Pregnancy in Women Previously Treated for an Adrenocortical Carcinoma
Pauline de Corbière, * Katrin Ritzel,* Laure Cazabat, Jacques Ropers, Matthias Schott, Rosella Libé, Ann-Cathrin Koschker, Sophie Leboulleux, Timo Deutschbein, Christine Do Cao, Stefanie Hahner, Delphine Drui, Konstanze Miehle, Philippe Caron, Jens Waldmann, Olivier Chabre, Marcus Quinkler, Philippe Touraine, Maria C. Villares Fragoso, Jérôme Bertherat, Xavier Bertagna, Martin Fassnacht, and Marie-Laure Raffin-Sanson+
Context: Adrenocortical carcinomas (ACCs) are rare, aggressive tumors, of which some express receptors for estradiol, progesterone, and/or human chorionic gonadotoropin. Because this dis- ease is encountered frequently in young women, pregnancy is a relevant issue.
Objective: to evaluate the impact of pregnancy on outcome of patients previously treated for ACC.
Design/Setting: retrospective observational multicenter study of the European Network for the Study of Adrenal Tumors.
Patients: Seventeen ACC patients (21 pregnancies), becoming pregnant at least 3 months after the initial treatment, were compared with 247 nonpregnant ACC patients less than 47 years old. A control group of 34 patients matched for age, sex, and tumor stage was used for survival analysis.
Main Outcome Measure(s): Overall survival, tumors characteristics at diagnosis, pregnancy outcome.
Results: All 17 patients with pregnancies had localized ACC. The median time between surgery and conception was 4 years (0.3-12 y). Two pregnancies were terminated at 8 weeks. Sixteen women gave birth to 19 live infants. With exception of 1 (presumably unrelated) cardiac malformation, no severe fetal or maternal complication was observed. After a median follow-up time of 8.36 years and 5.26 years after the first conception, 1 of the 17 patients had died and 5 had experienced a recurrence, among whom 3 occurred before conception. Overall survival was not significantly different between the “pregnancy group” and the matched controls.
Conclusion: Pregnancy in patients previously treated for ACC seems to not be associated with worse clinical outcome, although a “healthy mother effect” cannot be excluded. (J Clin Endocrinol Metab 100: 4604-4611, 2015)
A drenocortical carcinoma (ACC) is a rare malignancy, with an estimated annual incidence between 1 and 2 cases per million inhabitants. These tumors are more fre- quent in women, with a sex ratio of 1.5 (1, 2), and are ob- served earlier in female than in male patients (3). The disease is frequently diagnosed in women of reproductive age. Al- though the general prognosis is poor, some patients may survive for long period (1, 4, 5). For this reason, few patients wish to become pregnant after the treatment of the tumor.
In general, being pregnant at the diagnosis of a cancer, or the occurrence of a pregnancy after the diagnosis of a cancer, does not increase cancer mortality, with some no- table exceptions such as breast cancer (6). For ACC, sev- eral clinical studies have suggested a relationship between adrenocortical proliferation and pregnancy. Benign and malignant adrenocortical tumors are responsible for 70% of all cases of Cushing’s syndrome diagnosed during preg- nancy (7, 8). A few cases of pregnancy-dependent cortisol
Received May 19, 2015. Accepted October 5, 2015.
First Published Online October 13, 2015
* P.d.C. and K.R. contributed equally to this work.
+ Author affiliations are shown at the bottom of the next page. Abbreviations: ACC, adrenocortical carcinoma; ENSAT, European Network for the Study of Adrenal Tumors; WG, weeks of gestation; yo, years old.
production by adrenal tumors have been reported (9-11). In vitro studies also point out the role of hormones in adrenal proliferation. Certain strains of mice develop ad- renocortical tumors in response to gonadectomy through LH receptor expression (12). The stimulation of the LH/ human chorionic gonadotropin receptor expressed into adrenocortical cells may be sufficient to induce prolifer- ation (13). Progesterone and estradiol receptors are ex- pressed by adrenal tumors at levels comparable with those observed in normal or tumoral breast cells (14), and the adrenocortical tumoral cell line H295R exhibits estrogen- sensitive proliferation (15, 16).
Moreover, we have shown that ACC diagnosed during pregnancy or in the immediate postpartum period is as- sociated with a worse prognosis than ACC diagnosed in nonpregnant women (3). This finding raises the question as to whether a pregnancy occurring after the initial treat- ment of the tumor would also increase the risk of relapse or progression. Women who survived ACC are frequently counseled, as a precautionary advice, against future con- ception. However, this is made in the absence of any sup- porting evidence, because no study was conducted until now to address this question.
In order to help substantiate the guidance given to pa- tients previously treated for an ACC who ask about the risk of childbearing, we undertook a retrospective study of all ACC patients included in the French, German, and Brazilian database of the European Network for the Study of Adrenal Tumors (ENSAT) who became pregnant at least 3 months after the initial treatment of their tumor.
We analyzed hormonal and radiological data collected at diagnosis and obstetrical outcome for the 17 women with ACC and at least 1 subsequent pregnancy. The evolution of these patients was compared with matched female controls with ACC, treated over the same period at one of the ENSAT centers.
Patients and Methods
Patients
The records of 270 women of reproductive age (15-47 y) diagnosed with ACC were retrieved from the French, German,
and Brazilian databases of the ENSAT. Six women diagnosed with ACC during a pregnancy or in the immediate postpartum were excluded from the study.
Among the remaining patients, 17 became pregnant at least once, more than 3 months after the diagnosis. Twenty-one preg- nancies were reported. Nonpregnant women were defined as the patients not pregnant at the time of the diagnosis and not preg- nant at any time during the follow-up (n = 247).
For each patient, ACC diagnosis was confirmed by histo- pathological examination of the tumor (3). Hormonal evalua- tion and assays were performed as previously described (3, 17). Tumor stage was determined with the ENSAT staging system (18). Metastases were diagnosed by imaging (mostly abdominal and chest computed tomography scans), and the results were confirmed by pathological examination when surgery was con- sidered appropriate.
The median duration from surgery to last follow-up was 8.4 years (3.2-33 y) for the patients with a pregnancy and 6.9 years (2.1-26.2 y) for the nonpregnant patients. The median follow-up from conception was 5.3 years (1.5-31.5 y).
The ENSAT ACC Registry was approved by the Ethics Com- mittees of the respective hospitals and by the French Commission Nationale Informatique et Liberté, and all patients gave in- formed consent.
Statistical analysis
Data are reported as mean (SE), median (min-max), or as percentages, as appropriate. Student’s t test and Fisher’s exact test were used to compare the pregnant patients (n = 17) and nonpregnant patients (n = 247) in terms of clinical presentation, hormonal secretion, and stage of the tumor at discovery. Due to substantial differences in age, year of diagnosis, and ACC stage between pregnant and nonpregnant patients, overall survival and disease-free survival were compared with a sample of 34 nonpregnant patients individually matched with pregnant women on age (±10 y), and ENSAT stage. Two nonpregnant controls were matched with each pregnant woman. The controls had also to have survived for at least as long as the time between ACC diagnosis and completion of pregnancy in the respective matched pregnant patient. Survival curves were constructed by the Kaplan-Meier method and were compared using stratified Log-rank test and Cox model.
Results
Age and presentation of patients at diagnosis (Table 1)
The mean age at diagnosis in the group of patients sub- sequently becoming pregnant was 28.12 ± 5.86 years
Department of Endocrinology (P.d.C., L.C., M .- L.R.S.), Ambroise Paré University Hospital, Assistance Publique Hôpitaux de Paris (AP-HP), 92100 Boulogne Billancourt, France; Medizinische Klinik IV (K.R., M.F.), Klinikum der Universität München, München 80336, Germany; Inserm U1173 (L.C., M .- L.R .- S.), Université de Versailles St-Quentin-en-Yvelines, 78280 Montigny- le-Bretonneux, France; Unité de Recherche Clinique (J.R.), Hospitalo Universitaire Paris Île-de-france Ouest, AP-HP, Hôpital Ambroise Paré, 92100 Boulogne, France; Funktionsbereich Spezielle Endokrinologie (M.S.), Universitätsklinikum Düsseldorf, 40225 Düsseldorf, Germany; Department of Endocrinology (R.L., J.B., X.B.), Cochin University Hospital, AP-HP, French Adrenal Cancer Network COrtico MEdullo-surrénale Tumeur Endocrines (COMETE)-Cancer and Unite Mixte de Recherche 8104, Institut Cochin, 75014 Paris, France; Comprehensive Cancer Center Mainfranken (A .- C.K., M.F.), University of Würzburg, Würzburg, 97080 Germany; Institut Gustave Roussy (S.L.), 94800 Villejuif, France; Department of Medicine I (T.D., M.F.), Endocrine Unit, University Hospital, University of Würzburg, Würzburg, 97080 Germany; Centre Hospitalier Régional Universitaire de Lille (C.D.C.), Lille, 59000 France; Endocrinology and Diabetes Unit (S.H), Department of Medicine I, University Hospital, University of Würzburg, 97080 Würzburg, Germany; Department of Endocrinology (D.D.), Nantes University Hospital, 44000 Nantes, France; Department of Endocrinology and Nephrology (K.M.), University of Leipzig, 04357 Leipzig, Germany; Service d’Endocrinologie Maladies Métaboliques et Nutrition (P.C.), Pôle Cardio-Vasculaire et Métabolique, Centre Hospitalier Universitaire Larrey, 10623 Toulouse, France; Division of Endocrinology and Diabetology (J.W.), University Hospital Giessen and Marburg, Campus Marburg, 35041 Marburg, Germany; Department of Endocrinology (O.C.), Michallon Teaching Hospital, 38043 Grenoble, France; Endocrinology in Charlottenburg (M.Q.), 10623 Berlin, Germany; Department of Endocrinology and Reproductive Medicine (P.T.), University Hospital Pitié Salpêtrière, 75013 Paris, France; and Adrenal Unit (M.C.V.F.), Neuroendocrine Unit and Laboratory of Medical Investigation 42, Division of Endocrinology and Metabolism, Hospital das Clinicas, University of São Paulo, 05403-000 São Paulo, Brazil
| Patients With Pregnancy After ACC (n = 17) (A) | Nonpregnant Women With ACC, 15-47 yo (n = 247) (B) | Women With ACC Matched 2:1 (n = 34) (C) | p (A) vs (B) | p (A) vs (C) | |
|---|---|---|---|---|---|
| Age, mean (SD) | 28.12 (5.86) | 34.28 (7.90) | 30.09 (7.42) | 0.002 | 0.344 |
| Period of diagnosis | 0.264ª | 1.000ª | |||
| ≤1999 | 3/17 (17.6) | 20/247 (8.1) | 5/34 (14.7) | ||
| 2000-2009 | 12/17 (70.6) | 173/247 (70) | 26/34 (76.5) | ||
| ≥2010 | 2/17 (11.8) | 54/247 (21.9) | 3/34 (8.8) | ||
| Size of the tumor | |||||
| Mean (SD) | 11.14 (5.28) | 11.02 (5.36) | 10.04 (4.56) | 0.927 | 0.442 |
| Staging (ENSAT) | 0.003b | 1.000b | |||
| Stage 1 | 2/17 (11.8) | 21/247 (8.5) | 4/34 (11.8) | ||
| Stage 2 | 13/17 (76.5) | 93/247 (37.7) | 26/34 (76.5) | ||
| Stage 3 | 2/17 (11.8) | 64/247 (25.9) | 4/34 (11.8) | ||
| Stage 4 | 0/17 (0.0) | 69/247 (27.9) | 0 (0.0) | ||
| Secretion (yes) | 13/17 (76.5) | 179/236 (75.8) | 22/34 (64.7) | 1.000 | 0.527 |
| Glucocorticoid (yes) | 7/17 (41.2) | 120/215 (55.8) | 17/34 (50.0) | 0.313 | 0.767 |
| Androgen (yes) | 10/17 (58.8) | 106/203 (52.2) | 17/34 (50.0) | 0.625 | 0.767 |
| Weiss | 0.296° | 0.575℃ | |||
| 0-2 | 0 | 9/135 (6.7) | 4/29 (13.8) | ||
| 3-5 | 9/16 (56.2) | 49/135 (36.3) | 13/29 (44.8) | ||
| 609 | 7/16 (43.7) | 77/135 (57.0) | 12/29 (41.4) | ||
| Ki67 | 0.128ª | 0.433ª | |||
| <5 | 4/11 (36.4) | 27/131 (20.6) | 7/25 (28.0) | ||
| 5-10 | 6/11 (54.5) | 44/131 (33.6) | 12/25 (48.0) | ||
| 11-20 | 0 (0) | 22/131 (16.8) | 5/25 (20.0) | ||
| >20 | 1/11 (9.1) | 38/131 (29) | 1/25 (4.0) | ||
| Curative surgery (R0) (%) | 17/17 (100.0) | 92/120 (76.7) | 31/31 (100.0) | 0.106 | 1.000 |
| Mitotane adjuvant (%) | 12/17 (70.6) | NA | 23 (69.7) | — | 1.000 |
The characteristics of the 34 matched controls allowing the survival analysis are also shown (group C).
a Value compares data from all diagnosis periods.
b Value compares data from all stages.
” Value compares data from all Weiss classes.
d Value compares data from all Ki67 classes.
(mean ± SD), significantly lower than the mean age of the nonpregnant patients 34.28 ± 7.90 (P = . 002). The pa- tients from both groups were mainly diagnosed with ACC between 2000 and 2010, with only 3/17 patients (17.7%) in the pregnancy group and 20/247 patients (8.1%) in the no pregnancy group, being diagnosed before 2000.
Hormonal data (Table 1)
Hormonal investigations showed that 13/17 patients in the pregnancy group had secreting tumors (76.5%), se- creting cortisol (n = 7) and/or androgens (n = 10), and/or mineralocorticoids or precursors (n = 2). We found that 179 (75.8%) of the 236 patients for whom the status is known in the “no pregnancy group” had a secreting tu- mor; 55.8% (120/215) of them secreted glucocorticoids and 52.2% (106/203) produced androgens (Table 1).
Tumor sizes, stage at diagnosis, and histological features (Tables 1)
The sizes of the tumors of patients from the pregnancy group, 11.14 ± 5.28 cm, were not different from those of the patients belonging to the no pregnancy group, 11.02 + 5.36 (P = . 927).
In contrast, patients becoming pregnant after the diag- nosis of ACC had less invasive tumors at diagnosis. Two
out of 17 (11.8%) patients with pregnancy had stage 1 tumors, 13 (76.5%) had stage 2 tumors, and 2 (11.8%) had stage 3 tumors. Thus, almost 90% of these patients displayed no local or metastatic extension at the time of diagnosis. In the no pregnancy group, less than half of the patients had a tumor limited to the adrenal gland at the time of diagnosis: 21/247 (8.5%) with stage 1 tumors and 93/247 (37.7%) with stage 2 tumors, whereas 133/247 (53.8%) patients displayed local or metastatic extension: 64 of 247 (25.9%) with stage 3 and 69 of 247 (27.9%) with stage 4 tumors. The distribution of ENSAT stages in the 2 groups was significantly different (P = . 003), clearly indicating the need of stage-matched controls (Table 1).
Weiss score was available for 16 patients with a preg- nancy: 9/16 tumors (56.3%) had a Weiss score between 3 and 5, whereas 7/16 (43.8%) had a Weiss score of 6 or more. In the no pregnancy group, Weiss score tended to be higher with 77/135 patients (57%) having a Weiss score of 6 or above (P = . 296) (see Table 1). Similarly, quantitation of the proliferation marker Ki67 distribution tended to show lower results in patients from the pregnancy group with P = . 061 (see Table 1). However, for both histo- pathological parameters the matched controls were quite similar to the pregnancy group (Table 1).
Treatment
All patients of the pregnancy group underwent surgical treatment and had a complete resection of the tumor. Twelve of these 17 patients (71%) also received adjuvant mitotane therapy after surgery (median duration 33 mo). Treatment began either 2 or 3 weeks before surgery for 3 patients, or within 3 months after surgery in the other patients. Five patients received mitotane at the time of a recurrence after an apparently curative initial surgery (see below). In the matched control group, 69.7% of patients were treated adjuvantly with mitotane.
Pregnancy outcome (Table 2)
Thirteen women had 1 pregnancy and 4 women 2 preg- nancies. The median interval between initial ACC surgery and conception was 48 months (4-144) for the first preg- nancy and 66 months (48-84) for the second pregnancy.
Mean maternal age at delivery or abortion was 33.3 years (23-47).
Two pregnancies were terminated at 8 weeks because of concerns about the negative impact of mitotane on the fetus in one case, and fears about a negative influ- ence of the pregnancy on tumor progression in the sec- ond case.
Sixteen women gave birth to 19 live infants after 19 full-term pregnancies: 9 males and 10 females, 5 of them
born by cesarean section. No pregnancy except 1 (ter- minated at 8 weeks of gestation [WG]) began under mitotane therapy. The treatment was interrupted either because of poor tolerance or long-term remission (n = 11), or to allow the pregnancy (n = 1). In this last case, the blood concentration of mitotane was monitored, and undetectable plasma levels were obtained before conception. Seven patients had persistent adrenal in- sufficiency during the gestation and were given appro- priate substitution. No severe maternal or fetal compli- cation was observed. Pregnancy was uncomplicated in 12 cases. Maternal morbidity consisted in self-limiting placental abruption, mild hypertension, gestational di- abetes (no insulin treatment required and no fetal con- sequence), threat of premature delivery at 33 WG (end- ing with term delivery, at 37 WG, of a healthy baby), Graves’ disease (controlled by medical treatment with- out fetal consequences), mild hydramnios, and postpar- tum hemorrhage (no transfusion required), each event occurring in 1 patient (see Table 2).
Fetal cases of morbidity included 1 small for age infant (2610 g at 40 WG) and 1 cardiac malformation (pulmo- nary artery [PA] stenosis and aortic stenosis) requiring cardiac surgery at day 7 with further normal development of an otherwise healthy child (7 years old [yo] at last fol- low-up). No preterm birth was observed.
| Patient | Delay Conception/1st Surgery (y) | Delivery | Delivery/Term (W) | Fetal Complications | Maternal Complications |
|---|---|---|---|---|---|
| P1-G1 | 2 | VD | 39.5 | None | None |
| P1-G2 | 4.5 | VD | 37 | None | None |
| P2 | 3 | VD | 40 | None | None |
| P3 | 2 | Cesar | 38 | None | Hypertension |
| P4 | 5 | Cesar | 40 | None | None |
| P5-G1 | 4 | TAª | 8.5 | — | |
| P5-G2 | 5 | VD | 36 | None | None |
| P6-G1 | 4.5 | VD | 41 | None | Postpartum hemorrhage |
| P6-G2 | 7 | VD | 38.5 | None | None |
| P7 | 0,6 | VD | 40 | Small for age (2620 g) | Hydramnios |
| P8 | 0.3 | VD | 40 | Aortic + PA stenosis“ | None |
| P9 | 2 | Cesar | 42 | None | None |
| P10 | 2 | TAb | 8 | — | |
| P11 | 1.5 | VD | 37 | None | Premature delivery threat (33 w) |
| P12 | 6 | VD | 38 | None | Relapse Graves' disease in PP |
| P13 | 8 | VD | 40.5 | None | None |
| P14 | 5 | VD | 40 | None | Gestational diabetes |
| P15-G1 | 4 | VD | 41 | None | None |
| P15-G2 | 6 | VD | 40 | None | Placental abruption |
| P16 | 12 | Cesar | 39 | None | Graves' disease |
| P17 | 4 | Cesar | 40 | None | None |
Cesar, cesarean section; VD, vaginal delivery; TA, therapeutic abortion; PA, pulmonary artery; PP, postpartum.
a Conception under mitotane therapy.
b Concerns about tumoral evolution.
” Surgery at day 7, in good health at 7 yo.
ACC-related outcome
One of the 17 women with pregnancies died during the follow-up period (median follow-up, 8.1 y [2.8-33]). Five of them experienced a relapse of the tumor.
In 3 patients, ACC relapsed before the pregnancy. P12 underwent surgery at 17 yo (2003) for an androgen and cortisol-secreting ACC, diameter 13 cm, ENSAT stage 3, Weiss score 3, and Ki67 = 10%. She received an irradi- ation of the tumor bed but no mitotane. In September 2006, surgical resection of a pulmonary metastasis was performed. In January 2010, she delivered a healthy child at 38 WG of an uncomplicated pregnancy. No further relapse was observed until now.
P16 underwent at 26 yo the presumed complete resec- tion of a cortisol- and androgen-secreting tumor measur- ing 15 cm, ENSAT stage 3, and Weiss score 8. Mitotane was begun shortly afterward. Three months later, a local recurrence with peritoneal invasion and hepatic metasta- sis were diagnosed. She was treated by further surgery and systemic chemotherapy (cisplatine and vepeside for 6 mo) together with mitotane and radiotherapy. Ten years later, no relapse was observed, and mitotane was stopped to allow a pregnancy. Blood mitotane concentration was un- detectable at the time of conception. The pregnancy was uncomplicated with substitution of adrenal insufficiency, except for Graves’ disease. She delivered a healthy girl at 39 WG. No further progression was noted at last follow- up, 17 months after conception.
P13 was diagnosed in 1998 at age 18, with a 4-cm ACC, Weiss score 5, and ENSAT stage 1. Adjuvant mitotane therapy was given. In June 2001, 2 hepatic metastases were treated by chemotherapy and chemoembolization with cisplatin and lipiodol. Six hepatic metastases in June 2002 and 16 metastases in July 2003 were treated the same way. No further tumor progression was observed. She conceived in 2006 and gave birth to a male infant after an uncomplicated pregnancy. In December 2013, a breast carcinoma was diagnosed. P53 sequence was normal.
Two patients relapsed after delivery. P7 was treated by surgical resection of a stage 2 ACC, (diameter, 6 cm; Weiss 5) in 2011. Eight months later, she became pregnant and delivered at 40 WG a healthy female infant. One month after delivery, lung metastases were detected and surgi- cally removed followed by adjuvant mitotane. However, few months later she developed mediastinal metastases, and in April 2013, cisplatine, etoposide, and doxorubicine (19) were added. After a period of stabilization, the disease progressed in fall 2013. She was then treated later with gemcitabine and capecitabine but died in August 2014.
P10 was diagnosed with a cortisol-secreting ACC at age 39 (2006) and underwent resection of a stage 2 tumor (Weiss 5, Ki67 score 30%). Mitotane was given for 1 year.
One year later, she became pregnant but decided to inter- rupt the pregnancy at 8 GW. Three years later, liver me- tastasis and retroperitoneal lymph node metastases were diagnosed and resected. Mitotane was resumed. At the last imaging (November 2014), she was without evidence for disease.
To assess the influence of the pregnancy on tumor evo- lution, each patient with pregnancy was matched with 2 patients without pregnancy, as indicated above.
As shown on Table 1, pregnant women and matched controls were quite similar for all parameters potentially influencing tumor outcome: age, period of diagnosis, stage (ENSAT), Weiss, Ki67, glucocorticoid secretion, surgical resection, and adjuvant mitotane treatment. The selected controls were to have experienced a survival at least as long as the time between ACC diagnosis and completion of pregnancy of their matched pregnant patient.
The occurrence of a pregnancy was not associated with a shorter survival. On the contrary, only 1 woman with pregnancy died during the follow-up period as compared with 8 women in the control group, stratified Log-rank test (P = . 15) (Figure 1A).
Disease-free survival tends also to be shorter in the 34 control patients without pregnancy than in the 17 patients with pregnancy, although the difference was not signifi- cant either; stratified Log-rank test, P = . 54 (Figure 1B). The median recurrence-free survival could not be esti- mated, because less than half of the patients in both groups experienced a relapse.
Discussion
We studied here for the first time, the outcome of young women treated for a rare and aggressive adrenal tumor and becoming pregnant during the follow-up.
Mammary and adrenal carcinogenesis are both af- fected by the hormonal context of pregnancy. For the pa- tients whose breast cancer is diagnosed during a preg- nancy, the risk of dying from the tumor is significantly greater than that of women who had never been pregnant, even when controlling for stage and hormone receptor status (6, 20). In a similar manner, it was demonstrated that 12 patients with ACC diagnosed during pregnancy or in the postpartum period had in comparison with a control group more advanced tumors at diagnosis and that these tumors displayed a worse prognosis. Even when compared with patients matched for age and stage, increased risk of cause-specific death remained (3).
In contrast, we observe here that a subsequent preg- nancy in patients previously treated for an ACC seems to not increase the risk of recurrence or death. This discrep-
A
Disease free survival
1.00
0.75
H+++
1
Survival
0.50
0.25
No pregnancy group Pregnancy group
0.00
0
5
10
15
20
25
30
35
B
Overall survival
1.00
I
l
+1-
I
0.75
Survival
0.50
0.25
No pregnancy group Pregnancy group
0.00
0
5
10
15
20
25
30
35
| Time (years) | |||||||
|---|---|---|---|---|---|---|---|
| Number at risk | |||||||
| No pregnancy group 34 | 18 | 9 | 3 | 1 | 1 | 0 | 0 |
| Pregnancy group 17 | 12 | 3 | 2 | 1 | 1 | 1 | 0 |
| Time (years) | ||||||||
|---|---|---|---|---|---|---|---|---|
| Number at risk | ||||||||
| No pregnancy group 34 | 25 | 11 | 3 | 1 | 1 | 0 | 0 | |
| Pregnancy group 17 | 15 | 6 | 3 | 1 | 1 | 1 | 0 | |
Figure 1. Disease-free survival (A) and overall survival (B) in women with ACC, pregnant during the follow-up (dotted curves), or matched controls not pregnant during the follow-up (solid curves).
ancy was also observed for breast cancer: a pregnancy occurring after the initial treatment of the tumor does not impair the overall survival (21). Thus, for both types of malignancy, the particular aggressiveness of pregnancy- associated tumors appears limited to those developing during a gestation.
Is it therefore possible to reassure women treated for an ACC that the pregnancy will not have a negative impact on her survival? A few points prompt us to remain cautious.
Women considering a future pregnancy are a self-se- lected population, introducing an obvious bias. One of those biases named “healthy mother effect” refers to the
fact that only women who feel in good health will undertake a preg- nancy. They may constitute a popu- lation with a notably better survival, which may mask an increased risk of relapse induced by the pregnancy.
We tried to avoid such a bias by matching the patients with pregnan- cies with controls having compara- ble prognosis; in fact, the controls had the same characteristics at diag- nosis as the women with pregnan- cies, even for the Ki67 and the pro- portion of cortisol-secreting tumors, not taken into account at the match- ing. The matched controls had to survive at least the duration between diagnosis and pregnancy of their matching cases. We voluntarily de- cided not to match controls with dis- ease-free survival for as long as the time from diagnosis to pregnancy of the matched cases because, in the group of patient with pregnancies, 3 had a relapse before becoming preg- nant. Thus choosing control patients without relapse would have intro- duced an obvious bias. Other fac- tors, not taken into account for the matching may be associated with a better prognosis in patients wishing child bearing. For example, being married has been shown in many studies to be a marker of a better sur- vival after a cancer as compared with the risk for single, separated, and di- vorced patients (22, 23). Socioeco- nomic factors have not been consid- ered in the present study.
Some women with pregnancy ex- perienced an unexpectedly long sur- vival after a recurrence. As some of them relapsed before the pregnancy, this should be partly explained by a self- selection bias. The effort to match patients with the same prognosis is probably more effective in selecting subjects with the same risk of relapse after a first treatment than subjects with the same risk of progression after a relapse.
It is noteworthy that the patients with pregnancies had initially localized tumors with a relatively “good” prog- nosis (ENSAT stage 1 or 2 have, respectively, a 5-y survival of 81% and 61%). It cannot be ascertained that the op- timistic results obtained here could be extended to more advanced tumors.
Even if the pregnancy does not impact negatively the prognosis of women with ACC, it must be emphasized that the overall survival remains poor and that some women may not be able to care for their children.
This study included 17 cases, which can be considered as a large study considering the rarity of this tumor; only a multicentric and international recruitment and the ex- istence of the ENSAT database made this work possible. Although, the small size of the series still remains a limi- tation that prevents a definitive reassuring conclusion, it is very unlikely that in the near future a much larger series will provide better data.
Although the difference is not significant, the women being pregnant during the follow-up tend to display a bet- ter prognosis. This tendency, also observed for pregnancy occurring after the treatment of a breast cancer, may be due to the above discussed healthy mother effect. How- ever, a long-term protective effect of the pregnancy cannot be excluded. The hormonal milieu of pregnancy includes hormones capable of exerting a stimulating effect on tu- moral cells such as human chorionic gonadotropin and estradiol. However, hormonal, metabolic, and immuno- logical changes associated with gestation are complex and their long-term effects cannot be predicted. Another hy- pothesis, also proposed for breast cancer, is referred to as “alloimmunization.” Tumoral adrenocortical cells might display a gene expression profile mimicking fetal adrenal cells (24, 25). During pregnancy, fetal cells reach the ma- ternal circulation, and the immune stimulation against these cells could stimulate the immune system against the dormant metastatic cells. However, this remains speculative.
The fetal outcome was good in this study, except for 2 therapeutic abortions. One patient was advised to inter- rupt the pregnancy because the conception occurred while taking mitotane. The available publications on the subject are limited to few case reports, without confirmation of this legitimate concern (26-29). However, the data are too few to draw conclusions. In the remaining 19 full-term pregnancies, no adverse effect was observed. Noteworthy, 12 of 17 patients had received mitotane for different pe- riods of time (most of them for several years) before con- ception with very few data about mitotane plasma levels. In all term pregnancies, the treatment was stopped more than 1 year before conception. One child was born with an aortic and PA stenosis requiring cardiac surgery at day 7. The risk of this malformation in the general population is 1/1500 live births and a link with previous mitotane ther- apy seems to be unlikely. No sign or symptom of adrenal insufficiency was observed in the newborns. However, this optimistic conclusion needs to be tempered with cau- tion, considering the small size of the series.
The decision regarding how long the patients should wait after the end of the treatment before they consider conceiving is difficult. For lysodren, it seems prudent to ensure undetectable mitotane plasma level, which usually takes from 1 to 3 years. For platinum derivates or etopo- side whose half-life is very short, the concern is more about the prognosis of the tumor; for ethical reasons, it is advised to obtain a long term remission (5 y?) before considering parenthood.
Thus, although patients being pregnant after the treat- ment of an ACC did not experience, in our study, a worse survival than women not pregnant during the follow-up, a definitive answer cannot be given, considering that se- lection bias may have contributed to the reduced risk of death observed here and that the small size of the study remains a limitation.
However, our data may suggest that women with ACC in the capacity to become pregnant may not only be un- harmed by pregnancy but may actually have an improved outcome as compared with women without such capacity. At this time, we are not able to evaluate whether this is caused by a self-selection mechanism or a biological im- pact of pregnancy. Disentangling these 2 potential mech- anisms will be a major hurdle for future epidemiological studies.
Acknowledgments
We thank the IT support of Anthony Stell (University of Mel- bourne, Australia), in retrieving the data from the ENSAT ACC registry. We also thank Michaela Haaf for documenting a sig- nificant proportion of patients.
Address all correspondence and requests for reprints to: Marie- Laure Raffin-Sanson, MD, PHD, Service d’Endocrinologie, Hôpital Ambroise Paré, 9 Avenue Charles de Gaulle, 92100, Boulogne, France. E-mail: marie-laure.raffin-sanson@aphp.fr.
This work was supported by European Network for the Study of Adrenocortical Tumors funding, the European Union (Sev- enth Framework Programme [FP7/2007-2013] under Grant Agreement 259735), and the Deutsche Forschungsgemeinschaft Grant FA 466/4-1 (to M.F.).
Disclosure Summary: The authors have nothing to disclose.
References
1. Tritos NA, Cushing GW, Heatley G, Libertino JA. Clinical fea- tures and prognostic factors associated with adrenocortical car- cinoma: Lahey Clinic Medical Center experience. Am Surg. 2000; 66(1):73-79.
2. Fassnacht M, Libé R, Kroiss M, Allolio B. Adrenocortical carci- noma: a clinician’s update. Nat Rev Endocrinol. 2011;7(6):323- 335.
3. Abiven-Lepage G, Coste J, Tissier F, et al. Adrenocortical carcinoma
and pregnancy: clinical and biological features and prognosis. Eur J Endocrinol. 2010;163(5):793-800.
4. Assié G, Letouzé E, Fassnacht M, et al. Integrated genomic charac- terization of adrenocortical carcinoma. Nat Genet. 2014;46(6): 607-612.
5. Fassnacht, M, Johanssen S, Fenske W, et al. Improved survival in patients with stage II adrenocortical carcinoma followed up pro- spectively by specialized centers. J Clin Endocrinol Metab. 2010; 95(11):4925-4932.
6. Guinee VF, Olsson H, Möller T, et al. Effect of pregnancy on prog- nosis for young women with breast cancer. Lancet. 1994; 343(8913):1587-1589.
7. Guillaume B, Sanson ML, Billaud L, Bertagna X, Laudat MH, Lu- ton JP. Cushing’s syndrome and pregnancy: aetiologies and prog- nosis in twenty-two patients. Eur J Med. 1992;1(2):83-89.
8. Lindsay JR, Nieman LK. Adrenal disorders in pregnancy. Endocri- nol Metab Clin North Am. 2006;35:1-20, v.
9. Wallace C, Toth EL, Lewanczuk RZ, Siminoski K. Pregnancy-in- duced Cushing’s syndrome in multiple pregnancies. J Clin Endocri- nol Metab. 1996;81:15-21.
10. Kasperlik-Zaluska AA, Szczupacka I, Leszczynska-Bystrzanowska J, Drus-Przybyszewska G. Pregnancy-dependent Cushing’s syn- drome in three pregnancies. BJOG. 2000;107:810-812.
11. Wy LA, Carlson HE, Kane P, Li X, Lei ZM, Rao CV. Pregnancy- associated Cushing’s syndrome secondary to a luteinizing hormone/ human chorionic gonadotropin receptor-positive adrenal carci- noma. Gynecol Endocrinol. 2002;16:413-417.
12. Bielinska M, Parviainen H, Porter-Tinge SB, et al. Mouse strain susceptibility to gonadectomy-induced adrenocortical tumor for- mation correlates with the expression of GATA-4 and luteinizing hormone receptor. Endocrinology. 2003;144:4123-4133.
13. Mazzuco TL, Chabre O, Feige JJ, Thomas M. Aberrant GPCR ex- pression is a sufficient genetic event to trigger adrenocortical tu- morigenesis. Mol Cell Endocrinol. 2007;265-266:23-28.
14. de Cremoux P, Rosenberg D, Goussard J, et al. Expression of pro- gesterone and estradiol receptors in normal adrenal cortex, adre- nocortical tumors, and primary pigmented nodular adrenocortical disease. Endocr Relat Cancer. 2008;15:465-474.
15. Montanaro D, Maggiolini M, Recchia AG, et al. Antiestrogens up- regulate estrogen receptor ß expression and inhibit adrenocortical H295R cell proliferation. J Mol Endocrinol. 2005;35:245-256.
16. Sirianni R, Zolea F, Chimento A, et al. Targeting estrogen receptor-« reduces adrenocortical cancer (ACC) cell growth in vitro and in vivo: potential therapeutic role of selective estrogen receptor modulators
(SERMs) for ACC treatment. J Clin Endocrinol Metab. 2012;97: E2238-E2250.
17. Ronchi CL, Sbiera S, Volante M, et al. CYP2W1 is highly expressed in adrenal glands and is positively associated with the response to mitotane in adrenocortical carcinoma. PLoS One. 2014;9:e105855.
18. Fassnacht M, Johanssen S, Quinkler M, et al. Limited prognostic value of the 2004 International Union Against Cancer staging clas- sification for adrenocortical carcinoma: proposal for a Revised TNM Classification. Cancer. 2009;115:243-250.
19. Fassnacht, M, Terzolo M, Allolio B, et al. Combination chemother- apy in advanced adrenocortical carcinoma. New Engl J Med. 2012; 366:2189-2197.
20. Rodriguez AO, Chew H, Cress R, et al. Evidence of poorer survival in pregnancy-associated breast cancer. Obstet Gynecol. 2008;112: 71-78.
21. Azim HA Jr, Santoro L, Pavlidis N, et al. Safety of pregnancy fol- lowing breast cancer diagnosis: a meta-analysis of 14 studies. Eur J Cancer. 2011;47:74-83.
22. Shor E, Roelfs DJ, Bugyi P, Schwartz JE. Meta-analysis of marital dissolution and mortality: reevaluating the intersection of gender and age. Soc Sci Med. 2012;75:46-59.
23. Wang L, Wilson SE, Stewart DB, Hollenbeak CS. Marital status and colon cancer outcomes in US Surveillance, Epidemiology and End Results registries: does marriage affect cancer survival by gender and stage? Cancer Epidemiol. 2011;35:417-422.
24. Rainey WE, Carr BR, Wang ZN, Parker CR. Gene profiling of human fetal and adult adrenals. J Endocrinol. 2001;171:209-215.
25. Rosenberg D, Groussin L, Jullian E, et al. Transcription factor 3’,5’- cyclic adenosine 5’-monophosphate-responsive element-binding protein (CREB) is decreased during human adrenal cortex tumori- genesis and fetal development. J Clin Endocrinol Metab. 2003;88: 3958-3965.
26. Kojori F, Cronin CM, Salamon E, Burym C, Sellers EA. Normal adrenal function in an infant following a pregnancy complicated by maternal adrenal cortical carcinoma and mitotane exposure. J Pe- diatr Endocrinol Metab. 2011;24:203-204.
27. Luton JP, Remy JM, Valcke JC, Laudat P, Bricaire H. [Recovery or remission of Cushing’s disease following long term administration of OP’DDD in seventeen patients (author’s transl)]. Ann Endocri- nol. 1973;34:351-376.
28. Knappe G, Gerl H, Ventz M, Rohde W. [The long-term therapy of hypothalamic-hypophyseal Cushing’s syndrome with mitotane (o,p’-DDD)]. Dtsch Med Wochenschr. 1997;122:882-886.
29. Tripto-Shkolnik L, Blumenfeld Z, Bronshtein M, Salmon A, Jaffe A. Pregnancy in a patient with adrenal carcinoma treated with mito- tane: a case report and review of literature. J Clin Endocrinol Metab. 2013;98:443-447.