Abstract
Central adrenal insufficiency (AI) due to isolated adrenocorticotropic hormone (ACTH) deficiency (IAD) has been recently associated with immune checkpoint inhibitor (ICI) therapy. Our aim was to analyze the prevalence, clinical characteristics, and therapeutic outcomes in cancer patients with IAD induced by ICI therapy. A retrospective and multicenter study was performed. From a total of 4447 cancer patients treated with ICI antibodies, 37 (0.8%) (23 men (62.2%), mean age 64.7 ± 8.3 years (range 46–79 years)) were diagnosed with IAD. The tumor most frequently related to IAD was lung cancer (n = 20, 54.1%), followed by melanoma (n = 8, 21.6%). The most common ICI antibody inhibitors reported were nivolumab (n = 18, 48.6%), pembrolizumab (n = 16, 43.2%), and ipilimumab (n = 8, 21.6%). About half of the patients (n = 19, 51.4%) had other immune-related adverse events, mainly endocrine adverse effects (n = 10, 27.0%). IAD was diagnosed at a median time of 7.0 months (IQR, 5–12) after starting immunotherapy. The main reported symptom at presentation was fatigue (97.3%), followed by anorexia (81.8%) and general malaise (81.1%). Mean follow-up time since IAD diagnosis was 15.2 ± 12.5 months (range 0.3–55 months). At last visit, all patients continued with hormonal deficiency of ACTH. Median overall survival since IAD diagnosis was 6.0 months. In conclusion, IAD is a rare but a well-established complication associated with ICI therapy in cancer patients. It develops around 7 months after starting the treatment, mainly anti-PD1 antibodies. Recovery of the corticotropic axis function should not be expected.
Introduction
Checkpoint inhibitor immunotherapy has been shown to be an effective therapy in a wide variety of malignant tumors, such as advanced melanoma, renal carcinoma, non-small cell lung cancer, head and neck cancer, urothelial carcinoma, and Hodgkin lymphoma (Ott et al. 2013, Ito et al. 2015). Among the main immune checkpoint inhibitor (ICI) monoclonal antibodies are anti-cytotoxic T lymphocyte antigen 4 antibodies (anti-CTLA-4), such as ipilimumab and tremelimumab; anti-programmed cell death 1 receptor (anti-PD-1) antibodies such as nivolumab and pembrolizumab; and the anti-programmed cell death ligand 1 (anti-PD-L1) antibodies such as atezolizumab and durvalumab (Topalian et al. 2012, Conforti et al. 2018, Brahmer et al. 2021).
ICI therapeutic efficacy has been fully demonstrated, however, given its mechanism of action, this therapy may be accompanied by adverse effects related to an activation of the immune system that can complicate the clinical course, therapeutic management, and prognosis in cancer patients (Johnson et al. 2020, Stelmachowska-Banaś & Czajka-Oraniec 2020). Among the immune-related adverse events (irAEs), autoimmune endocrine adverse effects, such as hypophysitis, thyroiditis, adrenalitis, insulitis, and parathyroiditis have been described (Iglesias et al. 2015, Cukier et al. 2017, Iglesias 2018, Tan et al. 2019, Grouthier et al. 2020, Anderson & Morganstein 2021, Wright et al. 2021, Yoon et al. 2021).
Secondary adrenal insufficiency (SAI) due to isolated adrenocorticotropic hormone (ACTH) deficiency (IAD) has also been recently associated with ICI immunotherapy (Iglesias et al. 2021). In fact, in the last 5 years, around 60 patients have been reported, mainly as isolated clinical cases (Okano et al. 2016, Fujimura et al. 2017, Kastrisiou et al. 2017, Takaya et al. 2017, Zeng et al. 2017, Ariyasu et al. 2018, Boudjemaa et al. 2018, Oristrell et al. 2018, Takebayashi et al. 2018, Martins Machado et al. 2019, Ohara et al. 2019, Okahata et al. 2019, Takeno et al. 2019, Tanaka et al. 2020) or small series (Cho et al. 2017, Kitajima et al. 2017, Otsubo et al. 2017).
Given the paucity of information on IAD induced by ICI therapy and its important therapeutic implications, as well as the increase in its use in these patients in the future, we believe of interest to learn in more detail about this endocrine irAE. Therefore, we aimed to retrospectively analyze the prevalence, clinical characteristics, diagnostic methodology, treatments, and therapeutic outcomes in patients with IAD induced by ICI therapy followed in different reference centers in our country.
Materials and methods
Study subjects
Patients diagnosed with IAD developed after starting ICI immunotherapy in cancer patients from different reference centers in Spain were retrospectively studied. For this aim, the collaboration of endocrinologists who worked together with oncologists in different hospitals in our country were requested to participate in the recruitment of patients. Those patients diagnosed with IAD were included regardless of the tumor type and ICI used, both in monotherapy and in combination therapy.
Inclusion criteria were the presence of all the following: (1) Histological diagnosis of a malignant tumor; (2) Treatment of the tumor with one or more ICI antibodies before IAD development; and (3) Development of IAD after initiation of ICI therapy.
Exclusion criteria were at least one of the following: (1) Diagnosis of other anterior pituitary hormone deficiencies in addition to ACTH; (2) SAI due to steroid therapy (concomitant glucocorticoid therapy or abrupt withdrawal of chronic steroidal therapy in the last 3 months before IAD diagnosis); and (3) Holo-cranial radiotherapy in the past 10 years.
Design
All patients were studied retrospectively. Clinical and analytical data were obtained from clinical records until the last follow-up visit. Census date of the study was December 31, 2020.
The present study is a retrospective analysis of the clinical data from the medical records of cancer patients treated with immunotherapy either on a healthcare basis or as part of a clinical trial. In both cases, the therapeutic guidelines were logically followed in accordance with the Declaration of Helsinki. In relation to our study, approval was requested by the ethical committees of the Hospital Universitario Puerta de Hierro Majadahonda (PIL-NIV-2020-01) and the Hospital Universitari de Bellvitge (EPANA040/20).
Study variables
The clinical, analytical, and imaging data were collected from the medical records of the patients. IAD diagnosis was defined as the presence of low basal serum cortisol (<5 µg/dL) between 08:00 and 09:00 h associated with low/normal basal plasma ACTH concentrations in the absence of concomitant steroid treatment or a history of steroid treatment that could have produced suppression of the HPA axis (Joseph et al. 2016). These results were confirmed in successive medical visits. On the other hand, the rest of the anterior pituitary function must have been normal. For this, the baseline hormonal values (thyrotropin; free thyroxine; follicle stimulating hormone; luteinizing hormone; testosterone (men), 17-beta-estradiol (women); insulin-like growth factor type I, and prolactin) were analyzed. Hormonal measurements were performed in each laboratory using standard radioimmuno assays, immunoradiometric assays, or enzyme-immunometric assay methods, with their respective reference ranges. The tumor response to immunotherapy was classified according to those criteria considered adequate for each type of cancer by the physician responsible for each of the patients. The recovery of the corticotropic axis function was considered when basal serum cortisol concentrations at 8:00 h were equal to or greater than 12 µg/dL after at least 24 h without hormone replacement therapy with glucocorticoids (Sbardella et al. 2017).
Statistical analysis
The quantitative data were expressed by mean ± s.d. for variables with normal distribution and as median and interquartile range (IQR) for non-normal variables. The normality of each variable was studied using the Kolmogorov–Smirnov test. The qualitative data were expressed by proportions or percentages. For the comparison of means between two groups, we used the Student’s t-test for variables of normal distribution and the Mann–Whitney U test for nonparametric data. For the comparison of means in groups with paired data, we used the Student’s t-test for variables that followed a normal distribution or the Wilcoxon signed-rank test for variables that did not follow a normal distribution. Analysis of variance (ANOVA) and Kruskal–Wallis test were used to determine the significance of multiple groups of subjects for normally and non-normally distributed data, respectively. For comparison of proportions, the chi-square test and Fisher’s exact test were used. Survival analysis was performed using Kaplan–Meier method. The presence of statistical significance was considered for values of P < 0.05.
Results
From a total of 4447 cancer patients treated with ICI antibodies, 37 (0.8%) patients (23 men (62.2%), mean age 64.7 ± 8.3 years (range 46–79 year)s) were diagnosed with IAD. Clinical features of the patients at the time of IAD diagnosis are shown in Tables 1 and 2. There were 18 (48.6%), 13 (35%), and four (10.8%) patients with hypertension, dyslipidemia, and type 2 diabetes mellitus, respectively. Nine (24.3%) patients showed a previous endocrine disease (multinodular goiter (n = 4), silent thyroiditis (n = 2), primary hypothyroidism (n = 1), hyperthyroidism (n = 1), and normocalcemic hyperparathyroidism (n = 1). Four (10.8%) patients had a history of non-endocrine immune diseases (lichen planus (n = 2), vitiligo (n = 1), and psoriasis (n = 1)).
Clinical features of IAD induced by ICI therapy in cancer patients.
| IAD | |
|---|---|
| Number of patients | 37 |
| Sex, male, n (%) | 23 (62.2) |
| Age (years), mean ± s.d. (range) | 64.7 ± 8.3 (46–79) |
| Personal history, n (%) | |
| Hypertension | 18 (48.6) |
| Dyslipidemia | 13 (35.0) |
| Type 2 diabetes | 4 (10.8) |
| Endocrine disease | 9 (24.3) |
| Multinodular goiter | 4 (10.8) |
| Silent thyroiditis | 2 (5.4) |
| Primary hypothyroidism | 1 (2.7) |
| Hyperthyroidism | 1 (2.7) |
| Normocalcemic hyperparathyroidism | 1 (2.7) |
| Non-endocrine immune disease | 4 (10.8) |
| Lichen planus | 2 (5.4) |
| Vitiligo | 1 (2.7) |
| Psoriasis | 1 (2.7) |
| Tumor type, n (%) | |
| Lung cancer | 20 (54.1) |
| Melanoma | 8 (21.6) |
| Head and neck | 3 (8.1) |
| Urothelial | 2 (5.4) |
| Kidney | 1 (2.7) |
| Colon | 1 (2.7) |
| Others | 2 (5.4) |
| ICI, n (%) | |
| Nivolumab | 18 (48.6) |
| Pembrolizumab | 16 (43.2) |
| Ipilimumab | 8 (21.6) |
| Durvalumab | 1 (2.7) |
| Atezolizumab | 1 (2.7) |
| Tremelimumab | 1 (2.7) |
| Spartalizumab | 1 (2.7) |
| Others | 3 (8.1) |
| ICI regimen, n (%) | |
| Monotherapy | 27 (73.0) |
| Combined therapy | 10 (27.0) |
| Time to develop IAD after starting ICI (months), median (IQR) | 7.0 (5.0–12.0) |
IAD, isolated ACTH deficiency; ICI, Immune checkpoint inhibitor; IQR, interquartile range.
Physical examination, analytical evaluation, and imaging study in cancer patients diagnosed with IAD induced by ICI therapy at the time of diagnosis.
| IAD | |
|---|---|
| Number of patients | 37 |
| Physical examination at IAD diagnosis, mean ± s.d. | |
| BMI (kg/m2) | 26.5 ± 4.1 |
| SBP (mmHg) | 112.1 ± 20.4 |
| DBP (mmHg) | 70.3 ± 14.5 |
| Analytical determinations | |
| Glucose (mg/dL), mean ± s.d. (range) | 96.4 ± 18.8 (64–135) |
| Hypoglycemia (<50 mg/dL) (%) | 0 (0) |
| Creatinine (mg/dL) | 0.85 (0.80–1.00) |
| Sodium (mmol/L), mean ± s.d. (range) | 139.0 (133.7–141.0) |
| Hyponatremia (Na<135 mmol/L), n (%) | 9 (24.3) |
| Potassium (mmol/L) | 4.1 ± 0.6 (2.2–5.4) |
| Hyperkalemia (K>5.0 mmol/L), n (%) | 1 (2.7) |
| Eosinophils (%) | 4.0 (2.0–7.0) |
| Eosinophilia, n (%) | 11 (30.6) |
| ACTH (pg/mL) | 4.0 (1.8–6.0) |
| Cortisol (µg/dL) | 0.6 (0.3–1.1) |
| ACTH (pg/mL) last visit | 4.0 (1.3–5.2) |
| Cortisol (µg/dL) last visit | 1.0 (0.4–1.9) |
| Abnormal pituitary MRI, n (%) | 3 (8.1) |
DBP, diastolic blood pressure; IAD, isolated ACTH deficiency; ICI, Immune checkpoint inhibitor; IQR, interquartile range; SBP, systolic blood pressure.
Tumor- and immunotherapy-related data
The tumor most frequently related to IAD was lung cancer (n = 20, 54.1%), followed by melanoma (n = 8, 21.6%), head and neck cancer (n = 3, 8.1%), urothelial cancer (n = 2, 5.4%), renal cell carcinoma (n = 1, 2.7%), colon cancer (n = 1, 2.7%), and others (n = 2, 5.4%).
Twenty-eight (75.7%) patients had received prior treatment as cancer therapy (surgery, n = 23, 62.2%; conventional chemotherapy, n = 15, 40.5%; radiotherapy, n = 8, 21.6%; and tyrosine kinase inhibitors, n = 2, 5.4%).
The most commonly ICI antibody reported was nivolumab (n = 18, 48.6%), followed by pembrolizumab (n = 16, 43.2%), ipilimumab (n = 8, 21.6%), durvalumab (n = 1, 2.7%), atezolizumab (n = 1, 2.7%), and tremelimumab (n = 1, 2.7%). Most patients were treated with monotherapy (n = 27, 73%) and the rest with combined treatment (n = 10, 27%), mainly a combination of nivolumab plus ipilimumab (n = 6).
About half of the patients (n = 19, 51.4%) had other irAEs associated with ICI therapy. As shown in Fig. 1, the most common were endocrine (n = 10, 27.0%) (thyroiditis (n = 6) and hypothyroidism (n = 4)), followed by digestive (n = 8, 21.6%) (colitis (n =4), diarrhea (n = 3), and pancreatitis (n = 1)) and cutaneous (n = 5, 13.5%) (skin rash (n = 2), pruritus (n = 1), photosensitivity (n = 1), and lichen planus (n = 1)) adverse effects.
Percentage distribution of adverse effects associated with isolated ACTH deficiency secondary to ICI therapy.
Citation: Endocrine-Related Cancer 28, 12; 10.1530/ERC-21-0228
Percentage distribution of adverse effects associated with isolated ACTH deficiency secondary to ICI therapy.
Citation: Endocrine-Related Cancer 28, 12; 10.1530/ERC-21-0228
Percentage distribution of adverse effects associated with isolated ACTH deficiency secondary to ICI therapy.
Citation: Endocrine-Related Cancer 28, 12; 10.1530/ERC-21-0228
IrAEs generally preceded the diagnosis of IAD. Endocrine irAEs preceded (n = 7; 6.7 ± 6.9 months, range 1–21 months), coincided (n = 1) or followed (n = 2; 14.0 ± 12.7 months, range 5–23 months) the IAD diagnosis. Digestive irAEs preceded (n = 4; 6.2 ± 4.1 months, range 1–10 months), coincided (n = 1) or followed (n = 3; 3.0 ± 2.6 months, range 1–6 months) IAD. Cutaneous irAEs preceded (n = 4; 3 ± 2.5 months, range 0–6 months) or coincided (n = 1) with IAD.
Isolated ACTH deficiency-related data
IAD was diagnosed at a median time of 7.0 months (IQR, 5–12) after starting immunotherapy. We did not find significant differences in relation to the time to diagnosis between patients treated with monotherapy (7.0 months, IQR, 5–12) and those treated with combined treatment (8.0 months, IQR, 4.5–12.5). There were also no significant differences in the time to develop IAD in relation to the type of antibody used (nivolumab, 7.5 months (IQR, 5–13); ipilimumab 6.0 months (IQR, 3–13.5); and pembrolizumab 7.0 months (IQR, 5.2–10.5)).
Physical examination, analytical evaluation, and pituitary MRI study of the patients at the time of IAD diagnosis are shown in Table 2.
According to the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0 for endocrine disorders (35) most of the patients (n = 22, 59.5%) presented grade 2 adrenal insufficiency (AI) (moderate symptoms), approximately one-third (n = 12, 32.4%) showed grade 3 (severe symptoms) and 8.1% (n = 3) were asymptomatic (grade 1). We found no association between the severity of the AI and the age at diagnosis, history of diabetes, hypertension, hyperlipidemia, cancer type, treatment used prior to immunotherapy, ICI used, type of ICI therapy (mono- or combined therapy), and other associated irAEs.
The main reported symptom was fatigue (97.3%), followed by anorexia (81.8%), general malaise (81.1%), hypotension/syncope (45.9%), and nausea/vomiting (45.9%) (Fig. 2).
Percentage distribution of clinical symptoms associated with isolated ACTH deficiency secondary to ICI therapy.
Citation: Endocrine-Related Cancer 28, 12; 10.1530/ERC-21-0228
Percentage distribution of clinical symptoms associated with isolated ACTH deficiency secondary to ICI therapy.
Citation: Endocrine-Related Cancer 28, 12; 10.1530/ERC-21-0228
Percentage distribution of clinical symptoms associated with isolated ACTH deficiency secondary to ICI therapy.
Citation: Endocrine-Related Cancer 28, 12; 10.1530/ERC-21-0228
Eosinophilia (upper limit of normal (ULN)) and hyponatremia (lower limit of normal (LLN)) were present in 30.6 and 24.3% of the patients, respectively. Only 3 (8.1%) patients showed alterations on the pituitary MRI (two patients showed empty sella and one a pituitary microadenoma).
In most patients, IAD was treated at diagnosis with hydrocortisone (n = 31, 83.8%), followed by prednisone (n = 5, 13.5%) and methylprednisolone (n = 1, 2.7%). The median daily doses were 50 mg (30–100), 60 mg (22.5–77.5), and 50 mg, respectively. These doses were later reduced to usual replacement doses. During follow-up, 36 patients were treated with hydrocortisone (median dose 25.0 mg/day (20–30)), and one patient with prednisone (2.5 mg/day).
Clinical situation at the last visit
Mean follow-up time since IAD diagnosis was 15.2 ± 12.5 months (range 0.3–55 months). At last visit, no patient had recovered HPA axis function.
The type of corticosteroid used was hydrocortisone (n = 31, 83.8%, median daily dose 20 mg (20–30)), dexamethasone (n = 2, 5.4%, mean dose 6 mg/day), prednisone (n = 1, 2.7%, dose 10 mg/day), and methylprednisolone (n = 1, 2.7%, dose 70 mg/day).
Both basal serum cortisol and plasma ACTH concentrations at the last visit were similar to those obtained at the diagnosis of IAD (Table 2). The same occurred when only patients who were treated with replacement doses of hydrocortisone were compared (ACTH, 4.0 pg/mL (3.2–7.7) and 4.0 pg/mL (1.4–5.4), basal vs last visit, ns; and cortisol 0.8 µg/dL (0.31.1) vs 1.0 µg/dL (0.5–1.8), basal vs last visit, ns).
Tumor response evaluation at the last visit showed complete response (CR) in 11 patients (29.7%), partial response (PR) in eight (21.6%), stable disease (SD) in five (13.5%), and progression (P) in the remaining 13 (35.1%). In this IAD population, no significant association was found between tumor response and several variables analyzed such as age, sex, type of cancer, type of ICI and ICI therapy (monotherapy or combined therapy), and other associated irAEs.
At the end of the study, eight patients (21.6%) had died. The median overall survival since cancer diagnosis, initiation of immunotherapy, and IAD diagnosis was 28.0, 17.0, and 6.0 months, respectively (Fig. 3). In seven of the eight patients, death was related to cancer (tumor progression (n = 4), severe respiratory insufficiency (n = 2), and cardiorespiratory arrest (n = 1)). In one patient, the cause of death was unknown. There were no known cases of death directly related to AI. Patient mortality was only associated with a history of immunotherapy-induced pneumonitis (chi-square value 7.7; P = 0.042) and tumor progression (chi-square value 12.3; P = 0.001).
Kaplan–Meier survival analysis in 37 cancer patients who developed isolated IAD induced by ICI therapy since tumor diagnosis (left); the beginning of immunotherapy (center); and diagnosis of central AI to death (right).
Citation: Endocrine-Related Cancer 28, 12; 10.1530/ERC-21-0228
Kaplan–Meier survival analysis in 37 cancer patients who developed isolated IAD induced by ICI therapy since tumor diagnosis (left); the beginning of immunotherapy (center); and diagnosis of central AI to death (right).
Citation: Endocrine-Related Cancer 28, 12; 10.1530/ERC-21-0228
Kaplan–Meier survival analysis in 37 cancer patients who developed isolated IAD induced by ICI therapy since tumor diagnosis (left); the beginning of immunotherapy (center); and diagnosis of central AI to death (right).
Citation: Endocrine-Related Cancer 28, 12; 10.1530/ERC-21-0228
Discussion
To date, only 60 patients with isolated ACTH deficiency induced by immunotherapy have been reported. Main clinical and biochemical features of these patients have recently been summarized in a systematic review (Iglesias et al. 2021). Our contribution from this national cohort of 37 patients allows us to make a comparison of the results obtained with those previously reported and add the new information that we have been able to collect.
This is the largest study published to date on the IAD syndrome induced by ICI immunotherapy in cancer patients. This study analyzes in detail the clinical and diagnostic features and therapeutic outcomes through a retrospective and multicenter study. Our results show that IAD appears in 0.8% of patients treated with ICI, mainly in males in the seventh decade of life and approximately 7 months after starting immunotherapy. In our population, the most frequently reported tumor was lung cancer, followed by melanoma, while the type of ICI most commonly used was anti-PD1, preferably nivolumab, followed by pembrolizumab, and primarily as monotherapy. No patient recovered HPA axis function. Lastly, median survival after IAD diagnosis was around 6 months.
IAD is a well-differentiated syndrome whose pathogenesis is unknown in most patients. It was first reported by Steinberg et al. (1954) and is characterized by the presence of a secondary AI with the rest of the anterior pituitary function preserved and without structural pituitary alterations. IAD can be congenital or acquired. The former is generally due to mutations in T-box pituitary restricted transcription factor (TBX19) gene or TPIT gene, while the second form is usually due to exogenous administration of glucocorticoids which produce a prolonged suppression of the corticotropic axis function. Other less frequent etiologies are trauma, Sheehan’s syndrome, and radiotherapy. On other occasions, autoimmune pathogenesis appears to be responsible for IAD. This is supported by the fact that IAD is frequently associated with other autoimmune endocrinopathies and/or autoimmune non-endocrine diseases, the onset in the postpartum period in women, the presence of anti-pituitary antibodies, and its association with autoimmune paraneoplastic syndromes (Yamamoto et al. 1992, Bando et al. 2018, Kobayashi et al. 2021). Moreover, it has been also suggested that IAD could be caused by autoimmunity to corticotrophs (Fujita et al. 2019, Yamamoto et al. 2020).
IAD induced by ICI therapy in cancer patients is a recently reported syndrome. Since its first description by Okano et al. (2016), the number of reported cases have rapidly increased so far (Iglesias et al. 2021) and is expected to become even more prevalent as immunotherapy use with ICI increases. In a retrospective cohort study of a tertiary cancer center recently published in 2020 by Percik et al. (2020), 14 patients (0.86%) of 1615 treated with ICI were diagnosed with IAD. This prevalence was similar to that obtained in our study (0.8%; 37 of 4447 patients ICI treated) showing that IAD is a rare event associated to ICI therapy; however, it should always be kept in mind in these patients due to its important clinical implications.
To date, the only two reported series have shown discordant results regarding sex predominance associated with ICI-induced IAD. The first one published in 2020 (Percik et al. 2020) showed a higher prevalence in women (n = 14; 10 women; 71.4%); however, in a more recently published systematic review based on cases published in the literature, men were the most frequently affected (n = 60; 37 men; 61.7%) (Iglesias et al. 2021). The latter was similar to that found in our Spanish population (n = 37; 23 men; 62.2%). Considering the three series together, men constitute the majority of the reported patients with this complication associated with immunotherapy (n = 106; 73 men, 68.9%). However, male preponderance is more likely related to epidemiological differences in each tumor and the different ICI used. In fact, in our population the most frequently evaluated type of cancer was lung cancer (n = 20), followed by melanoma (n = 8), with a male:female ratio of 3:1 and 1.7:1, respectively. Lastly, something similar happened with the type of ICI antibody used. The most frequently used were nivolumab (n =11) and pembrolizumab (n =12) with a male:female ratio of 4.5:1 and 1.4:1, respectively.
The presence of other irAEs associated with immunotherapy are common in patients who develop ICI-induced IAD. They have been reported in 35.7% of IAD patients, all of which preceded IAD (Percik et al. 2020). Our study showed a higher prevalence of irAEs (51%), mainly endocrine, digestive, and cutaneous irAEs. Although most of them preceded IAD, some coincided and even followed (up to 23 months) the diagnosis of IAD. Therefore, the possibility of the development of other irAEs associated with immunotherapy should be borne in mind even after developing IAD.
The previously reported median time from the start of ICIs to IAD development has been 5.8 months (range 3–16 months) in one study (Percik et al. 2020) and 6.0 months (4–8) in another one (Iglesias et al. 2021). In our population this parameter was slightly longer 7.0 months (5–12). Although a previous study showed that combined treatment with anti-PD1 and anti-CTLA4 developed IAD significantly (P= 0.023) earlier (3.0 months (3.0–3.7)) than with anti-PD1 monotherapy (6.0 months (4.8–7.0) for nivolumab and 6.0 months (3.0–12.5) for pembrolizumab) (Iglesias et al. 2021), these findings could not be demonstrated in our study. The cause of this discrepancy is not clear, although multiple factors such as the type of cancer, previous tumor treatments, and the type of immunotherapy, etc. could have played a role. With all these data we can establish that IAD usually takes several months to develop, even reaching up to 16 months after starting immunotherapy. Therefore, it is necessary to be aware of this possible complication and have a high index of suspicion during the long-term clinical follow-up of these patients.
The main symptoms associated with IAD were fatigue, followed by anorexia, general malaise, hypotension/syncope, and nausea/vomiting. All of them are nonspecific symptoms that can appear in cancer patients, without the need for an underlying AI. This fact, together with the late appearance (several months) after the initiation of immunotherapy, can make the clinical suspicion of the development of secondary AI more difficult. Both hyponatremia (57–68%) and eosinophilia (34–36%) have been related to IAD; however, in our population their prevalence was lower (24 and 31%, respectively) than that reported in other series (Percik et al. 2020, Iglesias et al. 2021). Although the cause of this discrepancy is not known, their presence should always force us to rule out the diagnosis of IAD. On the other hand, as in previous studies, hypoglycemia did not appear in any of our patients. Lastly, when the pituitary imaging study is normal in a cancer patient with a history of ICI treatment and no concomitant or past treatment with glucocorticoids, the presence of central AI would support the diagnosis of an immunotherapy-induced IAD. However, we should keep in mind that there may be coincident pituitary lesions that do not explain an ACTH deficiency, such as pituitary incidentalomas, small cysts of Rathke’s pouch or an empty sella syndrome, among others.
According to CTCAE, it has been proposed that patients with AI with mild (grade 1) or moderate (grade 2) symptoms should begin oral hydrocortisone replacement; whereas those patients with severe (grade 3) or life-threating symptoms (grade 4) should be treated with high-dose parenteral glucocorticoids as in an adrenal crisis along with i.v. rehydration with isotonic saline to later switch them to oral hydrocortisone until reaching replacement doses (Stelmachowska-Banaś & Czajka-Oraniec 2020). In our population, hydrocortisone was the glucocorticoid most frequently used at AI diagnosis, at stress doses of glucocorticoids in most of the patients. The maintenance doses were around 20 mg/day of hydrocortisone or equivalent.
It is noteworthy that in our study the function of the HPA axis was not recovered. Although the corticotropic damage could become permanent from the diagnosis of IAD, it is possible that hormone replacement therapy interferes with the possibility for adrenal axis regeneration due to the negative inhibition feedback loop of cortisol on the hypothalamus and pituitary function, as suggested by some authors (Percik & Shoenfeld 2020).
Our study constitutes the first series of cancer patients treated with ICI in which survival is evaluated after the diagnosis of IAD. Median survival was 6 months with a wide range from 3 to 29 months. However, no predictors of better survival related to the IAD diagnosis were found. Only a history of immunotherapy-induced pneumonitis and tumor progression during ICI therapy were factors related to mortality in our patients. We must bear in mind that with only 37 patients with different cancers and in different lines of treatment it is difficult to find any predictive factor, therefore more studies are necessary in order to find these factors.
Currently, two well-differentiated forms of pituitary involvement associated with ICI in cancer patients should be considered. On the one hand, the development of hypophysitis and, on the other, IAD. While the former is usually associated with anti-CTLA-4 antibodies, mainly ipilimumab, the latter is more frequently described with anti-PD-1 antibodies, essentially nivolumab (Faje et al. 2014, Faje 2016, Percik et al. 2020, Iglesias et al. 2021). The main distinctive characteristics of both forms of pituitary involvement are reflected in Table 3. The prevalence of hypophysitis associated with anti-CTLA-4 is very high, having been reported in up to 10–17% of patients treated with ipilimumab (Faje et al. 2014). However, IAD appears in less than 1% of patients treated with anti-PD1, mainly nivolumab (Percik et al. 2020). Mean time from the start of immunotherapy to the development of pituitary hormonal involvement is also different for both clinical forms; whereas for ipilimumab-induced hypophysitis it is approximately 9 weeks (Faje et al. 2014, Faje 2016, Iglesias 2018), IAD is much later, generally up to 6–7 months (Faje et al. 2014, Percik et al. 2020, Iglesias et al. 2021). The clinical presentation is also different. While in ipilimumab-induced hypophysitis the most common symptom is headache, in IAD it is fatigue. Lastly, in hypophysitis, several hormonal axes are generally affected and an enlarged pituitary gland is commonly seen on a pituitary MRI; however, in IAD the pituitary morphology is usually normal (Faje et al. 2014, Iglesias et al. 2021).
Clinical characteristics of the different forms of pituitary involvement in cancer patients treated with ICI antibodies.
| Hypophysitis | Isolated ACTH deficiency (IAD) | |
|---|---|---|
| Type of ICI | Anti-CTLA-4 (ipilimumab) | Anti-PD-1 (nivolumab) |
| Prevalence | 10–17% | <1% |
| Time since ICI therapy initiation | 9 weeks | 7 months |
| Main symptom | Headache | Fatigue |
| Hyponatremia | 47% | 24–68% |
| Pituitary hormone alterations | ≥2 | ACTH |
| Pituitary MRI | Pituitary enlargement | Normal |
Anti-CTLA-4, anti-cytotoxic T-lymphocyte antigen 4; Anti-PD-1, Anti-programmed cell death protein 1; ICI, Immune checkpoint inhibitor.
The main limitations of our study were: (1) its retrospective nature; (2) the absence of cortisol stimulation tests for the diagnosis of ACTH deficiency; and (3) the possibility of developing another pituitary hormone deficiency in the long-term follow-up. On the other hand, the strengths were: (1) its multicenter nature; (2) large number of cancer patients (more than 4000) treated with ICI; (3) the detailed analysis of the symptoms at the time of the IAD diagnosis; and (4) the evaluation of the long-term follow-up of the patients with IAD diagnosis. We are aware that it is possible that of the more than 4000 cancer patients analyzed, there was some with IAD that went unnoticed or in which IAD was not suspected and, therefore, it was not diagnosed, so the prevalence of immunotherapy-induced IAD may be somewhat higher than that reported so far.
In conclusion, our study shows that, although rare (<1%), IAD seems to be a well-established complication associated with ICI treatment in cancer patients. It develops around 7 months after starting treatment with anti-PD1 monoclonal antibodies, mainly nivolumab, and it is frequently associated with other endocrine diseases, especially thyroid diseases, and non-endocrine diseases. Lastly, recovery of the corticotropic axis function should not be expected.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This work did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector.
Acknowledgement
A F P is a Rio Hortega researcher (ISCIII; CM20/00155).
References
Anderson B & Morganstein DL 2021 Endocrine toxicity of cancer immunotherapy: clinical challenges. Endocrine Connections 10 R116–R124. (https://doi.org/10.1530/EC-20-0489)
Ariyasu H, Inaba H, Ota T, Yamaoka H, Furukawa Y, Iwakura H, Doi N, Yamamoto Y & Akamizu T 2018 Thyrotoxicosis and adrenocortical hormone deficiency during immune-checkpoint inhibitor treatment for malignant melanoma. In Vivo 32 345–351. (https://doi.org/10.21873/invivo.11244)
Bando H, Iguchi G, Kanie K, Nishizawa H, Matsumoto R, Fujita Y, Odake Y, Yoshida K, Suda K & Fukuoka H et al.2018 Isolated adrenocorticotropic hormone deficiency as a form of paraneoplastic syndrome. Pituitary 21 480–489. (https://doi.org/10.1007/s11102-018-0901-7)
Boudjemaa A, Rousseau-Bussac G & Monnet I 2018 Late-onset adrenal insufficiency more than 1 year after stopping pembrolizumab. Journal of Thoracic Oncology 13 e39–e40. (https://doi.org/10.1016/j.jtho.2017.10.023)
Brahmer JR, Abu-Sbeih H, Ascierto PA, Brufsky J, Cappelli LC, Cortazar FB, Gerber DE, Hamad L, Hansen E & Johnson DB et al.2021 Society for Immunotherapy of Cancer (SITC) clinical practice guideline on immune checkpoint inhibitor-related adverse events. Journal for ImmunoTherapy of Cancer 9 e002435. (https://doi.org/10.1136/jitc-2021-002435)
Cho KY, Miyoshi H, Nakamura A, Kurita T & Atsumi T 2017 Hyponatremia can be a powerful predictor of the development of isolated ACTH deficiency associated with nivolumab treatment [Letter to the Editor]. Endocrine Journal 64 235–236. (https://doi.org/10.1507/endocrj.EJ16-0596)
Conforti F, Pala L, Bagnardi V, De Pas T, Martinetti M, Viale G, Gelber RD & Goldhirsch A 2018 Cancer immunotherapy efficacy and patients’ sex: a systematic review and meta-analysis. Lancet: Oncology 19 737–746. (https://doi.org/10.1016/S1470-2045(1830261-4)
Cukier P, Santini FC, Scaranti M & Hoff AO 2017 Endocrine side effects of cancer immunotherapy. Endocrine-Related Cancer 24 T331–T347. (https://doi.org/10.1530/ERC-17-0358)
Faje A 2016 Immunotherapy and hypophysitis: clinical presentation, treatment, and biologic insights. Pituitary 19 82–92. (https://doi.org/10.1007/s11102-015-0671-4)
Faje AT, Sullivan R, Lawrence D, Tritos NA, Fadden R, Klibanski A & Nachtigall L 2014 Ipilimumab-induced hypophysitis: a detailed longitudinal analysis in a large cohort of patients with metastatic melanoma. Journal of Clinical Endocrinology and Metabolism 99 4078–4085. (https://doi.org/10.1210/jc.2014-2306)
Fujimura T, Kambayashi Y, Furudate S, Kakizaki A, Hidaka T, Haga T, Hashimoto A, Morimoto R & Aiba S 2017 Isolated adrenocorticotropic hormone deficiency possibly caused by nivolumab in a metastatic melanoma patient. Journal of Dermatology 44 e13–e14. (https://doi.org/10.1111/1346-8138.13532)
Fujita Y, Bando H, Iguchi G, Lida K, Shichi H, Kanie K, Matsumoto R, Suda K, Fukuoka H & Ogawa W et al.2019 SAT-LB079 clinical heterogeneity of acquired idiopathic ACTH deficiency: a new classification based on the clinical characteristics and autoantibodies. Journal of the Endocrine Society 3 SAT-LB079. (https://doi.org/10.1210/js.2019-SAT-LB079)
Grouthier V, Lebrun-Vignes B, Moey M, Johnson DB, Moslehi JJ, Salem JE & Bachelot A 2020 Immune checkpoint inhibitor-associated primary adrenal insufficiency: WHO VigiBase report analysis. Oncologist 25 696–701. (https://doi.org/10.1634/theoncologist.2019-0555)
Iglesias P 2018 Cancer immunotherapy-induced endocrinopathies: clinical behavior and therapeutic approach. European Journal of Internal Medicine 47 6–13. (https://doi.org/10.1016/j.ejim.2017.08.019)
Iglesias P, Soria A & Diez JJ 2015 Autoimmune endocrinopathies induced by immunomodulating antibodies in the treatment of cancer. Medicina Clinica 145 264–268. (https://doi.org/10.1016/j.medcli.2015.02.010)
Iglesias P, Sánchez JC & Díez JJ 2021 Isolated ACTH deficiency induced by cancer immunotherapy: a systematic review. Pituitary 24 630–643. (https://doi.org/10.1007/s11102-021-01141-8)
Ito A, Kondo S, Tada K & Kitano S 2015 Clinical development of immune checkpoint inhibitors. BioMed Research International 2015 605478. (https://doi.org/10.1155/2015/605478)
Johnson DB, Jakubovic BD, Sibaud V & Sise ME 2020 Balancing cancer immunotherapy efficacy and toxicity. Journal of Allergy and Clinical Immunology in Practice 8 2898–2906. (https://doi.org/10.1016/j.jaip.2020.06.028)
Joseph RM, Hunter AL, Ray DW & Dixon WG 2016 Systemic glucocorticoid therapy and adrenal insufficiency in adults: a systematic review. Seminars in Arthritis and Rheumatism 46 133–141. (https://doi.org/10.1016/j.semarthrit.2016.03.001)
Kastrisiou M, Kostadima FL, Kefas A, Zarkavelis G, Kapodistrias N, Ntouvelis E, Petrakis D, Papadaki A, Vassou A & Pentheroudakis G 2017 Nivolumab-induced hypothyroidism and selective pituitary insufficiency in a patient with lung adenocarcinoma: a case report and review of the literature. ESMO Open 2 e000217. (https://doi.org/10.1136/esmoopen-2017-000217)
Kitajima K, Ashida K, Wada N, Suetsugu R, Takeichi Y, Sakamoto S, Uchi H, Matsushima T, Shiratsuchi M & Ohnaka K et al.2017 Isolated ACTH deficiency probably induced by autoimmune-related mechanism evoked with nivolumab. Japanese Journal of Clinical Oncology 47 463–466. (https://doi.org/10.1093/jjco/hyx018)
Kobayashi T, Iwama S, Sugiyama D, Yasuda Y, Okuji T, Ito M, Ito S, Sugiyama M, Onoue T & Takagi H et al.2021 Anti-pituitary antibodies and susceptible human leukocyte antigen alleles as predictive biomarkers for pituitary dysfunction induced by immune checkpoint inhibitors. Journal for ImmunoTherapy of Cancer 9 e002493. (https://doi.org/10.1136/jitc-2021-002493)
Martins Machado C, Almeida Santos L, Barroso A & Oliveira MJ 2019 Nivolumab-induced hypothyroidism followed by isolated ACTH deficiency. BMJ Case Reports 12 e231236. (https://doi.org/10.1136/bcr-2019-231236)
Ohara N, Kobayashi M, Ohashi K, Ito R, Ikeda Y, Kawaguchi G, Yoneoka Y, Hasegawa G & Takada T 2019 Isolated adrenocorticotropic hormone deficiency and thyroiditis associated with nivolumab therapy in a patient with advanced lung adenocarcinoma: a case report and review of the literature. Journal of Medical Case Reports 13 88. (https://doi.org/10.1186/s13256-019-2002-2)
Okahata S, Sakamoto K, Mitsumatsu T, Kondo Y, Noso S, Ikegami H & Shiba T 2019 Fulminant type 1 diabetes associated with isolated ACTH deficiency induced by anti-programmed cell death 1 antibody-insight into the pathogenesis of autoimmune endocrinopathy. Endocrine Journal 66 295–300. (https://doi.org/10.1507/endocrj.EJ18-0328)
Okano Y, Satoh T, Horiguchi K, Toyoda M, Osaki A, Matsumoto S, Tomaru T, Nakajima Y, Ishii S & Ozawa A et al.2016 Nivolumab-induced hypophysitis in a patient with advanced malignant melanoma. Endocrine Journal 63 905–912. (https://doi.org/10.1507/endocrj.EJ16-0161)
Oristrell G, Baneras J, Ros J & Munoz E 2018 Cardiac tamponade and adrenal insufficiency due to pembrolizumab: a case report. European Heart Journal: Case Reports 2 yty038. (https://doi.org/10.1093/ehjcr/yty038)
Otsubo K, Nakatomi K, Furukawa R, Ashida K, Yoneshima Y, Nakanishi Y & Okamoto I 2017 Two cases of late-onset secondary adrenal insufficiency after discontinuation of nivolumab. Annals of Oncology 28 3106–3107. (https://doi.org/10.1093/annonc/mdx497)
Ott PA, Hodi FS & Robert C 2013 CTLA-4 and PD-1/PD-L1 blockade: new immunotherapeutic modalities with durable clinical benefit in melanoma patients. Clinical Cancer Research 19 5300–5309. (https://doi.org/10.1158/1078-0432.CCR-13-0143)
Percik R & Shoenfeld Y 2020 Check point inhibitors and autoimmunity: why endocrinopathies and who is prone to? Best Practice and Research: Clinical Endocrinology and Metabolism 34 101411. (https://doi.org/10.1016/j.beem.2020.101411)
Percik R, Shlomai G, Tirosh A, Tirosh A, Leibowitz-Amit R, Eshet Y, Greenberg G, Merlinsky A, Barhod E & Steinberg-Silman Y et al.2020 Isolated autoimmune adrenocorticotropic hormone deficiency: from a rare disease to the dominant cause of adrenal insufficiency related to check point inhibitors. Autoimmunity Reviews 19 102454. (https://doi.org/10.1016/j.autrev.2019.102454)
Sbardella E, Isidori AM, Woods CP, Argese N, Tomlinson JW, Shine B, Jafar-Mohammadi B & Grossman AB 2017 Baseline morning cortisol level as a predictor of pituitary-adrenal reserve: a comparison across three assays. Clinical Endocrinology 86 177–184. (https://doi.org/10.1111/cen.13232)
Steinberg A, Shechter FR & Segal HI 1954 True pituitary Addison’s disease, a pituitary unitropic deficiency; fifteen-year follow-up. Journal of Clinical Endocrinology and Metabolism 14 1519–1529. (https://doi.org/10.1210/jcem-14-12-1519)
Stelmachowska-Banaś M & Czajka-Oraniec I 2020 Management of endocrine immune-related adverse events of immune checkpoint inhibitors: an updated review. Endocrine Connections 9 R207–R228. (https://doi.org/10.1530/EC-20-0342)
Takaya K, Sonoda M, Fuchigami A & Hiyoshi T 2017 Isolated adrenocorticotropic hormone deficiency caused by nivolumab in a patient with metastatic lung cancer. Internal Medicine 56 2463–2469. (https://doi.org/10.2169/internalmedicine.8548-16)
Takebayashi K, Ujiie A, Kubo M, Furukawa S, Yamauchi M, Shinozaki H, Suzuki T, Naruse R, Hara K & Tsuchiya T et al.2018 Isolated adrenocorticotropic hormone deficiency and severe hypercalcemia after destructive thyroiditis in a patient on nivolumab therapy with a malignant melanoma. Journal of Clinical Medicine Research 10 358–362. (https://doi.org/10.14740/jocmr3257w)
Takeno A, Yamamoto M, Morita M, Tanaka S, Kanazawa I, Yamauchi M, Kaneko S & Sugimoto T 2019 Late-onset isolated adrenocorticotropic hormone deficiency caused by nivolumab: a case report. BMC Endocrine Disorders 19 25. (https://doi.org/10.1186/s12902-019-0335-x)
Tan MH, Iyengar R, Mizokami-Stout K, Yentz S, MacEachern MP, Shen LY, Redman B & Gianchandani R 2019 Spectrum of immune checkpoint inhibitors-induced endocrinopathies in cancer patients: a scoping review of case reports. Clinical Diabetes and Endocrinology 5 1. (https://doi.org/10.1186/s40842-018-0073-4)
Tanaka S, Kushimoto M, Nishizawa T, Takubo M, Mitsuke K, Ikeda J, Fujishiro M, Ogawa K, Tsujino I & Suzuki Y et al.2020 Isolated ACTH deficiency during single-agent pembrolizumab for squamous cell lung carcinoma: a case report. Clinical Diabetes and Endocrinology 6 1. (https://doi.org/10.1186/s40842-019-0092-9)
Topalian SL, Hodi FS, Brahmer JR, Gettinger SN, Smith DC, McDermott DF, Powderly JD, Carvajal RD, Sosman JA & Atkins MB et al.2012 Safety, activity, and immune correlates of anti-PD-1 antibody in cancer. New England Journal of Medicine 366 2443–2454. (https://doi.org/10.1056/NEJMoa1200690)
Wright JJ, Powers AC & Johnson DB 2021 Endocrine toxicities of immune checkpoint inhibitors. Nature Reviews: Endocrinology 17 389–399. (https://doi.org/10.1038/s41574-021-00484-3)
Yamamoto T, Fukuyama J, Hasegawa K & Sugiura M 1992 Isolated corticotropin deficiency in adults. Report of 10 cases and review of literature. Archives of Internal Medicine 152 1705–1712.
Yamamoto M, Iguchi G, Bando H, Kanie K, Hidaka-Takeno R, Fukuoka H & Takahashi Y 2020 Autoimmune pituitary disease: new concepts with clinical implications. Endocrine Reviews 41 bnz003. (https://doi.org/10.1210/endrev/bnz003)
Yoon JH, Hong AR, Kim HK & Kang HC 2021 Characteristics of immune-related thyroid adverse events in patients treated with PD-1/PD-L1 inhibitors. Endocrinology and Metabolism 36 413–423. (https://doi.org/10.3803/EnM.2020.906)
Zeng MF, Chen LL, Ye HY, Gong W, Zhou LN, Li YM & Zhao XL 2017 Primary hypothyroidism and isolated ACTH deficiency induced by nivolumab therapy: case report and review. Medicine 96 e8426. (https://doi.org/10.1097/MD.0000000000008426)
