The main adverse events following alemtuzumab include well-recognised infusion reactions, infections and acquired autoimmune disease including thyroid disorders, idiopathic thrombocytopaenic purpura (ITP) and immune-mediated nephropathies. Other rarely occurring autoimmune disease observed in the CARE-MS trials included neutropaenia, haemolytic anaemia, agranulocytosis and pancytopenia [
7,
8].
Autoimmune Disease
A unique observation in the use of alemtuzumab in MS is the occurrence of acquired autoimmune disease (AID), which is not observed in other conditions treated with this drug [
51]. A range of autoimmune conditions has been described and in our own long-term follow-up cohort amounted to 47 % of patients [
49]. Autoimmune disease of the thyroid gland is the most commonly occurring condition, with others including ITP and anti-glomerular basement membrane (anti-GBM) disease seen less frequently [
7‐
9]. This development of novel autoimmunity is currently the focus of active research, offering a window into the genesis of human autoimmune disease. Whilst the underlying mechanism is still not fully understood, it is thought to occur as a result of subsequent homeostatic T cell proliferation following lymphodepletion [
52], and where homeostatic proliferation predominates over thymic reconstitution, autoimmune disease is more likely [
52]. Of interest, thyroid AID following alemtuzumab treatment is not seen in patients treated with B cell chronic lymphocytic leukaemia despite being given at higher doses suggesting a disease-specific phenomenon [
51].
Early experience suggested an occurrence of thyroid AID in approximately 30 % of patients [
34,
53]. Similar rates were seen in CAMMS223, CARE-MSI and CARE-MSII at 23, 18 and 16 %, respectively [
7‐
9]. A large multi-centre cohort subsequently demonstrated thyroid AID rates of 17 % [
54]. In recently published long-term follow-up cohorts, total AID specific to the thyroid has been shown to occur at a higher rate and probably reflects longer follow-up. In our own multi-centre cohort, 47 % developed AID of which 35 % were thyroid [
49]. Similarly, in the Cambridge cohort, 48 % patients experienced any AID with thyroid AID occurring in 41 % of patients [
48]. The incidence of thyroid AID appears to be greatest in the first 3 years following the initial treatment course [
42,
43,
48,
49] and responds to conventional therapies and management [
51]. Importantly, the risk of developing AID has been shown to be unaffected by the cumulative dose, dosage interval or dosage frequency [
48,
54].
A variety of other autoimmune conditions have also been documented [
49]. The most significant of which is ITP. CAMMS223 was suspended after three patients developed ITP with one patient dying of an intracerebral haemorrhage before a diagnosis could be made. During the suspension period, three further patients were diagnosed. In the subsequent phase III and open-label cohorts, the rate of ITP has been shown to be between 1 and 3 % [
7,
8,
48,
49,
54]. ITP associated with alemtuzumab therapy is thought to be characterised by delayed presentation after drug exposure, responsiveness to conventional therapies (corticosteroids, intravenous IgG, anti-Rh(D), platelet transfusion, rituximab) and prolonged remission [
55].
Although even less common, renal disease has been reported in four patients within the clinical trials—one case of anti-glomerular basement membrane (anti-GBM) disease and three cases of membranous nephropathy (one of which also had low-level anti-GBM antibodies). Each patient responded to prompt medical treatment with plasmapharesis, steroids and cyclophosphamide and diuretics for anti-GBM disease and membranous nephropathy, respectively [
56]. Three further patients treated outside of the clinical trials have also developed anti-GBM disease. Despite aggressive immunotherapy, two patients became dialysis-dependent and ultimately required renal transplantation, whilst the other although not requiring renal replacement has end-stage renal disease (A Coles, personal communication). Higher levels of Il-21 have been shown to be associated with developing autoimmune disease post-alemtuzumab [
57], but a test for this has so far not reached standard clinical practice [
58].
Infections
For a therapy whose main biological effect is lymphopaenia, rates of serious opportunistic infections might reasonably be expected to be high. In the CARE-MS studies, mild to moderate infections were seen in 67–77 % patients receiving the 12-mg dose of alemtuzumab (vs. 45 and 66 % of those receiving interferon β1a). Upper respiratory tract, urinary tract and herpes simplex and zoster virus infections remain the most commonly observed [
7,
8,
48,
49]. Infections attributable to immunosuppression in open-label studies include spirochaetal gingivitis, pyogenic granuloma and listeria meningitis. In addition, one further case of listeria meningitis was observed in the CAMMS223 study [
9] with two more cases also recently being reported [
59].
Although it is clear that vigilance needs to be maintained for opportunistic infections which can be severe, the relatively low frequency given the mode of action of alemtuzumab is intriguing and likely to represent a relative sparing of cells of the innate immune system in addition to haemopoetic precursor cells in the bone marrow and thymus. Subsequent lymphocyte repopulation following treatment also aids immune surveillance, and murine models have demonstrated that the function of the remaining lymphocytes following treatment is unimpeded [
60].
Malignancy
Although malignancies were seen in the clinical trials, the studies were not powered to detect changes between the groups. In CARE-MSI and CARE-MSII, malignancies were observed in 0.5 vs. 0 % and 0.6 vs. 1.5 % in the 12-mg alemtuzumab and interferon beta-1a-treated patients, respectively [
7,
8]. Of 1486 total patients treated with alemtuzumab in the clinical development programme, 29 (2 %) have developed neoplastic disease. Thyroid cancer accounted for six cases, which may relate to the increased thyroid surveillance undertaken in treated patients [
61,
62]. All were stage 1 papillary type carcinomas and discovered 10–41 months after last dose. One patient had a preexisting lymph node at baseline that was subsequently found to contain papillary thyroid cancer [
62]. Basal cell carcinoma (six patients), breast cancer (five) and malignant melanoma (four) occurred at a frequency >1 %.
Outside of clinical trials, one further case of papillary thyroid carcinoma [
63], one case of malignant melanoma [
64] and one case of Castleman’s disease (a prelymphomatous condition) [
38] have separately been reported. In addition, in our recent long-term follow-up cohort of 100 patients, 10 patients developed malignant/pre-malignant conditions. Three patients had monoclonal gammopathy of uncertain significance (MGUS), one of which was also diagnosed with a meningioma, two basal cell carcinomas and five patients had cervical dysplasia. Despite the theoretical risk of cervical malignancy following T lymphocyte depletion, there is currently no definitive available evidence that individual risk is increased compared to that of the background population. Despite this, it is recommended that female patients undergo annual cervical screening following treatment [
65]. It should be noted that the risk of malignancy following immunosuppression may take many years to manifest and therefore detailed post-marketing surveillance will be essential to better define long-term risks.
Pregnancy
Alemtuzumab may cross the placental barrier and therefore potentially produce harmful effects on the foetus. In animal studies, reproductive toxicity has been demonstrated although data in humans is lacking. It is unknown therefore whether alemtuzumab administration effects reproductive capacity or teratogenicity [
65]. Because of the theoretical risk of pregnancy adverse events, the parent company recommends effective contraception during and for 4 months following treatment. The most recently reported data of 1486 patients (64.8 % female) treated in the clinical trials reveals 179 pregnancies occurring in 131 patients. Of completed pregnancies, 104 (66 %) were live births. No congenital abnormalities have been observed in delivered infants. Excluding pregnancies with unknown outcomes, there were 36 (21 %) spontaneous abortions, 16 (9 %) elective abortions and 1 (0.6 %) stillbirth. Rates of spontaneous abortions are similar to those in the general population [
66]. Although CD52 is expressed in the epididymis, seminal vesicle, sperm and seminal fluid, suggesting a theoretical risk of reduced male fertility, a small study has not demonstrated any effects on sperm motility, morphology or count [
67].