Increased Incidence of De Novo Malignancies Compared With Malignancy Recurrences in Survivors of Cancer Undergoing Heart Transplantation

Abstract

Background

A proportion of cancer survivors (CS) with heart failure develop advanced heart failure. In select CS with advanced heart failure, heart transplantation (HT) remains the gold standard treatment. CS have been shown to have increased incidence of subsequent primary cancers (SPC). This study investigated the incidence and characteristics of SPC in CS undergoing HT at a high‐volume HT center.

Methods

A retrospective analysis of adult CS (age≥18) who underwent HT at Columbia University Irving Medical Center between January 1, 2000, and August 1, 2020, was conducted. Patients with nonmelanoma skin cancer were excluded. Kaplan–Meier analysis was used to measure cancer‐free survival.

Results

Of 1332 consecutive patients undergoing HT at our institution between January 1, 2000, and August 1, 2020, 104 patients were CS (7.8%). CS had improved survival compared with the non‐CS group, with a 5‐, 10‐, and 15‐year survival of 89%, 70%, and 60% in the CS group compared with 78%, 64%, and 41% in the non‐CS group. Twenty‐two (21.2%) developed SPC at a median of 4 years (interquartile range, 3–7.75) posttransplant. Eighteen of 22 SPC (82%) were de novo malignancies, and 4 were recurrences.

Conclusions

CS are 4 times more likely to have de novo malignancies unrelated to their primary malignancy rather than a recurrence. Despite the high rate of SPC, survival in this cohort was comparable to contemporary survival in HT patients. Further studies are required to understand optimal prevention and screening for SPC in CS undergoing HT beyond recommendations in the general population.

Nonstandard Abbreviations and Acronyms

CS

cancer survivors

HT

heart transplantation

SOT

solid organ transplantation

SPC

subsequent primary cancers

Clinical Perspective.

What Is New?

• Survivors of cancer undergoing heart transplantation are 4 times more likely to develop a de novo malignancy compared with a recurrence of their primary malignancy.

• Although survivors of cancer are at higher risk of subsequent primary malignancies, their overall survival appears to be superior to that of those without a history of cancer.

What Are the Clinical Implications?

• Survivors of cancer undergoing heart transplantation require a personalized risk assessment for heightened surveillance and screening as indicated for subsequent primary malignancies not related to their primary malignancy.

Advances in cancer therapeutics paired with an aging population have led to 18 million cancer survivors (CS) in the United States. ¹ CS have been found to have an increased risk of comorbidities compared with the general population, including cardiac comorbidities such as heart failure (HF). ² A proportion of CS with HF progress to requiring advanced therapies. ³ , ⁴ Heart transplantation (HT) remains the gold standard therapeutic intervention in suitable candidates with advanced HF.

New consensus statements, based largely on expert opinion, have been released to help guide patient selection for HT in CS and patients with advanced HF with active cancer. ⁵ , ⁶ HT in CS presents some unique concerns. Exposure to a growing armamentarium of cancer therapies, including chemotherapy, targeted therapies, radiation therapy, immunotherapy, as well as surgical therapy, has led to an increased burden of medical and psychosocial comorbidities that have the potential to increase perioperative and post‐HT complications. ⁷

CS are known to be at higher risk of developing subsequent primary cancers (SPC) as compared with the general population. ⁸ Recently, we published the outcomes of CS in HT recipients in the United Network for Organ Sharing database and demonstrated the increased incidence of SPC in this cohort. ⁹ However, more granular data regarding the specific types of SPC that develop in CS undergoing HT is lacking. The purpose of the current study was to describe the clinical characteristics and outcomes of HT in CS over a 20‐year period at our high‐volume transplant center with a focus on SPC.

METHODS

The data that support the findings of this study are available from the corresponding author upon reasonable request.

This study was a retrospective analysis of adults (age≥18) with a prior malignancy who received HT at Columbia University Irving Medical Center between January 1, 2000, and August 1, 2020. Patients with prior squamous or basal cell carcinoma of the skin were excluded.

Statistical Analysis

Categorical variables are presented as counts and percentages, with group comparisons performed using χ² tests. Continuous variables are reported as means±SDs and were compared using 2‐sample Student t tests for parametric variables and median and interquartile range (IQR) and comparison using the Wilcoxon rank test for nonparametric variables. Survival analyses were conducted using Kaplan–Meier methods, with comparisons assessed using the log‐rank test, and Cox proportional‐hazards regression models were performed to compare differences in outcomes between the groups. For the multivariable Cox regression analyses, only complete‐case data were included, including variables such as induction therapy. Statistical analysis was performed using R (R core team, 2019; version 4.2.3).

This study was performed under a protocol approved by the Columbia University Medical Center’s Institutional Review Board. Informed consent was waived.

RESULTS

Study Cohort

Of 1332 consecutive patients who underwent HT at our institution between January 1, 2000, and August 1, 2020, a total of 104 patients (7.8%) were identified as CS, including 2 with multiple malignancies. A total of 1196 patients (89.8%) comprised the non‐CS group, and 32 patients had incomplete data and were excluded from the final analytic cohort.

Baseline Characteristics

The median age at transplant for the entire cohort was 55 years (IQR, 45–62). CS were older at the time of transplant (58 [IQR, 49–65] years) compared with non‐CS (55 [IQR, 44–62] years; P=0.008). Overall, 978 patients (75.2%) were male. Within groups, 67 patients who were CS (64.4%) were male, compared with 911 male patients (70.1%) in the non‐CS cohort (P=0.01).

The percentage of HT recipients who are CS at our institution increased over the study period. In the decade from 2001 to 2010, 7.0% of HT recipients were CS, and in the decade from 2011 to 2020 that number increased to 9.9% (Table 1). Leukemia and lymphoma were the most common types of pretransplant malignancies (n=28, 26.7%) (Figure 1), followed by breast cancer (n=18, 17.1%) and prostate cancer (n=18, 17.1%). Other prominent pretransplant malignancies included colon (n=8, 7.6%), testicular (n=6, 5.7%), and sarcoma (n=6, 5.7%).

Table 1.

Baseline Characteristics for Survivors of Cancer

	N=104
Age, y	55.3 (11.6)
Age, y, n (%)
18–39	13 (12.5)
40–64	65 (62.5)
≥65	26 (25.0)
Male sex, n (%)	67 (64.4)
United Network for Organ Sharing status at time of transplant, n (%)
1A (old system)	54
1B (old system)	30
2–3 (old system)	6
1 (new system)	1
2 (new system)	6
3 (new system)	1
4 (new system)	6
Prior anthracycline exposure, n (%)	45 (43.2%)
Prior radiation therapy, n (%)	41 (39.4%)
Heart failure cause, n (%)
Ischemic	26 (25%)
Chemotherapy	32 (30.7%)
Dilated cardiomyopathy	31 (29.8%)
Restrictive cardiomyopathy	7 (6.7%)
Congenital	3 (2.8%)
Cardiac allograft vasculopathy	5 (4.8%)
Years since malignancy,* n (%)	(* 4 patients with missing data)
<2 y	18 (17.3%)
2–5 y	12 (11.5%)
>5 y	70 (67.3%)
Year of transplant, n (%)
2000–2005	19 (18.3)
2006–2010	28 (26.9)
2011–2015	29 (27.9)
2016–2020	28 (26.9)

Figure 1. Types of pretransplant malignancies in cancer survivors.

A, Pie chart of most common malignancies. B, Other cancers broken down by type.

The cause of HF in CS was nonischemic in 75 patients (72.1%) and ischemic in 29 patients (27.9%; Table 1). Forty‐five CS (43.3%) had received anthracyclines before transplant, and 34 patients (32.7%) received mediastinal radiation before transplant. The median time from remission to transplantation was 8 years (IQR, 3–16.5). The majority of the patients (70, 67.3%) had completed cancer therapy >5 years before HT. Twelve patients (11.5%) were transplanted 2 to 5 years after treatment, and 18 patients (17.3%) were transplanted within 2 years of treatment. Data regarding time since remission were missing for 4 (3.8%) patients.

Outcomes

Mean follow‐up of the total cohort was 8.4 (SD ±6.2) years. The patients who were CS had shorter follow‐up of 5.9 (SD ±4.9) years compared with the non‐CS patients with 8.6 (SD ±6.2 years; P<0.001). In the CS cohort, 22 (21.2%) developed SPC at a median of 4 years (IQR, 3–7.75 years) posttransplant. The incidence of SPC in CS was 13% (14/104) at 5 years, 20% (21/104) at 10 years, and 21.1% (22/104) at 15 years. In comparison, of the patients without a history of malignancy (non‐CS), 214 patients (17.9%) had SPC. SPC were detected in 9% (108/1196) at 5 years, 13.5% (136/1196) at 10 years, and 15.6% (186/1196) at 15 years in this cohort of patients. There was no difference in the rate of SPC between the different types of pretransplant malignancy (P=0.73). However, there were differences in the types of posttransplant malignancies.

Among the 28 patients with leukemia/lymphoma pretransplant, there were no cases of recurrence, but 6 patients developed a de novo malignancy posttransplant, for an SPC rate of 21.4%. Among the 18 patients with breast cancer pretransplant, 2 patients (11.1%) developed recurrence, and 3 (16.7%) developed de novo cancers for a total of 27.8% SPC. Among the 18 patients with prostate cancer pretransplant, only 1 (5.6%) had recurrence, and 2 patients (11.1%) developed de novo malignancy for an SPC rate of 16.7%. Compared with 1176 non‐CS patients with data available regarding cancer incidence who underwent HT, CS were at higher risk of SPC, as shown in Figure 2.

Figure 2. Comparison of new cancer diagnoses in cancer survivors and noncancer survivors after heart transplantation.

A, Freedom from cancer in cancer survivors and noncancer survivors. B, Incidence of cancer and death in cancer survivors and non‐cancer survivors after heart transplantation. HT indicates heart transplantation.

Twenty‐one patients in the CS cohort died. Five‐year survival was 89.3±3.2% (Figure 3). Among the causes of death, infection was the most common, occurring in 8 patients (38%), followed by cancer‐related complications in 28.5% (n=6) (Table 2). Specifically, death occurred due to breast cancer (n=2), colon cancer (n=2), lymphoma (n=1), and lung cancer (n=1). Other causes of death included primary graft failure (n=2), rejection (n=1), HF (n=1), renal failure (n=1), stroke (n=1), and abdominal hemoperitoneum (n=1). In the non‐CS cohort, 563 patients died over a mean follow‐up of 8.6 (SD ±6.2 years), with a 5‐year survival of 77.8%±1.2%. Despite the higher risk of SPC, CS showed overall improved survival compared with the non‐CS group (Figure 3A). There was no difference in event‐free survival in the groups (Figure 3B).

Figure 3. Survival after heart transplantation in cancer survivors and noncancer survivors after heart transplantation.

A, Overall survival and (B) event free survival. HT indicates heart transplantation.

Table 2.

Characteristics of Patients With Subsequent Primary Cancers

Patient	Age at transplant, y	Pretransplant malignancy	Recurrence vs new malignancy vs both	Type of posttransplant malignancy	Posttransplant malignancy, y	Cause of death	Follow‐up, y
1	35	Leukemia	New	Anal cancer	11	Anal cancer	17.7
2	41	Hodgkin’s lymphoma	New	Lung cancer	2		2.5
3	58	Breast cancer	New	Lung cancer	11		11.9
4	57	Gastrointestinal stromal tumor	New	Prostate cancer	3		7.0
5	63	Hodgkin’s lymphoma	New	Bladder cancer	9	C‐diff toxic megacolon	10.1
6	59	Prostate cancer	New	Lung cancer	2		4.4
7	56	Primary polycythemia	New	Breast cancer	9	Breast cancer	14.3
8	50	Waldenstrom’s macroglobulinemia	Both	Light chain amyloidosis	0		0.64
9	58	Prostate cancer	New	Lymphoma	7	Lymphoma	7.97
10	65	Colon cancer	New	Prostate	3		13.98
11	69	Colon cancer	New	Bladder	13		13.30
12	60	Hodgkin’s lymphoma	New	Leiomyosarcoma	1		3.51
13	67	Colon cancer	New	Thyroid cancer	3		10.94
14	64	Breast cancer	Both	Breast cancer, Thyroid cancer	5		10.02
15	66	Melanoma	New	Prostate cancer	4		9.46
16	56	Breast cancer	Recurrence	Breast cancer	5		9.23
17	64	Thyroid cancer	New	Renal cell cancer	4		8.00
18	67	Colon cancer	Recurrence	Colon cancer	8	Colon cancer	7.88
19	61	Diffuse large B‐cell lymphoma	New	Metastatic undifferentiated squamous cell	4		7.34
20	65	Breast cancer	New	Lung cancer	5	Lung cancer	5.53
21	66	Non‐Hodgkin’s lymphoma	New	Posttransplant lymphoproliferative disorder	3		5.67

Among the 22 CS who subsequently developed SPC, 17 of them developed new primary cancers, 3 experienced recurrence, and 2 patients had both recurrence and a new primary (Table 2). Subsequent primary cancers occurred at a median of 4 years posttransplant (IQR, 3–7.75 years). In this cohort, 7 patients died (31.8%), and the average survival posttransplant of CS with SPC was 10.4 years. Of the 7 posttransplant malignancy deaths, 5 died of cancer‐related complications, 1 of infection, and 1 of hypoxic respiratory failure.

In the non‐CS cohort, complete‐case data, including induction therapy were available for 637 patients, whereas all 104 patients in the CS cohort had complete data. A multivariable Cox proportional hazards analysis including demographic, clinical, and treatment‐related covariates was performed for both the patients who were CS and non‐CS. In the CS cohort, a history of induction was found to be significant (hazard ratio, 4.31 [95% CI, 2.58–10.85]; P<0.001) compared with the non‐CS cohort (Table S1). The other covariates, such as age, sex, race, and history of cigarette use, were nonsignificant (Table S1).

DISCUSSION

In this single‐center study of >100 CS who underwent HT over 20 years, we found (1) CS represent about 8% of all HT recipients; (2) about 1 in 5 CS developed SPC (21%) at a median of 4 years; (3) de novo malignancies were 4 times more common than recurrences of known pretransplant malignancies; and (4) the overall survival of CS undergoing HT was 89% at 5 years, which was greater than the non‐CS HT population (78%).

The number of CS continues to increase worldwide with more than 18 million in the United States alone. ¹ It is thought that this is a consequence of an aging population paired with remarkable improvement in early detection and cancer therapeutics. This population, however, has been shown to be at higher risk of developing cardiovascular disease and HF in particular compared with the general population. ¹⁰ A proportion of these CS who develop advanced HF, and the population of CS with advanced HF, continues to rise.

HT remains the gold standard for treatment of suitable candidates, with advanced HF with a contemporary median survival of 12.5 years and a conditional survival of 14.8 years for those who survive the first year. ¹¹ As donor hearts are a limited resource, it is the responsibility of heart transplantation centers to select candidates for HT anticipated to have favorable outcomes after HT. The International Society for Heart and Lung Transplantation listing criteria offers guidance to heart transplantation centers with regard to comorbidities in patients with advanced HF undergoing evaluation for HT. ¹² One of these comorbidities is cancer. Current International Society for Heart and Lung Transplantation guidelines state that in “patients needing cardiac transplantation, collaboration with oncology specialists should occur to stratify each patient as to their risk of tumor recurrence. Cardiac transplantation should be considered when tumor recurrence is low based on tumor type, response to therapy, and negative metastatic workup.”

Following HT, cancer represents the leading cause of death in recipients after 5 years. ¹¹ Solid organ transplantation (SOT) recipients have been found to be at increased risk of developing certain malignancies and dying from these malignancies, with risk varying depending on the cancer type. ¹³ , ¹⁴ The primary cause of this increased risk is thought to be due to compromised immune‐surveillance due to use of requisite immunosuppressive therapy. Other factors such as shared risk factors for HF and cancer, as well as the possible procarcinogenic effects of immunosuppressive medications like cyclosporine as noted in preclinical models unrelated to immunosuppression, may also play a role. ¹⁵ , ¹⁶

CS at baseline have been found to be at increased risk for the development of several SPC as well as increased mortality once these SPC develop. ¹⁴ This increased risk is thought to be mediated by several factors including genetic predisposition, cancer treatment exposures, as well as lifestyle and host factors that are common to primary and SPC. CS undergoing HT are exposed to the “double hit” of post‐HT immunosuppression combined with their innate higher risk of several SPC. Previous recommendations used empiric cutoffs of 2 or 5 years wait time in CS postcancer treatment before SOT, on the basis of findings from the Israel Penn Registry. ¹⁷ Recently, newer consensus statements have been released to guide candidate selection for SOT in patients with several previous solid organ malignancies as well as hematological malignancies and melanoma. ⁵ , ⁶ Most of these recommendations were based solely on expert opinion, as there are still limited data regarding outcomes post‐SOT in CS, and specifically post‐HT in CS.

In the current study, we reviewed HT outcomes at our center in adult CS over 20 years. CS comprised 7.8% of all HT recipients, with increasing numbers over the past 2 decades. This is consistent with our recent analysis of the United Network for Organ Sharing database, which showed a doubling of the number of CS undergoing HT to 8.2% in 2020. ⁹ This increase appears consistent with improvements in cancer therapies as well as newer consensus statements that have liberalized the approach to offering HT to CS. Previous consensus statements before 2000 had listed “recent malignancy” as a contraindication to HT. ¹⁸ As there are currently >18 million CS in the United States with a projected increase to >22 million in 2030, the number of CS undergoing HT is likely to continue to increase further.

CS with a previous history of lymphoma, leukemia, breast cancer, or prostate cancer formed >60% of the cohort. This is consistent with a previous United Network for Organ Sharing analysis showing that CS with a previous history of lymphoma, breast, and prostate cancer included 46% of all CS. ¹⁹ This distribution represents the heterogeneity of CS with advanced HF undergoing HT. Among survivors of breast cancer, 94.4% had a history of anthracycline exposure, and 72.2% had a history of chest radiation. Similarly, among survivors of lymphoma/leukemia, 77.8% had a history of anthracycline exposure, and 63.0% had a history of chest radiation. This may explain the high progression to HF in these cohorts of CS. Breast cancer and prostate cancer represent the most common noncutaneous cancers in women and men, respectively, which may also explain the higher proportion of these CS. ¹ CS underwent HT at a median of 8 years after known remission from their primary cancer. However, about 17% of the patients were <2 years after completing cancer therapy. Considering reduction in wait time recommendations for HT in the most common primary cancer subtypes (2 years remission for lymphoma, 2 years for stage 2 breast cancer, no wait time for low‐risk patients with prostate cancer), the majority of patients in this study appear to be well within current recommendations of wait time before HT and thus potentially may represent a cohort with a lower risk of SPC due to cancer recurrence compared with current consensus statements.

About 1 in 5 CS developed SPC (21.1%) at a median of 4 years after HT. The incidence of SPC was 13% at 5 years, 20% at 10 years, and 21.1% at 15 years. In comparison, of the 1196 patients without a history of malignancy, SPC were detected in 9% at 5 years, 13.5% at 10 years, and 15.6% at 15 years. It should be noted that this does not include nonmelanoma skin cancers such as squamous cell carcinoma or basal cell carcinoma. Although rates of malignancy have been reported as low as 2.9% in the general HT population at 1‐year post‐HT, prevalence has been shown to increase with time since transplant. A study by Youn et al of >17 000 HT patients reported an incidence of malignancy of 10.6% between years 1 and 5 post‐HT, which was associated with worse survival compared with those HT patients who did not develop a malignancy. ²⁰ Our study in comparison showed a small increase in malignancies in CS compared with this large study. Similarly, a study by Van Keer showed a cumulative incidence of malignancies of 14% in the general HT population at 5 years, slightly higher than the incidence in our CS cohort. ²¹

Among CS who developed SPC, the rates of de novo cancers were significantly higher compared with cancer recurrences. De novo SPC were 4 times more likely than a cancer recurrence (Graphical Abstract). This highlights the shortcomings of current screening practices for malignancies in CS after HT. There are limited data for cancer screening in SOT recipients and specifically the general HT population, with most previous consensus statements in agreement with guidelines for the general population other than annual screening for skin cancer. ²² , ²³ A recent consensus statement describes more frequent screening for recurrent malignancies in CS. ²⁴ Considering that CS are 4 times more likely to develop a de novo malignancy rather than a recurrence, there is likely a need to intensify screening for secondary malignancies unrelated to the primary malignancy of the CS. It should be noted that there are certain groups of CS in which there are recommendations for more intensive cancer screening, for example screening for colorectal cancer every 5 years with colonoscopy after age 35 in CS who previously received radiation to the abdomen, pelvis, spine, or total body irradiation. ²⁵ CS undergoing HT represent a high‐risk group who will likely benefit from the rapid developments in personalized risk assessment for SPC using advances in oncology like circulating tumor DNA, artificial intelligence, and biosensor technology. ²⁶

In spite of the high rate of SPC in CS, the 5‐year survival of this cohort was 89%. This survival was higher than the survival found in the non‐CS population (78%) (Figure 2) and appears comparable to, if not better than, the International Society for Heart and Lung Transplantation contemporary 5‐year survival of 75% at 5 years. ¹¹ This survival is consistent with our United Network for Organ Sharing analysis, which showed similar survival for CS compared with the general population except for patients with a history of hematologic malignancies. ⁹ It is also similar to previous studies that reviewed survival in CS undergoing HT. ²⁷

How does one reconcile the higher incidence of SPC paired with improved survival in the CS population compared with the general HT population? It is unclear at this stage, but a few factors may be at play. Several types of SPC (for example, breast cancer and prostate cancer) in our cohort have not been noted to be associated with increased mortality in HT patients. In addition, patients who have undergone HT have a high frequency of interactions with the health care system, which may increase early detection and treatment of malignancies. ²⁸ Finally over time there have been marked improvements in all modalities of cancer therapeutics, including traditional chemotherapy, radiation oncology, oncological surgery, targeted therapies, and immunotherapies. ²⁸ , ²⁹

Considering that the survival of CS appears similar to, if not better than, the general HT population with the data available, HT appears to be a reasonable therapy in selected CS with advanced HF. However, to optimize outcomes in this population, as shown in this study, it is vital for providers to be aware of the increased risk of SPC in this population, the majority of which are unrelated to the CS primary malignancy. It is essential for the transplant community to harness the latest developments in personalized risk assessment for cancer, and it is incumbent on us to advocate for this cohort of patients to be part of clinical trials to best understand optimal screening and modulation of immunosuppression. In our program we conduct an individualized risk assessment for SPC in all CS undergoing HT with the goal of active surveillance and reduction in immunosuppression balancing the risk of rejection. It should be noted that for certain CS with the highest risk of SPC or certain active cancers and concurrent advanced HF, mechanical circulatory support with a left ventricular assist device may be an option, which obviates the need for immunosuppression.

LIMITATIONS

Our study has certain limitations. It is a single‐center retrospective study of CS undergoing HT. Granular data regarding immunosuppression changes in patients transplanted from earlier in the study period were not available. However, as compared with multicenter registry data, these CS undergoing HT were exposed to similar institutional protocols with regard to immunosuppression and testing. Our institutional protocol consisted of triple therapy with cyclosporine, mycophenolate mofetil, and prednisone, with tacrolimus replacing cyclosporine in 2006. Considering that cancer becomes the leading cause of death in HT recipients after 5 years, longer follow‐up within this cohort will provide further insight into long‐term outcomes.

CONCLUSIONS

In our institutional experience, de novo malignancies were about 4 times more common than recurrent primary malignancies in CS undergoing HT. Further study is required to understand long‐term outcomes and optimal cancer screening and prevention strategies in this unique population of patients. In spite of the higher rate of SPC, the survival post‐HT was superior to non‐CS, and HT remains a viable option in carefully selected CS.

Sources of Funding

The authors report no funding source specific to this study.

Disclosures

N. Uriel‐Abbott (grant support for institution), Abiomed (grant support for institution), Livemetric (medical advisory board), Leviticus (medical advisory board), Doraya (medical advisory board).

Supporting information

Table S1