MINI REVIEW
Malignancy-associated pruritus B. Rowe, G. Yosipovitch Department of Dermatology and Temple Itch Center, Temple University School of Medicine, Philadelphia, Pennsylvania
Correspondence Gil Yosipovitch E-mail: [email protected] Funding sources None. Conflicts of interest None declared.
Accepted for publication 31 May 2015 doi:10.1002/ejp.760
Abstract Malignancy-associated pruritus can be the result of a neoplasm’s local effect on tissue or due to the systemic reaction to malignancy. A systemic reaction to malignancy has been termed ‘paraneoplastic itch’ and can be the first sign of an underlying malignancy. Paraneoplastic itch is most commonly caused by lymphoproliferative malignancies, and severity of itch correlates with stage of disease in Hodgkin’s lymphoma and polycythemia vera. Non-melanoma skin cancer is the most common type of malignancy-associated pruritus, and recent data indicate that pruritus is associated with more than one-third of non-melanoma skin cancers. Cutaneous T-cell lymphomas (CTCL), particularly more advanced stages, cause intractable pruritus and recent investigations into the pathophysiology of CTCL-associated itch have implicated cyotokine interleukin-31 as a putative mediator. Treatments that reduce itch in CTCL patients, such as histone deacetylase inhibitors (HDACi), Mogamulizumab, a novel monoclonal antibody against chemokine receptor type-4, and oral corticosteroids, have demonstrated a correlation between their anti-pruritic effect and reduced serum levels of interleukin-31.
1. Introduction Pruritus has been associated with malignancies for decades; in particular, haematological malignancies such as Hodgkin’s lymphoma, leukaemia and cutaneous T-cell lymphoma (CTCL) as well as biliary malignancies (Mandal et al., 2007; Polat et al., 2008; Vij and Duvic, 2012). Part of the work-up for patients with chronic pruritus in the absence of a primary skin disease is to rule out an underlying malignancy (Yosipovitch and Bernhard, 2013). However, there are few large-scale epidemiological studies to guide this practice. Pruritus associated with malignancy can be divided into two broad categories: itch induced by local reaction to malignancy and paraneoplastic itch. The Special Interest Group of the International Forum on the Study of Itch defines paraneoplastic itch as ‘the sensation of itch as a systemic (not local) reaction to the presence of a tumour or a haemato© 2015 European Pain Federation - EFICâ
logical malignancy neither induced by the local presence of cancer cells nor by tumour therapy’ (Weisshaar et al., 2015). A second classification of paraneoplastic itch is pruritus that occurs in the natural progression of malignancy or preceding evidence of an underlying malignancy, which is not caused by neoplastic mass invasion or compression, and subsides after resolution of the malignancy (Yosipovitch, 2010).
2. Prevalence of itch in malignancy Although malignancy has been associated with pruritus for decades, there is limited prospective epidemiological data. One study of 700 solid tumour and haematological cancer patients reported generalized pruritus was present in 13% of cancer patients with concurrent dermatological conditions (Kilic et al., 2007). Two recent population studies have tried to address the associated risk of an underlying malignancy in patients who suffer from chronic pruritus. Eur J Pain 20 (2016) 19--23
19
Malignancy-associated pruritus
What does this review add? • We review the epidemiology, pathophysiology and treatments for malignancy-associated pruritus. • We present the latest epidemiological data on skin cancer pruritus, and the association of interleukin-31 with cutaneous T-cell lymphoma. • We illustrate the importance of pruritus as a paraneoplastic sign for initial diagnosis as well as prognosis.
The first study followed 8743 patients with chronic itch in the absence of observable skin lesions for 5 years, and compared the incidence of developing malignancy to control subjects that did not suffer from chronic itch. They found that chronic itch without concomitant skin changes increased the overall risk for the diagnosis of a subsequent haematological (hazard ratio 2.0) or bile duct malignancy (hazard ratio 3.0) when compared to the control group (Fett et al., 2014). The study authors concluded that intractable itch without concomitant skin changes could be used in coordination with clinical signs and symptoms to determine if additional screening for malignancy should be pursued. The second investigation was a nationwide Danish cohort study, which recruited 12,813 patients diagnosed with pruritus. Patients were followed from 1978 to 2011 and first-time cancer diagnoses were recorded. Standardized incidence ratios (SIRs) for the development of malignancy compared to the general population and 1-year absolute risk of developing cancer were calculated. The overall incidence of cancer was not high (SIR 1.13), and was more common in men (SIR 1.22) than women (SIR 1.05). Both haematological (SIR 1.68) and various solid tumours were observed at increased rates compared to the general population. Overall, the highest SIRs were observed during the first 3 months of follow-up (SIR 2.14), and declined rapidly thereafter. The 1-year absolute risk of a cancer diagnosis was 1.6%; therefore, 155 patients with pruritus would need to be evaluated for malignancy to detect one additional neoplasm (Johannesdottir et al., 2014).
3. Skin cancer pruritus Pruritus associated with skin cancer is due to the local cutaneous reaction to malignancy. A recent 20 Eur J Pain 20 (2016) 19--23
B. Rowe and G. Yosipovitch
study investigating itch in 478 patients with biopsyproven non-melanoma skin cancer found the prevalence of pruritus in squamous cell carcinoma and basal cell carcinoma to be 43% and 33%, respectively (Mills et al., 2012). With an incidence of four million new cases of non-melanoma skin cancers per year in the United States, this study suggests there are over one million new cases of malignancy-associated pruritus due to non-melanoma skin cancer per year. This makes non-melanoma skin cancer the most prevalent type of malignancy-associated itch. Interestingly, the severity of itch in non-melanoma skin cancer as measured by a 10-point visual analogue scale (VAS) was low (3.9), and in most cases did not require treatment as pruritus resolved with tumour resection (Yosipovitch et al., 2014). Pruritus is a common clinical finding in patients with CTCL. One retrospective study analysing bedside surveys from 486 patients with CTCL found 66% of patients suffered from itch. The average itch severity, as rated on a 10-point VAS, was found to be 4.2 across the entire patient population. Additionally, severity of itch was found to correlate with stage of disease with VAS ratings of 3.4, 6.6 and 7.7 found in early-stage mycosis fungoides, late-stage mycosis fungoides and S ezary syndrome, respectively (Vij and Duvic, 2012).
4. Paraneoplastic itch Paraneoplastic itch can be a common clinical finding in several lymphoproliferative neoplasms, including multiple myeloma, non-Hodgkin’s lymphoma (15% prevalence) and Hodgkin’s lymphoma (25% prevalence) (Gobbi et al., 1983; Mandal et al., 2007). Additionally, severe pruritus has been correlated with advanced stage of Hodgkin’s disease (Gobbi et al., 1983). Similar findings have been observed in patients with CTCL (Vij and Duvic, 2012). Intractable itch not only serves as a marker of disease progression but can also be a harbinger of neoplasm. One retrospective study investigating the cutaneous manifestation of Hodgkin’s disease in 1049 patients found generalized itch preceded the diagnosis of Hodgkin’s disease in 17 patients (Rubenstein and Duvic, 2006). The paraneoplastic itch associated with polycythemia vera has the unique characteristic of occurring predominately after patients come into contact with water. This pattern of itch has been termed ‘aquagenic pruritus’ and is characterized as a ‘prickling-like’ itch that occurs in the absence of observable skin lesions. Aquagenic pruritus is a common © 2015 European Pain Federation - EFICâ
B. Rowe and G. Yosipovitch
finding in polycythemia vera with a reported prevalence of 48–68% (Siegel et al., 2013). It is less frequently observed in other lymphoproliferative disorders, such as lymphoma and leukaemia, but the exact prevalence is unknown. While paraneoplastic itch in haematological malignancies can often occur in the absence of skin lesions, several observable pruritic skin conditions can occur secondary to malignancy, including erythroderma, acanthosis nigricans, dermatomyositis, Grover’s disease and eruptive seborrhoeic keratoses (Yosipovitch, 2010). Pruritus in the setting of solid tumours is an uncommon form of paraneoplastic itch, with the exception of obstructive tumours in the pancreatic head and liver cancers resulting in cholestasis (Wang and Yosipovitch, 2010).
5. Pathophysiology The pathophysiology of malignancy-associated itch is poorly understood, but recent investigations have provided clues to mediators in this process. When investigators examined the histological features of skin cancer, they found the degree of inflammation and the presence of eosinophils were significantly associated with itch (Yosipovitch et al., 2014). Eosinophils may play a role in pruritus through their ability to amplify CD4 helper T cells via antigen presentation (Shi, 2004). In vivo experiments demonstrated the cytokine-mediated induction of major histocompatability complex II (MHC-II) on eosinophils. MHC-II expression allows for activation of CD4 helper T cells, which have received recent attention for their role in generalized pruritus.
5.1 Interleukin-31’s role in cutaneous T-cell lymphoma itch Interleukin-31 (IL-31) is a four-helix cytokine from the Interleukin-6 family and is predominately produced by CD4 helper T cells. One investigation implicates IL-31 in the pathophysiology of pruritic skin conditions by showing that IL-31 is capable of attaching to two receptor sites: IL-31 receptor A and the oncostatin M receptor (Dillon et al., 2004). IL-31 receptor A is predominantly found on the sensory dorsal root ganglion, while oncostatin M has been linked to pruritus by its association with primary localized cutaneous amyloidosis, an extremely pruritic disease (Sonkoly et al., 2006). IL-31 was found to induce itch when overexpressed in transgenic mice (Dillon et al., 2004). © 2015 European Pain Federation - EFICâ
Malignancy-associated pruritus
Additionally, IL-31 mRNA is elevated in pruritic skin diseases such as atopic dermatitis, allergic contact dermatitis and prurigo nodularis (Sonkoly et al., 2006). An additional study found serum levels of IL31 correlated with the severity of atopic dermatitis (Raap et al., 2008), thus demonstrating that IL-31 is not only associated with pruritic skin disease, but the intensity of itch experienced by patients is also linked to serum levels of IL-31. IL-31 has recently shown promise in elucidating the pathophysiology of CTCL pruritus. Serum levels of IL-31 were found to be elevated in patients with CTCL when compared to healthy control subjects (Ohmatsu et al., 2012). Given the correlation of IL31 to pruritic skin lesions and CTCL, a recent case– control study set out to determine if IL-31 serum levels correlated with the presence of itch in CTCL patients. Researchers compared serum levels of IL-31 in patients with CTCL with pruritus to CTCL patients without concurrent itch and found a significantly elevated IL-31 level in the pruritic group (p = 0.008) (Singer et al., 2013). A subsequent study found that reducing IL-31 using histone deacetylase inhibitors, prednisone or anti-CCR4 monoclonal antibodies in advanced stage CTCL patients significantly reduced pruritus (Cedeno-Laurent et al., 2015). However, a previous study did not find a link between serum levels of IL-31 and itch intensity in patients with early stage CTCL (Malek et al., 2015). The study authors suggest that these contradictory findings may be reconciled by the different mechanisms by which IL-31 is produced in early- and late-stage CTCL. Early-stage CTCL has a predominance of Th1produced cytokines, while IL-31 is produced mainly by Th2 lymphocytes in late-stage disease. Production of IL-31 by two separate pathways could help explain its association with itch in late-stage CTCL, while having no relationship with pruritus in earlystage disease. We recently studied the expression of IL-31 and its receptor in the skin of CTCL patients. A comparison of CTCL patients with mild and severe itch uncovered a significant correlation with keratinocyte expressions of IL-31 and its receptor to itch severity (Nattkemper et al., unpublished data). More research is needed to investigate the potential role of IL-31 and mechanism of action in CTCL and malignancy-associated itch.
6. Treatment The most effective treatment for malignancy-associated itch is to treat the underlying malignancy. Large-scale randomized clinical trials have yet to be Eur J Pain 20 (2016) 19--23
21
Malignancy-associated pruritus
done on treatments for paraneoplastic itch, but smaller studies have demonstrated the efficacy of a number of therapies. One randomized clinical trial demonstrated a significant reduction in paraneoplastic itch intensity when given paroxetine 20 mg daily compared to placebo (Zylicz et al., 2003). Another case report detailed successfully treating paraneoplastic itch secondary to Hodgkin’s lymphoma with thalidomide (Goncalves, 2010). A nightly thalidomide dose of 200 mg was used after a 22-year-old woman with nodular sclerosis type II Hodgkin lymphoma had failed multiple therapies for pruritus. Her rated itch severity on a 10-point VAS improved from ratings of 8–10 prior to therapy to an itch severity of 3 after initiation of thalidomide therapy. Special care must be taken in prescribing thalidomide to women of childbearing age due to its well-known teratogenic effect, which can be seen after just one dose. Peripheral neuropathy from prolonged use of more than 3 months is another adverse effect, which has a reported 2-year incidence of 25–56% of patients treated with thalidomide (Bastuji-Garin et al., 2002). Kappa-opioid agonists and mu-opioid antagonists have known anti-pruritic effects (Cowan et al., 2015; Ko, 2015). This has lead to use of butorphanol, a mixed kappa-opioid agonist and mu-opioid antagonist, for the treatment of paraneoplastic itch. One case report details a 42-year-old woman who suffered from pruritus secondary to non-Hodgkin’s lymphoma and failed multiple treatments for itch. She noted improvement in itch after a single dose of butorphanol (1 mg) and her dose was titrated to 4 mg per day for improved itch control (Dawn and Yosipovitch, 2006). Gabapentin and mirtazapine have been reported to be efficacious in the treatment of itch for CTCL (Demierre and Taverna, 2006). Study authors suggest a starting gabapentin dose of 300 mg nightly and titrating up to maximum of 2400 mg per day, and a starting dose of 7.5 mg for mirtazapine administered at night, titrating to a maximum dose of 15 mg nightly. A prednisone taper starting at 40 mg and tapering over 3 weeks has reported effectiveness in alleviating pruritus in lymphoma patients (Wang and Yosipovitch, 2010). Two novel classes of therapeutic agents, histone deacetylase inhibitors (HDACis) and monoclonal antibodies against chemokine receptor type-4 (CCR4), have demonstrated efficacy in treating CTCL pruritus by targeting the malignant IL-31-producing T cells. Vorinostat and romidepsin are HDACis recently approved by the Food and Drug Administration for the treatment of CTCL. The mechanism of 22 Eur J Pain 20 (2016) 19--23
B. Rowe and G. Yosipovitch
action of HDACis is through their disruption of histone formation, a key process in the production of chromatin and a vital step needed for rapidly dividing malignant cells. Vorinostat and romidepsin have shown efficacy in reducing pruritus in patients with CTCL (Cedeno-Laurent et al., 2015). Mogamulizumab, a monoclonal antibody against CCR4, was recently approved in Japan for the treatment of CTCL. CCR4 is found on malignant IL-31producing T cells in CTCL, and blockade of this receptor with mogamulizumab has been shown to reduce pruritus in CTCL (Cedeno-Laurent et al., 2015). HDACis and mogamulizumab could serve as potential treatments of CTCL and associated pruritus.
7. Conclusion This mini-review provided recent data on malignancy-associated pruritus. Itch can present as the first paraneoplastic sign of an underlying malignancy or as a correlate to disease progression. We reviewed recent data, which demonstrated non-melanoma skin cancer to be the most prevalent form of malignancy-associated pruritus. The pathophysiology behind malignancy-associated itch remains elusive, but research is currently being conducted to further elucidate this process. A better understanding of the pathophysiology of malignancy-associated pruritus will lead to improved therapies in the future. Author contributions B.R. was involved in manuscript writing and G.Y. was involved in manuscript preparation and writing.
References Bastuji-Garin, S., Ochonisky, S., Bouche, P., Gherardi, R.K., Duguet, C., Djerradine, Z., Poli, F., Revuz, J., Thalidomide Neuropathy Study Group (2002). Incidence and risk factors for thalidomide neuropathy: a prospective study of 135 dermatologic patients. J Invest Dermatol 119, 1020–1026. Cedeno-Laurent, F., Singer, E.M., Wysocka, M., Benoit, B.M., Vittorio, C.C., Kim, E.J., Yosipovitch, G., Rook, A.H. (2015). Improved pruritus correlates with lower levels of IL-31 in CTCL patients under different therapeutic modalities. Clin Immunol 158, 1–7. Cowan, A., Kehner, G., Inan, S. (2015). Targeting itch with ligands selective for kappa opioid receptors. In Pharmacology of Itch. Cowan, A., Yosipovitch, G., eds. (Berlin Heidelberg: Springer) 291– 314. Dawn, A.G., Yosipovitch, G. (2006). Butorphanol for treatment of intractable pruritus. J Am Acad Dermatol 54, 527–531. Demierre, M.F., Taverna, J. (2006). Mirtazapine and gabapentin for reducing pruritus in cutaneous T-cell lymphoma. J Am Acad Dermatol 55, 543–544. Dillon, S.R., Sprecher, C., Hammond, A., Bilsborough, J., RosenfeldFranklin, M., Presnell, S.R., Haugen, H.S., Maurer, M., Harder, B., Johnston, J., Bort, S., Mudri, S., Kuijper, J.L., Bukowski, T., Shea, P.,
© 2015 European Pain Federation - EFICâ
B. Rowe and G. Yosipovitch
Dong, D.L., Dasovich, M., Grant, F.J., Lockwood, L., Levin, S.D., LeCiel, C., Waggie, K., Day, H., Topouzis, S., Kramer, J., Kuestner, R., Chen, Z., Foster, D., Parrish-Novak, J., Gross, J.A. (2004). Interleukin 31, a cytokine produced by activated T cells, induces dermatitis in mice. Nat Immunol 5, 752–760. Fett, N., Haynes, K., Propert, K.J., Margolis, D.J. (2014). Five-year malignancy incidence in patients with chronic pruritus: A population-based cohort study aimed at limiting unnecessary screening practices. J Am Acad Dermatol 70, 651–658. Gobbi, P.G., Attardo-Parrinello, G., Lattanzio, G., Rizzo, S.C., Ascari, E. (1983). Severe pruritus should be a B-symptom in hodgkin’s disease. Cancer 51, 1934–1936. Goncalves, F. (2010). Thalidomide for the control of severe paraneoplastic pruritus associated with hodgkin’s disease. Am J Hosp Palliat Care 27, 486–487. Johannesdottir, S.A., Farkas, D.K., Vinding, G.R., Pedersen, L., Lamberg, A., Sorensen, H.T., Olesen, A.B. (2014). Cancer incidence among patients with a hospital diagnosis of pruritus: A nationwide Danish cohort study. Br J Dermatol 171, 839–846. Kilic, A., Gul, U., Soylu, S. (2007). Skin findings in internal malignant diseases. Int J Dermatol 46, 1055–1060. Ko, M.C. (2015). Neuraxial opioid-induced itch and its pharmacological antagonism. In Pharmacology of Itch. Cowan, A., Yosipovitch, G., eds. (Berlin Heidelberg: Springer) 315–335. Malek, M., Glen, J., Rebala, K., Kowalczyk, A., Sobjanek, M., Nowicki, R., Ruckemann-Dziurdzinska, K., Sokolowska-Wojdylo, M. (2015). Il-31 does not correlate to pruritus related to early stage cutaneous Tcell lymphomas but is involved in pathogenesis of the disease. Acta Derm Venereol 95, 283–288. Mandal, S., Varma, K., Jain, S. (2007). Cutaneous manifestations in non-Hodgkin’s lymphoma. Acta Cytol 51, 853–859. Mills, K.C., Kwatra, S.G., Feneran, A.N., Pearce, D.J., Williford, P.M., D’Agostino, R.B., Yosipovitch, G. (2012). Itch and pain in nonmelanoma skin cancer: Pain as an important feature of cutaneous squamous cell carcinoma. Arch Dermatol 148, 1422–1423. Ohmatsu, H., Sugaya, M., Suga, H., Morimura, S., Miyagaki, T., Kai, H., Kagami, S., Fujita, H., Asano, Y., Tada, Y., Kadono, T., Sato, S. (2012). Serum IL-31 levels are increased in patients with cutaneous T-cell lymphoma. Acta Derm Venereol 92, 282–283. Polat, M., Oztas, P., Ilhan, M.N., Yalcin, B., Alli, N. (2008). Generalized pruritus: A prospective study concerning etiology. Am J Clin Dermatol 9, 39–44.
© 2015 European Pain Federation - EFICâ
Malignancy-associated pruritus
Raap, U., Wichmann, K., Bruder, M., Stander, S., Wedi, B., Kapp, A., Werfel, T. (2008). Correlation of IL-31 serum levels with severity of atopic dermatitis. J Allergy Clin Immunol 122, 421–423. Rubenstein, M., Duvic, M. (2006). Cutaneous manifestations of Hodgkin’s disease. Int J Dermatol 45, 251–256. Shi, H.Z. (2004). Eosinophils function as antigen-presenting cells. J Leukoc Biol 76, 520–527. Siegel, F.P., Tauscher, J., Petrides, P.E. (2013). Aquagenic pruritus in polycythemia vera: Characteristics and influence on quality of life in 441 patients. Am J Hematol 88, 665–669. Singer, E.M., Shin, D.B., Nattkemper, L.A., Benoit, B.M., Klein, R.S., Didigu, C.A., Loren, A.W., Dentchev, T., Wysocka, M., Yosipovitch, G., Rook, A.H. (2013). IL-31 is produced by the malignant T-cell population in cutaneous T-cell lymphoma and correlates with CTCL pruritus. J Invest Dermatol 133, 2783–2785. Sonkoly, E., Muller, A., Lauerma, A.I., Pivarcsi, A., Soto, H., Kemeny, L., Alenius, H., Dieu-Nosjean, M.C., Meller, S., Rieker, J., Steinhoff, M., Hoffmann, T.K., Ruzicka, T., Zlotnik, A., Homey, B. (2006). IL31: A new link between T cells and pruritus in atopic skin inflammation. J Allergy Clin Immunol 117, 411–417. Vij, A., Duvic, M. (2012). Prevalence and severity of pruritus in cutaneous T cell lymphoma. Int J Dermatol 51, 930–934. Wang, H., Yosipovitch, G. (2010). New insights into the pathophysiology and treatment of chronic itch in patients with endstage renal disease, chronic liver disease, and lymphoma. Int J Dermatol 49, 1–11. Weisshaar, E., Weiss, M., Mettang, T., Yosipovitch, G., Zylicz, Z. (2015). Paraneoplastic itch: An expert position statement from the special interest group (SIG) of the international forum on the study of itch (IFSI). Acta Derm Venereol 95, 261–265. Yosipovitch, G. (2010). Chronic pruritus: A paraneoplastic sign. Dermatol Ther 23, 590–596. Yosipovitch, G., Bernhard, J.D. (2013). Clinical practice. Chronic pruritus. N Engl J Med 368, 1625–1634. Yosipovitch, G., Mills, K.C., Nattkemper, L.A., Feneran, A., Tey, H.L., Lowenthal, B.M., Pearce, D.J., Williford, P.M., Sangueza, O.P., D’Agostino, R.B. Jr (2014). Association of pain and itch with depth of invasion and inflammatory cell constitution in skin cancer: Results of a large clinicopathologic study. JAMA Dermatol 150, 1160–1166. Zylicz, Z., Krajnik, M., Sorge, A.A., Costantini, M. (2003). Paroxetine in the treatment of severe non-dermatological pruritus: A randomized, controlled trial. J Pain Symptom Manage 26, 1105–1112.
Eur J Pain 20 (2016) 19--23
23
Malignancy-associated pruritus B. Rowe, G. Yosipovitch Department of Dermatology and Temple Itch Center, Temple University School of Medicine, Philadelphia, Pennsylvania
Correspondence Gil Yosipovitch E-mail: [email protected] Funding sources None. Conflicts of interest None declared.
Accepted for publication 31 May 2015 doi:10.1002/ejp.760
Abstract Malignancy-associated pruritus can be the result of a neoplasm’s local effect on tissue or due to the systemic reaction to malignancy. A systemic reaction to malignancy has been termed ‘paraneoplastic itch’ and can be the first sign of an underlying malignancy. Paraneoplastic itch is most commonly caused by lymphoproliferative malignancies, and severity of itch correlates with stage of disease in Hodgkin’s lymphoma and polycythemia vera. Non-melanoma skin cancer is the most common type of malignancy-associated pruritus, and recent data indicate that pruritus is associated with more than one-third of non-melanoma skin cancers. Cutaneous T-cell lymphomas (CTCL), particularly more advanced stages, cause intractable pruritus and recent investigations into the pathophysiology of CTCL-associated itch have implicated cyotokine interleukin-31 as a putative mediator. Treatments that reduce itch in CTCL patients, such as histone deacetylase inhibitors (HDACi), Mogamulizumab, a novel monoclonal antibody against chemokine receptor type-4, and oral corticosteroids, have demonstrated a correlation between their anti-pruritic effect and reduced serum levels of interleukin-31.
1. Introduction Pruritus has been associated with malignancies for decades; in particular, haematological malignancies such as Hodgkin’s lymphoma, leukaemia and cutaneous T-cell lymphoma (CTCL) as well as biliary malignancies (Mandal et al., 2007; Polat et al., 2008; Vij and Duvic, 2012). Part of the work-up for patients with chronic pruritus in the absence of a primary skin disease is to rule out an underlying malignancy (Yosipovitch and Bernhard, 2013). However, there are few large-scale epidemiological studies to guide this practice. Pruritus associated with malignancy can be divided into two broad categories: itch induced by local reaction to malignancy and paraneoplastic itch. The Special Interest Group of the International Forum on the Study of Itch defines paraneoplastic itch as ‘the sensation of itch as a systemic (not local) reaction to the presence of a tumour or a haemato© 2015 European Pain Federation - EFICâ
logical malignancy neither induced by the local presence of cancer cells nor by tumour therapy’ (Weisshaar et al., 2015). A second classification of paraneoplastic itch is pruritus that occurs in the natural progression of malignancy or preceding evidence of an underlying malignancy, which is not caused by neoplastic mass invasion or compression, and subsides after resolution of the malignancy (Yosipovitch, 2010).
2. Prevalence of itch in malignancy Although malignancy has been associated with pruritus for decades, there is limited prospective epidemiological data. One study of 700 solid tumour and haematological cancer patients reported generalized pruritus was present in 13% of cancer patients with concurrent dermatological conditions (Kilic et al., 2007). Two recent population studies have tried to address the associated risk of an underlying malignancy in patients who suffer from chronic pruritus. Eur J Pain 20 (2016) 19--23
19
Malignancy-associated pruritus
What does this review add? • We review the epidemiology, pathophysiology and treatments for malignancy-associated pruritus. • We present the latest epidemiological data on skin cancer pruritus, and the association of interleukin-31 with cutaneous T-cell lymphoma. • We illustrate the importance of pruritus as a paraneoplastic sign for initial diagnosis as well as prognosis.
The first study followed 8743 patients with chronic itch in the absence of observable skin lesions for 5 years, and compared the incidence of developing malignancy to control subjects that did not suffer from chronic itch. They found that chronic itch without concomitant skin changes increased the overall risk for the diagnosis of a subsequent haematological (hazard ratio 2.0) or bile duct malignancy (hazard ratio 3.0) when compared to the control group (Fett et al., 2014). The study authors concluded that intractable itch without concomitant skin changes could be used in coordination with clinical signs and symptoms to determine if additional screening for malignancy should be pursued. The second investigation was a nationwide Danish cohort study, which recruited 12,813 patients diagnosed with pruritus. Patients were followed from 1978 to 2011 and first-time cancer diagnoses were recorded. Standardized incidence ratios (SIRs) for the development of malignancy compared to the general population and 1-year absolute risk of developing cancer were calculated. The overall incidence of cancer was not high (SIR 1.13), and was more common in men (SIR 1.22) than women (SIR 1.05). Both haematological (SIR 1.68) and various solid tumours were observed at increased rates compared to the general population. Overall, the highest SIRs were observed during the first 3 months of follow-up (SIR 2.14), and declined rapidly thereafter. The 1-year absolute risk of a cancer diagnosis was 1.6%; therefore, 155 patients with pruritus would need to be evaluated for malignancy to detect one additional neoplasm (Johannesdottir et al., 2014).
3. Skin cancer pruritus Pruritus associated with skin cancer is due to the local cutaneous reaction to malignancy. A recent 20 Eur J Pain 20 (2016) 19--23
B. Rowe and G. Yosipovitch
study investigating itch in 478 patients with biopsyproven non-melanoma skin cancer found the prevalence of pruritus in squamous cell carcinoma and basal cell carcinoma to be 43% and 33%, respectively (Mills et al., 2012). With an incidence of four million new cases of non-melanoma skin cancers per year in the United States, this study suggests there are over one million new cases of malignancy-associated pruritus due to non-melanoma skin cancer per year. This makes non-melanoma skin cancer the most prevalent type of malignancy-associated itch. Interestingly, the severity of itch in non-melanoma skin cancer as measured by a 10-point visual analogue scale (VAS) was low (3.9), and in most cases did not require treatment as pruritus resolved with tumour resection (Yosipovitch et al., 2014). Pruritus is a common clinical finding in patients with CTCL. One retrospective study analysing bedside surveys from 486 patients with CTCL found 66% of patients suffered from itch. The average itch severity, as rated on a 10-point VAS, was found to be 4.2 across the entire patient population. Additionally, severity of itch was found to correlate with stage of disease with VAS ratings of 3.4, 6.6 and 7.7 found in early-stage mycosis fungoides, late-stage mycosis fungoides and S ezary syndrome, respectively (Vij and Duvic, 2012).
4. Paraneoplastic itch Paraneoplastic itch can be a common clinical finding in several lymphoproliferative neoplasms, including multiple myeloma, non-Hodgkin’s lymphoma (15% prevalence) and Hodgkin’s lymphoma (25% prevalence) (Gobbi et al., 1983; Mandal et al., 2007). Additionally, severe pruritus has been correlated with advanced stage of Hodgkin’s disease (Gobbi et al., 1983). Similar findings have been observed in patients with CTCL (Vij and Duvic, 2012). Intractable itch not only serves as a marker of disease progression but can also be a harbinger of neoplasm. One retrospective study investigating the cutaneous manifestation of Hodgkin’s disease in 1049 patients found generalized itch preceded the diagnosis of Hodgkin’s disease in 17 patients (Rubenstein and Duvic, 2006). The paraneoplastic itch associated with polycythemia vera has the unique characteristic of occurring predominately after patients come into contact with water. This pattern of itch has been termed ‘aquagenic pruritus’ and is characterized as a ‘prickling-like’ itch that occurs in the absence of observable skin lesions. Aquagenic pruritus is a common © 2015 European Pain Federation - EFICâ
B. Rowe and G. Yosipovitch
finding in polycythemia vera with a reported prevalence of 48–68% (Siegel et al., 2013). It is less frequently observed in other lymphoproliferative disorders, such as lymphoma and leukaemia, but the exact prevalence is unknown. While paraneoplastic itch in haematological malignancies can often occur in the absence of skin lesions, several observable pruritic skin conditions can occur secondary to malignancy, including erythroderma, acanthosis nigricans, dermatomyositis, Grover’s disease and eruptive seborrhoeic keratoses (Yosipovitch, 2010). Pruritus in the setting of solid tumours is an uncommon form of paraneoplastic itch, with the exception of obstructive tumours in the pancreatic head and liver cancers resulting in cholestasis (Wang and Yosipovitch, 2010).
5. Pathophysiology The pathophysiology of malignancy-associated itch is poorly understood, but recent investigations have provided clues to mediators in this process. When investigators examined the histological features of skin cancer, they found the degree of inflammation and the presence of eosinophils were significantly associated with itch (Yosipovitch et al., 2014). Eosinophils may play a role in pruritus through their ability to amplify CD4 helper T cells via antigen presentation (Shi, 2004). In vivo experiments demonstrated the cytokine-mediated induction of major histocompatability complex II (MHC-II) on eosinophils. MHC-II expression allows for activation of CD4 helper T cells, which have received recent attention for their role in generalized pruritus.
5.1 Interleukin-31’s role in cutaneous T-cell lymphoma itch Interleukin-31 (IL-31) is a four-helix cytokine from the Interleukin-6 family and is predominately produced by CD4 helper T cells. One investigation implicates IL-31 in the pathophysiology of pruritic skin conditions by showing that IL-31 is capable of attaching to two receptor sites: IL-31 receptor A and the oncostatin M receptor (Dillon et al., 2004). IL-31 receptor A is predominantly found on the sensory dorsal root ganglion, while oncostatin M has been linked to pruritus by its association with primary localized cutaneous amyloidosis, an extremely pruritic disease (Sonkoly et al., 2006). IL-31 was found to induce itch when overexpressed in transgenic mice (Dillon et al., 2004). © 2015 European Pain Federation - EFICâ
Malignancy-associated pruritus
Additionally, IL-31 mRNA is elevated in pruritic skin diseases such as atopic dermatitis, allergic contact dermatitis and prurigo nodularis (Sonkoly et al., 2006). An additional study found serum levels of IL31 correlated with the severity of atopic dermatitis (Raap et al., 2008), thus demonstrating that IL-31 is not only associated with pruritic skin disease, but the intensity of itch experienced by patients is also linked to serum levels of IL-31. IL-31 has recently shown promise in elucidating the pathophysiology of CTCL pruritus. Serum levels of IL-31 were found to be elevated in patients with CTCL when compared to healthy control subjects (Ohmatsu et al., 2012). Given the correlation of IL31 to pruritic skin lesions and CTCL, a recent case– control study set out to determine if IL-31 serum levels correlated with the presence of itch in CTCL patients. Researchers compared serum levels of IL-31 in patients with CTCL with pruritus to CTCL patients without concurrent itch and found a significantly elevated IL-31 level in the pruritic group (p = 0.008) (Singer et al., 2013). A subsequent study found that reducing IL-31 using histone deacetylase inhibitors, prednisone or anti-CCR4 monoclonal antibodies in advanced stage CTCL patients significantly reduced pruritus (Cedeno-Laurent et al., 2015). However, a previous study did not find a link between serum levels of IL-31 and itch intensity in patients with early stage CTCL (Malek et al., 2015). The study authors suggest that these contradictory findings may be reconciled by the different mechanisms by which IL-31 is produced in early- and late-stage CTCL. Early-stage CTCL has a predominance of Th1produced cytokines, while IL-31 is produced mainly by Th2 lymphocytes in late-stage disease. Production of IL-31 by two separate pathways could help explain its association with itch in late-stage CTCL, while having no relationship with pruritus in earlystage disease. We recently studied the expression of IL-31 and its receptor in the skin of CTCL patients. A comparison of CTCL patients with mild and severe itch uncovered a significant correlation with keratinocyte expressions of IL-31 and its receptor to itch severity (Nattkemper et al., unpublished data). More research is needed to investigate the potential role of IL-31 and mechanism of action in CTCL and malignancy-associated itch.
6. Treatment The most effective treatment for malignancy-associated itch is to treat the underlying malignancy. Large-scale randomized clinical trials have yet to be Eur J Pain 20 (2016) 19--23
21
Malignancy-associated pruritus
done on treatments for paraneoplastic itch, but smaller studies have demonstrated the efficacy of a number of therapies. One randomized clinical trial demonstrated a significant reduction in paraneoplastic itch intensity when given paroxetine 20 mg daily compared to placebo (Zylicz et al., 2003). Another case report detailed successfully treating paraneoplastic itch secondary to Hodgkin’s lymphoma with thalidomide (Goncalves, 2010). A nightly thalidomide dose of 200 mg was used after a 22-year-old woman with nodular sclerosis type II Hodgkin lymphoma had failed multiple therapies for pruritus. Her rated itch severity on a 10-point VAS improved from ratings of 8–10 prior to therapy to an itch severity of 3 after initiation of thalidomide therapy. Special care must be taken in prescribing thalidomide to women of childbearing age due to its well-known teratogenic effect, which can be seen after just one dose. Peripheral neuropathy from prolonged use of more than 3 months is another adverse effect, which has a reported 2-year incidence of 25–56% of patients treated with thalidomide (Bastuji-Garin et al., 2002). Kappa-opioid agonists and mu-opioid antagonists have known anti-pruritic effects (Cowan et al., 2015; Ko, 2015). This has lead to use of butorphanol, a mixed kappa-opioid agonist and mu-opioid antagonist, for the treatment of paraneoplastic itch. One case report details a 42-year-old woman who suffered from pruritus secondary to non-Hodgkin’s lymphoma and failed multiple treatments for itch. She noted improvement in itch after a single dose of butorphanol (1 mg) and her dose was titrated to 4 mg per day for improved itch control (Dawn and Yosipovitch, 2006). Gabapentin and mirtazapine have been reported to be efficacious in the treatment of itch for CTCL (Demierre and Taverna, 2006). Study authors suggest a starting gabapentin dose of 300 mg nightly and titrating up to maximum of 2400 mg per day, and a starting dose of 7.5 mg for mirtazapine administered at night, titrating to a maximum dose of 15 mg nightly. A prednisone taper starting at 40 mg and tapering over 3 weeks has reported effectiveness in alleviating pruritus in lymphoma patients (Wang and Yosipovitch, 2010). Two novel classes of therapeutic agents, histone deacetylase inhibitors (HDACis) and monoclonal antibodies against chemokine receptor type-4 (CCR4), have demonstrated efficacy in treating CTCL pruritus by targeting the malignant IL-31-producing T cells. Vorinostat and romidepsin are HDACis recently approved by the Food and Drug Administration for the treatment of CTCL. The mechanism of 22 Eur J Pain 20 (2016) 19--23
B. Rowe and G. Yosipovitch
action of HDACis is through their disruption of histone formation, a key process in the production of chromatin and a vital step needed for rapidly dividing malignant cells. Vorinostat and romidepsin have shown efficacy in reducing pruritus in patients with CTCL (Cedeno-Laurent et al., 2015). Mogamulizumab, a monoclonal antibody against CCR4, was recently approved in Japan for the treatment of CTCL. CCR4 is found on malignant IL-31producing T cells in CTCL, and blockade of this receptor with mogamulizumab has been shown to reduce pruritus in CTCL (Cedeno-Laurent et al., 2015). HDACis and mogamulizumab could serve as potential treatments of CTCL and associated pruritus.
7. Conclusion This mini-review provided recent data on malignancy-associated pruritus. Itch can present as the first paraneoplastic sign of an underlying malignancy or as a correlate to disease progression. We reviewed recent data, which demonstrated non-melanoma skin cancer to be the most prevalent form of malignancy-associated pruritus. The pathophysiology behind malignancy-associated itch remains elusive, but research is currently being conducted to further elucidate this process. A better understanding of the pathophysiology of malignancy-associated pruritus will lead to improved therapies in the future. Author contributions B.R. was involved in manuscript writing and G.Y. was involved in manuscript preparation and writing.
References Bastuji-Garin, S., Ochonisky, S., Bouche, P., Gherardi, R.K., Duguet, C., Djerradine, Z., Poli, F., Revuz, J., Thalidomide Neuropathy Study Group (2002). Incidence and risk factors for thalidomide neuropathy: a prospective study of 135 dermatologic patients. J Invest Dermatol 119, 1020–1026. Cedeno-Laurent, F., Singer, E.M., Wysocka, M., Benoit, B.M., Vittorio, C.C., Kim, E.J., Yosipovitch, G., Rook, A.H. (2015). Improved pruritus correlates with lower levels of IL-31 in CTCL patients under different therapeutic modalities. Clin Immunol 158, 1–7. Cowan, A., Kehner, G., Inan, S. (2015). Targeting itch with ligands selective for kappa opioid receptors. In Pharmacology of Itch. Cowan, A., Yosipovitch, G., eds. (Berlin Heidelberg: Springer) 291– 314. Dawn, A.G., Yosipovitch, G. (2006). Butorphanol for treatment of intractable pruritus. J Am Acad Dermatol 54, 527–531. Demierre, M.F., Taverna, J. (2006). Mirtazapine and gabapentin for reducing pruritus in cutaneous T-cell lymphoma. J Am Acad Dermatol 55, 543–544. Dillon, S.R., Sprecher, C., Hammond, A., Bilsborough, J., RosenfeldFranklin, M., Presnell, S.R., Haugen, H.S., Maurer, M., Harder, B., Johnston, J., Bort, S., Mudri, S., Kuijper, J.L., Bukowski, T., Shea, P.,
© 2015 European Pain Federation - EFICâ
B. Rowe and G. Yosipovitch
Dong, D.L., Dasovich, M., Grant, F.J., Lockwood, L., Levin, S.D., LeCiel, C., Waggie, K., Day, H., Topouzis, S., Kramer, J., Kuestner, R., Chen, Z., Foster, D., Parrish-Novak, J., Gross, J.A. (2004). Interleukin 31, a cytokine produced by activated T cells, induces dermatitis in mice. Nat Immunol 5, 752–760. Fett, N., Haynes, K., Propert, K.J., Margolis, D.J. (2014). Five-year malignancy incidence in patients with chronic pruritus: A population-based cohort study aimed at limiting unnecessary screening practices. J Am Acad Dermatol 70, 651–658. Gobbi, P.G., Attardo-Parrinello, G., Lattanzio, G., Rizzo, S.C., Ascari, E. (1983). Severe pruritus should be a B-symptom in hodgkin’s disease. Cancer 51, 1934–1936. Goncalves, F. (2010). Thalidomide for the control of severe paraneoplastic pruritus associated with hodgkin’s disease. Am J Hosp Palliat Care 27, 486–487. Johannesdottir, S.A., Farkas, D.K., Vinding, G.R., Pedersen, L., Lamberg, A., Sorensen, H.T., Olesen, A.B. (2014). Cancer incidence among patients with a hospital diagnosis of pruritus: A nationwide Danish cohort study. Br J Dermatol 171, 839–846. Kilic, A., Gul, U., Soylu, S. (2007). Skin findings in internal malignant diseases. Int J Dermatol 46, 1055–1060. Ko, M.C. (2015). Neuraxial opioid-induced itch and its pharmacological antagonism. In Pharmacology of Itch. Cowan, A., Yosipovitch, G., eds. (Berlin Heidelberg: Springer) 315–335. Malek, M., Glen, J., Rebala, K., Kowalczyk, A., Sobjanek, M., Nowicki, R., Ruckemann-Dziurdzinska, K., Sokolowska-Wojdylo, M. (2015). Il-31 does not correlate to pruritus related to early stage cutaneous Tcell lymphomas but is involved in pathogenesis of the disease. Acta Derm Venereol 95, 283–288. Mandal, S., Varma, K., Jain, S. (2007). Cutaneous manifestations in non-Hodgkin’s lymphoma. Acta Cytol 51, 853–859. Mills, K.C., Kwatra, S.G., Feneran, A.N., Pearce, D.J., Williford, P.M., D’Agostino, R.B., Yosipovitch, G. (2012). Itch and pain in nonmelanoma skin cancer: Pain as an important feature of cutaneous squamous cell carcinoma. Arch Dermatol 148, 1422–1423. Ohmatsu, H., Sugaya, M., Suga, H., Morimura, S., Miyagaki, T., Kai, H., Kagami, S., Fujita, H., Asano, Y., Tada, Y., Kadono, T., Sato, S. (2012). Serum IL-31 levels are increased in patients with cutaneous T-cell lymphoma. Acta Derm Venereol 92, 282–283. Polat, M., Oztas, P., Ilhan, M.N., Yalcin, B., Alli, N. (2008). Generalized pruritus: A prospective study concerning etiology. Am J Clin Dermatol 9, 39–44.
© 2015 European Pain Federation - EFICâ
Malignancy-associated pruritus
Raap, U., Wichmann, K., Bruder, M., Stander, S., Wedi, B., Kapp, A., Werfel, T. (2008). Correlation of IL-31 serum levels with severity of atopic dermatitis. J Allergy Clin Immunol 122, 421–423. Rubenstein, M., Duvic, M. (2006). Cutaneous manifestations of Hodgkin’s disease. Int J Dermatol 45, 251–256. Shi, H.Z. (2004). Eosinophils function as antigen-presenting cells. J Leukoc Biol 76, 520–527. Siegel, F.P., Tauscher, J., Petrides, P.E. (2013). Aquagenic pruritus in polycythemia vera: Characteristics and influence on quality of life in 441 patients. Am J Hematol 88, 665–669. Singer, E.M., Shin, D.B., Nattkemper, L.A., Benoit, B.M., Klein, R.S., Didigu, C.A., Loren, A.W., Dentchev, T., Wysocka, M., Yosipovitch, G., Rook, A.H. (2013). IL-31 is produced by the malignant T-cell population in cutaneous T-cell lymphoma and correlates with CTCL pruritus. J Invest Dermatol 133, 2783–2785. Sonkoly, E., Muller, A., Lauerma, A.I., Pivarcsi, A., Soto, H., Kemeny, L., Alenius, H., Dieu-Nosjean, M.C., Meller, S., Rieker, J., Steinhoff, M., Hoffmann, T.K., Ruzicka, T., Zlotnik, A., Homey, B. (2006). IL31: A new link between T cells and pruritus in atopic skin inflammation. J Allergy Clin Immunol 117, 411–417. Vij, A., Duvic, M. (2012). Prevalence and severity of pruritus in cutaneous T cell lymphoma. Int J Dermatol 51, 930–934. Wang, H., Yosipovitch, G. (2010). New insights into the pathophysiology and treatment of chronic itch in patients with endstage renal disease, chronic liver disease, and lymphoma. Int J Dermatol 49, 1–11. Weisshaar, E., Weiss, M., Mettang, T., Yosipovitch, G., Zylicz, Z. (2015). Paraneoplastic itch: An expert position statement from the special interest group (SIG) of the international forum on the study of itch (IFSI). Acta Derm Venereol 95, 261–265. Yosipovitch, G. (2010). Chronic pruritus: A paraneoplastic sign. Dermatol Ther 23, 590–596. Yosipovitch, G., Bernhard, J.D. (2013). Clinical practice. Chronic pruritus. N Engl J Med 368, 1625–1634. Yosipovitch, G., Mills, K.C., Nattkemper, L.A., Feneran, A., Tey, H.L., Lowenthal, B.M., Pearce, D.J., Williford, P.M., Sangueza, O.P., D’Agostino, R.B. Jr (2014). Association of pain and itch with depth of invasion and inflammatory cell constitution in skin cancer: Results of a large clinicopathologic study. JAMA Dermatol 150, 1160–1166. Zylicz, Z., Krajnik, M., Sorge, A.A., Costantini, M. (2003). Paroxetine in the treatment of severe non-dermatological pruritus: A randomized, controlled trial. J Pain Symptom Manage 26, 1105–1112.
Eur J Pain 20 (2016) 19--23
23
