Evaluation of Serum Cortisol in the Postmortem Diagnosis of Acute Adrenal Insufficiency
Adam Clapper, MD, Marcus Nashelsky, MD, and Morris Dailey, MD, PhD
Abstract: Normal adrenocortical activity is necessary for electrolyte regulation and the maintenance of cardiovascular function. Although chronic adrenal insufficiency generally presents with the gradual onset of a set of characteristic symptoms and signs, the more sudden loss of adrenal activity can present with acute, rapidly progressive cardiovascular dysfunction that can be fatal if not recognized and treated promptly. We herein describe a patient who had most of his adrenal tissue removed during resection of metastatic renal carci- noma, conventional clear cell type, with much of the remaining adrenal tissue undergoing necrosis during or shortly after surgery. Although the patient appeared to be stable and progressing ade- quately well, he died suddenly 2 days postoperatively. When the gross autopsy findings suggested the possibility of adrenal insuffi- ciency, clinical laboratory assessment of adrenocortical activity was sought. Analysis of stored antemortem serum samples and of blood obtained at autopsy demonstrated a progressive decrease in cortisol levels which, in this stressed postsurgical patient, proved fatal. The use of both antemortem and postmortem blood in the demonstration of acute adrenal insufficiency at autopsy is discussed.
Key Words: acute adrenal insufficiency, autopsy, cortisol
(Am J Forensic Med Pathol 2008;29: 181-184)
C hronic adrenocortical insufficiency (AI) is a relatively rare syndrome that can be difficult to recognize clini- cally, requiring approximately 1 year to diagnose in half of the cases.1 Commonly, AI is symptomatic only in times of physiologic stress and the signs of AI can be nonspecific and subtle. The early signs and symptoms of chronic AI include weakness, fatigue, fever, and nausea. Hyperpigmentation may be seen in chronic primary AI. If not recognized, AI can progress to abdominal pain, mental status changes, hypoten- sion, and cardiovascular collapse.
In contrast to chronic AI, many of the early signs are absent when the onset of AI is acute, often making this diagnosis difficult when it arises in the absence of a previous history of the chronic syndrome. In addition, hyponatremia
and hyperkalemia, common early indicators of chronic AI, often do not occur before death ensues in rapidly progressive acute AI.2 Causes of acute AI include adrenal hemorrhage, such as that which occurs with sepsis, which can result in the rapid cessation of cortisol production. More commonly, how- ever, acute AI occurs in stressed patients in which the blood cortisol levels are within normal limits. In patients with conditions such as burns, sepsis, major trauma, or abdominal surgery, adequate physiological response to the severe stress requires a large increase in cortisol output and the absence of this increase can be fatal.2-4 There has been increased rec- ognition of the importance of glucocorticoid administration for preventing relative adrenal insufficiency in patients un- dergoing severe physiological stress.2,4,5
The autopsy diagnosis of acute AI as a cause of sudden death may not be considered unless it had been previously suspected clinically. Sometimes it is readily apparent in an infected patient in which adrenal hemor- rhage and necrosis are found at autopsy. Often, however, adrenal crisis is not accompanied by electrolyte abnormal- ities or other clinical signs of insufficiency2 or there is no antecedent history of chronic AI.
Consideration of AI may be especially important in the forensic setting, where many cases arrive with truncated, absent or unreliable medical histories, thereby requiring con- sideration of a broad differential diagnosis. In the absence of obvious internal findings, there are few clues to the presence of AI at autopsy except for the hyperpigmentation observed in some Addisonian patients. Hyperpigmentation, however, may be obscured by postmortem changes. The postmortem measurement of serum cortisol may be the only definitive way to assess these patients, but this test is not routinely performed in autopsy examinations. We present a case in which the measurement of cortisol in postmortem blood and in stored antemortem serum was essential to the evaluation of a sudden unexpected postoperative death of a patient follow- ing partial adrenalectomy.
CASE REPORT
A 47-year-old man presented to our institution in Feb- ruary 2003 complaining of left flank and lower back pain and hematuria for about 1 month. His medical history was sig- nificant for renal carcinoma, conventional clear cell type, of the right kidney with a single metastasis to the ipsilateral adrenal gland, for which a radical nephrectomy and adrenal- ectomy were performed in 1993. The surgery was performed at another institution and other details of the neoplasm and
From the Department of Pathology, University of Iowa, Iowa City, Iowa.
Reprints: Adam Clapper, MD, Department of Pathology, University of Iowa, Iowa City, IA 52242. E-mail: adam-clapper@uiowa.edu.
DOI: 10.1097/PAF.0b013e318174e7c8
baseline adrenal function were not known. Evaluation in 2003 revealed a 3.3 cm mass within the left adrenal gland. The morning serum cortisol at this time was 8.8 µg/dL (242.8 nmol/L) (normal: 4.3-22.4 µg/dL, 118.6-620.8 nmol/L). Serum electrolytes were within normal limits, as were urine norepinephrine, epinephrine, metanephrine, normetaneph- rine, and vanillylmandelic acid. It was decided to defer surgical resection to avoid possible AI. Approximately 12 months later a computed tomography scan revealed that the mass had increased in size to 4.6 cm. The patient and surgeon elected to proceed with surgical resection of the adrenal mass. The patient underwent surgery and tolerated the procedure well. The tumor was removed and approximately 5 g of viable adrenal gland were left in situ to provide adequate adrenal function. The resected mass contained metastatic renal carcinoma, conventional clear cell type; the surgical margins were free of tumor.
Postoperatively, the patient experienced intermittent tachycardia and tachypnea that were attributed to his consid- erable postsurgical pain. Vital signs were recorded on a regular basis. Postoperative blood pressure remained within normal limits. Blood drawn approximately 4 and 16 hours postoperatively (37 and 25 hours before death, respectively) contained normal levels of sodium and potassium; no addi- tional phlebotomy was performed. Cortisol levels were not measured and the patient was not given exogenous cortico- steroids.
The patient was reportedly awake and had a clear senso- rium 2 hours after vital signs were last recorded. Ninety- five minutes later the patient was found apneic and pulseless; resuscitation efforts were unsuccessful. Death occurred approx- imately 41 hours after conclusion of the adrenal tumor resection. The body was refrigerated within 4 hours of discovery of death. Blood was obtained from the femoral vessels and immediately frozen at the beginning of the autopsy, which commenced 25 hours after death.
Autopsy Findings
A complete autopsy revealed a 420 g heart (normal male, approximately 280-340 g) with mild biventricular hypertrophy, severe biventricular dilatation, and focal mild coronary artery atherosclerosis. There was no evidence of an acute or remote myocardial infarct and no pulmonary throm- boemboli were found. There was mild diffuse interstitial pulmonary fibrosis with retained architecture; inflammation or active fibrogenic foci were not present. There was severe pulmonary edema; the right and left lungs weighed 890 and 830 g, respectively. No central nervous system abnormalities were found.
No residual tumor was identified. There was recent soft tissue hemorrhage in the surgical bed and no remaining adrenal tissue could be identified grossly. However, multiple histologic sections of the hemorrhagic surgical bed revealed that the majority of the residual adrenal tissue had undergone hemorrhagic necrosis, with only a thin rim of viable tissue remaining (Fig. 1). Because some viable gland was still present, additional assessment of the patient’s antemortem adrenal function was sought. Serum remaining from the blood samples drawn 37 and 25 hours before death (stored
| TABLE 1. Antemortem and Postmortem Cortisol Levels* | ||
|---|---|---|
| Time Sample Obtained | Serum Cortisol (µg/dL) | Serum Cortisol (nmol/L) |
| 14 mo before surgery (morning) | 8.8 | 243 |
| 13 mo before surgery (afternoon) | 15.0 | 414 |
| 37 h PTD+ (4 h after surgery) | 2.56 | 70.6 |
| 25 h PTD (16 h after surgery) | 1.15 | 31.7 |
| Postmortem blood | 0.2 | 5.5 |
The antemortem serum specimens had been refrigerated within a few hours of phlebotomy. Upon notification of the patient’s death, all remaining antemortem serum samples were moved to the freezer, according to institutional protocol. The postmortem sample was collected 25 hours after death and was centrifuged and frozen before analysis.
*Normal values: morning: 4.3-22.4 µg/dL (118.6-620.8 nmol/L); afternoon: 3.1-16.6 µg/dL (85.5-458 nmol/L).
*PTD, prior to death.
frozen in the clinical chemistry laboratory) was retrieved and analyzed for cortisol. In addition, serum from postmortem femoral blood was also analyzed. The cortisol levels in the preoperative, postoperative, and postmortem samples are shown in Table 1.
The data in Table 1, together with the previous clinical history, indicate that adrenocortical activity after the 1993 adrenalectomy-nephrectomy had been normal. However, by 4 hours after the second procedure (resection of the left adrenal mass), cortisol had fallen to below normal, and this level declined further during the following 12 hours. These low values are particularly remarkable in a patient undergoing major postoperative stress. Cortisol at the time of death was barely detectable, as determined by analysis of the serum collected at autopsy. Thus, the viable adrenal tissue identified at autopsy was insufficient to provide almost any glucocor- ticoid activity. The findings of cardiac ventricular dilatation and pulmonary edema and the very low cortisol levels sug- gest that this patient died of acute cardiovascular failure in the setting of acute adrenal insufficiency.
DISCUSSION
This case illustrates the following 2 important points: (1) the suddenness with which an apparently stable postop- erative patient can die of acute adrenal insufficiency and (2) the potential usefulness of postmortem measurement of cor- tisol in assessing a patient who has died of an unexplained cause. In cases such as this, in which adrenal insufficiency had not been suspected clinically and at least some adrenal tissue is found at autopsy, the direct assessment of adrenal function is crucial in determining the cause of death. The availability of appropriately stored antemortem serum sam- ples can facilitate the retrospective analysis of many analytes. In addition, some analytes, including cortisol, are sufficiently stable to be measured in blood samples collected at autopsy. Cortisol in freshly prepared serum is stable for 1 day at room temperature, for 7 days when refrigerated, and indefinitely when frozen.6 As the antemortem samples in this case were drawn no more than 2 days before freezing, there was minimal likelihood of sample degradation.
After death, cortisol is stable within the body during the early postmortem period, before the onset of recognizable decomposition.7 Finlayson8 examined cortisol levels as a function of time after death and showed that cortisol is stable for at least 20 hours postmortem, the longest time interval tested; all of these bodies had been refrigerated within 4 hours of death. Although the postmortem interval in the current case was slightly longer (25 hours), it is very likely that the serum cortisol level in the blood collected at autopsy accu- rately represents the extremely low level of cortisol at the time of death and is consistent with the rapidly declining antemortem levels.
Postsurgical cortisol levels such as those seen in this patient would be interpreted as low even under conditions of low physiologic stress. Cortisol levels typically rise consid- erably as an adaptive response to stress and the degree of steroid elevation correlates with the severity of the illness or injury. In one study of 12 patients, uncomplicated cholecys- tectomy caused a rise in serum cortisol levels to a maximum of 46 ± 5.3 µg/dL (1268 ± 147 nmol/L) and 37.7 ± 0.7 ug/dL (1039.5 ± 169 nmol/L) (conventional and laparo- scopic, respectively) shortly after surgery.9 In another study of 9 patients who underwent major abdominal operations, the serum cortisol rose to 30 µg/dL (828 nmol/L) within 4 hours and did not return to baseline until 6 days postoperatively.10 Randomly obtained cortisol levels in severely stressed pa- tients (eg, postsurgical, hypoxemia, trauma, or pain) should be at least 25 µg/dL (690 nmol/L),2 although some authorities believe it should be even higher. Maintenance of normal adrenal responses in severely ill patients is critically impor- tant to outcome and even a small degree of adrenal insuffi- ciency adversely affects the mortality of such patients.2 Thus, a moderately elevated serum cortisol may represent an inad- equate response in the setting of extensive endogenous stres- sors, resulting in relative adrenal insufficiency which, like absolute insufficiency, requires rapid treatment with high doses of corticosteroids.5,11 Therefore, when considering a postmortem determination of acute AI, the absolute cortisol levels must be interpreted in conjunction with the severity of
the underlying stress, realizing that functional AI can be fatal even when the adrenal glands are not necrotic. Low absolute blood cortisol levels may be survivable in otherwise healthy persons, but severe insufficiency such as in this case is likely to be rapidly fatal in postsurgical patients.
The clinical syndrome of acute AI generally develops over the course of several hours. Initial findings commonly include severe abdominal pain, nausea, and fever. The pa- tient may then become tachycardie and hypotensive and, if AI is not recognized, cardiovascular collapse may follow. The resulting cardiovascular collapse can appear similar to hypo- volemic shock, with low preload and high peripheral vascular resistance. More commonly, however, there is high output heart failure in which there is low peripheral resistance.2,11 Such hyperdynamic cardiac failure may be mistaken as shock resulting from sepsis.
The early recognition of AI may be difficult clinically. The rapid progression of acute adrenal crisis to death is usually not heralded by the altered electrolyte levels observed in chronic insufficiency.4 Furthermore, abdominal pain and tachycardia are common in patients after abdominal surgery, thus masking some of the earliest symptoms of acute adrenal insufficiency. Two clues to the diagnosis of adre- nal crisis are hypoglycemia and hypotension unresponsive to fluids. The latter is at least partially attributable to the requirement of glucocorticoids for the maintenance of cardiac function, vascular tone, and blood pressure.2 The terminal course in the current patient was especially rapid. This may be explained by a sudden, near total loss of adrenal activity after the residual cortical tissue underwent necrosis, resulting in extremely low cortisol levels, com- bined with the severe stress.
Pathologists should include acute AI in the postmor- tem differential diagnosis of patients who die suddenly in the setting of stress or who had been chronically treated with exogenous corticosteroids. As this case illustrates, the ability to reliably quantitate the cortisol level in blood at autopsy, as well as in appropriately stored antemortem serum samples, can be essential in demonstrating both absolute and relative adrenal insufficiency as a cause of unexpected death.
ACKNOWLEDGMENTS
The authors gratefully acknowledge Dr. Ronald Feld for advice regarding cortisol measurements.
REFERENCES
1. Arlt W, Allolio B. Adrenal insufficiency. Lancet. 2003;361:1881-1893.
2. Marik PE, Zaloga GP. Adrenal insufficiency in the critically ill. A new look at an old problem. Chest. 2002;122:1784-1796.
3. Naito Y, Fukata J, Tamai S, et al. Biphasic changes in hypothalam pituitary-adrenal function during the early recovery period after major abdominal surgery. J Clin Endocrinol Metab. 1991;73:111-117.
4. Cooper MS, Stewart PM. Corticosteroid insufficiency in acutely ill patients. N Engl J Med. 2003;348:727-734.
5. de Herder WW, van der Lely AJ. Addisonian crisis and relative adrenal failure. Rev Endocr Metab Disord. 2003;4:143-147.
6. Wilding P, Silva JF, Wilde CE. Transport of specimens for clinical chemistry analysis. Ann Clin Biochem. 1977;14:301-306.
@ 2008 Lippincott Williams & Wilkins
7. Coe J. Time of death and changes after death, part II (chemical considerations). In: Spitz and Fisher’s Medicolegal Investigation of Death: Guidelines for the Application of Pathology to Crime Investiga- tion. Spitz WU, ed. Springfield, IL: Charles C. Thomas; 1993:59.
8. Finlayson NB. Blood cortisol in infants and adults: a postmortem study. J Pediatr. 1965;67:248-252.
9. Donald RA, Perry EG, Witter GA, et al. The plasma ACTH, AVP, CRH,
and catecholamine responses to conventional and laparoscopic chole- cystectomy. Clin Endocrinol. 1993;38:609-615.
10. Hamanaka Y, Manabe H, Tanaka H, Monden Y, Uozumi T, Matsumoto K. Effects of surgery on plasma levels of cortisol, corticosterone and non-protein-bound cortisol. Acta Endocrinol. 1970;64:439-451.
11. Lamberts SW, Bruining HA, de Jong FH. Corticosteroid therapy in severe illness. N Engl J Med. 1997;337:1285-1292.