ENDOCRINE SOCIETY
OXFORD
Expression and Prognostic Relevance of PD-1, PD-L1, and CTLA-4 Immune Checkpoints in Adrenocortical Carcinoma
Laura-Sophie Landwehr,10 Barbara Altieri,10 luliu Sbiera,1 Hanna Remde,10 Stefan Kircher,2 Julie Olabe, 30 Silviu Sbiera, 1,4[D Matthias Kroiss, 1,4,5[D and Martin Fassnacht1,4,6 [D
1Department of Internal Medicine I, Division of Endocrinology and Diabetes, University Hospital Würzburg, 97080 Würzburg, Germany 2Institute of Pathology, University of Würzburg, 97080 Würzburg, Germany
3Institute GRED (Genetics, Reproduction and Development), University Clermont Auvergne, 63001 Clermont-Ferrand, France 4Comprehensive Cancer Center Mainfranken, University of Würzburg, 97080 Würzburg, Germany
5Department of Medicine IV, LMU University Hospital, LMU Munich, 80336 München, Germany
6Clinical Chemistry and Laboratory Medicine, University Hospital Würzburg, 97080 Würzburg, Germany
Correspondence: Martin Fassnacht, MD, University Hospital Würzburg, Department of Internal Medicine I, Division of Endocrinology and Diabetes, Oberdürrbacher Str. 6, 97080 Würzburg, Germany. Email: Fassnacht_M@ukw.de.
Abstract
Context: Adrenocortical carcinoma (ACC) is a rare endocrine malignancy with poor prognosis in advanced stages. While therapies targeting the checkpoint molecules programmed cell death 1 (PD-1), its ligand PD-L1, and the cytotoxic T lymphocyte-associated protein 4 (CTLA-4) have revolutionized treatment in many cancers, the results in ACCs were heterogeneous.
Objective: Their expression in ACC has not been systematically studied and might explain the variable response to immune checkpoint inhibitors. Methods: The expression of PD-1, PD-L1 and CTLA-4 was examined in 162 tumor samples from 122 patients with ACC by immunohistochemistry (threshold of >1%) and correlated with tumoral T lymphocyte infiltration and clinical endpoints. Finally, univariate and multivariate analyses of progression-free and overall survival were performed.
Results: PD-1 and PD-L1 were expressed in 26.5% and 24.7% of samples, respectively, with low expression in most tumor samples (median positive cells: 2.1% and 21.7%). In contrast, CTLA-4 expression was observed in 52.5% of ACC with a median of 38.4% positive cells. Positive PD-1 expression was associated with longer progression-free survival (HR 0.50, 95% CI 0.25-0.98, P =. 04) even after considering prognostic factors. In contrast, PD-L1 and CTLA-4 did not correlate with clinical outcome. Additionally, PD-1 and PD-L1 expression correlated significantly with the amount of CD3+, CD4+, FoxP3+, and CD8+ T cells.
Conclusion: The heterogeneous expression of PD1, PD-L1, and CTLA-4 in this large series of well-annotated ACC samples might explain the heterogeneous results of the immunotherapies in advanced ACC. In addition, PD-1 expression is a strong prognostic biomarker that can easily be applied in routine clinical care and histopathological assessment.
Key Words: adrenocortical carcinoma, immune checkpoints, PD-1, PD-L1, CTLA-4, immunotherapy, glucocorticoids, prognostic marker
Abbreviations: ACC, adrenocortical carcinoma; CTLA-4, cytotoxic T lymphocyte-associated protein 4; ENSAT, European Network for the Study of Adrenal Tumors; GC, glucocorticoid; OS, overall survival; PD-1, programmed cell death 1; PD-L1, programmed cell death ligand 1; PFS, progression-free survival; RFS, recurrence-free survival; TIL, tumor-infiltrating lymphocyte.
Adrenocortical carcinoma (ACC) is a rare and often aggres- sive endocrine malignancy with an incidence of 0.5 to 2 cases per million per year and a dismal prognosis (1). In adults, about 50% of patients with ACC present with tumor-related hypercortisolism, 25% of tumors secrete sex hormones, while others are endocrine inactive (2, 3). Even after complete surgi- cal resection, which represents the only curative approach for ACC, patients remain at high risk for local recurrences and metastatic disease. For advanced ACC, standard treatment is a cytotoxic chemotherapy including etoposide, doxorubi- cin, and cisplatin plus mitotane (4, 5). However, the objective response rate is less than 25% and median overall survival (OS) is only 15 months (4); hence, new treatment options for advanced ACC are urgently required.
Cancer immunotherapies targeting the immune checkpoint molecules, programmed cell death 1 (PD-1), its ligand PD-L1, and the cytotoxic T lymphocytes-associated protein 4
(CTLA-4) have led to impressive results in many tumor entities (6). Physiologically, these immune checkpoint regulators are crucial for self-tolerance; however, this crucial mechanism is harnessed by tumors that overexpress these molecules to evade immune surveillance (7). Therefore, anti-tumoral immune checkpoint therapies have been developed that act by neutral- izing PD-1, its ligand PD-L1, and CTLA-4, while reversely en- hancing T cell cytotoxicity toward tumors (8). In turn, the tumoral expression of immune checkpoint molecules was rap- idly recognized as a potential prognostic and/or predictive bio- marker of response to cancer immunotherapy but also as general markers for patient survival.
While successes have been achieved in other tumor entities, the first results of immunotherapies in ACCs were disappoint- ing. Within 5 phase 1 and phase 2 clinical trials, 121 patients with ACC have been treated with different immune check- point inhibitors but an objective response was seen in only
17 patients (9-13). However, in 16 patients with ACC treated with pembrolizumab targeting PD-1, Raj et al observed a clin- ically relevant disease control rate of 52% and a median OS of almost 25 months, suggesting that a subgroup of patients with ACC may experience a benefit from an immunotherapeutic approach (12). In contrast, recent real-world data from Germany were less promising with only 7 partial responses in 52 evaluated patients (14). Accordingly, the identification of potentially relevant predictors of response to immune checkpoint inhibitors in ACC would be an important step to improve treatment options for patients with this rare disease. Results from several other (but not all) solid cancers indi- cate that the expression levels of the different immune check- point molecules might serve as such predictive markers (15-17). However, data on the expression of these molecules in ACC are very limited. Overall, 7 studies analyzing 187 pa- tients with ACC investigated the expression of PD-L1 and the
percentage of PD-L1 positive ACC ranged from 11.1 to 78.6% (9-13, 18, 19). Furthermore, PD-1 was heterogeneous- ly expressed only in 10 patients (60.0%) (10), while CTLA-4 was not evaluated in ACC so far.
Here, we report on the expression of immune checkpoint molecules in a large cohort of 162 samples of patients with clinically well-annotated ACC and correlate these results with clinical characteristics and outcome.
Materials and Methods
Patient Cohort
The study followed the principles of the declaration of Helsinki and good clinical practice guidelines and was ap- proved by the ethics committee of the University of Würzburg (#88/11) as part of the European Network for the Study of Adrenal Tumors (ENSAT) ACC registry (20). All pa- tients provided written informed consent. Anonymous non- adrenal tissue samples served as controls.
In total, 162 human tumor samples, including 107 primary tumors, 17 local recurrences, and 38 metastases, of 122 pa- tients with ACC in high accordance with a cohort of which 113 were reported in a previous study (21) were included in this study. Metastases were localized in the abdomen (n =6), bone (n =2), liver (n = 3), lung (n = 5), and lymph no- des (n =22). Endocrine workup confirmed an autonomous hypercortisolism diagnosed by means of the pathological 1 mg dexamethasone test (cortisol > 5 µg/dL) and in the pres- ence of suppressed adrenocorticotropin (2, 22) in 43% of pa- tients with ACC (Table 1). In general, patients with ACC were treated according to international guidelines (5). Accordingly, most of the primary tumors were from patients who had not previously received systemic therapy at the time of surgery, whereas the majority of distant metastases samples had seen mitotane and/or cytotoxic therapies (details see Table S1 (23)). None of the patients have been treated with any immune therapy prior surgery.
Immunohistochemical Staining
Chromogenic immunohistochemical staining was performed on full formalin-fixed, paraffin-embedded tumor sections. Tissue specimens were deparaffinized in xylene and sequen- tially rehydrated in 100%, 90%, 80% and 70% ethanol, rinsed in distilled water and washed twice in Dulbecco’s phosphate-buffered saline. Antigen retrieval was performed
| n | Age (year) | Sex (m/f) (n) | Size of tumor (cm) | Ki67 index (%) | GC + other steroid | Sex hormone | Other steroid pattern | No steroid excess | Hormone analysis n/a | |
|---|---|---|---|---|---|---|---|---|---|---|
| All ACC samples | 162 | 46 (18-77) | 55/107 | 9.0 (0.4-30.0) | 18 (0-80) | 80 | 12 | 8 | 28 | 34 |
| Primary tumor | 107 | 46 (18-77) | 36/71 | 12.0 (3.3-30.0) | 17 (0-80) | 49 | 9 | 4 | 21 | 24 |
| ENSAT I/IIª | 43 | 48 (18-77) | 13/30 | 11.0 (3.3-28.0) | 12 (0-40) | 16 | 6 | 2 | 11 | 8 |
| ENSAT IIIª | 40 | 42 (18-75) | 15/25 | 11.7 (5.0-30.0) | 22 (1-80) | 19 | 3 | 2 | 7 | 9 |
| ENSAT IVª | 24 | 46 (24-72) | 8/16 | 14.0 (7.0-25.0) | 16 (1-30) | 14 | 0 | 0 | 3 | 7 |
| Local recurrenceb | 17 | 44 (20-70) | 6/11 | 5.0 (1.5-9.4) | 13 (5-20) | 8 | 1 | 1 | 3 | 4 |
| Lymph node metastasisb | 19 | 47 (18-66) | 8/11 | 1.9 (0.4-4.5) | 40 (30-50) | 12 | 1 | 1 | 0 | 5 |
| Distant metastasisb | 19 | 45 (19-72) | 5/14 | 1.9 (0.5-4.0) | 21 (5-40) | 11 | 1 | 2 | 4 | 1 |
Data represent total number and median values with ranges.
Abbreviations: ENSAT, European Network for the Study of Adrenal Tumors; n/a, not available.
“Tumor stage at the time of diagnosis according to the ENSAT classification (20).
In patients who experienced local recurrence or metastases during mitotane treatment, endocrine activity was classified according to the information available at primary diagnosis.
Tumor size was not applicable in 10 cases, while proliferation status was not determined in 54 and hormone data in 34 cases.
in 10 mM citric acid monohydrate buffer (pH 6.5) under pres- sure for 13 minutes.
Afterwards, all incubations were performed under humidi- fied conditions without exposure to light. Endogenous perox- idase activity was blocked by incubating tissue sections with methanol containing 10% hydrogen peroxide at room tem- perature (RT) for 10 minutes. Unspecific binding sites were blocked using 20% human AB serum at RT for 1 hour, fol- lowed by incubation with the primary antibodies (PD-1: EH33, Cell Signaling, RRID:AB_2728836, 1:300; PD-L1: E1L3N, Cell Signaling, RRID:AB_2687655, 1:400; CTLA-4: CLA49, Abcam, RRID:AB_2905652, 1:500) over- night at 4 ℃. Universal isotype anti-mouse and anti-rabbit controls on human tonsils served as negative controls. After washing with Dulbecco’s phosphate-buffered saline, signal amplification was visualized by HiDef Detection HRP Polymer System for 20 minutes at RT. Antigen targeting was completed with 3,3’diaminobenzidine substrate-chromogen (DAB substrate kit; Cell Marque) for 10 minutes at RT and by horseradish peroxidase enzymatically visualizing colored precipitate. Nuclei counterstaining was performed using Mayer’s hematoxylin. Stained slides were dehydrated in 100% ethanol and dried at 56 ℃. Finally, tissue slides were preserved using ProLong Gold Antifade (Thermo Fisher, RRID:SCR_015961) and coverslips and stored at RT until microscopic assessment.
Fluorescence staining for tumor-infiltrating T lymphocytes were performed as described in Landwehr et al (21).
Quantitative Microscopy Analysis
For the evaluation of immune checkpoint molecule expres- sion, the full section of each tumor was microscopically scanned using a Leica Aperio Versa system (RRID: SCR_021016) with 20x magnification objective. Two inde- pendent investigators quantified the membranous expression of PD-1, its ligand PD-L1 and CTLA-4 by using Aperio eSlide Manager (Leica Biosystems, RRID:SCR_021016) and a final score was generated by consensus. Spearman correl- ation for interobserver reliability was >0.85.
The expression of PD-1 was analyzed as ratio of PD-1+ T cells to total number of cells with a threshold for positivity of >1%. For the quantification of PD-L1 and CTLA-4, the combined positive score, which defined the ratio of PD-L1+ tu- mor cells and immune cells to total number of viable tumor cells, was used with a threshold of >1%.
Endpoints and Statistics
OS describes the length of time from primary diagnosis (sur- gery or first diagnosis) to last follow-up or death. The period between primary surgery and the first occurrence of any re- lapse was defined as recurrence-free survival (RFS). For this endpoint, only patients with complete surgical removal of the tumor (R0 resection) were included. At the time of diagno- sis and during follow up examinations at an interval of 3 to 6 months, the presence of local recurrence or metastasis was assessed in clinical routine at least by computerized tomog- raphy of chest and abdomen. Progression-free survival (PFS) was defined as time interval between the surgery of the given tumoral lesion and the next documented progression or death, whatever occurred first. Patients who were still alive and with- out progression were censored at the time of last follow-up.
To estimate and compare event-free survival, the Kaplan- Meier method and log-rank test was performed. Cox propor- tional hazard regression modelling with evaluation of hazard ratio (HR) and 95% confidence interval (CI) was applied for the identification of factors that independently impacted pa- tients’ outcome. In a first step, a univariate analysis was per- formed for all known or potentially relevant prognostic factors (age, sex, ENSAT tumor stage, resection status, Ki67 proliferation index, and autonomous glucocorticoid [GC] se- cretion). Additionally, all three immune checkpoint markers (PD-1, PD-L1, and CTLA-4) were analyzed in the same man- ner. For the multivariate analysis, only parameters with a sig- nificant impact on patients’ prognosis and survival in the univariate analyses were included.
For differences in immune checkpoint molecules expression considering tumor localization, ENSAT tumor stage, resec- tion status, tumor size, Ki67 proliferation index, and GC ex- cess, the unpaired nonparametric Mann-Whitney test or Kruskal-Wallis test were performed, as appropriate.
The correlations between categorical parameters were as- sessed by using the Pearson’s Chi-squared (X2) test, while im- mune checkpoint expression (%) and the amount of tumor-infiltrating T lymphocytes (#) were assessed using Spearman test (R) with Phi (+) coefficient as a measure for ef- fect size.
Statistical analyses were performed using Prism (Version 9.5.1, GraphPad Software Inc., La Jolla, CA, USA, RRID: SCR_002798) and SPSS (Version 29.0, IBM Corporation, Armonk, New York, USA, RRID:SCR_002865).
All statistically generated data were specified as median and range. A threshold of P < . 05 was considered to be statistically significant.
Results
Immune Checkpoint Molecules in Adrenocortical Carcinoma
The expression of immune checkpoint molecules PD-1, PD-L1, and CTLA-4 was analyzed in 162 tumor samples of 122 patients with ACC. The cohort was representative for ACC (regarding sex, age, tumor size, and Ki67 index) (1) with exception of a slightly lower proportion of tumors of pa- tients with ENSAT stage IV as these patients underwent sur- gery less frequently (Table 1). Due to the low number of patients that have been pretreated prior surgical resection, we could not reliably analyze the influence of any pretreat- ment on the expression of PD-1, PD-L1, and CTLA-4.
Programmed Cell Death 1
The immune checkpoint molecule PD-1 was detectable on T lymphocytes (and also on few tumor cells) in 26.5% of all 162 ACC tumors (Fig. 1 and Table 2). There were no signifi- cant differences regarding primary tumors, which were infil- trated by PD-1+ T lymphocytes in 29.0%, while 21.4% of local recurrences, 30.8% of lymph node metastases and 18.8% of distant metastases were positive for PD-1 (Table 2). Moreover, the expression of PD-1 was independent of tumor size (PD+ 9.7 cm vs PD- 9.0 cm; R = 0.01, P = . 97) and resection status (R0/RX vs R1/R2, R = 0.24, P = .13) and did not correlate with the Ki67 proliferation index (0-19% vs >20%, R = 0.25, P =. 13). However, higher infil- tration of PD-1+ tumor-infiltrating lymphocytes (TILs) was
PD-1
PD-L1
CTLA-4
Primary Tumor (< 1%)
A
B
C
D
E
F
Primary Tumor
(> 1%)
G
H
I
Lymph node metastasis
associated with lower ENSAT tumor stage I/II, while patients with more advanced tumor stage III or IV presented with less PD-1 expression (46.5% vs 15.0% and 27.8%, respectively, X2 = 6.28, P = . 04; Table 2). Furthermore, glucocorticoid ex- cess appeared to be associated with the expression of PD-1 as ACC without hypercortisolism presented with higher rates of PD-1+ TILs (GC+ vs GC-, X2 = 5.36, P =. 02).
Programmed Cell Death Ligand 1
The expression of the corresponding immune checkpoint lig- and PD-L1 was identified on antigen-presenting cells and tu- mor cells in 24.7% of ACC specimens (PD-L1+ 40 vs PD-L1- 116; Fig. 1 and Table 2). In detail, 24.3% of the pri- mary tumors (PD-L1+ 26 vs PD-L1-74), 17.7% of local recur- rences (PD-L1+ 3 vs PD-L1- 14), 31.6% of the lymph node metastases (PD-L1+ 3 vs PD-L1- 13), and 26.3% of distant metastases (PD-L1+ 5 vs PD-L1- 14) had PD-L1 expression above the threshold of >1% (Fig. 1, Table 2).
PD-L1 expression was independent of tumor size (PD-L+ 9.7 cm vs PD-L- 8.8 cm; R = 0.30, P =. 08) and resection sta- tus (R0/RX vs R1/R2, X2 = 0.18, P = .30). In addition, a Ki67 proliferation index of >20% was significantly associated with higher expression of PD-L1, while ACC tumors with a lower Ki67 proliferation index had lower PD-L1 positivity (0-19%; X2 = 0.94, P <. 0001). In contrast to PD-1+ TILs, PD-L1 ex- pression was not associated with hypercortisolism (X2 = 0.56, P =. 45; Table 1).
A significant correlation between the expression of PD-1 and its ligand PD-L1 was observed in primary tumors as well as local recurrences and metastases (X2 = 36.24, P < 0.001).
Cytotoxic T Lymphocytes Antigen 4
The immune checkpoint marker CTLA-4 was expressed on T lymphocytes (and also on few tumor cells) in 52.5% of ACC specimens (CTLA-4+ 85 vs CTLA-4- 63; Fig. 1 and Table 2). In detail, 53.3% of the primary tumors (CTLA-4+ 57 vs CTLA-4- 40), 35.3% of local recurrences (CTLA-4+ 6 vs CTLA-4- 10), 73.7% of the lymph node metastases (CTLA-4+ 14 vs CTLA-4- 5) and 42.1% of the distant metas- tases (CTLA-4+ 8 vs CTLA-4- 8) presented CTLA-4 expression (Table 2). Its expression was not correlated with tumor size (CTLA-4+ 9.2 cm vs CTLA-4- 9.2 cm, R =0.22, P =. 10), resection status (RO/RX vs R1/R2, X2=0.16, P =. 69), or Ki67 proliferation index (0-19%; X2 = 0.01, P =. 91). Again, CTLA-4 expression is not affected by hormone secretion status in ACC (X2 = 2.29, P =. 13).
The expression of PD-1, PD-L1, and CTLA-4 in primary tu- mors, local recurrences, lymph node metastases, and distant metastases was analyzed (Fig. 2). Interestingly, PD-1+ tumor- infiltrating cells were significantly more frequent in lymph node metastases (mean 5.0% ± 3.2) than in local recurrences (mean 1.7% ±2.1, P =. 003), while there was no significant difference to primary tumors and distant metastases (Fig. 2A). For PD-L1, a significantly higher presence was de- termined in primary tumors (mean 14.3% ± 22.0) and ACC metastases within distant organs (mean 15.0% ± 10.8) com- pared with local recurrences and metastases in lymph nodes (mean 3.8% ±7.8, P =. 005 and mean 7.17% ± 5.52, P = . 02, respectively, Fig. 2B). In contrast, CTLA-4 was highly present in primary tumors (mean 42.2% ±29.3) and distant metastases (mean 57.5% ± 26.6) compared with local recur- rences (P>.001) and lymph node metastases (P>.001 and
| PD-1+ | PD-1 | PD-1+ ACC (%) | PD-L1+ | PD-L1- | PD-L1+ACC (%) | CTLA-4+ | CTLA-4 | CTLA-4+ ACC (%) | |
|---|---|---|---|---|---|---|---|---|---|
| All ACC samples | 43 | 114 | 26.5 | 40 | 116 | 24.7 | 85 | 63 | 52.5 |
| Primary tumor | 31 | 71 | 29.0 | 26 | 75 | 24.3 | 57 | 40 | 53.3 |
| ENSAT I/IIª | 20 | 22 | 46.5 | 12 | 30 | 27.9 | 27 | 13 | 62.8 |
| ENSAT IIIª | 6 | 31 | 15.0 | 7 | 29 | 17.5 | 17 | 19 | 42.5 |
| ENSAT IVª | 5 | 18 | 27.8 | 7 | 16 | 29.2 | 13 | 8 | 54.2 |
| Local recurrenceb | 3 | 14 | 21.4 | 3 | 14 | 17.7 | 6 | 10 | 35.3 |
| Lymph node metastasis® | 6 | 13 | 30.7 | 6 | 13 | 31.6 | 14 | 5 | 73.7 |
| Distant metastasis6 | 3 | 16 | 18.8 | 5 | 14 | 26.3 | 8 | 8 | 42.1 |
Data represent total numbers or percentage (%). n/a: PD-1 (5), PDL-1 (6), and CTLA-4 (14).
Abbreviations: ACC, adrenocortical carcinoma; CTLA-4, cytotoxic T lymphocyte-associated protein 4; ENSAT, European Network for the Study of Adrenal Tumors; n/a, not available; PD-1, programmed cell death 1; PD-1L, PD-1 ligand.
“Tumor stage at the time of diagnosis according to the ENSAT classification (20).
“In patients who experienced local recurrence or distant metastases during mitotane treatment.
P =. 005), while its expression was diminished in local recur- rences (mean 2.9% ±4.3) and lymph node metastases (mean 3.8% ± 4.7, Fig. 2C).
Association of Immune Checkpoint Molecules With Immune Cells and the Impact of Glucocorticoids
Since the immune checkpoint molecule PD-1 was primarily expressed on tumor-infiltrating T lymphocytes (Fig. 1), their association was analyzed in the total cohort of 162 ACCs. Here, a significant correlation between the PD-1 expression (%) and the proportion of CD3+ T cell tumor infiltration was detected (R = 0.52, Ø = 2.61, P = 0.04). In detail, T cell subtypes including CD4+ TH and CD8+ cytotoxic T cells significantly correlated with PD-1 expression (R =. 57, Ø= 2.60, P <. 001 and R = 0.55, Ø = 2.56, P = . 001, respect- ively; Fig. S1 (23)).
Similarly, immune checkpoint molecule PD-L1 correlated with the total number of TILs in ACC, according to CD3+, CD4+, FoxP3+, and CD8+ TILs (R =0.41, Ø=3.666, P =. 002, R = 0.44, Ø = 3.38, P =. 05, R = 0.48, Ø = 2.29, P =. 02, and R = 0.52, Ø = 3.65, P =. 001, respectively).
In contrast, CTLA-4 expression (%) did not correlated sig- nificantly with the levels of CD3+ T cell infiltrates (R = 0.11, Ø = 4.12, P = . 91).
Association of Immune Checkpoint Molecule Expression With Overall, Recurrence-, and Progression-Free Survival
To evaluate the impact of PD-1, PD-L1 and CTLA-4 expres- sion on the clinical outcome of patients affected by ACC, OS was investigated in 96 patients with primary tumors. While the median OS in patients with ACC with a PD-1+ pri- mary tumor was not reached, PD-1” ACCs were characterized by a median OS of only 47.0 months (Fig. 3 and Table 3). Using univariate analyses, patients with ACC with PD-1+ infil- trating T lymphocytes had a better OS compared with ACC tumors lacking PD-1 expression (HR 0.33, 95% CI 0.16-0.68, P =. 003; Fig. 3 and Table 3), but this was not con- firmed in the multivariate analysis.
In contrast, its ligand PD-L1 and CTLA-4 expression did not influence OS of patients with ACC (HR 0.79, 95% CI
0.40-1.57, P =. 49 and HR 1.47, 95% CI 0.81-2.67, P = . 20, respectively, Fig. 3 and Table 3).
Next, in patients with completely resected, non-metastatic ACC (n = 63), we found a significant association of PD-1 ex- pression and a longer median RFS (HR 0.49, 95% CI 0.25-0.95, P = . 04; Table S2 (23)). However, there was no as- sociation of the presence of the immune checkpoint molecules PD-L1 and CTLA-4 regarding RFS in this subcohort of pa- tients (HR 1.21, 95% CI 0.63-2.29, P =. 57 and HR 1.53, 95% CI 0.83-2.82, P =. 17, respectively, Table S2 (23)).
Lastly, a significantly better PFS of 25 months in patients with ACC with PD-1 expression in primary tumors, local recurrences, and distant metastases was found compared with patients with tumors lacking PD-1 expression (34 vs 9 months, respectively, P =. 01; Fig. 3 and Table 4). This was also valid in a multivariate analysis, where the presence of immune checkpoint molecule PD-1 correlated significantly with longer PFS (HR 0.54, 95% CI 1.02-3.94, P = . 04), inde- pendently of other significant prognostic markers, including ENSAT tumor stage, resection status, Ki67 proliferation in- dex, and GC excess (Fig. 3G and Table 4). In contrast, PD-L1 and CTLA-4 had both no impact on PFS of patients with ACC (Table 4).
Discussion
Immune checkpoint inhibitors are an immense advancement in the treatment of many tumor entities, and in some tumors their efficacy even revolutionized the entire treatment concept (24).
Here, we demonstrate for the first time in the largest cohort of ACCs that the expression of PD-1 but not PD-L1 and CTLA-4 has prognostic impact, and can serve as a marker of tumoral T cell infiltration and inversely correlates with clin- ical cortisol excess associated with some ACC tumors.
In our study on ACC, the presence of PD-1-expressing TILs was associated with a significantly better OS and PFS com- pared with PD-1- ACC, the latter even independently of other established prognostic markers. Hence, patients with ACC with PD-1+-infiltrated primary tumors are characterized by a 3-fold lower risk of mortality. When local recurrences and metastases were included, the risk for progression was 2-fold lower in PD-1-expressing tumors. This is in contrast
A
B
C
✱
100
100
100
PD-1 Expression (%)
75
50
PD-L1 expression (%)
50
CTLA-4 Expression (%)
10
75
25
50
5
25
0
0
0
Primary Tumors
Local Recurrences
Lymph Node Metastases
Distant Metastases
Primary Tumors Local Recurrences
Lymph Node Metastases
Distant Metastases
Primary Tumors Local Recurrences
Lymph Node Metastases Distant Metastases
to several studies that focused on the immunosuppressive properties of PD-1, the best example being the negative correl- ation of PD-1 expression and number of TILs in breast cancer, where PD-1+ TILs were associated with a worse OS and a 2.7x higher risk for death (25).
However, other studies have highlighted the ambiguous role of PD-1 expression in T cell response. Although its immu- noregulatory character was commonly established, PD-1 is first and foremost a biomarker for T cell receptor strength and T cell activation (26).
These observations were confirmed in other tumor entities, for instance, in colorectal carcinoma, non-small-cell lung can- cer or breast cancer, in which higher expression of PD-1 sig- nificantly correlated with an improved clinical outcome, and serves as independent prognostic factor (27-29) similar to what we found in ACC.
In contrast to PD-1, the expression of its ligand PD-L1, while present in 24.69% of ACCs when considering a thresh- old of ≥1%, was not associated with clinical outcome and was not suitable as a prognostic biomarker. This is in line with the study by Fay et al, who demonstrated in 28 patients with ACC that 10.7% of the tumor cells stained positive for PD-L1 when applying a threshold of ≥5% positivity and were not related to 5-year survival by Kaplan-Meier analysis (18).
In clinical trials on immune checkpoint inhibitor monother- apy, PD-L1 expression varied from 0% to 75% when thresh- olds of ≥1% or ≥5% tumor proportional score were applied (9-12), but the largest of these studies included only 42 patients (9). In addition, based on mRNA analyses, the heterogeneous expression of PD-L1, which tends to be asso- ciated with younger age and a higher Ki67 proliferation in- dex, was confirmed in another retrospective study including
79 patients (30). Furthermore, these data were consistent with findings from our larger cohort.
Nevertheless, several studies considering the prognostic im- plication of PD-L1 in different tumor types indicated a hetero- geneous correlation of its expression and patients’ prognosis.
While PD-L1 expression did also not predict patients’ out- come in cervical carcinoma (31), other studies found evidence for its prognostic power. For instance, patients with PD-L1+ hepatocellular carcinoma had a significantly worse OS of only 29.6 months compared with 59.4 months for tumors with absent or low PD-L1 expression, and its expression was established as an independent prognostic marker in hep- atocellular carcinoma (32, 33). In contrast, in non-small-cell lung cancer Velcheti et al observed a significantly better out- come for patients in PD-L1-expressing tumors, with a mortal- ity risk reduction of 39% (34).
When considering CTLA-4, the absence of a prognostic ef- fect of CTLA-4 was demonstrated in different cancer entities (35, 36), similar to our study. This might be due to its broader immunoregulatory impact compared with the PD-1 axis that co-opted by tumor cells via ligand expression (37).
To summarize, since the presence of tumor-resident T cells is crucial for mounting a sufficient anti-tumoral immune re- sponse (38, 39) and modern immunotherapies target immune checkpoint molecules, we studied in a large ACC cohort the presence of CD3+, CD4+, FoxP3+, and CD8+ tumor- infiltrating T lymphocytes and expression of the PD-1, PD-L1, and CTLA-4 immune checkpoint molecules as pre- dictive biomarkers for the outcome of patients with ACC. Interestingly, in the present study, we could show not only the presence of CD3, CD4, FoxP3 and CD8 and PD- 1-expressing infiltrating T cells, but also their prognostic impact
PD-1+
PD-L1+
CTLA-4+
A
p=0.003
B
p=0.494
C
1.0
p=0.201
Overall Survival
0.8
PD-1+ (n=30)
0.6
PD-L1+ (n=24)
CTLA-4- (n=35)
0.4
0.2
PD-1- (n=64)
PD-L1- (n=69)
CTLA-4+ (n=53)
Progression Free Survival
Univariate
0.0
0
50
100
150
0
50
100
150
0
50
100
150
D
Time (Months)
E
Time (Months)
F
Time (Months)
1.0
p=0.009
p=0.709
p=0.499
0.8
0.6
0.4
PD-1+ (n=33)
PD-L1+ (n=31)
CTLA-4- (n=51)
0.2
0.0
PD-1- (n=92)
PD-L1- (n=93)
CTLA-4+ (n=64)
0
50
100
150
0
50
100
150
0
50
100
150
G
Time (Months)
Time (Months)
Time (Months)
1.0
Progression Free Survival
p=0.045
0.8
Multivariate
0.6
0.4
PD-1+ (n=33)
0.2
PD-1- (n=92)
0.0
0
50
100
150
Time (Months)
on patients’ survival. Very interestingly, Thommen et al de- fined via transcriptional analyses in human non-small-cell lung cancer, three distinct CD8+ cytotoxic T cell subtypes when considering PD-1 expression. The CD8+ PD-1+ T cells recruited CD4+ TH cells with highly proliferative abilities serv- ing as strong predictors for both clinical outcome and re- sponse to PD-1 inhibition when compared with CD8+ PD-1-flow TILs (40). The predictive impact of both, TILs and PD-1/PD-L1/CTLA-4, respectively, was certainly con- firmed in other tumor entities, such as melanoma (41-43) and non-small-cell lung cancer (40, 44).
Moreover, the presence of TILs expressing PD-1 might also be favorable to patient outcome as they are associated with therapeutically promising neoantigens in other tumoral en- tities. Hence, the presence of TILs expressing PD-1 enabled the identification of tumor-reactive mutant-specific CD8+ cytotoxic T cells that were used for targeting patient-specific neoantigen with high affinity T cell receptor (45-47). These T cells were able to eliminate autologous tumor cells and might be of therapeutic relevance for personalized therapies
using neoantigen-reactive T lymphocytes derived from periph- eral blood in ACC as already demonstrated in patients with melanoma (47).
To get a definitive answer to the question whether immune checkpoint molecules are reliable biomarkers predicting clin- ical benefit through immunotherapy, investigating a larger co- hort of patients with ACC treated with immune checkpoint inhibitors is necessary, because all small trials to date have not found any correlation (48), probably because of the low number of investigated patients.
In conclusion, it is obvious that current immunotherapies are efficient only in a subset of patients with ACC. This study provides several potential explanations for the heteroge- neous results of the immune checkpoint therapy in advanced ACC. In addition, we could show that presence of intratu- moral PD-1-expressing T cells is a new and possible strong prognostic biomarker that can easily be applied in histo- pathological assessment. The fact that it can also serve as a marker of T cell infiltration makes it even more attractive for clinical use. Thus, in the future, reporting the exact
| Univariate analyses | Multivariate analyses | |||||||
|---|---|---|---|---|---|---|---|---|
| n | Hazard ratio | 95% CI | P | Hazard ratio | 95% CI | P | ||
| Age | — | 0.995 | 0.98-1.01 | .59 | — | — | — | |
| Sex | Female | 64 | ||||||
| Male | 32 | 1.13 | 0.66-1.96 | .66 | — | — | — | |
| ENSAT stage | I/II | 42 | ||||||
| III | 34 | 2.37 | 1.26-4.46 | .01 | ||||
| IV | 20 | 4.75 | 2.36-9.57 | <. 001 | 2.40 | 1.30-4.45 | .01 | |
| Resection status | R0/RX | 54 | ||||||
| R1/R2 | 32 | 5.13 | 2.87-9.18 | <. 001 | 1.05 | 0.65-1.69 | .85 | |
| Ki67 index | 0-19% | 50 | ||||||
| >20% | 35 | 2.35 | 1.29-4.28 | .01 | 2.60 | 1.12-6.00 | .03 | |
| GC excess | — | 35 | ||||||
| + | 39 | 2.10 | 1.12-3.96 | .02 | 1.48 | 0.62-3.56 | .38 | |
| PD-1 (> 1%) | — | 64 | ||||||
| + | 30 | 0.33 | 0.16-0.68 | .003 | 0.84 | 0.34-2.11 | .71 | |
| PD-L1 (>1%) | — | 69 | ||||||
| + | 24 | 0.79 | 0.40-1.57 | .49 | — | — | — | |
| CTLA-4 (>1%) | — | 35 | ||||||
| + | 53 | 1.47 | 0.81-2.67 | .20 | — | — | — | |
For overall survival analysis, only samples from patients with ACC with primary tumors and applicable clinical data were included (n = 96).
| Univariate analyses | Multivariate analyses | |||||||
|---|---|---|---|---|---|---|---|---|
| n | Hazard ratio | 95% CI | P | Hazard ratio | 95% CI | P | ||
| Age | - | 0.99 | 0.98-1.00 | .12 | — | — | — | |
| Sex | Female | 86 | ||||||
| Male | 40 | 1.18 | 0.79-1.76 | .43 | — | — | — | |
| ENSAT stage | I/II | 61 | ||||||
| III | 41 | 1.45 | 0.94-2.25 | .09 | ||||
| IV | 24 | 2.97 | 1.78-4.98 | <. 001 | 1.47 | 0.93-2.34 | .10 | |
| Resection status | R0/RX | 76 | ||||||
| R1/R2 | 34 | 0.39 | 0.25-0.61 | <. 001 | 1.10 | 0.75-1.62 | .64 | |
| Ki67 index | 0-19% | 56 | ||||||
| > 20% | 40 | 2.45 | 1.52-3.94 | <. 001 | 2.38 | 1.30-4.33 | .01 | |
| GC excess | — | 46 | ||||||
| + | 55 | 1.63 | 1.05-2.52 | .03 | 0.95 | 0.52-1.73 | .87 | |
| PD-1 (>1%) | — | 92 | ||||||
| + | 33 | 0.54 | 1.17-2.98 | .01 | 0.50 | 0.25-0.98 | .04 | |
| PD-L1 (>1%) | — | 93 | ||||||
| + | 31 | 0.92 | 0.69-1.71 | .71 | — | — | — | |
| CTLA-4 (>1%) | — | 51 | ||||||
| + | 64 | 1.15 | 0.58-1.31 | .50 | — | — | — | |
For progression-free survival analysis, primary tumors, local recurrences, and metastases from patients with ACC with available clinical data were included. For multivariate analyses, complete available data from 67 ACC tumors were included. ENSAT stage I/II, RO/RX resection status, low proliferating Ki67 (0-19%), and lack of glucocorticoid excess (GC-) was classified as reference category in multivariate analyses.
In 22 cases, no data about resection status were available. Clinical data considering proliferation were not determined in 53 and considering glucocorticoid secretion state in 35 cases.
Abbreviations: ACC, adrenocortical carcinoma; CTLA-4, cytotoxic T lymphocyte-associated protein 4; ENSAT, European Network for the Study of Adrenal Tumors; GC, glucocorticoid; PD-1, programmed cell death 1; PD-L1, PD-1 ligand.
PD-1 expression in ACC samples might be helpful in select- ing patients for immunotherapy.
Funding
The study was funded by the German Research Foundation (Deutsche Forschungsgemeinschaft (DFG); project numbers: 314061271-CRC/TRR 205).
Author Contributions
M.K. and M.F. received the funding by the German Research Foundation. L .- S. L., S. S., M.K. and M.F. contributed to the study conception and design. Material preparation, data col- lection and analysis were performed by L .- S.L., I.S., B.A., H.R., S.K., and S.S. All authors read and approved the final manuscript.
Disclosures
Martin Fassnacht and Matthias Kroiss are investigators of the SPENCER clinical trial sponsored by Enterome Pharma. Other coauthors declare no potential conflicts of interest.
Data Availability
The authors confirm that the data supporting the findings of this study are available within the article, its supplementary materials (23) or available from the corresponding author, Martin Fassnacht, upon reasonable request.
Ethics Approval
The study followed the principles of the declaration of Helsinki and good clinical practice guidelines and was ap- proved by the ethics committee of the University of Würzburg (#88/11).
Consent to Participate
All patients as part of the European Network for the Study of Adrenal Tumors (ENSAT) ACC registry provided written in- formed consent.
References
1. Else T, Kim AC, Sabolch A, et al. Adrenocortical carcinoma. Endocr Rev. 2014;35(2):282-326.
2. Fassnacht M, Dekkers OM, Else T, et al. European society of endo- crinology clinical practice guidelines on the management of adreno- cortical carcinoma in adults, in collaboration with the European network for the study of adrenal tumors. Eur J Endocrinol. 2018;179(4):G1-G46.
3. Puglisi S, Calabrese A, Ferraù F, et al. New findings on presentation and outcome of patients with adrenocortical cancer: results from a national cohort study. J Clin Endocrinol Metab. 2023;108(10): 2517-2525.
4. Fassnacht M, Terzolo M, Allolio B, et al. Combination chemother- apy in advanced adrenocortical carcinoma. N Engl J Med. 2012;366(23):2189-2197.
5. 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.
6. Johnson DB, Nebhan CA, Moslehi JJ, Balko JM. Immune-checkpoint inhibitors: long-term implications of toxicity. Nat Rev Clin Oncol. 2022;19(4):254-267.
7. Han Y, Liu D, Li L. PD-1/PD-L1 pathway: current researches in cancer. Am J Cancer Res. 2020;10(3):727-742.
8. Iwai Y, Ishida M, Tanaka Y, Okazaki T, Honjo T, Minato N. Involvement of PD-L1 on tumor cells in the escape from host im- mune system and tumor immunotherapy by PD-L1 blockade. Proc Natl Acad Sci U S A. 2002;99(19):12293-12297.
9. 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.
10. Carneiro BA, Konda B, Costa RB, et al. Nivolumab in metastatic adrenocortical carcinoma: results of a phase II trial. J Clin Endocrinol Metab. 2019;104(12):6193-6200.
11. 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.
12. Raj N, Zheng Y, Kelly V, et al. PD-1 blockade in advanced adreno- cortical carcinoma. J Clin Oncol. 2019;38(1):71-80.
13. 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.
14. Remde H, Schmidt-Pennington L, Reuter M, et al. Outcome of im- munotherapy in adrenocortical carcinoma: a retrospective cohort study. Eur J Endocrinol. 2023;188(6):485-493.
15. Konishi J, Yamazaki K, Azuma M, Kinoshita I, Dosaka-Akita H, Nishimura M. B7-H1 expression on non-small cell lung cancer cells and its relationship with tumor-infiltrating lymphocytes and their PD-1 expression. Clin Cancer Res. 2004;10(15):5094-5100.
16. Obeid JM, Erdag G, Smolkin ME, et al. PD-L1, PD-L2 and PD-1 expression in metastatic melanoma: correlation with tumor- infiltrating immune cells and clinical outcome. Oncoimmunology. 2016;5(11):e1235107.
17. Gil Del Alcazar CR, Huh SJ, Ekram MB, et al. Immune escape in breast cancer during. Cancer Discov. 2017;7(10):1098-1115.
18. Fay AP, Signoretti S, Callea M, et al. Programmed death ligand-1 expression in adrenocortical carcinoma: an exploratory biomarker study. J Immunother Cancer. 2015;3(1):3.
19. Zhang Z, Li M, Wang J, et al. Expression and clinical significance of VISTA and PD-L1 in adrenocortical carcinoma. Endocr Relat Cancer. 2022;29(7):403-413.
20. Fassnacht M, Johanssen S, Quinkler M, et al. Limited prognostic value of the 2004 international union against cancer staging classi- fication for adrenocortical carcinoma: proposal for a revised TNM classification. Cancer. 2009;115(2):243-250.
21. 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.
22. Fassnacht M, Arlt W, Bancos I, et al. Management of adrenal inci- dentalomas: European society of endocrinology clinical practice guideline in collaboration with the European network for the study of adrenal tumors. Eur J Endocrinol. 2016;175(2):G1-G34.
23. Landwehr L-S, Altieri B, Sbiera I, et al. Supplemental Material: Expression and Prognostic Relevance of PD-1, PD-L1 and CTLA-4 Immune Checkpoints in Adrenocortical Carcinoma. OSF; 2024. 07 Februar 2024, OSF, DOI 10.17605/OSF.IO/ XH2WQ [Internet].
24. Waldman AD, Fritz JM, Lenardo MJ. A guide to cancer immuno- therapy: from T cell basic science to clinical practice. Nat Rev Immunol. 2020;20(11):651-668.
25. Muenst S, Soysal SD, Gao F, Obermann EC, Oertli D, Gillanders WE. The presence of programmed death 1 (PD-1)-positive tumor- infiltrating lymphocytes is associated with poor prognosis in human breast cancer. Breast Cancer Res Treat. 2013;139(3):667-676.
26. Simon S, Labarriere N. PD-1 expression on tumor-specific T cells: friend or foe for immunotherapy? Oncoimmunology. 2017;7(1): e1364828.
27. Li Y, Liang L, Dai W, et al. Prognostic impact of programed cell death-1 (PD-1) and PD-ligand 1 (PD-L1) expression in cancer cells and tumor infiltrating lymphocytes in colorectal cancer. Mol Cancer. 2016;15(1):55.
28. Kumagai S, Togashi Y, Kamada T, et al. The PD-1 expression bal- ance between effector and regulatory T cells predicts the clinical ef- ficacy of PD-1 blockade therapies. Nat Immunol. 2020;21(11): 1346-1358.
29. Matikas A, Zerdes I, Lövrot J, et al. PD-1 protein and gene expres- sion as prognostic factors in early breast cancer. ESMO Open. 2020;5(6):e001032.
30. Billon E, Finetti P, Bertucci A, et al. PDL1 expression is associated with longer postoperative, survival in adrenocortical carcinoma. Oncoimmunology. 2019;8(11):e1655362.
31. Karim R, Jordanova ES, Piersma SJ, et al. Tumor-expressed B7-H1 and B7-DC in relation to PD-1+ T-cell infiltration and survival of patients with cervical carcinoma. Clin Cancer Res. 2009;15(20): 6341-6347.
32. Nomi T, Sho M, Akahori T, et al. Clinical significance and thera- peutic potential of the programmed death-1 ligand/programmed death-1 pathway in human pancreatic cancer. Clin Cancer Res. 2007;13(7):2151-2157.
33. Gao Q, Wang XY, Qiu SJ, et al. Overexpression of PD-L1 signifi- cantly associates with tumor aggressiveness and postoperative re- currence in human hepatocellular carcinoma. Clin Cancer Res. 2009;15(3):971-979.
34. Velcheti V, Schalper KA, Carvajal DE, et al. Programmed death ligand-1 expression in non-small cell lung cancer. Lab Invest. 2014;94(1):107-116.
35. Paulsen EE, Kilvaer TK, Rakaee M, et al. CTLA-4 expression in the non-small cell lung cancer patient tumor microenvironment: diver- ging prognostic impact in primary tumors and lymph node metas- tases. Cancer Immunol Immunother. 2017;66(11):1449-1461.
36. Hu P, Liu Q, Deng G, Zhang J, Liang N, Xie J. The prognostic value of cytotoxic T-lymphocyte antigen 4 in cancers: a systematic review and meta-analysis. Sci Rep. 2017;7(1):42913.
37. Dong H, Strome SE, Salomao DR, et al. Tumor-associated B7-H1 promotes T-cell apoptosis: a potential mechanism of immune eva- sion. Nat Med. 2002;8(8):793-800.
38. Gide TN, Quek C, Menzies AM, et al. Distinct immune cell popu- lations define response to anti-PD-1 monotherapy and anti-PD-1/ anti-CTLA-4 combined therapy. Cancer Cell. 2019;35(2): 238-55.e6.
39. Badalamenti G, Fanale D, Incorvaia L, et al. Role of tumor- infiltrating lymphocytes in patients with solid tumors: can a drop dig a stone? Cell Immunol. 2019;343:103753.
40. Thommen DS, Koelzer VH, Herzig P, et al. A transcriptionally and functionally distinct PD-1. Nat Med. 2018;24(7):994-1004.
41. Tumeh PC, Harview CL, Yearley JH, et al. PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature. 2014;515(7528):568-571.
42. Uryvaev A, Passhak M, Hershkovits D, Sabo E, Bar-Sela G. The role of tumor-infiltrating lymphocytes (TILs) as a predictive biomarker of response to anti-PD1 therapy in patients with metastatic non- small cell lung cancer or metastatic melanoma. Med Oncol. 2018;35(3):25.
43. Daud AI, Loo K, Pauli ML, et al. Tumor immune profiling predicts response to anti-PD-1 therapy in human melanoma. J Clin Invest. 2016;126(9):3447-3452.
44. Haratani K, Hayashi H, Tanaka T, et al. Tumor immune micro- environment and nivolumab efficacy in EGFR mutation-positive non-small-cell lung cancer based on T790 M status after disease progression during EGFR-TKI treatment. Ann Oncol. 2017;28(7):1532-1539.
45. Inozume T, Hanada K, Wang QJ, et al. Selection of CD8+ PD-1+ lymphocytes in fresh human melanomas enriches for tumor- reactive T cells. J Immunother. 2010;33(9):956-964.
46. Gros A, Robbins PF, Yao X, et al. PD-1 identifies the patient- specific CD8+ tumor-reactive repertoire infiltrating human tumors. J Clin Invest. 2014;124(5):2246-2259.
47. Gros A, Parkhurst MR, Tran E, et al. Prospective identification of neoantigen-specific lymphocytes in the peripheral blood of melan- oma patients. Nat Med. 2016;22(4):433-438.
48. Altieri B, Ronchi CL, Kroiss M, Fassnacht M. Next-generation therapies for adrenocortical carcinoma. Best Pract Res Clin Endocrinol Metab. 2020; 34(3):101434.
Downloaded from https://academic.oup.com/jcem/article/109/9/2325/7615509 by National Library of Medicine user on 06 April 2026