You are here

Article Text

Article menu
Heart failure and cardiomyopathies
Original article
Cardiopulmonary exercise test and sudden cardiac death risk in hypertrophic cardiomyopathy
  1. Damiano Magrì1,
  2. Giuseppe Limongelli2,
  3. Federica Re3,
  4. Piergiuseppe Agostoni4,5,
  5. Elisabetta Zachara3,
  6. Michele Correale6,
  7. Vittoria Mastromarino1,
  8. Caterina Santolamazza1,
  9. Matteo Casenghi1,
  10. Giuseppe Pacileo2,
  11. Fabio Valente2,
  12. Beatrice Musumeci1,
  13. Antonello Maruotti7,8,
  14. Massimo Volpe1,9,
  15. Camillo Autore1
  1. 1Department of Clinical and Molecular Medicine, University of Rome “La Sapienza”, Rome, Italy
  2. 2Cardiologia SUN, Monaldi Hospital, II University of Naples, Naples, Italy
  3. 3Cardiology Division, Cardiac Arrhythmia Center and Cardiomyopathies Unit, San Camillo-Forlanini Hospital, Rome, Italy
  4. 4Centro Cardiologico Monzino, IRCCS, Milan, Italy
  5. 5Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy
  6. 6Department of Cardiology, University of Foggia, Foggia, Italy
  7. 7Department of Scienze Economiche, politiche e delle lingue moderne, Libera Università SS Maria Assunta, Rome, Italy
  8. 8Centre for Innovation and Leadership in Health Sciences, University of Southampton, Southampton, UK
  9. 9IRCCS, Neuromed, Pozzilli (IS), Italy
  1. Correspondence to Professor Damiano Magrì, Department of Clinical and Molecular Medicine, “Sapienza” University, Rome, Italy; Cardiology Unit, Sant'Andrea Hospital, Via di Grottarossa 1037, Rome 00189, Italy; damiano.magri{at}uniroma1.it

Abstract

Background In hypertrophic cardiomyopathy (HCM), most of the factors associated with the risk of sudden cardiac death (SCD) are also involved in the pathophysiology of exercise limitation. The present multicentre study investigated possible ability of cardiopulmonary exercise test in improving contemporary strategies for SCD risk stratification.

Methods A total of 623 consecutive outpatients with HCM, from five tertiary Italian HCM centres, were recruited and prospectively followed, between September 2007 and April 2015. The study composite end point was SCD, aborted SCD and appropriate implantable cardioverter defibrillator (ICD) interventions.

Results During a median follow-up of 3.7 years (25th–75th centile: 2.2–5.1 years), 25 patients reached the end point at 5 years (3 SCD, 4 aborted SCD, 18 appropriate ICD interventions). At multivariate analysis, ventilation versus carbon dioxide relation during exercise (VE/VCO2 slope) remains independently associated to the study end point either when challenged with the 2011 American College of Cardiology Foundation/American Heart Association guidelines-derived score (C index 0.748) or with the 2014 European Society of Cardiology guidelines-derived score (C index 0.750). A VE/VCO2 slope cut-off value of 31 showed the best accuracy in predicting the SCD end point within the entire HCM study cohort (sensitivity 64%, specificity 72%, area under the curve 0.72).

Conclusions Our data suggest that the VE/VCO2 slope might improve SCD risk stratification, particularly in those HCM categories classified at low-intermediate SCD risk according to contemporary guidelines. There is a need for further larger studies, possibly on independent cohorts, to confirm our preliminary findings.

https://doi.org/10.1136/heartjnl-2015-308453

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Hypertrophic cardiomyopathy (HCM), the most common inherited heart disease, is characterised by markedly different clinical spectra but sudden cardiac death (SCD), albeit relatively rare, remains the most devastating clinical manifestation in this setting. Accordingly, given the efficacy of the implantable cardioverter defibrillator (ICD) for primary prevention of SCD, a systematic approach to risk stratification has become mandatory.1–3 In the last decades, there was a substantial agreement in considering SCD risk based on five binary risk factors, such as family history of SCD (FH-SCD), massive left ventricular (LV) hypertrophy, unexplained syncope, non-sustained ventricular tachycardia (NSVT) and abnormal blood pressure response at exercise (ABPRE).4 ,5 However, this approach tends to overestimate SCD risk and misses a significant number of fatal events in the low-intermediate risk categories.6 More recently, a 5-year risk prediction model, namely the HCM Risk-SCD, has been proposed and promptly incorporated in the 2014 European Society of Cardiology (ESC) guidelines.7 ,8 Notwithstanding, this strategy also suffers from a high percentage of missed events in the low-intermediate risk categories and possible implementation with additional variables have been hypothesised. In this context, an adjunctive cardiopulmonary exercise test (CPET) assessment might improve contemporary strategies, being most of the SCD risk factors implied in the pathophysiology of HCM exercise limitation.9–11

    The current multicentre study, besides challenging the HCM Risk-SCD model in an independent HCM cohort, investigated the ability of CPET-derived variables in improving SCD stratification in HCM over the contemporary strategies, particularly focusing on a possible incremental prognostic role in the low-intermediate SCD risk categories.

    Methods

    Study sample

    A total of 683 consecutive outpatients with HCM were recruited and prospectively followed in five tertiary HCM Italian centres between September 2007 and April 2015: Azienda Ospedaliera Sant'Andrea—‘Sapienza’ University–Rome (n.379); Azienda Ospedaliera San Camillo Forlanini–Rome (n.221); Ospedale Monaldi—Second University of Naples–Naples (n.52); Centro Cardiologico Monzino—University of Milan–Milan (n.20); Ospedali Riuniti—University of Foggia–Foggia (n.11). The diagnosis of HCM was based on a maximum wall thickness (MWT) ≥15 mm unexplained by abnormal loading conditions or in accordance with published criteria for the diagnosis of disease in relatives of patients with unequivocal disease.5 ,8 Patients with known metabolic diseases or syndromic causes of HCM (ie, Noonan syndrome) were excluded from the study as well as those with history of prior ventricular fibrillation or sustained ventricular tachycardia.

    The study complied with the ethical standards of the Declaration of Helsinki and was reviewed and approved by the institutional ethics committee of each of the five participating centres. Written informed consent was obtained from all participants.

    Patients clinical assessment

    Data were independently collected at each participating centre using uniform methodology. Each patient with HCM who fulfilled initial inclusion criteria underwent a de novo clinical assessment, including history taking with pedigree analysis and New York Heart Association classification, 24-h ECG Holter monitoring, transthoracic Doppler echocardiography and maximal CPET. The five usual SCD risk factors were collected: (A) FH-SCD (history of HCM-related SCD in at least one first-degree or other relatives ≤50 years); (B) massive LV hypertrophy (MWT ≥30 mm); (C) at least one run of NSVT (≥3 consecutive ventricular beats at a rate of ≥120 bpm and <30 s in duration on 24-h ECG Holter monitoring); (D) recent (≤5 years) unexplained syncope judged inconsistent with neurocardiogenic origin; (E) ABPRE (failure to increase systolic blood pressure (SBP), by at least 20 mm Hg from rest to peak exercise or a fall of ≥20 mm Hg from SBP).4 ,5 ,8

    The following echocardiographic measurements were considered: LV end-diastolic diameter (parasternal long axis), the greatest LV thickness (MWT, measured at any LV site), left atrial diameter (parasternal long axis), the highest maximal LV outflow tract gradient among those measured at rest, in orthostatic position and after Valsalva manoeuvre (apical four-chamber view), and LV ejection fraction with Simpson's biplane methods (LVEF, apical four-chamber view).12

    All CPETs were performed using an electronically braked cycle ergometer equipped with metabolic carts (see online supplementary table 1S). A personalised ramp exercise protocol was performed, aiming at a test duration of 10±2 min.13 The exercise was preceded by few minutes of resting breath-by-breath gas exchange monitoring and by a 3-min unloaded warm-up. In the absence of clinical events, CPET was interrupted when patients stated that they had reached maximal effort. However, we considered metabolic maximal effort as achieved if the respiratory exchange ratio was above 1.05. A 12-lead ECG, diastolic blood pressure and SBP were recorded during CPET, in order to obtain the following parameters: rest heart rate (HR), peak HR, %pHR ((peak HR/(220−age))×100), and ΔSBP (peak SBP−rest SBP). A breath-by-breath analysis of expiratory gases and ventilation (VE) has been performed, and peak values were obtained in the last 20 s of exercise. The predicted peak oxygen consumption (VO2) was determined by using the gender-adjusted, age-adjusted and weight-adjusted formula.14 Circulatory power (CP=peak VO2×SBP) was obtained both considering the peak VO2 value in terms of ml/kg (CP) as well as in terms of percentage of predicted (CP%).15 VO2/workload relationship was measured throughout the entire exercise. Anaerobic threshold (AT) was measured by V-slope analysis of VO2 and VCO2, and confirmed by ventilator equivalents and end-tidal pressures of CO2 and O2. The end of the isocapnic buffering period was identified when VE/VCO2 increased and end-tidal pressure of CO2 decreased. VE/VCO2 slope was calculated as the slope of the linear relationship between VE and VCO2 from the 1st minute after the beginning of the loaded exercise and the end of the isocapnic buffering period.14 ,16 Within each centre, all CPETs were analysed by a single expert physician blinded to the main clinical data and, when it was needed, data were reviewed again by a second reader from one of the remaining participating centres.

    Clinical outcomes

    All patients had planned clinical reviews every 6–12 months or earlier according to the clinical status. Follow-up duration was defined as the time interval between CPET execution and either the first event or the last visit/telephone interview in case of no events. The primary composite end point was represented by SCD or an equivalent event. SCD was defined as witnessed sudden death with or without documented ventricular fibrillation or death within 1 h of new symptoms or nocturnal deaths with no antecedent history of worsening symptoms.2 ,3 Aborted SCD during follow-up and appropriate ICD therapies (defined as intervention triggered by ventricular fibrillation or rapid ventricular tachycardia at >180 bpm) were considered equivalent to SCD in accordance with previous studies.6 ,7 ,17 The causes of death, as well as the other events, were ascertained by experienced cardiologists at each centre using hospital and primary healthcare records, death certificates, postmortem reports, and interviews with relatives and/or physicians.

    Statistical analysis

    Statistical analysis was performed using R (R Development Core Team, 2014). A p value lower than or equal to 0.05 was generally considered as statistically significant. Preliminarily, an extension of the Shapiro-Wilk test of normality was performed. Unless otherwise indicated, all data are expressed as mean±SD. Categorical variables were compared with a difference between proportion tests; a two-sample t test was used to compare the continuous data between the group with the SCD end point (SCD group) and the one without (no SCD group). In comparing the two populations, the variance was estimated separately for both groups and the Welch-Satterthwaite modification to the degrees of freedom was used. The HCM Risk-SCD score was obtained by using the calculator provided with the 2014 ESC guidelines at (http://www.escardio.org/guidelines-surveys/esc-guidelines/Pages/hypertrophic-cardiomyopathy.aspx).8 Then, based on the 2011 American College of Cardiology Foundation (ACCF)/American Heart Association (AHA) guidelines, a continuous risk score was estimated by considering NSVT and ABPRE as minor risks factors and codifying them as major risk factors only when they occurred simultaneously with at least one of the other risk factors. We therefore focused on the distribution of the survival times by adopting the Cox proportional-hazards regression model. We performed a backward selection of the predictors to be included in the model. A 15% significance level was used in the backward elimination procedure to select covariates for the final multivariate model. To avoid the inclusion of collinear variables in the multivariate Cox analysis, we built several models in which VO2-derived variables, known to be collinear, were added to the prognostic model one at a time (maximum number of variables included in the model equal to nine, see table 3 legend). We retain the model with the best trade-off between model complexity and model fit judged by the log-likelihood. Other approaches have been also investigated as the backward elimination procedure may lead to overfitting issues. Indeed, following Pavlou and colleagues,18 we used lasso regression to perform variable selection and model fitting. However, although the lasso regression penalty shrinks away the less important variables yielding interpretable results, even this approach may suffer from important drawbacks.19 ,20 We also performed a calibration analysis. We computed the average calibration error for both approaches and tested the observed versus average predicted probabilities for each class of risk. The Brier quadratic error score and a χ2 test of goodness of fit based on the Brier score were also checked. We did not find any clear indication of overfitting from the post hoc analysis and, consequently, in the present paper we simply reported results based on the backward elimination procedure only. Discrimination of variables included in the final multivariate model specification was performed by Harrell's C index. Therefore we investigated proportional hazards assumption by tests and graphical diagnostics based on scaled Schoenfeld residuals. Test of proportional hazards assumption for each covariate was obtained by correlating the corresponding set of scaled Schoenfeld residuals with the Kaplan-Meier estimate of the survival distribution. To check for the presence of influential observations, we produced a matrix of estimated changes in the regression coefficients upon deleting each observation in turn and comparing the magnitudes of the largest values to the regression coefficients. Lastly, a possible incorrect specification of the parametrical part of the model is investigated, that is, we checked for non-linear effects of the VE/VCO2 slope. To detect non-linearity, we provided plots of martingale residuals against covariates. Harrell's C and the equivalent parameter Somers’ D were proposed as measures of discrimination. A receiver-operating characteristic (ROC) analysis has been considered to determine the predictive capability of the VE/VCO2 slope in identifying the SCD end point and its cut-off value was identified according to the accuracy ((true positive+true negative)/total sample) of the highest value. At last, to quantify the improvement of our proposal with previous suggested approaches, we also provided results on the integrated discrimination improvement (IDI) and net reclassification improvement (NRI).21

    Results

    Starting from an initial sample of 683 consecutive outpatients with HCM, 60 patients were excluded because metabolic maximal effort was not achieved or because of poor quality CPET data. Notably the excluded patients’ characteristics did not differ from those enrolled in the study except for a significantly older age and a lower male prevalence (respectively 56±9 years and 51%, p<0.001 for both). No major adverse events were reported during CPET execution. A total 623 patients was therefore considered and the main clinical data are reported in table 1. Within the enrolled sample, 211 patients (34%) had family history of HCM and 44 patients (7%) had an HCM diagnosis in accordance with published criteria for the diagnosis of disease in relatives of patients with unequivocal disease.5 ,8 Atrial fibrillation was documented in 33 patients (5%) and 22 patients (3%) had an HCM at the end-stage phase (defined as echocardiographic evidence of dilated LV cavity with LVEF lower than 50%). Furthermore 160 patients (26%) had systemic hypertension, 36 (6%) had coronary artery disease and 30 (5%) had diabetes. During the follow-up period, an ICD was further implanted in 63 patients (10%) and 44 (7%) underwent a surgical myectomy.

    View this table:
    Table 1

    Main clinical variables of the entire study sample at the study run-in (623 patients)

    SCD end point analysis

    Median follow-up was 3.7 years (25th–75th centile: 2.2–5.1 years) with a total of 2355 patient-years. During the follow-up period, 25 (4%) patients reached the primary SCD end points with an SCD 5-year cumulative hazard equal to 7.5% (95% CI 4.0% to 11.0%). Specifically, 3 patients died suddenly, 4 patients experienced aborted SCD and 18 had appropriate ICD therapies (12 ICD shocks and 6 antitachycardia pacing for ventricular tachycardia at >200 bpm). During the follow-up, 16 patients with ICD (11.5%) had at least one inappropriate shock due to fast atrial arrhythmias, the total number of inappropriate ICD shocks being equal to 21 (15%). Each patient who ended the follow-up period before the 5th year was considered censored at the time of the last visit or telephone interview.

    Table 2 shows a detailed comparison between the two study groups. Within the SCD group, two patients (8%) had atrial fibrillation and four (16%) had an HCM at the end-stage phase. Moreover, during the follow-up, within the SCD group, an ICD was implanted in seven more patients (28%) whereas nobody underwent a surgical myectomy.

    View this table:
    Table 2

    Comparison between HCM subgroup according the primary end point at 5 years

    The univariate model based on 2014 ESC guidelines produced a C index of 0.687 and, similarly, the model with a continuous risk score based on 2011 ACCF/AHA guidelines produced a C index of 0.665. Several single guidelines proposed SCD risk factors as well as most CPET-derived covariates were found significantly associated at the univariate Cox regression model (table 3). Notwithstanding, covariates showing significant independent effects remained the following: 2011 ACCF/AHA or 2014 ESC guidelines-derived scores and VE/VCO2 slope (table 4).

    View this table:
    Table 3

    Significant univariate Cox proportional survival analysis according to the main variables for the SCD end point

    View this table:
    Table 4

    Significant multivariate Cox proportional survival analysis and test for proportional hazards assumption for the SCD end point

    The multivariate Cox proportional survival analysis model with a risk score based on 2014 ESC guidelines and VE/VCO2 slope produced a C index of 0.750. The model with a continuous risk score based on the 2011 ACCF/AHA guidelines and VE/VCO2 slope produced a C index of 0.748 (table 4). According to the selected models, Harrell's C index was 0.75, and Somers’ D index 0.5. We also executed a sensitivity analysis of model assumptions: proportional hazards cannot be rejected; atypical data do not influence results and considering linear relationships in the model specification results appropriate (see online supplementary file, model assumptions). The ROC analysis identified a VE/VCO2 slope=31 as the cut-off value with the best accuracy in predicting the SCD end point within the entire HCM study cohort (figure 1). A total of 183 patients (29%) has a VE/VCO2 slope ≥31, their prevalence increasing in parallel with the SCD risk according to contemporary algorithms (see online supplementary table 2S). Data of accuracy for the historical five conventional SCD risk factors were also explored and supplied in table 5.

    View this table:
    Table 5

    Accuracy of the historical five conventional SCD risk factors and of the proposed cut-off value for the ventilatory efficiency (VE/VCO2 slope)

    Figure 1
    Figure 1

    Upper panel: Receiver-operating characteristic (ROC) curve showing the point with the best sensitivity and specificity of ventilatory efficiency (VE/VCO2 slope) value in the entire hypertrophic cardiomyopathy (HCM) study sample (equal to 31). Lower panel: Kaplan-Meier estimator of sudden cardiac death (SCD) and its surrogate events (aborted SCD and appropriate implantable cardioverter defibrillator therapies). VE/VCO2 slope, relation between ventilation versus carbon dioxide production, area under curve (AUC).

    Table 6 shows the simulated effect of using each of the contemporary strategies with different thresholds as well as the simulated effect of using the VE/VCO2 slope ≥31 to select the possible adjunctive candidate to the ICD implant in those categories currently classified at low-intermediate SCD risk (figure 1). The expected number of patients needed to treat with an ICD implant to save one patient at 5 years (NNTx-5 years) was also supplied. The latter was calculated according to the standard formula: 1/((patients receiving an ICD in the SCD group/total patients receiving an ICD)−patients who are not receiving an ICD in the SCD group/total patients not receiving an ICD).

    View this table:
    Table 6

    The simulated effect of using different strategies and thresholds of 5-year SCD end point to implant an ICD in the study cohort

    An improvement was estimated with respect to ESC 2014 guidelines (NRI=0.241 (0.007 to 0.451), p=0.040; IDI: 0.037 (0.003 to 0.103), p=0.040)) and, albeit less pronounced, with respect to 2011 ACCF/AHA guidelines (NRI=0.214 (0.001 to 0.432), p=0.044; IDI: 0.017 (−0.001 to 0.070), p=0.139).

    Discussion

    The present study, conducted on a well characterised multicentre cohort of consecutive outpatients with HCM, represents the first attempt to introduce CPET-derived variables in HCM SCD risk management and suggests that the VE/VCO2 slope might improve the accuracy of contemporary strategies, mainly in those categories currently classified at low-intermediate risk.

    The identification of patients with HCM who deserve an ICD therapy for SCD primary prophylaxis still remains a debated topic.1–3 ,6 ,22 The HCM Risk-SCD model7 is certainly young and, up to now, there is only one validation paper showing that its clinical application improves SCD risk management with respect to 2003 AHA/ESC and the 2011 ACCF/AHA strategies.23 The present paper confirms a satisfying HCM Risk-SCD model predictive power in the high-risk categories but it contextually shows that a correct application of the 2011 ACCF/AHA strategy yields similar results. Nonetheless, neither the American nor the new European guidelines exhaustively answer the thorny question about the low-intermediate SCD risk categories. Accordingly, the HCM Risk-SCD paper acknowledged 49% of fatal events in patients with <6% risk, this finding is in line with the recent validation paper as well with our data. Furthermore, even using a 4% HCM Risk-SCD cut-off value, the original7 and its validation paper23 acknowledged 30% of potentially missed SCD events, this percentage rising up to 40% in our cohort. Similarly, we also showed a rate of potentially missed SCD events equal to 32% for the 2011 ACCF/AHA guidelines. In this context, investigating possible implementation of contemporary strategies, the actual study evaluated advantages coming from CPET-derived variables. Indeed, albeit there are well known strong outcome predictors in patients with heart failure with and without systolic dysfunction,24–26 up to now, the HCM guidelines relegate the CPET analysis in patients in whom uncertainties persist with regards to presence or severity of heart failure symptoms.5 ,8 Furthermore, two recent retrospective studies demonstrated an original association between peak VO2 and the VE/VCO2 slope and overall mortality in large HCM cohorts.27 ,28 Conversely, no study has yet challenged the CPET parameters specifically with SCD risk, notwithstanding most of the substrates implied into the pathophysiology of exercise limitation in HCM are in common with those thought to increase SCD risk, including LV hypertrophy with myocardial fibre disarray, diastolic dysfunction, left ventricular outflow tract obstruction, subendocardial ischaemia, as well myocardial interstitial fibrosis.9–11 Interestingly we found that, among several CPET-derived variables, the VE/VCO2 slope remained the only one independently associated to the SCD end point at the multivariate analysis including the guidelines algorithms. From a pathophysiological viewpoint, the VE/VCO2 slope has been shown as the CPET variable with the best accuracy for resting pulmonary capillary wedge pressure and LV diastolic properties in patients with HCM.29 Supporting a possible link between this CPET variable and SCD risk, it has been recently shown that the abrupt energetic deficit in HCM during exercise represents the mechanism underlying the exercise-related diastolic dysfunction and it has been hypothesised that it might also represent a potential arrhythmic mechanism (ie, sarco/endoplasmic reticulum Ca2+ ATPasi dysfunction with Ca2+ overload).30 Thus, albeit merely speculative, it is conceivable that the VE/VCO2 slope has the potential to unmask an exercise-induced diastolic function derangement given that the ventilation/perfusion mismatch phenomena due to pulmonary arterial constriction increase in parallel with the exercise-related LV filling pressure increase.31 ,32 Although the VE/VCO2 slope should be physiologically evaluated as a continuous variable in stratifying HCM SCD risk, we also obtained a cut-off value equal to 31 as most accurate for the study end point and we investigated its utility in identifying further ICD indications among the intermediate SCD risk categories according to the guidelines. Nonetheless, considering the high rate of potentially missed events in the 2014 ESC 3–4% category,9 ,29 we included the latter subgroup to the official 2014 ESC intermediate risk category, thus arbitrarily creating a low-intermediate risk group. Indeed, the implementation of the above mentioned VE/VCO2 cut-off value would have resulted in adjunctive useful ICD implants in three patients (12%) of the 2011 ACCF/AHA intermediate risk category, in two patients (8%) of the 2014 ESC intermediate risk category and, even, in five patients (20%) of the proposed 2014 ESC low-intermediate risk category (figure 2). Noteworthy, albeit the absolute number of ICD candidates significantly increased, the NNTx-5 years remained acceptable, the simulated number of ICDs implanted to abort one life-threatening event at 5 years ranging from 11 to 14.

    Figure 2
    Figure 2

    Diagram showing the simulated effect, in the present study cohort (623 patients), of using a ventilatory efficiency (VE/VCO2 slope) cut-off values ≥31 to improve the strategies for implantable cardioverter defibrillator (ICD) implantation in primary prevention in the 2014 European Society of Cardiology (ESC) (A) and 2011 American Heart Association (AHA) (B) intermediate SCD risk categories. (C) Showed possible advantages coming from the adoption of the suggested VE/VCO2 slope cut-off values into a 2014 ESC low-intermediate SCD risk category. HCM: hypertrophic cardiomyopathy; Pts: patients; SCD: sudden cardiac death; VE/VCO2 slope, relation between ventilation versus carbon dioxide production.

    Limitations

    The small number of patients enrolled, mainly derived from three out of the five participating centres, together with the low number of SCD events represents an obvious limitation that allows us to suggest, rather than to affirm, that the VE/VCO2 slope analysis might improve HCM SCD risk stratification in the low-intermediate risk categories. However, it is remarkable that no risk stratification strategy will ever be able to predict SCD with absolute certainty as well as that each scoring system should always be used by physicians highly expert in HCM management and, possibly, in CPET analysis. The low positive predictive value of every established SCD risk factor in our cohort (table 5) reinforces this concept. Nonetheless, it should be underlined that the sensitivity and specificity of the VE/VCO2 slope cut-off point derived in the same data where it is being tested in will naturally have superior performance compared with externally derived metrics. The lack of data on cardiac resonance imaging as well as the fact that a possible impact of different gene mutations has not been investigated represent obvious limitations.33 ,34 Furthermore, we acknowledge that our SCD risk analysis considered syncopal episodes occurred ≤5 years instead of <1 year.5 At last we acknowledge that we examined the prognostic impact of several variables at a single time point, thus we cannot exclude that changes in some variables, as for instance an upgrading of treatment during follow-up or upcoming risk factors, had altered our survival analysis.

    Key messages

    What is already known on this subject?

    • Sudden cardiac death (SCD) risk stratification in hypertrophic cardiomyopathy (HCM) still remains a challenging topic and is currently managed according to the 2011 American College of Cardiology Foundation/American Heart Association guidelines and the early released 2014 European Society of Cardiology guidelines.

    What might this study add?

    • Besides a validation of the new risk prediction model (HCM Risk-SCD) in a large and independent cohort of patients with HCM, the present represents the first attempt to introduce cardiopulmonary exercise test-derived variables in SCD risk management in HCM.

    How might this impact on clinical practice?

    • HCM Risk-SCD confirms its power in stratifying SCD risk in the high-risk categories. However, the ventilatory efficiency (ventilation versus carbon dioxide relation during exercise, VE/VCO2 slope) significantly improves the accuracy of contemporary SCD risk stratification strategies, particularly in those HCM categories classified at low-intermediate risk.

    References

    1. ↵
      1. Maron BJ,
      2. Maron MS
      . Hypertrophic cardiomyopathy. Lancet 2013;381:242–55. doi:10.1016/S0140-6736(12)60397-3
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. O'Mahony C,
      2. Elliott PM
      . Prevention of sudden cardiac death in hypertrophic cardiomyopathy. Heart 2014;100:254–60. doi:10.1136/heartjnl-2012-301996
      OpenUrlFREE Full Text
    3. ↵
      1. Elliott PM,
      2. Poloniecki J,
      3. Dickie S, et al
      . Sudden death in hypertrophic cardiomyopathy: identification of high risk patients. J Am Coll Cardiol 2000;36:2212–18. doi:10.1016/S0735-1097(00)01003-2
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Maron BJ,
      2. McKenna WJ,
      3. Danielson GK, et al
      . American College of Cardiology/European Society of Cardiology Clinical Expert Consensus Document on Hypertrophic Cardiomyopathy: a report of the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents and the European Society of Cardiology Committee for Practice Guidelines. J Am Coll Cardiol 2003;42:1687–713. doi:10.1016/S0735-1097(03)00941-0
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Gersh BJ,
      2. Maron BJ,
      3. Bonow RO, et al
      . 2011 ACCF/AHA Guideline for the Diagnosis and Treatment of Hypertrophic Cardiomyopathy: Executive Summary: A Report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation 2011;124:2761–96. doi:10.1161/CIR.0b013e318223e230
      OpenUrlFREE Full Text
    6. ↵
      1. O'Mahony C,
      2. Tome-Esteban M,
      3. Lambiase PD, et al
      . A validation study of the 2003 American College of Cardiology/European Society of Cardiology and 2011 American College of Cardiology Foundation/American Heart Association risk stratification and treatment algorithms for sudden cardiac death in patients with hypertrophic cardiomyopathy. Heart 2013;99:534–41. doi:10.1136/heartjnl-2012-303271
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. O'Mahony C,
      2. Jichi F,
      3. Pavlou M, et al
      . A novel clinical risk prediction model for sudden cardiac death in hypertrophic cardiomyopathy (HCM Risk-SCD). Eur Heart J 2014;35:2010–20. doi:10.1093/eurheartj/eht439
      OpenUrlAbstract/FREE Full Text
    8. ↵
      Authors/Task Force MembersElliott PM, Anastasakis A, et al. 2014 ESC Guidelines on diagnosis and management of hypertrophic cardiomyopathy: the Task Force for the Diagnosis and Management of Hypertrophic Cardiomyopathy of the European Society of Cardiology (ESC). Eur Heart J 2014;35:2733–79. doi:10.1093/eurheartj/ehu284
      OpenUrlFREE Full Text
    9. ↵
      1. Sachdev V,
      2. Shizukuda Y,
      3. Brenneman CL, et al
      . Left atrial volumetric remodeling is predictive of functional capacity in nonobstructive hypertrophic cardiomyopathy. Am Heart J 2005;149:730–6. doi:10.1016/j.ahj.2004.07.017
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Payá E,
      2. Marín F,
      3. González J, et al
      . Variables associated with contrast-enhanced cardiovascular magnetic resonance in hypertrophic cardiomyopathy: clinical implications. J Card Fail 2008;14:414–19. doi:10.1016/j.cardfail.2008.02.006
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Magrì D,
      2. Agostoni P,
      3. Cauti FM, et al
      . Determinants of peak oxygen uptake in patients with hypertrophic cardiomyopathy: a single-center study. Intern Emerg Med 2014;9:293–302. doi:10.1007/s11739-012-0866-x
      OpenUrlCrossRefPubMed
    12. ↵
      1. Lang RM,
      2. Bierig M,
      3. Devereux RB, et al
      . Recommendation for chamber quantification: a report from the American Society of Echocardiography's Guidelines and Standard Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Society of Echocardiograph, a brunch of the European Society of Cardiology. J Am Soc Echocardiogr 2005;18:1440–63. doi:10.1016/j.echo.2005.10.005
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Agostoni P,
      2. Bianchi M,
      3. Moraschi A, et al
      . Work-rate affects cardiopulmonary exercise test results in heart failure. Eur J Heart Fail 2005;7:498–504. doi:10.1016/j.ejheart.2004.06.007
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Wasserman K,
      2. Hansen JE,
      3. Sue DY, et al
      . Principles of exercise testing and interpretation. 4th edn. Baltimore: Lippincott Williams & Wilkins, 2005; 10–65.
    15. ↵
      1. Corrà U,
      2. Mezzani A,
      3. Giordano A, et al
      . Exercise haemodynamic variables rather than ventilatory efficiency indexes contribute to risk assessment in chronic heart failure patients treated with carvedilol. Eur Heart J 2009;30:3000–6. doi:10.1093/eurheartj/ehp138
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Balady GJ,
      2. Arena R,
      3. Sietsema K, et al
      . Clinician's Guide to cardiopulmonary exercise testing in adults: a scientific statement from the American Heart Association. Circulation 2010;122:191–225. doi:10.1161/CIR.0b013e3181e52e69
      OpenUrlFREE Full Text
    17. ↵
      1. Maron BJ,
      2. Rowin EJ,
      3. Casey SA, et al
      . Hypertrophic cardiomyopathy in adulthood associated with low cardiovascular mortality with contemporary management strategies. J Am Coll Cardiol 2015;65:1915–28. doi:10.1016/j.jacc.2015.02.061
      OpenUrlCrossRefPubMed
    18. ↵
      1. Pavlou M,
      2. Ambler G,
      3. Seaman R, et al
      . How to develop a more accurate risk prediction model when there are few events. BMJ 2015;351:h3868. doi:10.1136/bmj.h3868
      OpenUrlFREE Full Text
    19. ↵
      1. Tibshirani R
      . Regression shrinkage and selection via de lasso. JRSS-B 1996;58:267–88.
      OpenUrl
    20. ↵
      1. Goeman JJ
      . L1 penalized estimation in the Cox proportional hazards model. Biom J 2010;52:70–84. doi:10.1002/bimj.200900028
      OpenUrlPubMedWeb of Science
    21. ↵
      1. Pencina MJ,
      2. D'Agostino RB,
      3. Vasan RS
      . Statistical methods for assessment of added usefulness of new biomarkers. Clin Chem Lab Med 2010;48:1703–11. doi:10.1515/CCLM.2010.340
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Ostman-Smith I,
      2. Wisten A,
      3. Nylander E, et al
      . Electrocardiographic amplitudes: a new risk factor for sudden death in hypertrophic cardiomyopathy. Eur Heart J 2010;31:439–49. doi:10.1093/eurheartj/ehp443
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Vriesendorp PA,
      2. Schinkel AF,
      3. Liebregts M, et al
      . Validation of the 2014 European Society of Cardiology guidelines risk prediction model for the primary prevention of sudden cardiac death in hypertrophic cardiomyopathy. Circ Arrhythm Electrophysiol 2015;8:829–35. doi:10.1161/CIRCEP.114.002553
      OpenUrlAbstract/FREE Full Text
    24. ↵
      1. Agostoni P,
      2. Corrà U,
      3. Cattadori G, et al
      . Metabolic exercise test data combined with cardiac and kidney indexes, the MECKI score: a multiparametric approach to heart failure prognosis. Int J Cardiol 2013;167:2710–18. doi:10.1016/j.ijcard.2012.06.113
      OpenUrlCrossRefPubMed
    25. ↵
      1. Guazzi M,
      2. Myers J,
      3. Arena R
      . Cardiopulmonary exercise testing in the clinical and prognostic assessment of diastolic heart failure. J Am Coll Cardiol 2005;46:1883–90. doi:10.1016/j.jacc.2005.07.051
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Guazzi M,
      2. Labate V,
      3. Cahalin LP, et al
      . Cardiopulmonary exercise testing reflects similar pathophysiology and disease severity in heart failure patients with reduced and preserved ejection fraction. Eur J Prev Cardiol 2014;21:847–54. doi:10.1177/2047487313476962
      OpenUrlAbstract/FREE Full Text
    27. ↵
      1. Finocchiaro G,
      2. Haddad F,
      3. Knowles JW, et al
      . Cardiopulmonary responses and prognosis in hypertrophic cardiomyopathy: a potential role for comprehensive noninvasive hemodynamic assessment. JACC Heart Fail 2015;3:408–18. doi:10.1016/j.jchf.2014.11.011
      OpenUrlCrossRefPubMed
    28. ↵
      1. Masri A,
      2. Pierson LM,
      3. Smedira NG, et al
      . Predictors of long-term outcomes in patients with hypertrophic cardiomyopathy undergoing cardiopulmonary stress testing and echocardiography. Am Heart J 2015;169:684–92. doi:10.1016/j.ahj.2015.02.006
      OpenUrlCrossRefPubMed
    29. ↵
      1. Arena R,
      2. Owens DS,
      3. Arevalo J, et al
      . Ventilatory efficiency and resting hemodynamics in hypertrophic cardiomyopathy. Med Sci Sports Exerc 2008;40:799–805. doi:10.1249/MSS.0b013e31816459a1
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Dass S,
      2. Cochlin LE,
      3. Suttie JJ, et al
      . Exacerbation of cardiac energetic impairment during exercise in hypertrophic cardiomyopathy: a potential mechanism for diastolic dysfunction. Eur Heart J 2015;36:1547–54. doi:10.1093/eurheartj/ehv120
      OpenUrlAbstract/FREE Full Text
    31. ↵
      1. Arena R,
      2. Myers J,
      3. Abella J, et al
      . Development of a ventilatory classification system in patients with heart failure. Circulation 2007;115:2410–17. doi:10.1161/CIRCULATIONAHA.107.686576
      OpenUrlAbstract/FREE Full Text
    32. ↵
      1. Guazzi M,
      2. Adams V,
      3. Conraads V, et al
      . EACPR/AHA Scientific Statement. Clinical recommendations for cardiopulmonary exercise testing data assessment in specific patient populations. Circulation 2012;126:2261–74. doi:10.1161/CIR.0b013e31826fb946
      OpenUrlFREE Full Text
    33. ↵
      1. Briasoulis A,
      2. Mallikethi-Reddy S,
      3. Palla M, et al
      . Myocardial fibrosis on cardiac magnetic resonance and cardiac outcomes in hypertrophic cardiomyopathy: a meta-analysis. Heart 2015;101:1406–11. doi:10.1136/heartjnl-2015-307682
      OpenUrlAbstract/FREE Full Text
    34. ↵
      1. Lopes LR,
      2. Syrris P,
      3. Guttmann OP, et al
      . Novel genotype-phenotype associations demonstrated by high-throughput sequencing in patients with hypertrophic cardiomyopathy. Heart 2015;101:294–301. doi:10.1136/heartjnl-2014-306387
      OpenUrlAbstract/FREE Full Text

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Contributors All authors approved the version submitted to the journal and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. DM, GL, FR, CA, PA and MC contributed to the conception of the work, drafted it and/or contributed to the acquisition, analysis or interpretation of data. AM analysed the entire data set and revised it critically for statistical analysis.

    • Competing interests None declared.

    • Ethics approval The following institutions provided approval for this study, University La Sapienza-Rome; II University of Naples-Naples; San Camillo-Forlanini Hospital-Rome; Centro Cardiologico Monzino-IRCCS-Milan; University of Foggia-Foggia.

    • Provenance and peer review Not commissioned; externally peer reviewed.