A systematic review of ipilimumab-induced hypophysitis

Abstract

Graphical abstract

Abstract

Background

Ipilimumab, a CTLA-4-targeting monoclonal antibody, enhances T-cell activation and improves outcomes in various malignancies. However, it is associated with immune-related adverse events (IRAEs), including hypophysitis – a rare but potentially life-threatening condition. This review characterizes the clinical features, diagnostic approaches, and therapeutic strategies for ipilimumab-induced hypophysitis and explores its underlying pathophysiology through a case report and literature synthesis.

Methodology

We conducted a systematic review of published cases of ipilimumab-induced hypophysitis, extracting data on demographics, comorbidities, cancer types, treatment regimens, imaging findings, endocrine dysfunctions, and therapeutic outcomes. In addition, we present a detailed case report of a 60-year-old male with renal cell carcinoma who developed hypophysitis following ipilimumab–nivolumab combination immunotherapy.

Results

The literature review included 92 patients (mean age 57; 68% male), most commonly treated for melanoma. MRI revealed pituitary abnormalities in 46 patients. The most frequent symptoms were headache and fatigue, with panhypopituitarism and secondary adrenal insufficiency being the most common endocrine manifestations. Glucocorticoids were administered in 86 patients, and 62 required hormone replacement. Only 15/92 patients had full pituitary function recovery. Our case report mirrored these findings, with symptom onset after the third immunotherapy cycle and partial hormonal recovery following steroids.

Conclusions

Ipilimumab-induced hypophysitis is a significant IRAE with a variable clinical course and often irreversible endocrine dysfunction. Early recognition and management with glucocorticoids are critical, although long-term hormone replacement is frequently required. The autoimmune pathogenesis, linked to CTLA-4 expression in pituitary cells, underscores the need for further research into predictive markers and preventive strategies.

Introduction

Historical background

The body of evidence regarding molecular pathogenesis of cancer has been growing, as well as knowledge about potential molecular targets against cancer allowing for the development of novel selective therapies. One notable target is cytotoxic T-lymphocyte antigen 4 (CTLA-4), as its primary action is downregulation of T-cell activity. Antibody-mediated CTLA-4 blockage was first under investigation in 1996 with murine models, in order to increase T-cell activity and elicit an antitumor immune response. To further elucidate the effect of CTLA-4 inhibition in humans, researchers investigated how experimental models with primates responded to monoclonal antibodies in 2003. Given promising results of monoclonal antibodies in primate models, researchers decided to move forward with developing a CTLA-4 antibody that could be targeted toward human malignancies. There was a specific interest in developing an immunotherapy targeted against metastatic melanoma, given its aggressive nature, high mortality, and increasing prevalence. Eventually in 2011, ipilimumab emerged as a first-in-class drug, FDA-approved to treat metastatic melanoma (1). Since its initial FDA approval, ipilimumab has been further investigated in various other malignancies and has been FDA-approved as a first-line treatment against renal cell carcinoma and non-small cell lung cancer in combination with nivolumab, another widely used immunomodulatory drug (2, 3).

Mechanism of action and off-target effects

Ipilimumab binds to CTLA-4 and renders it unable to bind to its normal ligand (Fig. 1). Normally, CTLA-4 is expressed on activated T-cells. In order for T-cells to be completely activated, they need to bind to antigen-presenting cells (APCs). This process requires at least two signals. Prototypically, the first signal is between the major histocompatibility complex (MHC) on the APC and the T-cell receptor (TCR) on the T-cell. The second costimulatory signal is between B7 on the APC and CD28 on the T-cell. However, when CTLA-4 binds to B7 instead of CD28, the necessary costimulatory signal is disrupted and the T-cell is not activated. Therefore, by binding to CTLA-4, ipilimumab is able to prevent the inactivation of T-cells, increasing T-cell activity against cancer cells (1).

Figure 1.

Ipilimumab mechanism of action. (A) T-cell activation necessitating two signals: first between MHC and TCR, and second between B7 and CD28. (B) T-cell inactivation due to CTLA-4 binding to B7. (C) T-cell activation due to ipilimumab binding to CTLA-4.

However, in addition to creating an antitumor immune response, ipilimumab can create an autoimmune response by inhibiting CTLA-4, which acts as an immune checkpoint inhibitor, resulting in immune-related adverse events (IRAEs). Among the many IRAEs that have been recorded, autoimmune hypophysitis has emerged as an IRAE that can have a wide array of manifestations (4). We present a case of ipilimumab-induced hypophysitis and give an overview of previously reported cases through a literature review.

Methodology

Search strategy

A comprehensive literature search was conducted across the PubMed databases. The search included all studies published from January 2003 to January 2025. Keywords and MeSH terms used included the following: ‘ipilimumab’, ‘hypophysitis’, ‘immune checkpoint inhibitors’, ‘CTLA-4’, and ‘immune-related adverse events’. Boolean operators were applied to combine terms and maximize retrieval. Reference lists of relevant articles were also screened manually to identify additional eligible studies. The review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Supplementary Table S1 (see section on Supplementary materials given at the end of the article)).

Inclusion and exclusion criteria

We included case reports, case series, observational studies, and clinical trials that reported on patients diagnosed with ipilimumab-induced hypophysitis. Articles were included regardless of language, with non-English texts (Czech, Danish, Dutch, Spanish, German, Russian, Chinese, Japanese, Danish, and French) translated using Google Translate when necessary.

Studies were excluded if they lacked sufficient clinical details, focused on hypophysitis from non-ipilimumab agents, or were inaccessible after reasonable attempts to obtain them. Cases with multiple potential causes of hypophysitis were included only if ipilimumab was the most probable etiology based on clinical judgment and available data.

Data extraction

Two independent reviewers screened titles, abstracts, and full texts (Fig. 2). Discrepancies were resolved by consensus with a third reviewer. Extracted data included patient demographics, cancer type, ipilimumab dosage and cycles, timing of symptom onset, clinical presentation, imaging findings, hormonal profiles, treatment strategies, and outcomes. A standardized data collection form was used to ensure consistency.

Figure 2.

Flowchart of the screening process.

Statistical analysis

Descriptive statistics were used to summarize findings. Categorical variables were reported as frequencies and percentages, while continuous variables were expressed as means, medians, and ranges. Data analysis was performed using Microsoft Excel (version 16.0, Microsoft Corp., USA).

Results

Case report

We report the case of a 60-year-old male with a medical history of type 2 diabetes mellitus and metastatic renal cell carcinoma who presented to the Emergency Department of our hospital with acute onset of intractable headache. His oncologic history was notable for grade 2 clear cell renal carcinoma with metastases to the left lower lung lobe and left adrenal gland, both of which had been surgically resected. He was undergoing combination immunotherapy with ipilimumab and nivolumab and had completed three cycles at the time of presentation. His home medications included metformin 1,000 mg twice daily and atorvastatin 20 mg daily.

Five days following his latest immunotherapy infusion, the patient developed a progressively worsening headache. Two days prior to hospital presentation, he was evaluated at an urgent care clinic, diagnosed with sinusitis, and prescribed cephalexin. Despite antibiotic therapy, his headache worsened and was associated with nuchal rigidity on neurological examination.

Upon presentation, the headache was described as persistent, non-positional, and non-localized (holocephalic) and associated with nausea. He denied fever, photophobia, phonophobia, and vomiting. He was alert and oriented to person, place, and time. Neurological examination revealed no focal deficits. Cranial nerves II–XII were intact, motor strength was 5/5 in all extremities, and both coordination and gait were normal. Notably, he had a history of transient, brief headaches following previous immunotherapy sessions.

Given the concern for a meningeal process, a lumbar puncture was performed. Cerebrospinal fluid (CSF) analysis revealed pleocytosis consistent with aseptic meningitis: glucose 100 mg/dL, protein 72 mg/dL, 9 red blood cells/μL, and 7 nucleated cells/μL with 86% lymphocytes. Viral PCR and cryptococcal antigen testing were negative. Serum glucose was 170 mg/dL. He was empirically treated with intravenous acyclovir (650 mg every 8 h for 4 days, 10 doses in total), without clinical improvement.

Subsequent brain MRI revealed an enlarged and heterogeneous pituitary gland. Pituitary imaging confirmed a heterogeneously enhancing pituitary consistent with hypophysitis (Fig. 3). Given the temporal relationship with immunotherapy, ipilimumab-induced hypophysitis was suspected. The patient was initiated on corticosteroid therapy, resulting in a symptomatic improvement.

Figure 3.

Dedicated pituitary magnetic resonance imaging. Pituitary enlargement with heterogeneous enhancement. The pituitary stalk is midline. There is no evidence of optic chiasm compression. Craniocaudal diameter measures approximately 12 mm, significantly increased from prior neuroimaging when it measured approximately 5 mm.

Endocrine evaluation revealed the following: FSH (follicle-stimulating hormone) 3.3 mIU/mL, LH (luteinizing hormone) 1.5 mIU/mL, prolactin 7.2 ng/mL, TSH (thyroid-stimulating hormone) 0.12 μIU/mL, free T4 1.1 ng/dL, ACTH 52.5 pg/mL, and cortisol 3.8 μg/dL. He was discharged in stable condition with a steroid taper and scheduled outpatient oncology follow-up.

Literature review

Demographics and clinical characteristics

Since 2013, we found 80 articles about ipilimumab-induced hypophysitis with 92 patients being reported. The mean age of patients diagnosed with ipilimumab-induced hypophysitis was approximately 57 years, with a male predominance (68%) (Supplementary Tables S2 and S3). Comorbidities were reported in 10 patients, with hypertension being the most common (n = 5), followed by type 2 diabetes mellitus, hypothyroidism, multinodular goiter, and psoriasis (n = 2 each). Less frequent comorbidities included hyperthyroidism, meningioma, thromboembolic events, chronic obstructive pulmonary disease, familial ovarian cancer, cerebral aneurysm, obesity, hyperlipidemia, and gastroesophageal reflux disease (n = 1 each).

Among 92 patients with several cancer types, melanoma was the most prevalent (84%, n = 77), followed by renal cell carcinoma (8%, n = 7), prostate cancer (5%, n = 5), and lung cancer (2%, n = 2). Metastatic disease was documented in 51 patients, with the lungs (n = 20) and liver (n = 14) being the most common sites. Other metastatic sites included the brain, bone, adrenal glands, lymph nodes, and various soft tissues.

Imaging and endocrine findings

Brain MRI findings were available for 61 patients. Of these, 46 demonstrated pituitary abnormalities, including pituitary enlargement (n = 20), stalk thickening (n = 10), optic chiasm compression (n = 2), and heterogeneous or necrotic lesions. Interestingly, eleven patients had normal MRI findings, and a few exhibited unrelated abnormalities, such as cerebellar or calvarial metastases.

Endocrine dysfunction was a prominent feature, with 77 patients experiencing ipilimumab-related endocrine adverse effects. These included 33 cases of panhypopituitarism (defined as ≥3 hormone deficiencies) (incidence 35.9%; 95% CI: 26.6–46.2%), 27 cases of central adrenal insufficiency (incidence 29.3%; 95% CI: 20.8–39.5%), 22 cases of central hypothyroidism (incidence 23.9%; 95% CI: 16.2–33.9%), and 11 cases of central hypogonadism (incidence 12.0%; 95% CI: 6.5–20.7%) (Fig. 4). Of note, additional endocrine complications included SIADH (n = 2), hyperthyroidism, thyrotoxicosis with suppressed TSH, primary hypothyroidism, hypocortisolism with normal ACTH, primary hypergonadotropic hypogonadism, and immune-mediated hypoparathyroidism with associated hypocalcemia, hypomagnesemia, and hyperphosphatemia. Electrolyte disturbances included hyponatremia (n = 6), hypoglycemia (n = 2), and hyperglycemia (n = 1). Two patients had unspecified pituitary hypophysitis, one had clinically silent disease, and one was asymptomatic but had abnormal thyroid function tests.

Figure 4.

Forest plot showing the incidence of endocrine dysfunction in patients with ipilimumab-induced hypophysitis. The points represent the proportion of patients experiencing each hormonal deficiency, and the horizontal lines indicate 95% confidence intervals calculated using the Wilson method. Among 92 patients, panhypopituitarism was the most common manifestation, followed by central adrenal insufficiency, central hypothyroidism, and central hypogonadism.

Pituitary function recovery was observed in 15 patients, while 33 showed no recovery. Several patients experienced partial recovery or persistent deficiencies, particularly in ACTH and thyroid hormone axes. One patient had chronic hypophysitis without neurological recurrence.

Treatment and outcomes

Glucocorticoid therapy was administered in 86 patients, and 62 required hormone replacement therapy. Most received unspecified hormone regimens, while a few were treated with thyroid hormone, testosterone, or calcium supplementation.

The average number of ipilimumab cycles was three. Doses ranged from 1 to 10 mg/kg, with 3 mg/kg being the most common. Clinical responses included disease stabilization (n = 7), partial response (n = 4), and progression (n = 6). One patient died due to disease progression.

Adverse events

Headache was the most frequently reported symptom (n = 48), followed by fatigue (n = 19), nausea (n = 17), and emesis (n = 8). Gastrointestinal symptoms included diarrhea (n = 7), colitis (n = 6), and hepatitis (n = 5). Cutaneous manifestations (e.g., rash and pruritus) and neurological complications (e.g., confusion, hallucinations, and Guillain–Barré syndrome) were also noted.

Other reported complications included ophthalmologic disturbances (e.g., diplopia and visual field defects), renal impairment (e.g., interstitial nephritis), cardiovascular events (e.g., hypotension and pericardial effusion), hematologic abnormalities (e.g., thrombocytopenia and TTP), and systemic symptoms, such as muscle weakness, decreased libido, and emotional lability.

Discussion

General clinical course and comparison to previous literature

Ipilimumab-induced hypophysitis (IIH) typically manifests 8–10 weeks after treatment initiation, often following the second or third dose (5). The median time to diagnosis has been reported to be 8.4 weeks (range: 6.9–10.3 weeks) or around 2.3 months (6, 7). This timing generally aligns with the later appearance of endocrine immune-related adverse events observed with checkpoint inhibitors, which typically occur after the sixth or seventh week of treatment, with a median time to onset of 7–20 weeks (8).

IIH is characterized by a range of symptoms. The clinical course is indolent, commonly including symptoms such as headache and fatigue (4, 9). Visual disturbances or visual field defects may occur, similar to symptoms attributed to mass effect such as in the case of pituitary adenoma (10). Various other nonspecific symptoms have been reported, including generalized weakness, anorexia, GI disturbances, temperature intolerance, loss of libido, or altered mental status (5, 11). Abnormalities including hypotension, hypoglycemia, or hyponatremia indicate the need for further endocrine evaluation (12). Endocrine adverse effects are prominent, with panhypopituitarism being the most common in our review. More specifically, deficiencies of ACTH, TSH, and gonadotropins were the most common. While this is consistent with previous studies showing that adrenal insufficiency and hypothyroidism occurred in 60–100% of patients and hypogonadotropic hypogonadism in 83–87% of male patients, our review emphasizes the severity of endocrine dysfunction by demonstrating that these hormone deficiencies often present concurrently rather than in isolation (5, 6). The morbidity associated with IIH has been predominantly linked to secondary adrenal insufficiency, a potentially life-threatening condition if not promptly recognized and treated (4). Recovery of ACTH secretion is rare, often necessitating indefinite replacement (13). Pituitary enlargement with homogeneous or heterogeneous contrast enhancement on MRI is common, although normal MRI does not rule out hypophysitis (5, 14).

The prevalence of IIH has been reported to range from 0.4 to 17%, while the incidence of IIH has been reported to be 5.6% for ipilimumab alone and 8.8–10.5% for combination therapy (5, 9). Generally speaking, patients receiving combination therapy have been shown to have a significantly higher tendency to experience hypophysitis compared to those receiving ipilimumab monotherapy, which is consistent with our case report (15, 16). However, it is important to note that hypophysitis is rarer when chemotherapy includes nivolumab without ipilimumab, in comparison with chemotherapy including ipilimumab (17). Hence, even though patients receiving combination therapy tend to experience hypophysitis more than patients receiving ipilimumab monotherapy, the main contributor to the development of hypophysitis has been suggested to be ipilimumab in both scenarios.

There is strong alignment with our review in comparison with Araujo et al. (4), in which the authors found a higher frequency of IIH among older and male patients. Our review supports this demographic, showing a mean age of approximately 57 years and a 68% male predominance. Fatigue and headache were also the most frequently initial clinical manifestations reported in their review, affecting 58.6 and 61.0% of cases, respectively. Fatigue and headache were the most frequently reported non-endocrine adverse effects in our review as well. Regarding clinical course, Araujo et al. suggested a cumulative effect in which IIH symptoms typically emerged after the third cycle of treatment, which was the case for our patient. Furthermore, the mean number of cycles with ipilimumab received was approximately 3 in our review. Regarding treatment outcomes, Araujo et al. concluded that glucocorticoid therapy and hormone replacement were required in the majority of patients and that most patients (63.6%) experienced permanently impaired pituitary function. Our review reinforces this pattern: while glucocorticoid therapy was administered to the majority of patients (86 patients), complete pituitary function recovery was achieved in only 15, with many (62 patients) requiring continued hormone replacement therapy.

Pathophysiological hypothesis for ipilimumab-induced hypophysitis

The precise mechanism of IIH has not been entirely elucidated but is widely believed to be due to an autoimmune process. Studies have shown the development of circulating autoantibodies in patients with IIH against TSH-, FSH-, and ACTH-secreting cells, aligning with the observed patterns of hormone deficiencies (11, 18). In addition, autoantibodies against guanine nucleotide-binding protein G (olf) subunit alpha (GNAL) and integral membrane protein 2B (ITM2B), both normally expressed in the pituitary, have been associated with the development of checkpoint inhibitor-induced hypophysitis (13).

An interesting hypothesis suggests that CTLA-4 is expressed ‘ectopically’ on pituitary endocrine cells. Given that ipilimumab functions to activate T cells by binding to CTLA-4 in order to release its inhibitory effect, ipilimumab is hypothesized to bind not only to the primary target of cancer cells but also to off-target CTLA-4 receptors, such as on pituitary endocrine cells (11). This lies in contrast to PD-1 (e.g., nivolumab and pembrolizumab) and PD-L1, which have not been shown to be expressed in the pituitary (13). This understanding is consistent with the finding that hypophysitis is considerably more frequent with CTLA-4 inhibitors, such as ipilimumab, compared to PD-1/PD-L1 inhibitors (15).

Autopsy studies further suggest type II (IgG-mediated) and type IV (T-lymphocyte-dependent) hypersensitivity reactions in the pituitary gland of affected individuals (5). Ipilimumab is known for its ability to activate the classical complement pathway, even in comparison with other agents in its class (e.g., tremelimumab) (11). Although both agents target CTLA-4, ipilimumab is an IgG1 monoclonal antibody, whereas tremelimumab is an IgG2b monoclonal antibody. Furthermore, PD-1/PD-L1 inhibitors are often IgG4. In this way, the difference in IgG subclass has been hypothesized to be involved in the development of IIH (13).

Overall, the varying incidence of hypophysitis among the different checkpoint inhibitors suggests distinct pathological mechanisms (13). Combination therapy with both CTLA-4 and PD-1/PD-L1 inhibitors consistently shows the highest incidence of hypophysitis, further reinforcing the non-redundant roles of these pathways in immune regulation (13, 14, 15). However, the overall finding that CTLA-4 inhibitors as a whole are significantly more likely to induce hypophysitis compared to PD-1 and PD-L1 inhibitors supports the hypothesis that hypophysitis is a class effect of anti-CTLA-4 therapy, with ipilimumab being the most potent inducer due to its immunoglobulin subclass and pituitary-specific off-target effects.

Glucocorticoid effect

The current treatment of choice for IIH is glucocorticoids, showing the most benefit in acute symptomatic relief (4). Although the value of glucocorticoids in addressing the life-threatening potential of adrenal crisis is undisputed, the benefit of high-dose glucocorticoids for IIH has been a subject of debate. Many authors recommend high doses with subsequent taper followed by maintenance dosing for severe cases, especially those with mass effect symptoms such as severe headache and/or visual disturbances, in order to acutely reduce inflammation (9, 19). However, high-dose glucocorticoids do not seem to play a significant role in reversing hypopituitarism, particularly in the case of secondary adrenal insufficiency (9, 11, 20). Based on this, some studies suggest lower physiological doses of glucocorticoids in most cases, with high doses reserved for only the most severe forms (11, 19, 21). Furthermore, one retrospective study suggested that high-dose glucocorticoids for IIH may be associated with reduced survival in melanoma patients, although no other studies have found a similar negative association (17). While the benefit of glucocorticoids must be weighed against their potential immunosuppressive effect, most studies show that short-term use of high-dose glucocorticoids does not appear to negatively impact the anti-cancer efficacy of checkpoint inhibitors (4, 7, 11, 12). Regardless of the initial high-dose steroid regimen, lifelong glucocorticoid replacement therapy is generally required, as secondary adrenal insufficiency is known to be permanent in nearly all cases of IIH (7, 9, 19).

On the other hand, pituitary function has been shown to recover both spontaneously and with steroid treatment in the case of idiopathic lymphocytic hypophysitis (10, 11). This contrasts with the seemingly irreversible nature of IIH-induced hormonal deficiencies, highlighting a distinction in the utility of steroids for functional recovery between different etiologies of hypophysitis. While glucocorticoids effectively manage acute inflammation and symptoms across various forms of hypophysitis, their capacity to restore long-term endocrine function seems more limited in the context of IIH, as permanent hormone replacement is frequently necessary.

Future studies

As the use of immune checkpoint inhibitors continues to expand across various malignancies, future research must address several critical gaps in the understanding and management of ipilimumab-induced hypophysitis (IIH). First, there is a pressing need to identify predictive biomarkers that could help stratify patients at higher risk of developing IIH. These may include genetic predispositions, baseline immune profiles, or early serologic markers, such as pituitary-specific autoantibodies.

Second, prospective studies are needed to better define the natural history of IIH, particularly regarding the timeline of symptom onset, progression of hormonal deficiencies, and potential for recovery. Longitudinal endocrine follow-up in larger patient cohorts would help clarify which patients may benefit from long-term hormone replacement versus those who may experience partial or full recovery.

Third, the optimal glucocorticoid regimen remains unclear. Comparative studies evaluating high-dose versus physiological-dose steroid therapy could help determine the most effective and least harmful approach, especially in balancing symptom control with preservation of antitumor immunity.

Fourth, mechanistic studies exploring the immunopathology of IIH – particularly the role of CTLA-4 expression in pituitary tissue and the involvement of specific IgG subclasses – could lead to targeted preventive strategies or safer checkpoint inhibitor designs.

Finally, future trials should assess whether prophylactic interventions (e.g., low-dose steroids and immune modulation) could reduce the incidence or severity of IIH without compromising the efficacy of cancer immunotherapy. Collaborative registries and multicenter studies will be essential to gather sufficient data on this rare but impactful condition.

Conclusion

Ipilimumab-induced hypophysitis (IIH) represents a significant immune-related adverse event associated with CTLA-4 blockade in cancer immunotherapy. While relatively rare, its clinical impact is substantial due to the potential for irreversible endocrine dysfunction, particularly secondary adrenal insufficiency. Our systematic review, supported by a representative case, highlights the typical onset of IIH after the second or third cycle of treatment, with symptoms such as headache and fatigue often preceding hormonal abnormalities.

Neuroimaging frequently reveals pituitary enlargement, although a normal MRI does not exclude the diagnosis. Hormonal evaluation remains essential, as panhypopituitarism and isolated deficiencies – especially of ACTH, TSH, and gonadotropins – are common. Glucocorticoid therapy is the mainstay of acute management, offering symptomatic relief and preventing adrenal crisis. However, the recovery of pituitary function is uncommon, and most patients require long-term hormone replacement therapy.

The pathogenesis of IIH appears to involve both humoral and cell-mediated autoimmune mechanisms, potentially triggered by ectopic CTLA-4 expression on pituitary cells. This distinguishes IIH from other forms of hypophysitis and underscores the unique immunological profile of CTLA-4 inhibitors compared to PD-1/PD-L1 agents.

Given the increasing use of immune checkpoint inhibitors, clinicians must maintain a high index of suspicion for IIH in patients presenting with nonspecific symptoms during or after treatment. Early recognition and appropriate endocrine evaluation are critical to reducing morbidity, which may be allowed for if screening for endocrine abnormalities is done before and after each cycle of chemotherapy. Future research should focus on identifying predictive biomarkers, refining diagnostic criteria, and evaluating strategies to prevent or mitigate IIH without compromising antitumor efficacy.

Declaration of interest

The authors declare that this work was conducted without commercial or financial relationships that could create a conflict of interest.

Funding

This work did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector.