Adult thymectomy is associated with increased mortality risk from cancer but not cardiovascular disease Open Access

TO THE EDITOR:

The role of the thymus in adult health is unclear, and the thymus is often removed during cardiothoracic surgery because it has been considered vestigial in adults.¹ We recently reported that adult thymectomy associates with increased mortality, cancer, and autoimmunity risk in conjunction with reduced new T-cell production and increased inflammatory cytokines compared with control patients who underwent cardiothoracic surgery without thymectomy.² 

The impact of losing new T-cell production on cardiovascular disease (CVD) is unclear.³ T cells can have beneficial and detrimental roles in CVD based on experimental mouse models and observational human studies.⁴ Single-cell sequencing analyses have shown that T cells contribute to both atherogenesis and atheroprotection.^5-10 Regulatory T cells are activated by hypercholesterolemia¹¹ and dampen atherogenic inflammation.¹² Although there is controversy on the precise roles of different CD4⁺ helper T-cell (Th) subtypes, it appears some can abrogate the inflammatory response (Th2 phenotypes)¹³ whereas others can encourage it (Th1 phenotypes).¹⁴ CD8⁺ lymphocytes appear to have multifaceted roles in atherogenesis^15-17 and atheroprotection,^18,19 although elevated levels of CD8⁺ lymphocytes in the blood of patients after myocardial infarction is a poor prognostic factor.²⁰ 

We sought to define the impact of thymus removal on CVD with particular emphasis on whether CVD was responsible for the significantly increased risk of mortality in patients who have received thymectomy.

All adult patients who underwent a thymectomy procedure at the Massachusetts General Hospital between 1 January 1993 and 1 March 2020 (N = 1532) were identified using the Mass General Brigham (MGB) Research Patient Data Registry. Patients who died within 90 postoperative days (N = 9) were excluded from all analyses. No patients with 22q11.2 deletion syndrome (DiGeorge syndrome) were identified. Patients who underwent a thymectomy procedure with <5 MGB interactions in the 5 years before matching were excluded. Controls included all patients who received care from the Massachusetts General Hospital Internal Medicine Associates and had at least 5 separate MGB system interactions (laboratory tests, medication prescriptions, diagnoses, etc) over any time period to reflect sustained MGB interaction (N = 62 422 patients). Patients who underwent a thymectomy procedure and controls were matched 5:1 (with replacement) at time of thymectomy based on demographics (age within 5 years, sex, and race), smoking status (never, former, and current), prescription of aspirin or statins, and comorbidities (hypertension, cancer, autoimmunity, hyperlipidemia, and CVD; data on body mass index were too scarce among patients who underwent a thymectomy procedure to match). Comorbidities were based on International Classification of Diseases (ICD) 9 and ICD 10 codes. For all patients, demographics, diagnostic history, and mortality status were collected from the Research Patient Data Registry, with data being valid up to 1 May 2023. Postmatch rates of mortality, cancer, and major adverse cardiovascular events (MACE; myocardial infarction, heart failure, and stroke) were analyzed using Kaplan-Meier curves and Cox proportional–hazards models, with patients censored at the date of their last MGB system interaction. All statistical analysis was performed in MATLAB. Patient data were collected using the MGB Research Patient Data Registry under an approved institutional review board protocol (number 2020P002958).

To determine the effect of thymectomy on life expectancy, the incidences of all-cause mortality and cancer after matching were compared between the 2 groups. The 5-year risk of all-cause mortality in the thymectomy group was more than twice that of the control group (relative risk, 2.4; 95% confidence interval [CI], 1.6-3.7), as was the risk of cancer (2.4 [95% CI, 1.3-4.4]; Figure 1A). Over multiple decades, all-cause mortality remained higher in the thymectomy group (hazard ratio, 1.9; 95% CI, 1.5-2.3; Figure 1B), as did cancer (hazard ratio, 1.3; 95% CI, 1.01-1.6; Figure 1C). Collectively, these results support the findings of our previous study of thymectomy risk but, here, compared with a novel, healthy control group, further demonstrating the importance of the adult thymus to human health.

We hypothesized that MACE would significantly contribute to the increased risk of mortality after thymectomy. However, after matching, the groups had similar rates of MACE (relative risk over 5 years, 0.29 [95% CI, 0.07-1.2]; Figure 1A; hazard ratio over 20 years, 0.75 [95% CI, 0.56-1.02]; Figure 1D), other diagnoses of CVD, and coronary revascularizations (coronary artery bypass grafting, 0.88 [95% CI, 0.36-2.16], and percutaneous coronary intervention, 0.84 [95% CI, 0.49-1.46]; Figure 1E). The 5-year risks of MACE were similar between controls and the United States population aged 45 to 65 years.²¹ 

To determine the causes of death after thymectomy, a detailed medical record review was performed for all 112 patients who underwent a thymectomy procedure and died after surgery, excluding patients who died within the first 90 postoperative days. This analysis matched on postoperative rates of infections, malignancies, and autoimmune disease. Primary causes of death for patients who underwent a thymectomy procedure and control patients were found to be different: cancer represented 63.8% of deaths in the thymectomy group compared with 29.8% in control patients (P < .001; Figure 2A), who largely died of CVD or incidental causes. The types of cancers that caused death in each group were different (Figure 2B). Patients who underwent a thymectomy procedure developed thymomas, sarcomas, and a host of various cancers from different tissue types across the body, vs control patients who largely developed lung, breast, gastrointestinal, and pancreatic cancers, the most common types of cancers worldwide.²² Before death, patients who underwent a thymectomy procedure experienced higher rates of cancer relapse despite definitive treatment (60.4% vs 31.5% of cancers, P < .001; Figure 2C) and the types of cancers that relapsed leading up to death were also distinct between groups. Taken together, these findings demonstrate that cancer, not CVD, is the overwhelming driver of mortality in patients who underwent a thymectomy procedure, who experience more diverse and recurrent cancers than controls.

Overall, the adult thymus is important for human longevity. Retention of the thymus appears to protect against the development of cancer and consequently reduces mortality. This study demonstrates that cancer is the most significant driver of death after thymectomy; the types of cancers that develop in patients who underwent a thymectomy procedure and contribute to death are more diverse and prone to relapse than those in control patients. This aligns with a model in which the thymus continuously produces new T cells that can maintain a broad repertoire of reactivity and thereby surveillance against emerging mutant cells. In the absence of new T-cell production and a contracted T-cell receptor repertoire in patients who underwent a thymectomy procedure, the risk of a diverse host of lethal cancers increases. We previously reported an increase in autoimmunity with thymectomy and show here that autoimmunity comprises 10.3% of the causes of mortality in patients who underwent a thymectomy procedure; this less clearly relates to the T-cell receptor repertoire. Rather, this may reflect imbalances in T regulatory cells, because the thymus is a critical source for their production. Our previous findings of elevated proinflammatory cytokines in the blood of patients who underwent a thymectomy procedure support that possibility. It was these proinflammatory findings that raised the question of whether CVD could be exacerbated by thymectomy. Given that patients who underwent a thymectomy procedure lose the ability to produce new T cells, studying patients who underwent a thymectomy procedure enables evaluation of newly produced T cells in CVD. This contrasts with the designs of many of the studies that have previously defined the role of the T-cell compartment in CVD, because they are based on experimental mouse models and observational human studies.^3,4,23 Our findings indicate that thymectomy is not associated with increased MACE, CVD, cardiac proceduralization, or cardiovascular death, supporting the conclusion that loss of new T-cell production does not exacerbate CVD. There may be a trend toward reduced CVD with thymectomy (Figure 1D-E) that future studies can further define. Our data do not exclude a role for existing T cells in CVD.

This study may help physicians counseling patients on the risks and benefits of thymectomy and indicates that thymectomy performed during cardiothoracic surgery does not exacerbate CVD but instead is associated with a significant increase in the risk of dying from cancer. As this study is retrospective and observational, its findings cannot demonstrate causation.

Contribution: K.A.K. and B.H.F. conducted the analysis; K.A.K., B.H.F., and R.A.B. analyzed the data; and all the authors designed the study and wrote the manuscript.

Conflict-of-interest disclosure: D.A.Z. has a consultancy with Bristol Myers Squibb; and receives research funding from Abbott Laboratories. T.G.N. has consultancies with Bristol Myers Squibb, Genentech, Sanofi, and Roche; and receives research funding from Bristol Myers Squibb, AstraZeneca, and Abbott Laboratories. D.T.S. has memberships on the boards of directors of Agios Pharmaceuticals, Editas Medicine, Carisma Therapeutics, Lightning Biotherapeutics, Garuda Therapeutics, and Sonata Therapeutics; receives stock options from Lightning Biotherapeutics, Sonata Therapeutics, Garuda Therapeutics, and Clear Creek Bio; and has a consultancy with VCanBio. The remaining authors declare no competing financial interests.

Correspondence: David T. Scadden, 185 Cambridge St, Simches Research Building, Massachusetts General Hospital, Boston, MA 02114; email: dscadden@mgh.harvard.edu.