EJE
Outcome of immunotherapy in adrenocortical carcinoma: a retrospective cohort study
Hanna Remde, 1,+(D Laura Schmidt-Pennington,2,1 Miriam Reuter,1 Laura-Sophie Landwehr,10 Marie Jensen,2 Harald Lahner,3 Otilia Kimpel,10D Barbara Altieri,1 Katharina Laubner,4 Jochen Schreiner,5 Joerg Bojunga,6 Stefan Kircher,7 Catarina Alisa Kunze,8 Anne Pohrt,9 Maria-Veronica Teleanu, 1º Daniel Hübschmann, 11,12,13 Albrecht Stenzinger, 14
Hanno Glimm, 15,16,17,18 Stefan Fröhling, 10,13 Martin Fassnacht, 1,19[D Knut Mai,2,*+(D and Matthias Kroiss 1,5,t
1Department of Internal Medicine I, Division of Endocrinology and Diabetes, University Hospital, University of Würzburg, Josef-Schneider- Straße 2, 97080 Würzburg, Germany
2Department of Endocrinology & Metabolism, Charité-Universitätsmedizin Berlin, corporate member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Berlin 10117, Germany
3Department of Endocrinology and Metabolism, University Hospital Duisburg-Essen, Hufelandstraße 55, 45147 Essen, Germany 4Division of Endocrinology and Diabetology, Department of Medicine II, Faculty of Medicine, Medical Centre University of Freiburg, University of Freiburg, Hugstetter Straße 55, 79106 Freiburg, Germany
5Department of Internal Medicine IV, University Hospital Munich, Ludwig-Maximilians-University München, Ziemssenstraße 5, 80336 München, Germany
6Department of Internal Medicine 1, Division of Endocrinology, Goethe University Frankfurt, Faculty 16 Medicine, Theodor-Stern-Kai 7, 60596 Frankfurt am Main, Germany
7Institute of Pathology, University of Würzburg, Josef-Schneider-Str. 2, 97080 Würzburg, Germany
8Institute of Pathology, Charité-Universitätsmedizin Berlin, corporate member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Charitéplatz 1, 10117 Berlin, Germany
9Institute of Biometry and Clinical Epidemiology, Charité-Universitätsmedizin Berlin, Charitéplatz 1, 10117 Berlin, Germany
10National Center for Tumor Diseases Heidelberg and German Cancer Research Center, Im Neuenheimer Feld 460, Heidelberg 69120, Germany 11Computational Oncology Group, Molecular Precision Oncology Program, National Centre for Tumour Diseases (NCT), Heidelberg and German Cancer Research Centre (DKFZ), Im Neuenheimer Feld 580, 69120 Heidelberg, Germany
12Pattern Recognition and Digital Medicine Group, Heidelberg Institute for Stem cell Technology and Experimental Medicine (HI-STEM) gGmbH, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany
13German Cancer Consortium, Im Neuenheimer Feld 280, Heidelberg 69120, Germany
14Institut für Pathologie, Universitätsklinikum Heidelberg, Im Neuenheimer Feld 224, Heidelberg 69120, Germany
15Department for Translational Medical Oncology, National Centre for Tumour Diseases (NCT/UCC), Dresden, Germany
16Translational Medical Oncology, Faculty of Medicine and University Hospital Carl Gustav Carus, Technische Universität Dresden, Dresden, Germany
17Translational Functional Cancer Genomics, National Centre for Tumour Diseases (NCT) and German Cancer Research Center (DKFZ), Heidelberg, Germany
18 Department for Translational Medical Oncology (DD05), National Center for Tumor Diseases Dresden (NCT/UCC), German Cancer Consortium (DKTK), Fetscherstraße 74/PF 64, 01307 Dresden, Germany
19Comprehensive Cancer Center Mainfranken, University of Würzburg, Würzburg, Berlin, Germany
*Corresponding author: Department of Endocrinology & Metabolism, Charite-Universitätsmedizin Berlin, Chariteplatz 1, Berlin 10117, Germany. Email: knut. mai@charite.de
Abstract
Objective: Clinical trials with immune checkpoint inhibitors (ICI) in adrenocortical carcinoma (ACC) have yielded contradictory results. We aimed to evaluate treatment response and safety of ICI in ACC in a real-life setting.
Design: Retrospective cohort study of 54 patients with advanced ACC receiving ICI as compassionate use at 6 German reference centres between 2016 and 2022.
Methods: Objective response rate (ORR), disease control rate (DCR), progression-free survival (PFS), overall survival (OS), and treatment-related adverse events (TRAE) were assessed.
Results: In 52 patients surviving at least 4 weeks after initiation of ICI, ORR was 13.5% (6-26) and DCR was 24% (16-41). PFS was 3.0 months (95% CI, 2.3-3.7). In all patients, median OS was 10.4 months (3.8-17). 17 TRAE occurred in 15 patients, which was associated with a longer PFS
+ These authors contributed equally to this work.
of 5.5 (1.9-9.2) vs 2.5 (2.0-3.0) months (HR 0.29, 95% CI, 0.13-0.66, P= 0.001) and OS of 28.2 (9.5-46.8) vs 7.0 (4.1-10.2) months (HR 0.34, 95% CI, 0.12-0.93). Positive tissue staining for programmed cell death ligand 1 (PD-L1) was associated with a longer PFS of 3.2 (2.6-3.8) vs 2.3 (1.6-3.0, P < 0.05) months. Adjusted for concomitant mitotane use, treatment with nivolumab was associated with lower risk of progression (HR 0.36, 0.15-0.90) and death (HR 0.20, 0.06-0.72) compared to pembrolizumab.
Conclusions: In the real-life setting, we observe a response comparable to other second-line therapies and an acceptable safety profile in ACC patients receiving different ICI. The relevance of PD-L1 as a marker of response and the potentially more favourable outcome in nivolumab-treated patients require confirmation.
Keywords: immune checkpoint inhibitor, treatment, mitotane, PD-L1, adverse drug reaction
Significance
The tumour microenvironment in most adrenocortical carcinoma (ACC) cases is immunologically cold. Clinical trials of different immune checkpoint inhibitors (ICI) have reported highly divergent response rates. We here retrospectively analyse the outcome of ACC treatment with ICI in a multi-centre real-world setting in a series that exceeds the largest previous trial in size. We find a modest objective response rate (ORR) of 13.5%, identify programmed cell death ligand 1 (PD-L1) expres- sion as a potential treatment-related prognostic marker, and demonstrate association of response with immune-related ad- verse events in ACC. Given its overall manageable toxicity profile, ICI can be considered as an appropriate second-line treatment for ACC.
Introduction
Adrenocortical carcinoma (ACC) is a rare endocrine malig- nancy originating from the adrenal cortex.1 About 50% of ACC patients are diagnosed with unresectable disease.2 Even after complete resection, 50% of ACCs recur locally or develop distant metastases.3 Due to limited treatment op- tions in metastatic disease, the median overall survival (OS) is as low as 15 months in these patients. Current guidelines recommend mitotane, the only officially approved drug in ACC, as first-line therapy in advanced, non-resectable ACC either as monotherapy or in combination with etopo- side, doxorubicin, and cisplatin (EDP-M).1,4 In the seminal FIRM-ACT trial, median progression-free survival (PFS) still was only 5 months with EDP-M and rates of objective response and disease control were 23.2% and 58.3%, re- spectively.5 Gemcitabine and capecitabine6,7 or streptozoto- cin8 constitute the most commonly used second- and third-line therapy options,9 but response rates are around 10% only. Recently, promising response rates have also been reported for multikinase inhibitors (MKI) in retro- spective series.10,11 Further-line treatments are only effective in small patient cohorts.12
While immune checkpoint inhibitors (ICI) have revolution- ized the treatment of many solid tumours, 13 5 prospective tri- als so far investigated the safety and efficacy of different ICI in a limited number of ACC patients.14-18 Outcome was hetero- geneous with an objective response rate (ORR) between 33% in a very small trial with 6 patients18 and 6% in the currently largest study investigating 50 patients.16 However, these re- sults suggest that at least a subgroup of ACC patients benefits from ICI,1 but predictive factors for therapy response remain to be identified.
Immune cell infiltration has recently been shown to be asso- ciated with better OS and recurrence-free survival in localized ACC, whereas glucocorticoid excess was associated with im- mune cell depletion and unfavourable prognosis.19 Microsatellite instability-high or mismatch repair (MMR) de- ficient tumours have been shown to respond to pembrolizu- mab across entities,20,21 while expression of the immune checkpoint molecule PD-L1 has been demonstrated to be asso- ciated with response in some malignancies.22 For none of these
markers established in other malignancies, an association with response in ACC could so far be demonstrated.
Given that selection bias may account for some of the het- erogeneity in prospective trials and that relevant clinical data were only partially reported, we here challenged these trials by retrospectively analysing a well-characterized cohort of more than 50 ACC patients receiving ICI outside of prospect- ive trials. We aimed (1) to evaluate the efficacy (assessed as best objective response) of ICI treatment in advanced ACC in a “real-world setting” and (2) to determine a potential prog- nostic impact of steroid hormone excess, mitotane therapy and its plasma levels, programmed cell death 1 (PD-1) and its ligand PD-L1 expression status, MMR status, and tumour mutational burden (TMB) on treatment response.
Materials and methods
Study population
We retrospectively evaluated all patients who received ICI out- side of prospective trials as compassionate treatment and for whom a minimum dataset was available at 6 reference centres in Germany between 2016 and 2022. Data on the course of ICI treatment, local radiographic response assessment, treatment-related adverse events (TRAE), tumour histology, hormonal assessment, and previous therapies were obtained from the European Network for the Study of Adrenal Tumours (ENSAT) registry or by retrospective chart review. Diagnosis of ACC was histologically confirmed in all cases, and all patients had metastatic disease at ICI treatment initi- ation. The inclusion of patients was independent of previously administered anti-tumour therapies including concomitant mitotane use.
The choice of the ICI administered was made by the treating physician according to the recommendation of the local multi- disciplinary tumour board.
Treatment costs were covered by the patients’ health insur- ances after a request for compassionate use in most cases.
Response assessment
Response was assessed retrospectively according to radiologic reports of computed tomography (CT), magnetic resonance
imaging (MRI), and/or positron emission tomography (PET) with CT or MRI. For the assessment of PFS, we applied a land- mark approach of 4 weeks without documented clinical pro- gression to account for the delay observed in anti-tumoural activity of ICI. Overall, there were 74 staging examinations, of which 38 (51.4%) were CT, 10 (13.5%) were MRI + CT, 25 (33.8%) were PET-CT, and 1 (1.4%) was MRI + PET-CT. Out of 14 patients receiving more than 1 staging, 10 had the same modality throughout the study, while 4 had changing modalities.
After each staging, radiologic assessment was conducted by the local treating physicians and response was classified as progressive disease (PD), stable disease (SD), partial response (PR), complete response (CR), and mixed response (MR). To operationalize the clinical judgement of MR for analysis of re- sponse similar to RECIST 1.123 and in view of the phenom- enon of pseudo-progression during ICI treatment,24 MR was assessed according to the following scheme: If the majority of tumour lesions had increased since the last evaluation or new lesions had occurred, the respective staging was consid- ered as PD, even if some lesions were smaller than before. Otherwise, the consecutive staging under treatment was taken into account. In case of MR followed by PD (n =4), the re- spective staging was considered as PD. In case of MR followed by PR (n = 2), the respective staging was judged as SD assum- ing pseudo-progression. A single case of MR followed by progression after discontinuation of treatment was classified as SD.
Study endpoints
The primary endpoint was ORR, defined as the proportion of patients with CR or PR. Secondary endpoints were disease control rate (DCR: the proportion of patients with CR, PR, or SD), PFS (time from the first day of treatment to first docu- mented disease progression or death, whichever occurred first), and OS (time from the first day of treatment to death from any cause). Patients with ongoing response at the end of the study period or lost to follow-up were censored at the date of last assessment.
Ethics approval
Written informed consent for the retrospective study and pub- lication was obtained from all included patients via the ENSAT registry (approval by the ethics committee of the University of Würzburg 88/13, Charité Berlin EA2/021/20, LMU Munich 379/10) or the local ethics committee waived approval of anonymized data usage. The study was conducted in accordance with the Helsinki declaration.
Statistical analysis
Data are displayed as median (range) for continuous variables and as percentage for categorical variables if not stated other- wise. Average mitotane levels were obtained by adding up all available levels during ICI therapy and dividing the total by their number. Clopper-Pearson confidence intervals (CI) were calculated for ORR and DCR. To assess survival, Kaplan-Meier plots were created and log-rank tests were com- puted to assess differences between groups. Cox regression analysis was performed to calculate hazard ratios (HR) for PFS and OS. Spearman’s coefficient was calculated to assess correlation between ordinally scaled non-parametric
variables. Fisher’s exact test was used to assess group differen- ces between categorical variables. The Mann-Whitney U test was used to compare groups regarding continuous variables. P < . 05 was considered significant. Statistical analysis was car- ried out using IBM SPSS statistics.25
Supplemental methods
Further information on “mitotane treatment”, “hormone ex- cess”, “immunohistochemistry”, “potential tissue markers of treatment response”, and “tumour mutational burden” can be found in the supplements.
Results
Patients and response
A total of 57 patients received ICI therapy between April 2016 and February 2022. After exclusion of 3 patients with essential data missing, 54 patients were included into the study (Table 1). The most commonly used ICI was pembrolizumab (n=32, 59%) followed by nivolumab (n=13, 24%). Detailed information on treatment and response can be ob- tained from Table 2.25 52/54 patients (96%) had documented PD before ICI therapy was started, and 2/54 had SD. Six pa- tients did not undergo any staging due to early clinical pro- gression and were considered as PD; 2 of them did not survive 4 weeks after treatment initiation and were excluded from response assessment as per pre-specified landmark. In the remaining 48 patients, imaging was performed after a me- dian of 2.5 months (range 0.5-6.2), which was judged clinical- ly as PR in 4, MR in 5, and SD in 7 cases, respectively. Thirty-two patients (67%) had PD. A second staging was per- formed in 14 patients 5.3 months (range 3.5-7.8) after start of therapy and demonstrated ongoing PR in 2 and new PR in 3, PD in 7, and MR in 2 cases. One patient with MR therapy was discontinued due to pneumonitis, and the other had a third im- aging consistent with classification of PD.
Out of 52 patients surviving at least 4 weeks after initiation of ICI treatment, the best objective response was PR in 7 of them, leading to an ORR of 13.5% (95% CI, 6-26). Seven pa- tients showed SD resulting in a DCR of 26.9% (95% CI, 16-41). Median PFS was 3.0 months (range 0.2-18; censored in 5 patients with ongoing disease control at data cut-off), and median OS was 10.4 months (range 0.2-56; censored in 24 patients alive at data cut-off; Figure 1). In patients with an objective response (n=7, Table 3), median OS was 28.4 months (95% CI, 2.6-54.1) and PFS was 10.4 months (95% CI, 4.9-15.9).
Potential factors of treatment-related prognosis
No association of PFS and OS was observed with age, tumour size, Weiss score, and Ki-67 index at first diagnosis as well as time since first diagnosis. Female patients had a significantly longer PFS than male patients (median: 3.2 [95% CI, 2.1-4.3] vs 2.3 [95% CI, 1.6-3.0] months, log-rank P <. 05), but OS was not different. Glucocorticoid excess at first diag- nosis (Supplemental methods) was neither associated with PFS nor OS. Furthermore, no significant differences in OS and PFS were observed between patients with more than 3 sys- temic therapies and those with 3 or less.
| Characteristics | Data availability (n) | |
|---|---|---|
| Female sex n (%) | 54 | 31 (57%) |
| Age at diagnosis (years) | 54 | 46 (19-70) |
| ENSAT stage at first diagnosis | 50 | |
| ENSAT I | 2 (4%) | |
| ENSAT II | 19 (38%) | |
| ENSAT III | 10 (20%) | |
| ENSAT IV | 19 (38%) | |
| Histology | ||
| Initial tumour size (cm) | 42 | 14 (3-26) |
| Weiss score | 22 | 7.5 (3-9) |
| ki67 | 48 | 30 (5-80) |
| Steroid hormone excess at first | 47 | |
| diagnosisª | ||
| Glucocorticoids ± others | 22 (47%) | |
| Sex steroids only | 6 (13%) | |
| Aldosterone only | 2 (4%) | |
| No steroid excess | 17 (36%) | |
| Previous therapies | 54 | |
| Initial surgery | 50 (93%) | |
| Mitotane | 54 (100%) | |
| Number of additional systemic | 3 (1-8) | |
| therapies | ||
| Etoposide/doxorubicin/cisplatin | 53 (98%) | |
| Streptozotocin | 42 (78%) | |
| Gemcitabine/capecitabine | 37 (69%) | |
| Otherb | 14 (26%) | |
| Line of ICI therapy | ||
| Second line | 9 (17%) | |
| Third line | 10 (18%) | |
| Fourth line | 22 (41%) | |
| Fifth line | 10 (18%) | |
| >Fifth linec | 3 (6%) | |
| Surgery for recurrence/metastasis | 29 (54%) | |
| Radiotherapy for recurrence or | 20 (37%) | |
| metastasis | ||
| Other local therapiesd | 9 (17%) | |
| Documented progression after | 52 (96%) | |
| preceding therapy n (%)e | ||
| Sites of metastases before treatment | 50 | |
| initiation | ||
| Lung | 39 (78%) | |
| Liver | 32 (64%) | |
| Other intra-abdominalf | 36 (72%) | |
| Bone | 13 (26%) | |
| Soft tissue | 9 (18%) | |
| Brain | 2 (4%) | |
| Pleural carcinosis | 2 (4%) | |
Data are median (range) or frequency (%).
aCombined clinical and laboratory data.
bCabozantinib (n =7), trofosfamide (n = 3), docetaxel (n= 1),
(R)-1-[1-(4-[123I]iodophenyl)ethyl]-1H-imidazole-5-carboxylic acid azetidinyl amide therapy (IMAZA)25 (n = 1), sunitinib (n = 1), temozolomide (n = 1), thalidomide (n = 1), trabectedin + olaparib (n= 1), trametinib (n = 1), and different study agents.
“ICI for 2/54 patients as sixth line, for 1/54 patients as nineth line. dBrachytherapy (n = 4), radio frequency ablation (n = 1), hyperthermia (n = 1), microwave ablation (n = 1), transarterial chemoembolization (n = 1), and histoacryl embolization (n= 1).
eIn 1 patient termination of EDP after 8 cycles presumably due to toxicity. “Including abdominal lymph nodes, intra- and retroperitoneal tumour bulks, local recurrence, and unresected primary tumour.
Value of tumour markers of immune response
Tissue markers of immunohistochemistry and TMB are re- ported in Table 4. Patients with positive PD-L1 staining dem- onstrated significantly longer PFS compared to those negative for PD-L1 (median: 3.2 [95% CI, 2.6-3.8] vs 2.3 [95% CI,
| Characteristics | Data availability | (n) |
|---|---|---|
| Time from first diagnosis to start of ICI (months) | 54 | 31 (3-231) |
| Immune checkpoint inhibitora | 54 | |
| Pembrolizumab | 32 (59%) | |
| Nivolumab | 13 (24%) | |
| Avelumab | 6 (11%) | |
| Atezolizumab | 1 (2%) | |
| Ipilimumab & nivolumab | 2 (4%) | |
| Number of doses administered | 50 | 4 (1-49) |
| Concomitant mitotane therapy n (%) | 54 | 13 (24%) |
| Other concomitant therapiesb n (%) | 54 | 8 (15%) |
| Best objective response | 52 | |
| Partial response | 7 (13.5%) | |
| Stable disease | 7 (13.5%) | |
| Progressive disease | 38 (73%) |
Data are median (range) or frequency (%).
aNivolumab was administered every 2 weeks intravenously with a dose of 240 mg. Pembrolizumab was administered every 3 weeks intravenously with a dose of 200 mg. Avelumab was administered every 2 weeks intravenously with doses calculated with 10 mg/kg body weight.
bRadiation therapy (n =2), EDP (n = 1), gemcitabine (n = 1), lenvatinib (n = 1), surgery for metastasis (n= 1), (R)-1-[1-(4-[123I]iodophenyl) ethyl]-1H-imidazole-5-carboxylic acid azetidinyl amide therapy (IMAZA) (n=1),25 transarterial chemoembolization (n =1), and radiotherapy with carboplatin weekly (n = 1).
progression-free survival [%] >
100
Median PFS: 3.0 months (95% CI 2.3-3.7)
50
0
0
5
10
15
20
time since treatment initiation [months]
B
overall survival [%]
100
Median OS: 10.4 months (95% CI 3.8-17.0)
50
0
0
20
40
60
time since treatment initiation [months]
1.6-3.0] months, log-rank P <. 05; Figure S1). However, no such difference was observed for OS. PD-1 staining was posi- tive in only 1 out of 18 patients. Three out of 32 patients
| ID | Age at diagnosis | Time since first diagnosis | Number of previous systemic treatments | Metastasis locations | Hormonal activity | Substance | Concomitant mitotane | Other concomitant therapies | IHC | Molecular analysis | Time to progression | survival Time of | Adverse events | Reason for discontinuation of ICI therapy |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 4 | 33 | 2.2 yrs. | 3 | PUL, LYM, | Glucocorticoids | Nivolumab | No | Radiochemotherapy | PD-L1: TPS = 0, IC-S =5, | TMB high | 10 months | 23 | Skin rash, | AE |
| other (abdominal tumour | + other steroids | months | cryptogenic organizing pneumonia | |||||||||||
| bulk) | CPS = | |||||||||||||
| 5; | ||||||||||||||
| MMR | ||||||||||||||
| 6 | 70 | 3.6 yrs. | 2 | HEP, LYM | Aldosterone | Nivolumab | Yes (average | Debulking surgery | pos. PD-L1 | N.A. | 5 months | Alive at | Colitis | AE |
| level 16.1) | neg .; | 12 | ||||||||||||
| MMR | months | |||||||||||||
| 7 | 64 | 2.8 yrs. | 3 | HEP, PUL | None | Avelumab | Yes (average level 13.9) | None | pos. PD-L1: TPS = | TMB high | Ongoing at 18 months | Alive at 18 | Increase of creatinine | Ongoing at data cut-off (18 months) |
| 2, IC-S =2, CPS = | months | |||||||||||||
| 4, | ||||||||||||||
| MMR | ||||||||||||||
| 9 | 20 | 2.6 yrs. | 5 | HEP, PUL, | Glucocorticoids | Avelumab | No | None | pos. PD-L1 | TMB low | 5 months | 10 | None | PD |
| LYM | + other steroids | neg., | months | |||||||||||
| MMR | ||||||||||||||
| 14 | 39 | 16.9 yrs. | 4 | BRA, PUL, | Glucocorticoids + other steroids | Nivolumab | No | None | pos. PD1 | TMB low | 18 months | 28 | Polymyalgia | AE |
| other | neg, | months | ||||||||||||
| (abdominal | PD-L1 | |||||||||||||
| mass) | lung | |||||||||||||
| neg., brain | ||||||||||||||
| TC | ||||||||||||||
| 25%, | ||||||||||||||
| MMR | ||||||||||||||
| 25 | 66 | 0.3 yrs | 1 | OSS, HEP, PUL, (+primary tumour) | None | Pembrolizumab | No | None | pos. N.A. | N.A. | 10 months | Alive at 11 months | Adrenal insufficiency | PD |
| 27 | 28 | 5.4 yrs | 3 | HEP, PUL, other (pleural carcinosis) | Glucocorticoids + other steroids | Pembrolizumab | No | None | N.A. | N.A. | Ongoing at 5.5 months | Alive at 5.5 months | Colitis | Ongoing at data cut-off (5.5 months) |
Response to previous treatment was PD in all patients with objective response to immune checkpoint inhibitor therapy.
AE, adverse event; BRA, brain; CPS, combined positive score, TPS + IC-S; HEP, hepatic; IC-S, immune cell score; therapy, immunohistochemistry; LYM, lymph nodal; MMR, mismatch-repair; N.A., not available; PD, progressive disease; PD-1, programmed death 1; PD-L1, programmed death ligand 1; PUL, pulmonary; OSS, bone; TMB, tumour mutational burden; TPS, tumour proportion score.
| Characteristics | Data availability (n) | n (%) |
|---|---|---|
| Immunohistochemistry | ||
| PD-1 | 18 | 1 (6%) |
| PD-L1 | 33 | 8 (24%) |
| Loss of MMR protein expressiona | 32 | 3 (9%) |
| molecular markers | ||
| Tumour mutational burden high | 36 | 11 (31%) |
| presence of a molecularly informed | 51 | 18 (35%) |
| treatment rationale for ICIb |
ICI, immune checkpoint inhibitor; MMR, mismatch-repair; PD-1, programmed death 1; PD-L1, programmed death ligand 1; MSH2, MutS homolog 2; MSH6, MutS homolog 6; MLH1, MutL protein homolog 1; PMS2, PMS1 homolog 2.
aOne case each of isolated loss of MSH6, MSH6, and MSH2. One case of a MSH2 kilobase single nucleotide variant (SNV) and 1 case of loss of MLH1 and PMS2, as well as a germline mutation in MLH1 (Lynch syndrome). bAt least 1 out of the immunohistochemistry markers or tumour mutational burden.
| Event | Grade 1-2 (n = all/ responders) | Grade 3 (n = all/ responders) |
|---|---|---|
| Adrenal insufficiency | 1/1 | 0 |
| Colitis | 0 | 2/2 |
| Creatinine increased | 1/1 | 0 |
| Cryptogenic organizing pneumonia | 0 | 1/1 |
| Diarrhoea | 1/0 | 0 |
| Oedema and fever | 1/0 | 0 |
| Haemolytic anaemia | 0 | 1/0 |
| Hypothyroidism | 1/0 | 0 |
| Itching | 1/0 | 0 |
| Pneumonitis | 0 | 2/1 |
| Polymyalgia | 0 | 1/1 |
| QT interval prolongation | 1/0 | 0 |
| Skin rash | 3/1 | 0 |
showed loss of MMR protein expression at immunohisto- chemistry. These patients experienced PD at first staging or even before. TMB was considered high in 11 out of 36 patients (31%). No significant difference in PFS or OS was observed between patients with and without a high TMB. When classifying results of immunohistochemistry and TMB (Supplemental Methods) with at least 1 of them being patho- logic (reflecting any therapy rationale), a non-significant trend towards longer median PFS (3.2 [95% CI, 2.4-4.1] vs 2.5 [95% CI, 1.9-3.2] months, P =. 098) was observed in those with molecular alteration but OS was not different.
Adverse events
A total number of 17 adverse events at least possibly related to treatment occurred in 15 patients (28%) including 7 grade III TRAE (Table 5). No grade IV or V TRAE was observed. TRAE led to postponing of treatment cycles in 2 patients and treatment discontinuation in 7 patients (13%). TRAE oc- curred significantly more frequently in patients with concomi- tant mitotane treatment compared to those without (54% vs 20%, P <. 05) and in patients treated with nivolumab com- pared to pembrolizumab (54% vs 13%, P <. 01). Patients with at least 1 TRAE had a significantly longer median dur- ation of therapy of 5.4 months (0.5-30.2) vs 2.1 months
progression-free survival [%] >
100
Median PFS (months):
TRAE 5.5 (95% CI 1.9-9.2)
No TRAE 2.5 (95% CI 2.0-3.0)
Log rank P=0.001
50
0
0
5
10
15
20
time since treatment initiation [months]
B
overall survival [%]
100
Median OS (months):
TRAE 28.2 (95% CI 9.5-46.8)
No TRAE 7.2 (95% CI 4.1-10.2)
Log rank P=0.030
50
0
0
20
40
60
time since treatment initiation [months]
(0.2-8.5, P <. 05) and received a significantly higher median number of 11 (2-49) administered doses of ICI therapy vs 4 (1-12, (P <. 01) in patients without TRAE. The occurrence of an adverse event was associated with a significantly longer median PFS of 5.5 (95% CI, 1.9-9.2) vs 2.5 (95% CI, 2.0-3.0) months (P =. 001; HR 0.29, 95% CI, 0.13-0.66) and signifi- cantly longer median OS of 28.2 (95% CI, 9.5-46.8) vs 7.2 (95% CI, 4.1-10.2) months (P <. 05; HR 0.34, 95% CI, 0.12-0.93; Figure 2).
Impact of immune checkpoint inhibitor and mitotane
All patients had received prior mitotane treatment, which was continued during ICI in 13 cases (24%). Mitotane plasma con- centrations were available for 25 patients (46%) independent of concomitant mitotane intake. No significant difference in PFS and OS was observed between patients with (potentially) therapeutic and those with sub-therapeutic or non-therapeutic mitotane levels (Supplemental Methods).
Mitotane use during immunotherapy differed significantly between patients treated with different ICI (P =. 043) and was more frequently applied in patients treated with nivolu- mab compared to other ICIs (7/13 vs 6/41 patients, P =. 040). After adjustment for concomitant mitotane use, we observed a significantly lower risk of death (HR 0.29, 95% CI, 0.09-0.87) and higher risk of tumour progression (HR 0.62, 95% CI, 0.30-1.27) in patients treated with
nivolumab compared to other ICIs (Figure S2A, B). Furthermore, risks of death (HR 0.20, 95% CI, 0.06-0.72) and progression (HR 0.36, 95% CI, 0.15-0.90) were signifi- cantly lower in patients treated with nivolumab compared to those treated with pembrolizumab (Figure S2C, D). After ad- justment for nivolumab use, no significant impact of mitotane on OS and PFS was observed (Figure S3A, B). However, when considering only the 2 most frequently used ICIs pembrolizu- mab and nivolumab, concomitant mitotane treatment was as- sociated with a significantly higher risk of death (HR 3.12, 95% CI, 1.06-9.18) and of progression (HR 2.28, 95% CI, 0.93-5.56) (Figure S3C, D). Age, sex, number of previous sys- temic therapies, Ki-67, initial tumour size, Weiss score, time since first diagnosis, and staging intervals did not differ be- tween patients treated with nivolumab and pembrolizumab. Exploratory analyses for potential confounders did not reveal association of survival with age, number of previous therapies, or time since first diagnosis.
Discussion
The analysis of this so far largest published ACC cohort treated with ICIs, which also is the first in a real-life setting, shows a moderate efficacy and acceptable safety profile. We conclude that ICIs can be considered an appropriate second- line treatment option in advanced ACC.
Previous prospective clinical trials reported ORR and DCR from 6% to 33% and from 48% to 66%,15-18 respectively. We here demonstrate a compatible ORR of 13.5% (95% CI, 6-26) and DCR of 24% (95% CI, 16-41) in the real-life setting.
Across studies, median PFS ranged from 1.8 to 2.6 months14-16 which primarily reflects the interval to the first imaging in 2 of these studies.14,15 Similarly, we find a median PFS of 2.5 months (range 0.2-18) with a median time to first imaging of 2.6 months (range 0.5-6.3). The lower median OS of 10.4 months (range 0.2-56) of our study in comparison to previous reports (10.6 months16 and 24.9 months15) might be due to longer time since first diagnosis (2.6 vs 1.6 and 1.3 years, respectively) and a relatively high proportion of our pa- tients (44%) being still alive at last follow-up.
With a median number of 3 previous systemic treatments other than mitotane and a median time since first diagnosis of 31 months, our patient cohort likely suffered from selection bias towards less aggressive ACC. At the same time, the high number of previous therapies, the long time since first diagno- sis, and the mostly short time of survival demonstrate that in most of our patients, ICI was a salvage therapy initiated in a very late stage of disease when most of the “standard second- and third-line therapies” had already failed. Nevertheless, the therapy efficacy of ICI (ORR 13.5% and PFS 2.5 months) is at least comparable with the frequently applied second- and third-line therapy regimens of gemcitabine/capecitabine or streptozotocin, that result in response rates of 5%-10% and a median PFS of 2 to 4 months.9
Our series included patients treated with atezolizumab (n = 1), avelumab (n = 6), and combination of ipilimumab and ni- volumab (n = 2). Primarily, however, pembrolizumab (n = 32) and nivolumab (n = 13) were applied.
Until now, published data on nivolumab in ACC are scarce. A phase II study including 10 patients had to be terminated for futility because no confirmed PR could be observed. In that study, 2 patients (20%) had SD and 1 had unconfirmed
PR.14 In this study, efficacy of nivolumab may have been missed because of the limited sample size and concomitant mi- totane use. More frequent dosing of nivolumab might account for increased efficacy in the immunologically cold ACC.
Of note, we observed a more favourable outcome both in terms of overall and progression-free survival for nivolumab when focusing on the 45 patients treated with nivolumab or pembrolizumab. Although this finding needs to be treated with caution due to the retrospective nature of our study, nivo- lumab may have significant activity in ACC patients and war- rants further studies. Nivolumab in ACC is currently under investigation in a prospective trial in combination with a novel therapeutic vaccine or alone (NCT04187404).
Concomitant mitotane treatment was excluded in the pem- brolizumab trials,15,17 whereas half of the cohort in the avelu- mab trial continued mitotane treatment.16 Interestingly, in the latter study, 2 out of 3 responders had simultaneous mitotane treatment. In the present study, mitotane during ICI therapy appears to be associated with higher risk of progression and death although the latter was not significant when the entire population was taken into account. Furthermore, the preva- lence of TRAE was significantly higher among patients with concomitant mitotane treatment compared to those without. Discontinuation of mitotane should, therefore, be discussed in patients in whom control of steroid excess does not man- date mitotane use.
We observed TRAE in 31% of patients, which is less than previously reported in studies on ICIs in ACC.14-18 This is most likely due to underreporting of grade I and II TRAE in this retrospective setting, which also would explain the rela- tively high proportion of grade III TRAE. Most reported TRAE are likely to be immune related (Table 3) and led to dis- continuation in 7/15 patients. However, most patients recov- ered quickly after discontinuation of ICI (and in some cases glucocorticoid treatment) suggesting a safety profile of ICI in ACC comparable to other tumour entities. Most interestingly, patients with an adverse event showed a significantly longer PFS and OS. This is in accordance with previous studies re- porting higher rates of immune-related adverse events in res- ponders, both in ACC15,17 and in other entities.26,27
As only a small fraction of ACC patients benefits from ICI, the identification of potential responders is essential. Various predictors of treatment response have been proposed and ob- served in other tumour entities.20-22 In the present study, a moderately but nevertheless significantly longer PFS was observed in patients with positive PD-L1 staining. No associ- ation with treatment response was observed for loss of MMR proteins, PD-1 status, or TMB. Similarly, we did not detect a significant difference between glucocorticoid-producing and hormonally inactive tumours. Thus, no additional evidence could be found for an impact of glucocorticoid secretion in association with a more limited immune infiltrate on the response to immunotherapy as has been previously sug- gested.19,28 However, these results might not be generalizable as glucocorticoid excess was evaluated only at first diagnosis of ACC and 24% of patients were treated with mitotane dur- ing ICI therapy. Furthermore, the sample size might still be too small for such an analysis.
We here find a response rate above 10% for ICI monother- apy. At variance, no objective response was observed in pro- spective clinical trials of MKI.29,30 In light of the overall moderate efficacy of ICI alone, combination of pembrolizu- mab with the multi-kinase inhibitor (MKI) lenvatinib,
however, has yielded a response rate of 25% and a PFS of 5.5 months (n = 8) even after progression on 1 of the treatments in single-agent use in a retrospective case series.31 Together with the favourable outcome of the MKI cabozantinib in mono- therapy,10 combinations of checkpoint inhibitors with MKI could leverage on the efficacy of single drug regimens.
Finally, several limitations of this study should be high- lighted. This includes the retrospective evaluation with hetero- geneity of treatment regimens and staging procedures as well as partly reduced data availability. Furthermore, due to reim- bursement issues, ICI were employed after exhaustion of sev- eral prior lines of treatment. In addition to that, inclusion of several ICI regimes might conceal a stronger effect of selected ICIs and did not allow to sufficiently disentangle compound- specific effects. However, especially in rare diseases like ACC, studies like this are crucial to evaluate the current clin- ical practice and provide a basis for future trials and further improvement of treatment options.
In conclusion, we demonstrated a moderate treatment re- sponse and an acceptable safety profile of ICI in extensively pre-treated ACC patients. Outcome of ICI is at least compar- able with frequently applied therapy regimens like gemcita- bine/capecitabine or streptozotocin. Therefore, we consider ICI as an alternative second line to those possibly more toxic regimens as well as to MKI. Although no clear-cut prognostic factors of treatment response could be defined, the occurrence of TRAE was associated with longer PFS and OS. This sup- ports efforts to continue effective ICI therapy in case of man- ageable TRAE. Furthermore, nivolumab which had been studied in only 10 patients previously as well as the impact of mitotane on response to ICI requires further studies to dis- entangle their respective effects and to come to a firmer conclu- sion regarding the discontinuation of mitotane prior start of ICI therapy.
Acknowledgments
The authors thank Michaela Haaf and Martina Zink for main- taining the ENSAT registry database and biobank.
Author’s contributions
Study conception: H.R., L.S .- P., M.J., M.F., K.M., and M.K. Data and sample collection: H.R., L.S .- P., M.R., L.L., M.J., H.L., O.K., B.A., K.L., J.S., J.B., F.M., K.M., and M.K. Tissue and data analysis: H.R., L.S .- P., S.K., A.P., M .- V.T., D.H., A.S., H.G., S.F., K.M., M.K., and K.C.A. Manuscript draft: H.R. and M.K. All authors provided critical input and contributed to the manuscript.
Supplementary material
Supplementary material is available at European Journal of Endoicrinology online.
Funding
This work has been supported by the DFG German Research Foundation Project 314061271-TRR 205, by the European Reference Network on Rare Endocrine Conditions (Endo-ERN), and the European Reference network on rare adult solid cancers (EuRaCan). This work was supported by the National Center for Tumor Diseases (NCT) Molecular Precision Oncology Program.
Conflicts of interest: K.M. received research support by Ipsen (paid to the University Hospital Würzburg) and consulting fees or speaker honoraria by HRA Pharma, Roche, Eisai, Bayer, Advanz Pharma, and Recordati. The University Hospitals in Würzburg and Munich are participating in a clinical trial spon- sored by Enterome, Paris, France. S.F. received research fund- ing from AstraZeneca, Pfizer, PharmaMar, and Roche; consulting fees from Bayer, Illumina, and Roche; honoraria from Amgen, Eli Lilly, PharmaMar, and Roche; and travel or accommodation expenses from Amgen, Eli Lilly, Illumina, PharmaMar, and Roche. F.M. received research grants from Enterome, MSD, and Recordati and consulting fee (paid to the University Hospital Würzburg) from HRA Pharma and Bayer Pharma. K.C.A. received speaker honoraria by Ipsen. The other authors declare to have no competing interests.
Data availability
All datasets generated and/or analysed during the current study are not publicly available but are available from the cor- responding author upon reasonable request.
References
1. Fassnacht M, Assie G, Baudin E, et al. Adrenocortical carcinomas and malignant phaeochromocytomas: ESMO-EURACAN Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2020;31(11):1476-1490. https://doi.org/10.1016/j.annonc. 2020.08.2099
2. Kebebew E, Reiff E, Duh QY, Clark OH, McMillan A. Extent of disease at presentation and outcome for adrenocortical carcinoma: have we made progress? World J Surg. 2006;30(5):872-878. https:/ doi.org/10.1007/s00268-005-0329-x
3. Berruti A, Grisanti S, Pulzer A, et al. Long-term outcomes of adju- vant mitotane therapy in patients with radically resected adreno- cortical carcinoma. J Clin Endocrinol Metab. 2017;102(4): 1358-1365. https://doi.org/10.1210/jc.2016-2894
4. Fassnacht M, Dekkers OM, Else T, et al. European Society of Endocrinology Clinical Practice Guidelines on the management of adre- nocortical carcinoma in adults, in collaboration with the European Network for the Study of Adrenal Tumors. Eur J Endocrinol. 2018;179(4):G1-G46. https://doi.org/10.1530/EJE-18-0608
5. Fassnacht M, Terzolo M, Allolio B, et al. Combination chemother- apy in advanced adrenocortical carcinoma. N Engl J Med. 2012;366- (23):2189-2197. https://doi.org/10.1056/NEJMoa1200966
6. Henning JEK, Deutschbein T, Altieri B, et al. Gemcitabine-based chemotherapy in adrenocortical carcinoma: a multicenter study of efficacy and predictive factors. J Clin Endocrinol Metab. 2017;102- (11):4323-4332. https://doi.org/10.1210/jc.2017-01624
7. Sperone P, Ferrero A, Daffara F, et al. Gemcitabine plus metronom- ic 5-fluorouracil or capecitabine as a second-/third-line chemother- apy in advanced adrenocortical carcinoma: a multicenter phase II study. Endocr Relat Cancer. 2010;17(2):445-453. https://doi.org/ 10.1677/ERC-09-0281
8. Khan TS, Imam H, Juhlin C, et al. Streptozocin and o, p’DDD in the treatment of adrenocortical cancer patients: long-term survival in its adjuvant use. Ann Oncol. 2000;11(10):1281-1287. https://doi. org/10.1023/A:1008377915129
9. Ardolino L, Hansen A, Ackland S, Joshua A. Advanced adrenocort- ical carcinoma (ACC): a review with focus on second-line therapies. Horm Cancer. 2020;11(3-4):155-169. https://doi.org/10.1007/ s12672-020-00385-3
10. Kroiss M, Megerle F, Kurlbaum M, et al. Objective response and prolonged disease control of advanced adrenocortical carcinoma with cabozantinib. J Clin Endocrinol Metab. 2020;105(5): 1461-1468. https://doi.org/10.1210/clinem/dgz318
11. Miller KC, Chintakuntlawar AV, Hilger C, et al. Salvage therapy with multikinase inhibitors and immunotherapy in advanced ad- renal cortical carcinoma. J Endocr Soc. 2020;4(7):bvaa069. https://doi.org/10.1210/jendso/bvaa069
12. Megerle F, Kroiss M, Hahner S, Fassnacht M. Advanced adreno- cortical carcinoma-what to do when first-line therapy fails? Exp Clin Endocrinol Diabetes. 2019;127(2-03):109-116. https://doi. org/10.1055/a-0715-1946
13. Vaddepally RK, Kharel P, Pandey R, Garje R, Chandra AB. Review of indications of FDA-approved immune checkpoint inhibitors per NCCN Guidelines with the level of evidence. Cancers (Basel). 2020;12(3):738. https://doi.org/10.3390/cancers12030738
14. Carneiro BA, Konda B, Costa RB, et al. Nivolumab in metastatic adrenocortical carcinoma: results of a phase 2 trial. J Clin Endocrinol Metab. 2019;104(12):6193-6200. https://doi.org/10. 1210/jc.2019-00600
15. Raj N, Zheng Y, Kelly V, et al. PD-1 Blockade in advanced adreno- cortical carcinoma. J Clin Oncol. 2020;38(1):71-80. https://doi.org/ 10.1200/JCO.19.01586
16. Le Tourneau C, Hoimes C, Zarwan C, et al. Avelumab in patients with previously treated metastatic adrenocortical carcinoma: phase 1b results from the JAVELIN solid tumor trial. J Immunother Cancer. 2018;6(1):111. https://doi.org/10.1186/s40425-018-0424-9
17. Habra MA, Stephen B, Campbell M, et al. Phase II clinical trial of pembrolizumab efficacy and safety in advanced adrenocortical car- cinoma. J Immunother Cancer. 2019;7(1):253. https://doi.org/10. 1186/s40425-019-0722-x
18. Klein O, Senko C, Carlino MS, et al. Combination immunotherapy with ipilimumab and nivolumab in patients with advanced adreno- cortical carcinoma: a subgroup analysis of CA209-538. Oncoimmunology. 2021;10(1):1908771. https://doi.org/10.1080/ 2162402X.2021.1908771
19. Landwehr LS, Altieri B, Schreiner J, et al. Interplay between gluco- corticoids and tumor-infiltrating lymphocytes on the prognosis of adrenocortical carcinoma. J Immunother Cancer. 2020;8(1): e000469. https://doi.org/10.1136/jitc-2019-000469
20. Diaz L, Marabelle A, Kim TW, et al. Efficacy of pembrolizumab in phase 2 KEYNOTE-164 and KEYNOTE-158 studies of microsatel- lite instability high cancers. Ann Oncol. 2017;28(5):v128-v129. https://doi.org/10.1093/annonc/mdx367.020
21. Boyiadzis MM, Kirkwood JM, Marshall JL, Pritchard CC, Azad NS, Gulley JL. Significance and implications of FDA approval of pembrolizumab for biomarker-defined disease. J Immunother Cancer. 2018;6(1):35. https://doi.org/10.1186/s40425-018-0342-x
22. Davis AA, Patel VG. The role of PD-L1 expression as a predictive biomarker: an analysis of all US Food and Drug Administration (FDA) approvals of immune checkpoint inhibitors. J Immunother Cancer. 2019;7(1):278. https://doi.org/10.1186/s40425-019-0768-9
23. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evalu- ation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer. 2009;45(2):228-247. https://doi.org/10.1016/j. ejca.2008.10.026
24. Hyo JP, Kyung WK, Junhee P, et al. Incidence of pseudoprogression during immune checkpoint inhibitor therapy for solid tumors: a sys- tematic review and meta-analysis. Radiology. 2020;297(1):87-96. https://doi.org/10.1148/radiol.2020200443
25. Hahner S, Hartrampf PE, Mihatsch PW, et al. Targeting 11-beta hy- droxylase with [131I]IMAZA: a novel approach for the treatment of advanced adrenocortical carcinoma. J Clin Endocrinol Metab. 2022;107(4):e1348-e1355. https://doi.org/10.1210/clinem/dgab895
26. Maher VE, Fernandes LL, Weinstock C, et al. Analysis of the asso- ciation between adverse events and outcome in patients receiving a programmed death protein 1 or programmed death ligand 1 anti- body. J Clin Oncol. 2019;37(30):2730-2737. https://doi.org/10. 1200/JCO.19.00318
27. Eggermont AMM, Kicinski M, Blank CU, et al. Association be- tween immune-related adverse events and recurrence-free survival among patients with stage III melanoma randomized to receive pembrolizumab or placebo: a secondary analysis of a randomized clinical trial. JAMA Oncol. 2020;6(4):519-527. https://doi.org/10. 1001/jamaoncol.2019.5570
28. Greenstein AE, Habra MA, Wadekar SA, Grauer A. Adrenal tumors provide insight into the role of cortisol in NK cell activity. Endocr Relat Cancer. 2021;28(8):583-592. https://doi.org/10.1530/ERC- 21-0048
29. Berruti A, Sperone P, Ferrero A, et al. Phase II study of weekly pa- clitaxel and sorafenib as second/third-line therapy in patients with adrenocortical carcinoma. Eur J Endocrinol. 2012;166(3): 451-458. https://doi.org/10.1530/EJE-11-0918
30. Kroiss M, Quinkler M, Johanssen S, et al. Sunitinib in refractory adrenocortical carcinoma: a phase II, single-arm, open-label trial. J Clin Endocrinol Metab. 2012;97(10):3495-3503. https://doi.org/ 10.1210/jc.2012-1419
31. Bedrose S, Miller KC, Altameemi L, et al. Combined lenvatinib and pembrolizumab as salvage therapy in advanced adrenal cortical car- cinoma. J Immunother Cancer. 2020;8(2):e001009. https://doi.org/ 10.1136/jitc-2020-001009