Abstract
Background Pulmonary endarterectomy (PEA) is the recommended treatment for eligible patients with chronic thromboembolic pulmonary hypertension (CTEPH). The Cambridge Pulmonary Hypertension Outcome Review (CAMPHOR) score is an internationally validated patient-reported outcome (PRO) measure for CTEPH. It assesses three domains: activity, quality of life (QoL) and symptoms. We assessed PROs in patients with CTEPH undergoing PEA.
Methods This retrospective observational study of consecutive CTEPH patients undergoing PEA at the UK national PEA centre between 2006 and 2017 assessed change in CAMPHOR score from baseline (pre-PEA) until up to 5 years post-PEA. CAMPHOR scores were compared between 1) those with and without clinically significant residual pulmonary hypertension and 2) those undergoing PEA and propensity-matched CTEPH patients who were not operated on. The minimally clinically important difference (MCID) was calculated using an anchor-based method.
Results Out of 1324 CTEPH patients who underwent PEA, 1053 (80%) had a CAMPHOR score recorded pre-PEA, 934 (71%) had a score recorded within a year of PEA and 784 (60%) had both. There were significant improvements between pre- and post-PEA in all three CAMPHOR domains (median±interquartile range activity −5±7, QoL −4±8, symptoms −7±8; all p<0.0001). Improvements in CAMPHOR score were greater and more sustained in those without clinically significant residual pulmonary hypertension. CTEPH patients undergoing PEA had better CAMPHOR scores than those not operated on. The MCID in CAMPHOR score was −3±5 for activity, −4±7 for QoL and −6±7 for symptoms.
Conclusions PROs are markedly improved by PEA in patients with CTEPH, more so in those without clinically significant residual pulmonary hypertension.
Abstract
Patients with CTEPH report significant improvement in patient-reported CAMPHOR scores after pulmonary endarterectomy compared with patients not operated on, but those with clinically significant residual pulmonary hypertension have less benefit https://bit.ly/2yU8V1v
Introduction
Chronic thromboembolic pulmonary hypertension (CTEPH) is an infrequent, but important, complication of pulmonary embolism. Organisation and fibrosis of thrombotic material leads to obstruction of proximal pulmonary arteries which, together with a secondary small-vessel vasculopathy, results in pulmonary hypertension [1]. CTEPH can lead to debilitating symptoms and functional impairment that affects quality of life (QoL) [2]. CTEPH patients who are eligible for surgery should be offered pulmonary endarterectomy (PEA), which involves removing obstructive thromboembolic material from surgically accessible pulmonary arteries [3]. PEA leads to improved symptoms and better survival in CTEPH [4, 5]. However, after PEA up to half of patients will have residual pulmonary hypertension (PH) that is associated with lower post-operative 6-min walk distance (6MWD) and worse World Health Organization functional class (WHO FC) [6, 7]. Treatment options for residual PH may include licenced drug therapy or balloon pulmonary angioplasty (BPA) [8–11].
Since 2008, the World Symposium on Pulmonary Hypertension has recommended including patient-reported outcome (PRO) measures as secondary end-points in clinical PH trials to assess health-related quality of life (HRQoL) [12]. However, PROs remain an underutilised outcome measure in clinical trials despite their relevance to patients [2]. This is partly because changes in PROs have often been more modest than other clinical trial end-points, such as 6MWD and pulmonary haemodynamics [13–15]. In addition, there has been a reliance on generic PRO measures in PH clinical trials, which may lack sensitivity in PH including CTEPH [16].
Several HRQoL measures have been assessed in CTEPH including generic (e.g. 36-Item Short Form Survey (SF-36) [17–19] and EuroQol-5-dimension (EQ-5D) [20, 21]) and PH-specific (EmPHasis-10 [22] and the Living with Pulmonary Hypertension questionnaire [23]) scores. Disease-specific PROs for PH/CTEPH may be more effective than generic measures in assessing HRQoL and predicting clinical deterioration [13–15]. The Cambridge Pulmonary Hypertension Outcome Review (CAMPHOR) score is the only PRO that was specifically developed in a cohort that included CTEPH and was validated in an adequate sample size [24]. It assesses three domains: activity, QoL and symptoms and is an independent predictor of clinical deterioration in CTEPH [14]. The minimally clinical important difference (MCID) is the smallest difference in outcome score that patients perceive as important and has not yet been defined for CAMPHOR [2].
The traditional clinical outcome measures in PH and CTEPH (6MWD, WHO FC, pulmonary haemodynamics and mortality) have several limitations. Survival following PEA has substantially improved, with in-hospital mortality <2.5% in some expert centres, limiting its usage as a future outcome measure [6]. Clinical symptoms may be disproportional to resting pulmonary haemodynamics, which is exemplified by chronic thromboembolic disease with a mean pulmonary artery pressure (mPAP) <25 mmHg [25]. Finally, functional outcome measures in PH may not correlate with meaningful outcomes such as hospitalisation, initiation of rescue therapy and death [26].
Treatment options for managing CTEPH have broadened to include PEA, BPA, riociguat [3, 8, 11], and medical therapies for pulmonary arterial hypertension (PAH) used off-label [4, 27, 28]. Current and future clinical trials will assess combination treatments particularly for residual PH post-PEA. Therefore, there is an increasing need for robust outcome measures that patients deem important including PROs in CTEPH clinical trials.
The aims of this study were to 1) investigate the change in CAMPHOR score in a large cohort of CTEPH patients undergoing PEA, 2) determine whether this varies in those with and without clinically significant residual pulmonary hypertension post-PEA, 3) compare CAMPHOR scores in those undergoing PEA with CTEPH patients not operated on, and 4) determine the MCID for CAMPHOR score in CTEPH.
Methods
Study participants
This retrospective study was approved by the Health Research Authority, UK (Integrated Research Application System project ID: 238805). Consecutive adult patients with CTEPH who underwent PEA at the national PEA centre (Royal Papworth Hospital, Cambridge, UK) from January 2006 to June 2017 were eligible. Study inclusion/exclusion criteria are summarised in supplementary figure S1. CTEPH was diagnosed according to international guidelines [29] and PEA was performed as described previously [5, 29]. Following PEA, patients were reviewed at 6 and 12 months at either Royal Papworth Hospital or their local specialist PH centre and annually thereafter for ≥5 years. Data were prospectively collected using a dedicated PH database. Clinical data (not including survival) and CAMPHOR scores were only available for patients at the time of review at Royal Papworth Hospital.
Patient characteristics at baseline and CAMPHOR score
Baseline data for all patients were recorded at the time of diagnostic right heart catheterisation and pre-operatively for patients with CTEPH undergoing PEA. This included CAMPHOR score, demographics, comorbidities, pulmonary vasodilator therapy, WHO FC, 6MWD, N-terminal pro-B-type natriuretic peptide (NT-proBNP) and haemodynamics. Operative data included type of surgical disease (Jamieson classification [30]), concomitant surgery, complications and length of stay. Outcome data, including CAMPHOR score, were recorded within 1 year of PEA. The CAMPHOR questionnaire assesses symptoms (25 questions), activity (15 questions) and QoL (25 questions). It takes 8–10 min to complete and is negatively weighted, with higher scores indicating poorer outcome. Symptoms and QoL are both scored out of 25, and activity out of 30.
Clinically significant residual PH post-PEA
To assess the effect of residual PH on longitudinal CAMPHOR score, patients were dichotomised into those with (mPAP ≥30 mmHg) and without (mPAP <30 mmHg) clinically significant residual PH post-PEA, and this threshold has been proposed in previous studies [6, 7]. This relationship was further investigated by stratifying the cohort into subgroups (mPAP post-PEA <25, 25–30, 31–36, 37–42, 43–48 and ≥49 mmHg). Additional analyses were performed by stratifying patients by post-PEA pulmonary vascular resistance (PVR) (supplementary material).
CAMPHOR score for operated and not-operated CTEPH
CAMPHOR scores from CTEPH patients undergoing PEA were compared with CTEPH patients who were not operated on between 2005 and 2015 at Royal Papworth Hospital. The groups were propensity matched for baseline CAMPHOR score, mPAP, age, sex and year of diagnosis. Matching was performed using the nearest neighbour method and a ratio of 3:1 (PEA:no-PEA).
MCID
The MCID in CAMPHOR score was calculated using an anchor-based method. Participants were asked to answer an “anchor” question that assessed their global change in health status since the last review (seven-point scale ranging from “very much better” to “very much worse”) in addition to completing the CAMPHOR questionnaire. The median change in CAMPHOR score from baseline (pre-PEA) to the first follow-up post-PEA was calculated for each point on this scale. The MCID was defined as the median change in CAMPHOR score associated with a “moderately better” health status change. The MCID was confirmed by statistical distribution methods and additionally by utilising the MCID for 6MWD that has previously been defined for PAH (supplementary material) [31]. Sensitivity analyses were performed in different subsets of operated CTEPH patients to confirm that MCIDs were robust (supplementary material).
Survival analysis
Survival from PEA until a census date of April 2018 was recorded using centralised national records. Multivariable Cox proportional hazards models were used to calculate hazard ratios and 95% confidence intervals and associations were confirmed by multiple imputing missing data (supplementary material). Cox models were checked for proportional hazards assumptions, influential observations and nonlinearity.
Statistical analyses
Groups were compared using the Mann–Whitney U-test for continuous data and the Cochran–Armitage test for WHO FC. A false discovery rate adjusted p-value was used to account for multiple testing. Data averages are described as median±interquartile range (IQR) unless otherwise specified; 95% confidence intervals of the median were calculated using bootstrapping (100 replicates) and the percentile method. Change from baseline (pre-PEA) were calculated as the median of differences. Pearson correlation coefficient described associations between complete pairs of outcomes or CTEPH severity variables including CAMPHOR score. Variables associated with CAMPHOR were assessed using multivariable linear models. Multiple imputation was used for a sensitivity analysis of longitudinal CAMPHOR scores to assess for bias in the complete cases analysis.
Results
In total, 1324 patients with CTEPH underwent PEA and were included in the main analysis. The cohort characteristics are summarised in table 1. Of these patients, 1053 (80%) had a CAMPHOR score recorded at baseline (pre-PEA), 934 (71%) had a score recorded at follow-up within a year of PEA and 784 (60%) had scores at both time points. The median±IQR time to follow-up CAMPHOR score was 139±79 days.
Patient demographics and characteristics at baseline
Median±IQR change from baseline (pre-PEA) to post-PEA follow-up (within 1 year) for the three CAMPHOR subscores indicated a significant improvement in all three domains: activity −5±7; QoL −4±8; and symptoms −7±8 (all p<0.0001; table 2). In addition, there were significant improvements in 6MWD, NT-proBNP, WHO FC and haemodynamics (all p<0.0001; table 2). The improvements in CAMPHOR score were sustained for 5 years post-PEA (figure 1, supplementary table S1). To investigate potential confounding from CTEPH patients with missing pre- and post-PEA CAMPHOR scores, multiple imputation was performed. As the change in CAMPHOR score from baseline in the multiple-imputed analysis was consistent with the complete cases analysis there was no significant bias from missingness (supplementary table S2).
Outcome measures pre-pulmonary endarterectomy (PEA), post-PEA and change from baseline in patients with chronic thromboembolic pulmonary hypertension (CTEPH)
Longitudinal summary of median Cambridge Pulmonary Hypertension Outcome Review (CAMPHOR) scores among patients with chronic thromboembolic pulmonary hypertension (CTEPH) undergoing pulmonary endarterectomy (PEA). a) Activity; b) quality of life (QoL); c) symptoms. Data are median score (95% CI) at each time point. The y-axis is truncated at 20 to improve visualisation (maximum scores: activity 30, QoL and symptoms 25). p-values are calculated with respect to baseline values. ****: p≤0.0001. The number of individuals and summary data for each time point are summarised in supplementary table S1.
Clinically significant residual PH post-PEA
Patients were dichotomised into those with (n=369), and those without (n=685) clinically significant residual PH post-PEA. Patient characteristics in these groups are summarised in supplementary table S3. There were significant post-operative improvements in CAMPHOR compared with baseline in the no residual PH group (median±IQR change from baseline: activity −5±7, QoL −5±9, symptoms −8±8; all p<0.0001) and residual PH group (activity −4±6, QoL −3±6, symptoms −6±7; all p<0.0001). Patients with no residual PH had greater improvements than those with residual PH (figure 2a). Furthermore, the improvements in the no residual PH group were better sustained over 5 years than in those with residual PH, who experienced a worsening trend in CAMPHOR score 2–3 years after PEA (figure 2a). When patients were stratified by mPAP or PVR post-PEA, there was a worsening of CAMPHOR scores as mPAP or PVR increased (figure 2b, supplementary figure S2).
a) Groups were dichotomised into those with (n=369) and without (n=685) clinically significant residual pulmonary hypertension (PH) (≥30 mmHg) assessed by right heart catheterisation within 1 year of pulmonary endarterectomy (PEA). i) Activity; ii) quality of life; iii) symptoms. Data are presented as median (95% CI) at each time point. p-values compare residual PH post-PEA and no residual PH post-PEA at each time point using the Mann–Whitney U-test. The number of individuals and summary data for each time point are summarised in supplementary table S3. b) Groups were subdivided by post-PEA mean pulmonary arterial pressure (mPAP) groups: <25 mmHg (n=410), 25–30 mmHg (n=181), 31–36 mmHg (n=120), 37–42 mmHg (n=78), 43–48 mmHg (n=43), >49 mmHg (n=33). i) Activity; ii) quality of life; iii) symptoms. The median score is plotted at each time point. 95% confidence intervals are not displayed, to improve visualisation. *: p≤0.05; **: p≤0.01; ***: p≤0.001; ****: p≤0.0001. CAMPHOR: Cambridge Pulmonary Hypertension Outcome Review.
CAMPHOR score for operated and not-operated CTEPH
There were 198 patients with CTEPH not operated on at Royal Papworth Hospital during 2005–2015. As the PEA and no-PEA CTEPH groups differed in baseline variables, including CAMPHOR score (supplementary table S4), propensity matching was performed. This resulted in 132 patients not operated on (no-PEA) matched to 396 patients who underwent PEA. The not operated group was heterogeneous and the reasons for not undergoing PEA included distal, surgically inaccessible disease (n=56), comorbidities (n=31), patient declined the operation (n=19) and other reasons (n=26), including a combination of these factors or limited disease/symptoms. The characteristics of the two groups following propensity matching are summarised in supplementary table S5. CAMPHOR scores at follow-up years 1 and 2 were significantly worse in the no-PEA group than in the PEA group (p<0.0001, all domains) (figure 3; supplementary table S5).
Cambridge Pulmonary Hypertension Outcome Review (CAMPHOR) score over time among propensity-matched individuals with chronic thromboembolic pulmonary hypertension who underwent (n=396) and did not undergo (n=132) pulmonary endartectomy (PEA). a) Activity; b) quality of life (QoL); c) symptoms. Groups were propensity matched (3:1) by baseline CAMPHOR score, mean pulmonary arterial pressure, age, sex and year of diagnosis. The median (95% CI) scores are plotted at each time point. p-values are calculated between PEA and no-PEA groups at each time point using the Mann–Whitney U-test. The number of individuals and summary data for each time point are summarised in supplementary table S5.
CAMPHOR correlation and association
There was a moderate negative correlation between the change from baseline to 1-year post-PEA in CAMPHOR activity score and 6MWD (Pearson correlation coefficient −0.4; supplementary figure S3). Correlations between the change from baseline to 1 year post-PEA in CAMPHOR score and changes in both WHO FC and haemodynamic parameters were relatively weak. A multivariable linear regression analysis of CAMPHOR domain scores at baseline demonstrated that pre-PEA CAMPHOR score (activity, QoL and symptoms) was associated with age, and the CAMPHOR activity domain was associated with 6MWD, but there was no association with haemodynamics (supplementary table S6).
MCID
The median±IQR MCID in CAMPHOR scores associated with a moderately better health status change following PEA were as follows: activity −3±5; QoL −4±7; and symptoms −6±7 (table 3). Importantly, the median change in CAMPHOR score from baseline (pre-PEA) to post-PEA follow-up within a year (described earlier; figure 1 and supplementary table S1) exceeded these MCID thresholds. The MCIDs were consistent across different subgroups in the sensitivity analyses (supplementary material). The changes in CAMPHOR score at 1 year post-PEA among patients with residual PH (described earlier; figure 2 and supplementary table S2) exceeded or equalled the MCID threshold for activity and symptoms, but not QoL.
Minimally clinically important difference (MCID) in Cambridge Pulmonary Hypertension Outcome Review (CAMPHOR) score for patients with chronic thromboembolic pulmonary hypertension undergoing pulmonary endarterectomy (PEA)
Survival
Age was the only variable at baseline (pre-PEA) that was significantly associated with 1-year survival post-PEA: hazard ratio (95% CI), 1.17 (1.02–1.35) per 5-year age increase (p=0.0288) (table 4). The three domains of the CAMPHOR score at baseline were not associated with 1-year survival following PEA. The same findings occurred when multiply imputed data were considered with the exception of a nominal association for sex (supplementary table S7). Additional survival analyses to investigate wether CTEPH patients with greater CAMPHOR improvement post-PEA also had improved long-term survival are presented in the supplementary material.
Cox proportional hazards model of baseline variables and survival 1 year post-pulmonary endarterectomy (PEA) in chronic thromboembolic pulmonary hypertension using complete cases
Discussion
This is the largest study to date of PROs in patients with CTEPH and in patients undergoing PEA. There were significant improvements in the CAMPHOR domains of activity, QoL and symptoms following PEA, and importantly these were sustained for up to 5 years. CTEPH patients with no clinically significant residual PH (mPAP <30 mmHg) had greater improvements in CAMPHOR compared to those with residual PH. In addition, CAMPHOR scores were better in patients who had undergone PEA compared with those who were not operated on. The MCID in each domain of the CAMPHOR score was established as activity −3 points, QoL −4 points and symptoms −6 points.
In contrast to previous findings [32], correlations between CAMPHOR scores and other clinical outcomes (WHO FC, pulmonary haemodynamics) were relatively weak. The only moderate correlation was between an improvement in CAMPHOR activity score and 6MWD. This suggests that improved physiological and functional outcome measures may not translate into improvements deemed important to patients such as QoL. Therefore, the CAMPHOR score provides important additional utility to assessing outcomes for CTEPH patients undergoing PEA.
Previous studies have reported improved PROs following PEA; however, they have been limited by reliance on generic PROs, small sample sizes or absence of MCIDs for the PRO tool [18, 33–35]. An advantage of the CAMPHOR score over generic and PH-specific PROs is its development and validation in CTEPH with multiple domains that are specific to CTPEH patients. The present study, in a large cohort of CTEPH patients undergoing PEA demonstrated an improvement in CAMPHOR score that exceeded the defined MCID and was sustained for up to 5 years. Patients with clinically significant PH following PEA had CAMPHOR activity and symptom improvements greater than the MCID. This is consistent with previous smaller studies using generic PROs [18, 34]. However, in the residual PH group the QoL change did not quite exceed the MCID threshold and improvements over 5 years were less sustained. As the improvements in CAMPHOR score following PEA were lower in patients with residual PH, additional treatment modalities including BPA and licenced pulmonary artery vasodilators need consideration in this group [8, 10]. Furthermore, as there was a graduation of worsening CAMPHOR scores with increasing post-PEA mPAP and PVR, additional treatment modalities following PEA may benefit from a stratified approach.
We have previously reported improved CAMPHOR score in a cohort of patients with chronic thromboembolic disease undergoing PEA and these improvements would exceed the MCID defined in the present study [25]. This confirms that in selected patients with chronic thromboembolic disease, PEA offers significant improvements in PROs including QoL, which is a more meaningful outcome than some traditional measures including haemodynamics. PROs such as CAMPHOR score should be utilised more widely in future studies of CTEPH as recommended by the World Symposium on Pulmonary Hypertension [36]. This is particularly relevant in scenarios where more modest improvements in functional, haemodynamic and survival outcome measures are expected, including residual PH post-PEA, chronic thromboembolic disease and clinical trials of additive and combination therapies in CTEPH.
Patients with CTEPH undergoing PEA had better longitudinal CAMPHOR scores than a propensity matched group of not-operated patients. This is consistent with a recent study in CTEPH that reported improved survival in patients undergoing PEA compared to those with operable disease but declining PEA surgery [37]. Our CAMPHOR data further substantiate that PEA is the guideline recommended treatment for patients with CTEPH as it results in improved PROs in addition to the known improved functional, haemodynamic and survival outcomes [4, 5].
This study has a number of strengths including the large sample size and prospective data collection of consecutive patients using a dedicated PH database, and it is the first study to establish the MCID for the CAMPHOR score in CTEPH. While patients were referred by a number of specialist PH centres, one limitation is that the majority of data included in the study were from a single national PEA centre. This contributed to incomplete longitudinal data for the CAMPHOR score and a different number of patients at each time point. This was addressed by multiple imputation that produced results similar to complete case analyses. Another limitation is that the not-operated CTEPH group were heterogenous in the reasons that PEA was not undertaken, and the analysis in this group should be considered exploratory. The MCIDs were defined for operated CTEPH patients and may not apply to the not-operated CTEPH cohort. Finally, there were incomplete data on follow-up medications, which limited an assessment of the effect of pulmonary vasodilator therapy on CAMPHOR score in the operated and not-operated groups.
In conclusion, PROs are markedly improved by PEA in patients with CTEPH, more so in those without clinically significant residual PH or those not operated on. The MCID was established for CTEPH, and PEA resulted in improvements greater than this threshold. This provides a benchmark against which future therapeutic interventions can be assessed and PROs should be utilised more widely in future CTEPH studies.
Supplementary material
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material ERJ-02096-2019.SUPPLEMENT
Shareable PDF
Supplementary Material
This one-page PDF can be shared freely online.
Shareable PDF ERJ-02096-2019.Shareable
Acknowledgements
We would like to thank the patients and staff at all participating pulmonary hypertension centres: Pulmonary Hypertension Centre and Adult Congenital Heart Centre, Royal Brompton Hospital, London, UK; Pulmonary Hypertension Service, Royal Free Hospital, London, UK; Scottish Pulmonary Vascular Unit, Golden Jubilee Hospital, Glasgow, UK; National Pulmonary Hypertension Service, Hammersmith Hospital, Imperial College Healthcare NHS Trust, London, UK; Pulmonary Vascular Disease Unit, Royal Hallamshire Hospital, Sheffield, UK; Northern Pulmonary Vascular Unit, Freeman Hospital, Newcastle, UK. Medical writing support was provided by Victoria Atess (Watermeadow Medical, an Ashfield Company, Oxfordshire, UK), funded by Actelion Pharmaceuticals Ltd.
Footnotes
This article has supplementary material available from erj.ersjournals.com
Conflict of interest: M. Newnham has nothing to disclose.
Conflict of interest: K. Bunclark has nothing to disclose.
Conflict of interest: N. Abraham has nothing to disclose.
Conflict of interest: S. Ali has nothing to disclose.
Conflict of interest: L. Amaral-Almeida has nothing to disclose.
Conflict of interest: J.E. Cannon reports grants from Actelion, during the conduct of the study; meeting travel support from Actelion and GSK, outside the submitted work.
Conflict of interest: N. Doughty has nothing to disclose.
Conflict of interest: C. Ng has nothing to disclose.
Conflict of interest: A. Ponnaberanam has nothing to disclose.
Conflict of interest: K. Sheares reports educational support from Actelion, Bayer and GSK, personal fees for consultancy from Actelion, outside the submitted work.
Conflict of interest: N. Speed has nothing to disclose.
Conflict of interest: D. Taboada reports other (speaker honoraria and education/travel grants) from Actelion, Bayer, GlaxoSmithKline, Lilly, MDS and Pfizer, outside the submitted work.
Conflict of interest: M. Toshner reports personal fees from Actelion/J&J and GSK, grants and personal fees from Bayer, grants from Merck, outside the submitted work.
Conflict of interest: S. Tsui has nothing to disclose.
Conflict of interest: D.P. Jenkins reports grants and personal fees for lectures from Bayer, personal fees for lectures and adjudication committee work from Actelion, outside the submitted work.
Conflict of interest: J. Pepke-Zaba has participated in advisory board activities for Actelion, GSK and MSD, given paid lectures for Actelion and MSD, and via institution has received research, education grants from Actelion and MSD.
Support statement: M. Toshner is supported by the National Institute for Health Research Cardiorespiratory Biomedical Research Centre.
- Received October 26, 2019.
- Accepted May 7, 2020.
- Copyright ©ERS 2020
References
