Characteristics of patients meeting the new definition of pre-capillary pulmonary hypertension (Nice 2018) in a single Japanese pulmonary hypertension center

Background The 6th World Symposium on Pulmonary Hypertension (Nice 2018) proposed a new definition of pre-capillary pulmonary hypertension (PH) as a condition with mean pulmonary artery pressure (mPAP) > 20 mmHg, pulmonary artery wedge pressure  ≤ 15 mmHg, and pulmonary vascular resistance (PVR) ≥ 3 Wood units (WU). The characteristics and prognosis of patients with pre-capillary PH, according to this new definition, is unclear. Therefore, we determined the characteristics and survival of patients with borderline pre-capillary PH. Methods We retrospectively enrolled 683 patients who underwent their first right heart catheterization at Chiba University, Japan. Among them, 489 patients met the pre-capillary PH requirement with mPAP ≥ 25 mmHg (conventional pre-capillary PH group), while 22 patients met the borderline pre-capillary PH criteria (borderline pre-capillary PH group). Additionally, 16 patients with a mean PAP of 20–25 and PVR of 2–3 WU were also examined. Results The borderline pre-capillary PH group comprised 4.3% of the total patients with pre-capillary PH, and the majority was in Group 3 (40.9%) or 4 (45.5%). The survival of the borderline pre-capillary PH group tended to be better than that of the conventional pre-capillary PH group. The prognosis of Group3 PH was the worst among the patients with borderline precapillary PH. There was no significant difference in survival between the borderline pre-capillary PH group with PVR ≥ 3 WU and that with PVR of 2–3 2WU, although none of the patients in the latter group died due to right heart failure. Conclusions This is the first study conducted in a PH center in an Asian country to reveal the characteristics of patients with pre-capillary PH, according to the Nice 2018 definition. They comprised 4.3% of the total population with pre-capillary PH, and the majority of the pre-capillary PH cases were in either Group3 or 4. The prognosis may be affected by the patients’ underlying diseases. Further prospective studies are needed to determine whether the new definition, including the PVR cut-off, is beneficial in clinical practice.

a new definition for PH and new treatment strategies. The general purpose of the clinical classification of PH is to categorize clinical conditions associated with PH based on similar pathophysiological mechanisms, clinical presentation, hemodynamic characteristics, and therapeutic management. A comprehensive and simplified updated version of the clinical classification of PH was presented in Nice 2018 [1]. To date, since the first WSPH in 1973, PH has been arbitrarily defined as mean pulmonary arterial pressure (mPAP) ≥ 25 mmHg at rest. However, recent data obtained for patients undergoing right heart catheterization (RHC) have shown that normal mPAP was 14.0 ± 3.3 mmHg in healthy subjects, and two standard deviations above this mean value would suggest that mPAP > 20 mmHg is above the upper normal limit [1,2]. In addition, the mPAP value is inadequate to define pulmonary vascular disease since this value can be affected by the cardiac output (CO) or pulmonary arterial wedge pressure (PAWP). Considering these aspects, the task force of the 6th WSPH in 2018 proposed that if all three criteria are met, namely, mPAP > 20 mmHg, PAWP ≤ 15 mmHg, and pulmonary vascular resistance (PVR) ≥ 3 Wood units (WU), then the new definition of pre-capillary PH would be satisfied [1]. A PAWP > 15 mmHg is defined as post-capillary PH, which is considered as isolated PH when PVR < 3 WU, and combined pre-and post-capillary PH when PVR ≥ 3WU. In terms of pulmonary vascular disease, the reliability and validity of this new PH definition have not been defined. Further, the characteristics and survival of patients with pre-capillary PH, diagnosed according to this new definition, are unclear. Moreover, a recent study has revealed that patients with PVR ≥ 2WU and scleroderma had a significantly poor prognosis [3].Therefore, in our cohort of patients with PH, we determined the characteristics and prognosis of patients with pre-capillary PH who had pulmonary vascular disease with a modest elevation in mPAP (borderline pre-capillary PH group). This study was conducted in a Japanese PH center that is associated with respiratory medicine and is one of the high-volume pulmonary endarterectomy (PEA) centers in Japan.

Ethics approval and consent to participate
Patient identity was concealed, and all data were compiled according to the requirements of the Japanese Ministry of Health, Labour and Welfare, which is dedicated to privacy, information technology, and civil rights. Based on the Japanese legislation, the need for informed consent was waived. The study protocol was approved by the Research Ethics Committee of Chiba University School of Medicine (approval number: 2,584). Since 2009, all survivors have provided written informed consent for a prospective cohort study (approval number 826). For patients who died before 2008, written informed consent was not required, in line with the guidelines for retrospective studies in Japan and in accordance with the criteria of the ethics committee of Chiba University Hospital. The study database was anonymized and all experiments were performed in accordance with the relevant guidelines and regulations.

Patients
Our patients were mainly referred from other hospitals or other departments of Chiba University Hospital. Our PH center is associated with respiratory medicine. Therefore, patients with respiratory diseases suspected of PH during follow-up were also included. From among them, we conducted a retrospective analysis of the data from patients who had undergone RHC. Indication criteria for RHC were as follows: RHC was conducted when a patient's hypoxia or exercise limitation was considered to have stemmed from PH, and the outcome of RHC seemed to have influenced the patient's treatment, including their eligibility for clinical trials. Among patients with respiratory diseases, RHC was conducted if transplantation was indicated or when RHC would be beneficial. Among the 1542 patients who underwent RHC between 1999 and 2020 at Chiba University, we enrolled 683 patients who were catheterized for the first time at the first diagnosis. The reason for choosing patients from 1999 was because this was the year epoprostenol was approved in Japan; since then, patients have been prescribed selective pulmonary vasodilators. Among these, 531 patients had mPAP ≥ 25 mmHg (conventional PH group) and 50 patients had mPAP ranging from 20 to 24 mmHg (borderline PH group). The non-PH group consisted of 102 patients with mPAP < 20 mmHg (Fig. 1a). We then chose patients with PH accompanied with pulmonary vascular disease who met the criteria for pre-capillary PH (mPAP > 20 mmHg with PAWP ≤ 15 mmHg and PVR ≥ 3WU). Among the patients with pre-capillary PH (n = 511), mPAP ≥ 25 mmHg was observed in 489 (conventional pre-capillary PH group) and 25 > mPAP > 20 mmHg was observed in 22 patients (borderline pre-capillary PH group) (Fig. 1b). We also categorized our cohort into Group 1-5 according to the Nice 2018 classification [1]; the patients were allocated to the groups after evaluation by two pulmonologists. Despite following the new Nice 2018 recommendation, classifying Group 1 and 3 was occasionally difficult, especially when judging morphological or physiological severity.
For Study 1, we compared the characteristics of the conventional PH, borderline PH, and non-PH groups. The non-PH group was used as a reference since the main aim was to clarify the characteristics of patients diagnosed with PH based on the new definition of precapillary PH.
For Study 2, to focus on the pre-capillary nature of PH, the characteristics and survival of the conventional pre-capillary PH and borderline pre-capillary PH groups were compared. For Group 3, the PH prognostic factors were also examined. Fig. 1 Selection of patients. a Among 1542 cases of right heart catheterization (RHC) handled at the Chiba University, Japan, we enrolled 683 patients who underwent the first RHC. Among these, 531 patients had mPAP ≥ 25 mmHg (conventional PH group), and 50 patients had 25 > mPAP > 20 mmHg (borderline PH group). There were 102 patients with mPAP < 20 mmHg (non-PH group). PH, pulmonary hypertension; mPAP, mean pulmonary arterial pressure. b We chose patients with PH and pulmonary vascular disease, namely those with pre-capillary PH (mPAP > 20 mmHg with PAWP ≤ 15 mmHg and PVR ≥ 3 WU). Among the patients with pre-capillary PH (n = 501), 489 had mPAP ≥ 25 mmHg (conventional pre-capillary group) and 22 patients had 25 > mPAP > 20 mmHg (borderline pre-capillary group). PH, pulmonary hypertension; mPAP, mean pulmonary arterial pressure; PAWP, pulmonary arterial wedge pressure; PVR, pulmonary vascular resistance; WU, Wood units Additionally, we compared the survival between the patients with borderline pre-capillary PH (25 > mPAP > 20 mmHg and PAWP ≤ 15 mmHg) with PVR ≥ 3WU and those with 2WU ≤ PVR < 3WU.
With respect to survival, all-cause mortality or lung transplantation was determined in all participants. We excluded patients with a history of PEA as it has been shown to improve the prognosis of patients with chronic thromboembolic pulmonary hypertension [4].
The follow-up data of RHC in the borderline PH and borderline pre-capillary PH groups was also examined.

Statistical analysis
We used univariate regression analysis to evaluate the baseline characteristics in each group. Student's t-tests and chi-square tests were used to compare continuous variables and categorical variables, respectively. The differences among the three groups were evaluated using one-way analysis of variance (ANOVA). The results are displayed as mean ± standard deviation or median (interquartile range) for continuous variables, and the number (%) for categorical variables. The Kaplan-Meier method, log-rank test, and Cox proportional hazards model were used to analyze the mortality and prognostic factors. Multivariate logistic regression analysis was used to identify the factors contributing to poor outcomes. A p-value < 0.05 was considered statistically significant. All analyses were performed using the JMP Pro software 13.2.0, Japanese version (SAS Institute Inc).

Conventional PH and borderline PH group (Study 1)
Among the patients who underwent RHC for the first time (n = 683), the number of patients in the conventional PH group was 531 (77.7%), and that in the borderline PH group was 50 (8.6% of total PH patients) (Fig. 1a).
Most of the patients with borderline PH were in Group 3 (30%) and Group 4 (52%) ( Table 1). However, according to the clinical classification, most of the patients with borderline PH among the total patients in their specific Groups were in Group 3 (15%), and only 6.8% were in Group 4 (Table 1).
Detailed analysis, based on etiology, showed that hemodynamics, alveolar-arterial oxygen difference (AaDO 2 ), gas exchange impairment, and 6-min walk distance were worst in the conventional group (Table 2). In addition, the partial pressure of arterial carbon dioxide (PaCO 2 ) was the lowest in the conventional PH group. In Group 3, no significant differences in parameters of the ventilatory function were observed among the conventional, borderline PH, and non-PH groups.
Next, we compared the survival of the conventional PH, borderline, and non-PH groups (Fig. 2). The survival of the conventional PH group was worse than that of the borderline group, and the worst among all three groups. Analogical tendencies were observed in Groups 1 and 4. However, in Group 3, the 10-year survival of all groups was < 40%. Group 3 showed poor prognosis, even in the non-PH group.
We divided the patients into two sub-groups according to the time of diagnosis (diagnosed in 1999-2009 or 2010-2020), since we recently tended to perform RHC only in patients who may benefit from treatment using vasodilators. The ratio of patients in the non-PH group diagnosed in 1999-2009 was higher than that diagnosed in 2010 (Table 3). Moreover, we analyzed the cause of death among the patients in Group 3. The number of patients who died due to malignant disease or who underwent lung transplantation was higher in the non-PH group than in the PH group (Table 4).   The number of patients in the conventional pre-capillary PH group was 489 (71.6%). Twenty-two patients (3.2% of the total patients, including the patients with non-PH; 4.3% of the patients with total pre-capillary PH) were included in the borderline pre-capillary PH group (Fig. 1b and Table 5). Similar to Study 1, most of the patients with borderline pre-capillary PH belonged to Groups 3 (40.9%) and 4 (45.5%). However, among the total patients with pre-capillary PH, most of the patients with borderline pre-capillary PH belonged to Group 3 (13.2%), and only 3.2% patients belonged to Group 4 ( Table 5).
Regarding baseline characteristics, in addition to hemodynamics, partial pressure of arterial oxygen (PaO 2 ), partial pressure of mixed venous oxygen (PvO 2 ), and AaDO 2 in the conventional pre-capillary PH group were significantly worse than those in the borderline precapillary PH group. In addition, PaCO 2 was significantly lower in the conventional pre-capillary PH group. There was no significant difference in the ventilatory function between the two groups.
Focusing on each etiological group, the conventional pre-capillary PH group showed more severe hemodynamics and gas exchange impairment than the borderline pre-capillary PH group in Groups 1, 3, and 4. A significantly lower PaCO 2 in the conventional pre-capillary PH group was observed only in Group 1 relative to that in the other Groups. Even in Group 3, there was no significant difference in the ventilatory function parameters between the two groups (Table 6).
Furthermore, the survival of the conventional precapillary PH group was worse than that of the borderline pre-capillary PH group; however, it did not reach statistical significance. The same tendency was observed in Group 1, 3, and 4 (Fig. 3). The 10-year-survival was the worst in Group 3 in the conventional PH group (Group 1, 50.1%; Group 3, 0.00%; Group 4, 68.0%) and the borderline pre-capillary PH group (Group 1, 50.9%; Group 3, 0.0%; Group 4, 67.0%). Age and conventional PH vs. borderline PH were poor prognostic factors; however, no parameter was statistically significant (Table 7).

Follow-up RHC data of patients with borderline PH and those with borderline pre-capillary PH
Among the patients with borderline PH, follow-up RHC was performed when the clinical condition was considered to worsen. Among 50 patients with borderline PH and 22 with borderline pre-capillary PH, only five patients underwent follow-up RHC (borderline PH = 4; borderline pre-capillary PH = 1), and only two patients were prescribed pulmonary vasodilators (borderline PH = 1; borderline pre-capillary PH = 1).
Only one patient's condition improved after treatment (Table 8).

Discussion
This is the first study conducted in a PH center in an Asian country to reveal the characteristics and survival of patients diagnosed with pre-capillary PH, according to the new diagnostic criteria proposed at the WSPH 2018. The borderline pre-capillary PH group accounted for 3.2% (22 in 683) of the total patient population who underwent their first RHC, and 4.3% of all patients with pre-capillary PH (22 in 511). The borderline PH group comprised 8.6% (50 in 581) of all patients with PH. Most of the patients in the borderline pre-capillary PH group belonged to Group 3 and 4. The survival of the borderline pre-capillary PH group tended to be better than that of the conventional pre-capillary PH group. Further, the prognosis of PH in Group 3 was the worst among the patients with borderline precapillary PH.
Most of the previous studies have focused on patients with borderline PH (25 > mPAP > 20 mmHg), and not on the pre-capillary nature of PH. In previous studies, the percentage of borderline PH has ranged from 4.2 to 18% among all patients and 4.5-22.6% of all patients with mPAP > 20 mmHg [5][6][7]. Further, Group 2 PH corresponds to the most common form of conventional PH due to left heart failure [8]. Previous studies have shown    Data are expressed as mean ± SD or n (%). mPAP, mean pulmonary arterial pressure; PVR, pulmonary vascular resistance; PAWP, pulmonary arterial wedge pressure; CO, cardiac output; 6MWD, 6-min walk distance; %VC, percent vital capacity; FEV1.0%, percent predicted forced expiratory volume in one second; %DLCO/VA, diffusing capacity of carbon monoxide by the alveolar volume; PaO 2 , partial pressure of arterial oxygen; PaCO 2 , partial pressure of arterial oxygen; PvO 2 , mixed venous oxygen tension; AaDO 2 , alveolar-arterial oxygen difference; WHO-FC, World Health Organization Functional Class; IPAH, idiopathic pulmonary hypertension; HPAH, hereditary pulmonary hypertension; PVOD, pulmonary veno-occlusive disease; PCH, pulmonary capillary hemangiomatosis; CTD, connective tissue disease; ILD, interstitial lung disease; COPD, chronic obstructive pulmonary disease; BE, bronchiectasis; PEA, pulmonary endarterectomy; PE, pulmonary embolism showed that among all patients, the percentage of those with mPAP between 19 and 24 mmHg, including non-PH patients (20.1%) was 18%, and that the majority of them belonged to Group 2 due to the presence of cardiovascular disease in > 70% of the patients [5]. Douschan et al. observed that 11.7% of all patients, including those without PH (35.2%), had borderline PH [6]. In their study, 20.3% of patients belonged to Group 2 with overt PH, and patients with borderline PH and those with overt PH showed a higher risk of cardiac disease. However, in these studies, detailed demographic data, including number of patients without PH, were not known [5,6]. Another study showed that the total percentage of patients with borderline PH was only 4.2%, including non-PH patients (5.7%) [7]. In that study, a relatively lower percentage of left heart disease (16.2%) and a relatively higher percentage of respiratory disease (29.7%) was observed among the patients with borderline PH and those without PH. However, the background status of all patients, including those with overt PH, is not known (Table 9) [7]. Only one study showed the proportion of patients with borderline pre-capillary PH; however, the number was quite small and did not show the prognosis [9].
In our study, the number of patients with borderline PH, including those in PH Groups 1−5, was 8.6% lower than that reported in Assad's and Douschan's studies (22.6% and 18.1%, respectively) ( Table 9), and most of the total patients belonged to Group 3 (14.6%) and 4 (55.6%). The relatively higher ratio of these groups can be explained by our PH center being associated with respiratory medicine and being one of the high-volume PEA centers in Japan. Furthermore, the proportion of patients in Group 2, which is the most common form, was quite low. In our study, the ratio of borderline PH was relatively high in Group 3. Similarly, a study on 15 patients with severe COPD who underwent lung transplantation also showed a mild elevation of mPAP (20−25 mmHg) in most patients [10]; therefore, the majority of patients with severe lung disease tended to have a mild elevation of mPAP. In contrast, our data showed that the ratio of borderline PH in Group 4 was low. The Papworth hospital study, which was also conducted in a PEA center, reported chronic thromboembolic disease with mPAP < 25 mmHg in only 42 of 1019 patients (4.1%) who underwent PEA [11]. The higher percentage of Group 4 patients in our center may explain the lower percentage of the borderline PH group in total. Overall, the number of patients who met the new diagnostic criteria depended on their background status. Survival (Study 2: Conventional PH and borderline PH group in pre-capillary PH, cut-off ≥ 3 WU). The survival of the conventional pre-capillary PH group was worse than that of the borderline pre-capillary PH group; however, the difference was not statistically significant. The same tendency was observed in Groups 1, 3 and 4. In Group 3, the 5-year and 10-year survival rates in the conventional pre-capillary PH population were 39.8% and 22.4%, respectively. The survival rates in the borderline pre-capillary PH group were 100% and 50%, respectively. PH, pulmonary hypertension; WU, Wood units Regarding baseline characteristics, in addition to hemodynamics, the PaO 2 , PvO 2 , and AaDO 2 were better in the borderline pre-capillary PH group than in the conventional pre-capillary PH group. Lower PaCO 2 in Group 1 of the conventional PH group might be suggestive of hyperventilation compensating for gas exchange impairment.
Several studies have shown little correlation between ventilatory function and severity of PH in patients with lung disease [12][13][14]. Similarly, in our study, there was no significant difference in ventilatory function between the conventional and the borderline pre-capillary PH groups in Group 3 (Table 6).
Regarding survival, in Study 1, the survival of the conventional PH group was worse than that of the borderline PH group. Similarly, in Study 2, the survival of the conventional pre-capillary PH group was worse than that of the borderline pre-capillary PH group; however, no significant difference was observed between the borderline PH and non-PH groups.
Previous data has suggested that mild elevation of PH is associated with poor prognosis in idiopathic pulmonary fibrosis [15] or chronic obstructive pulmonary disease [16,17]. Assad et al. also showed poor prognosis in patients with borderline PH, and the majority of patients seemed to be in Group 2 [5]. Douchan et al. revealed poorer prognosis and increased cardiopulmonary comorbidities in patients with mPAP of 17-26 mmHg than in those with mPAP < 17 mmHg [6]. They chose patients having a similar background status in both the PH and non-PH groups. Although a report including patients with relatively heterogeneous background diseases also showed poor prognosis of patients with borderline PH, the difference in the prognosis between patients with overt PH and those with borderline PH was detected when they focused on patients with Group 1 PH [7]. In our study, the patients in Group 3 had a poor prognosis, even in the non-PH group. Further, the number of non-PH patients was higher during 1999-2009 than that during 2010-2020, since we recently tended to perform RHC only in patients who may benefit from treatment with vasodilators. This means that most non-PH patients in 1999-2009 may not have received better treatment, compared with those diagnosed in 2010-2020. In addition, the number of patients who died due to malignant disease or who underwent lung transplantation was higher in the non-PH group. These underlying conditions may have affected the poor prognosis of the non-PH group in Group 3. Additionally, even when focusing on pre-capillary PH in Group 3, conventional PH vs. borderline PH and age were poor prognostic factors; however, the factors were not statistically significant. These data suggest that hemodynamics alone did not always determine the patients' prognosis in Group 3. Similarly, the Japanese Group 3 PH registry revealed that in combined pulmonary fibrosis and emphysema, and interstitial pneumonia, the major cause of death was either respiratory failure or progression or acute exacerbation of underlying disease [18]. The higher percentage of patients in Group 3 in our study may also explain why there was no significant difference in survival between the conventional PH and non-PH groups (Fig. 2).
Concerning survival with different PVR cut-offs, Xanthouli et al. recently showed that patients with precapillary borderline PH with PVR ≥ 2 WU had a significantly poorer prognosis than those with PVR < 2 WU in patients with systemic sclerosis [3]. Following this study, we compared the survival between patients with borderline pre-capillary PH (25 > mPAP > 20 mmHg and PAWP ≤ 15 mmHg) with PVR ≥ 3WU and those with PVR of 2-3 WU. There was no significant difference in survival between these two definition groups (Fig. 4). Additionally, both patients died due to underlying diseases other than right heart failure. These data indicated that patients with PVR of 2-3 WU may not always have a good prognosis, which corresponds to the findings of Xanthouli et al. regarding patients with scleroderma. Further studies are needed to confirm whether this new definition is beneficial in clinical practice, including the cut-off for PVR. Furthermore, the follow-up data for borderline PH and borderline pre-capillary PH was small, and very few patients were administered vasodilators during the follow-up ( Table 8). The poor prognosis and unknown cause of death in these patients warrants the necessity for a closer follow-up to detect the progression of PH. Accordingly, prospective studies are needed to evaluate whether the new definition is valuable in identifying patients with PH and those who require prescription of vasodilators.

Limitation
This was a retrospective single-center study, and the sample size was relatively small to evaluate the pre-capillary PH group effectively. Furthermore, the possibility of selection bias could not be discounted, since our PH center specializes in respiratory medicine and the incidence of PH due to heat failure in this study was low. Further, we could not examine extensive lung disease by computed tomography in Group 3.

Conclusions
This is the first study performed in a PH center in an Asian country to reveal the characteristics of patients with pre-capillary PH, according to the Nice 2018 definition. The Nice 2018 definition accounted for 4.3% of the patients with pre-capillary PH, and most of them were in Groups 3 and 4. It was suggested that hemodynamics alone may not determine the patients' prognosis. Further prospective studies are needed to determine whether this new definition is beneficial in clinical practice and provides relevant information regarding prescription of PHspecific treatment.