Patients with acute severe TBI and a Glasgow Coma Scale (GCS) score ≤ 8 after resuscitation and stabilization were entered into the study. Two hundred and thirty patients from the Neurotrauma Center of our teaching hospital were included. Male or female patients between the ages of 18 and 65 years were studied. The patients received progesterone within 8 hours after the documented time of injury. All patients admitted to the Neurotrauma Center, Clinical Medical College of Hangzhou between March 2004 and February 2007 were consecutively eligible for enrollment.

We excluded patients who had received any investigational drugs 30 days prior to the enrollment, such as progesterone, estrogen and investigational compound, patients with severe anoxic intracerebral damage or brain death, and patients whose clinical condition was unstable (partial pressure of oxygen < 60 mmHg or a systolic blood pressure < 90 mmHg, or both). We also excluded pregnant patients and lactating female patients, and those for whom there was doubt whether the neurological status resulted from head trauma or acute or chronic spinal cord injury.

The study was conducted in compliance with the clinical protocol approved by the Institutional Review Board and the ethical committees of Clinical Medical College of Hangzhou, according to Good Clinical Practice standards. Because of the nature of patients' injuries, subjects in this clinical study were incapable of granting informed consent. Investigators therefore obtained informed consent from the subject's legal guardian or health proxy before administering the drug. Given the urgent circumstances, we were unable to obtain permission from a legal guardian or health proxy within the stipulated time window for some patients (n = 53). Investigators therefore sought approval from the Institutional Review Board to use deferred consent. If the Institutional Review Board determined that these regulatory criteria were satisfied, the investigators were permitted to enroll subjects without consent. When the drug was administered without proxy consent, the Institutional Review Board was notified within 2 working days. We continued to try to contact the proxy consent after drug administration, and documented those attempts to the Institutional Review Board. Once contacted, the family or legally authorized representative was notified of the patient's enrollment and asked to provide written approval for the patient's continued participation. If attempts to contact proxy consent were unsuccessful, or if the patient died before the family could be contacted, we notified the Institutional Review Board and placed a full report in the patient's record and study file.

After head computerized tomography scanning, the patients were delivered to the neurosurgical intensive care unit of the teaching hospital immediately or following surgical evacuation of an intracranial hematoma. All patients received the standard treatment for management of severe TBI based on the guidelines for the management of severe head injury of the American Association of Neurologic Surgeons 20. Particular emphasis was placed on the prevention and treatment of secondary insults, the avoidance of intracranial hypertension, maintenance of a normovolemic state as well as normothermia and normoglycemia, with ventilation to maintain the oxygen pressure at a minimum of 100 mmHg and the carbon dioxide pressure at approximately 35 mmHg.

The prospective, randomized, placebo-controlled, double-blind study was conducted in our neurosurgical intensive care unit. Subjects enrolled in the study were randomized to receive either progesterone (Tianjing Jinyao Pharmaceutical Co. Ltd, Tianjing, China) or matching placebo within 8 hours of the documented time of injury. Qualifying patients were randomly assigned in a 1:1 manner to receive the matching treatment with random numbers. Patients for the treatment group were given progesterone at 1.0 mg/kg via intramuscular injection and then once per 12 hours for 5 consecutive days. A single-dosage volume equivalent to 0.05 ml/kg was used in each subject. In a double-blind manner, progesterone and placebo were supplied via identical-looking solutions in identical glass vials with or without progesterone. The appearance, packaging and administration of placebo and progesterone injections were the same for the two groups. All patients, treating physicians, nursing staff, and pharmacists were blinded throughout the study period.

The ICP was monitored continuously using ICP monitoring apparatus (CAMINO. MPM-1; Integra Co., San Diego, CA, USA). A computerized tomography scan was obtained in all patients at admission and this was categorized according to the modified Marshall computerized tomography scan classification: I, intracranial pathology not visible on the computerized tomography scan; II, cisterns present with shift ≤ 5 mm; lesions present, but no high-density or mixed-density lesions > 25 cm³, with bone fragments and foreign bodies; III, cisterns compressed or absent, shift ≤ 5 mm, with no high-density or mixed-density lesions > 25 cm³; IV, shift > 5 mm, with no high-density or mixed-density lesions >25 cm³; V, any surgically evacuated lesion; and VI, high-density or mixed-density lesions >25 cm³ without surgical evacuation.

The patient's condition – body temperature, heart rate and respiratory rate, blood pressure, and pulse blood oxygen saturation – was monitored continuously at the bedside with monitoring apparatus (Hewlett-Packard, Palo Alto, CA, USA). Daily evaluations of neurologic status over the initial 14-day period were performed via the GCS score, adverse experiences, surgical procedures, and intracranial complications. Intake and output of fluids were also recorded.

Laboratory tests including hematology, the coagulation profile and clinical chemistry were performed daily and then for 1 week after injury. A urine pregnancy test was performed at enrollment for female patients (as necessary).

The neurologic outcome was evaluated according to the Glasgow Outcome Scale (GOS) score, which contains five levels of outcome: good recovery, moderate disability, severe disability, vegetative survival, or death. For statistical analysis, GOS scores were dichotomized into favorable or unfavorable outcomes. Patients in the upper two GOS outcome groups (good recovery and moderate disability) were considered of favorable outcome, and patients in the other groups (severe disability, vegetative state, or death) were considered of unfavorable outcome.

Secondary efficacy endpoints were the modified Functional Independence Measure (FIM) score and mortality. Based on previous reports 2122, the modified FIM measurements of disability in three areas of activity (domains of self-care, motor function, and cognitive function) were chosen from the 18 items in the full FIM. Each of three items (expression, feeding, and locomotion) includes four possible levels of function ranging from total dependence (1) to independence (4). The total modified FIM scores therefore ranged from 3 to 12. The patients were assessed using the same measures both at 3 and 6 months in the follow-up protocol.

Descriptive statistics, including proportions, means and standard deviations, were compiled for all demographic and outcome measures. Demographic and clinical data were analyzed using Fisher's exact test. The statistical analyses were conducted to assess the differences between the treatment group and the control group on specific variables. Statistical analysis was performed using contingency analysis (chi-squared) for categorical data and Student's t test for continuous data. P < 0.05 was considered statistically significant. SPSS 11.0 software package (SPSS Inc., Chicago IL, USA) was used for statistical analysis.

Between March 2004 and February 2007, a total of 230 patients were screened in the present study. Of these, 159 patients meeting the protocol stipulation and condition were recruited and randomized to receive either progesterone (n = 82) or placebo (n = 77). Data were available for 154 patients (96%) at the 3-month follow-up and for 135 patients (84%) at the 6-month follow-up. Nineteen patients (11%) were lost to follow-up, three patients (1%) refused follow-up, and two patients (1%) withdrew from the trial. No subjects were enrolled in violation of the protocol stipulations (Figure 1).

The demographics of the progesterone and placebo groups are presented in Table 1. The cohorts were well balanced with no significant differences between the two groups. The medication history of patients, medication administration, and medical procedures were not significantly different among treatment groups.

The 3-month and 6-month GOS scores between the progesterone and placebo groups are summarized in Table 2. There was a better recovery rate for the patients who were given progesterone than for those in the control group at 3-month follow-up (P = 0.044). A dichotomized analysis revealed significant differences in neurologic outcome between the treatment and control groups (Figure 2). The analysis using the dichotomization of GOS scores at 3 months post injury revealed a favorable outcome in 47% of the patients receiving progesterone and in 31% of the placebo group (P = 0.034). There was an unfavorable outcome in 53% of the patients receiving progesterone and in 70% of the placebo group (P = 0.022). At 6-month follow-up, the dichotomized GOS scores also showed a significant statistical difference between the two groups, similar to those 3 months after injury. The percentage of favorable outcome was 58% for the patients who were given progesterone and was 42% in the placebo group (P = 0.048). Forty-one percent of patients who were given progesterone and 57% of the placebo group exhibited an unfavorable outcome (P = 0.048).

Subgroup analysis for women also showed a significant difference in the percentage of favorable outcome between the two groups at 6-month follow-up (35% in the placebo group and 66% in the progesterone group, P = 0.036). The patients who were given progesterone in the group with GCS of 6 to 8 showed a more favorable outcome (43%) compared with the placebo group (28%) at 6-month follow-up (P = 0.044). There was no significant difference, however, in dichotomized outcomes in the group with GCS of 3 to 5 (P > 0.05).

Figure 3 shows the modified FIM scores at 3-month and 6-month follow-up. There was a significant difference in the mean modified FIM score between two groups both at 3-month and 6-month follow-up. At the 3-month follow-up, the scores were 7.35 ± 1.89 for the placebo group and 8.02 ± 1.73 for the progesterone group (P < 0.05). Six months after injury, the placebo group showed a score of 8.95 ± 1.05 and the progesterone group presented 9.87 ± 1.17 (P < 0.01), suggesting good functional outcome in the patients treated with progesterone.

During the 6 months of follow-up, a total of 40 patients (25%) died in the present study (37 patients died during their hospital stay). Seventy percent of deaths occurred within 1 week after trauma. Mortality was attributed to the heavy head injury in each case. The mortality rate in the progesterone treatment group was significantly lower at 6-month follow-up compared with the placebo group (18% versus 32%, P = 0.039).

Figure 4 shows the ICP in the progesterone group patients and in the placebo group patients at 24 hours, 72 hours and 7 days after injury. The ICP was monitored continuously for 75 patients (47%), 40 in the progesterone group and 35 in the placebo group. The mean ICP shows no apparent difference at 24 hours after trauma between the two groups (progesterone group, 22.1 ± 4.3 mmHg versus placebo group, 23.2 ± 4.6 mmHg; P = 0.121). At 72 hours and 7 days after injury, the mean ICP of patients who were given progesterone was slightly lower than those of patients who received placebo, but the differences were not statistically significant (16.9 ± 3.8 mmHg and 14.8 ± 3.8 mmHg for progesterone-treated patients versus 18.2 ± 5.1 mmHg and 15.9 ± 4.1 mmHg for placebo-treated patients, respectively; P > 0.05).

The mean GCS scores increased progressively in the two groups during the 14-day acute phase of the study, with no apparent differences among the treatment groups. Meanwhile, there was no obvious difference in average body temperature, heart and respiratory rates, blood pressure, pulse blood oxygen saturation, and laboratory testing between the progesterone and placebo groups.

Progesterone was well tolerated in the treated patients with acute severe TBI. No complication and adverse event associated with the administration of progesterone was found in this clinical study during the hospitalization periods.