A systematic review on the management of pruritus in patients with cutaneous T-cell lymphoma

Introduction: Cutaneous T-cell lymphomas (CTCLs) represent a rare group of primary cutaneous lymphomas. Pruritus is common in patients with CTCL and is severe and intractable in the subtypes Sézary syndrome (SS) and folliculotropic mycosis fungoides (MF). Materials and methods: We conducted a systematic review on interventions demonstrating efficacy in reducing pruritus in patients with CTCL. The primary aim of our study was to identify disease-directed and itch-directed therapies effective in reducing CTCL-associated pruritus. Our secondary aim was to outline various tools used to quantify itch in clinical studies. Results: Our study identified multiple disease-directed therapies effective in reducing CTCL-associated pruritus. Most evidence supported the use of histone deacetylase inhibitors. For the agents romidepsin, vorinostat, and quisinostat, reduction in pruritus was achieved in both disease responders and nonresponders. Various therapies were effective in managing pruritus associated with SS and folliculotropic MF. Vorinostat is effective in reducing pruritus in patients with SS. Extracorporeal photophoresis, total skin electron beam therapy, and romidepsin are effective in reducing pruritus in patients with folliculotropic MF. The antiemetic aprepitant is an effective targeted treatment of CTCL-associated pruritus. Aprepitant demonstrated efficacy in reducing pruritus in patients with all stages of MF, including patients with SS. Lignocaine administered via continuous subcutaneous infusion is effective in reducing pruritus in patients with advanced-stage MF, including patients with SS. The most frequently used tools to quantify itch were the Visual Analogue Scale and Numerical Rating Scale. Definitions of a significant reduction in pruritus were extremely varied between studies. Discussion: To our knowledge, this is the first systematic review specifically addressing the management of itch in patients with CTCL. Patients with all stages of CTCL were represented across included studies, including patients with folliculotropic MF and SS. A wide range of treatment options were identified, including options appropriate for patients with end-stage disease.


Introduction
Cutaneous T-cell lymphomas (CTCLs) represent a rare group of primary cutaneous lymphomas. Multiple subtypes of CTCL are recognized, of which mycosis fungoides (MF) and Sézary syndrome (SS) are the most common. In MF, neoplastic cells are derived from noncirculating skin resident effector memory T cells, whereas in SS, neoplastic T cells have a central memory T-cell phenotype. CTCL can range from an indolent to highly aggressive disease, and management should be tailored according to the disease stage [1] .
Early-stage disease is usually managed with skin-directed therapy, such as topical corticosteroids, topical chemotherapy, topical retinoids, phototherapy, and radiotherapy. Advanced-stage or transformed disease warrants systemic treatment, which may include systemic retinoids, interferon, histone deacetylase (HiDAC) inhibitors, systemic chemotherapy, and denileukin diftitox. Extracorporeal photophoresis and various investigational therapies also have a role in the treatment of advanced-stage disease [1] .
Patients with CTCL have a considerable burden of physical symptoms that affect their physical, social, and emotional wellbeing. Pruritus is the most commonly reported physical symptom in patients with CTCL, affecting up to 88% of patients [2,3] . Itch has a major negative impact on quality of life, interfering with daily activities, sleep, and mood [4] . The severity of CTCL-associated itch generally increases with disease progression. Pruritus is also more common in certain subtypes of CTCL, namely SS and folliculotropic MF [4,5] .
Pruritus is defined as an unpleasant sensation that provokes the desire to scratch. Pruritus is a multidimensional symptom, with effects on cognition and emotion. Pruritus is mediated by histaminedependent and histamine-independent pathways [6] . Conventional antipruritic therapies such as antihistamines are largely ineffective in relieving CTCL-associated pruritus, suggesting unique pathophysiology [2] . Recent research has aimed to elucidate the mechanism of CTCL-associated pruritus. Possible mediators of pruritus in CTCL include interleukin (IL)-4, IL-31, and substance P [5] .
Treatments for relieving pruritus in CTCL have mainly focused on disease-directed therapies, aimed at controlling the underlying lymphoma. Improvement in pruritus has been included as a specific endpoint in recent clinical trials of disease-directed therapies for CTCL. As research on the pathophysiology of CTCL-associated pruritus is relatively recent, few studies on targeted itch-directed therapies are available [5] .
There are a lack of standardized methods used to measure and report itch in clinical studies. This makes it challenging to compare results from studies. A wide range of tools are used to measure the quality of life in patients with dermatological conditions, however, there is no agreed gold standard [3] .
We conducted a systematic review on interventions demonstrating efficacy in reducing pruritus in patients with CTCL. The primary aim of our study was to identify disease-directed and itch-directed therapies effective in reducing CTCL-associated pruritus. Our secondary aim was to outline the various tools used to quantify itch in clinical studies.

Materials and methods
A study protocol was finalized in April 2020 and submitted to the PROSPERO register of systematic reviews. (Prospero registration number CRD42020149210, published on April 28, 2020).

Eligibility criteria
We included all clinical studies published in a peer-reviewed journal that included itch as a specific study endpoint. Studies were included only if a tool was used to objectively quantify itch. We excluded nonhuman studies, conference abstracts, and studies published in a language other than English. Given the rarity of CTCL as a disease entity, we placed no restriction on study design.

Information sources and search strategy
A literature search was performed on the September 10, 2019 for the purpose of identifying relevant published articles. The search was run in Ovid Medline, Ovid Embase, and Ovid Emcare. The date of coverage was not restricted and results were limited to articles in English only.
In Medline, the search strategy consisted of subject headings (MESH) and various text words to identify the literature. Subject headings used in Ovid Medline included: "Lymphoma, T-Cell, Cutaneous," "Pruritus," and "Treatment outcome." Subject headings were combined with keywords (and their word variations) such as "mycosis fungoides," "itch," and "therapy" using the "OR" Boolean operator to cluster all similar indexed and free terms together. The "AND" Boolean operator was applied to link the different concepts.
Searches in Embase and Emcare followed a similar format to the Medline search with variations according to each database's subject thesaurus.
Complete search strategies for each database are outlined in Appendix I.

Study selection
The clinical librarian (S.L.) performed the literature search, retrieved all abstracts, and deduplicated results. Titles and abstracts of identified studies were independently screened by the review team members (G.F., O.S.), according to the inclusion and exclusion criteria. Full-text articles were obtained for studies meeting the inclusion criteria, and their eligibility for inclusion was independently assessed by the review team (G.F., O.S.). The final agreement on study eligibility was resolved by consensus among the collaborators.

Data extraction
Data were independently extracted by the review team (G.F., O.S.) using a pilot-tested form. Data extracted included study design, number of study participants, subtype of CTCL in study patients, intervention, a tool used to quantify pruritus, improvement in itch due to treatment, and if other pruritus treatments were permitted.

Study selection
A total of 1452 records were retrieved by the search strategy. After adjusting for duplicates, 1129 records remained. Of these, 1051 studies were discarded based on a review of titles and abstracts. The full texts of 78 articles were reviewed, of which 27 were eligible for inclusion in accordance with the selection criteria. A flow diagram detailing the inclusion and exclusion of studies from the systematic review is shown in Figure 1.

Study characteristics
Our review identified a total of 18 studies on disease-directed therapies, and 9 studies on itch-directed therapies. Studies were categorized into levels of evidence, according to the National Health and Medical Research Council's framework [7] .

Bias assessment
All subtypes of CTCL were represented in studies, ranging from early-stage to advanced-stage disease. Most studies were unblinded and lacked a placebo control arm. A single randomized controlled trial (RCT) on the use of aprepitant was included, however, the study was small and underpowered. Almost all studies utilized descriptive statistics. The use of confounding antiitch medications was permitted in 16 of the studies, and not permitted in 3 studies. It was unclear if other confounding treatments were used in 8 of the included studies.

Further assessments
Studies were heterogeneous in terms of methods used to quantify itch. As a result, a meta-analysis could not be performed.

Disease-directed therapies
Eighteen studies were identified on various disease-directed therapies. Most studies investigated the use of HiDAC inhibitors (7 studies).

HiDAC inhibitors
Six open-label interventional trials and 1 case series were identified on the use of HiDAC inhibitors, including the drugs romidepsin, vorinostat, quisinostat, and belinostat. The HiDAC studies included larger patient numbers than other studies and included patients with the advanced, treatment-refractory disease. A significant reduction in pruritus was achieved with HiDAC inhibitors for patients with moderate and severe pruritus at baseline. A reduction in pruritus was found in both disease responders and nonresponders for the agents romidepsin [8] , vorinostat [9,10] , and quisinostat [11] .
Vorinostat was effective in reducing pruritus in patients with SS [9,10] . Although patients with SS were included in other HiDAC inhibitors trials, there was no specific comment on improvement in this subset of patients for other agents. Romidepsin demonstrated benefit in pruritus in patients with folliculotropic MF [12] , a variant typically associated with intractable pruritus.

Denileukin diftitox
One open-label interventional trial was found on the use of denileukin diftitox, an IL-2-diphtheria toxin fusion protein. The study included patients with the heavily pretreated disease.
Denileukin diftitox only significantly reduced pruritus in patients with an objective disease response [13] .

Alemtuzumab
One open-label interventional trial was found on the use of alemtuzumab, a humanized monoclonal antibody against CD52. A small trial in patients with treatment-refractory, advancedstage disease demonstrated a reduction in pruritus with alemtuzumab. Reduction in pruritus was achieved for both disease responders and nonresponders, however, the benefit in nonresponders was modest [14] .

Extracorporeal photopheresis (ECP)
One case report demonstrated a long-term reduction in pruritus with ECP in a patient with generalized syringotropic MF, a rare variant of folliculotropic MF [15] .

Total skin electron beam therapy (TSEBT)
One case series reported on the use of low-dose TSEBT in patients with MF, including 1 patient with folliculotropic MF. The patient with folliculotropic MF experienced a modest improvement in pruritus as a result of treatment [16] .

Thalidomide
One case report was identified on the use of thalidomide in a patient with early-stage, treatment-refractory MF. Pruritus relief was sustained long-term, even after cessation of thalidomide [17] .

Retinoids
One open-label interventional trial was found on the oral retinoid, bexarotene. The study included patients with advanced-stage, treatment-refractory disease. A modest reduction in pruritus was noted as a result of treatment with oral bexarotene [18] .
Two open-label interventional trials were identified on topical retinoids (tazarotene and bexarotene), including patients with early-stage disease. Topical retinoids demonstrated no benefit on pruritus in the studies identified [19,20] .

Combination therapy
Three open-label interventional trials were identified that used combination treatment. A trial on the use of vorinostat with bexarotene demonstrated improvement in pruritus in both disease responders and nonresponders [21] . Combination therapy with a HiDAC inhibitor and DNA methyltransferase inhibitor (valproate plus hydralazine) was investigated in one trial, with pruritus improving in all patients [22] . A trial investigating the use of combination therapy with bexarotene and rosiglitazone showed no significant reduction in pruritus [23] .

Itch-directed therapies
A total of 9 studies were identified on itch-directed therapies. Of these, 8 were on the use of the antiemetic aprepitant. One study was identified on the use of lignocaine administered via continuous subcutaneous infusion.

Aprepitant
Eight studies were identified on the use of aprepitant, including 1 RCT, 1 open-label interventional trial, 4 case series, and 2 case reports. Studies included patients with all stages of MF, including patients with SS. Aprepitant was effective in reducing pruritus, with a rapid time of onset (within hours to days). Reduction in pruritus could be sustained long term, for up to 13 months (until the patient passed away from disease) in 1 case report [24] . One RCT concluded that aprepitant was not beneficial in patients with SS. However, the study included only 5 patients, was underpowered, and only followed patients for 21 days [25] .

Lignocaine via continuous subcutaneous infusion
A retrospective case series was identified on the use of lignocaine via continuous subcutaneous infusion. This study included patients with advanced CTCL, including patients hospitalized for end-of-life care. Lignocaine via continuous subcutaneous infusion was effective in reducing pruritus in patients with the advanced-stage disease [6] . Definitions of a significant reduction in pruritus were extremely varied between studies, and are shown in Table 1. Some studies graded improvement in pruritus as "partial" or "complete," again using varied definitions. Definitions of a partial response included a ≥ 3-point reduction in patients with baseline pruritus > 3 on 10-point VAS (n = 1), a 25%-50% reduction in 10-point VAS compared with baseline (n = 1), and a VAS 3-6 on 10-point VAS on a treatment day (n = 1). Definitions of a complete response included VAS of 0 on 100-mm VAS for at least 2 treatment cycles (n = 1), the disappearance of pruritus for ≥ 4 continuous weeks without an increase in the use of antipruritic medications (n = 1), and VAS 0-2 on 10-point VAS on a treatment day (n = 1).

Tools used to quantify pruritus in studies
In addition to quantitative tools, several studies utilized qualitative tools to measure the quality of life. Qualitative tools included the Functional Assessment in Cancer Therapy-General questionnaire (n = 3), patient-rated 7-point global skin severity scale (n = 1), Dermatology Life Quality Index (n = 6), Spitzer quality of life questionnaire (n = 2), a nonvalidated CTCL-specific questionnaire (n = 2), Skindex-16 (n = 1) and Qualitative Patient's Global Assessment (n = 1)

Discussion
Pruritus is a common symptom of CTCL and has a major impact on the quality of life for both patients and their caregivers. A variety of disease-directed and itch-directed therapies demonstrated efficacy in relieving pruritus associated with CTCL. Pruritus relief can be achieved in patients with advanced-stage and treatment-refractory disease, including patients with folliculotropic MF and SS.
Our systematic review identified multiple disease-directed therapies effective in reducing CTCL-associated pruritus. Most evidence supported the use of HiDAC inhibitors. Other effective treatments include denileukin diftitox, alemtuzumab, ECP, TSEBT, thalidomide, bexarotene, combination therapy with vorinostat, and bexarotene, and combination therapy with valproate and hydralazine (HiDAC inhibitor and DNA methyltransferase inhibitor).
In some studies, pruritus was reduced in both patients achieving an objective disease response, and nonresponders. Pruritus relief was achieved for both disease responders and nonresponders with the agents romidepsin, vorinostat, quisinostat, alemtuzumab, and combination therapy with vorinostat and bexarotene. These results suggest that the clinical benefit of agents may extend beyond objective disease responses, and should be taken into consideration when assessing a patient's response to treatment.
SS and folliculotropic MF are subtypes of CTCL associated with severe pruritus. Vorinostat was effective in reducing pruritus in patients with SS [9,10] . Folliculotropic MF involves deeper periadnexal structures and is generally considered resistant to skin-directed therapies. Our results demonstrate that the skin-directed therapies ECP and TSEBT are effective in reducing pruritus in patients with folliculotropic MF. The systemic therapy romidepsin also demonstrated efficacy in reducing pruritus in patients with folliculotropic MF.
A significant proportion of patients with CTCL develop the treatment-refractory disease, with severely debilitating pruritus. There is a need for targeted symptomatic approaches for those patients who do not respond to disease-directed therapies, or for when disease-directed therapies are not appropriate.
Our results support the use of the antiemetic aprepitant as a targeted treatment of CTCL-associated pruritus. Substance P is a neuropeptide released from sensory nerve endings in the skin. Substance P activates neurokinin-1 receptors present on keratinocytes, and on neurons in the dorsal root ganglia. High levels of substance P in keratinocytes are associated with pruritus. Aprepitant is a neurokinin-1 antagonist, inhibiting the effect of substance P [2,5] . Aprepitant demonstrated efficacy in reducing pruritus in patients with all stages of MF, including patients with SS. The effect is rapid in onset, and reduction in pruritus was sustained in some studies. These findings add to the body of evidence suggesting that substance P is an important mediator of pruritus in CTCL.
Lignocaine administered via continuous subcutaneous infusion is effective in reducing pruritus in patients with advancedstage MF, including patients with SS. Lignocaine blocks sodium channels in neuronal cell membranes, and antagonizes kappaopioid antagonist-induced scratching in mouse models [34] . Lignocaine administered by subcutaneous infusion should be considered in patients with intractable pruritus. This intervention also provides an important treatment option for symptom control in patients receiving end-of-life care.
Our systematic review highlights the variety of tools used to quantify itch in clinical studies and the subsequent difficulties in comparing results across studies. An ideal tool to quantify a symptom must be well-validated, appropriate for progressive conditions, simple to complete, and multidimensional [3] .
Our study identified the VAS as the most commonly used tool for measuring pruritus, using either a 100-mm VAS or 10-point VAS. The VAS was originally developed to assess the intensity of pain and has been adapted for use in the measurement of itch. The VAS is a 100-mm long line, on which patients indicate the severity of pruritus by crossing the line at the point that corresponds to their pruritus intensity. The VAS is a simple and reproducible tool and is one of the most commonly used methods of assessing pruritus severity. Studies have demonstrated that the VAS is a reliable method of pruritus assessment. The VAS is quick to perform and suitable for repeat assessments in clinical settings. The VAS demonstrates good test-retest reproducibility in patients with pruritic skin conditions. However, the VAS is a unidimensional tool, and only measures itch intensity. The tool is unsuitable for patients with motor deficits or cognitive issues, which may exclude the elderly or patients receiving end-of-life care [35] .
The next most frequently used measure of pruritus identified by our study was the NRS. The NRS is a similar method of pruritus measurement to the VAS. Patients rate pruritus intensity on a scale from 0 to 10 (no pruritus to worst imaginable pruritus) [36] . The NRS has been validated as a tool for assessing pain severity. In the study of itch, high concurrent validity in pruritus intensity assessment has been shown with both the VAS and NRS. However, discrimination of pruritus intensity using the VAS may be more sensitive than with the NRS [37] .
Studies included in our systematic review varied greatly in the definition of significant pruritus at baseline, and what constitutes a clinically meaningful reduction in pruritus. More recent trials on the use of HiDAC inhibitors defined a clinically meaningful reduction in pruritus as a ≥ 30 mm (or ≥ 3 points) improvement in VAS in patients with a baseline pruritus score of ≥ 30 mm (or ≥ 3 points), without an increase in the use of antipruritus medications, over a defined time period. This varied between 2 consecutive treatment cycles, 3 continuous weeks, and 4 continuous weeks [8][9][10][11]22] . In these studies, a ≥ 30 mm reduction in pruritus was prospectively selected as a clinically significant reduction in pruritus based on expert opinion, and previous use in other clinical trials [8] . A clinically significant reduction in pruritus should be defined more vigorously, ideally by incorporating the patient's perception of pruritus relief.
The need for breakthrough medication can be considered a surrogate marker of a patient experiencing inadequate symptom relief. In the acute pain setting, the action of a patient not requesting breakthrough analgesia can be considered a marker that the patient perceives the therapy as effective. This marker has been used in previous studies to determine thresholds in pain scales that are considered clinically significant. This method places a stronger focus on the patient's experience of pain, rather than arbitrarily assigning a numerical value as significant. This approach could be translated to research on pruritus. This information is simple to capture in the setting of a trial, and clinically appropriate. Although the choice of scales and analyses for the study of itch would differ, the same marker could be used to calculate a patient-determined clinically significant reduction in pruritus [38][39][40] .
To our knowledge, this is the first systematic review specifically addressing the management of itch in patients with CTCL. Patients with all stages of CTCL were represented across included studies, including patients with folliculotropic MF and SS, specific subtypes associated with severe and intractable pruritus. A wide range of treatment options were identified, including options appropriate for patients with end-stage disease.
Our systematic review has limitations. We limited results to articles published in English, excluded conference abstracts, and did not contact authors for further information regarding studies. Our systematic review retrieved no studies on certain treatments that are commonly used for the management of pruritus in clinical practice, including wet wraps, emollients, narrow-band UVB, topical and oral corticosteroids, naloxone, naltrexone, butorphanol, mirtazapine, and gabapentin [5] . We also found no studies on the disease-directed therapy mogamulizumab. This may be a reflection of our inclusion criteria.
The HiDAC inhibitors may be overrepresented in our results. As recent studies on the use of HiDAC inhibitors for CTCL included a reduction in pruritus as a study outcome, more of these studies met our inclusion criteria. This does not necessary imply that the HiDAC inhibitors are the most effective treatment for CTCL-associated pruritus.
We acknowledge that most included studies fall into low levels of evidence, particularly for itch-directed studies. Evidence supporting some agents was based only on case reports, and we acknowledge the risk of publication bias associated with case reports. However, given the rarity of CTCL as a disease entity, well-designed RCTs for this patient population are impracticable.
CTCL-associated pruritus is a debilitating symptom and requires a systematic treatment approach. We have identified multiple disease-directed and itch-directed therapies demonstrating efficacy in reducing pruritus in patients with CTCL.
Translating a subjective experience into an objective measure used for research and treatment evaluation is challenging. Our study identified the VAS and NRS as the most commonly used measures of pruritus in patients with CTCL. However, a unidimensional numerical scale alone does not adequately capture the patient's experience of pruritus.
By drawing on the pain literature, various research techniques could be translated to the study of itch. The VAS or NRS should be validated to define a patient-determined clinically significant reduction in pruritus. For example, a patient not requiring breakthrough medication could be considered a surrogate marker of a patient experiencing adequate symptom relief. Validating the VAS or NRS for use in the study of itch would allow comparison of results between studies, as well as pooling of results across studies. This would assist in adding to the body of research on this rare and debilitating disease.