Keywords
severe traumatic brain injury - progesterone - biochemical marker - outcome
Introduction
Traumatic brain injury (TBI) is one of the leading causes of injury-related death
and severe disability worldwide, resulting in large direct and indirect costs to society.
The management of TBI currently includes preventing further neurological insults,
managing the intracranial pressure (ICP), and surgical procedures.[1] Despite advances in research and improved neurological intensive care in recent
years, the clinical outcome of severely head injured patients is still poor. Evaluation
of direct and indirect costs reveals that TBI is a 60 billion dollar “industry” in
the United States.[2]
[3]
[4]
[5] In contrast to the primary insult, secondary brain injury evolves over time. These
are characterized by a complex cascade of molecular and biochemical events that lead
to neuroinflammation, brain edema, and delayed neuronal death.[6] Apart from the evidence-based guidelines for treatment of TBI published by Brain
Trauma Foundation,[7] no specific pharmacological therapy is currently available which prevents the development
of secondary brain injuries, and most therapeutic strategies have failed in translation
from “bench to bedside”.[8] The recent disappointing results of the Progesterone for the Treatment of Traumatic
Brain Injury (ProTECT) and Study of a Neuroprotective Agent, Progesterone, in Severe
Traumatic Brain Injury (SyNAPSe) phase III trials for progesterone in TBI is one example
of such failed translation from bench to bedside.[9] Progesterone showed considerable efficacy as a neuroprotective agent in the treatment
of TBI in animal studies[10]
[11]
[12]
[13]
[14] and two moderately encouraging phase II clinical reports, but two similarly designed
phase III TBI clinical trials recently resulted in negative outcomes.[15]
[16]
[17]
[18] Pending the validation of the role of progesterone in ameliorating the secondary
injury in TBI, two recent prospective studies have concluded that serum progesterone
levels may have a role as biochemical marker to independently predict the global outcome
in cases of TBI.[19]
[20] We conducted a prospective observational study of a cohort of 100 patients with
severe TBI (Glasgow coma scale [GCS] score ≤ 8 at arrival) and analyzed the predictive
value of this biomarker (serum progesterone) on two validated and dichotomized measures
of outcome: Glasgow outcome scale (GOS) score and functional independence measure
(FIM). We found that admission serum progesterone levels significantly (p value ≤ 0.05) predicted the outcome on dichotomized (favorable vs. unfavorable, dependent
vs. independent) measures of outcome at 3 months.
Materials and Methods
Study Design
We conducted a prospective observational statistician blinded outcome assessor study
on a cohort of patients who had suffered severe TBI at the time of admission to SKIMS,
which is the tertiary care center for the management of TBI in the region of J&K.
The cohort included all the eligible patients of severe TBI presented to the Dept.
of Neurosurgery SKIMS between November 2013 to November 2015.
Eligibility
A total of 100 male and female patients between 16 and 70 years of age presenting
within 12 hours of suffering closed head injury with GCS score of 4 to 8 (on a scale
from 3 to 15), assessed after resuscitation, a TBI diagnosed by history and examination,
visible pathologic changes consistent with TBI as assessed on the admission CT scan,
and at least one reactive pupil were included in the study. Patients were excluded
if they were less than 16 years or greater than 70 years of age, had a GCS score of
3, bilaterally fixed and dilated pupils, a life expectancy of less than 24 hours,
prolonged or uncorrectable hypoxemia (partial pressure of arterial oxygen, < 60 mm
Hg), hypotension (systolic blood pressure, < 90 mm Hg) at the time of admission, spinal
cord injury, pregnant and lactating females, only an isolated epidural hematoma, or
coma that was suspected to be due primarily to other causes or comorbid illnesses
like diabetes mellitus, ischemic heart disease or known malignancy.
Study Procedure
Peripheral blood (≠ 2.5 mL) was collected from the sample drawn at the time of presentation
for general investigations in sterile serum collection tubes and allowed to clot.
The tubes were then subjected to centrifugation at 2500 rpm for 5 minutes to separate
the serum. The serum was aspirated and poured in 1.5 mL microfuge tubes after proper
labeling and stored at − 20°C, which were then subjected to the enzyme-linked immunosorbent
assay (ELISA) technique for estimation of serum progesterone levels.
General treatment and the treatment of raised ICP were in accordance with published
international guidelines.[7]
Outcome Measures
GOS score was used as the measure of primary outcome. The scale has five levels: 1
(death), 2 (vegetative state), 3 (severe disability), 4 (moderate disability), and
5 (good recovery). For statistical clarity, the GOS was further dichotomized into
favorable outcome (GOS 4 and 5) and unfavorable outcome (GOS 1, 2 and 3).
The FIM scale was used as secondary measure of outcome which assesses physical and
cognitive disability. FIM is scored from 1 to 7 based on level of independence: 7
complete independence–fully independent; 6 modified independence–requiring the use
of a device but no physical help; 5 supervision–requiring only standby assistance
or verbal prompting or help with set-up; 4 minimal assistance–requiring incidental
hands-on help only (subject performs > 75% of the task); 3 moderate assistance–subject
still performs 50 to 75% of the task; 2 maximal assistance–subject provides less than
half of the effort (25–49%); 1 total assistance–subject contributes < 25% of the effort
or is unable to do the task.
For statistical clarity, the FIM scale was further dichotomized into independent outcome
(FIM scale of 6 and 7) and dependent outcome (FIM scale of 1–5).
Statistical Analysis
The baseline parameters, serum progesterone levels, and outcome are presented as either
number/percentage, mean (± standard deviation [SD]), or median (interquartile [IQR]),
wherever appropriate. Statistical analysis of serum progesterone level and dichotomized
outcome was performed using the t-test considering p < 0.005 to be significant.
Ethical Considerations
The blood samples were drawn as a part of general investigations on arrival to emergency
department, with 2.5 mL saved for the study. The study was a prospective observational
study, so there were no ethical issues related to the study.
Results
Demographic Profile
A total of 100 patients suffering from severe TBI (GCS score 4–8) with a mean age
of 41.20 ± 16.20 years mean, and GCS score of 6.48 ± 1.61 at admission were analyzed.
The demographic profile of the study is shown in [Table 1].
Table 1
Demographic profile (n = 100)
|
Abbreviations: DAI, diffuse axonal injury; GCS, Glasgow coma scale; IQR, interquartile
range; SAH, subarachnoid hemorrhage.
|
|
Age (16–70 years)
|
|
|
Mean ± SD
|
41.20 ± 16.20
|
|
IQR
|
29
|
|
Sex (male)
|
90 (90)
|
|
Cause of injury (%)
|
|
|
Road traffic accidents
|
58
|
|
Falls
|
32
|
|
Assaults
|
10
|
|
Average time since injury (min)
|
151.90 ± 86.43 (range: 15–420 min)
|
|
GCS score at admission
|
|
|
Mean ± Y
|
6.48 ± 1.61
|
|
IQR
|
3
|
|
Polytrauma (%)
|
53
|
|
CT findings
|
|
|
Contusions
|
54
|
|
Acute subdural hematoma
|
30
|
|
Fractures
|
25
|
|
SAH
|
20
|
|
DAI
|
5
|
|
Extradural hematoma
|
5
|
|
Decompressive craniectomy (%)
|
48 (48)
|
|
Length of hospital stay (days)
|
7.96 ± 7.01 (range: 1–32 days)
|
|
Mortality rate at 3 months (n = 100)
|
30%
|
Analysis of Outcome
The GOS score of the study cohort at 3 months is shown in [Table 2]. The favorable outcome is seen in 60 of the 100 patients, whereas only 40 had an
unfavorable outcome. The mean serum progesterone levels were 2.48 ± 1.97 ng/mL. The
dichotomized GOS score ([Table 3]) showed a mean serum progesterone level of 7.55 ± 2.55 ng/mL in favorable outcome
group (GOS 4–5) and a mean level of 2.58 ± 1.04 in unfavorable outcome group (GOS1–3).
On statistical analysis, we found that the difference of mean serum progesterone level
was significantly different between the two groups (p < 0.0001), thus predicting a favorable outcome (GOS 4–5) at 3 months in patients
with higher mean serum progesterone levels at admission.
Table 2
GOS score at 3 months
|
GOS score
|
No. of patients (n = 100)
|
Serum progesterone levels
Mean ± SD (ng/mL)
|
|
Abbreviation: GOS Glasgow outcome scale,.
|
|
1
|
30
|
0.78 ± 0.50
|
|
3
|
10
|
1.99 ± 0.23
|
|
4
|
32
|
2.95 ± 2.08
|
|
5
|
28
|
3.93 ± 1.80
|
|
Total
|
100
|
2.48 ± 1.97
|
Table 3
Dichotomized GOS score at 3 months versus serum progesterone levels
|
GOS score (n = 100)
|
No. of patients
|
Serum progesterone levels
Mean ± SD (ng/mL)
|
p-Value
|
|
Abbreviation: GOS, Glasgow outcome scale.
|
|
Favorable[4], [5]
|
60 (60%)
|
7.88 ± 2.55
|
0.000
|
|
Unfavorable[1]
[2]
[3]
|
40 (40%)
|
2.58 ± 1.04
|
The FIM score of the study cohort at 3 months is shown in [Table 4]. The independent outcome is seen in 42 of the 70 patients, whereas only 28 had an
unfavorable outcome. The mean serum progesterone levels were 3.2 ± 1.91 ng/mL. The
dichotomized FIM score ([Table 5]) showed a mean serum progesterone level of 3.81 ± 2.27 ng/mL in independent outcome
group (FIM 6–7) and a mean level of 1.51 ± 0.88 ng/mL in dependent outcome group (FIM
1–5). On statistical analysis, we found that the difference of mean serum progesterone
level was significantly different between the two groups (p < 0.001), thus predicting functionally an independent outcome (FIM 6–7) at 3 months
in patients with higher mean serum progesterone levels at admission.
Table 4
FIM score at 3 months
|
Score
|
No. of patients (n = 70)
|
Serum progesterone levels Mean ± SD (ng/mL)
|
|
Abbreviation: FIM, functional independence measure.
|
|
2
|
1
|
1.60
|
|
3
|
3
|
1.89 ± 0.20
|
|
4
|
10
|
2.1 ± 0.25
|
|
5
|
14
|
2.6 ± 0.26
|
|
6
|
22
|
4.4 ± 2.95
|
|
7
|
20
|
3.2 ± 0.92
|
Table 5
Dichotomized FIM score at 3 months versus serum progesterone levels
|
FIM score (n = 70)
|
No. of patients
|
Serum progesterone levels
Mean ± SD (ng/mL)
|
p-Value
|
|
Abbreviation: FIM, functional independence measure.
|
|
Independent[6], [7]
|
42 (60%)
|
3.81 ± 2.27
|
< 0.001
|
|
Dependent[1]
[2]
[3]
[4]
[5]
|
28 (40%)
|
1.51 ± 0.88
|
Note was also made that none of the patients with serum progesterone levels less than
1 ng/mL (n = 20) at admission survived.
Discussion
TBI is one of the leading causes of injury-related death and severe disability worldwide,
resulting in large direct and indirect costs to society.[1] Apart from primary injury, the secondary brain injury is the result of complex interplay
of brain tissue-specific antigens, hormonal imbalances, and cytokine-mediated humoral
and cellular immune reactions,[21]
[22]
[23] eventually leading to a complex cascade of molecular and biochemical events that
lead to neuroinflammation, brain edema, and delayed neuronal death, adding to the
quantum of severity of brain tissue damage.[24] Due to limitations of clinical and radiological methods in delineating the exact
severity and prognosis of TBI, there has been a considerable paradigm shift toward
the development and use of biochemical markers to delineate the extent of brain tissue
damage and to independently predict the outcome.[19]
Dhandapani et al found that unfavorable neurological outcome at 6 months was associated
with 15% fall of serum albumin or low urinary creatinine at 3 weeks.[25] Takahashi et al identified fibrinolytic parameters, α2-plasmin inhibitor-plasmin complex (PIC) levels, and D-dimer levels as admission prognostic
markers of head injury in patients and found that higher levels predicted worse prognosis.[26] Specific brain-tissue antigens like S100-B,[27] neuron-specific enolase (NSE),[28]
[29] and glial fibrillary acidic protein (GFAP), have been studied extensively over the
past decade in various laboratory and clinical studies as potential markers of severity
of brain injury. Similarly, the role of proinflammatory (interleukin [IL]-6, IL-1,
tumor necrosis factor–a [TNF-a]) and anti-inflammatory (IL-4, IL-10, transforming
growth factor) cytokines have been studied in recent trials.[30]
[31]
[32] Many of these markers either have a very short half-life, have issue of specific
window period, serum levels are affected by other confounding factors like sepsis,
hypo-perfusion, liver damage, etc., have poor blood brain penetration requiring cerebrospinal
fluid (CSF) sampling, or studies have shown ambiguous results.[19] A recent study by Raheja et al concluded that out of NSE, GFAP, S100-B, TNFα and
IL 6, only serum GFAP and IL6 levels at day 7 of injury could significantly predict
long-term outcome in severe TBI at 1 year.
Recently, prospective studies have concluded that serum progesterone levels may have
a role as biochemical marker to independently predict the global outcome in cases
of TBI. Progesterone, a gonadal hormone and neurosteroid naturally distributed in
human brains, has potent neuroprotective properties.[8] Progesterone is produced by the ovaries and the placenta and predominantly considered
as female hormone. In the brains of both sexes, progesterone synthesized by oligodendrocytes
and some neurons plays an important role in neuronal development during gestation.[33]
[34] Progesterone receptors are widely expressed in the developing and adult brain, so
various brain regions are the normal targets of progesterone.[35] Indeed, the 10-fold increase in progesterone synthesis during fetal growth is considered
an evolutionary mechanism of progesterone to protect the fetus during gestation. Because
many processes involved in CNS repair after brain injury are thought to recapitulate
what occurs during normal brain development,[36] progesterone may be actively involved in TBI recovery and thus involved in neuroprotection.[8] Progesterone is thought to mediate its neuroprotective effects by reducing cerebral
edema, lipid peroxidation, isoprostanes, and expression of proinflammatory genes;
generating metabolites that reduce proapoptotic and increase antiapoptotic enzymes;
and modifying the expression of vascular endothelial growth factor, brain-derived
neurotrophic growth factor, and aquaporins responsible for development of edema.[37] In literature, more than 200 articles have reported the neuroprotective effects
of progesterone in four species, including humans, and in 22 different injury models,
including TBI, stroke, spinal cord injury and neurodegenerative disorders.[24] Also two phase II clinical reports from a level 1 trauma centers in Atlanta USA
in 2001 and Hangzhou, China, in 2004, respectively showed encouraging results for
progesterone as a neuroprotective agent in TBI.[17]
[18] In contrast, recently two similarly designed phase III TBI clinical trials failed
to demonstrate any positive effect of progesterone as a neuroprotective agent on outcomes.[15]
[16] It also triggered considerable speculation about the reasons for the negative outcomes
of these two studies in particular and for those of all previous phase II TBI clinical
trials in general.[9]
Despite contradictory results of progesterone as a neuroprotective agent, two reports[19]
[20] have shed light on value of serum progesterone at admission in predicting the outcome
of TBI patients. Our study also showed a very significant role of serum progesterone
levels as a biomarker of outcome on outcomes in TBI patients.
Dasgupta et al[20] in a prospective observational study of 100 male patients sought to investigate
whether lower levels of baseline serum progesterone at admission within 10 hours of
injury can be used to predict worse outcome in severe-to-moderate TBI patients (GCS
4–12), as measured by GOS and Functional FIM scores at 30 days postinjury. Mean age
was 35 ± 11.4 years, average time taken to reach the hospital was 3.14 ± 1.8 hours,
and index GCS score was 4 to 8 in 84 patients and 9 to 12 in 16 patients. The mean
progesterone level in the group with poor outcome (GOS 1–3) was 0.89 ± 033 ng/mL and
the mean progesterone level in the group with good outcome (GOS 4–5) was 1.03 ± 0.18
ng/mL. Despite the mean serum progesterone levels being higher in patients with good
outcome (GOS 4–5), they did not find this difference statistically significant (p value 0.16). In contrast, when the FIM score was analyzed, the mean progesterone
level in the group with poor outcome (FIM 1–4) was 0.9 ± 0.45 ng/mL and the mean progesterone
level in the group with good outcome (FIM 5–7) was 1.27 ± 0.32 ng/mL, which was a
statistically significant difference (p value 0.048). So, their findings suggested potential application of serum progesterone
levels as an admission prognostic marker in moderate-to-severe male TBI patients.
Our study had female patients (10%) as well and included only patients who had suffered
severe TBI (GCS 4–8). We used GOS score as a measure of global outcome (favorable
vs. unfavorable) and FIM score as measure of functional outcome (independent vs. dependent)
and found highly significant difference in the mean serum progesterone levels for
both measures of outcome respectively whereas Dasgupta et al[20] used good and poor outcomes for both measures. We think that since Dasgupta et al[20] included patients with moderate TBI (GCS 9–12), and this might be responsible for
not getting statistically significant results for the measure of GOS scale for outcome,
as moderate TBI is a distinct entity. Similar to their results, we also made note
of extremely poor outcome (GOS 1) for patients with serum progesterone levels less
than 1 ng/mL.
Raheja et al[19] analyzed 86 patients (84.9% of them male) with severe TBI (GCS 4–8) with mean age
32.3 years as a part of a prospective, outcome-assessor, and statistician-blinded,
randomized, placebo-controlled phase II trial of progesterone with or without hypothermia.
The average ICP measured serially for 5 days was 11.8 mm Hg. Decompressive craniectomy
was performed for refractory intracranial hypertension in 32.6% of patients. Favorable
GOS scores were found in 71.4% of patients at 6 months and in 78.5% of patients at
12 months. Also, 63.3% and 81.5% of patients were functionally independent at 6 and
12 months, respectively. The mortality rate was 14.3% and 16.9% at the 6- and 12-month
follow-up, respectively. In multivariate logistic regression analysis, they found
that day 7 serum progesterone levels were independent predictors of unfavorable outcome
at 1 year and admission IL-6 and Day 7 GFAP were independent predictors of favorable
outcome at 1 year. They concluded that serial serum progesterone, GFAP, and IL-6 monitoring
could aid in prognosticating outcomes in patients with severe TBI. Thus, progesterone
had a significant role in predicting long-term favorable outcome. They also suggested
whether a cause-and-effect relationship or a mere association of these biomarkers
to outcome needs to be further studied for better understanding of the pathophysiology
in severe TBI and for choosing potential therapeutic targets.
Our study found admission serum progesterone levels a very significant predictor of
favorable global (GOS) as well functional outcomes. In contrast, Raheja et al[19] found only day 7 serum progesterone levels to be a significant predictor of favorable
outcome. So, whether serial measurements of serum progesterone levels from day 1 to
day 7 will give a better picture is only speculative. Another difference was the number
of patients who underwent decompressive craniectomy for raised ICP. In our study group,
more patients underwent surgical intervention (48%) as compared with study done by
Raheja et al (32.6%).[19] The reason could be lack of ICP monitoring in our cohort and depending only on clinical
and radiological findings for surgical decisions. Also, hypothermia and therapeutic
progesterone administration was used in a good proportion of patients. The same reasons
could explain a higher mortality in our patients (30%). Overall, the results imply
the significant role of serum progesterone levels as a predictor of global as well
as functional outcome in severe TBI.
We speculate that serum progesterone levels are a reflection of degree of suppression
of hypothalamic-pituitary-gonadal axis, the degree of which depends on the severity
of trauma, which ultimately influences the outcome.
Conclusion
Serum progesterone levels have a potential to serve as biomarker of outcome in cases
of TBI. Low serum progesterone levels at admission predict bad overall as well as
functional outcomes. Whether single measurements (day 1 vs. day 7) or serial measurements
(day 1 to day 7) have better predictive values has to be validated in further larger
studies.
