Health‐related quality of life in cutaneous T‐cell lymphoma: A cross‐sectional survey study

Summary Background Patients with cutaneous T‐cell lymphoma (CTCL) often have indolent but symptomatic disease. Objective Assessment of the health‐related quality of life (HRQoL) of patients with CTCL. Methods Cross‐sectional survey study. HRQoL was measured by Skindex‐16 and FACT‐G. Results A total of 372 responses were received; 80 incomplete/ineligible responses were excluded. A majority of respondents identified as white (87%; 250/288) and female (67%; 193/286) with a mean age of 57 ± 14 years. Most patients had early‐stage (IA–IIA) (74%; 162/203) mycosis fungoides (87%; 241/279). There were 33 (12%; 33/279) patients with Sézary syndrome. Mean itch score (visual analogue scale; VAS) was 3.2 ± 2.8, overall; 2.7 ± 2.6 for early, and 4.2 ± 2.9 for advanced disease (p = 0.008). Thirty‐eight percent (108/284) and 24% (69/284) reported head/neck and groin/genital involvement, respectively. Overall HRQoL was 46 ± 27 (Skindex‐16) and 71 ± 19 (FACT‐G), with worse HRQoL for patients with advanced versus early disease (Skindex‐16: 67 vs. 40; p=<0.001, FACT‐G: 62 vs. 76; p = 0.001). Predictors of worse HRQoL included head/neck, hand/foot or groin/genital involvement, younger age and spending >15 min daily treating CTCL. Limitations Include anonymous survey methodology, underrepresentation of certain CTCL subtypes and non‐white respondents. Conclusions Patients with CTCL, particularly those with advanced disease or involvement of the head/neck, acral or groin/genital sites, experience significant impact on HRQoL.

survivors. 4 For example, survivors of non-Hodgkin lymphoma have worse HRQoL compared to age and sex-matched normative controls. 5,6 The Cutaneous Lymphoma Foundation (CLF) is an independent, non-profit patient advocacy organization dedicated to supporting people affected by cutaneous lymphoma (https://www.clfoundation.org/about-us). A large-scale study done in 2005 by Demierre et al. in partnership with the CLF found that CTCL has a profound impact on HRQoL, with worse HRQoL reported for more advanced disease. 7 We partnered with the CLF to examine QoL in patients with CTCL, with a focus on physical, social and emotional well-being. We hypothesized that despite advances in CTCL therapy since the work of Demierre et al. more than 15 years ago, patients with CTCL still experience lower HRQoL. We also examined the impact of disease stage, location of skin involvement, time since diagnosis and itch severity on HRQoL for those with CTCL.

| MATERIALS AND METHODS
A cross-sectional, anonymous electronic survey was administered between February and April 2019. The survey was distributed via a link posted on the CLF Facebook page and email lists. At the time of distribution, the CLF Facebook group had approximately 1900 members, while the email listerv had approximately 1200 members. The study team did not directly access membership lists. Survey responses were collected and managed using REDCap electronic data capture tools hosted at the University of Washington. 8,9 This study was determined to be exempt from institutional review board review by the University of Washington Human Subjects Division (STUDY00005784).
Independent measures included demographics, CTCL type and stage, time since diagnosis, and comorbid conditions (adapted from the self-reported Charlson Comorbidity Index 10 ). Four independent measures, three of which were similar to those in Demierre's original study and one novel question, assessed the emotional and financial burden of CTCL (collectively termed the 'Burden Score'). Itch was measured using VAS. Two validated HRQoL instruments were included: Skindex-16 11 and Functional Assessment of Cancer Therapy: General (FACT-G). 12 Skindex-16 scores were calculated per Chren et al., 11 transforming responses to a linear scale from 0 to 100. Total score and subscale scores were considered valid if at least 70% of the items had responses. Missing items in each subscale were imputed with the mean of the non-missing items in that subscale. FACT-G scores were calculated according to the FACT-G Scoring Guidelines Version 4. Scores were calculated for each subscale and summed together to derive the total score, with a range of 0-108 (with lower scores indicating worse HRQoL). Subscale scores were considered valid if >50% of the items had responses (i. e., ≥4 of 6 or 7 items per subscale). The total score was considered valid if >80% (≥22 of 27 items) had responses and all subscales were valid. Missing items in each subscale were imputed using the mean of nonmissing items in that subscale.
HRQoL scores and demographic variables were compared between these groups using the Wilcoxon rank-sum test or Fisher's exact test. Linear regression models were used to examine mean differences in Skindex-16 or FACT-G. p-Values less than 0.05 were considered statistically significant. Β-values are mean differences in Skindex-16 or FACT-G per change in the corresponding variable; estimates were derived using linear regression models. All statistical analyses were performed using STATA (version 14.0; StataCorp) and R software (version 4.0.0; R Foundation for Statistical Computing).
Mean overall Skindex-16 and FACT-G scores and scores for MF/SS by stage are shown in Table 3. Association of HRQoL as measured by Skindex-16 and FACT-G is shown in Tables 4 and 5, respectively. Age, advanced stage, higher itch scores by VAS, head/neck, groin/genital or acral involvement were all significantly associated with worse HRQoL in both instruments, as were requiring help from a caregiver, spending >15 min daily treating CTCL (vs < 15 min daily), each of the independent measures of the Burden Score and the overall Burden Score. Time since diagnosis of CTCL (within 1 year vs. longer) was not significantly associated with HRQoL by either instrument. Patients who reported groin/genital involvement had lower satisfaction with their sex life as assessed by FACT-G, with 46% (25/54) of patients with groin/genital involvement reporting they are 'not at all' satisfied with their sex life compared to 25% (41/161) of patients with other body areas involved (p = 0.006) (data not included in tables).

| DISCUSSION
Overall HRQoL among our cohort of patients with CTCL as assessed by Skindex-16 is similar compared to patients with other chronic dermatologic diseases, including eczematous dermatitis and psoriasis. 13 Our cohort reported worse HRQoL as measured by FACT-G compared to CTCL patients from previous studies by Demierre et al. 14 The inclusion of the FACT-G instrument in our survey also allows us to contrast our results with other cancer patients. Our cohort had worse HRQoL scores than long-term survivors of indolent and aggressive non-Hodgkin lymphoma. 15 The mean total FACT-G score in our patients was equivalent to cancer patients with an Eastern Cooperative Oncology Group Performance Status Rating (ECOG PSR) category of 2 ('require bed rest for <50% of waking day'). 16 As a disorder that spans chronic skin disease and cancer, the optimal instrument for assessing overall HRQoL for CTCL is not clear. There is variability in the existing CTCL HRQoL literature around instruments used, with most groups using the Skindex-29, 1,14,17-19 but FACT-G 14 and EORTC QLQ-C30 18 have also been used. We chose to use the shorter Skindex-16 given the high validity in other skin diseases, 11 which impacts the direct comparability of our numerical results but not trends or associations. We found similar associations and trends between the Skindex-16 and FACT-G. One important distinction is as a cancer-specific instrument, the FACT-G does not specifically assess skin symptoms such as itching. This was evident when comparing the overall HRQoL scores between patients with early and advanced disease across subscales, in particular, the symptoms subscale of the Skindex-16, for which there was a nearly twofold difference in scores (29 and 50 for early and advanced disease, respectively) compared to a smaller difference in the physical well-being subscale of the FACT-G (23 and 20 for early and advanced disease, respectively). We recommend that when HRQoL of patients with CTCL is assessed using FACT-G, additional questions around skin symptoms, particularly itch, are added. Alternatively, validation of a CTCL-specific FACT instrument could address these concerns, while maintaining the benefits of comparability with the existing data.   Patients with more advanced CTCL had worse HRQoL compared to those with early disease in this study. The association with stage was significant for both the Skindex-16 (p < 0.001) and FACT-G (p = 0.001) instruments, even when adjusted for age, gender and number of medical comorbidities (data not shown). These results are consistent with clinical experience and prior studies, suggesting patients with later-stage disease generally experience more symptoms and greater morbidity than those with earlystage disease. 1,7,17,19 Demierre et al. reported that patients with more advanced CTCL reported more effects on general health, particularly in the physical, emotional and functional domains of the FACT-G. 14 In our cohort, patients with advanced disease had statistically significantly worse HRQoL on all of the Skindex-16 subscales (symptoms, emotions and functioning), but only on two of the four FACT-G subscales (physical and functional well-being), suggesting that later disease stage most severely impairs patients' overall physical health and ability to perform daily tasks, while the emotional impact is high regardless of stage.
We hypothesized that patients who were newly diagnosed with their CTCL might experience worse HRQoL, positing that patients who had more time to acclimate to their diagnosis might fare better. This did not prove to be the case, with no significant difference in HRQoL between those diagnosed <1 year, 1--5 years or >5 years. We found that the time spent treating CTCL, in particular, spending >15 min daily, was associated with worse HRQoL. Spending >60 min daily treating CTCL had a large impact on HRQoL (Tables 4 and 5). Asking patients about the amount of time spent daily treating their CTCL may provide important insight into their disease-related HRQoL.
The mean VAS itch score in our cohort was 3.2 on a 10-point scale, representing moderate pruritus. 20 Similar to others, 1,17 we found that higher itch scores were significantly associated with worse HRQoL on the Skindex-16 (r s = 0.56; p < 0.001) and FACT-G (r s = −0.36; p < 0.001) instruments. Existing literature supports that patients with chronic pruritus have lower HRQoL, and that pruritus may contribute to other symptoms (e.g., sleep disturbance, depression) which further compound the problem. 21 It is important to note that our cohort had slightly lower VAS scores compared to a larger cohort of CTCL patients (mean of 4.2). 22 This suggests our data might underestimate HRQoL in the general CTCL population.
Patients who reported involvement of their CTCL in the head/neck, hands/feet or groin/genitals had significantly worse HRQoL compared to those with other body parts affected (Tables 4 and 5). This association was seen globally and across all three Skindex-16 subscales (p < 0.001 for each), as well as the global and physical and functional well-being subscales on the FACT-G (p < 0.001 for each) (Tables 4 and 5). Our findings add to those of the Prospective Cutaneous Lymphoma International Prognostic Index (PROCLIPI) study, which demonstrated a worse QoL in patients with alopecia. 19 We found that nearly one-quarter (24%) of patients had involvement of their CTCL in the groin/genitals. Previous studies have reported on the impairment on sexual life reported by patients with CTCL, 18 and we confirm that worse HRQoL is reported by patients with
a Regression coefficient, corresponding to the mean change in quality of life score per change in the associated variable.
b Dichotomized as sometimes, often or very often versus rarely or never (reference); see Table 2 for the text of the original question.
SHINOHARA ET AL. groin/genital involvement by their CTCL. This finding of worse HRQoL with groin/genital involvement is not surprising when put in the context of the impact on HRQoL of other skin conditions that impact the groin/ genitals. Groin/genital involvement in patients with hidradenitis suppurativa (HS) is associated with decreased HRQoL generally, 23 and patients with HS report lower sexual health and sexual function. 24 Our findings reinforce the importance of asking patients about and evaluating for the involvement of CTCL in the groin/genitals, as focusing on the treatment of these areas could significantly impact patient's sexual health and overall HRQoL.
Based on the findings of Demierre's group, 7 we were inspired to further evaluate the themes from the independent measures in that study along with one additional measure ( Table 2). The Burden Score questions address possible stigma or shame that patients with CTCL might experience (such as needing to keep the condition private and being treated unfairly because of CTCL). Collectively, the Burden Score was significantly associated with overall HRQoL by both instruments used, with an impact on par with the stage of disease (Tables 4 and 5). This suggests that patients with CTCL may experience significant emotional distress from shame and stigma associated with their disease which, when added to the impact of physical symptoms such as itching and poor sexual function, contribute to worse overall quality of life, and studies to further validate the Burden Score are warranted.
This study is one of the largest describing HRQoL in CTCL, 7,17-19 though there are factors which might limit the generalizability of our findings. Given the rarity of CTCL, this survey was distributed electronically via disease-specific social media groups and email listservs with overlapping membership, making calculation of a response rate not feasible. Our sample was of a similar age (mean 57 years) to the average age of CTCL patients 25 ; however, our patient group had more who identified as white and female than the overall population of patients with CTCL 26 (though in similar proportions to that of Demierre et al. 7 ). There are known racial disparities in the age of presentation and outcomes of non-white patients with CTCL [27][28][29] ; given the low overall number of nonwhite respondents in our group we could not assess racial disparities in HRQoL. Our sample also contains fewer patients with CD30 LPDs compared to the known frequency of these disorders, 25 possibly because this group of patients may not be as active in support groups. Importantly, the data in this study are patient-reported, and subject to bias both at the level of self-selection and self-reporting. Lastly, we did not ask about specific treatments for CTCL and cannot comment on the relationship of particular treatments to HRQoL. Dichotomized as sometimes, often or very often versus rarely or never (reference); see Table 2 for the text of the original question. SHINOHARA ET AL.