- L-lactate dehydrogenase in mitochondria.
dictive factors, and influence on clinical out- tients with acute ischemic stroke or transient Arch Biochem Biophys. 1986; 246(2): 673–
come: results of a prospective multicenter ischemic attack. Ann Palliat Med. 2021; 80.
study. Stroke. 2008; 39(8): 2249–56. 10(10): 10185–95. 29 Peters T Jr, Davis JS. Serum heat-stable lac-
5 Young A, Oldford C, Mailloux RJ. Lactate 17 Wu LW, Kao TW, Lin CM, Yang HF, Sun YS, tate dehydrogenase in the diagnosis of myo-
dehydrogenase supports lactate oxidation in Liaw FY, et al. Examining the association be- cardial infarction. JAMA. 1969; 209(8):
mitochondria isolated from different mouse tween serum lactic dehydrogenase and all- 1186–90.
tissues. Redox Biol. 2020; 28: 101339. cause mortality in patients with metabolic 30 Masumoto A, Kitai T, Matsumoto S, Kuroda
6 Tjokrowidjaja A, Lord SJ, John T, Lewis CR, syndrome: a retrospective observational S, Kohsaka S, Tachikawa R, et al. Impact of
Kok PS, Marschner IC. Pre- and on-treat- study. BMJ Open. 2016; 6(5): e011186. serum lactate dehydrogenase on the short-
ment lactate dehydrogenase as a prognostic 18 Kong LL, Gao L, Wang KX, Liu NN, Liu C, term prognosis of COVID-19 with pre-exist-
and predictive biomarker in advanced non- Ma GD, et al. Pinocembrin attenuates hem- ing cardiovascular diseases. J Cardiol. 2022;
small cell lung cancer. Cancer. 2022; 128(8): orrhagic transformation after delayed t-PA 79(4): 501–8.
1574–83. treatment in thromboembolic stroke rats by 31 Shi J, Li Y, Zhou X, Zhang Q, Ye X, Wu Z, et
7 Datta S, Chakrabarti N. Age related rise in regulating endogenous metabolites. Acta al. Lactate dehydrogenase and susceptibility
lactate and its correlation with lactate dehy- Pharmacol Sin. 2021; 42(8): 1223–34. to deterioration of mild COVID-19 patients:
drogenase (LDH) status in post-mitochon- 19 Cummings BS, Schnellmann RG. Measure- a multicenter nested case-control study.
drial fractions isolated from different re- ment of cell death in mammalian cells. Curr BMC Med. 2020; 18(1): 168.
gions of brain in mice. Neurochem Int. 2018; Protoc. 2021; 1(8): e210. 32 Wu MY, Yao L, Wang Y, Zhu XY, Wang XF,
118: 23–33. 20 Figueroa EG, Gonzalez-Candia A, Caballe- Tang PJ. Clinical evaluation of potential use-
8 Rho JM. Inhibition of lactate dehydrogenase ro-Roman A, Fornaguera C, Escribano-Fer- fulness of serum lactate dehydrogenase
to treat epilepsy. N Engl J Med. 2015; 373(2): rer E, Garcia-Celma MJ. Blood-brain barrier (LDH) in 2019 novel coronavirus (CO-
187–9. dysfunction in hemorrhagic transformation: VID-19) pneumonia. Respir Res. 2020; 21(1):
9 Wang CY, Zhang YB, Wang JQ, Zhang XT, a therapeutic opportunity for nanoparticles 171.
Pan ZM, Chen LX. Association between se- and melatonin. J Neurophysiol. 2021; 125(6): 33 Van Wilpe S, Koornstra R, Den Brok M, De
rum lactate dehydrogenase level and hema- 2025–33. Groot JW, Blank C, De Vries J, et al. Lactate
toma expansion in patients with primary 21 Arba F, Rinaldi C, Caimano D, Vit F, Busto dehydrogenase: a marker of diminished an-
intracerebral hemorrhage: a propensity- G, Fainardi E. Blood-brain barrier disrup- titumor immunity. Oncoimmunology. 2020;
matched analysis. World Neurosurg. 2022; tion and hemorrhagic transformation in 9(1): 1731942.
160: e579–90. acute ischemic stroke: systematic review 34 Hermans D, Gautam S, Garcia-Canaveras
10 Chu H, Huang C, Dong J, Yang X, Xiang J, and meta-analysis. Front Neurol. 2020; 11: JC, Gromer D, Mitra S, Spolski R, et al. Lac-
Dong Q. Lactate dehydrogenase predicts 594613. tate dehydrogenase inhibition synergizes
early hematoma expansion and poor out- 22 Jin H, Bi R, Hu J, Xu D, Su Y, Huang M, et with IL-21 to promote CD8(+) T cell stem-
comes in intracerebral hemorrhage pa- al. Elevated serum lactate dehydrogenase ness and antitumor immunity. Proc Natl
tients. Transl Stroke Res. 2019; 10(6): 620–9. predicts unfavorable outcomes after rt-PA Acad Sci U S A. 2020; 117(11): 6047–55.
11 Buckner SL, Loenneke JP, Loprinzi PD. thrombolysis in ischemic stroke patients. 35 Wulaningsih W, Holmberg L, Garmo H,
Cross-sectional association between nor- Front Neurol. 2022; 13: 816216. Malmstrom H, Lambe M, Hammar N, et al.
mal-range lactate dehydrogenase, physical 23 Adams HP Jr, Bendixen BH, Kappelle LJ, Serum lactate dehydrogenase and survival
activity and cardiovascular disease risk Biller J, Love BB, Gordon DL. Classification following cancer diagnosis. Br J Cancer.
score. Sports Med. 2016; 46(4): 467–72. of subtype of acute ischemic stroke. Defini- 2015; 113(9): 1389–96.
12 Wu Y, Lu C, Pan N, Zhang M, An Y, tions for use in a multicenter clinical trial. 36 Bernardo-Castro S, Sousa JA, Bras A, Cecilia
Xu M, et al. Serum lactate dehydrogenase TOAST. Trial of Org 10172 in Acute Stroke C, Rodrigues B, Almendra L, et al. Patho-
activities as systems biomarkers for 48 types Treatment. Stroke. 1993; 24(1): 35–41. physiology of blood-brain barrier permea-
of human diseases. Sci Rep. 2021; 11(1): 24 Hacke W, Kaste M, Fieschi C, von Kummer bility throughout the different stages of is-
12997. R, Davalos A, Meier D, et al. Randomised dou- chemic stroke and its implication on hemor-
13 Ferriero R, Nusco E, De Cegli R, Carissimo ble-blind placebo-controlled trial of thrombo- rhagic transformation and recovery. Front
A, Manco G, Brunetti-Pierri N. Pyruvate lytic therapy with intravenous alteplase in Neurol. 2020; 11: 594672.

Prediction of Hemorrhagic Gerontology 2023;69:571–580 579


Transformation DOI: 10.1159/000528951
37 Arba F, Piccardi B, Palumbo V, Biagini S, that exacerbate hemorrhagic transforma- 42 Gordon GRJ, Choi HB, Rungta RL, Ellis-Da-
Galmozzi F, Iovene V, et al. Blood-brain bar- tion in stroke. Circ Res. 2021; 128(1): 62–75. vies GCR, MacVicar BA. Brain metabolism
rier leakage and hemorrhagic transforma- 40 Desilles JP, Syvannarath V, Ollivier V, dictates the polarity of astrocyte control over
tion: the Reperfusion Injury in Ischemic Journe C, Delbosc S, Ducroux C, et al. Exac- arterioles. Nature. 2008; 456(7223): 745–9.
StroKe (RISK) study. Eur J Neurol. 2021; erbation of thromboinflammation by hyper- 43 Hein TW, Xu W, Kuo L. Dilation of retinal
28(9): 3147–54. glycemia precipitates cerebral infarct growth arterioles in response to lactate: role of nitric
38 Spronk E, Sykes G, Falcione S, Munsterman and hemorrhagic transformation. Stroke. oxide, guanylyl cyclase, and ATP-sensitive
D, Joy T, Kamtchum-Tatuene J. Hemorrhag- 2017; 48(7): 1932–40. potassium channels. Invest Ophthalmol Vis
ic transformation in ischemic stroke and the 41 Yamanishi S, Katsumura K, Kobayashi T, Sci. 2006; 47(2): 693–9.
role of inflammation. Front Neurol. 2021; 12: Puro DG. Extracellular lactate as a dynamic 44 Wu H, Wang Y, Ying M, Jin C, Li J, Hu X. Lactate
661955. vasoactive signal in the rat retinal microvas- dehydrogenases amplify reactive oxygen species
39 Shi K, Zou M, Jia DM, Shi S, Yang X, culature. Am J Physiol Heart Circ Physiol. in cancer cells in response to oxidative stimuli.
Liu Q, et al. tPA mobilizes immune cells 2006; 290(3): H925–34. Signal Transduct Target Ther. 2021;6(1):242.

Downloaded from http://karger.com/ger/article-pdf/69/5/571/3949659/000528951.pdf by Univ. of California Davis user on 01 April 2024

580 Gerontology 2023;69:571–580 Chen/Xu/Huang/Wen/Ren


DOI: 10.1159/000528951

You might also like