Treatment and side effect management of CTLA-4 antibody therapy in metastatic melanoma


Dr. med. Katharina C. Kähler
Klinik für Dermatologie und
Schleswig-Holstein, Campus Kiel
Schittenhelmstraße 7
D-24105 Kiel
Tel.: +49-431-597-1512
Fax: +49-431-597-1853


Immune-modifying monoclonal antibodies may induce or enhance the natural immune response against tumor cells. The complex interaction between antigen-presenting cells and T lymphocytes as an immune response is strongly affected by anti-CD152 (CTLA-4)-antibodies. The cytotoxic T-lymphocyte (CTLA-4) receptor binds molecules of the B7-family which leads to a suppression of T cells. Specific CTLA-4 antibodies induce an unrestrained T-cell activation. Treatment with the CTLA-4 antibodies ipilimumab and tremelimumab has been investigated in metastatic melanoma only within clinical trials. Currently, the critical phase III trial on ipilimumab is in the final analysis process and expected to lead to approval.

CTLA-4 antibodies belong to the most promising new molecules for the treatment of advanced melanoma. During treatment with CTLA-4 antibodies, distinct adverse events may occur. Treating physicians must be familiar with their appropriate treatment and prophylaxis. The most frequently observed side effects are diseases such as an autoimmune colitis which is typically characterized by a mild to moderate, but occasionally also severe and persistent diarrhea. Other autoimmune-mediated side effects like hypophysitis, hepatitis, iridocyclitis or an exacerbation of lupus nephritis have been reported in the literature. Their early recognition and treatment are mandatory to reduce the risk of sequelae for CTLA-4-antibod-treated patients. Autoimmune-mediated side effects are reported to correlate positively with treatment response. We review the mechanisms of action, provide an update on clinical trials with the two CTLA-4-antibodies for metastatic melanoma, and present detailed recommendations for managing the side effects of these new agents.


CTLA-4 antibody therapy has been examined in clinical studies and compassionate use programs at various skin cancer centers in Germany. As with any other complex cancer therapy, the treating and consulting dermatologists should be familiar with the typical side effects of CTLA-4 antibody therapy and should be able to administer appropriate therapy. During the past two years, international cancer experts from various fields of medicine, as well as pharmaceutical companies (Medarex/Bristol-Myers Squibb GmbH, USA/Munich; Pfizer Pharma GmbH, Berlin), have been studying the use of CTLA-4 antibody therapies. Several treatment algorithms have been developed, in particular for the management of adverse effects in patients with autoimmune diseases. In Germany, other areas of medicine have not yet been involved in studies on the use of CTLA-4 antibody therapy in metastatic melanoma. It may be presumed, however, that in compassionate use programs, and if ipilimumab is approved, internist oncologists will also begin prescribing CTLA-4 antibodies for patients with metastatic melanoma.

CTLA-4 antibody: mechanism of action

Basically speaking, there are two different signals that are needed for a cellular immune response to occur. First, the T-cell receptor must identify a tumor antigen, and second there must be co-stimulation through binding of CD28 to the T cell and to B7 complex (Figure 1a, b). To regulate the immune response, the inhibitory cytotoxic T lymphocyte antigen 4 (CTLA-4; CD152) binds via a receptor to molecules in the B7 family. CTLA-4 is a type I membrane protein that is expressed on T cells and monocytes [1]. Specific CTLA-4 antibodies block the CTLA-4 receptor, reversing inhibition of the T-cell immune response. The result is a virtually uninhibited T-cell response.

Figure 1.

CTLA-4-antibodies: mode of action (a, b): APZ: antigen presenting cell, TZR: T-cell receptor, MHC: major histocompatibility complex.

Two different CTLA-4 antibodies have been developed simultaneously by two different pharmaceutical companies and are currently undergoing clinical testing in patients with metastatic melanoma. Ipilimumab (also known as MDX010; Medarex, USA; Bristol-Myers Squibb GmbH, Munich) is an IgG1 antibody with a plasma half-life of 12 to 14 days. Tremelimumab (CP-675,206) is an IgG2 antibody made by Pfizer (Karlsruhe). It has a plasma half-life of 22 days. Both companies used metastatic melanoma as the model for clinical development of the drug. A small number of patients with other solid tumors have been treated as well in previous clinical studies.

The enhanced T-cell response has been associated with an increase in autoimmune-related effects such as rashes, diarrhea, colitis, hepatitis, and hypophysitis [2].

Clinical data on tremelimumab

The response rate for tremelimumab has been reported in clinical studies as 7–15%[3]. Results from a large phase III study were presented at the ASCO Congress in 2008. This study, which included 665 patients, compared tremelimumab (15 mg/kg body weight every 12 weeks) with standard chemotherapy (DTIC) and temozolomide. An analysis of the main objective of the study (median survival) revealed a slight advantage of 10.7 months compared with 11.7 months in patients who were given tremeli-mumab versus those who were given conventional chemotherapy; the difference was not statistically significant [4]. Complete remission occurred in 1.5% of patients who were given tremelimumab, in the same range as for patients receiving standard chemotherapy (1.8%). In patients who were given tremelimumab, partial remission occurred in 7.6%; in patients who were given chemotherapy it was 8.3%[4]. At present there is no sign that the antibody will be tested in further phase III studies on melanoma patients.

Clinical data on ipilimumab

A phase II study on 217 previously treated patients with stage IV disease found that a dosage of 10 mg/kg of body weight was the most effective level with a still acceptable side effect profile [5]. In this study, patients were given the drug four times every three weeks for 12 weeks, and then every three months. In other phase II studies, with a total of 343 patients, the same dosage achieved control of the disease (disease control rate: proportion of patients with partial, complete remission of metastases or stabilization of disease) in 27–35% of patients [5–7]. A recent study reported a median overall survival rate of 10.2 to 19.3 months [5–7]. Two years after beginning treatment, the proportion of surviving patients was 30 to 42%[5–7]. The characteristic response to CTLA-4 antibodies varies from that seen with conventional chemotherapy. Studies have shown that in the initial weeks of therapy there may be apparent growth of the tumor. This may be partly explained by a drug-related inflammatory infiltrate due to activation of T lymphocytes. In addition, new metastases as well as target legions may respond during therapy. This demonstrates that the traditional RECIST criteria are insufficient for evaluating treatment success. A recent analysis of three phase II studies on ipilimumab has reported that, unlike in conventional chemotherapy, an increase in LDH values as an expression of a high tumor burden does not necessarily rule out the possibility of a successful response to ipilimumab [8]. A phase III placebo-controlled study is currently underway with 500 patients that will examine the use of DTIC with or without ipilimumab (10 mg/kg body weight). The study is currently in the follow-up observational phase. The final results, which should determine approval of ipilimumab as a first-line therapy for metastatic melanoma, are expected in late 2010.

At the 2010 American Society of Clinical Oncology (ASCO) meeting, promising results from a large, randomized phase III study on pre-treated melanoma patients were presented. This three-arm study compared the use of ipilimumab (3 mg/kg body weight) together with a peptide vaccine (gp 100) versus either one alone. Ipilimumab was given four times over 12 weeks at 3-week intervals. The results showed that administration of ipilimumab with or without the peptide vaccine led to significantly longer overall survival compared with patients who were given the vaccine only [9]. Median overall survival rates for patients who were given ipilimumab or ipilimumab and the peptide vaccine were 10.1 and 10.0 months. Patients who were given the vaccine alone had a median overall survival of 6.4 months. This was the first time that a positive effect of a drug has been shown on overall survival rates in patients with metastatic melanoma.

Promising results have also been reported from a preliminary prospective study on the effectiveness of ipilimumab (10 mg/kg of body weight) in patients with metastatic melanoma with cerebral involvement [10]. Ipilimumab was administered four times in intervals of three weeks. If stabilization or remission occurred, therapy could be continued in intervals of 12 weeks. In 51 patients who were not taking steroids at the time of treatment, control of the disease (disease control rate, DCR) occurred during week 12, that is, there was stabilization or partial or complete remission in 23.5% with regard to cerebral lesions and in 27.5% with regard to extracranial metastases [10]. The treatment response rate was significantly worse for the 21 patients who were taking steroid therapy. In week 12 the DCR was 9.5% for cerebral lesions and in 4.8% for extracranial metastases. These results show that ipili-mumab is also effective against cerebral metastases, but not in patients who are taking steroids. Furthermore, there is no significant difference between the response of intracerebral or extracerebral metastases. There was no evidence in this study of neurotoxic adverse effects related to ipilimumab.

Current use of ipilimumab

In regard to clinical studies, ipilimumab is currently being tested in the adjuvant EORTC 18071 study. The 950 participating patients with stage III disease are being given 10 mg/kg of body weight -ipilimumab or i.v. placebo four times every 3 weeks for 3 months, and then every 3 months as maintenance therapy. It should be noted that with micrometastasis to the lymph nodes (AJCC stage IIIA), the tumor diameter of the micrometastasis should be at least 1 mm. Patients who have already been treated with interferon or other immunomodulatory drugs are excluded as are those with lymph node metastasis with satellite lesions or in-transit metastases.

Side effect profile

The results of a dose escalation study showed that increasing dosage (0.3 vs. 3 vs. 10 mg/kg body weight) is associated with increased effectiveness, but also with an increase in immune-related side effects [5, 9]. At a dosage of 3 mg compared with 10 mg/kg body weight there was no significant difference, however, in the frequency of immune-related side effects. The occurrence of severe side effects was more common, however, especially those involving the gastrointestinal tract.

In a pooled analysis of 325 patients who were given ipilimumab 10 mg/kg body weight, drug-specific side effects were reported in 84.6%[12]. In 32.6% the effects were more severe (CTC grade 3–5). Immune-related side effects, which may be explained by a direct effect of the immune therapy on the target tissue, occur in 72.3% of patients. Severe side effects (grades 3–4) have been reported in 25.2% of patients, especially in four organ systems: the gastrointestinal tract (12%), liver (7%), skin (3%), and endocrine system (3%) (Table 1) [12]. These adverse effects have a characteristic pattern of onset (Figure 2). After an average of 3–4 weeks, adverse effects affecting the skin were seen, after 6–7 weeks side effects involving the gastrointestinal tract and liver appeared, and only after an average of 9.2 weeks were endocrine effects apparent [12]. Time to improvement of side effects for gastrointestinal symptoms was a median of 2 weeks, for liver toxicity 4 weeks, for the skin 6 weeks, and for endocrine disorders 20 weeks. Algorithms are available for the treatment of immune-related side effects. Management of symptoms depends on their severity (Figure 3a–d). These algorithms should provide even less experienced practitioners with some assurance in terms of treatment in actual clinical practice.

Table 1.  Toxicity profile of the CTLA-4-antibody ipilimumab (10 mg/kg KG).
Ipilimumab 10 mg/kg body weightAny gradeGrade 3–4
Skin, exanthema, pruritus47–68%0–4%
Gastrointestinal tract, diarrhea, colitis31–46%8–23%
Inflammatory hepatotoxicity3–9%3–7%
Figure 2.

Time course of typical adverse events during a therapy with CTLA-4-antibodies.

Figure 3.

Treatment algorithms for immune-related adverse events (a–d).

Organ-specific side effects of CTLA-4 antibody therapy

Side effects involving the skin and mucous membranes

In 47–68% of patients a clinically apparent maculopapular rash has been reported, generally appearing after 3.6 weeks (Figure 4) [5–7]. In only 4% of patients were the symptoms severe [5–7]. Histo-logical analyses have revealed a perivascu-lar lymphocytic infiltrate extending into the epidermis. Immunohistochemistry has shown CD4+- and melan-A-specific CD8+ T lymphocytes near apoptotic melanocytes [10]. The appearance of a rash could be explained by an ipilimumab-stimulated immune response to melanocytes [13]. This is supported by the appearance of vitiligo in 11% of patients in a phase II study on ipilimumab [7].

Figure 4.

Maculopapular rash during a treatment with the CTLA-4-antibody tremelimumab.

The rash, which is caused by CTLA-4 antibodies, as well as the pruritus often associated with it, generally respond well to therapy. There is usually not need to stop therapy or to reduce the dosage. Topical corticosteroids such as betamethasone 0.1% cream or, later, urea-based topical therapies in combination with antipru-ritic agents such as polidocanol (e.g., Op-tiderm® fatty cream) are recommended.


One study on ipilimumab (10 mg/kg body weight) reported that 44%[12] of patients experienced diarrhea as a side effect of the drug. About 18% of patients had severe diarrhea (grade 3–4) (Table 2) [12]. This is clinically relevant given related weakness, an electrolyte imbalance, and prolonged weight loss. Over time, colitis can occur. In the worse case bowel perforation can result – in as many as 0.5% of patients – sometimes even requiring a colectomy. Earlier studies reported that intervention was required in 0.9% of patients; the decline in this rate was due to introduction and consistent use of algorithms for side effect management [14]. Unlike in chronic inflammatory bowel diseases (IBD) such as ulcerative colitis, in immune-related colitis the descending colon is more often affected than the sigmoid colon or rectum [7]. Histology shows three pathological patterns: in 46% there is a neutrophilic infiltrate, in 15% a lymphocytic infiltrate, and in 38% a mixed infiltrate [2].

Table 2.  CTCAE (Common Terminology Criteria for Adverse Events, v.3.0) grading of diarrhea.
Diarrhea: Categorization based on CTCAE
Grade 1Grade 2Grade 3Grade 4Grade 5
Increased stool frequency of less than 4 more compared to normal4–6 stools more than normal; i.v. fluid substitution for less than 24 hoursMore than 7 stools more than normal i.v.-fluid substitution 24 hours or longerLife-threateningFatal

In patients with mild diarrhea treatment is symptomatic and includes loperamide and electrolyte substitution. If symptoms are prolonged or severe, other causes such as a gastrointestinal infection or manifestation of IBD should be excluded using a stool test (calprotectin and leukocytes in stool, stool structure). The therapy of choice is budesonide which targets the colon (9 mg orally daily, e.g., Budenofalk, alternatively I mg prednisone/kg of body weight). Endoscopy to confirm or exclude colitis is also indicated. While the presence of colitis does not alter the treatment approach, it does present a risk factor for bowel perforation.

In patients with grade 3 or grade 4 diarrhea (Table 1), high-dose steroids such as methyl prednisolone 2 mg/kg of body weight 1–2 X daily or dexametha-sone in an equivalent dosage (0.33 mg/kg of body weight) are indicated. Symptoms generally improve dramatically after 1–2 weeks. It is important to slowly taper corticosteroid therapy over a course of at least four weeks to ensure that that symptoms have completely resolved. In rare instances, there may be no response to high-dose corticosteroid therapy; treatment may be attempted with infliximab (Remicade) similar to the procedure for IBD [15].

Prevention of diarrhea and colitis

A double-blind, placebo-controlled phase II study on 115 patients has addressed the question of prevention of diarrhea and colitis with simultaneous administration of 9 mg of budesonide daily during ipilimumab therapy (10 mg/kg of body weight). Patients who were given budesonide therapy had diarrhea or colitis just as often as patients who were given placebo [7]. Hence, based on these data budesonide cannot be recommended to prevent diarrhea. Another interesting point was that the administration of corticosteroids in this study was not related at all to a decrease in effectiveness of ipilimumab in terms of a negative impact on adverse immunomodulatory effects [7].

Inflammatory hepatotoxicity

In various studies on CTLA4 antibodies, 3–9% of patients had immune-related liver toxicity [5–7]. This was apparent as an asymptomatic increase in transaminase or as immune-related hepatitis. Progressive metastasis to the liver should be ruled out as well as other causes of liver damage such as viral hepatitis or a toxic drug reaction. The algorithm for management of inflammatory liver toxicity contains information on further autoantibody testing (antimitochondrial antibodies [AMA], liver-pancreas-specific antigen [LP], liver-kidney microsomes [LKM1]), and dose administration as well as a recommendation for the use of high-dose corticosteroids such as methyl prednisolone 2 mg/kg of body weight 1–2 X daily (Figure 3). If the condition persists, immunosuppressant therapy with my-cophenolate mofetil (1 g i.v. or 1.5 g orally 2 X daily) may also be considered.


Hypophysitis is a rare immune-related side effect that has been reported in 1–6% of patients treated with ipilimumab 10 mg/kg of body weight [5–7]. Clinical symptoms include headache, nausea, dizziness, vision disturbances (e.g., double vision), and weakness. Symptoms usually occur during later stages of therapy, on average after 6 weeks. The most important differential diagnosis is cerebral metastasis. Imaging studies are crucial as they allow precise evaluation of the region around the pituitary gland. Ideally, an MRI of the brain should be performed. Before initiating treatment, blood tests for endocrine status should also be ordered (serum cortisol levels, ACTH, fT3, fT4, TSH, prolactin, testosterone, FSH, LH). Hypophysitis is visible on imaging studies as an enlargement of the hypophysis with thickening of the hypophyseal stalk (Figure 5). A comparison with a previous MRI of the brain before beginning therapy can help facilitate radiological evaluation. Laboratory chemical tests typically show a reduction in the laboratory parameters mentioned above.

Figure 5.

Hypophysitis with thickened pituitary gland and stalk.

The therapy of choice is high-dose steroid therapy, e.g., with dexametha-sone at a dosage of 4 mg every 6 hours. Improvement of clinical symptoms is generally seen after only a few days, and decreased swelling of the hypophysis is also evident at MRI [16]. Given the complexity of endocrine disorders, cooperation and consultation with an endocrinologist is advisable throughout the diagnosis and treatment process.

Hypophyseal functioning may be impaired for a longer period of time following high-dose corticosteroid therapy. Previous studies have shown that the time needed for symptoms to resolve and the necessary duration of substitution therapy with physiological hydrocortisone dosages (on average 20 weeks) is considerably longer (Figure 2) [12]. Long-term substitution therapy with physiological dosages of hydrocortisone may be required.


Severe ophthalmological side effects have been reported in fewer than 1% of treated patients. Previous studies have reported that uveitis and episcleritis occurred in patients who also had diarrhea or colitis during CTLA-4 antibody therapy [17]. Adverse effects appeared on average 2 months after beginning therapy and resolved within a week regardless of therapy.

Neurological side effects

Less than 1% of patients taking CTLA-4 antibody experience side effects such as Guillain-Barré syndrome, sensory or motor neuropathy, or myas-thenia gravis. If neuromotor symptoms persist for more than 5 days, the patient should be examined for a neurological disorder. Corticosteroids may be given to treat symptoms. Depending on the severity of the side effects, further steps may be undertaken based on the algorithm for neurotoxicity (Figure 3).

Correlation between immune-related adverse effects and treatment success

Smaller studies have indicated a possible relationship between immune-related side effects and successful therapy with CTLA-4 antibodies [18, 19]. In 56 patients with stage IV disease who were given ipilimumab, 36% with high-grade (grades 3 or 4) side effects exhibited a clinical response (partial or complete remission) compared with only 5% of patients who did not experience any immune-related side effects [18]. A recently published pooled analysis of three phase II studies on ipilimumab with a total of 343 patients with no or only mild immune-related side effects (grade 1 CTC) reported a disease control rate of 20–24% during ipilimumab therapy. In patients with immune-related side effects with at least grade 2 CTC, the disease control rate was 34–43%. This positive trend was not statistically significant, however. In terms of overall survival, there was also a positive trend in favor of patients who had immune-related side effects in the first 12 weeks. It was also not statistically significant. Average survival was 14.8 months compared with 8.21 months in patients who did not have any immune-related side effects [20]. The results of a large phase III study on ipilimumab are still needed before any conclusions may be drawn.

A pooled analysis of the studies A3671008 and A3671009 with 571 patients who were given tremelimumab 15 mg/kg of body weight every three months reported that 50% of patients with a clinical response (response rate: 8.9%) had had an exanthem in the first cycle. Patients in whom therapy was unsuccessful were less often affected. Yet given that patients who did not have an exanthem also responded to therapy in the first cycle, an exanthem alone cannot be viewed as a marker to help decide whether to continue or stop therapy [21].


CTLA-4 antibodies are presently the most promising immuno-stimulatory drug available for the treatment of metastatic melanoma. The range of potential side effects is well known from detailed studies on the adverse effects of CTLA-4 antibodies, and appropriate treatment algorithms have been developed by experts from various disciplines. In most cases, prompt and consistent administration of corticosteroids can ensure control of immune-related adverse effects and alleviation of symptoms. Not only dermato-oncologists, but also internist oncologists may be confronted with the possibility of administering CTLA-4 antibody therapy. The primary aim is to ensure the safety of melanoma patients with what is now an experimental therapy which may in the future become routine.

Conflict of interest

Dr. Hauschild received study funding as well as compensation for advisory work and lectures from Bristol-Myers Squibb and Pfizer. Dr. K.C. Kähler received compensation for lectures from Bristol-Myers Squibb.