uo~

ITHE ADRENAL RESPONSE TO TRAUMA, OPERATION')"',1'$

AND COSYNTROPIN STIMUlATION

MichaelJ. Harris, M.D., Richard T. Baker,J. William McRoberts, M.D.,

Lexington, Kentucky, andJames L. Mohler, M.D., ChapelHill, North Carolina

Thirtyjaur patients who underwent abdominal or tha-racic, or both, aperation for injuries were studied. Serumand urinary cortisollevels were elevated periaperativelyand returned to normallevels by the faurth postaperativeday. Postaperative complication was associated with per-sistent elevation of levels of serum and urinary cortisol.Baseline serum cortisol levels correlated with InjurySeverity Scores. Cosyntrapin stimulation produced asig!tificant rise in serum cortisol to a peak level of47.4:!: 10.2 (S.D.) micrograms per deciliter in patientsafter emergent operations for traumatic injury. Use ofthe cosyntroPin stimulation test for the diag!tosis ofpostaperative adrenal insufficienl) can be extended topatients who undergo aperation for traumatic injury.

THE ADRE:-.'AL RESPO:-.'SE of patients to fue str~ss oftrauma or operation has been characterize~ (1-5). In addition, the cosyntropin stimulatio* testhas be en performed and normal values ~stab-lished in patients who have undergone el~ctiveoperations (3,6,7). Although rare, postoperativeadrenal insufficiency is usually fatal if not diag-nosed and treated. The recognition of adrenalinsufficiency in patients operated upon for ex-tensive traumatic injury is often difficult becauseof fue una'v-ailability of historical informationand some-times tllrbulent postoperative courses. Recognition ofadrenal insufficiency in fuese patients would befacilitated by fue establishment of the normaladrenal response to cosyntropin stimulation afteroperation for extensive traumatic injury. Herein,we evaluate fue adrenal response to trau~aticinjury and operation, and we apply fue ci:osyn-tropin stimulation test to patients who were !oper-ated upon in an emergency situation for tra~matic

injury.MATERIAL A1'.:D METHODS

Thirty-four patients who underwent emergentabdominal or thoracic, or both, operation while

From the Deparunent of Surgery, University of Kentucky MedicalCenter. Lexington, and the Deparunent of Sltrgery. University ofNorth Carolina at Chapel Hill, Chapel Hill.

Reprint requests: Dr. Michael J. Harris, Deparunent of Surgery,Division of UrolOg)., Vniversity of Kentucky Medical Center, 800Rose Street. Lexington, Kentucky 40536-0084.

513

~

recei~ing general anesthesia for traumatic injurieswere evaluated. Injury Severity Scores (8) werecalculated for all patients. AlI operations wereat least one hour in duration. Physical findingsand history, when obtainable, on admission didnot suggest adrenal insufficiency in any patient.Patients were interviewed postoperatively and thehospital course was monitored for symptoms orsigns of adrenal insufficiency.

Four patients had serial serum and 24 hoururine cortisol determinations from admission tothe emergency room through the seventh post-operative day. Serum cortisol levels ,vere dra,Vilbefore, during and immediately after operationand at 8 A.M. and 8 P.M. for seven days. Urinewas collected from the end of the operation forseven days in 24 hour aliquots.

Thirty patients underwent a cosyntropin stim-ulation test at 12, 24 or 48 hours -after inductionof anesthesia. Mter a baseline serum cortisollevelwas obtained, 200 micrograms of cosyntropin ,vasadministered intravenously. Two hours later, a"stimulated" cortisol level ,vas obtained.

AlI serum cortisol levels were drawn from anintra-arterial catheter, if present, or by routinevenipuncture. The blood was centrifuged and theserum stored at zero to 4 degrees C. Urine wascollected by a Foley catheter, when present, andstored. AlI specimens from an individual patientwere analyzed simultaneously by radioimmunoas-say (Gammacoat 1 125 cortisol radioimmunoassaykit, Clinical Assays, Travenol Laboratories, Incor-porated). Serum cortiso} levels were expressedin micrograms per deciliter and urinary cortisolproduction in micrograms per day.

In four patients studied with serial serum andurinary levels, values were p)otted against timeand correlated with the postoperative course ofthe patient. In 30 patients whose adrenal functionwas tested after traumatic injury and after opera-tion, fue relationship between Injury Sev~rity Scoreand baseline level of serum cortigol ,vas analyzedby linear regression analysis. The serum cortisolresponse to cosyntropin stimulation ,vas compared

'1

514 SURGERY, Gynecology & Obstetrics. June 1990. VolurnR 170

Serum Cortisol (ug/dl) Urinary Cortlsol (ug/d)60. -.4000

\,50

,~

300040,

2000

100010

oo 4 6 8

daysFIG. l. Serum and urinary cortisol levels in patient with an unremarkable

recovery and a patient with a complicated recovery. Solid and oren circles,setum and urinary cortisollevels, respectively, in a patient \\ith an uncomplicatedrecovery, and solid and oren triangles, serum and urinary cortisol levels, respec-tively, in a patient with postpperative complications.

2

with baseline \'alues by paired Student's t tests.

Similarity of degree of trauma was verified by

analysis of \'afiance of Injury Severity Scores. Adrenal

response to coS)'Iltropin was compared at different

intef\1als after operation by analysis of variknce.

In another study, 125 patients were stibulated

with COs)'Iltropin at different intervals after elec-

tive extensive operations to a mean seru~ cortiSOI level of 43.9~12.4 micrograms per decil ter (7).

We compared stimulated cortisol levels om the

30 patients we operated upon after traumatic

injury to the pre\iously established normals by

the two-tailed Student's t test. A new expected

stimulated serum cortisol level in patients after

emergent operations for traumatic injury or after

elective operation was established by weighted

averaging of the means.

RESULTS

Serum and uri~ary cortisol levels measured in

four patients after operations for trauma were

elevated in the perioperative periodo Diurnal vari-

ation was reco\'ered by the third postoperative

day. In Figure 1, a patient who had an uncom-

plicated convalescence is compared with a patient

who had a complicated recovery. Hepatocellular

failure recognized by increased values from liver

function tests and confirmed by histologic ex-

amination of specimens taken at biopsy developed

in the latter patient. No cause for hepatocellularfailure was found. The patient recovered fromthe injury, operaf;ion and hepatocellular dysfunc-tion without further complication.

Baseline serum cortisollevels obtained 12 and48 hours after operation for trauma correlatedwith Injury Severity Scores (linear regression anal-ysis, r=O.73, p=0.017, and r=O.63, p=0.049, respec-tively). Twen ty-four hours after trauma andoperation, fue baseline cortisollevels of ten palien ts'\Tere related to their Injury Severity Scores, butthe association did not reach statistical signifi-cance (r=O.52, p=0.12).

Three groups of ten patients who received ca-syntropin adrenal stimulation at difIerent inter-vals after operation had suffered similar degreesof trauma. Injury Severity Scores (Table 1) weresimilar (analysis of variance, p=O.18). Baselinecortisol levels were 28.4::t:12.0 micrograms perdeciliter, 25.2::t:8.3 micrograms per deciliter and20.4::t:10.2 micrograms per deciliter at 12, 24 and48 hours after induction of anesthesia, respec-tively. Two hour stimulated levels were 44.7::t:9.4micrograms per deciliter, 48.6::t:12.5 microgramsper deciliter and 48.9::t:8.7 micrograms per de-ciliter (Table 11). Neither baseline le,'els nor stim-ulated levels differed significantly among fue threegroups (analysis of variance, p=O.23 and 0.60,respecti,'ely) (Fig. 2).

The stimulated serum cortisollevels in patientsafter elective extensive operation (43.9::t:12.4 mi-crograms per deciliter [7]) and in patients aftertraumatic injury and operation (47.4::t:10.2 micra-grams per deciliter) were similar (Student's t test,p=O.l). The data were combined, and a normal

TABLE I.-IKJURY SEVERITY SCORES

n Mean :i: S.D.

."'. 10 19.1:1:5.010 23.7:i:8.110 24.6:1:7.530 22.5:i:7.1

Group12 hrs.24 hrs.48 hrs.Totals

Range13 lO 2717 lO 4113 lO 3413 lO 41

~

Harris et al.: ADRENAL ~PONSE TO TRAUMA, OPERATION AND COSYNTROPIN 515

Serum Cortisol (ug/dl)

12 hour 24 hour 48 hourFIG. 2. Baseline and stimulated serum cortisol levels (mean plus or minus

standard error of the mean) at 12, 24 and 48 hours postinduction of anesthesiafor operative treatment of traumatic injuries. Oren columns, baseline, and solidcolumns, stimulated leveIs.

droxycorticosteroids (11, 12). More recently, serumcortisol has been measured directly by radioim-munoassay techniques in elective extensive opera-tions (3), operations upon fue neck (5), cardiacoperations (4) and surgical procedures for trau-matic injury (2, 13). Serum and urinary cortisollevels were elevated immediately after operationbut returned to the normal range in one to twodays. Persistent or transient increases in serumcortisol corresponded to periods of postoperativecomplications; serum cortisol values retumed tofue normal range after resolution of fue com-plications (3, 14).

Our results confirm previously described serumand urinary cortisol elevations after trauma andoperation. Patients wifu uncomplicated recoverieshad a prompt retum to normal serum and uri-nary cortisol levels; postoperative complicationwas associated wifu persistent elevation of cortisollevels.

In one study, serum cortisol was found to beincreased with severity of injury until injury wassevere and cortisol levels decreased. Decreasedadrenal blood flow secondary to shock was sug-

TABLE II.-SERUM CORTISOL LEVELS BEFORE AND AFfERCOSTh'TROPIN SnMULAnON

HOUTS post induction

12 24 48n 10 10 10Base line 28.4:112.0 25.2:t8.3 20.4:110.2Stimulated 44.7I9.4 48.6.t12.5 48.9:1:8.7Increment 16.1 23.4 28.6

Lcv"ls ar" micrograms p"r d"ciliter, mean plus or minus Ihe standard

deviau<Jn.

serum cortisol response to cosyntropin based oq155 patients treated by extensive operation elec~tively or after traumatic injury was 44.5j:12.0 mi~crograms per deciliter.

DISCUSSION

Postoperative adrenal insufficiency is rafe butusually fatal if not diagnosed and treated. In-cidences of one instance in 8,109 urologic sur-gical procedures (9) and five in 4,364 cardiacsurgical procedures (10) have been reported. Theactual incidence is probably higher because ofa lack of diagnosis. Postoperative adrenal insuf-ficiency is not obvious; its S)'Illptoms and signsare similar to those of many more common post-operative processes. Preoperative adrenal insuf-ficiency is the most common cause of postoperativeadrenal insufficiency. Therefore, a careful historyaqd physical examination are critical for diagnosisof adrenal insufficiency preoperatively. When doubtexists preoperatively, formal adrenal evaluationcan prevent disastrous postoperative results. Inthe population of patients ,\rith traumatic injury,medical history is usually unavailable and thepatient is often unable to provide any historythat might suggest preoperati,'e adrenal insuf-ficiency. Patients with traumatic injury who re-quire surgical treatment are often taken quicklyto the operating room before elaborate evalua-tions are undertaken.

The stress of operation has been described wellin the literature. In the 1950s, serum cortisol,as a measure of adrenal function, was determinedindirecdy by quantitating total and free 17-hy-

(516 SURGERY, Gynecology & Obstetrics' june 1990. Volume 170

gested as an explanation for the reversal of this the adrenal response is adequate, then steroidrelationsñip (1). Our data demonstrate~ a cor- therapy and its attendant risk of immwlosuppres-relation bet\\'een the Injury Severity Scpre and sion, wound healing impairment and metabolicbaseline cortisol levels. derangement can be discontinued and the patient

In other studies, patients were stimclated at carefully observed.various intervals after elective extensiv~ opera- The adrenal response to operation is not alteredtions with coS)'l1topin; a similar stimulated serum by antecedent trauma. The cosyntropin stimula-cortisollevel ,vas found regardless of the baseline tion test can be used in patients ,vith traumaserum cortisol or the interval between induction who undergo emergent operations for the diag-and stimulation (6, 7). We, toa, found !no cor- nosis of postoperative adrenal insufficiency.relation bet\veen baseline or stimulated i cortisolvalues and the interval after operation! for ex-tensive trauma. However, patients dem~strated significant adrenal response to cosyntro in stim-

ulation at 12, 24 and 48 hours after o erationfor n-auma, which were similar to those dem-onstrated in patients undergoing elective extensive ...operations (7). The superimpositíon of serioustrauma on operation affected the baseline but .not the stimulated serum cortisollevel. Therefore, -~.we combined the t\vo series of patients to char-acterize more precisely the cosyntropin stimulationtest A normal postoperative cosyntropin stimulatedtest defined in 155 patients required a serumcortisol level of 44.5:t12.0 micrograms per de-ciliter t\vo hours after intravenous administrationof 200 micrograms of cosyntropin.

SUM?1.-\RY

When one is faced with a patient in extremisafter trauma, elective operation or emergent oper-ation, dexamethasone should be administered anda cosyntropin test performed. Once the stimulatedserum cortisolle,'el is drawn, hydrocortisone, whichwould interfere ~ith cortisol measurement, marbe substituted for dexamethasone to pro\ide amineralocorticoid effect not present ~ith dexa-methasone. AlI patients, regardless of their ageand sex, duration and invasiveness of the opera-tive procedure, the interval since operation orfue presence of preoperative trauma, were stimu-lated to a mean serum cortisollevel of 44.5::!: 12.0micrograms per deciliter. Therefore, a stimulatedserum cortisol level of ~20.5 micrograms perdeciliter will include 97.5 per cent of normalpatients. If the stimulated cortisol level is <20.5micrograms per deciliter, adrenal insuffiáency is like-ly and steroid therapy should be continued. If

REFERENCESl. BARTON, R. N., STONER, H. B., and WATSON, S. M.

Relationships among plasma cortisol, adrenocorticotro-phin, and severity of injury in recently injured patients..

J. Trauma, 1987, 27: 384-392.2. G.~""N, D. S., and LILLY, M. P. The neuroendocrine

response to multiple trauma. World J. Surg., 1983, 7.101-118.

3. MOHLER, J. L., MICHAEL, K A, FREED~{A.'1, A. M., andothers. The serum and urinary cortisol response tooperative trauma. Surg. Gynecol. Obstet., 1985, 161445-449

4. OKA, Y, WAKAYAMA, S., O~\.fA, T., and others. Cortisoland antidiuretic hormone responses to stress in cardiacsurgical patients. Can. Anaesth. SOCo l., 1981, 28: 334-338.

5. UDELSMAN, R., NORTON, J. A., HELE~"ICH, S. E., andothers. Responses of the hypothalamic-pituitary-adrenaland renin-angiotensin axes and the sympathetic systemduring controlled surgical and anesthetic stress. J. Clin.Endocrinol. Metab., 1987, 64: 986-994.

6. MOHLER, J. L., MICHAEL, K A, FREEDM.~~, A. M., andothers. The e\oaluation of postoperative function of theadrenal gland. Surg. Gynecol. Obstet., 1985, 161: 551-556.

7. Idem. EvaIuation of adrenal function in the periopera-tive periodo Surg. Forum, 1984, 35: 55-57.

8. B.~R, S. P., O'NEILL. B., HADDEN. W., and LONG. W.B. The Injury Severity Score: A method for describingpatients wiili multiple injuries and evaluating emer-gency careo J. Trauma, 1974, 14: 187-196.

9. MOHLER, J. L., FLUECK, l. A., and McRoBERTS, J. w.Adrenal insuffidency following unilateral adrenalectomy:A case reporto J. Urol., 1986, 135: 554-556.

10. ALFORD, W.,JR, MEADOR, C., MlHALEVICH.J., and others.Acute adrenal insufficiency following cardiac surgicalprocedures.J. Thora~. c:ardio\oasc. Surg., 1979, 78: 489.

11. HELMREICH, M. L.,JENKINS, D., and SWA."J. H. The adrenalc°r,tical response to surgery. Surge.ry, 1957,41: 895-909.

12. SA DBERG, A. A., EIK-NEW. K, SA.\1UELS, L. T., and 1\LER,F. M. The effects of surgery on the blood levels andmetabolism of 17-hydroxycorticosteroids in mano l. Clin.Invest.., 1954, 33: 1509-1516.

13. CAREY, L. C., CLounER, C. T., and Lo\\'ERY, B. D. GroWthhormone and adrenal cortica!. response to shock andtrauma in the humano Ann. Surg., 1971, 174: 451-458.

14. GILL, G., PRUDHOE, K, CoOK, D., and UTKER, A. Effectof surgical trauma on plasma concentrations of cyclicA..\1P and cortisol. Br. l. Surg., 1983, 62: 441-443.~