HLA-Cw6 Status and Treatment Responses Between Psoriasis Patients.

Background: The number of treatment options in psoriasis has increased considerably, so biomarkers should be searched to assist in the selection of the optimal treatment agent. The most common of these biomarkers is HLA-Cw6. Objective: The aim was to determine whether there is a relationship between HLA-Cw6-positivity (HLA-Cw6-POS) and the response of treatment agents in psoriasis.
Methods: Blood samples of 124 patients were subjected to genetic study for HLA-Cw6.
Results: Psoriasis area severity index (PASI) score of more than 75% (PASI75) response was received in 34 (73.9%) of the HLA-Cw6-POS methotrexate-treated patients, 30 (78.9%) of the HLA-Cw6-POS cyclosporine-treated patients, and 8 (37.5%) of the HLA-Cw6-POS acitretin-treated patients. The differences were not statistically significant (P = 0.634-0.071-0.409). PASI75 response was received in 73 (68.2%) of the HLA-Cw6-POS patients in patients treated with conventional agents. In adalimumab-treated patients, PASI75 response was received in 8 (53.3%) of the HLA-Cw6-POS patients, 6 (75%) of the HLA-Cw6-POS infliximab-treated patients, and 4 (57.1%) of the HLA-Cw6-POS ustekinumab-treated patients. The differences were not statistically significant (P = 0.245-1.00-0.322). PASI75 response was received in 24 (64.9%) of the HLA-Cw6-POS and 33 (84.6%) of HLA-Cw6 negative (HLA-Cw6-NEG) patients with biological agents. The biological agent response was statistically significantly lower in HLA-Cw6-POS.
Conclusion: None of the agents were affected by HLA-Cw6. When biological agents were evaluated collectively, the treatment response of HLA-Cw6-POS patients was lower. Copyright:

Introduction

Psoriasis is a chronic inflammatory disease that is characterized by attacks and remission periods. It is suggested that psoriasis is a polygenic disease caused by a polymorphism in both genes involved in immune functions and keratinocyte biology.[1]

The first gene discovered to be associated with a predisposition to psoriasis is HLA-Cw6 in chromosome 6p21.3 in PSORS1.[1]

Henseler and Christophers identified Type 1 and Type 2 psoriasis for the first time. Although the clinical features of these forms are the same, the prognosis, age of onset, pathogenetic mechanisms, and genetic features are different. Type 1 psoriasis has the characteristic of early-onset and positive family history.[2]

The prevalence of HLA-Cw6 worldwide is between 14.1% and 59.1%, but between 10.5% and 77.4% among patients with psoriasis.[3]

Psoriasis treatment aims to eliminate lesions and chronic systemic inflammation, and ensure the patient remains in remission. With the development of biological agents, the number of treatment options has significantly increased. This increase in treatment options has led to the investigation of biomarkers to help select the optimal treatment agent.[4] The most commonly known of these biomarkers is HLA-Cw6. We investigated the clinical demographic characteristics of our patients with psoriasis and the treatment responses, which we obtained with the drugs using in their treatment and HLA-Cw6. The aim was to reveal whether there is a relationship between HLA-Cw6 positivity (HLA-Cw6) and responses of the psoriasis patients.

Material and Methods

Patients

A hospital-based case study was conducted in the Gazi University Faculty of Medicine Clinic of Dermatology from August 2018 to August 2019. One hundred twenty-four patients with psoriasis vulgaris were included. The patients' files were reviewed retrospectively. HLA-Cw6 was genetically studied by using a PCR test. Patients were selected from those who used at least one of the conventional or biological agents for at least 3 months. Those who reached the PASI75 score in the 12th week were considered as responders. The distribution of HLA-Cw6-POS in the group responders or nonresponders was examined pursuant to the type of treatment agent. As well as age, gender, concomitant systemic disease, body mass index (BMI), family history, smoking and alcohol use, disease duration, and age of onset were evaluated.

HLA-C typing

A total of 0.2 μL PCR tube was used for each sample with respect to the number of tests to be performed. A total of 4 μL of DNA was pipetted to the bottom of the PCR tubes. Then, a mixture of 6 μL master Mix (Lifecodes HLA-C eRes, Germany), 10 μL dH2O, 0.2 μL Taq polymerase (Lifecodes, Germany) was pipetted for each tube. The tubes were closed, put into the thermal loop device (Sensequest, Germany), and were run.

At the end of the PCR process, 5 μL PCR product was taken, and the costar plate was pipetted. Based mix (Probe mix) was heated in the heat block at 55°C for 7 min. After sonicating for 15 s, it was vortexed for 15 s. A total of 15 μL Baed mix (Probe mix) was pipetted into all wells of the relevant test. At the end of the pipetting, the top of the plate was covered with adhesive tape. It was put into the thermal loop device for hybridization.

In the last 5 min of the third stage of hybridization, 170 μL dilution solution + 0.85 μL Streptavidin mixture was prepared for each well. When the dilution solution was delivered to the laboratory, it was heated once at 45°C for 5 min and stored at room temperature. At the last stage of hybridization, the tape on the plate was removed, and 170 μL streptavidin was pipetted into the eyes. Read on Luminex-Life-Match (Life Match Fluoroanlyzer Tepnel, USA). It was analyzed in the MATCH-IT DNA program.

Statistics strategy and data analysis

The research data were evaluated via SPSS 22.0. Descriptive statistics were recorded as mean (±) standard deviation, median (min-max) frequency distribution, and percentage. For categorical variables, whether there was a difference in frequency between groups, they were compared using Pearson Chi-square and Continuity correction tests. Independent samples t-test was used in independent groups between two independent groups. All other parameters determined to show no normal distribution were compared between the two groups by the Mann–Whitney U test. The statistical significance value of this study was accepted as P < 0.05

Results

Demographic findings, disease features

A total of 70 male (56.5%) and 54 female (43.5%) patients were included in the study. The age of onset in patients aged between 18 and 65 years (mean = 48.02 ± 13.30 years) was between 4 and 58 years, and the mean was 26.92 ± 13.29 years. The disease duration of patients was between 1 and 53 years, and the mean was 21.1 ± 13.01 years. The BMIs of patients were between 18 and 39, and the mean was 27.57 ± 4.77 kg/m2 [Table 1].

Table 1

Demographic and Disease Information of the Study

	Psoriasis (n=124)
Age (year), avg±SD (min-max)	48.02±13.30 (18-65)
Gender, n (%)	70 (56.5)
Male
Duration of disease (year), median (min-max)	21.1±13.01 (1-53)
Age onset of disease (year), avg±SD (min-max)	27.92±13.29 (1-61)
Body mass index, (kg/m2)	27.57±4.77 (18-39)
Type of Psoriasis, n (%)
Plaque	111 (89.5)
Erythrodermic	9 (7.3)
Pustular	3 (2.4)
Palmoplantar	1 (0.8)
Family History Positivity, n (%)	70 (56.5)
Smoking Habit, n (%)	67 (54)
Alcohol Habit, n (%)	14 (11.3)
Patients with at least one additional disease, n (%)	49 (39.5)
Concomitant diseases, n (%)	68 (100)
Psoriatic arthritis	7 (10.2)
Hypertension	28 (41.2)
Diabetes Mellitus	20 (29.5)
Coronary Heart Disease	5 (7.3)
Dyslipidemia	8 (11.2)

Plaque psoriasis (n = 111, 89.5%) was the most common psoriasis, followed by erythrodermic psoriasis in nine (7.3%) patients, pustular psoriasis in three (2.4%) patients, and palmoplantar psoriasis in one (0.8%) patient. [Table 1].

Seventy-five (60.5%) patients had no concomitant systemic disease. Concomitant diseases, smoking, and alcohol use are shown in Table 1.

HLA-Cw6-POS was present in 50.8% (n = 63) of the patients. Whereas 26.5% (n = 63) of the HLA-Cw subtypes were as follows; 15.5% (n = 37) were found as Cw4-POS and Cw12-POS.

HLA-Cw6 and variables

In the current study, while the average age of the disease onset was 20.94 ± 9.91 (5–58) in HLA-Cw6-POS patients, it was 32.15 ± 13.78 (1–61) in HLA-Cw6- negative (HLA-Cw6-NEG) patients. Statistical analysis of HLA-Cw6-POS patients showed that they tended to be young (P = 0.001). In spite of the fact that HLA-Cw6-POS patients tended to be weaker than HLA-Cw6-NEG, there was no statistical significance (P = 0.81). A total of 33 (52.4%) of the HLA-Cw6-POS patients were male, and 37 (60.7%) of the HLA-Cw6-NEG patients were male. The relationship between the distribution of genders and HLA-Cw6 did not show a statistically significant difference (P = 0.454). A total of 21 of the HLA-Cw6-POS patients (33.3%), and 28 of the HLA-Cw6-NEG patients (45.9%) had concomitant systemic disease. There was no statistical significance found (P = 0.212). In our cases, family history was present in 46 (73%) of the HLA-Cw6-POS patients and 24 (39.3%) of the HLA-Cw6-NEG patients. The existence of psoriasis in the family was statistically higher in HLA-Cw6-POS patients. (P < 0.05) [Table 2].

Table 2

The Relationship of HLA-Cw6 and Variables

	HLA-Cw6-POS	HLA-Cw6-NEG	P
Age of onset (year) avg±SD (min-max)	20.94±9.91 (5-58)	32.15±13.78 (1-61)	0.001*
Body mass index (kg/m2) avg±SD (min-max)	26.84±4.18 (18-39)	28.34±5.23 (18.2-38.6)	0.81£
Gender, n (%)
Male	33 (52.4)	37 (60.7)	0.454¶
Female	30 (47.6)	24 (39.3)
Concominant Disease, n (%)
None	42 (66.7)	33 (54.1)	0.212¶
At least one	21 (33.3)	28 (45.9)
Family history, n (%)
No	17 (27)	37 (60.7)	<0.01¶
Yes	46 (73)	24 (39.3)

*Mann-Whitney U test £T- test ¶Continuity Correction Test

HLA-Cw6 and responses of treatment agents

In patients treated with methotrexate, PASI75 response was received in 34 (73.9%) of the HLA-Cw6-POS patients and 33 (67.3%) of the HLA-Cw6-NEG patients. Even though the rate of response from methotrexate was higher in those with HLA-Cw6-POS, the difference was not statistically significant (P = 0.634) [Table 3].

Table 3

The Relationship of HLA-Cw6 and Treatment Agents

Treatment	Response, n (%)			P ¶
	PASI Status	HLA-Cw6-POS	HLA-Cw6-NEG
Methotrexate	PASI75 (+)	34 (73.9)	33 (67.3)	0.634¶
	PASI75 (-)	12 (26.1)	16 (32.7)
	Total	46 (100)	49 (100)
Cyclosporine	PASI75 (+)	30 (78.9)	18 (58.1)	0.071¶
	PASI75 (-)	8 (21.1)	13 (41.9)
	Total	38 (100)	31 (100)
Acitretin	PASI75 (+)	9 (37.5)	10 (55.5)	0.409¶
	PASI75 (-)	15 (62.5)	8 (44.5)
	Total	24 (100)	18 (100)
Infliximab	PASI75 (+)	6 (75)	7 (77.8)	1.00¶
	PASI75 (-)	2 (25)	2 (22.2)
	Total	8 (100)	9 (100)
Adalimumab	PASI75 (+)	8 (53.3)	12 (80)	0.245¶
	PASI75 (-)	7 (46.7)	3 (20)
	Total	15 (100)	15 (100)
Ustekinumab	PASI75 (+)	4 (57.1)	9 (90)	0.322¶
	PASI75 (-)	3 (42.9)	1 (10)
	Total	7 (100)	10 (100)

¶Continuity Correction Test

In patients treated with cyclosporine, PASI75 response was received in 30 (78.9%) of HLA-Cw6-POS patients and 18 (58.1%) of the HLA-Cw6-NEG patients. Even though the rate of response from cyclosporine was higher in those with HLA-Cw6-POS, the difference was not statistically significant (P = 0.071) [Table 3].

In patients treated with acitretin, PASI75 response was received in 9 (37.5%) of the HLA-Cw6-POS patients and 10 (55.5%) of the HLA-Cw6-NEG patients. Even though the rate of response from acitretin was lower in those with HLA-Cw6-POS, the difference was not statistically significant (P = 0.409) [Table 3].

In patients treated with infliximab, PASI75 response was received in 6 (75%) of the HLA-Cw6-POS patients and 7 (77.8%) of the HLA-Cw6-NEG patients. Even though the rate of response from infliximab was lower in those with HLA-Cw6-POS, the difference was not statistically significant (P = 1.00) [Table 3].

In patients treated with adalimumab, PASI75 response was received in 8 (53.3%) of the HLA-Cw6-POS patients and 12 (80%) of the HLA-Cw6-NEG patients. Even though the rate of response from adalimumab was lower in HLA-Cw6-POS patients, the difference was not statistically significant (P = 0.245) [Table 3].

In patients treated with ustekinumab, PASI75 response was received in 4 (57.1%) of the HLA-Cw6-POS patients and in 9 (90%) of the HLA-Cw6-NEG patients. Even though the rate of response from ustekinumab was lower in those with HLA-Cw6-POS, the difference was not statistically significant (P = 0.322) [Table 3].

HLA-Cw6 and conventional/biologic treatment agents

In patients treated with conventional agents totally, PASI75 response was received in 73 (68.2%) of the HLA-Cw6-POS patients and 61 (62.6%) of the HLA-Cw6-NEG patients. The difference was not statistically significant (P = 0.425). In patients treated with biological agents totally, PASI75 response was received in 24 (64.9%) of the HLA-Cw6-POS patients and 33 (84.6%) of the HLA-Cw6-NEG patients. Treatment response was statistically significantly lower in HLA-Cw6-POS patients (P = 0.047) [Table 4].

Table 4

The Relationship of HLA-Cw6 and Conventional/Biologic Treatment Agent

	HLA-Cw6 (+)-POS	HLA-Cw6 (-)-NEG	P ¶
Conventional
PASI75 (+)	73 (68.2)	61 (62.6)	0.425¶
PASI75 (-)	34 (31.8)	36 (37.4)
Total	107 (100)	97 (100)
Biologic
PASI75 (+)	24 (64.9)	33 (84.6)	0.047¶
PASI75 (-)	13 (35.1)	6 (15.4)
Total	37 (100)	39 (100)

¶Pearson Chi-Square test

Discussion

In the current study, the prevalence of HLA-Cw6 positivity was found in 50.8% of our psoriasis patients, whereas it varied between 14.1% and 59.1% worldwide, 10.5% and 77.2% in patients with psoriasis.[5] The most common HLA-Cw subtype was Cw6 (n = 63, 26.5%), the rest (n = 37, 15.5%) were either Cw4 or Cw12. Similar to our findings, Onsun et al.[6] also showed that the most common identified alleles in Turkish psoriasis patients were HLA-Cw6 (55.7%), HLA-Cw12 (35%), and HLA-Cw4 (32%).

In our study, whereas the average age of the disease onset was 20.94 ± 9.91 (5–58) in HLA-Cw6-POS patients, it was 32.15 ± 13.78 (1–61) in HLA-Cw6-NEG patients (P < 0.05). As expected, statistical analysis showed that the disease in HLA-Cw6-POS patients begins at a younger age, consistent with type 1 psoriasis. Many studies have been carried out in the literature to support our data.[78] Even though HLA-Cw6-POS patients tended to be weaker than HLA-Cw6-NEG patients, there was no statistical significance (P = 0.81). In conflict with our results, Jin et al.[9] reported that the risk is 35 times higher in overweight patients with HLA-Cw6 than a nonoverweight HLA-Cw6-NEG patient. It is obvious that obesity is not affected by the HLA-Cw6 status alone. As a result, our data were found different from other literature. In our study, the rate of HLA-Cw6-POS male patients was found to be lower than that of HLA-Cw6-NEG male patients (P = 0.454). On the contrary, in a study by Fan et al.,[10] it was found that the rate of HLA-Cw6-POS men is higher, which does not support our data. A total of 21 of the HLA-Cw6-POS patients (33.3%) and 28 of HLA-Cw6-NEG patients (45.9%) had concomitant systemic disease. When the relationship of concomitant systemic disease with HLA-Cw6 was evaluated, no statistical significance was found (P = 0.212). These systemic diseases were hypertension, diabetes mellitus, psoriatic arthritis, dyslipidemia, and coronary heart disease. Inflammatory bowel diseases, psychiatric disorders, and malignancies, which are among the other diseases accompanying patients with psoriasis, were not determined in our cases. Due to the retrospective design of our study, the frequency of these diseases could be evaluated to the extent they were recorded, so we think that our data may not reflect the true frequency of the concomitant diseases. Gulliver et al. found that HLA-Cw6-POS patients with psoriasis had a significantly higher rate of hospitalization for neoplasm compared to HLA-Cw6-NEG patients (16.6% vs 8.6%; P = 0.028), whereas HLA-Cw6-NEG patients had a significantly higher rate of hospitalization for circulatory system disease (29.1% vs 19.5%; P = .034)[11] In our study, the rate of HLA-Cw6-POS psoriasis patients with a positive family history was statistically significantly higher than the HLA-Cw6-NEG patients. (P < 0.05) In the literature, there are studies with a large patient population that supports our data.[12]

In our study, at 12 weeks, PASI75 response was maintained in 73.9% of HLA-Cw6-POS, and 67.3% of HLA-Cw6-NEG patients with methotrexate (P = 0.634). Whereas Indhumathi et al.[13] (n = 189) and West et al.[14] (n = 333) reported statistically significant higher treatment responses in HLA-Cw6-POS patients, supporting our study. Yan et al.[15] (n = 222) found that methotrexate's response was irrespective of HLA-Cw6 status in patients with psoriasis.

At 12 weeks, 78.9% of the HLA-Cw6-POS patients and 67.3% of the HLA-Cw6-NEG psoriasis patients achieved PASI 75 response with cyclosporine treatment (P = 0.071). Interestingly, despite the fact that the higher percentage of the HLA-Cw6-NEG psoriasis patients (55.5%) achieved PASI75 responses to acitretin when compared with the HLA-Cw6-POS (37.5%) patients, the difference was not statistically significant. Currently, studies evaluating the relationship between HLA-Cw6 and treatment responses to cyclosporin or acitretin are lacking. Thus, according to our findings, determination of HLA status was not necessary to predict their response to cyclosporine and acitretin.

We found that HLA-Cw6 status is predictive of response to biological treatment. In our study, the rate of response from biological agent treatment in HLA-Cw6-POS patients was statistically significantly lower. To our knowledge, there is no study in which biological agents are evaluated together, but there are few studies searching the relationship between anti-TNFs and HLA-Cw6 status. As evident from the literature, in a study by Batalla et al.,[16] Coto-Segura et al.,[17] and Gallo et al.,[18] anti-TNFs responses tend to be statistically nonsignificant higher in HLA-Cw6-POS patients. However, in a study by Talamonti et al.,[19] HLA-Cw6-POS patients were shown to respond statistically nonsignificant lower to anti-TNF's. Our study has demonstrated the importance of HLA-Cw6 in predicting biological treatment response.

In our study, the rate of response from adalimumab in patients with HLA-Cw6-POS was 53.3%, whereas this ratio was 80% in HLA-Cw6-NEG patients (P = 0.245). In a study by Gallo et al.,[18] supporting our study, adalimumab responses tend to be statistically non-significant lower in HLA-Cw6-POS patients. On the contrary, in a study, the treatment responses of HLA-Cw6-POS patients were found to be statistically significantly higher at 24 weeks. No statistical significance was detected at week 12.[20] In another study, there was a nonsignificant trend suggesting that HLA-Cw6-POS patients were more likely to respond to adalimumab.[21] In this study, it was found that HLA-Cw6 was not predictive of response to adalimumab

In our study, the rate of response from infliximab in HLA-Cw6-POS patients was 75%, whereas this ratio was 77.8% in HLA-Cw6 - NEG patients (P = 1.00). Apart from Talamonti et al.' s study,[19] there are a number of studies in the literature that are contrary to our data. A study by Talamonti et al.[19] supporting our study, showed less response to these drugs in HLA-Cw6-POS patients. In a study by Gallo et al.[18] and Gulliver et al.,[22] unlike to our study, infliximab responses tend to be statistically nonsignificant higher in HLA-Cw6-POS patients In this study, it was found that HLA-Cw6 was not predictive of response to infliximab.

Lastly, in our study, the rate of response from ustekinumab in patients with HLA-Cw6-POS was 57.1%, whereas the ratio was 90% in HLA-Cw6-NEG patients (P = 0.322). In a study by Talamonti et al., Galuzzo et al., Li et al., and Chiu et al., HLA-Cw6-POS patients were shown to respond statistically significantly higher to ustekinumab.[2324252627] However, in a study by Chiu et al.,[28] supporting our study, it was stated that HLA-Cw6 was not an effective factor in evaluating the ustekinumab treatment response. In our study, it was concluded that there was no need to look at the status of HLA-Cw6 in cases treated with ustekinumab.

Conclusion

In our study, it was shown that HLA-Cw6-POS patients responded better to methotrexate and cyclosporine and responded less to ustekinumab, adalimumab, infliximab, and acitretin. The current data show that HLA-Cw6-POS patients have low responses to biological agent treatment statistically significant. When the drugs are evaluated separately, statistical significance is lost due to the small number of patients in each drug group. Our results suggest that new research is needed in which more patients are treated with each agent.

Financial support and sponsorship

Turkish Society of Dermatology.

Conflicts of interest

There are no conflicts of interest.

Approval of Ethics Committee: Yes (Gazi University Ethics Committee)