World J. Surg. 24, 579–582, 2000DOI: 10.1007/s002689910095 WORLD

Journal ofSURGERY© 2000 by the Societe

Internationale de Chirurgie

Natural Course of Adrenal Incidentalomas: Seven-year Follow-up Study

Jukka Siren, M.D.,1 Pekka Tervahartiala, M.D.,2 Arto Sivula, M.D.,1 Reijo Haapiainen, M.D.1

1Department of Surgery, Helsinki University Central Hospital, Haartmaninkatu 4, 00290 Helsinki, Finland2Department of Radiology, Helsinki University Central Hospital, Haartmaninkatu 4, 00290 Helsinki, Finland

Abstract. Most adrenal incidentalomas are nonfunctioning cortical ade-nomas that do not require surgery. Operative treatment is indicated if theincidentaloma has malignant features as seen during the radiologicworkup, is hormonally active, or grows during follow-up. A conservativeattitude is justified in asymptomatic patients if we know that the naturalcourse of these tumors is not disadvantageous. We followed a group ofpatients treated by observation to determine the natural course of adrenalincidentalomas. Altogether 30 incidentalomas in 27 patients were de-tected in the Helsinki University Central Hospital from June 1981through December 1992 and were re-examined during 1997. The patientswere evaluated clinically, and hormonal testing was done by performing a1-mg overnight dexamethasone suppression test; the next day we mea-sured the 24-hour urinary excretion of vanillylmandelic acid, metaneph-rines and normetanephrines, and serum potassium. Magnetic resonanceimaging (MRI) was used to determine the size of the tumor. If the patienthad died during the follow-up period, the causes of death from the deathcertificates and autopsy reports were reviewed. The mean follow-up was 7years. Nine patients had died, with none of deaths related to the inciden-taloma. Sixteen patients were evaluated clinically and by the basic hor-monal tests and MRI; none showed any signs of hormonal activity orfeatures of malignancy. Two patients did not want to be reexamined butstated that they were asymptomatic. This follow-up study supports theconclusion that conservative therapy is justified because the growth ten-dency of adrenal incidentalomas is slow and they seem to remain hor-monally inactive.

Adrenal tumors are incidentally detected in 0.35% to 1.30% ofupper abdominal computed tomography (CT) scans [1, 2]. Inautopsy series the incidence of adrenocortical adenomas is evenhigher: 1.4% to 8.7% [3, 4]. Hormonally active tumors and thosewhere radiologic examinations suggest malignancy should be sur-gically removed. Regarding hormonally inactive tumors with abenign appearance on imaging studies and that do not increase insize at follow-up, the traditional recommendation has been thatincidentalomas larger than 3 cm [1] or 6 cm [5] in diameter shouldbe removed, and that the others should be followed by repeatedCT or magnetic resonance imaging (MRI) scans. Staren and Prinzsuggested that masses between 3 and 6 cm should be removed ifthe patient is less than 50 years of age to avoid prolonged obser-vation [6]. The advantages of laparoscopic adrenalectomy havechanged some investigators’ opinions, and they now recommendoperative treatment also for small incidentalomas [7]. Few long-term follow-up studies have been reported so we can assess the

justification of the nonoperative management. We present a seriesof 27 patients followed for 7 years to gain information about thenatural course of incidentalomas of the adrenal gland.

Materials and Methods

At Helsinki University Central Hospital from June 1981 throughDecember 1992, a series of 27 patients with nonoperativelytreated incidentalomas were identified. In 2 cases the incidenta-loma was primarily found during an ultrasonography (US) exam-ination and in 25 cases on a CT scan. In the cases found by US, thefinding had been verified with CT. If no malignant features in CTexamination and no hormonal activity had been found, the pa-tients were observed by CT or MRI 1 year later. If the inciden-taloma had not changed in size or appearance, the follow-up wasstopped. During 1997 the medical records and CT images wereanalyzed retrospectively, and the results of the hormonal investi-gations and the size and appearance of the incidentalomas wererecorded. Nine patients had died before 1997. The death certifi-cates were obtained from the national register, and the cause ofdeath was recorded. If a postmortem examination had been done,the report was obtained and the findings were recorded. All theliving patients were contacted and asked to participate in thefollow-up program that had been approved by the ethical com-mittee of Helsinki University Central Hospital. The programconsisted of the patients’ history, physical examination, hormonalevaluation, and MRI scan of the adrenal glands. The hormonalevaluation included 24-hour urine excretion of metanephrinesand normetanephrines, 24-hour urine excretion of vanillylman-delic acid, a 1-mg overnight dexamethasone suppression test, andserum potassium measurement.

The MRI examination was performed with a 1.5T supercon-ducting imager (Siemens Magnetom Vision 1.5T; Siemens, Erlan-gen, Germany). A phased array local body coil was used in all thestudies. The slices were acquired in the transaxial plane with a 256imaging matrix. Fast breath-hold sequences were utilized to re-duce the imaging time and moving artifacts.

Fat-suppressed T1-weighted axial images were acquired beforeand 25 and 55 seconds after intravenous injection of a bolus ofgadolium DTPA (Magnevist; Schering, Berlin, Germany) (dose0.1 g/kg). The imaging was continued with slices in the coronalplane. The duration of the entire imaging procedure was about 20Correspondence to: J. Siren, M.D.

minutes. The MR images were reviewed by a blinded observerexperienced in abdominal MRI. Visualization of the adrenalglands was assessed. The size of the existing adrenal tumor wasmeasured and compared with the data of earlier CT studies. TheT1 and T2 signal intensity of the tumor and the contrast enhance-ment of the lesion were compared with those of normal adrenaltissue.

Results

All 27 patients had an incidentally discovered adrenal mass; 3patients had a bilateral incidentaloma. There were 7 male pa-tients, and 20 were female. The mean age of the patients at thetime of the incidental discovery of the adrenal mass was 59 years(range 19–80 years). The mean size of the incidentalomas was 25mm (range 15–40 mm). Eight of the incidentalomas were 3 cm orlarger. Hormonal evaluation had been done in all cases, and theincidentalomas were all found to be hormonally inactive. Themean follow-up time was 85 months (range 23–196 months). Themean follow-up time for the alive patients was 98 months (range53–196 months).

Nine patients had died before the follow-up program started.The mean age at death was 72.8 years. The cause of death wasacute myocardial infarct in three cases and one death each due toventricular arrhythmia after aortic and mitral valve reconstruc-tion, pulmonary edema caused by cardiac failure, thrombosis ofthe femoral vein and pulmonary embolism, acute cerebrovascularinfarct, bolus suffocation related to multiinfarct dementia, andlung carcinoma with cerebral metastases. When found, the inci-dentalomas had been hormonally evaluated and inactive. Ondeath certificates and available medical records nothing suggestedthat these deaths would be related to the incidentalomas. Anautopsy had been performed on four patients, and the incidenta-loma had been documented in three: two cortical adenomas andone myelolipoma. No change in the size of the incidentalomascould be noted.

One patient had been investigated for hypertension in anotherhospital just prior to being asked to participate in our follow-upprogram. The hospital records were analyzed, and it was con-cluded that the hypertension was not related to the hormonallyinactive incidentaloma, which had not changed in size during the12 years of follow-up. One patient did not want to participate butstated that he was in normal health and had no symptoms.

Sixteen patients were examined as described above. None hadany symptoms related to the incidentaloma. In all cases the results

of the hormonal tests were normal. MRI revealed no adrenalpathology in four patients, although visualization of the adrenalglands was considered excellent. In one of these patients theincidentaloma had had a cystic appearance and had been emptiedby aspiration. One adrenal gland was horizontally positioned,which had caused misinterpretation of an adrenal tumor at theoriginal CT examinations. In two cases the incidentalomas haddiminished in size during the original 1-year follow-up and wethought to be adrenal hematomas. The size of seven incidentalo-mas had not changed. Five incidentalomas were larger and threewere smaller than at the time of detection; the change was 4 to 10mm. The results are summarized in Figure 1.

When compared with the MR signal of normal adrenal tissue,the T1 and T2 signals of the observed tumors were isointense in allbut one case. This particular tumor appeared homogeneouslyhyperintense on T2-weighted images. The lesion showed no con-trast enhancement. Such a tumor correlates well with a simplecyst. All the other adrenal tumors demonstrated a minor rimenhancement of the lesion 25 seconds after the contrast injection.The following slices, 55 seconds after the injection, showed slight,equalized enhancement of the tumor analogous to the surround-ing normal adrenal tissue.

Discussion

Adrenal incidentalomas are found in more than 1% of abdominalCT scans. Most of them are benign, hormonally inactive, andnonfunctioning; and operative treatment is not indicated. Currentcriteria for surgical treatment are based on endocrine function,possible malignant features of the CT appearance, and tumor size[6]. The goal is to remove malignant and hormonally active le-sions, while avoiding unnecessary operations. Endocrinologic ac-tivity must be tested in each incidentaloma, because encounteringhormonally active incidentalomas is not rare [8]. Most adrenalcarcinomas are larger than 6 cm [9, 10], although some smallercarcinomas have been reported [11]. If the incidentaloma is hor-monally inactive and has no malignant features on the CT scan,the choice between operation and observation is made by evalu-ating the risk for malignancy, which increases when the diameterof the incidentaloma enlarges. Current recommendations for acutoff point for operative treatment vary between 2.5 and 6.0 cm,with 4 cm being the value most often used [12]. Recommendationsfor the follow-up of patients with incidentalomas vary among

Fig. 1. Results of the 7-yearfollow-up of 27 patients with 30incidentalomas.

580 World J. Surg. Vol. 24, No. 5, May 2000

centers, and the natural course of adrenal incidentalomas has notbeen clear. When we decided to start this study, no long-termfollow-up studies had been reported. There is little if any evidencethat repeated follow-up by imaging or functional studies is bene-ficial, and the likelihood of function or malignancy during long-term follow-up is low [6, 12].

In the present series the mean follow-up for 27 incidentalomapatients was more than 7 years. Nine patients had died, but wefind it unlikely that the causes of death could have been related tothe incidentalomas. No signs of cortisol overproduction or pheo-chromocytoma had been found at initial hormonal tests, andnothing on the death certificates or available medical recordssuggested the presence of hormonally active adrenal pathology.An autopsy was done for four patients, and in three of them theincidentaloma was found and documented. Sixteen patients werereexamined clinically; basic adrenal hormonal testing was per-formed; and MRI was used to determine the size of the inciden-taloma. Because MRI is a well documented technique for inves-tigating adrenal pathology [13] it was chosen as the imaging studyinstead of CT to avoid exposing healthy subjects to ionizingradiation. No signs of excessive hormone production could befound on physical examination or basic hormonal tests. In fourpatients the MRI could not find any adrenal pathology, althoughvisualization of the gland was considered excellent. When thehospital records and primary CT scans were reevaluated, theexplanations were as follows: One was incorrectly diagnosed aspositive because of anatomic variation, two were considered to beprimarily adrenal hematomas; and in one case the incidentalomahad had a cystic appearance and had been emptied by aspiration.Ten incidentalomas were smaller or of the same size as at the timeof diagnosis. In five cases the adrenal mass was larger but notmore than 1 cm.

Barry et al. [12] have reported a series of 231 patients identifiedfrom the records of all abdominal or thoracic CT scans performedfrom 1985 to 1989 at the Mayo Clinic. The clinical follow-up(mean 7 years) was obtained by review of medical records, per-sonal telephone interviews, or correspondence with patients ortheir physicians. The results showed that four tumors had showna . 1 cm increase in size and had been surgically excised butproven benign. No patients developed any clinical evidence ofadrenal hyperfunction or malignancy. The findings of Barry et al.are in line with our results, and both indicate that conservativemanagement of small adrenal incidentalomas is appropriate.

Conclusions

The present results indicate that the clinical significance of thesmall nonfunctioning adrenal incidentalomas is limited. They tendto grow slowly, and a conservative attitude toward treatment isjustified. MRI is suitable as an imaging method for the follow-upof these patients.

Resume

La vaste majorite d’incidentalomes surrenaliens sont desadenomes non-fonctionnels de la corticale pour lesquels lachirurgie n’est pas indiquee. Le traitement chirurgical est indiquesi la tumeur a des caracteres maligns aux investigationsradiographiques, possede une activite hormonale propre ou

augmente de volume pendant le suivi. Une attitude conservatriceserait justifiee chez le patient asymptomatique si nousconnaissions l’histoire naturelle, benigne, de ces tumeurs. Nousavons suivi un groupe de patients pour determiner l’histoirenaturelle des incidentalomes de la surrenale. On a detecte 30incidentalomes chez 27 patients a l’Hopital Universitaire deHelsinki entre juin 1981 et decembre 1992, qui ont ete reexaminespendant l’annee 1997. Ces patients ont ete reevalues du point devue clinique et hormonal au moyen du test de suppression dedexamethasone nocturne, 1 mg, et en dosant l’excretion urinairede 24 h de l’acide vanillylmandelique, les metanephrines et lesnormetanephrines, ainsi que le potassium serique. L’imagerie parresonance magnetique a ete utilisee pour determiner la taille de latumeur. Si le patient etait decede pendant la periode de suivi, ona determine les causes de deces a partir des certificats de deces,et/ou a l’autopsie. Le suivi moyen a ete de 7 ans. Neuf patientssont decedes, aucun en rapport avec leur incidentalome. Seizepatients ont ete evalues cliniquement, par des test hormonaux etpar l’IRM; aucune des tumeurs n’avait de signes d’activitehormonale ou de malignite. Deux patients ont refuse la re-examination, mais ont dit qu’ils etaient asymptomatiques. Cetteetude soutient qu’une attitude conservatrice est tout a faitjustifiee, car la croissance des incidentalomes est lente, et ilsrestent inactifs du point de vue hormonal.

Resumen

La mayorıa de los incidentalomas de las capsulas suprarrenalesson adenomas corticales no funcionantes y, por ello, no requierentratamiento quirurgico. Este, solo esta indicado sı: se observansignos de malignidad en el estudio radiologico, aparece actividadhormonal o crece el tumor durante el periodo de seguimiento. Enpacientes asintomaticos se justifica el tratamiento conservadorpor la benigna evolucion natural de estos tumores. Efectuamos unseguimiento de pacientes asintomatico con incidentalomastratados conservadoramente, con objeto de averiguar la evolucionespontanea de estos tumores. En el Hospital Central de laUniversidad de Helsinki, se diagnosticaron entre junio de 1981 ydiciembre de 1992, 30 incidentalomas en 27 pacientes; todosfueron revisados en 1997. Los enfermos fueron estudiadosmediante exploracion clınica y el test de supresion realizado trasinfundir durante la noche 1 mg de dexametasona, a continuacion,se recogıa la orina de 24 horas, determinandose los niveles deacido vinilmandelico, metanefrinas y normetanefrinas. Tambien,se valoro la potasemia. La resonacia magnetica permitio evaluarel tamano del tumor. Cuando el paciente habıa fallecido serecogieron las causas de muerte registradas no solo en elcertificado de defuncion sino tambien, en los protocolos de lasnecropsias. El seguimiento medio fue de 7 anos. Nueve pacientesfallecieron por causas ajenas al incidentaloma. Dieciseis pacientesfueron evaluados clınicamente y ni los tests basicos hormonales, nila resonancia magnetica demostraron actividad hormonal alguna,ni signos de malignizacion. Dos pacientes rechazaron un nuevoestudio, pero ambos estaban asintomaticos. Este seguimientopermite llegar a la conclusion de que esta justificada la terapiaconservadora, pues el crecimiento de los incidentalomas es muylento y no desarrollan actividad hormonal alguna.

Siren et al.: Follow-up of Incidentalomas 581

References

1. Prinz, R.A., Brooks, M.H., Churchill, R., Graner, J.L., Lawrence,A.M., Paloyan, E., Sparagana, M.: Incidental asymptomatic adrenalmasses detected by computed tomographic scanning: is operationrequired? J.A.M.A. 248:701, 1982

2. Abecassis, M., McLoughlin, M.J., Langer, B., Kudlow, J.E.: Serendip-itous adrenal masses: prevalence, significance, and management.Am. J. Surg. 149:783, 1985

3. Kokko, J.P., Brown, T.C., Berman, M.M.: Adrenal adenoma andhypertension. Lancet 1:468, 1967

4. Hedeland, H., Ostberg, G., Hokfelt, B.: On the prevalence of adre-nocortical adenomas in an autopsy material in relation to hyperten-sion and diabetes. Acta Med. Scand. 184:211, 1968

5. Gajraj, H., Young, A.E.: Adrenal incidentaloma. Br. J. Surg. 80:422,1993

6. Staren, E.D., Prinz, R.A.: Selection of patients with adrenal inciden-talomas for operation. Surg. Clin. North Am. 75:499, 1995

7. Linos, D.A.: Adrenaloma (incidentaloma). In: Textbook of Endocrine

Surgery, O.H. Clark, Q. Duh, editors. Philadelphia, Saunders, 1997,pp. 475–482

8. Siren, J.E., Haapiainen, R.K., Huikuri, K.T., Sivula, A.H.: Incidenta-lomas of the adrenal gland: 36 operated patients and review of liter-ature. World J. Surg. 17:634, 1993

9. Herrera, M.F., Grant, C.S., van Heerden, J.A., Sheedy, P.F., Ilstrup,D.M.: Incidentally discovered adrenal tumors: an institutional per-spective. Surgery 110:1014, 1991

10. Khafagi, F.A., Gross, G.M., Shapiro, B., Glazer, G.M., Francis, I.,Thompson, N.W.: Clinical significance of the large adrenal mass. Br. J.Surg. 78:828, 1991

11. Linos, D.A., Stylopoulos, N., Raptis, S.A.: Adrenaloma: a call formore aggressive management. World J. Surg. 20:788, 1996

12. Barry, M.K., van Heerden, J.A., Farley, D.R., Grant, C.S., Thompson,G.B., Ilstrup, D.M.: Can adrenal incidentalomas be safely observed?World J. Surg. 22:599, 1998

13. Peppercorn, P.D., Reznek, R.H.: State-of-the-art CT and MRI of theadrenal gland. Eur. Radiol. 7:822, 1997

Invited Commentary

J.F. Moley, M.D.

Department of Surgery, Washington University School of Medicine, St.Louis, Missouri, USA

Surgeons are often required to assist in the management of pa-tients with adrenal Incidentalomas. Detection of these lesions,which are encountered in as many as 9% of autopsies, will un-doubtedly become more common in the future as the sensitivity ofimaging studies improve. The surgeon must weigh the cost andmorbidity of diagnostic testing and surgery against the risk ofmalignancy and function of the adrenal lesion. The ability to makeintelligent, safe decisions regarding management of incidentalo-mas requires a knowledge of the types of pathology that occur inthe adrenal gland, biochemical workup, imaging characteristics,and natural history.

Cysts and lipomas have characteristic computed tomographic(CT) and magnetic resonance imaging (MRI) characteristics, andmay generally be observed safely. Adrenal hemorrhage also has acharacteristic appearance on imaging studies and occurs morecommonly on the right, in the setting of trauma or anticoagula-tion. Adrenal adenomas are the most common neoplasms ac-counting for incidentalomas, and they are usually non-functional.Functional adenomas may produce aldosterone or cortisol; andscreening for these lesions should include measurement of bloodpressure and potassium levels and an overnight dexamethasonesuppression test. “Subclinical” Cushing syndrome may be causedby adenomas that produce cortisol at levels that suppress thecontralateral gland without causing clinical Cushing syndrome.Removal of these lesions without steroid coverage postoperativelymay be dangerous, and it is recommended that an overnightdexamethasone suppression test be performed in all patients inwhom excision of an adrenal adenoma is planned. Adrenocorticalcarcinomas are usually large and have features of invasiveness.

These tumors may cause Cushing syndrome and masculinizationin women. Pheochromocytomas are associated with symptoms ofcatecholamine excess and may be present in patients with a familyhistory of multiple endocrine neoplasia type II, neurofibromatosistype 1, or von Hippel-Lindau disease. Catecholamine and metab-olite levels should be determined in 24-hour urine specimens torule out this lesion. Adrenal metastases account for up to 30% ofincidentalomas discovered during the course of a metastaticworkup in a patient with cancer.

Advances in imaging technology has improved the ability todistinguish these lesions. Pheochromocytomas are often bright onT2-weighted MRI imaging. Adenomas have characteristic signalintensity characteristics on opposed phase chemical shift MRIimaging—the signal “drops out” in opposed phase owing to thehigh fat content of adenomas compared to carcinomas, metasta-ses, or pheochromocytomas.

The series reported by Siren and colleagues is an interesting,important contribution to the literature regarding management ofadrenal incidentalomas. Although the authors do not provideinformation regarding adrenal lesions that were not observed—those that were surgically removed—they demonstrate that intheir practice all lesions managed nonoperatively were stable anddid not cause morbidity in the patient. In our practice, we recom-mend surgical removal of adrenal incidentalomas that are 4 cm ormore in size, functional, or less than 4 cm with suspicious imagingcharacteristics [1]. We recommend observation of small, nonfunc-tional lesions that are clearly benign radiologically. Adrenal le-sions should be biopsied only if the results can influence manage-ment and only after a 24-hour urine determination has ruled outthe possibility of a pheochromocytoma.

Reference

1. Brunt, L.M., Doherty, G.M., Norton, J.A., Soper, N.J., Quasebarth,M.A., Moley, J.F.: Laparoscopic adrenalectomy compared to openadrenalectomy for benign adrenal neoplasms. J. Am. Coll. Surg. 183:1,1996

582 World J. Surg. Vol. 24, No. 5, May 2000