Biologic therapies in rheumatic diseases: drug and anti-drug antibody levels and clinical efficacy

Background: The aim of this study was to evaluate the relevance of drug and anti-drug antibody detection in the clinical management of patients with rheumatoid arthritis (RA) and spondyloarthropaties (SpA) in treatment with antitumor necrosis factor alpha (TNFα) biologics. Methods and results: The study included 192 adult consecutive patients treated for at least 6 months with adalimumab (ADA) or etanercept (ETN) or infliximab (IFX); patients underwent clinical observations in the Rheumatologic Unit of 5 Hospitals in Tuscany. Their demographic and clinical characteristics to calculate DAS28 and BASDAI scores were collected. Drug levels and anti-drug antibodies (anti-drug Ab) were evaluated immediately before drug injection or infusion. A total of 192 patients were studied: 62 receiving IFX, 64 ADA, and 66 ETN with a mean age of 57 years (range 18-86 years); the study group was composed of 51% women. Forty percent of the patients were affected by RA, 60% by SpA. Altogether, 81% of patients demonstrated therapeutic drug levels. Anti-drug Ab were found in 19% of patients taking IFX, 10% taking ETN and 5% taking ADA. No significant correlation was found between anti-drug Ab presence and low drug levels, between anti-drug Ab and high DAS28 and BASDAI scores, as well as between low drug levels and high DAS28 and BASDAI scores. Conclusions: Low drug levels were found in 19% of the rheumatic patients and there were not correlations with presence of anti-drug Ab or patient’s clinical status.


Introduction
ADA, ETN and IFX, all TNFα blockers, actually represent an important therapeutic aid for RA and SpA patients. In a third of cases such drugs are ineffective and this is due to primary treatment failure (drug inefficacy or serious side effects) or secondary treatment failure (drug loss of efficacy after initial good response). It has been shown that serum levels of ADA, ETN and IFX correlate with the clinical response in RA, while in ankylosing spondylitis (AS) data are controversial [1][2][3][4]. Similarly, anti-drug Ab levels inversely correlate with drugs levels and therapeutic response [4][5][6]. The mechanism of drug inactivation is linked to anti-drug Ab formation (against biological medications) which causes the neutralization of the drug and its increased clearance [5]: only 4% of patients with anti-adalimumab antibodies (anti-ADA Ab) achieved clinical remission compared to 34% anti-ADA Ab negative patients [6]. ETN has shown a less immunogenic effect with anti-etanercept antibodies (anti-ETN Ab) not detectable or present only in a low number of patients; as a consequence anti-ETN immune response has a minimal effect on clinical response [4,[7][8][9]. A clinically relevant impact of anti-drug Ab is described in about half of the patients receiving repeated TNFα monotherapy and therefore immune suppression by concomitant administration of metrotrexate (MTX) is indicated in RA and SpA patients [8,[10][11][12][13][14][15][16][17][18][19][20][21]. Detectable anti-drug Ab decrease TNFα blockers response by as much as 80% [22]. This has been shown with only one doi: 10.7243/2054-989X-4-1 or two biologic treatments without comparing differences in patients suffering from different inflammatory diseases.
The aims of our study were: a) to assess the relationship between clinical response and TNFα blockers serum levels and anti-drug Ab concentrations in RA and SpA (AS and psoriatic arthritis (PsA)) patients treated with ADA, ETN, and IFX b) to verify if a single serological determination could be useful in clinical practice for patient management.

Study population
Between September 2014 and February 2015, consecutive patients with RA and SpA, that were treated with anti-TNFα biologics were invited to participate to the present study. Patients were followed by the Rheumatology Units of 5 Tuscany hospitals participating in the study. The study included 192 adults ≥18 years who were treated with either ADA, IFX or ETN monotherapy or concomitant MTX therapy. Exclusion criteria were: patients treated with a different TNFα blocker or for less than 6 months. Written consent was obtained from all participants and the study was approved by the Ethics Committee of our hospital. Clinical data and blood samples were collected from patients just before they received a new intravenous or subcutaneous dose of the anti-TNFα agents. Samples were kept a -20°C until processed.

Assessment of disease activity
RA disease activity was assessed using the 28-joint disease activity score (DAS28) (swollen joint count, tender joint count, ESR and general health) as described elsewhere [23]. The disease activity of SpA was assessed using the Bath ankylosing spondylitis disease activity index (BASDAI) on a 100-mm visual analogue scale as previously described [24].

Definition of clinical response
The clinical response to anti-TNFα treatment in RA patients was assessed using the EULAR response criteria [25]

Dosage of the anti-TNFα agents
The dosage of IFX was 3 mg/Kg per dose for RA and 5 mg/Kg for SpA, given as intravenous infusion at baseline, week 2, 6, and then every 6-8 weeks. ETN was given subcutaneously at 50 mg per dose every week. ADA was given subcutaneously at 40 mg per dose every 2 weeks.

Drug concentration and anti-drug antibody assays
Serum drug levels and anti-drug Abs were evaluated immediately before next drug injection or infusion. Concentrations of free IFX and ADA were measured in serum samples by IDK monitor Infliximab or Adalimumab drug level ELISA, respectively, according to the manufacturer's instructions (Immunodiagnostik AG, Bensheim, Germany). Concentrations of serum free ETN were measured by Shikari Q-ETA (Matriks Biotek Laboratories, Israel). Briefly, standards and serum samples were incubated in the microtitre plate coated with the specific monoclonal anti-IFX or ADA or ETN antibody. Following incubation wells were washed and peroxidaselabeled anti-IFX or ADA or ETN specific antibody was added. After incubation, wells were washed and the bound enzymatic activity was detected by addition of chromogenic substrate. A dose response curve, specific for each biologic, was generated using the values obtained from standards. The concentrations of free biologic in the samples were determined directly from the curve. Results were express as ng/ml for ADA and IFX and as µg/ml for ETN. The ELISA detection limits were: 2.3 ng/ ml and 2.6 ng/ml for IFX and ADA assays respectively, and 0.2 µg/ml for ETN assay. No cross reactivity to other plasma proteins or anti TNFα-blockers, different from the investigated one, was reported.
The presence in serum samples of free human antibodies against IFX, ADA and ETN was detected by IDK monitor Infliximab or Adalimumab total ADA ELISA and IDK monitor Etanercept free ADA ELISA, respectively (Immunodiagnostik AG, Bensheim, Germany). Briefly, in the first incubation step the free anti-drug antibodies from the sample were bound to the drug coated on the plate. To remove all unbound substances, a washing step was carried out. In a further incubation step, peroxidase-labeled therapy antibody was added. After another washing step, to remove all unbound substances, the solid phase was incubated with chromogenic substrate. The optical density was directly proportional to the amount of bound anti-drug Ab from sample. Results were evaluated by a cut-off control. Samples with optical density higher than that of the specific drug cut off were considered positive.

Statistical analysis
Unless otherwise indicated, values were expressed as mean± standard deviation. Comparison of continuous variables between two groups was performed by the non-parametric Mann-Whitney rank sum test. Discrete variables were compared, by group, using chi square test. When the frequency was <5, the Fisher's exact test was used. Statistical significance was defined as a two-tailed p value of <0.05.

Characteristics of the study population
The study enrolled a total of 192 patients of which 51% were women. The mean age at the time of blood collection was 60±12 years for women and 55±13 years for men.
The underlying diseases of these patients that warranted anti-TNFα treatment were: RA (N=77; 40%) and SpA (N=115; 60%). The proportion of patients treated with IFX, ADA and ETN, was 32% (N=62), 33% (N=64) and 35% (N=66), respectively. The mean duration of therapy at the time of blood collection was 56±35, 42±36 and 57±39 months for IFX, ADA and ETN users, respectively. Table 1 shows the gender and age distribution of the studied patients. As shown, male subjects are predominant in the 18-45 age group, representing 66.7% of the patients, while an opposite figure is found in the >67 age group where women represent 60.9% of the patients. No difference in gender distribution is observed in the 46-66 age group. Subdividing the 18-45 and >67 age groups by gender and disease ( Table 2), it was observed that in the 18-45 age group the prevalence of SpA is significantly higher (p=0.01) than RA in male patients, while in the >67 years group the prevalence of RA is lightly higher (64.3%) in female patients. Taking into account gender and disease distribution of the 192 patients, women represent 64.9% (50 out 77) of RA patients with a 1.9:1 female/male ratio.

Drug and anti-drug antibody in RA and SpA patients
The relationship between serum drug levels and anti-drug Ab presence for each drug treatment was evaluated in both RA and SpA patients ( Table 3). Antibodies against IFX were demonstrated in 12 (19.4%) patients, 5 RA and 7 SpA patients; anti-ADA Ab were detected only in 3 (4.7%) patients all belonging to the RA group; anti-ETN Ab were present in 6 (9%) patients, 4 RA and 2 SpA. As shown in Table 3, patients who developed anti-drug Ab don't had drug levels significantly different from patients who did not developed anti-drug Ab.

Serum drug level, anti-drug Ab status and clinical response in RA and SpA patients
In order to analyze the relationship between serum drug levels and clinical response, RA and SpA patients were divided  into three groups according to DAS28 and BASDAI scores, respectively. As shown in Table 4, there were no significant differences in serum drug levels between subjects of the three DAS28 and BASDAI categories both in RA and SpA patients. Moreover, for each drug treatment no significant difference was observed in the clinical response of anti-drug Ab positive and negative subjects in each disease group ( Table 5).  Table 4. Relationship between serum drug levels and clinical efficacy in between RA and SpA patients.

Serum drug level, anti-drug Ab status and clinical response in RA and SpA patients treated with anti-TNFα monotherapy or concomitant MTX therapy
In order to verify the effects of the concomitant use of MTX, RA and SpA patients were divided into MTX users and MTX non-users, and serum drug levels, anti-drug Ab status and clinical response were analyzed (

Discussion
TNF blockers represent a breakthrough in the management of rheumatic diseases, including RA and SpA, having a relevant effect on their clinical course and prognosis. These drugs, however, have some limitations due to primary or secondary loss of efficacy. The latter maybe determined by the phenomenon of immunogenicity [26,27] which lowers the blood levels of effective drug and consequently decreases the clinical response. The aim of this study was to evaluate the relevance of drug and anti-drug antibody detection on the clinical management The study population included 192 patients of which 98 were females and 94 males. Fifty-two % of female population was affected by RA and 48% by SpA, while only 28.8% of male patients was RA; the incidence of RA, as expected [28], is higher in the female population with a ratio female/male 2:1. When the population was subdivided in age groups, it was observed that in the intermediate age groups  of 67% and in the ≥67 years females reached 61%. In the 18-45 group the SpA is the disease with the highest prevalence of 79.5%; in particular, in the male population SpA is significantly higher (74% p=0.01) than RA (26%) while in females the two diseases are roughly equivalent. Moreover the female:male ratio of patients affected by SpA is 1:3 in agreement with data from the literature that show a higher prevalence of SpA in young males with a ratio female:male of 1:3 [29]. In the ≥67 age group females and males are 60.9% and 39.1% respectively. In the female group RA shows a prevalence of 64.3% and SpA 35.7%. A higher prevalence of RA in females is reported in the literature: in women RA develops between 30-50 years with an incidence 3-5 times higher in females than males and this incidence tends to decrease with increasing age [29]. Therefore our results demonstrate that the study population is epidemiologically representative of the two autoimmune diseases both for sex and for age distribution.
In the literature, the percentage of patients who develop anti-drug Ab may vary by sex, different autoimmune inflammatory diseases and among different anti-TNF treatments. In this study, patients with anti-drug Ab were 21/192 (10.9%), 8 females and 13 males (data not shown) indicating that there is no correlation between drug immunogenicity and sex of the patient in contrast with what it was described by Mok [5] who reported a higher prevalence of anti-drug Ab in females than in males.
Anti-drug Ab are generally reported as detected in up to one third of RA and about 25% of SpA patients [2,[30][31][32][33]. Chimericdrugs (mouse-human), such as IFX, have a greater likelihood of inducing anti-drug Abs production compared to fully human antibodies [34,35]. Not all patients treated with anti-TNF therapy develop anti-drug Ab and this seems to be multi-factorial: the treatment, the patient, external factors [33]. We have found that the phenomenon of immunogenicity is present for all used drug treatments in RA patients while in SpA patients is limited to IFX and ETN. Anti-IFX Ab were present in 28% of RA and 16% of SpA patients in accordance with previous studies revealing anti-IFX Ab from 12 to 44% in RA [36] and from 6 to 61% in SpA [37]. Anti-ETN Ab were found in 12.5% of RA and 6% of SpA patients and anti-ADA Ab in 11% of RA patients; ETN is considered the less immunogenic one and only few studies [38] report the presence of anti-ETN Ab in SpA patients. In our hands, on the contrary, ADA is the treatment that shows the lowest rate of anti-drug Ab (4.7%) compared to IFX (19.4%) and ETN (9.0%).
We did not find statistically significant differences between serum TNFα blockers concentrations in RA and SpA patients that did and did not develop anti-drug Ab, although in previous studies low serum drug levels resulted inversely correlated with anti-drug Ab presence [5]. Moreover, no significant differences were found between drug levels and clinical status in RA and SpA classified by DAS28 or BADSAI score; furthermore, no significant difference was observed in the clinical response of anti-drug Ab positive and negative subjects of each disease group. Most of the studies refer that therapy failure occurred more frequently in positive anti-drug Ab patients [3,32,37,39], although there are previous reports which did not find a similar relation [40][41][42]. It is also reported that the concomitant use of MTX with anti-TNFα drugs reduces the incidence of anti-drug Ab [6,36]; in our study, in agreement with opposite results [13,39] concomitant MTX therapy was irrelevant on anti-drug Ab development.
A possible explanation of the discrepancies observed in the present study could be found in the recruitment criteria of the patients population; the patients of this study represent a consecutive population in the normal daily clinical setting without selection based on the presence or absence of antidrug Ab. Moreover, in order to simulate more precisely the normal clinical use of these diagnostic tools in real clinical practice it has been decided to collect blood only once.
It has been described that anti-TNF Ab titres can decrease and increase over time, and vice versa [37,43,44], causing a gradual increase in incidence over time when anti-drug Ab status is presented cumulatively, but not when assessed at each time point independently [44]. This explains why the time of measuring can influence the relevance and interpretation of anti-drug Ab status. It is therefore very difficult to draw a well-defined statement regarding the relationship between clinical response and anti-drug Ab measurement in clinical practice.

Conclusion
Our results indicate that the link between either serum drug levels or anti-drug Ab and clinical response is not as strong as previously assumed. This argues against the use of these parameters in monitoring drug efficacy. Although testing immunogenicity in clinical trials is standard practice and may yield interesting scientific insights, in clinical practice the real added value of the presence or absence of anti-drug Ab and the detection serum drug levels in an individual patient remains to be demonstrated.