| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Centre de Référence Maladies Rares de la Surrénale, Service des Maladies Endocriniennes et Métaboliques (G.A., L.G., X.B., J.B.), Service de Biostatistique (J.C.), Service d’Orthopédie B (P.A.), Service d’Anatomo-pathologie (F.T.), Service de Radiologie A (P.L.), Service de Chirurgie Digestive et Endocrine (B.D.), Hôpital Cochin, Assistance Publique-Hôpitaux de Paris, Université Paris 5, Paris, France; and Institut National de la Santé et de la Recherche Médicale Unité 567 and Centre National de la Recherche Scientifique Unité Mixte de Recherche 8104 (L.G., F.T., X.B., J.B.), Institut Cochin, Paris, France
Address all correspondence and requests for reprints to: Prof. Jérôme Bertherat, Service d’Endocrinologie, Hôpital Cochin, 27, rue du Faubourg Saint-Jacques, 75014, Paris, France. E-mail: jerome.bertherat{at}cch.aphp.fr.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Objective: We aimed to determine the clinical characteristics and outcome of ACC and to identify prognostic factors.
Design and Setting: This study is a descriptive and multivariate analysis of a cohort from a single endocrinology center.
Patients: A total of 202 consecutive patients with ACC were studied.
Results: The sex ratio (female to male) was 2.7. Mean age at diagnosis was 44 ± 16 yr (range, 11–88 yr). We found that 154 patients (76%) had hypersecreting tumors [mostly cortisol and androgens (47%), cortisol alone (27%), or androgens alone (6%)] and 43 patients (21%) had metastases at diagnosis. At initial staging or during follow-up, 85 patients (42%) had liver metastases, 79 patients (39%) had lung metastases, and 20 patients had bone metastases (10%). The survival rate was 37% at 5 yr. Multivariate analysis identified the following independent prognostic factors associated with shorter survival: older age at diagnosis [hazard ratio (HR), 1.03; P < 0.0001], initial MacFarlane extension stages 3 (HR, 4.42; P = 0.005) and 4 (HR, 7.93; P < 0.0001), and cortisol hypersecretion (HR, 3.90; P < 0.0001). Treatment with 1,1-dichlorodiphenildichloroethane (o,p’DDD) in the 3 months after surgery increased the survival rate of patients with cortisol-secreting tumors (HR, 0.40; P = 0.04).
Conclusion: This study highlights the better prognosis of ACC diagnosed at a noninvasive local stage, the particularly poor prognosis of patients with cortisol-secreting tumors, and the beneficial effect of o,p’DDD therapy in this subgroup of patients.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This study of a large cohort of consecutive patients followed in a single clinical center was designed to analyze the clinical and hormonal characteristics of ACC at diagnosis, to identify prognostic factors for the occurrence of metastasis and survival, and to evaluate the effect of o,p’DDD on survival.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
A total of 202 patients with ACC over the age of 10 yr at diagnosis were referred to the Endocrine Department of Cochin Hospital from 1963–2003. For each patient, diagnosis was confirmed by pathological examination of adrenal tumor tissue removed during surgery and/or malignant history and endocrine investigations, as previously reported (18, 19, 20, 21). The minimal endocrine work-up included 24-h urinary 17-OH steroids and/or cortisol, plasma and/or salivary cortisol circadian rhythm, dexamethasone test and/or ACTH plasma assay, plasma androgen assay (testosterone and/or dehydroepiandrosterone sulfate or dehydroepiandrosterone), and when appropriate, plasma renin or renin activity and aldosterone assays as well as plasma estradiol, 17-OH-progesterone, S compound, and desoxycorticosterone assays. All patients were investigated in our department at the time of initial diagnosis before adrenal surgery and/or when tumor recurrence occurred. Assays were performed as previously reported (18, 19, 20). Staging was performed using MacFarlane’s criteria, as modified by Sullivan et al. (22) based on surgical and/or imaging results, as previously reported (18). Metastases were diagnosed by systematic imaging investigations—principally abdominal and chest computed tomography scans—and by symptom-oriented radiological investigations in some cases. They were confirmed by pathological examination if surgery was appropriate. The mean duration of follow-up was 3.4 ± 4.4 yr (range 0.3–26). The study was approved by and performed according to the recommendations of the ethics committee of Cochin Hospital.
Statistical analysis
Data are reported as means ± 1 SD or percentage as appropriate. Due to nonnormal distribution of several variables and the small number of subjects in certain groups of interest, nonparametric statistical methods were used to analyze the relationships between variables when appropriate (exact 
2 test, Wilcoxon test, and Kruskall-Wallis ANOVA). We used two-tailed tests, and P values less than 0.05 were considered significant. We plotted survival curve, disease-free survival curve, and metastasis-free survival curves, according to the Kaplan-Meier method. Prognostic factors were identified by means of Cox proportional hazards regression models, which were used to estimate crude and adjusted hazard ratios (HRs) and their 95% confidence intervals. The final model for outcome prediction was constructed in several stages. First, a backward procedure was used to keep in the final model only the factors with significant explanation of outcome (remove P values 
0.05). Second, all first-order interactions between variables kept in the model were tested. The last stage was to test interactions between gender and all the variables kept in the preceding stage. o,p’DDD therapy initiated in the 3 months after surgery as well as an adjuvant or a palliative treatment, was studied in intention-to-treat analyses. We defined three periods of time for analysis of the effect on survival of the development of new methods of patient management: patients diagnosed before 1975, from 1975–1988, and since 1988. Statistical analysis was carried out with the SAS package (SAS Institute Inc., Cary, NC).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The sex ratio (147 females, 55 males) was 2.7. The mean age at diagnosis was 44.2 ± 16.2 yr (Table 1
). Female patients were younger: 42.7 ± 16.5 vs. 47.9 ± 14.6 yr (P = 0.04).
|
The mean delay from initial symptoms to diagnosis was 12.1 ± 18.8 months. The disease was revealed by endocrine features in 109 patients (54%), by local or regional manifestations (abdominal pain, palpable tumor mass, or compression signs) in 49 patients (24%). Twenty-seven patients (13%) were investigated for an adrenal incidentaloma: 16 females and 11 men, mean age was 46 yr ± 18.3. All incidentally detected ACC were diagnosed after 1983. Four ACC were revealed by bone metastases, 13 by general manifestations such as weakness and weight loss, three by noninfectious fever, three by metastatic pulmonary emboli, one by hypoglycemia. One patient had paraneoplasic symptoms with pruritus. One patient had an ectopic tumor in the pelvis with an ACTH-independent Cushing syndrome. Her adrenals were atrophied, and a large mass was found in the pelvis corresponding to an adrenal carcinoma. She died 6 months later with metastases. In 10 female patients, ACC was diagnosed during pregnancy or in the postpartum period.
Secretions (Table 2)
Hormonal investigations showed that 154 patients (76%) had a secreting tumor. Only 70% of these patients presented endocrine symptoms (109 patients). Eighty-two percent of the 147 females presented a secreting tumor (vs. 58% of the males). The steroids most commonly oversecreted were cortisol and androgens: 72 patients (47% of secreting tumors) had tumors secreting both cortisol and androgen, 41 patients (27%) had a tumor secreting only cortisol, and 10 patients (6%) had a tumor secreting only androgens. Aldosterone- and estradiol-secreting tumors were less frequent. Overall, 135 patients with a secreting tumor (87%) presented cortisol oversecretion, 97 (63%) presented androgen oversecretion, 20 (15%) presented aldosterone oversecretion, and six presented estradiol oversecretion. Six patients had only an excess of steroid precursors. All the patients diagnosed during pregnancy or in the postpartum period had a secreting tumor with cortisol secretion alone or associated to other steroid.
|
Mean tumor size was 11.3 ± 5.2 cm (range, 4–30 cm). Mean tumor weight was 615.7 ± 757.1 g (range, 22–4300 g). Six percent of patients had a MacFarlane stage 1 ACC (tumor diameter < 5 cm) and 51% had a stage 2 tumor (>5 cm); 19% had a stage 3 tumor (local invasion or invasion of regional lymph nodes) and 24% had a stage 4 tumor (distant metastases and/or invasion of adjacent organs plus nodes). Local invasion was mostly venous, in the inferior vena cava. Patients diagnosed at a localized stage (MacFarlane stage I and II) were 40% of the 25 diagnosed before 1975, 51% of the 76 patients diagnosed between 1975–1988, and 61% of the 101 patients diagnosed since 1989. Eighty-four percent (22 of 26) of the patients with incidentally detected ACC were diagnosed at localized stage. Cortisol-secreting tumors distribution by stages were not different (P = 0.09): 72% in stage 1, 59% in stage 2, 68% in stage 3, and 81% in stage 4.
Metastases: description and predictive factors
Sixty-seven patients had no recurrence during the follow-up. During the study period, 52 patients developed local relapse and 123 metastases (Fig. 1E): 85 patients (42%) were diagnosed with liver metastases, 79 (39%) with lung metastases, 20 (10%) with bone metastases, and 14 patients with abdominal lymph nodes (7%). Two years after diagnosis, 40% of MacFarlane stage 1–3 ACC patients had developed distant metastases: 37% liver metastases, 32% lung metastases, and 8% bone metastases (Fig. 1A). The percentage of patients who developed metastasis during follow-up was 27% for stage 1, 46% for stage 2, and 63% for stage 3. Multivariate analysis identified three factors significantly associated with a higher risk of metastasis at least one of these three sites: being male [HR, 0.35; (0.21–0.61); P < 0.001], metastasis at least one of the other two sites [HR, 4.50; (1.62–12.48); P = 0.004], and cortisol hypersecretion [HR, 4.62; (1.43–14.91); P = 0.01]. The size and weight of the primary tumor, patient age, mineralocorticoid or androgen secretion, o,p’DDD treatment, and study period were not significantly associated with the risk of developing metastases.
|
One hundred eighty-two patients had a surgery, with complete tumor removal in 142 cases. o,p’DDD was initiated in the 3 months after removal of the primary tumor in 99 patients. An additional 63 patients received o,p’DDD at a later time from the initial surgery. The objective was to give the maximal tolerable dose starting from 6 g up to 12 g/d of mitotane (Roussel UCLAF and Pharmacie Centrale des Hôpitaux, Paris, France) capsules containing 0.5 g of micronized mitotane mixed with cellulose acetylphtalate. The mean treatment duration was 729.57 ± 853.45 d (range, 2–4194 d). Cytotoxic chemotherapy was given to 13 patients with various drug regimens.
Survival
The survival rate was 71% at 1 yr, 57% at 2 yr, 37% at 5 yr, and 31% at 8 yr (Fig. 1B
). Two patients died from other neoplasia and one died from suicide. In univariate analysis (Table 3), poor survival was associated with older age (P < 0.001; HR, 1.02), large tumor size (P = 0.02; HR, 1.04), MacFarlane stage 3 (P < 0.001; HR, 4.27) or stage 4 (P < 0.001; HR, 5.83) (Fig. 1C). Secreting tumors were associated with a poor prognosis (P = 0.04; HR, 1.63), especially for glucocorticoid-secreting tumors (alone or associated with other steroids) (P < 0.001; HR, 2.33) and mineralocorticoid-secreting tumors (P = 0.01; HR, 1.96). Patients with more recent diagnoses since 1988 had higher survival rates (P < 0.001; HR, 0.58) as well as patients diagnosed as adrenal incidentalomas (P = 0.004; HR, 0.32). o,p’DDD treatment and the duration of symptoms before diagnosis were not associated with an effect on survival. However, in multivariate analysis (Table 4), the factors associated with a poor prognosis were: older age at diagnosis [HR, 1.03; (1.02–1.05); P < 0.0001], MacFarlane stage 3 [HR, 4.42; (1.91–10.22); P = 0.0005] and stage 4 [HR, 7.93; (3.36–18.72); P < 0.0001] tumors, and cortisol oversecretion [HR, 3.94; (2.22–6.99); P < 0.0001]. The effect of the period of diagnosis was not significant after adjusting for age, stage, cortisol secretion, and o,p’DDD treatment in the first 3 months after surgery. Nevertheless, an interaction between o,p’DDD treatment and cortisol secretion was found (P = 0.04), indicating that o,p’DDD decreased the risk of death [HR, 0.40; (0.17–0.95)] for cortisol-secreting tumors only (Table 4 and Fig. 1D). No significant interaction was found between gender and any variable included in the final model (Table 4).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The mean age of the patients at diagnosis (44.1 yr) was similar to that in most other series (5, 22). Some differences between studies in mean age at diagnosis may have arisen from the inclusion of children in some series. As adrenocortical cancer in children may have a different pathogenesis [for instance TP53 germline mutation (1)] and a different prognosis, we included only patients over the age of 10 yr at diagnosis in this study. A bimodal age distribution at diagnosis of ACC has been reported, with peaks occurring in the first and fourth decades of life (9, 23).
The clinical features of the patients in this series are similar to those of most other series (2, 5, 6, 7, 10, 22, 24, 25, 26). Clinically apparent endocrine symptoms were moderate or absent in some patients with high levels of steroid precursor or active hormone production. Only 109 of the 154 patients with a functional carcinoma presented endocrine symptoms. The prevalence of secreting tumors was higher in this series than in previously reported series of patients. This difference may be due to differences in the extent of hormonal investigations or to a referral bias, as hypersecreting tumors may be more frequently diagnosed in endocrine centers (6, 10) and nonfunctional tumors at oncology centers (2, 25). Functional tumors were more frequent in women, in accordance with previous reports (5, 6, 7, 22, 25). Glucocorticoids and/or androgens were the steroids most frequently oversecreted. The hypersecretion of aldosterone or estradiol was rare.
Recurrence may occur long after initial treatment. Venkatesh et al. (5) reported a case of recurrence 13 yr after initial treatment. In our series, one patient suffered metastases 12 yr after initial surgery. However, most recurrences and/or metastases were diagnosed within 5 yr of diagnosis and initial surgery. This observation may have important implications for patient information and follow-up. It also provides us with a rationale for determining the duration of adjuvant therapy after curative surgery. The most frequent sites of metastases were the liver, lung, and bone. Risk factors for the development of lung and liver metastases were similar and included cortisol hypersecretion.
Size was a prognostic factor in univariate analysis but not in multivariate analysis. This is most likely because patients with metastatic disease or a local invasion already have a large tumor and that tumor stage has a major impact on prognosis. Sex was not a predictive factor in our series, contrary to other series.
Age and initial stage of the tumor were among the best predictors of survival in this series, as in previous reports (23). Older patients and patients with tumors not limited to the adrenal gland (i.e. MacFarlane stages 3 and 4) at diagnosis had lower survival rates. In this cohort, cortisol secretion was a strong, independent factor associated with poor prognosis. The prognostic value of steroid secretion was unclear in previous studies; however, a recent report by Berruti et al. (28) also suggests that, in 72 patients with ACC, cortisol secretion is a bad prognostic factor (6, 7, 26, 27, 28). However, we were able to analyze this effect more precisely due to the large number of secreting tumors in this study. The poorer prognosis of cortisol-secreting tumors may be related to comorbidity with Cushing’s syndrome. It is also possible that the immunosuppressive effects of excess cortisol favor the development of the tumor and its metastases. Alternatively, the pathophysiology of cortisol-secreting ACC may lead to the growth of a more aggressive tumor. The lack of predictive value of tumor size and weight for survival in multivariate analysis may be accounted for by large tumors being more likely to be invasive. A previous report on the predictive value of tumor size did not include adjustment for tumor stage (26). Some studies have reported that sex affects survival rate (5, 6, 29). We found no such effect in this study.
Univariate analysis showed that survival rates were higher for patients diagnosed since 1988. However, this effect was not observed in multivariate analysis adjusted for stage at diagnosis. This suggests that tumors now may be diagnosed at earlier stages than in the past. Incidentally detected ACC also had a better survival, and this might be related to the fact that they were diagnosed at a earlier stage and in more recent years. Consistent with this hypothesis, 50% of patients presented metastases at diagnosis in older studies (22, 24) vs. less than 25% in more recent studies (6, 7). The investigation of adrenal incidentalomas may have facilitated the earlier diagnosis of ACC, thereby increasing survival rates. However, a French multicenter study carried out by surgeons suggested that the higher rates of survival observed after 1988 were due to advances in perioperative care (anesthesiology, intensive care) (7). Radical surgery with extensive resection is currently the best therapy for adrenal cancer. Radiotherapy does not seem to be very effective but may be useful as a palliative treatment for metastases (29, 30, 31). The antineoplastic action of o,p’DDD has been considered unpredictable, inconsistent, and restricted. Objective tumor regression has been reported with this treatment (5, 6, 8, 17). Pooled data obtained before 1993 suggest that about 35% of tumors respond to o,p’DDD (9). It has been suggested that patients in whom circulating o,p’DDD levels reach 14 µg/liter or more may be more responsive to therapy (17, 32). Previous studies have shown that o,p’DDD does not increase recurrence-free survival rates (11). Icard et al. (7) reported that o,p’DDD increased survival rates when given as a palliative treatment, but not as an adjuvant therapy after "curative" surgery. In the Anderson Cancer Center study, o,p’DDD was associated with a poorer prognosis when given as an adjuvant treatment for localized or regional tumors (12). However, Kasperlik-Zaluska et al. (10) reported that survival rates were higher for patients treated with o,p’DDD immediately after surgery than for patients treated later (2–15 months later). We demonstrate in this study, for the first time, that o,p’DDD can increase survival rates in the subgroup of patients with cortisol-secreting tumors. Venkatesh et al. (5) found no difference in response to o,p’DDD treatment of oversecreting and nonsecreting tumors, but they did not study specifically the subgroup of patients with cortisol oversecretion. Our study shows that cortisol oversecretion may be an adverse factor in itself. The beneficial effects of o,p’DDD treatment in these patients may be due to anticortisol effects, decreasing cortisol secretion, and attenuating Cushing’s syndrome. Excess glucocorticoid levels may play a role in the comorbidity of patients with ACC, affecting survival rates regardless of tumor growth. It is also possible that the adrenolytic action of o,p’DDD requires CYP11B activity within the tumor (33). This enzyme is probably expressed in cortisol-secreting tumors, accounting for the more potent adrenolytic effect of o,p’DDD in cortisol-secreting tumors. This suggests that the o,p’DDD could be more effective in patients with cortisol-secreting tumors due to the improvement of Cushing’s syndrome but also to a specific stronger anti-tumor effect.
In conclusion, this study of a large cohort of ACC patients from an endocrine center confirms the high prognostic value of age at diagnosis and initial stage and suggests that improvements in the outcome of ACC in the last 15 yr may have resulted mainly from earlier diagnosis. This study demonstrates, for the first time, that cortisol-secreting tumors are associated with poorer prognosis and that o,p’DDD treatment may be beneficial in patients with cortisol-secreting tumors.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
First Published Online May 2, 2006
Abbreviations: ACC, Adrenocortical carcinoma; HR, hazard ratio; o,p’DDD, 1,1-dichlorodiphenildichloroethane.
Received December 15, 2005.
Accepted April 24, 2006.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. Quinkler, S. Hahner, S. Wortmann, S. Johanssen, P. Adam, C. Ritte, C. Strasburger, B. Allolio, and M. Fassnacht Treatment of Advanced Adrenocortical Carcinoma with Erlotinib plus Gemcitabine J. Clin. Endocrinol. Metab., June 1, 2008; 93(6): 2057 - 2062. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Bertherat, J. Coste, X. Bertagna, G. Dickstein, C. Shechner, O. Nativ, J. E. Lee, A. Machens, H. Dralle, M. Terzolo, et al. Adjuvant Mitotane in Adrenocortical Carcinoma N. Engl. J. Med., September 20, 2007; 357(12): 1256 - 1259. [Full Text] [PDF]
|
|
|
|
 |
|
 R. Libe, A. Fratticci, and J. Bertherat Adrenocortical cancer: pathophysiology and clinical management Endocr. Relat. Cancer, March 1, 2007; 14(1): 13 - 28. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Libe, L. Groussin, F. Tissier, C. Elie, F. Rene-Corail, A. Fratticci, E. Jullian, P. Beck-Peccoz, X. Bertagna, C. Gicquel, et al. Somatic TP53 Mutations Are Relatively Rare among Adrenocortical Cancers with the Frequent 17p13 Loss of Heterozygosity Clin. Cancer Res., February 1, 2007; 13(3): 844 - 850. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Assie, G. Antoni, F. Tissier, B. Caillou, G. Abiven, C. Gicquel, S. Leboulleux, J.-P. Travagli, C. Dromain, X. Bertagna, et al. Prognostic Parameters of Metastatic Adrenocortical Carcinoma J. Clin. Endocrinol. Metab., January 1, 2007; 92(1): 148 - 154. [Abstract] [Full Text] [PDF]
|
|
Read all eLetters
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
