Thyroid autoimmunity following alemtuzumab treatment in multiple sclerosis patients: a prospective study

ALZ, a humanized anti-CD52 monoclonal antibody, is approved for highly active RRMS, as it has been demonstrated that its administration decreases relapse rate and disability progression either in treatment-naive patients or in patients not responding to first-line immunomodulatory treatments [1]. CD52 is a cell surface marker expressed on T and B lymphocytes, natural killer cells, dendritic cells, and on most monocytes, but importantly not on hematopoietic precursors [2]. ALZ exerts its immunomodulatory action through binding to the CD52 antigen and the subsequent lysis and depletion of mature circulating CD52 + immune cells; thus, a transient but profound immunosuppression status with relatively brief B cell lymphopenia and prolonged T-cell lymphopenia, overlaps and is followed by an immune reconstitution phase. The prolonged modulation of lymphocyte composition and the long-lasting shift of the immunological balance seems to be relevant to the efficacy of this drug [3].

Alemtuzumab in MS is administered intravenously in two treatment cycles, 12 months apart; the first cycle includes 12 mg/day for five consecutive days, and the second, the same dose for three consecutive days; additional cycles may follow if needed. Apart from the infusion-associated reactions and mild to moderate infections, the principal adverse effect of ALZ is the development of secondary autoimmune disorders that can be observed mainly during the immune reconstitution period post-ALZ. The most common secondary autoimmune phenomenon is AITD.

In the general population AITD has been reported to occur in approximately 1–11%. In the Whickham survey, for example, 5% of women and 1% of men had both positive anti-thyroglobulin (anti-Tg) and anti-thyroperoxidase (anti-TPO) antibody tests and a serum thyroid stimulating hormone (TSH) value > 6 [4]. In the NHANES survey, anti-Tg antibodies were positive in 10.4% and anti‐TPO antibodies in 11.3% of the population [5] GD occurs more often in women and has a general population prevalence of 1–1.5%. Approximately 3% of women and 0.5% of men develop GD during their lifetime [6]. In the context of ALZ treatment of MS, AITD has been reported in approximately one-third of treated patients with no previous history of thyroid disorders, with most studies reporting a prevalence between 29.2% and 45.5% [7‐13]. In more detail, GD is the most common thyroid disorder (56–71%), followed by HT with hypothyroidism (6–34%), silent thyroiditis (ST) (3–12%), and hypothyroidism with positive TRAb (2–12.5%) [7, 8, 10‐12, 14‐16]. Some reports suggest that GD occurring in this context may be less aggressive than the non-ALZ-related-GD, while others indicate a more aggressive course with fluctuations between hypothyroidism and hyperthyroidism [7, 8, 12, 14, 15, 17]. The majority of ALZ-treated subjects are reported to develop thyroid dysfunction within 5 years, with a peak incidence in the third year from the first course [13, 18]. A threefold increased risk of AITD by smoking and a sevenfold increased risk by positive family history of thyroid dysfunction have also been reported, whereas the effect of gender is uncertain [12, 17]. Moreover, the presence of high baseline anti-TPO antibodies is associated with a greater risk of developing AITD[16, 19] Other secondary autoimmune conditions such as immune thrombocytopenic purpura (1–3%), Goodpasture syndrome or single cases of autoimmune neutropenia [20], hemolytic anemia [21] and type 1 diabetes have been rarely reported as well [16].

The exact pathogenic mechanism underlying post-ALZ AITD remains unclear; it is considered to be part of an immune reconstitution syndrome, an autoimmune phenomenon characterized by the recovery of the immune cells after lymphopenia. Similar manifestations have also been reported after highly active antiretroviral therapy (HAART) in human immunodeficiency virus (HIV) patients and after allogeneic or autologous hematopoietic stem cell transplantation (HSCT) [13, 22]. Interestingly, post-ALZ AITD has been described rarely or in very limited numbers when ALZ is administered for indications other than MS (e.g., vasculitis, Bechet’s disease, rheumatoid arthritis, neoplasms of the blood) [13]. In contrast, the prevalence of post-ALZ AITD is very high in MS, suggesting a disease-specific shift in the immunological environment during the reconstitution. Interestingly, no genetic association seems to exist between MS and GD. This high prevalence may be explained by the more rapid reconstitution of B cells without adequate regulatory control by T cells [22, 23]. Potentially as a consequence, a thyroid-specific humoral autoimmune response is induced [24]. It has also been suggested that patients who develop post-ALZ autoimmunity have higher basal levels of IL-21, a cytokine that is instrumental for the help T cells (T helper cells) offer to B cells so that their specificity toward the antigen increases. In T cells, cytokine expression indicates a shift toward a T helper cell 2 (Th2) profile, which could be related to high B cell numbers and IgG4 production [23, 25]. Interestingly, a recent study showed a distinct modulation pattern of serum IgG4 compared to the total IgG during the immune reconstitution after ALZ administration in MS patients, and the patients with high IgG4 levels were more prone to autoimmune manifestations; in addition, the highest IgG4 levels were found among patients developing GD[26].

Notably, real-world evidence studies are very few for ALZ, and real-life data concerning ALZ-induced AITD are lacking. In our study, we describe a cohort of RRMS patients who received ALZ courses at a tertiary academic center in Greece, aiming to define the laboratory characteristics, the timing of onset and the course of ALZ-induced AITD. Furthermore, herein we report two cases of successful pregnancies in ALZ-treated patients; one with GD and the other with HT and hypothyroidism.

We included 35 patients with a diagnosis of RRMS that were treated with ALZ at the Multiple Sclerosis and Demyelinating Diseases Unit of “Eginition” University Hospital from January 2016 to October 2021. In the following, all patients were referred to the tertiary Endocrinology Center at “Alexandra” University Hospital for evaluation and monitoring for possible AITD on follow-up. All patient data were collected prospectively. ALZ was administered in two cycles as described above; however, in one patient a third cycle of ALZ (12 mg/day × 3 days) was administered due to MS disease activity 27 months after the second cycle.

Evaluation included data collection of (1) patient demographic details, (2) previous disease-modifying treatments (DMTs) (3) baseline and quarterly (every three months) thyroid function tests including TSH, free thyroxine (fT4), triiodothyronine (T3), anti-TPO antibodies, anti-Tg antibodies and stimulating TRAb, (4) family history of thyroid autoimmunity and (5) smoking status history (current/ever/never), since the latter two parameters are reported to be risk factors for AITD [12, 17]. At baseline and during the quarterly follow-up a clinical review for thyroid-related symptoms was also performed. The primary outcome measure of our study was the number of patients developing AITD post-ALZ. Details of the time of onset of thyroid autoimmunity regarding treatment timeline, condition type, and patient outcomes were recorded.

AITD occurred slightly more frequently post-ALZ treatment in our cohort of MS patients (54.8%) compared to previous studies (29–45%)[7, 8, 10‐12, 14, 15, 17, 19, 30]; only Muller et al. [16] have reported a higher rate of 68.8%, although with a longer mean ± SD follow-up of 9 ± 2.5 years. In our cohort the median follow-up was 43.5 months (range, 12–69.5 months), similarly to most previous studies. In more detail, previous studies with a median follow-up of 36 [19], 84 [8], 57.3 [12], and 34 [17] months have reported AITD prevalence of 29.2%, 41%, 34%, and 16.5%, respectively. A metanalysis by Scapaticcio et al. [11] including seven studies with a median follow-up of 24–60 months has reported an estimated prevalence of 33%. Moreover, Pariani et al. [7], in their study with a mean follow-up of 67 months (range 6–251), reported a prevalence of 41.1%. Similarly, Manso et al. [14] in their retrospective study have reported a AITD prevalence of 39% during a mean follow-up period of 32 ± 17 months. Therefore, we report an increased rate of AITD compared to many studies with similar or longer mean follow-up.

Additionally, GD was the most common disorder (52.9%) we noted, in accordance with previous studies [7, 11, 12, 14, 17, 19]. The mean time to AITD onset was 19.4 months from the first ALZ course. Specifically, the mean onset of AITD was 10.2 months after the first ALZ course in seven cases and 13.5 months after the second course in the rest 10 cases, showing a time interval shorter than most previous studies have reported. According to these studies, the incidence peak was 28.5 months from the last ALZ dose and the median time of onset was 32 months after the first course or the 91% of AITD cases occurred within four years from the last ALZ course [7, 8, 11‐13, 17]. Only Manso et al. [14] in their real-life retrospective cohort have reported, similarly to our results, a mean time interval to AITD onset of 17 months after the first ALZ course and 10 months after the last. In our cohort, HT ± hypothyroidism was the most frequent thyroid disorder after GD (41.2%), followed by hypothyroidism with positive stimulating TRAb (5.9%), in line with previous studies [7, 11‐15]. No ST was recorded, while several reports mentioned a low prevalence of 4–12% [7, 11, 12, 14].

We further recorded a higher rate of patients with fluctuating GD (77.7%) than previous studies (prevalence of 15–67%) [7, 11, 12, 14, 16]. Possibly due to this fluctuation, we reported a very low remission rate of 11%. Permanent remission occurred only in 1 patient with fluctuating GD, who had been on “block and replace” ATD therapy for 24 months, 1 year after ocrelizumab treatment. Remission of GD 1 year after CD20 depletion points to possible involvement B cell deregulation in the pathogenesis of post-ALZ GD and highlights the need for an early switch to B cell depletion (if indicated by MS activity) or mitigation of B/T cell imbalance with low-dose rituximab [31]. On the last follow-up, the remaining GD patients were on ATD therapy ± levothyroxine with positive stimulating TRAb, while one patient needed radical treatment with radioactive iodine and another underwent total thyroidectomy. Similarly, Manso et al. [14] in their cohort recorded a low remission rate (18%) and Pariani et al. [7] recorded a remission rate of 34%. In contrast, initial clinical trials and earlier reports suggest that post-ALZ GD has a more favorable course than the “conventional” GD [8, 12, 17, 32] with a higher rate of response to ATD therapy [23, 33] and higher remission rates of 37–57% [8, 11, 12]. However, our findings are in accordance with a growing body of data indicating that GD after ALZ treatment seems to have a more unpredictable course compared to conventional GD, requiring long-term ATD therapy or possible radical treatment (radioactive iodine or surgery); it should be noted here that “conventional” GD has a remission rate of 50% [7, 12, 14, 34]. As previously reported, the fluctuating phenotype of GD is indicative of the alternating presence and/or coexistence of stimulating and inhibitory TRAb with a resultant predominant stimulation or inhibition of thyroid hormone secretion [7, 14, 16]. This fluctuation from hyperthyroidism to hypothyroidism and vice versa has been described in rare cases after levothyroxine treatment for hypothyroidism or after ATD treatment of conventional GD [35]. Accordingly, the presence of hypothyroidism with high stimulating TRAb titer may suggest the coexistence of a higher titer of blocking TRAb. Of note, the TRAb assay we performed only identifies stimulating TRAbs.

We found no correlation of AITD development with risk factors such as age at the time of ALZ treatment, sex, smoking habits, previous DMTs, positivity of baseline anti-TPO/anti-Tg antibody measurements, or history and family history of AITD. Scappaticcio et al. [11] report in their meta-analysis that data concerning risk factors for the development of post-ALZ AITD are scarce. Some studies have shown that smoking and positive family history of thyroid autoimmunity increase the risk of AITD [12, 17], while other studies report no correlation with sex, age, smoking status, or family history of thyroid dysfunction [14, 19]. However, most studies agree that baseline positivity of anti-TPO/ and or anti-Tg antibodies is significantly associated with increased risk for the development of AITD [12, 14, 16, 19, 30]. Few data concerning previous DMTs as a risk factor for AITD are found in the literature. Nevertheless, our results are in accordance with previous studies [7, 11, 14, 17, 19], which reported no significant association. Only the study by Pfeuffer et al.[36] demonstrated an increased risk of secondary autoimmunity among patients previously treated with fingolimod. In our study a trend for an increased occurrence of TRAbs in patients who received fingolimod as last treatment before ALZ was found as well, perhaps not reaching significance due to the number of patients.

In our cohort, the 17 patients who developed AITD had negative baseline anti-TPO/anti-Tg antibodies. Of note, only one patient with positive baseline anti-TPO antibodies developed fluctuating GD. Our negative results concerning baseline anti-TPO/anti-Tg antibody positivity as a risk factor may be due to the small number of such patients. Overall, 44.4% of our GD patients exhibited mild GO, contrary to previous studies which have shown a rate of 6.25–13.4% [7, 11, 12]. However, during quarterly follow-up, clinical examination in our study included screening for endocrine ophthalmopathy whereas in several other studies, ophthalmopathy screening was not routinely performed and therefore GO may have been underdiagnosed [7].

Interestingly, we also report on two pregnancies post-ALZ treatment; in the first case, the patient developed HT and hypothyroidism, and in the second GD hyperthyroidism. Information concerning incidence and course of post-ALZ AITD and especially ALZ-induced GD during pregnancy is limited to a few reports [37‐39] of such challenging cases. Herein, we report a pregnancy, in which GD hyperthyroidism with a high TRAb titer developed within the first trimester and required ATD treatment. Subsequently, hyperthyroidism improved allowing discontinuation of ATD treatment in the third gestational trimester and leading to the delivery of a healthy, euthyroid baby. On the contrary, similar case reports mention a more aggressive GD course with a rise or persistence of the TRAb titer in the third trimester necessitating escalation and continuation of ATD therapy with augmented risk of both fetal and neonatal thyrotoxicosis [37‐39]. In the general population, maternal hyperthyroidism occurs in about 0–2.5% of pregnancies [40], and in most cases the cause is GD. As mentioned above, the incidence of AITD is up to 45% of the patients treated with ALZ, of which 60–70% are GD cases [7, 14]. During pregnancy, hyperthyroidism may result in maternal, obstetrical, fetal, and neonatal complications. Fetal and neonatal hyperthyroidism occurs in 1–5% of pregnant mothers with GD, when maternal TRAbs cross the placenta stimulating the fetal thyroid gland, thus leading to excessive thyroid hormone secretion. The risk correlates with the TRAb titer; a maternal TRAb level > 5 IU/L or three times the upper reference range is associated with a higher risk for neonatal hyperthyroidism [41]. In our case, although the TRAb level in the beginning of pregnancy was three times over the upper reference range, hyperthyroidism resolved in the third trimester. In the general population, GD during pregnancy may improve in some patients and allow for the discontinuation of ATD treatment in the third trimester [42], however, in severe cases the risk for fetal and neonatal hyperthyroidism remains high [43]. Concerning future pregnancies, it is important to address the risk for secondary AITD in ALZ-treated women of childbearing age, mainly GD, in conjunction with their treating obstetrician.

The key strength of our study lies in the fact that it is based on real-world clinical and laboratory data prospectively collected by the same multidisciplinary medical team of an academic hospital center, with criteria and goals for treatment remaining constant during the whole study period. Although a limitation compared to other studies might be the lower number of patients, the present study provides real-world data that complement information derived from randomized controlled trials.

In conclusion, post-ALZ AITD may represent a dynamic spectrum of diseases with unique non-classical phenotypes, such as the unpredictable and fluctuating course of GD with resistance to ATD therapy necessitating radioiodine therapy or radical surgery.[7, 11, 12]. Further studies are needed to understand the underlying mechanisms that could be responsible for this immune dysregulation and the pathways related to AITD. Importantly, a detailed, thyroid-specific pre-treatment screening and a thorough long-term follow-up of all ALZ-treated patients can facilitate early diagnosis and effective treatment of possible secondary AITD and should therefore be considered standard of care in this setting. Recent guidelines recommend 3‐monthly thyroid function tests for at least 4 years after the last course of ALZ [13]. According to our results, for the fluctuating course of GD long-term “block and replace” ATD therapy in the minimum dose of each therapeutic agent or definitive treatment (radioiodine or thyroidectomy) may be needed.