Jonathan M. Hernandez, Resident, Department of Surgery; Michael H. Albrink, Associate Professor of Surgery, Department of Surgery; Emmanuel E. Zervos, Assistant Professor of Surgery, Department of Surgery; Alexander S. Rosemurgy, Professor of Surgery and Medicine, Department of Surgery, University of South Florida, Tampa General Hospital, Tampa, FL
Jonathan M. Hernandez, MD
Resident Department of Surgery
Michael H. Albrink, MD
Associate Professor of Surgery
Department of Surgery
Emmanuel E. Zervos, MD
Assistant Professor of Surgery
Department of Surgery
Alexander S. Rosemurgy, MD
Professor of Surgery and Medicine
Department of Surgery
University of South Florida
Tampa General Hospital
Objective: To determine whether the empiric administration of an intramuscular preparation of an anabolic steroid would improve clinical outcomes in severely injured adults. Mortality was the primary end point. Duration of mechanical ventilation and hospitalization were secondary end points.
Methods: The study included 140 consecutive, prospectively randomized trauma patients who had Injury Severity Scores greater than 19. Saline placebo or nandrolone decanoate was administered in a double-blind manner 72 hours after injury and each week thereafter until discharge from the hospital or death. Survival, duration of ventilator dependence, and duration of stay in the intensive care unit (ICU) and hospital were determined.
Results: Patients receiving nandrolone had no signiﬁcant improvement in survival or decreased duration of stay in the ICU or hospital. Members of the subgroup requiring mechanical ventilation who received nandrolone did not spend any fewer days on a ventilator than those who did not receive nandrolone.
Conclusion: Empiric administration of nandrolone after severe injury did not decrease the duration of ICU or hospital stays and did not affect ventilator dependence or mortality. The use of anabolic steroids in broad groups of severely injured patients is of no beneﬁt.
The catabolic response to severe injury can be profound, resulting in rapid loss of lean body mass and nutritional depletion. This breakdown of muscle protein is associated with poor wound healing, increased rates of infection, and higher mortality.1 Aggressive protein supplementation alone has proven ineffective at ameliorating the catabolic state.2
Anabolic steroids have been used in combination with nutritional support in an attempt to curb the ﬂux of metabolites through catabolic pathways. Several authors have associated steroid administration in the trauma setting with decreased nitrogen and amino acid losses, increased concentration of total plasma amino acids, and improved body composition.3-5 Anabolic steroids also have been shown to increase net protein synthesis in severely burned children.6,7 This occurred in delayed and acute settings, with the latter being associated with gene upregulation.
Despite the apparent nutritional success with anabolic steroids in the acute and rehabilitative phases of recovery, no study to date has shown improvement in clinical outcomes in the acute setting for critically injured adult patients. A randomized prospective trial by Gervasio and associates failed to show any quantiﬁable beneﬁt, whether clinical or nutritional, to administering oxandrolone in the acute trauma setting.8 Oxandrolone also was tested in patients requiring mechanical ventilation.9 In that trial, patients receiving oxandrolone remained on the ventilator signiﬁcantly longer than those receiving the placebo.
Despite the lack of clinical success with oral anabolic therapy in the literature, our prospective, double-blind, randomized trial sought to determine the effects of an intramuscularly administered anabolic steroid that is continued until discharge or death in adult patients with severe injury. We chose an intramuscular preparation—nandrolone decanoate—to ensure adequate delivery and circumvent the absorption difﬁculties that are inherent to severely injured adults. Our hypothesis was that nandrolone would improve survival and result in shorter hospital stays.
We conducted a prospective, randomized, double-blind, placebo-controlled trial of nandrolone therapy in 140 consecutive severely injured adult patients admitted to a university-afﬁliated level 1 trauma center. The university institutional review board approved the study, and informed consent was obtained from each patient before enrollment. Pregnancy, a patient age of younger than 20 years, and current use of anticoagulant therapy were immediate exclusion criteria. Patients were included in the study if they had an Injury Severity Score (ISS) greater than 19. The ISS was assigned by a trauma audit committee on admission and again at discharge or upon death. No patient had to be excluded from the study based on his or her ﬁnal ISS, such as an ISS lower than 19.
Patients were randomized using a random numbers table to receive either placebo (0.9% normal saline intramuscularly) or nandrolone. Nandrolone was administered 3 days after the injury at a dose of 2 mg/kg, followed by a dose of 1 mg/kg 1 week after the injury and a dose of 0.5 mg/kg at the beginning of each subsequent week until death or discharge from the hospital. All patients received supplemental nutrition early in their hospital stay, per protocol. Enteral feedings were used when possible to support caloric and protein needs. Total parenteral nutrition was used when enteral feeding was contraindicated or impractical.
Of the 140 patients enrolled in the study, 67 received the placebo and 73 received nandrolone. Patients in both groups were similar with regard to age, sex, mechanism of injury, and ISS and trauma scores (Table 1). All patients were followed prospectively for clinical outcome, including survival, length of stay in the intensive care unit (ICU), duration of ventilator dependence, and overall duration of hospital stay.
38 ± 13.3
36 ± 12.7
30 ± 9.3
30 ± 9.2
13.1 ± 3.1
13.0 ± 3.2
*Data presented as mean ± standard deviation where appropriate.
Data were collected on a spreadsheet, and statistical analyses were undertaken using the True Epistat Software Package. Comparisons within treatment groups were performed using unpaired analysis. Data are reported as the mean ? standard deviation, where appropriate.
Mortality rates were similar in both groups (Table 2). However, fatally injured patients who received nandrolone had higher ISS scores on admission than their placebo counterparts (43 ? 17.1 vs 32 ? 12.7; = 0.233). There was no difference between the two groups in duration of ICU stay or overall number of days in the hospital for surviving patients; thus, survival and duration of hospital stay were unaffected by the administration of nandrolone.
A subgroup of 67 trauma patients required mechanical ventilation during their hospital stay; 37 (55%) were randomized to receive nandrolone and 30 (45%) to receive the placebo. No statistical differences were recorded between either of these subgroups in terms of days on mechanical ventilation. Patients receiving nandrolone tended to require fewer days of mechanical ventilation; however, this ﬁnding did not achieve statistical signiﬁcance (Table 2).
Mortality, n (%)
12 ± 11.1
12 ± 12.1
Total hospital stay (days)
33 ± 24.5
32 ± 28.5
Mechanical ventilation (days)
11 ± 14.8
6 ± 5.3
*Data presented as mean ± standard deviation where appropriate.
Although improved nutritional indices and wound healing have been demonstrated acutely with anabolic steroid therapy in burn patients, their role in aiding other trauma patients, if any, remains ambiguous. Two previous studies have failed to demonstrate any clinical improvement with oral oxandrolone therapy over placebo.10,11 These trials used an oral formulation in a patient population with compromised functionality of the gastrointestinal tract. Our study is the largest trial to assess the use of anabolic steroids in the acute trauma setting and the ﬁrst to use an intramuscular formulation to avoid issues of absorption that typically arise in severely injured patients. We used nandrolone because it promotes the development of blood and bone, enhances splanchnic fuel metabolism, has a long half-life, and has fewer androgenic side effects compared with testosterone.12-17 The doses of nandrolone used in this study were chosen based on comparable doses of testosterone shown to be safe in a previous study.18
With mortality as the primary end point, we sought to include 70 patients in each arm of the study, anticipating an approximate mortality rate of 20% (16 patients) for the placebo. We hypothesized that the mortality rate would decrease by at least 50% among those receiving nandrolone decanoate versus those receiving the placebo; however, the mortality rates were similar in both groups, with 6 deaths occurring in each arm of the study. Interestingly, if the anticipated 16 deaths had occurred in the placebo group, a statistically signiﬁcant reduction in mortality would have been observed for those patients receiving nandrolone. It is difﬁcult to postulate what contributed to the low mortality rate observed among patients receiving the placebo. Perhaps it was due in part to the dedicated nursing and critical care staff in our level 1 trauma center.
The average ISS of fatally injured patients treated with nandrolone was higher than that for fatally injured patients receiving the placebo, documenting objectively that the nandrolone group included more severely injured patients. The implications of this latter ﬁnding are difﬁcult to interpret in the context of the overall survival data.
The average patient in this study was an approximately 40-year-old man who had sustained serious blunt traumatic injury during a motor vehicle accident. On average, patients spent more than 1 month in the hospital, and nearly half of that period was spent in the ICU. No signiﬁcant differences were observed between those receiving anabolic steroids or the placebo with regard to duration of hospital stay and number of days spent in the ICU. Our data, in this respect, are in agreement with the trial of 60 patients randomized to receive either the placebo or oral oxandrolone.8
Assessment of the duration of days spent on mechanical ventilatory support suggests that, overall, those patients who received nandrolone spent fewer days on a ventilator. Although this reduction is an interesting observation, it did not achieve statistical signiﬁcance in our study. This ﬁnding also does not correspond to the results of an earlier study by Bulger and associates, which was interrupted before completion due to the seemingly poor outcomes of those receiving oral oxandrolone compared with those receiving the placebo.9
Several inherent limitations to the current study should be noted. We sought to measure clinical outcomes as an end point, which are difﬁcult to affect with pharmacotherapy. Many uncontrollable variables culminate to determine a given patient's duration of hospital stay or mortality, which would not be affected by a single pharmacologic intervention. The objective of our study was to determine whether anabolic steroids should be applied to broad, diverse groups of patients, such as those sustaining traumatic injuries.
We chose to randomize 140 patients and anticipated a 20% mortality rate for patients in the placebo group, which was consistent with their ISS. Instead we observed an unexpected reduction in mortality in the placebo group, such that 20,000 patients would have required randomization to obtain a statistically signiﬁcant reduction, based on the mortality rates we encountered.
While quantitative analyses of nitrogen loss, catabolism, and nutritional indices in other studies have shown numerical improvement in patients treated with anabolic steroids, our study primarily assessed the efﬁcacy of anabolic steroids by analyzing outcome. In this broader framework, we observed no advantages in empirically administrating nandrolone to severely injured patients with regard to overall survival, duration of ventilatory support, or duration of ICU and hospital stays, and the use of anabolic steroids in this setting should be abandoned. Future studies of anabolic steroids should be limited to very speciﬁc patient populations in which clinical success has been documented previously.