Lung function data of North-African patients with rheumatoid arthritis: acomparative study between anti-citrullinated peptides antibodies positive and negative patients

Abstract

Introduction. Although lung damages are among the leading causes of death from Rheumatoid Arthritis (RA), few studies have assessed the spirometric and plethysmographic data and profile of patients with RA, particularly those with Anti-Citrullinated Peptides Antibodies Positive (ACPA+).

Aim. To compare the spirometric and plethysmographic data and profile of RA patients ACPA+ and ACPA-.

Methods. This comparative pilot study was performed over a two-year period (2018-2019) in Algiers (Algeria). The study included two groups of RA non-smoker patients: 26 ACPA+ and 33 ACPA-. RA was diagnosed according to the ACR/EULAR 2010 RA classification criteria. Spirometry and plethysmography were performed. The following definitions were applied: Obstructive Ventilatory Impairment (OVI): FEV1/FVC z-score < -1.645; Restrictive Ventilatory Impairment (RVI): Total Lung Capacity (TLC) z-score < -1.645; Mixed Ventilatory Impairment (MVI): FEV1/FVC z-score < -1.645 and TLC z-score < -1.645; lung- hyperinflation: residual volume z-score > +1.645; Nonspecific Ventilatory Impairment (NSVI): FEV1 z-score < -1.645, FVC z-score < -1.645, FEV1/FVC z-score ≥ -1.645, and TLC z-score ≥ -1.645.

Results. The ACPA- group was older than the ACPA+ one by ∼ 10 years (63±13 vs. 53±12 years, p=0.0025; respectively). The ACPA+ and ACPA- groups included comparative percentages of patients having RVI, MVI, and NSVI (23.1 vs. 45.5%, p=0.0745; 3.8 vs. 3.0%, p=0.8654; and 7.7 vs. 6.1%, p=0.8086; respectively). Compared to the ACPA- group, the ACPA+ group included a higher percentage of patients having OVI and lung-hyperinflation (9.1 vs. 38.5%, p=0.0069; 9.1 vs. 42.3%, p=0.0029; respectively).

Conclusion. Compared to the ACPA- group, the ACPA+ one had more lung-hyperinflation and OVI, and comparative percentages of RVI, MVI, and NSVI.

INTRODUCTION

Rheumatoid Arthritis (RA) is the most common inflammatory rheumatic disease, with a prevalence of 0.1 to 2.5% in the African continent (1 ). Females are twice to five times more susceptible to RA than males, and the disease is more common at the age of 40-50 years (2 ). Pulmonary involvement is one of the most frequent extra-articular manifestations of RA, and it is one of the leading causes of RA mortality (3 ). Several studies have evaluated the respiratory function and the bronchial inflammatory state in patients with RA (3, 4, 5, 6, 7, 8, 9, 10). It appears that RA is associated with increased odds of Restrictive and Obstructive Ventilatory Impairments (RVI, OVI, respectively) (7 ). Moreover, the existing body of evidence suggests that the lung is an organ of immune defense and has a role in the pathogenesis of the autoimmune disease Anti-Citrullinated Protein Antibody positive (ACPA+) RA (11 ). ACPAs are highly specific for RA (12 ), and ACPA+ represents a subclass of RA that differs in terms of pathogenesis, disease course, and response to therapy when compared with ACPA negative (ACPA-) (13, 14, 15). However, to the best of the authors’ knowledge, only a few studies have evaluated the relationship between lung impairments and immunological markers, such as the ACPA (3, 4, 5, 6).

ACPAs are considered as both sensitive and specific markers for rheumatoid disease ( 16, 17). They bind to proteins or peptides containing citrulline, and are detected in joints as well as in other tissues and organs, such as the lung (16 ). Although recent studies have suggested that the lung is the initial site for ACPA production, and that specific citrullination in the lung may precede the onset of joint involvement in RA (11, 13), the mechanism explaining the association of ACPA with pulmonary impairments has not been elucidated yet. Furthermore, there is not enough data explaining the different types of pulmonary involvement according to the serologic status of RA, whether ACPA+ or ACPA-. Indeed, the question whether ACPA levels could influence the clinical, Computed-Tomography (CT) and respiratory functional expressions of the rheumatoid lung is an issue worth raising (18 ).

According to Daha and Toes (18 ) “contrasting genetic backgrounds in ACPA+ and ACPA- support the notion that these are two distinct disease subsets, with different underlying pathogenesis”. Moreover, another study concluded that ACPA+ is a distinct entity from ACPA−, and that ACPA+ is independently associated with more severe erosive disease (19 ). Since the lung is a potential primary initiation site for ACPA+ (11 ), it is possible that pulmonary manifestations of ACPA+ and ACPA- are distinct on spirometry and plethysmography. Thus, it seems to be relevant to evaluate the spirometric and plethysmographic data and profiles of RA patients, and to determine their variabilities based on the patients’ ACPA status.

The objective of the present comparative pilot study, which was performed during the

2018-2019 period in Algiers, Algeria, was to compare the spirometric and plethysmographic data and profiles of RA non-smoker patients depending in the presence/absence of ACPA (ACPA+ vs. ACPA-). The primary outcome was the frequency of RVI, and the second outcomes were the frequencies of the following impairments: OVI; Mixed Ventilatory Impairment (MVI); Nonspecific Ventilatory Impairment (NSVI), and lung-hyperinflation.

METHODS

Study design

This was a cross-sectional comparative pilot study. The study was performed over a two-year period (2018-2019) in the Department of Pneumology, Phthisiology, and Allergology (Rouiba Hospital, Algiers, Algeria). The medical council of Rouiba hospital was consulted in reviewing and approving this manuscript for its design, conduct, reporting, and dissemination of this research work. The above-cited medical council approved this study (approval number 11/2018). As the spirometry and plethysmography are routine tests used for diagnostic activity, and after a discussion with the medical council of Rouiba hospital, the patients’ participation or individual ‘written’ consent from the study participants was not required.

Population

Only   patients   diagnosed   with   RA   were   included.   Patients   were   recruited20, 21

the age- and sex- adjusted prevalence of asthma was 3.45% (22 ).

The following non-inclusion criteria were applied: malignant tumor, history of pulmonary surgery, active infection, human immunodeficiency virus/acquired immune

deficiency syndrome, heart failure, chronic respiratory conditions such as COPD or asthma, active smoking, previous treatment by methotrexate (because of its possible impacts on lung interstitium and air trapping (23 )), destruction or fibrosis of the lung parenchyma due to previous infection or having pulmonary pathologies that can cause sequelae lung lesionssuch asandplethysmographymeasurements.PatientshavingInterstitialLungDiseases(ILDs)werenot excluded from the study. Files of RA patients with missing data were excluded from thefinal statistical analysis.presentsthestudyflowchart.

Positive diagnosis of RA

The positive diagnosis of RA was based on a body of clinical, biological, and radiological arguments, and in particular on the absence of clinical and biological arguments for another cause of inflammatory rheumatism (24 ). The criteria useful for diagnosis were those of the American college of rheumatology and European League of associations for rheumatology 2010 (25 ). However, in the absence of these clinico-biological criteria and the presence of erosions typical of RA identified at the standard x-rays of the hands-wrists and frontal forefoot, the disease is also classified as RA (25 ). Indeed, the concept of "typical RA erosion" has recently been defined as having at least three erosive joints among the metacarpophalangeal, proximal interphalangeal, wrist, and metatarsophalangeal bone (26 ).

Collected data

Clinical data were collected using a standard non-validated French medical questionnaire. The latter is part of normal clinical care at our department, and the questionnaire is used in the daily activity at the outpatient consultation Department. The questions, which were asked in Arabic language dialect, were essentially closed-ended and

most often dichotomous. Frequent comorbidities like arterial-hypertension, diabetes- mellitus, and heart diseases were noted. The following RA characteristics were noted from the patients’ medical records: length of time, age of symptoms onset, time of diagnosis, extra-articular symptomatology (Raynaud's phenomenon, morning stiffness), and Disease Activity Score (DAS) (27 ). The DAS was calculated at 28 joints using the level of C-Reactive Protein (CRP, DAS28-CRP) (28 ). The DAS28-CRP takes into account the number of painful and swollen joints, the overall evaluation of the disease by the patient, and CRP (28 ). The DAS28-CRP was considered severe, moderate, weak, or in remission if it was > 5.1, 3.2 to 5.1, 2.6 to 3.2, or ≤ 2.6 (29 ), respectively.

Cough and dyspnea were noted. Dyspnea was classified according to the modified medical research council scale [from grade 0 (breathless with strenuous exercise) to grade

4 (breathless to leave the house or when dressing)] (30 ). The smoking status [passive smoker, active smoker (smoking history > one pack-year)] was determined. The durations of the respiratory and extra-respiratory signs (in months) before RA diagnosis were noted.

Sex and some anthropometric data were determined. Age was calculated from the date of birth to the date of measurement. Standing height (m) and weight (kg) were recorded (seca medical scales and measuring systems, Germany). Body Mass Index (BMI, kg/m2) was calculated, and the following corpulence status were determined: underweight (BMI <

18.5 kg/m2), normal weight (BMI: 18.5-24.9 kg/m2), overweight (BMI: 25.0-29.9 kg/m2), and obesity (BMI ≥ 30.0 kg/m2) (31 ).

Some biological data such as Erythrocyte Sedimentation Rate (ESR, first hour, mm), CRP (mg/L), hemoglobin (g/L), white blood cells (/mm3), polynuclear neutrophils or eosinophils (/mm3), lymphocytes (/mm3), monocytes (/mm3)] were determined using standard hospital methods. Biological data were determined during the same week in which spirometry and plethysmography were performed. In patients aged < 50 years, ESR was considered elevated if it was > 15 in males or > 20 in females (32 ). In patients aged ≥ 50 years, ESR was considered elevated if it was > 20 in males or > 30 in females (32 ). CRP was considered elevated if its level was > 12 mg/L (33 ). Biological inflammatory syndrome was retained in the presence of an increase in CRP and/or ESR. Anemia was defined as hemoglobinemia < 12 g/L in females and <13 g/L in males (34 ). A number of white blood cells being > 11000/mm3 defined leukocytosis (35 ).

The following immunological data were determined: Rheumatoid Factor (RF, IU/ml), Anti-Nuclear Antibody (ANA, titer), and ACPA (U/ml). Immunoglobulin (Ig)M-RF, IgG-RF, and IgA-RF were detected by immunoturbidimetry (OCD VITROS, USA). According to the RF values, five RF levels were determined: negative (FR < 10 IU/mL), light (FR: 10 to 40), moderate (FR: 40 to 100), high (FR: 100 to 300), and very high (FR > 300) (36 ). Two groups of patients were identified: FR negative (FR-) and FR positive (FR+). A positive ANA test is

usually reported as a ratio (called a titer), and looked for using the indirect Hep-2 cell immunofluorescence method (Medical Biological Lab, Tokyo, Japan) with an initial dilution of 1:20, following a standardized protocol (37 ). Positive reactions were titrated by double dilution to the end-point. Serum reactivity at a dilution of at least 1:160 was considered a positive result for ANA in the present study (38 ). ACPAs were measured using an electro- chemiluminescence test on Cobas 6000 (ROCHE, Switzerland). RA patients were divided into two groups, ACPA+ (ACPA ³ 5 U/ml) and ACPA- (ACPA < 5 U/ml) (39 ). The ACPA+ and ACPA- groups were divided into subgroups according to the FR status: ACPA+-FR+, ACPA+- FR-, ACPA--FR+, and ACPA--FR- (Figure 1 ).

Figure 1. Study flowchart (Algiers (Algeria): 2018-2019).

usually reported as a ratio (called a titer), and looked for using the indirect Hep-2 cell immunofluorescence method (Medical Biological Lab, Tokyo, Japan) with an initial dilution of 1:20, following a standardized protocol (37 ). Positive reactions were titrated by double dilution to the end-point. Serum reactivity at a dilution of at least 1:160 was considered a positive result for ANA in the present study (38 ). ACPAs were measured using an electro- chemiluminescence test on Cobas 6000 (ROCHE, Switzerland). RA patients were divided into two groups, ACPA+ (ACPA ³ 5 U/ml) and ACPA- (ACPA < 5 U/ml) (39 ). The ACPA+ and ACPA- groups were divided into subgroups according to the FR status: ACPA+-FR+, ACPA+- FR-, ACPA--FR+, and ACPA--FR- (Figure 1 ).

Radiological data were determined using standard hospital methods [24 multi-bar 16- slice scanner (Optima 540 general electrique USA)]. No patient received intravenous contrast media. An experienced radiologist with high experience in thoracic CT and ILDs performed a semi-quantitative visual evaluation based on the lobe of the thoracic scanner, and a quantitative evaluation with computer-assisted calculation of emphysema and lung volume computed-assisted reading for pulmonary nodules. Inspiratory CT images were first evaluated, followed by expiratory CT images, with no time limit for visual evaluation. The radiologist was not informed about the patients’ clinical data, their biological, immunological, and spirometric/plethysmographic results. The radiologist was asked to apply the recommendations of the Fleischner society (40 ), and to determine the presence or not of an ILD, and if yes to determine its type, such as usual interstitial pneumonia, nonspecific ILD,

large airway abnormalities such as thickening of the bronchial wall, and bronchiectasis, or

small airway abnormalities like centrilobular nodules, and entrapment of expiratory air.

Lung Function Data (LFD) were measured by a plethysmograph (Body-box 5500, MediSoft, Belgium) between 9:00 a.m. and 3:00 p.m. Only one qualified person (KA in the authors’ list) performed the lung function measurements. The Lilly heated pneumotachograph (MICRO 5000 spirometer) combined to the plethysmograph, was used

to measure LFD. Calibration of the expiratory flow rates (daily calibration with a 3-liter syringe) and plethysmograph (daily calibration according to the manufactory guideline) were performed according to the international recommendations ( 41, 42 ). LFD were measured/calculated according to the international recommendations for spirometry and plethysmography (41, 42 ).

Spirometry was carried out in the sitting position. The following spirometric data were measured: Forced Vital Capacity (FVC, L), Forced Expiratory Volume in one second (FEV1, L), and Maximal Mid-Expiratory Flow (MMEF, L/s). The FEV1/FVC ratio (absolute value) was calculated. The spirometric data were expressed at ‘‘body temperature, barometric pressure saturated with water vapor’’ (41 ). The FVC maneuver had three distinct phases: maximal inspiration, a blast of exhalation, and continued complete exhalation to the end of testing. Patients were verbally encouraged to continue exhaling air at the end of the maneuver to obtain optimal effort. The criterion for end of testing was a volume-time curve showing no change in volume (25 mL) for one second, despite the patient’s effort to exhale for at least six seconds. The FVC maneuver repeatability was considered acceptable when the differences between the largest and the next largest FVC or between the largest and the next largest FEV1 were < 0.150 L (if FVC > 1 L) or < 0.100 L (if FVC ≤ 1 L) (41 ).

The steps for performing plethysmography were previously described ( 43, 44 ). Slow Vital Capacity (L), expiratory reserve volume (L), and Functional Residual Capacity (FRC, L) were measured, and the following static lung volumes were calculated (43, 44 ): Residual Volume (RV, L), Total Lung Capacity (TLC, L), and inspiratory capacity (L). The RV/TLC ratio (absolute value) was calculated.

LFD were expressed in three ways: i) absolute values, ii) percentages of predicted values from the global lung initiative norms for spirometry and static lung volumes (44, 45, 46, 47 ), and iii) z-scores (44, 45, 46, 47 ). Any LFD’ z-score below the Lower Limit of Normal (LLN = -1.645) or above the Upper Limit of Normal (ULN = +1.645) was considered abnormal (48 ).

The following definitions of ventilatory impairments were applied (48, 49, 50, 51):

• OVI: FEV1/FVC z-score < LLN;

• RVI: TLC z-score < LLN;

• MVI: FEV1/FVC z-score < LLN and TLC z-score < LLN;

• NSVI: FVC z-score < LLN and z-score FEV1 < LLN and FEV1/FVC z-score ≥ LLN and TLC z-score ≥ LLN;

• Lung-hyperinflation: RV z-score > ULN

The following three-level system assessing the severity of lung function impairments (ie; OVI and RVI) using FEV1 z-score values was used (50): i) mild: z-score between -1.65 and -2.50; ii) moderate: z-score between -2.50 and -4.00; and iii) severe: z-score < - 4.00.

Sample size calculation and statistical analyses

The sample size was calculated using the following formula (52 ): N= [(Z_α/2)² x P x (1- P) x D]/E2; where “ P ” was the frequency of the main event of interest (ie; frequency of RVI in RA patients), “ E ” was the margin of error (= 6.5%), “ Z α/2 ” was the normal deviate for two- tailed alternative hypotheses at a level of significance (Z α/2 = 1.96 for a 95%CI) and “ D ” was the design (= 1 for simple random sampling). Given the pioneer character of this pilot study, “ P ” was determined from a previous North-African study (5 ), where the frequency of RVI was 7% (p = 0.07). The total sample size was 59 consecutive RA patients. Assumption of 45% for the non-inclusion- and exclusion- criteria gave a corrected total sample of 108 patients [108 = 59/ (1-0.45)].

To assess the normal distribution of quantitative data, the Kolmogorov-Smirnov normality test was used. Quantitative data were expressed as mean±standard-deviation (and 95% CIfor ACPA values) if they had a normal distribution, and they were then compared using the Student’s T test for independent samples. If not, quantitative data were expressed as median [interquartile], and they were then compared using the Mann Whitney U Test. Categorical data were expressed as number (%). The two-sided Chi-square test was used

to compare the categorical data of two groups (ACPA+ vs. ACPA-) or subgroups (ACPA+- FR+ vs. ACPA+-FR-). No comparison was made between the ACPA--FR+ and ACPA--FR- subgroups since the ACPA--FR- subgroup included only two patients (Figure 1). Hedge value of TLC z-score was used for measuring the effect size (53 ). The latter was described as small, medium, large, and very large if it was ≤ 0.2, around 0.5, around 0.8, and more than 1.30, respectively (53 ). All mathematical computations and statistical procedures were performed using statistical software (Statistica, version 12). Significance level was set at 0.05.

RESULTS

An initial sample of 108 RA patients was included. After application of the non- inclusion- and exclusion- criteria, only 59 (8 males) were included in the final sample (Figure 1): 26 (4 males) in the ACPA+ group, and 33 (4 males) in the ACPA- group.

Table 1 presents some characteristics of the RA patients. Compared to the ACPA- group, the ACPA+ group was ∼ 10-years younger, and included a significantly higher percentage of patients with dyspnea grade 0.

Table 1. Characteristicsofpatientswithrheumatoid arthritis(RA):Algiers(Algeria):2018-2019.

Data	Outcome	Unit/Category	Total sample (n=59)	ACPA - (n=33)	ACPA + (n=26)	p-value
	Sex	(male)	8 (13.6)	4 (12.1)	4 (15.4)	0.7219
	Age	(year)	59±14	63±13	53±12	0.0025 *
Sex and anthropometric data	Weight	(kg)	69±15	67±14	71±15	0.3244
	Height	(m)	156±7	155±8	158±6	0.0897
	Body mass index	(kg/m2)	28.1±5.7	27.9±6.0	28.2±5.5	0.8507
		(underweight)	5 (8.5)	3 (9.1)	2 (7.7)	0.8480
		(normal weight)	9 (15.3)	5 (15.2)	4 (15.4)	0.9831
	Corpulence status	(overweight)	22 (37.3)	10 (30.3)	12 (46.2)	0.2099
		(obesity)	23 (39.0)	15 (45.5)	8 (30.8)	0.2505
Comorbidities and passive smoking	Arterial-hypertension	(yes)	14 (23.7)	8 (24.2)	6 (23.1)	0.9214
	Mellitus-diabetes	(yes)	10 (16.9)	5 (15.2)	5 (19.2)	0.6844
	Heart diseases	(yes)	8 (13.6)	5 (15.2)	3 (11.5)	0.6803
	Passive smoking	(yes)	11 (18.6)	7 (21.2)	4 (15.4)	0.5701
	Cough	(yes)	49 (83.1)	28 (87.9)	21 (80.8)	0.4511
		(grade 0)	4 (6.8)	0 (0.0)	4 (15.4)	0.0195 *
		(grade 1)	22 (37.3)	12 (36.4)	10 (38.5)	0.8685
Respiratory signs	Dyspnea (modified medical research council)	(grade 2)	18 (30.5)	13 (39.4)	5 (19.2)	0.0943
		(grade 3)	14 (23.7)	8 (24.2)	6 (23.1)	0.9214
		(grade 4)	1 (1.7)	0 (0.0)	1 (3.8)	0.2588
		(score)	1.8±1.0	1.9±0.8	1.6±1.1	0.2959
	Duration of signs before RA diagnosis	(months)	137±173	106±160	176±183	0.1261
Extra-respiratory signs and symptoms	Raynaud's phenomenon	(yes)	29 (49.2)	15 (45.5)	14 (53.8)	0.5267
	Morning stiffness	(yes)	51 (86.4)	29 (87.9)	22 (84.6)	0.4511
	Duration of signs before RA diagnosis	(months)	118±144	90±145	154±138	0.0944
RA activity	DAS28-CRP	(absolute value)	6.00±1.02	6.04±0.86	5.94±1.21	0.7279
	Disease activity status	(severe)	48 (81.4)	28 (84.8)	20 (76.9)	0.4396
		(moderate)	11 (18.6)	5 (15.2)	6 (23.1)	0.4396

ACPA: Anti-Citrullinated Peptides Antibodies. CRP: C-Reactive Protein. DAS28: Disease Activity Score 28. Quantitative and categorical data were mean±SD and number (%). *p-value < .05: 2 sided Chi-square test (comparison of categorical data between the 2 groups). Student's T test (comparison of quantitative data between the 2 groups).

Table 2 presents the immunological, biological, hematological, and radiological data of RA patients. The ACPA+ and ACPA- groups had comparative RF, ANA, biological, and hematological data, and comparative biological and hematological data/profiles. Compared to the ACPA- group, the ACPA+ group had a significantly higher value of ACPA [mean±standard-deviation (95%CI): 1.4±0.7 (1.2 to 1.7) vs. 133.1±247.7 (33.1 to 233.2), p=0.003, respectively], included lower percentages of patients RF+, or with a light RF level. The two groups had comparative radiological data. No patient has radiological signs of emphysema.

Table2.Immunological,biological,hematological,andradiologicaldataandprofileofpatientswithrheumatoidarthritis(RA):Algiers(Algeria):2018-2019.

Data	Outcome	Unit/Category	Total sample (n=59)	ACPA - (n=33)	ACPA + (n=26)	p-value
Immunological data and profile	ACPA	(U/ml)	59±175	1±1	133±248	0.0034 *
	ANA cut-off (titer)	≥1/160(%)	161±238	128±161	202±308	0.2453
	Rheumatoid factor (RF)	(IU/ml)	85±124	62±91	114±154	0.1122
	RF subgroups	Negative	13 (22.0)	2 (6.1)	11 (42.3)	0.009 *
		Positive	46 (78.0)	31 (93.9)	15 (57.7)	0.009 *
	RF levels	Light	22 (37.3)	18 (54.5)	4 (15.4)	0.0020 *
		Moderate	11 (18.6)	9 (27.3)	2 (7.7)	0.0551
		High	7 (11.9)	2 (6.1)	5 (19.2)	0.1225
		Very high	6 (10.2)	2 (6.1)	4 (15.4)	0.2412
Biological data and profile	ESR (1st hour)	(mm)	52±33	51±30	54±37	0.6831
	C-reactive protein	(mg/L)	28±43	21±33	36±51	0.1640
	High ESR	(yes)	15 (25.4)	8 (24.2)	7 (26.9)	0.8130
	High C-reactive protein	(yes)	28 (47.5)	13 (39.4)	15 (57.7)	0.1623
	Biological inflammatory syndrome	(yes)	48 (81.4)	25 (75.8)	23 (88.5)	0.2204
	Hemoglobin	(g/L)	12±2	12±2	12±2	0.9362
Hematological data and profile	White blood cells	(/mm3)	7810±2682	7953±2551	7627±2881	0.6469
	Polynuclear neutrophils	(/mm3)	4657±1916	4796±1939	4481±1910	0.5352
	Polynuclear eosinophils	(/mm3)	258±543	189±139	346±804	0.2746
	Lymphocytes	(/mm3)	2272±948	2304±846	2231±1080	0.7733
	Monocytes	(/mm3)	623±244	665±257	570±219	0.1373
	Anemia	(yes)	22 (37.3)	13 (39.4)	9 (34.6)	0.7050
	Leukocytosis	(yes)	8 (13.6)	5 (15.2)	3 (11.5)	0.6803
	Bronchiectasis	(yes)	16 (27.1)	7 (21.2)	9 (34.6)	0.2503
Radiological data	Bronchial wall thickening	(yes)	2 (3.4)	0 (0.0)	2 (7.7)	0.1048
	Thickened septal/non-septal lines	(yes)	45 (76.3)	23 (69.7)	12 (46.2)	0.0681
	Interstitial lung disease	(yes)	43 (72.9)	27 (81.8)	16 (61.5)	0.0817
	Nonspecific interstitial pneumonia	(yes)	32 (54.2)	19 (57.6)	13 (50.0)	0.5607
	Bronchiolitis	(yes)	18 (30.5)	8 (24.2)	10 (38.5)	0.2362
	Ground glass attenuation	(yes)	34 (57.6)	20 (60.6)	14 (53.8)	0.5998
	Usual interstitial pneumonia	(yes)	9 (15.3)	6 (18.2)	3 (11.5)	0.4772
	Micro nodule	(yes)	35 (59.3)	19 (57.6)	16 (61.5)	0.7621
	Air space consolidation	(yes)	32 (54.2)	19 (57.6)	13 (50.0)	0.5607

ACPA: Anti-Citrullinated Peptides Antibodies. ANA: Anti-Nuclear Antibody. ESR: Erythrocyte Sedimentation Rate. Quantitative and categorical data were mean±SD and number (%), respectively. *p-value < 0.05: 2 sided Chi-square test (comparison of categorical data between the 2 groups) Student's T test (comparison of quantitative data between the 2 groups).

Table 3 presents the LFD of RA patients. When LFD were expressed as z-scores, compared to the ACPA^- group, the ACPA⁺ one had significantly lower MMEF and FEV1/FVC ratio, and significantly higher values of FRC, RV, TLC, and RV/TLC. When LFD were expressed as percentage of predicted values, compared to the ACPA- group, the ACPA+ one had significantly higher values of FRC, RV, TLC, and RV/TLC, and a significantly lower FEV1/FVC ratio. The TLC z-score effect size was medium (Hedges’ unbiased d = -0.442).

Table3.Lungfunctiondataofpatientswithrheumatoidarthritis(RA):Algiers(Algeria):2018-2019.

Data	Outcome	Unit	Total sample (n=59)	ACPA - (n=33)	ACPA + (n=26)	p-value
	FVC	L	2.20±0.80	1.99±0.74	2.46±0.81	0.0260 *
Spirometric data		%	73±23	71±23	76±23	0.3837
		z-score	-1.91±1.61	-1.95±1.53	-1.86±1.73	0.8386
	FEV1	L	1.70±0.69	1.63±0.63	1.80±0.76	0.3470
		%	72±26	73±24	70±30	0.7308
		z-score	-1.92±1.80	-1.72±1.48	-2.17±2.14	0.3419
	MMEF	L/s	2.39±1.13	2.44±1.04	2.32±1.26	0.7044
		%	109±51	118±43	98±59	0.1396
		z-score	-0.06±1.61	0.33±1.13	-0.56±1.98	0.0349 *
	SVC	L	2.30±0.85	2.07±0.74	2.58±0.89	0.0200 *
		%	71±22	69±21	74±22	0.3723
		z-score	-2.25±1.68	-2.40±1.64	-2.06±1.75	0.4432
Static lung volumes	FRC	L	3.03±1.22	2.60±0.73	3.57±1.50	0.0018 *
		%	123±51	108±39	141±59	0.0114 *
		z-score	0.81±1.87	0.25±1.47	1.52±2.10	0.0083 *
	RV	L	2.24±1.15	1.86±0.73	2.73±1.40	0.0032 *
		%	150±96	117±51	192±121	0.0021 *
		z-score	1.18±2.03	0.43±1.47	2.13±2.27	0.0010 *
	TLC	L	4.51±1.46	3.90±0.91	5.27±1.67	0.0002 *
		%	94±28	85±24	106±30	0.0048 *
		z-score	-0.56±2.21	-1.26±1.81	0.33±2.37	0.0052 *
	IC	L	1.53±0.54	1.38±0.53	1.72±0.50	0.0146 *
		%	67±23	65±25	71±19	0.3152
		z-score	-1.69±1.17	-1.80±1.28	-1.54±1.01	0.4081
	FEV1/FVC	AV	0.77±0.13	0.81±0.10	0.72±0.14	0.0055 *
Ratios		%	97±17	103±13	90±18	0.0032 *
		z-score	-0.23±1.67	0.35±1.30	-0.97±1.83	0.0019 *
	RV/TLC	AV	48.89±13.00	47.82±13.14	50.25±12.94	0.4810
		%	153±51	137±34	174±61	0.0045 *
		z-score	2.22±1.90	1.67±1.56	2.92±2.09	0.0114 *

ACPA: Anti-Citrullinated Peptides Antibodies. AV: Absolute Value. FEV 1: Forced Expiratory Volume in one second. FRC: Functional Residual Capacity. FVC: Forced Vital Capacity. IC: Inspiratory Capacity. MMEF: Maximal-Mid Expiratory Flow. RV: Residual Volume. SVC: Slow Vital Capacity. TLC: Total Lung Capacity. %: percent of predicted values. Data were mean±SD. * p-value < 0.05: Student's T test (comparison between the 2 groups).

Table 4 presents the patients’ spirometric/plethysmographic profiles of RA patients. Compared to the ACPA- group, the ACPA+ group included higher percentages of patients with i) low MMEF (3.0 vs. 30.8%, respectively), ii) high FRC (9.1 vs. 38.5%, respectively),

Table4.Spirometricand plethysmographicprofilesofpatientswithrheumatoidarthritis(RA):Algiers(Algeria):2018-2019.

Data	Outcome	Total sample(n=59)	ACPA - (n=33)	ACPA + (n=26)	p-value
	FVC	33 (55.9)	18 (54.5)	15 (57.7)	.8058
	FEV1	32 (54.2)	16 (48.5)	16 (61.5)	.3197
	FEV1/FVC (OVI)	13 (22.0)	3 (9.1)	10 (38.5)	.0069 *
z-score < -1.645	MMEF	9 (15.3)	1 (3.0)	8 (30.8)	.032 *
	SVC	38 (64.4)	21 (63.6)	17 (65.4)	.8860
	FRC	2 (3.4)	2 (6.1)	0 (0.0)	.2000
	RV	2 (3.4)	2 (6.1)	0 (0.0)	.2000
	IC	30 (50.8)	18 (54.5)	12 (46.2)	.5267
	TLC (RVI)	21 (35.6)	15 (45.5)	6 (23.1)	.0745
	RV (lung-hyperinflation)	14 (23.7)	3 (9.1)	11 (42.3)	.0029 *
	TLC	9 (15.3)	3 (9.1)	6 (23.1)	.1377
z-score > +1.645	FRC	13 (22.0)	3 (9.1)	10 (38.5)	.0069 *
	IC	0 (0.0)	0 (0.0)	0 (0.0)	-
	RV/TLC	33 (55.9)	17 (51.5)	16 (61.5)	.4424
Ventilatory impairments	Mixed ventilatory impairment	2 (3.4)	1 (3.0)	1 (3.8)	.8654
	Nonspecific ventilatory impairment	4 (6.8)	2 (6.1)	2 (7.7)	.8086
Classifications of OVI	Moderate	8 (13.6)	2 (6.1)	6 (23.1)	0.0585
	Severe	5 (8.5)	1 (3.0)	4 (15.4)	0.0894
Classifications of RVI	Mild	8 (13.6)	7 (21.2)	1 (3.8)	0.0524
	Moderate	13 (22.0)	8 (24.2)	5 (19.2)	0.3226

OVI: Obstructive Ventilatory Impairment. RVI: Restrictive Ventilatory Impairment. For the remaining abbreviations, see table 3. Data were number (%). * p-value < 0.05: 2 sided Chi-square test: comparison between the 2 groups. Notes: 1. ACPA - group: among the 3 subjects with an OVI, 1 had lung-hyperinflation, and among the 3 subjects with lung-hyperinflation, 1 had an OVI. 2. ACPA + group: among the 10 subjects with an OVI, 9 had lung-hyperinflation, and among the 11 subjects with lung-hyperinflation, 9 had an OVI

iii) OVI (9.1 vs. 38.5%, respectively), and iv) lung-hyperinflation (9.1 vs. 42.3%, respectively). The two groups included comparative percentages of patients i) with low FVC,

FEV1, SVC, FRC, RV, and IC; ii) with high TLC, IC and RV/TLC; iii) with RVI, MVI, and NSVI; and iv) divided according to the severity of OVI and RVI.

Table 5 llustrates the lung function and radiological data of the ACPA+-FR+ and ACPA+-FR- subgroups. Whatever the expression mode of the LFD (ie; L, %, z-score), the two groups had comparative FVC, FEV1, MMEF, SVC, IC, and RV/TLC. Compared to the ACPA+-FR- subgroup, the ACPA+-FR+ subgroup had significantly higher FRC (z-score and

Table 5.Lung function and radiological data of patients with rheumatoid arthritis ACPA⁺(n=26)divided according to the rheumatoid factor (RF) status): Algiers (Algeria):2018-2019.

Data	Outcome	Unit/Category	RF - (n=11)	RF + (n=15)	p-value
	FVC	L	2.35±0.80	2.53±0.84	.5760
		%	75±23	77±23	.8298
		z-score	-1.89±1.68	-1.85±1.82	.9525
	FEV1	L	1.83±0.69	1.78±0.83	.8694
		%	75±30	67±30	.4998
		z-score	-1.76±2.07	-2.47±2.20	.4166
	MMEF	L/s	2.55±1.32	2.15±1.22	.4309
		%	116±66	84±51	.1750
		z-score	0.09±1.69	-1.03±2.10	.1571
	SVC	L	2.46±0.93	2.67±0.88	.5715
		%	73±25	75±21	.7860
Lung function data		z-score	-2.16±1.91	-1.99±1.69	.8108
	FRC	L	2.99±1.52	4.00±1.37	.0888
		%	115±47	161±60	.0443 *
		z-score	0.44±1.82	2.30±1.98	.0218 *
	RV	L	2.23±1.27	3.09±1.41	.1220
		%	137±62	233±139	.0423 *
		z-score	1.12±1.88	2.88±2.30	.0490 *
	TLC	L	4.62±1.86	5.74±1.39	.0916
		%	91±27	117±27	.0271 *
		z-score	-0.88±2.29	1.22±2.08	.0224 *
	IC	L	1.67±0.54	1.76±0.49	.6746
		%	69±22	72±18	.7144
		z-score	-1.63±1.13	-1.48±0.94	.7048
	FEV1/FVC	AV	0.79±0.13	0.68±0.13	.0424 *
		%	99±17	83±17	.0292 *
		z-score	-0.00±1.74	-1.68±1.59	.0178 *
	RV/TLC	AV	47.09±8.54	52.56±15.28	.2960
		%	151±37	191±70	.1029
		z-score	2.25±1.63	3.40±2.30	.1696
	Bronchiectasis	(yes)	3 (27.3)	6 (40.0)	.5013
	Bronchial wall thickening	(yes)	1 (9.1)	1 (6.7)	.8207
	Thickened septal/non- septal lines	(yes)	6 (54.5)	6 (40.0)	.4637
Radiological data
	Interstitial lung disease	(yes)	9 (81.8)	7 (46.7)	.0691
	Nonspecific interstitial pneumonia	(yes)	8 (72.7)	5 (33.3)	.0471 *
	Bronchiolitis	(yes)	6 (54.5)	4 (26.7)	.1500
	Ground glass attenuation	(yes)	7 (63.6)	7 (46.7)	.3931
	Usual interstitial pneumonia	(yes)	2 (18.2)	1 (6.7)	.3650
	Micro nodule	(yes)	7 (63.6)	9 (60.0)	.8521
	Air space consolidation	(yes)	6 (54.5)	7 (46.7)	.6943

ACPA: Anti-Citrullinated Peptides Antibodies. AV: Absolute Value. FEV 1: Forced Expiratory Volume in one second. FRC: Functional Residual Capacity. FVC: Forced Vital Capacity. IC: Inspiratory Capacity. MMEF: Maximal-Mid Expiratory Flow. RV: Residual Volume. SVC: Slow Vital Capacity. TLC: Total Lung Capacity. %: percent of predicted values. Quantitative and categorical data were mean±SD and number (%), respectively. * p-value < 0.05: 2 sided Chi-square test (comparison of categorical data between the 2 groups) Student's T test (comparison of quantitative data between the 2 groups).

%), RV (z-score and %), and TLC (z-score and %), and a significantly lower FEV1/FVC (absolute value, z-score, %). Compared to the ACPA+-FR- subgroup, the ACPA+-FR+ subgroup included lower percentages of patients with nonspecific interstitial pneumonia (72.7 vs. 33.3%, respectively).

DISCUSSION

The main results of the present comparative North-African pilot study were that compared to the ACPA- group (n=33), the ACPA+ group (n=26) included i) higher percentages of patients with OVI (9.1 vs. 38.5%, respectively) or lung-hyperinflation (9.1 vs. 42.3%, respectively), and ii) comparative percentages of patients with RVI, MVI, and NSVI. RA is a heterogeneous condition (11 ). The subdivision of RA by autoantibody status, conservatively RF and recently ACPA, has facilitated better understanding of possible pathogenic mechanisms in the genesis of RA (11 ). This is especially the case for ACPA+ RA, where RA initiating locations distant from the joints have been proposed within the gastrointestinal and respiratory systems (11 ). ACPAs are an interesting biomarker in RA (11 ). ACPA+ patients differ from ACPA- ones in terms of pathogenesis, disease course, and response to treatment (11, 12, 13 ). ACPA can develop before the onset of clinically detectable RA, and the lung can be an early site of autoimmune damage and ACPA production in the nonexistence of any clinical signs of the joint (17 ). Higher ACPA levels in bronchial alveolar lavage fluid have been reported compared with serum from ACPA+ patients, demonstrating that the lung is a local site of ACPA production (54 ). In addition to its higher prevalence in smokers (55 ), ACPA higher prevalence is also recognized in RA-related lung diseases, including patients with RA and bronchiectasis, a condition with a high prevalence of never smokers (56 ). On high-resolution CT, three commonly reported changes identified of RA populations are RA-bronchiectasis, RA-ILD, and RA-obstructive disease of smaller airways (57 ). To the best of the authors’ knowledge, this is the first study comparing the spirometric/plethysmographic data/profiles of RA patients divided according to their ACPA

status.

Spirometric and plethysmographic profiles of RA patients

Extra-articular manifestations of RA are frequent and dominated by pulmonary involvement (3 ). In this study including non-smokers RA patients free from asthma and

COPD, the most frequent spirometric/plethysmographic impairments were RVI (35.6%), lung-hyperinflation (23.7%), OVI (22.0%), NSVI (6.8%), and MVI (3.4%) (Table 4). According to the literature, the frequencies of spirometric/plethysmographic impairments vary largely from one study to another ( 5, 9, 10): RVI [from 5.8% to 22%], lung-hyperinflation

[3.4%], OVI [from 9% to 23.1%], NSVI [2.8%], and MVI [from 1% to 26.9%]. The divergence in the reported frequencies of spirometric/plethysmographic impairments can be explained

by the divergence of the patients’ profiles in terms of age, sex, smoking status, and corpulence status, presence of comorbidities such as asthma and/or COPD, RA

characteristics in terms of duration, stage, whether receiving treatment or not, different RF status (RF+ vs. RF-), patients’ recruitment mode (population-based setting (54 ) vs. rheumatology clinic), and mainly the applied criteria to retain the spirometric

plethysmographic impairments [for example LFD < 80% (58 ) vs. < LLN (present study)]. In the context of the hypothesis that RA-related autoimmunity is initiated at an extra-articular site (59 ), the high frequencies of some spirometric/plethysmographic impairments may strengthen the hypothesis that RA-related autoimmunity RA is initially generated in the lungs (17 ).

Spirometric and plethysmographic profiles of ACPA+ and ACPA- patients

In this study, the ACPA+ and ACPA- groups have comparative percentages of patients with RVI (main outcome), MVI and NSVI (secondary outcomes), and lung-hyperinflation and OVI (secondary outcomes) occur more frequently in patients with ACPA+ (Table 4). On the one hand, our results are “inconsistent” with those reported in a population-based study reporting no differences in the distribution of airway impairments between ACPA+ and ACPA- patients (54 ). On the other hand, our findings are “closer” to those reported by some authors (17, 60, 61 ). First, Huang et al. (60 ) investigated whether RA-related autoantibodies are associated with lung function impairments. RA serostatus was assessed by research assays for Cyclic Citrullinated Peptide (CCP) and RF (60 ). Among the 1272 analyzed patients

(mean age: 56.3±14.1 years, 17.8% were males, 69.5% were seropositive), 100 patients (7.9%) present lung function impairment (60 ). Compared with seronegativity, seropositivity is associated with increased odds of any lung function impairment (odds-ratio: 2.29, 95% CI: 1.30-4.03). When analyzing the type of lung function impairment, seropositivity was also associated with RVI (odds-ratio: 2.48; 95%CI: 1.26-4.87), OVI (odds-ratio: 3.12; 95%CI: 1.28-7.61), and diffusion impairment (odds-ratio: 2.30; 95%CI: 1.09-4.83) (60 ). When analyzing by CCP status, the associations were stronger for CCP+ (any lung function impairment odds-ratio: 1.67, 95%CI 1.03-2.69 for CCP+ vs. CCP−) (60 ). Huang et al. (60 ) concluded that CCP+ patients have a two-fold increased risk for impairments on lung function tests compared with CCP- patients. Second, Demoruelle et al. (17 ) reported the presence of subclinical lung impairments at high-resolution CT in more than 70% of ACPA+ patients without evidence of arthritis. The most frequent impairments are airways alterations, such as bronchial wall thickening and air trapping, with a minority of parenchymal alterations (17 ). Third, Fischer et al. (61 ) reported a prevalence of 54% of airways impairments, 14% of ILD, and 26% of a combination of both in ACPA+ patients without arthritis and with respiratory complaints.

In our study, the ACPA- and ACPA+ groups included comparable percentages of patients with the main outcome (frequency of a RVI: 45.5 vs. 23.1%, p = 0.0745) (Table 4). This result is “expected” since the ACPA- and ACPA+ groups included comparable percentage of patients with ILD (81.8 vs. 61.5%, p = 0.0817). However, the latter radiological finding is “unexpected”. In patients with RA-ILD, while one small study reported no association with ACPA+ (62 ), another study identified a strong relationship with ACPA+ and RF+ in 94% and 89% of patients (63 ). Moreover, a cohort including patients with early-RA and patients with longstanding-RA reported a higher frequency of both interstitial and airway diseases, with the latter being more noticeable in longstanding-RA (64 ). Finally, a study including 60 patients with early-RA concluded that the RF and ILD are linked (65 ).

In our study, compared to the ACPA- group, the ACPA+ group included a higher percentage of patients with lung-hyperinflation (9.1 vs. 42.3%, p=0.009, respectively) (Table 4). Lung-hyperinflation is a chief concern in the management of some chronic respiratory conditions (66 ). It leads to a rise in the relaxation volume due to the reduction of the lung’ elastic retraction forces (67 ). Lung-hyperinflation has deleterious clinical, functional, and radiological consequences that make it a major source of impaired quality of life ( 67, 68). In our ACPA+ group, the presence of lung-hyperinflation can explain the higher percentage of patients with dyspnea grade 0 (breathless with strenuous exercise) (68 ). Lung- hyperinflation can be understood as an aging index of the ventilatory mechanics, or an indirect sign towards an OVI and/or an expiratory muscle weakness (66 ). Lung- hyperinflation can also indirectly reflects damage to the distal airways (69 ). In the present, the last hypothesis is plausible since compared to the ACPA- group, the ACPA+ group included a higher percentage of patients with low MMEF (3.0 vs. 30.8%, respectively) (Table 4). MMEF is considered more sensitive (but nonspecific) in detecting early airway obstruction which tends to take place at lower lung volumes (70 ). Since lung-hyperinflation (eg, higher RV) can be observed in patients with expiratory muscle weakness, it was better to evaluate the maximal expiratory pressure (71 ).

Compared to the ACPA- group, the ACPA+ group included a higher percentage of patients with OVI (9.1 vs. 38.5%, p=0.009, respectively). This result is “unexpected” and some remarks should be noted. First, since the two groups had comparative inflammatory data/profile (Table 2), therefore the hypothetic relationship between inflammation and airway obstruction cannot be advanced in our study (72 ). Second, since the two groups of patients were no smokers, and included comparative percentages of passive smokers (Table 1), the role of smoking in the pathogenesis of OVI in the ACPA+ group cannot be forwarded. Concerning smoking, there is wide literature concerning its potential role in the pathogenesis of ACPA+ (11 ). Because of the robust relationship between smoking and

ACPA’ pathogenesis, the lung has become the attention of investigation to decide whether processes within the lung are connected to the generation of ACPA (11 ). Smoking was identified as an environmental factor that in the context of human leukocyte antigen-D related isotype shared epitope genes may trigger RA-specific immune reactions to citrullinated proteins (13 ). In one animal study (73 ), ACPA levels were lower in mice exposed to cigarette smoke. In another study (74 ), no patient with COPD has a carrier of ACPA, suggesting that smoking is not directly responsible for triggering ACPA production. Third, since no patient from the two groups were asthmatic and/or COPD, the presence of OVI in the ACPA+ group cannot be linked to these two obstructive lung diseases. However, the high frequency of OVI observed in the ACPA+ group compared to the ACPA- group is “similar” to the high frequency of COPD or asthma in ACPA+ (8, 75 ). Zaccardelli et al. (8 ) reported that i) 17.7% of pre-ACPA+ cases and 6.3% of matched controls reported physician-diagnosed asthma, ii) After adjusting for matching factors, smoking pack-years, passive smoking, and BMI, asthma remains significantly associated with pre-ACPA+ [odds- ratio: 3.57, 95%CI: 1.58-8.04], and iii) Asthma is significantly associated with seropositive RA (odds-ratio: 1.79, 95%CI: 1.01-3.18). Moreover, Zaccardelli et al. (75 ) analyzed 283 females with pre-RA (before RA diagnosis) and 842 controls, and reported that i) Pre-ACPA+ is associated with increased COPD risk [hazard-ratio: 3.04, 95%CI: 1.33-7.00] after adjusting for covariates, including smoking pack-years, and ii) Pre-ACPA+ has a hazard- ratio for asthma of 1.74 (95%CI: 0.72-4.24). The authors concluded that “females with elevated ACPA before RA diagnosis have increased risk for developing COPD compared to controls, and that females who later develop RA are more likely to develop asthma, regardless of pre-ACPA status” (75 ). Another study identified an association between ACPA and bronchial wall thickening on chest CT, suggesting that structural airway abnormalities may be present early in the course of RA, even though the patients are asymptomatic or have few respiratory symptoms (76 ). Finally, in 60 patients with early RA, the presence of

ACPAs is associated with bronchial wall thickening (65 ). One possible explanation of the OVI observed in the ACPA+ group could be the presence of an oxidative stress (77 ). One study reported that autoreactivity to carbamylated and acetylated epitopes is more commonly found within RA patients with citrulline-reactivity (77 ). NSVI is an old ventilatory impairment described in 1972 in a study involving asymptomatic asthmatic patients (78 ). The NSVI can reflect reduced effort, a RVI, or be an early consequence of small airway condition with emphysema and/or air trapping (79 ). The NSVI has been titled “preserved ratio-impaired spirometry” which, in follow-up has been shown to be associated with both more typical RVI or OVI (80, 81, 82 ). The NSVI occurs in 9.5%90 to 15% (83 ) of patients with complete lung function tests. The two frequent etiologies of NSVI are airway hyperreactivity (68%) and/or obesity (50%), and RA is identified only in 1% of patients (84 ). Since the two groups of ACPA+ and ACPA- have comparative corpulence status (Table 1), obesity cannot be advanced to explain the presence of NSVI. The latter may be secondary to changes, such as bronchial inflammation, bronchoconstriction, or mucus accumulation.

Compared to the ACPA+-FR- subgroup, the ACPA+-FR+ subgroup included lower

percentages of patients with nonspecific interstitial pneumonia and comparable frequencies of patients with ILD or bronchiectasis (Table 5). Our finding is opposite to the evidence suggesting an association between ACPA+, RF+ and RA-ILD (11 ). Perry et al. (11 ) identified that 94% and 97% of RA-bronchiectasis patients were ACPA+ and RF+, respectively. The aforementioned frequencies were significantly higher than would normally be seen in patients with RA alone (56).

Study limitations

This study has some limitations. First, our sample size (n=59) “appears” small. However, the sample size was calculated according to a predictive equation (52 ). Calculation of an optimal size is a crucial point since it helps avoid an inadequate power to detect statistical effects (85 ). On the one hand, using few participants in a study may lead

to lower "precision" in findings. On the other hand, a large sample size is, however, expensive and exposes more participants to procedures (85 ). Second, as the sample size was calculated only for the primary outcome measure (frequency of a RVI), our statements that there was difference or no difference between ACPA- and ACPA+ groups for the secondary outcomes (frequencies of lung hyperinflation, OVI, MVI, or NSVI) should be interpreted with a caution, as this could be a consequence of the low power of the study. Third, it was better to include two groups matched for age, since age is a predictor factor of adults’ LFD (86 ). To counterbalance this limitation, LFD were expressed as z-scores ( 45, 46). z-scores are free from bias due to age, height, sex and ethnic group ( 45, 46 ). Fourth, it could have been better to explore the alveolar-capillary diffusion capacity (eg; diffusing capacity for carbon monoxide), the submaximal aerobic capacity via the 6-minute walk test for example, and the oxidant/antioxidant status of the RA patients. Fifth, our study’ one- center character did not allow us to generalize our findings. Moreover, since our patients were selected from those referred to a pulmonology service, this will confound the results, as the patients studied will not be representative of all RA patients. Sixth, it could have been better if a bronchodilator test was performed. The latter would made possible to distinguish between true MVI (or NSVI) and RVI due to the eventual reduction of static lung volumes in terms of air trapping (48 ). Seventh, we have compared the subgroups according to RF serology status, but the study was not powered for these comparisons. Therefore, conclusion related to the RF status must be considered with caution. Finally, it was better to

evaluate if the ventilatory impairments reported in our study have a clinically meaningful, for

example; do they influence patients’ symptoms or walk distance or quality of life? For future studies, it will be interesting to evaluate the correlation between LFD and quality of life, evaluated via an Arabic validated short questionnaire such as the VQ11 (87 ), of ACPA+ and ACPA- patients (24 ).

CONCLUSION

Our study revealed significantly higher risks of occurrence of lung-hyperinflation and OVI in patients with ACPA+, but comparable risk of occurrence of RVI, MVI, and NSVI in both ACPA- and ACPA+ groups. Our study “confirms” the benefit of early screening for lung- hyperinflation and OVI in patients with ACPA+, and testing for ACPA in non-smokers patients with OVI and/or lung-hyperinflation, particularly in those with clinical signs suggestive of RA. This attitude could improve the management of RA patients and some patients with lung- hyperinflation and/or OVI. There is a significant clinical implication of our findings, since physicians would treat isolated lung-hyperinflation seen on lung function tests especially in dyspneic patients. Lung-hyperinflation is a “natural” target for bronchodilators, and it is by decreasing it that bronchodilators can reduce the dyspnea (66 ).