Combined Lung Transfer of NO and CO in Patients ReceivingMethotrexate or Bleomycin Therapy Compared to Normal Subjects

1 Pulmonary Function Testing Department, Lyon University Hospital, 165 Chemin du Grand-Revoyet, 69 495 Pierre-Bénite Cedex, France UCBL1 FRE CNRS 3310, 7 Passage du Vercors, 69367 Lyon Cedex, France 3 Respiratory Diseases Department, Lyon University Hospital, 165 Chemin du Grand-Revoyet, 69495 Pierre-Bénite Cedex, France 4 Lyon Sud-Charles Mérieux Medical Faculty, Claude Bernard Lyon 1 University, 69600 Oullins, France


Introduction
Methotrexate (MTX), a folic acid antagonist, is widely used for the treatment of many autoimmune diseases such as rheumatoid arthritis (RA), psoriasis, or lupus erythematous (LE). is agent acts as a folic acid antagonist and is known to be an effective anti-in�ammatory agent. However, it is also known for its pulmonary toxicity which is independent of the dose delivered [1][2][3][4][5][6][7]. MTX-induced pulmonary toxicity is an unpredictable, unusual, and mostly reversible event. Nevertheless, it is a serious adverse effect, since it may be fatal, particularly in patients with psoriatic arthritis [1]. Its incidence ranges from 1% to 5% [2,3], and its prognosis is usually favorable [4].
Bleomycin (BLM) is another agent known for its pulmonary toxicity [8][9][10][11][12][13]. is cytotoxic agent is successfully used in the treatment of several malignancies such as germ cell tumors, lymphomas, and some squamous cell carcinomas. e most frequent adverse effects of BLM are interstitial pneumonitis (ILD), followed by pulmonary �brosis. BLMinduced pneumonitis occurs in around 46% of patients treated with a BLM-based chemotherapy. is lung toxicity usually appears during treatment [13] but can also appear up to 10 years aer the incriminated treatment. Mulder et al. studied a cohort of childhood cancer survivor who received a BLM-based chemotherapy. Aer a followup of 18 years, 44% of patients presented with a pulmonary function impairment including 39.9% with a decreased carbon monoxide diffusing capacity (DLCO) [8]. Both MTX and BLM increase the risk to develop a pulmonary interstitial disease. Two components of the pulmonary alveolar membrane are concerned: the alveolar membrane (Dm) and the pulmonary capillary volume 2 Physiology Journal (Vc). Consequently, a strict pulmonary followup including pulmonary functional tests should absolutely be set up when MTX or BLM has been used in order to diagnose lung toxicity as early as possible [14].
Measurement of the gases transfers through the lung is one of the few tests enabling physicians to investigate alveolar function.
e single-breath carbon monoxide diffusing capacity (TLCO) is a well-known and helpful parameter used in many respiratory �elds. [15][16][17][18][19][20]. TLCO is the product of two distinct but concurrent measurements during breath: the rate constant for carbon monoxide uptake from alveolar gas (KCO) and the alveolar volume (VA) [21]. Recently, by using Roughton and Forster's equations [21,22], the combined measurement of TLCO and TLNO (nitric oxide diffusing capacity) allowed the measurement of two other parameters: the alveolarcapillary membrane conductance (Dm) and the pulmonary capillary blood volume (Vc) [23,24]. Actually, in this model, carbon monoxide switches through the alveolar and capillary structures and is split into two resistances. e �rst one is for the alveolar membrane (Dm). e other one is for the blood reacting with the gas (1/ COVc, in which COVc is the red cell conductance at a particular concentration and Vc the capillary volume). e resulting equation is 1/DLCO = (1/Dm) + (1/ COVC). Guenard et al. [24] reported measurements of Dm and Vc using TLNO and TLCO and considering NO as an in�nite value resulting in TLNO = DmNO. erefore, the measurement of DLNO leads to the calculation of DmCO and �nally of Vc.
Considering that the VA value depends on the performance of inspiratory muscles and of chest-lung mechanisms, the interpretation of KCO = TLCO/VA should, therefore, be taken into account with caution, as suggested by many authors [21,25,26]. VA is measured during a single breath, from a volume diluted with helium and aer subtracting an "estimated" anatomic space from the inspired volume (VI) [21]. VI is usually de�ned as the volume comprised between residual volume and maximal in�ation. It is measured by the device also used for TLNO and TLCO measurements. Finally, a high TLNO/TLCO ratio has been suggested to be a re�ection of a decrease in the thickness of the capillary membrane [27].
e aims of this study were to determine whether MTX and/or BLM have a speci�c toxicity either for Dm or for Vc compared to subjects not treated with such agents, and to determine whether KCO and TLCO/VI% (VI% = % of predicted VI) correlate better with Dm and/or Vc than KCO.

Studied
Population. e patients treated in our hospital and who underwent functional pulmonary tests as part of their regular followup were included in this study. ree groups were studied: a "normal" control group ( ), a group of patient treated with MTX ( ), and a last a group of patients treated with BLM ( ). All subjects knew the aim of the study and gave their informed consent. e study was conducted in accordance with the declaration of Helsinki (1964) and was approved by the Internal Review Board of the University of Lyon [11][12][13][14][15][16][17][18][19][20]. e control group was composed of patients who underwent pulmonary tests as part as their routine medical followup, and whom results were 00% ± 0% of the normal predicted values. e MTX group included 17 patients: two with dermatomyositis, one with lupus erythematous, three with rheumatoid arthritis, three with cutaneous psoriasis, two with psoriatic arthritis, one with scleroderma, one with in�ammatory rheumatism, three with ankylosing spondylitis arthritis, and one with a chronic lymphoid leukemia before undergoing a bone marrow autogra. Each patient received MTX for at least one year and was treated for a mean of .94 ± 0. 8 years. e BLM group included 12 patients: six suffering from a testis cancer, �ve from Hodgkin's lymphoma, and one who presented with an acute lymphoid leukemia (before bone marrow autogra). A mean of 5.8 (±1.7) cycles of bleomycinbased chemotherapy were administered to these patients (min: 2-max: 9). Hemoglobin in these 12 patients was 106, 111, 146, 126, 87, 103, 151, 86, 11.5, 143, and 131 g/L.

Pulmonary Function Tests.
Spirometry and plethysmography were performed using Bodybox 5 500 from Mediso cardiorespiratory instrumentation (Dinant, Belgium). In the whole cohort, the following parameters were measured: forced vital capacity (FVC), forced expiratory volume in one second (FEV1), FEV1/FCV ratio, and functional residual capacity FRC. TLCO and TLNO were measured simultaneously, as appropriate [25]. Patients were seated and wore a nose clip. A gas mixture containing 0.28% CO, 14% helium (He), and 21% oxygen (O ) balanced with nitrogen (N ) was mixed with a NO/N mixture (450 ppm NO in N ; Air Liquide Sant�,V�nissieux, France). e �nal concentrations of NO and O in inspired gas were 40 ppm and 19.1%, respectively. e device was calibrated for gas fractions using automated procedures as appropriate (dilution of the NO mixture with the CO/He/O mixture in order to check the NOanalyzer linearity). e pneumotachograph was calibrated daily using a 2 L syringe as speci�ed by the manufacturer. Patients breathed through a mouth bit committed to a singleuse �lter and connected to the pneumotachograph. At the right time, patients were requested to inspire the gas mixture in a rapid and deep inspiration followed by a brief (4 seconds) and full expiration. e �rst 0.8 L of expired gas was rejected and the following 0.6 L was sampled in a bag and then automatically analyzed for NO, CO, and He concentrations. is sample was analyzed for 35 seconds. VA during apnea was calculated using the He-dilution method. TLCO and TLNO measurements enabled physicians to determine Dm and Vc, according to Roughton and Forster's equations [21,22]. TLCO

Characteristics of the ree Studied
Groups. e main characteristics of the three studied groups are given in Table 1.
As expected, the sex ratio was different between the three groups. e rate of smokers was not signi�cantly different in the three groups. e BLM group showed a younger mean age than the MTX group. e mean height was greater in the BLM group than in the MTX one. e weight was not signi�cantly different between the three considered groups.  Table 3. TLNO was lower in the MTX (66 3% ± %) and BLM (66 % ± %) groups compared to the N one (9 9% ± 8%). VA was not signi�cantly different in the three groups. KCO was lower in the MTX and BLM groups compared to the N one and between the BLM and the MTX groups. TLCO/VI% was signi�cantly reduced in the MTX compared to the N group only. Dm and Vc were signi�cantly reduced in both of the MTX and the BLM groups compared to

Discussion
e difference observed in sex ratios may be explained by the type of pathology treated in the patients included in both groups such as testis cancer in the BLM group ( = 6) or autoimmune disorders (affecting mainly women). e younger mean age in the BLM group may be explained similarly (testis cancer and Hodgkin lymphoma affect mainly young adults).
Considering the �rst aim of the study, both MTX and BLM worsen FVC and FEV1, while FEV1/FVC ratio is worsened only in the BLM group, suggesting that BLM could cause more pulmonary obstruction than MTX. Both NO and CO diffusing capacities are decreased, much more in the BLM group. However, in animals, BLM-induced lung injury and �brosis are known to be more serious in males than in females [28]. So, the greater proportion of females in the N and MTX groups could induce a bias. In further studies of this kind, it should be relevant to stratify experimental groups on sex and age.
In order to avoid MTX-induced pulmonary toxicity, the lowest dose of MTX should be delivered. Actually, a low dose of MTX (7.5 mg/week) proved to be efficient and to induce fewer adverse effects in patients suffering from rheumatoid arthritis [29].
KCO is more strongly decreased in the BLM group. Both Dm and Vc are reduced by MTX and BLM. ese results match the literature data [1-3, 7, 8, 13]. However, in sarcoidosis, the limitation of gas diffusing capacities is mainly Physiology Journal 5 located in the membrane barrier, although recruitment of microvascular reserves is modestly impaired. ese results suggest that in sarcoidosis, the thickening of the alveolocapillary membrane is the main factor inducing a decline of the gas diffusing capacities in lungs [30].
e TLNO/TLCO ratio is another critical point. Actually, it is only increased in the BLM group. is result suggests that BLM affects pulmonary capillary more than MTX does. As a matter of fact, this ratio has been suggested by some other authors to be re�ective of a decreasing thickness of the capillary blood layer [27]. BLM could mimic this pattern. e TLNO/TLCO ratio is also known to increase aer obstruction of the pulmonary artery in animals [31]. e authors of this work suggest that this rise may be the consequence of a better sensitivity of TLCO than TLNO in response to a local reduction in capillary blood �ow.
Regarding the second aim of the study, KCO correlates neither with Dm nor with Vc in the pooled population. is observation may lead to a debate about the usefulness of this parameter to assess the alveolar membrane and capillary volume, as previously described [21,25,26]. is result could be explained by the way VA is calculated [21]. VA is calculated during the TLCO measurement and so expressed in standard temperature and pressure dry (STPD) conditions; while KCO calculation (KCO = TLCO/VA) is expressed in body temperature and pressure saturated (BTPS) conditions. By contrast, the VI is directly measured by the device used for TLNO and TLCO measurements.

Conclusion
is study shows that TLCO/VI% correlates signi�cantly with both Dm and Vc. is result suggests that this parameter could be taken into account to investigate pulmonary alveolar membrane impairment. Further studies are needed to con-�rm this observation. Several studies have demonstrated the great interest of a combined TLCO and TLNO measurement in many different diseases such as liver cirrhosis [27], idiopathic pulmonary hypertension [32], or chronic renal failure [33]. As far as we know, it has been poorly studied in patients receiving agents inducing lung toxicity such as MTX or BLM. is study shows that this technique is useful and con�rms that combined TLNO and TLCO could be a relevant test and could be routinely performed as suggested by other authors [34].

Con�ict of �nterest
e authors declare no con�ict of interests.
Authors' Contribution C. Viart-Ferber and S. Couraud are contributed equally to this work.