Abstract
To the Editor:
We think it is important to understand under what conditions progesterone (or any other drug) will work as a potential treatment for traumatic brain injury (TBI). In a recent paper in this journal, Allitt and colleagues 1 report that they tested progesterone in a model of severe diffuse TBI and found no benefit of the treatment. The authors induced this injury in rats by closed-skull, weight drop, impact acceleration technique. The skull was exposed and a metal disc (1 cm in diameter, 3 mm thick) was fixed to the skull between bregma and lambda with dental acrylic. A 450-g weight was then dropped from a height of 2 m through a vertical tube positioned above the rat's head onto the metal disc. “The impact velocity was calibrated to 6.15 m/sec to induce severe injury.” Allitt and associates cite Marmarou and colleagues 2 as the basis for their model with some modifications. With a 450-g weight dropped from a height of 2 m, the Marmarou study reported mortality levels between 44% and 58.6%, depending on anesthesia and respiration parameters. In the Allitt study, if there was any mortality from the injury, it was not reported so it was not clear what specific procedures might have been followed to reduce this high level of mortality.
Following the brain injury, the short-term treatment group received 16 mg/kg progesterone in 0.6 mL of peanut oil. Doses were administered within 1 h after injury and then again at 6 h and 24 h. After a 4-day recovery interval, electrophysiological recordings were obtained from sensory motor cortex. For the longer course of study, after the three injections on day 1, animals received injections at 1 week and then once a week until the recording session on the 8th week post-surgery. The authors do not provide a rationale for this treatment approach of one injection/week. “In the long-term group, sensorimotor function was assessed before and after surgery, using three behavioral tests. The rotarod and beam-walk tasks were primarily used to assess motor function and injury severity and were conducted on the 2 days prior to surgery and then from day 1 post-surgery daily for 1 week, then once per week for the next 7 weeks until terminal electrophysiology 8 weeks post-surgery” (p. 376). Electrophysiological recordings were taken after whisker stimulation to measure neuronal activation in the barrel field of the sensory cortex. The TBI essentially was associated with short-term suppression of supragranular layer activity and long-term hyperexcitability.
Progesterone was reported to exacerbate the effects of the injury in the short term and conferred no long-term benefits in re-establishing normal neuronal activity as measured by the unit recordings in the different laminae. We were surprised that the authors omitted any histological confirmation of the size and extent of the “very severe injury” they inflicted with their closed-head impact model. As others have noted previously, the injury parameters Allitt and colleagues report will almost certainly produce extensive diffuse cerebral damage, including skull fracture and intracranial and subdural bleeding. This is likely to create widespread injury to cortical and subcortical structures throughout the brain with an accompanying systemic inflammatory response and persistent behavioral impairments. Under these conditions, it is not surprising that there is a serious dysfunction and lack of a treatment effect.
We learn from this study that progesterone as applied and tested in such a severe injury model did not yield positive results. To draw relevant conclusions about treatment benefits from this article, it would have to provide mortality data or at least some histological indication of lesion size and locus. Under the circumstances and given the very high mortality rates reported in the Marmarou article, it is reasonable to think that the lack of any progesterone beneficial effects may have been due to the fact that the sensory and morphological deficits were simply too devastating to be effectively treated with progesterone or any other drug thus far tested for TBI. Further, without any histological verification of injury severity or other biomarkers of inflammation, there is nothing to suggest what mechanisms of action could have been manipulated to produce any evidence of a salutary effect in such a severe injury condition.
Allitt and collaborators propose that the changes they observed in neuronal excitability are likely the result of chronic administration of progesterone leading to the activation and downregulation of GABAA receptor expression, and this could lead to hyperactivity in the long term. We wonder whether a strategy of only one injection of progesterone/week could induce such relatively enduring changes seen in the sham-operated rats as well as in those with the severe TBI. Others who have examined unit activity in various areas of the hippocampus after injury and progesterone treatment report that progesterone receptor characteristics alter hippocampal synaptic activity and neurogenesis, 4 and that acute or chronic treatment with progesterone given by subcutaneous pellets could lead to the modification of hippocampal gene expression, but these studies used more chronic dosing, over a 4-day and a 30-day treatment regimen, than that used in the Allitt study. 5 Interestingly, another recent report indicates that combination therapy with estradiol and progesterone can reduce hippocampal cell death caused by ischemic injury both in cell culture and in live animal models of damage. 6 In these recent studies and in contrast to Allitt and colleagues' study, progesterone was shown to be beneficial.
We think it is possible that the very severe diffuse injury used by Allitt and associates could lead to pathological neurogenesis in the hippocampus resulting in increased seizure activity appearing as hyperexcitability. 7,8 Under these conditions it is not surprising that neither short- nor long-term sensory recovery was seen because: (1) the long-term part of the study used too long a delay between the first and second round of progesterone treatments, and (2) one progesterone treatment/week may simply not be enough to confer benefit. Our laboratory has found that, even with a much less severe model of TBI, 3 days compared with 5 days of treatment afford no significant beneficial results on spatial learning performance. 9 Finally, with respect to injury to the sensory motor cortex, over 15 years ago, using more typical dosing parameters of 5 days of daily treatment beginning 1 h after contusion injury, progesterone injections did not improve sensory and motor functions, indicating that there may be important regional differences in the brain's response to progesterone for neuroprotection. 10 Based on the evidence provided by Allitt and collaborators and for reasons not yet known, it may be a fact that progesterone treatment after sensory cortex injury does not produce beneficial outcomes. We can speculate that different injury loci and severities may require different dosing and duration of treatment schedules to be effective—and this may certainly be the take-home message of the Allitt article. However, at this stage of our knowledge, generalizing their negative findings under extreme injury conditions and with undefined brain pathology to every instance of progesterone application for central nervous system (CNS) injury would require the rejection of now hundreds of papers confirming its beneficial effects. 11
From the study by Allitt and collaborators we can learn much about: the complexities of unique dosing schedules, when and how best to measure behavioral recovery, hormonal actions and receptor mechanisms, and their effects on CNS plasticity under normal and abnormal conditions. The article also highlights how important it is in testing any neurosteroid treatment, or any other drug for that matter, to find the most appropriate pathology as well as to optimize the dosing, duration, and timing of treatment to translate to effective clinical evaluation. The same careful selection procedures and parameters should also apply to the design and implementation of clinical trials. 12