Pruritus in Patients Treated with Targeted Cancer Therapies: Systematic Review and Meta-Analysis

Abstract

Background

Pruritus has been anecdotally described in association with targeted cancer therapies. The risk of pruritus has not been systematically ascertained.

Objective

A systematic review and meta-analysis of the literature was conducted for axitinib, cetuximab, dasatinib, erlotinib, everolimus, gefitinib, imatinib, ipilimumab, lapatinib, nilotinib, panitumumab, pazopanib, rituximab, sorafenib, temsirolimus, tositumomab, vandetanib, and vemurafenib.

Methods

Databases from PubMed, Web of Science (01/1998–07/2012), and American Society of Clinical Oncology abstracts (2004–2012) were searched. Incidence and risk (RR) of pruritus were calculated using random or fixed effects model.

Results

The incidences of all-grade and high-grade pruritus were 17.4% (95% confidence interval (CI): 16.0%−19.0%) and 1.4% (95% CI: 1.2%−1.6%), respectively. There was an increased risk of all-grade pruritus (RR=2.90 (95% CI: 1.76–4.77, p<0.001)); and variation among different drugs (P<0.001).

Limitations

The reporting of pruritus may vary, resulting from concomitant medications, comorbidities, and underlying malignancies. We found a higher incidence of pruritus in patients with solid tumors, concordant with those targeted therapies with the highest pruritus incidences.

Conclusion

There is a significant risk of developing pruritus in patients receiving targeted therapies. In order to prevent suboptimal dosing and decreased quality of life, patients should be counseled and treated against this untoward symptom.

Introduction

Novel agents targeting specific cancer pathways or proteins have been shown to significantly increase the survival of patients with various malignancies¹. The increased lifespan alongside the expanded use has led to a variety of therapy-associated adverse events (AE). These novel agents are associated with lower systemic toxicity than conventional chemotherapy, yet dermatologic events may affect the majority of treated patients².

Dermatologic toxicities to targeted therapies manifest in cosmetically sensitive areas, are associated with symptoms, and can interfere with activities of daily living³. This results in a negative impact on quality of life, which may lead the physician to lower the dose⁴. Pruritus is a common but infrequently discussed AE; a survey of 379 cancer survivors reported 36% experienced pruritus during treatment, with 44% indicating a negative impact on quality of life (QoL)⁵. An analysis of anticancer therapies reported rash and pruritus to have the greatest negative impact on QoL among dermatologic AE including alopecia, nail changes, hand-foot syndrome, mucosal changes, and fissures⁶. Knowledge of these effects and which agents have a higher incidence of pruritus is important for patient counseling and directing supportive care efforts.

Whereas the acneiform (papulopustular) rash to EGFR inhibitors (EGFRIs) and hand-foot syndrome provoked by multikinase inhibitors have been extensively described, the overall risk of developing pruritus for patients receiving targeted therapies has not been systematically ascertained. We conducted a systematic review and meta-analysis of the literature to identify published clinical trials of targeted therapies to determine the incidence and risk of pruritus.

Methods

Data Source

The PubMed database was searched from January 1998 to July 2012 using the keywords of the name of the targeted agent (e.g. ‘axitinib’) and ‘clinical trials,’ and was limited to the English language and human studies. In addition, we reviewed abstracts and virtual meeting presentations that contained ‘axitinib’ presented at the American Society of Clinical Oncology (ASCO) annual meetings from 2004 through 2012. An independent search using the Web of Science database (a product developed by the Institute for Scientific Information) was also conducted to ensure that there were no additional studies. Only full publications from the Web of Science were added to the study selection. We reviewed each publication and used only complete or the most recent data reports when duplicate publications of the trial were identified. Information regarding patient characteristics, study design, treatment regimen, study results, and safety and tolerability were extracted from the publications. This systematic search was performed for axitinib, cetuximab, dasatinib, erlotinib, everolimus, gefitinib, imatinib, ipilimumab, lapatinib, nilotinib, panitumumab, pazopanib, rituximab, sorafenib, temsirolimus, tositumomab, vandetanib, and vemurafenib.

Study Selection

Each targeted therapy has been approved for treatment of malignancies in patients at a specific dose. It is therefore clinically significant to determine the incidence of pruritus at this dosing level. We excluded trials that treated at unapproved doses, including phase I studies. Since chemotherapy and radiation may cause pruritus, we excluded trials that combined targeted agents with chemotherapeutic agents and/or radiotherapy. Trials that met the following criteria were included for further analysis: (1) prospective phase II and phase III clinical trials in cancer patients; (2) assignment of participants to the treatment with and (3) clear data available for the incidence of pruritus.

Clinical End Points

The clinical end point of pruritus was extracted from the safety profile in each trial. Pruritus was recorded according to the National Cancer Institute Common Toxicity Criteria version 2 or Common Terminology Criteria for Adverse Events (CTCAE) version 3. We included the incidence of all patients with pruritus grade 1 and above. The grading of pruritus in version 2.0 is described with three grades, as follows: grade 1, mild or localized, relieved spontaneously or by local measures; grade 2, intense or widespread, relieved spontaneously or by systemic measures; grade 3, intense or widespread and poorly controlled despite treatment. In version 3.0, the descriptions of these three grades are updated to: grade 1, mild or localized; grade 2, intense or widespread; grade 3, intense or widespread and interfering with activities of daily living (ADL). Version 4.0 expands further upon the descriptions in version 3.0; however, none of the studies reviewed used version 4.0.

Statistical Analysis

All statistical analysis was performed using version 2 of the Comprehensive MetaAnalysis program (Biostat, Englewood, New Jersey, USA). The number of patients with all-grade and high-grade pruritus were extracted from the clinical trial data. For each study, the proportion of patients with pruritus was calculated and the 95% exact confidence interval (CI) was derived. For studies with a placebo-only control arm, the relative risk of rash among patients was also calculated.

For meta-analysis, both the fixed-effects model (weighted with inverse variance) and the random-effects model were considered ⁷. For each meta-analysis, the Cochran’s Q statistic was first calculated to assess the heterogeneity of the included trials. For p-value of Cochran’s Q statistic less than 0.1, the assumption of homogeneity was deemed invalid ⁸, and random-effects model was reported after exploring the causes of heterogeneity. Otherwise, both the fixed-effects model and the random-effects model results were reported. A two-tailed p-value of less than 0.05 was judged as statistically significant.

Results

Search Results

Our search yielded a total of 5065 potential articles on targeted therapies in the literature (see Figure 1 for the overall selection process). A total of 144 clinical trials were included for this analysis, including 116 phase II and 28 phase III trials^9–148.

Figure 1.

Selection process for studies included in the meta-analysis.

Patients

A total of 20,532 (treated: 17,375; controls: 3157) patients from 144 clinical trials were included for analysis (see Table I for number of patients receiving each targeted therapy). Of these studies, 114 were solid organ malignancies and 30 were hematologic.

Table I.

Incidences of all-grade and high-grade pruritus associated with targeted therapies

Targeted therapy	Number of Studies*	Number of Patients	All-grade (95% CI)	High-grade (95% CI)
mTOR inhibitors	9	672	23.8% (15.0–35.7)	1.2% (0.5–2.9)
Everolimus	5	486	14.3% (11.5–17.7)	1.3% (0.5–3.7)
Temsirolimus	4	186	37.7% (20.9–58.0)	1.0% (0.2–4.8)
Bcr-Abl inhibitors	26	5036	12.8% (10.4–15.7)	0.9% (0.6–1.3)
Dasatinib	2	196	9.7% (6.3–14.7)	0.8% (0.2–4.0)
Imatinib	14	2351	10.2% (7.4–13.9)	0.8% (0.5–1.3)
Nilotinib	10	2489	17.1% (13.2–21.8)	1.0% (0.7–1.6)
Raf kinase inhibitors	19	1944	18.3% (12.9–25.2)	1.3% (0.8–2.1)
Sorafenib	16	1448	18.2% (12.3–26.1)	1.0% (0.5–1.9)
Vemurafenib	3	496	18.5% (6.3–43.4)	1.7% (0.9–3.5)
VEGFR inhibitors	4	158	3.0% (1.1–7.8)	1.5% (0.4–5.7)
Axitinib Pazopanib	1	12	8.3% (1.2–41.3)	3.9% (0.2–40.3)
Pazopanib	3	146	2.2% (0.7–6.6)	1.1% (0.2–5.1)
EGFR inhibitors	57	6809	22.7% (17.8–28.6)	1.8% (1.5–2.3)
Cetuximab	6	217	18.2% (10.8–28.8)	2.1% (0.8–5.3)
Erlotinib	24	2742	20.8% (14.3–29.3)	2.3% (1.5–3.4)
Gefitinib	22	3002	21.0% (15.3–28.3)	1.0% (0.6–1.5)
Panitumumab	5	848	54.9% (46.9–62.7)	2.6% (1.7–4.0)
EGFR-HER2 inhibitors	9	501	14.6% (9.9–21.0)	1.0% (0.4–2.6)
Lapatinib
EGFR-VEGFR inhibitor	3	1261	9.1% (5.0–16.2)	0.5% (0.2–1.5)
Vandetanib
Monoclonal antibodies to CD20	13	658	11.3% (8.8–14.3)	1.2% (0.5–2.7)
Rituximab	10	519	10.2% (7.3–14.1)	1.2% (0.5–2.9)
Tositumomab	3	139	13.7% (8.9–20.5)	0.8% (0.05–11.8)
Monoclonal antibody to CTLA4	4	336	30.7% (15.9–51.0)	1.0% (0.3–3.9)
Ipilimumab
Overall	144	17,375	17.4% (16.0–19.0)	1.4% (1.2–1.6)

^*References: everolimus^9–13, temsirolimus^14–17, dasatinib⁴⁶,⁴⁷, imantinib⁵⁵,^58–70, nilotinib^48–57, sorafenib^18–33, vemurafenib^37–39, axitinib⁹³, pazopanib^34–36, cetuximab^40–45, erlotinib⁷²,^111–133, gefitinib^71–92, panitumumab^134–138, lapatinib³⁶,^139–146, vandetanib¹¹⁵,¹⁴⁷,¹⁴⁸, rituximab^94–103, tositumomab^108–110, ipilimumab^104–107

Incidence of all-grade pruritus

Data for all-grade pruritus was available for analysis from a total of 17,368 patients treated with targeted therapies as a single agent from 141 clinical trials. Among these studies, the incidence of all-grade pruritus ranged between 3.0% (95% CI: 1.1%−7.8%) and 30.7% (95% CI: 15.9%−51.0%), with the lowest incidence in patients treated with VEGFR inhibitors (axitinib and pazopanib) and the highest in patients treated with CTLA4 inhibitor ipilimumab. The incidence of pruritus was determined to be 19.2% (95% CI: 16.2%−22.6%) in solid organ malignancies and 13.0% (95% CI: 10.7%−15.7%) in hematologic malignancies (p=0.003). Meta-analysis (heterogeneity test: Q=45.308, I2=80.136, P<0.001) revealed that the overall summary incidence of all-grade pruritus was 17.4% (95% CI: 16.0%−19.0%), according to a random-effects model (Table I).

Incidence of high-grade pruritus

High-grade (grade 3) pruritus is considered severe and can lead to dose reduction or treatment interruption. Data for high-grade pruritus was available for analysis from a total of 15,927 patients treated with targeted therapies as a single agent from 132 clinical trials. Among these studies, the incidence of high-grade pruritus ranged between 0.5% (95% CI: 0.2%−1.5%) and 1.8% (95% CI: 1.5%−2.3%), with the lowest incidence in patients treated with EGFR-VEGFR inhibitor, vandetanib, and the highest in patients treated with EGFRIs (gefitinib, cetuximab, panitumumab, and erlotinib). The overall incidence of high-grade pruritus in patients treated with CTLA4 inhibitor, ipilimumab, was 1.0% (95% CI: 0.3%−3.9%). The overall incidence of high-grade pruritus for all patients was 1.4% (95% CI: 1.2%−1.6%) (Table I).

Incidence of pruritus in patients with different EGFRIs

We investigated whether the specific EGFRI used as therapy has an impact on the incidence of pruritus. The incidences of all-grade pruritus were determined among cetuximab (n=217), erlotinib (n=2717), gefitinib (n=3002), and panitumumab (n=848), and ranged from 18.2% (95% CI: 10.8%−28.8%) to 54.9% (95% CI: 46.9%−62.7%), with the lowest incidence in cetuximab and the highest in panitumumab. The overall incidence of high-grade pruritus was determined among cetuximab (n=217), erlotinib (n=2263), gefitinib (n=3002), and panitumumab (n=842). There was a significant variation among these EGFRIs (P<0.001). The incidences of high-grade pruritus ranged from 1.0% (95% CI: 0.6%−1.5%) and 2.6% (95%CI: 1.7%−4.0%), with the lowest incidence in patients treated with gefitinib and the highest in patients treated with panitumumab (Table I).

Relative risk (RR) of developing pruritus

A meta-analysis of RR for all-grade pruritus associated with targeted agents versus controls was performed on eleven randomized control trials, in which the incidence of pruritus was reported for 2261 patients receiving best supportive care (BSC) alone. According to the random-effects model, the overall RR for all-grade pruritus was calculated to be 2.90 (95% CI: 1.76–4.77, p<0.001) (Figure 2A). There was significant variation among different classes of targeted therapies (P<0.001) and different EGFRIs (P<0.001). The RR for all-grade pruritus associated with specific EGFRIs was found to be 1.77 (95% CI: 1.23–2.56, p<0.001) for gefitinib and 26.57 (95% CI: 11.08–63.70, p<0.001) for panitumumab. The summary RR for high-grade pruritus associated with targeted agents versus controls was performed and found to be 2.13 (95% CI: 0.61–7.48, p=0.452), according to the fixed-effects model (Figure 2B).

Figure 2.

Relative risk of all-grade (A) and high-grade (B) pruritus associated with targeted therapies versus controls. RR was calculated using a random-effects model. The relative risk for each study is displayed numerically on the left and graphically on the right. Under study name, the first author’s name and publication year were used to represent each trial. The size of the squares is directly proportional to the amount of information available. For individual trials: filled-in square, incidence; lines, 95% confidence interval; diamond plot, overall results of the included trials.

Discussion

Our study has demonstrated that patients treated with targeted therapies have a significantly increased risk of developing pruritus. The overall incidence of all-grade pruritus is 17.4% (95% CI: 16.0%−19.0%) with a RR of 2.90 (95% CI: 1.76–4.77, p<0.001). Therefore, it is important for physicians and patients to recognize the risk in order to monitor and treat the toxicity adequately.

The pathophysiology of pruritus remains unclear. Our meta-analysis determined the incidence of all-grade pruritus from EGFRIs to be 22.7% (95% CI: 17.8%−28.6%). These targeted agents inhibit the EGFR of basal keratinocytes, perturbing normal epidermal physiology^{149, 163}. During the first month of treatment with EGFRIs—cetuximab, erlotinib, or panitumumab—xerosis appears in 20% to 50% of patients^150–153. Amongst individual EGFRIs and individual targeted agents included in this study, the highest overall incidence of pruritus of 54.9% (95% CI: 46.9%−62.7%) was seen with panitumumab, when compared to patients treated with cetuximab, erlotinib, or gefitinib (incidences were 18.2% (95% CI: 10.8%−28.8%), 20.8% (95% CI: 14.3%−29.3%), and 21.0% (95% CI: 15.3%−28.3%), respectively). These summary incidences are lower than in panitumumab, but are higher than the incidences in patients treated with dual inhibitors, such as 14.6% (95% CI: 9.9%−21.0%) in EGFR-HER2 inhibitor lapatinib and 9.1% (95% CI: 5.0%−16.2%) in EGFR-VEGFR inhibitor vandetanib.

With ipilimumab, pruritus appears to be a direct result of CTLA4 inhibition and subsequent enhanced immune system activation¹⁵⁴. The incidence of all-grade pruritus in patients treated with ipilimumab was 30.7% (95% CI: 25.9%−51.0%). The skin is an immunologic organ, and dermatologic disorders may be caused by either exacerbation or reduction of cutaneous immune activity¹⁵⁵. Ipilimumab abrogates CTLA4-induced inhibition of T cells, and results in increased activated T-cell function and thus enhances the immune response¹⁰⁶. Cutaneous immune-related adverse events such as pruritus may be directly caused by this increased activation of the immune system. The incidence of pruritus with other monoclonal antibodies included in this study, rituximab and tositumomab, was found to be much lower than with ipilimumab (11.3%), likely due to their targeting of CD20 bearing cells.

Of patients treated with VEGFR inhibitors, axitinib and pazopanib had the lowest incidence of all-grade pruritus (3.0%), when compared to sorafenib. The incidences of pruritus among mTOR inhibitors (everolimus and temsirolimus), inhibitors of Bcr-Abl (dasatinib, imatinib, and nilotinib), and inhibitors of Raf (sorafenib and vemurafenib) were 23.8%, 12.8% and 18.3%, respectively.

Possible pathogenesis of pruritus may involve unmyelinated C fibers and neurotransmitters or receptor activation, such as serotonin, neurokinin 1 receptor, opioid receptors, and gamma-aminobutyric acid^{156, 157}. In some cases, pruritus may be indirectly caused by targeted therapies. Indeed, xerosis is cited as the most frequent cause of pruritus in oncology, and pruritus also accompanies papulopustular rash¹⁵⁶. Papulopustular (acneiform) rash is a common skin toxicity in patients treated with targeted therapies, and is the most common dermatologic AE that occurs in patients treated with EGFRIs^{156, 158}. Recent research has proposed that patients with EGFRI-induced rash and pruritus may be associated with an increased number of dermal mast cells surrounding adnexal structures. A continued increase in mediators released from these cells may activate sensory nerves, ultimately resulting in itch, both of which have been associated with the acneiform rash in 62% of cases^{159, 160}. Classically, mast cell mediators such as histamine are associated with non-allergic urticaria¹⁶¹.

Currently, management options for pruritus in cancer patients require a tailored approach, which includes patient education, topical and systemic treatments. Patient education is key, as severe itching leads to scratching, causing secondary skin changes such as excoriations and infections (Fig. 3). Patients should be informed of how to break the “itch-scratch” cycle, for example by keeping fingernails short, wearing light clothing, using a humidifier, restricting bath and shower time and using lukewarm water, and avoiding cleansers with a high pH or containing alcohol¹⁶². Regular moisturizing and use of emollients are central to the management of pruritus, especially when associated with xerosis. Treatments for mild to moderate pruritus include topical corticosteroids, anesthetics (ie. lidocaine, prilocaine), capsaicin, salicylic acid, and menthol and for severe pruritus, oral agents such as antihistamines, anticonvulsants, antidepressants, mu antagonists, aprepitant, and phototherapy. These therapies have shown benefit in uncontrolled studies^{162, 163}.

Figure 3.

Pruritus-induced excoriations associated with use of targeted therapies. Left panel: Excoriations induced by pruritus affecting anterior chest, abdomen, and upper legs; right panel: ventral arm and wrist, showing excoriations induced from pruritus.

Pruritus is common but often overlooked in cancer patients, as evidenced by it being reported in 2.8% of analyzed trials, likely due to seemingly more life-threatening side effects often taking precedence. Pruritus can impact on quality of life, often negatively affecting sleep, attention, and sexual function¹⁶². Preference-based quality of life measures have demonstrated that patients afflicted with pruritus would be willing to reduce their life expectancy by 13% to not have pruritus¹⁶⁴. Although it was only reported to occur in 3 of the analyzed trials, patients may withdraw from treatment due to intolerable pruritus^165–167.

There are several limitations to our meta-analysis. First, there is a large amount of heterogeneity amongst institutions in the assessment and reporting of pruritus. During our selection process, numerous studies either did not specify amongst dermatologic adverse events or would list pruritus in combination with rash and xerosis. Additionally, we came across a study that specified pruritus to be localized to the scalp of a patient with metastatic lymphoma¹⁸. Another study described that the only side effect seen in a patient treated with rituximab was “persistent itching”¹⁶⁸. This wide variability and inconsistency in reporting pruritus may affect our ability to accurately assess the safety profiles of targeted agents, estimate the risk of pruritus development, and assess patient response to therapies. The development of an improved system for classification of dermatologic adverse events, with greater reflection upon their complexity and variability, may improve upon these areas. Another limitation of this study involves the possibility of sampling bias, as only 144 of 5065 studies initially found were included in our analysis, based on our selection criteria (see Study Selection in Methods for details). Due to the aforementioned reasons, it is possible that the clinical studies used in this meta-analysis over- or underestimated the incidence of pruritus associated with the use of targeted therapies. Furthermore, the results of our analysis may not apply to patients in the real-world setting, where patient and clinician AE reporting may differ from that in clinical trials.

Further studies are needed to investigate other factors that contribute to pruritus in this patient population. Concomitant medications and comorbid conditions, such as hepatic and renal impairment, and cancer itself, are common in these patients and can result in pruritus¹⁶². The type of cancer is a potential variable that may contribute to pruritus; hematologic malignancies per se, have been associated with paraneoplastic pruritus. However, our analysis found a higher incidence of pruritus in solid organ malignancies (19.2%) than hematologic tumors (13.0%) (p=0.003). This is consistent with the association of pruritus with drug rather than tumor type. Whereas the development of acneiform rash to EGFR inhibitors has been correlated with response, this observation has not been examined with pruritus, however the findings described herein warrant additional analyses to define whether pruritus correlates with clinical outcome. Additionally, analysis of other factors that could give clinical significance of pruritus such as allergic reactions, infections, and environmental factors (hot or cold weather, low humidity, bathing too frequently) should be considered.

Conclusion

Our results demonstrate that targeted cancer therapies are associated with a significant risk of developing pruritus. In order to prevent suboptimal dosing and reduction in patients’ quality of life, further research is needed to improve upon the current understanding of the pathogenesis of pruritus, risk factors, and management strategies. Prophylactic treatment, early detection and intervention, and close monitoring of this untoward event are critical to ensure patient adherence and maximize clinical benefit from optimal dosing.

Capsule.

• Pruritus is frequent in patients receiving targeted anticancer therapies, but its incidence is unknown. Use of these novel agents may be hindered due to widespread dermatologic adverse events, such as pruritus.

• This manuscript assesses the incidence and relative risk of developing pruritus among patients treated with targeted anticancer therapies.

• Patients with pruritus should be counseled and symptomatically treated in order to prevent suboptimal dosing and significant reduction in quality of life.

Abbreviations and Acronyms List

CI

confidence interval

CD20

B-lymphocyte antigen CD20

CTLA4

T-lymphocyte antigen 4

CTCAE

Common Terminology Criteria for Adverse Events

EGFR

epidermal growth factor receptor

EGFRIs

epidermal growth factor receptor inhibitors

HER2

human epidermal growth factor receptor 2

Mtor

mammalian target of rapamycin

RR

relative risk

VEGFR

vascular endothelial growth factor receptor