While ICIs kill tumors by modulating the immune response, over-activated immune cells may also lead to autoimmune damage, i.e., immune-related adverse reactions (irAEs)[4].Studies[6] have shown that pablizumab is more likely to cause thyroid injury than other ICIs. It is now generally accepted that the onset of thyroid dysfunction due to ICIs is mainly between 2 and 6 weeks after ICIs administration. The natural course of hypothyroidism is similar to that of painless thyroiditis[4]. The natural course of the disease is similar to that of painless thyroiditis. Thyroiditis is a common disease, and the true prevalence of ICI-mediated thyroiditis has not been clearly reported. A retrospective study[7] of 657 patients being treated with ICI found that 67% of patients developed painless thyroiditis.
The pathogenesis of thyroiditis caused by PD-1 inhibitors is unclear. It may be related to the patient's underlying thyroid disease, TSH levels, autoantibody titers (underlying autoimmune disease), and other factors[8]. A prospective cohort study[9] confirmed that patients with positive anti-thyroglobulin antibodies/thyroid peroxidase antibodies were more likely to develop thyroid dysfunction after PD-1 inhibitor therapy. However, some studies have also[10] .It is believed that the mechanism of thyroid dysfunction triggered by PD-1 inhibitors involves T cells, natural killer cells and monocyte-related pathways, and is not related to thyroid autoantibodies.
ICI-associated pituitary inflammation occurs in 11–14% of patients treated with ICI, and the prevalence varies depending on the type of ICI[11–13]. The main manifestations of pituitary inflammation are pituitary enlargement and hypopituitarism, mostly in the form of anterior hypopituitarism, with some involvement of the posterior pituitary, mostly secondary to hyperaldosteronism, and also secondary to hypothyroidism[14]. The patient may also present with secondary hypothyroidism. In some patients recovering from pituitary inflammation, the pituitary gland may show vacuolation of the saddle[15, 16]. The pituitary gland may show vacuolated saddle. The mean onset of ICIs-related pituitary inflammation is currently considered to be mainly at 9 weeks after ICIs administration, with PD-1/PD-L1-induced pituitary inflammation at 3–5 months after administration[17], but is rare.
In this case, before pembrolizumab, T3 and T4 were below normal and tended to get progressively worse. TGAb and TPOAb were positive, but TSH was normal. The diagnosis should be considered: Hashimoto's disease with mild pituitary hypothyroidism[18]. The diagnosis should have been Hashimoto's disease with mild pituitary hypothyroidism. Unfortunately, the test was not definitive at that time. The patient had painless thyroiditis induced by the use of pembrolizumab, which is a common adverse drug reaction. After a brief period of hyperthyroidism, the patient develops hypothyroidism, so we use L-Thyroxine replacement therapy[18]. The special point of this case is that the starting labs support a mild pituitary hypothyroidism, however, the patient appeared to be a purely thyroidal hypothyroid with significant high TSH during the time of thyroiditis. However, during supplementation with L-Thyroxine, T3, T4 never reached the standard. TSH and T3, T4 did go into very low status. This cannot be explained by hypothyroidism. Combining the vacuolar saddle of the pituitary gland with normal reproductive hormone and adrenocorticotropic hormone indices, our interpretation is that the patient has isolated pituitary hypothyroidism (mild) due to the vacuolar saddle on top of Hashimoto's disease[19]. The TSH secretory function reserve of both the thyroid and pituitary glands is inadequate. When pembrolizumab induces painless thyroiditis, the hyperthyroid phase is very low. The subsequent hypothyroid phase is very severe, with extremely low T3, T4 ,stimulating the secretion of hypothalamic TRH (thyrotropin-releasing hormone), which in turn promotes abnormal proliferation and secretion of pituitary TSH cells. However, the persistent malignant hyperplasia eventually depletes the already fragile pituitary TSH cell population, resulting in severe pituitary hypothyroidism. The sketch of lesion development is shown in Fig. 2.
The patient's preoperative PET-CT showed no distant metastases and no obvious lesions in the head, and there were no symptoms such as headache and blurred vision during all phases of his immunotherapy. Although we did not perform pituitary-related tests, this case does not look very much like pituitary inflammation. During the period of his severely increased TSH, although there was no pituitary MRI examination to confirm the time of vacuolation saddle occurrence, and the presence of hyperplasia and pituitary inflammation of the pituitary gland for a short period of time could not be denied. However, T3 and T4 were extremely low and did not improve with L-Thyroxine supplementation. The core etiology was considered to be the ongoing injury to the thyroid gland in Hashimoto's disease by pablizumab. The initial Hashimoto's disease with hypothyroidism, if caused by the thyroid gland, should be pituitary hyperplasia[20]that should not be vacuolated pteroid saddle. Therefore, it is considered that there may be a combination of both thyroid and pituitary factors.
This patient has a reduced pituitary gland and a vacuolated butterfly saddle. Autoimmune thyroid disease with pituitary inflammation cannot be excluded, but it is difficult to confirm at this time. The patient had a month of severely reduced T3 and T4 and a significant increase in TSH, suggesting the presence of pituitary reactive hyperplasia[20]. After replacement therapy with L-Thyroxine tablets, the proliferating pituitary cells retract so that the funnel hole and pituitary fossa can leave some empty space[21].This can also produce a vacuolated pterygoid saddle. However, this could not be confirmed.
This patient had poor pituitary function prior to the use of PD-1 inhibitors, and the treatment did not exclude that PD-1 inhibitors caused damage to the pituitary gland itself. However, evidence for this is lacking. The TRH excitation test was not performed in this case, and the assessment of his pituitary and thyroid reserve function was inadequate.