Sixty-eight patients with ACHD and malignancy met inclusion criteria, out of a total of 6,963 patients seen at our center during the study period, thus yielding a cancer prevalence of 0.98% or 980 per 100,000 individuals. Baseline demographics are detailed in table 1. The median age of the cohort at the time of cancer diagnosis was 43.5 years (IQR 3-11), while age at most recent follow-up was 53.5 years (IQR 40-66). Four (6%) patients were under the age of 18 years at the time of cancer diagnosis. 59% of the cohort was female, while Caucasian (62%) and Hispanic (18%) races were the most represented. Hypertension and chronic liver disease was noted in 27% and 24% of patients, respectively, with Fontan-associated liver disease, found in 11 patients, being the most common liver pathology. Median follow-up after cancer diagnosis was 5 years (IQR 3-11, range 0.2-51).
Table 1: Key baseline cardiovascular data. ACE = angiotensin converting enzyme, ACHD = adult congenital heart disease, IQR = interquartile range, NYHA = New York Heart Association
Characteristic
|
Frequency (n=68)
|
Median age at most recent follow-up, yr (IQR)
|
53.5 (40-66)
|
Median age at first cancer diagnosis, yr (IQR)
|
43.5 (31-60)
|
Median follow-up since cancer diagnosis, yr (IQR)
|
5 (3-11)
|
Female sex (%)
|
40 (59)
|
ACHD Anatomic Complexity (%)
|
|
Class I
|
12 (18)
|
Class II
|
32 (47)
|
Class III
|
24 (35)
|
Single ventricle circulation (%)
|
15 (22)
|
Fontan palliation (%)
|
12 (18)
|
Systemic right ventricle in biventricular circulation (%)
|
4 (6)
|
Baseline cyanosis (%)
|
6 (9)
|
Prior cardiac surgery (%)
|
53 (78)
|
Arrhythmia prior to cancer diagnosis (%)
|
30 (44)
|
Supraventricular tachycardia
|
19 (28)
|
Atrial fibrillation
|
12 (18)
|
Ventricular arrhythmia
|
1 (2)
|
Sinus node dysfunction
|
2 (3)
|
Atrioventricular nodal block
|
3 (4)
|
Heart failure prior to cancer diagnosis (%)
|
17 (25)
|
NYHA Functional Class at baseline (%)
|
|
Class I
|
53 (78)
|
Class II
|
13 (19)
|
Class III
|
2 (3)
|
Medications prior to cancer diagnosis, n=66 (%)
|
|
ACE-inhibitor/ angiotensin receptor blocker
|
16 (24)
|
Beta-blocker
|
24 (36)
|
Diuretic
|
16 (24)
|
Clinical Comorbidities
|
|
Diabetes Mellitus
|
5 (7)
|
Hypertension
|
18 (27)
|
Chronic kidney disease
|
7 (10)
|
Coronary artery disease (> 50% luminal stenosis)
|
6 (9)
|
Smoking History
|
13 (19)
|
Chronic liver disease
|
16 (24)
|
Baseline Cardiovascular Data
Patients predominantly had moderate or great anatomic complexity of CHD, with 12 (18%) having class I (simple) complexity and 24 (35%) having class III (great) complexity (Central Illustration). Fifteen (22%) patients had single ventricle circulation and 12 (18%) had undergone a Fontan palliation. Prior to cancer diagnosis, 53 (78%) patients had cardiac surgery and 35 (52%) patients had fluoroscopic procedures. History of arrhythmia at the time of cancer diagnosis was present in 30 (44%) patients, with supraventricular tachycardia (28%) and atrial fibrillation (18%) being the most common arrhythmias. Median baseline BNP level was 68 pg/mL (IQR 37-174), while baseline NYHA functional class of I, II and III were noted in 78%, 19% and 3% patients respectively (Table 1).
Baseline echocardiographic data were available in 59 patients (Supplemental Table 1). Systemic and subpulmonary ventricular dysfunction were noted in 14% and 15% respectively. Diastolic function was not consistently reported, however, one patient had severe biventricular restrictive cardiomyopathy – all others marked as having ventricular dysfunction had systolic dysfunction. Moderate or greater valvular stenosis/ regurgitation was noted in 53% of echocardiograms, with 23% demonstrating severe valvular lesions. No MRI was performed within three months of cancer diagnosis.
Baseline Cancer-related Data
Eighty-two cancers were diagnosed in 68 patients, with 12 (18%) patients having been diagnosed with multiple cancers (Table 2). Hematologic malignancy was the most common primary cancer accounting for 24% of all cancers, followed by breast (21%) and skin (12%) cancer (Figure 1). Hepatocellular carcinoma (HCC) was the fourth most common cancer (9%) along with urologic malignancies; 5 out of the 7 patients with HCC had undergone a prior Fontan palliation.
Only 16% of the cohort had undergone any prior cancer screening, and cancer was diagnosed by screening in only 10% of all cases. 38% patients had stage I disease while 22% had stage IV disease at the time of diagnosis. Known genetic mutations prior to cancer diagnosis were present in 7 patients, including 4 with trisomy 21, of whom 3 developed leukemia.
Multiple cancers were diagnosed in 12 (18%) patients; lymphoma was the most common primary malignancy while the most common subsequent malignancies were of breast and skin origin. Two patients were deemed to have had secondary cancers due to treatment administered for their primary cancer (Supplemental Table 2).
Table 2: Cancer-related baseline characteristics and treatment data. VEGF = vascular endothelial growth factor, EGFR = epidermal growth factor receptor
Characteristic
|
Frequency, n = 68 (%)
|
Stage at initial cancer diagnosis
|
|
I
|
23 (38)
|
II
|
14 (23)
|
III
|
10 (17)
|
IV
|
13 (22)
|
Multiple cancer diagnoses
|
12 (18)
|
Genetic abnormalities
|
|
Trisomy 21
|
4 (6)
|
Noonan syndrome
|
1 (1)
|
BRCA mutation
|
2 (3)
|
JAK2 mutation
|
1 (1)
|
Any cancer treatment
|
64 (94)
|
Surgical resection alone
|
14 (21)
|
Radiation therapy
|
26 (38)
|
Mediastinal radiation
|
15 (22)
|
Systemic pharmacotherapy
|
41 (60)
|
Radiofrequency ablation
|
2 (3)
|
Taxane/Vinca Alkaloids
|
16 (24)
|
Platinum based
|
12 (18)
|
Anthracycline
|
11 (16)
|
Immune checkpoint inhibitors
|
7 (10)
|
HER2 inhibitor
|
5 (7)
|
Tyrosine kinase inhibitor
|
5 (7)
|
VEGF inhibitor
|
5 (7)
|
EGFR inhibitor
|
2 (3)
|
BRAF/MEK inhibitor
|
2 (3)
|
Mortality
Sixteen patients died over the course of follow-up, conferring a 24% all-cause mortality. Eleven deaths (69%) occurred due to non-cardiac causes, all of which were cancer-related (Table 3). Cardiovascular death occurred in 3 patients while 2 patients died of indeterminate causes. Details regarding mortality of specific patients can be found in the Supplemental Appendix.
Table 3: Outcome measures of the cohort (n = 68 unless otherwise specified). MACCE = Major Adverse Cardiovascular and Cerebrovascular Event, BNP = B-type Natriuretic Peptide, NYHA = New York Heart Association
Outcome measure
|
Frequency, n = 68 (%)
|
Mortality
|
16 (24)
|
Cardiac
|
3/16 (19)
|
Non-Cardiac
|
11/16 (69)
|
Indeterminate
|
2/16 (12)
|
MACCE*
|
40 (59)
|
Arrhythmia
|
29 (43)
|
Decompensated heart failure
|
22 (32)
|
Thromboembolism (arterial or venous)
|
7 (10)
|
Stroke
|
3 (4)
|
Hospitalization due to MACCE
|
34 (50)
|
Worsening systemic ventricular function
|
9/61 (15)
|
Worsening subpulmonary ventricular function
|
8/53 (15)
|
Worsening valvular function
|
25/62 (40)
|
Worsening BNP
|
28/33 (85)
|
Worsening NYHA Functional Class
|
19 (28)
|
Cardiac transplantation†
|
3 (4)
|
Cancer Outcome (index malignancy)
|
|
Progression
|
17 (25)
|
Relapse
|
3 (4)
|
Partial Remission
|
26 (38)
|
Complete Remission
|
12 (18)
|
Stable and undergoing treatment
|
10 (15)
|
*More than one MACCE event may have occurred in the same patient
†One patient underwent combined heart and liver transplantation
Survival probability at 1, 5, and 10 years was 93%, 82% and 77% respectively. Univariate predictors of mortality were incrementally advanced cancer stage at diagnosis (OR 2.37, 95% CI 1.32 – 4.25, p = 0.004), history of arrhythmia prior to cancer diagnosis (OR 3.82, 95% CI 1.15 - 12.67, p = 0.028) and diuretic use at baseline (OR 3.54, 95% CI 1.04-12.04, p = 0.044) (table 4). Immune checkpoint inhibitor (ICI) therapy use also significantly increased odds of mortality (OR 11.36, 95% CI 1.95 - 66.38, p = 0.004), however all patients with ICI use had cancer stage of stage II or greater at the time of administration and ICI use correlated with metastatic disease at the time of diagnosis (r = 0.34). Survival analysis showed significantly reduced survival in the subgroups with advanced cancer i.e. stage IV compared to stages I-III (p = 0.005), history of arrhythmia (p = 0.003) and baseline diuretic use (p = 0.009), while there was no significant difference in survival stratified by ACHD anatomic complexity (p = 0.14) (Figure 2).
Table 4: Univariate predictors of mortality and MACCE; rows in bold denote significance (p < 0.05). OR = odds ratio, CI = confidence interval. ACE-I = Angiotensin converting enzyme inhibitor, ARB = angiotensin receptor blocker, MACCE = Major Adverse Cardiovascular and Cerebrovascular Event, NYHA = New York Heart Association
Predictor variable
|
Mortality (n=16)
|
MACCE (n=40)
|
OR (95% CI)
|
p
|
OR (95% CI)
|
p
|
Male sex
|
0.87 (0.28 - 2.70)
|
0.811
|
1.03 (0.38 - 2.79)
|
0.952
|
ACHD complexity
|
1.22 (0.54 - 2.72)
|
0.632
|
1.52 (0.75 - 3.09)
|
0.246
|
Single ventricle circulation
|
1.24 (0.33 - 4.62)
|
0.748
|
6.02 (1.23 - 29.37)
|
0.011
|
Prior cardiac surgery
|
5.53 (0.67 - 45.81)
|
0.054
|
1.985 (0.62 - 6.34)
|
0.246
|
Number of fluoroscopic procedures
|
1.02 (0.74 - 1.40)
|
0.928
|
1.46 (1.04 - 2.06)
|
0.018
|
History of arrhythmia
|
3.82 (1.15 - 12.67)
|
0.028
|
2.90 (1.02 - 8.17)
|
0.039
|
Systemic ventricular dysfunction at baseline
|
2.5 (0.49 - 12.76)
|
0.271
|
13.33 (0.73 - 247.15)*
|
0.081
|
Subpulmonary ventricular dysfunction at baseline
|
0.56 (0.06 - 5.37)
|
0.601
|
6.63 (0.73 - 60.22)
|
0.093
|
Valvular dysfunction (> mild)
|
1.25 (0.38 - 4.11)
|
0.713
|
1.21 (0.42 - 3.47)
|
0.724
|
NYHA class at baseline
|
1.81 (0.65 - 5.09)
|
0.259
|
2.25 (0.71 - 7.09)
|
0.168
|
ACE-I/ARB use at baseline
|
1.06 (0.29 - 3.89)
|
0.935
|
1.15 (0.360 - 3.67)
|
0.814
|
Diuretic use at baseline
|
3.54 (1.04 - 12.04)
|
0.044
|
15.62 (1.91 - 127.64)
|
0.001
|
Beta blocker use at baseline
|
1.07 (0.33 - 3.42)
|
0.914
|
2.87 (0.94 - 8.66)
|
0.055
|
Multiple Cancer diagnoses
|
0.68 (0.13 - 3.54)
|
0.640
|
2.0 (0.48 - 8.35)
|
0.326
|
Cancer stage at diagnosis
|
2.37 (1.32 - 4.25)
|
0.004
|
0.80 (0.51 - 1.25)
|
0.325
|
Any systemic cancer therapy use
|
2.38 (0.68 - 8.36)
|
0.160
|
1.03 (0.38 - 2.79)
|
0.952
|
Platinum based compounds
|
0.59 (0.11 - 3.00)
|
0.521
|
0.89 (0.25 - 3.18)
|
0.859
|
Kinase inhibitor
|
2.44 (0.67 - 8.95)
|
0.178
|
1.05. (0.30 - 3.65)
|
0.939
|
Immune Checkpoint Inhibitor use
|
11.36 (1.95 - 66.38)
|
0.004
|
1.39 (0.24 - 8.17)
|
0.713
|
Anthracycline use
|
1.27 (0.29 - 5.49)
|
0.753
|
0.63 (0.16 - 2.42)
|
0.501
|
Mediastinal radiation therapy
|
0.77 (0.19 - 3.16)
|
0.711
|
1.47 (0.44 - 4.90)
|
0.529
|
*Firth logistic regression used due to complete separation
MACCE
Over the follow-up period, MACCE occurred in 40 (60%) patients and prompted hospitalization in 34 (50%). The events were driven mostly by arrhythmias, which occurred in 29 (43%) patients, and decompensated heart failure which occurred in 22 (32%) patients (table 4). Thromboembolic events occurred in 7 (10%) patients. The most common arrhythmias were atrial fibrillation and supraventricular tachycardia seen in 23% and 19% of the patients respectively. Twenty of the 29 patients with arrhythmia events did not have a baseline history of arrhythmia. Of the 22 patients who developed decompensated heart failure, 9 did not have a history of heart failure prior to cancer diagnosis. Heart failure hospitalization occurred in 17 (25%) patients. Two patients underwent heart transplantation, while a third underwent combined heart and liver transplantation and died post-operatively.
Freedom from MACCE at 1, 5 and 10 years after cancer diagnosis was 77%, 53% and 33% respectively (Central Illustration). Univariate predictors of MACCE were baseline diuretic use (OR 15.62, 95% CI 1.91 - 127.64, p = 0.001), single ventricle circulation (OR 6.02, 95% CI 1.23 - 29.37, p = 0.011), history of arrhythmia (OR 2.90, 95% CI 1.03 - 8.17, p = 0.039) and number of fluoroscopic procedures (OR 1.46, 95% CI 1.04 - 2.06, p = 0.018). (Table 4). Number of fluoroscopic procedures correlated with single ventricle circulation (r = 0.53) and baseline arrhythmia (r = 0.50). In multivariable analysis, baseline diuretic use was the only independent predictor of MACCE (OR 9.91, 95% CI 1.12 - 87.85, p = 0.039) (Table 5). A significantly reduced freedom from MACCE was seen in subgroups with single ventricle anatomy (p = 0.004), history of arrhythmia prior to cancer diagnosis (p < 0.001) and baseline diuretic use (p < 0.001). Ten-year freedom from MACCE was lower with class II and class III anatomic complexity, however this was not statistically significant (p = 0.09) (Figure 3).
Table 5: Multivariate predictors of MACCE. MACCE = Major Adverse Cardiovascular and Cerebrovascular Events
Variable
|
Odds Ratio
|
95% Confidence interval
|
Standard Error
|
p
|
Upper limit
|
Lower limit
|
History of arrhythmia
|
1.54
|
0.48
|
4.96
|
0.92
|
0.466
|
Baseline diuretic use
|
9.91
|
1.12
|
87.85
|
11.03
|
0.039
|
Single ventricle anatomy
|
2.73
|
0.48
|
15.51
|
2.42
|
0.258
|
Echocardiography
At most recent follow-up, systemic ventricular dysfunction was noted in 10/67 (15%) patients, while 11/57 (19%) patients with biventricular circulation had subpulmonary ventricular dysfunction. When compared to baseline, systemic ventricular dysfunction was worse in 9/10 patients and subpulmonary ventricular dysfunction worse in 8/8 patients with baseline data. Moderate or greater valve dysfunction was present in 38/67 (57%) patients at most recent follow-up (Table 3).
Cancer Therapeutics and Outcomes
Details of cancer therapy are presented in Table 2. Malignancy was treated in 64 patients; the remaining 4 patients either were managed conservatively or died before receiving treatment. Fourteen patients with locally advanced malignancy received surgical resection as their sole cancer treatment, with all achieving complete remission without additional treatment (Table 3). Systemic pharmacologic therapy was administered to 41 (60%) patients, with alkylating and alkylating-like agents being the most common (42% patients), followed by antimetabolites (30%) and taxanes/ vinca alkaloids (30%). Median cumulative dose of anthracyclines (doxorubicin equivalent) in 8 patients with available data was 179 mg/m2. Mediastinal radiation was administered to 15 (22%) patients. Stem cell transplantation was performed in 5 (7%) patients. No patient received chimeric antigen T-cell therapy.
Over the follow-up period, 17 (25%) patients had cancer progression, 26 (28%) patients achieved partial remission and 12 (18%) achieved complete remission. Cancer relapse occurred in 3 (4%) patients while 10 (15%) patients had stable disease with either ongoing or maintenance therapy at the time of most recent follow-up (Table 3). When stratified by primary malignancy, case fatality was highest in patients with lung cancer, with 2 deaths in the 3 diagnosed patients (Figure 4). HCC had the second highest fatality rate of 57%, while breast cancer carried the least case fatality of all fatal malignancies, with 2/17 (12%) deaths.
Cardiotoxicity
A total of 7 therapy-related MACCE occurred in 6 patients, i.e. 14% of patients on systemic cancer therapy, and 9% of the entire cohort. Out of the 7 events, 5 were decompensated heart failure, with one patient experiencing 2 episodes in response to different therapeutic agents. Ventricular function deteriorated in only 1 of these 5 events, in a patient with possible ICI-associated myocarditis. Three of these events occurred in patients with pre-existing cardiomyopathy and/or valvulopathy. In response to the MACCE event, cancer therapy had to be modified in 2 patients and completely withheld in one patient for two weeks until treatment of the heart failure episode. In most patients, therefore, continuation of the current therapy was made possible. Additional details of the presentation and management of therapy-associated MACCE are found in Table 6 and the supplementary appendix.
Of the 42 patients who received systemic cancer therapy, 2 patients were on cardioprotective medications at baseline while 3 others were pre-emptively started on a cardioprotective medication regimen (angiotensin-converting enzyme inhibitor and beta blocker) before chemotherapy. None of these patients experienced therapy-related MACCE. Additionally, none of the patients receiving anthracyclines, and one of the patients on anti-HER2 therapy had therapy-related MACCE.
Table 6: Cancer therapy-related MACCE events in the cohort, adjudicated as cardiotoxicity events. ASD = atrial septal defect, VSD = ventricular septal defect, SVT = supraventricular tachycardia, AF = atrial fibrillation, LAD = left anterior descending, LV = left ventricle, EF = ejection fraction
Patient ID
|
ACHD diagnoses
|
MACCE prior to cancer therapy
|
Primary cancer
|
Cancer Therapy
|
Therapy-related MACCE
|
Time of therapy-related MACCE episode from initiation of therapy
|
Management of suspected cardiotoxicity
|
Outcome of cardiotoxic episode
|
Cancer-related outcome
|
3
|
• ASD
• VSD
• Subaortic membrane
|
None
|
Melanoma
|
Ipilimumab/ nivolumab
|
New non-ischemic cardiomyopathy with heart failure. LVEF decreased from 53% to 36% after first cycle of immunotherapy
|
31 days
|
• Ipilimumab discontinued, switched to trametinib.
• Furosemide, carvedilol, spironolactone and lisinopril started
|
• Improvement in LV EF and volume status.
• Able to tolerate additional cycles of therapy
|
• New brain metastases discovered
• Suffered large intracranial hemorrhage due to gamma-knife treatment
• Died shortly after comfort care initiated
|
11
|
• Bicuspid aortic valve
• Mitral valve prolapse
|
• SVT, paroxysmal AF
• Known LV cardiomyopathy (LVEF 35%) due to ventricular pacing/ mid-LAD 70% plaque
|
Mantle cell lymphoma
|
Bendamustine/ rituximab (BR)
|
Decompensated heart failure, with 1.2L administration of fluids with first cycle
|
Same day as initiation
|
Prophylactic furosemide before each subsequent cycle
|
• Fluid status well-managed, no further decompensation or change in LV function.
• Completed 4 cycles of BR.
|
Remission followed by relapse
|
11
|
• Bicuspid aortic valve
• Mitral valve prolapse
|
Same as above
|
Mantle cell lymphoma
|
Ibrutinib
|
Decompensated heart failure. No changes to LV EF or valve function
|
30 days
|
• Discontinuation of ibrutinib,
• Initiation of bendamustine/ rituximab (BR) with scheduled hospitalization for management of fluid status with each dose
|
• Fluid status improved.
• Subsequent doses of BR given with intravenous furosemide during scheduled admissions
• Tolerated BR therapy well
|
• Progression to blast crisis
• Died shortly after comfort care initiated
|
21
|
Bicuspid aortic valve
|
None
|
Hodgkin's lymphoma
|
Mediastinal radiation
|
• Radiation pericarditis
• Radiation induced severe valvulitis of the aortic, tricuspid and mitral valves
|
Unknown
|
• Pericardial window
• surgical tricuspid valve replacement.
|
• Worsening heart failure due to severe mitral and aortic regurgitation.
• Deemed unintervenable due to poor functional status from metastatic lung cancer
|
• Hodgkin's lymphoma with remission
• Metastatic small-cell lung cancer diagnosed approximately 35 years later, progressed despite immunotherapy.
• Died shortly after being placed on palliative care for advanced cancer and heart failure
|
41
|
• Double outlet right ventricle (unintervened)
• Pulmonic stenosis
• VSD
|
Paroxysmal AF
|
Urothelial carcinoma
|
Gemcitabine/ cisplatin
|
• Acute chest pain with elevated troponin-I level to 0.86 ng/mL in setting of gastrocnemius vein thrombosis.
• No pulmonary embolism found
|
28 days
|
Conservative management advised since symptoms resolved.
|
• None, no recurrence in symptoms and hence no additional ischemia evaluation pursued
• Received another 2 cycles of gemcitabine/ cisplatin
|
• Completed 3 cycles of neoadjuvant gemcitabine/ cisplatin
• Underwent robotic nephroureterectomy and cystectomy. Currently under surveillance
|
44
|
Bicuspid aortic valve
|
• None
• Known severe aortic regurgitation with LV dilation and preserved LVEF
|
B-cell acute lymphoblastic leukemia
|
• Stem cell transplantation, thiotepa and fludarabine.
• Previously received ABFM induction including 18.75 mg/m2 of anthracycline + 2 cycles of blinatumomab.
|
Acute pulmonary edema resulting in hypoxic respiratory failure
|
10 days after stem cell transplantation, 6 months after induction
|
Aggressive high-dose intravenous diuretic therapy instituted
|
• Normalization of volume status
• Maintenance oral diuretic therapy instituted
• Continued on losartan and carvedilol
|
• Relapsed 6 months after stem cell transplantation
• Currently back on blinatumomab therapy
|
58
|
Pulmonary valve stenosis
|
• None.
• Known severe pulmonary regurgitation with normal right ventricular systolic function
|
Concomitant invasive ductal and lobular carcinoma in the same breast
|
TCH (taxotere/ carboplatin, herceptin)
|
• Progressive right-sided heart failure with each cycle, worst after completion of 6th and final cycle of TCH.
• No changes to ventricular systolic function on echocardiogram
|
21 days after 1st cycle
|
Oral diuretic therapy
|
• Normalization of volume status and resolution of heart failure
• Maintenance diuretic therapy instituted
|
• Underwent prophylactic right breast mastectomy.
• Currently in remission
|