Cancer incidence following bariatric surgery in renal transplant recipients: a retrospective multi-center analysis

Abstract

Background:

Obesity, a known independent risk factor for developing malignancy. Additionally, renal transplant recipients (RTR) confer a 2- to 4-fold increased risk of overall malignancies with an excess absolute risk of .7% per year. While transplant recipients are at risk for obesity and malignancy, the effect of bariatric surgery (BS) in the posttransplantation setting is not well known.

Objectives:

Our study primarily evaluated the impact of BS on cancer incidence in RTR with severe obesity in the posttransplantation setting. Weight loss outcomes were analyzed secondarily.

Setting:

University Hospital.

Methods:

A retrospective study using TriNetX database was developed to analyze cancer outcomes in RTR with posttransplantation BS versus RTR without BS from 2000 to 2023. After the exclusion process and propensity matching, both cohorts consisted of 153 patients.

Results:

RTR-BS had a significantly lower incidence of overall cancer and transplant-related cancers (P < .05). No significant difference was identified in cutaneous, gastrointestinal, and reproductive cancers. Percent Excess Weight Loss (%EWL) was significantly lower in RTR-only cohort (11.4%) versus RTR-BS cohort (57.8%) at 5 years. Sleeve gastrectomy (SG) patients (73.19%) had significantly higher %EWL than Roux en-Y gastric bypass (RYGB) patients (49.33%) at 3 years. No difference in cancer incidence was noted between SG and RYGB patients.

Conclusion:

Postrenal transplantation BS had a diminishing effect on overall and transplant-related cancer incidence in RTR with severe obesity. Significant weight loss was also demonstrated with post-renal transplantation BS.

Obesity is a common phenomenon following renal transplantation. Up to 50% of recipients gain weight with an estimated increase in body mass index (BMI) by 2–4 kg/m² within the first year following transplantation [1]. This post-transplant obesity is attributed to multiple factors, including immunosuppressive medications and behavioral changes. Transplant physicians typically optimize immunosuppressive regimens by minimizing the use of corticosteroids, transitioning to medications with fewer metabolic side effects, or encouraging lifestyle modifications. Both pretransplant and de novo posttransplant obesity can negatively impact transplant outcomes and increase the risk of graft dysfunction, cardiovascular disease, diabetes with associated insulin resistance, dyslipidemia, hypertension, and malignancy [2–4].

Obesity is also established as an independent risk factor for cancer and is estimated to surpass smoking as the leading cause of preventable cancers [5]. Bariatric surgery (BS) has been shown to not only reduce the incidence and severity of cardiovascular and endocrine pathologies, but also significantly reduce the risk of cancer related morbidity and mortality. Weight-loss induced changes to hormonal imbalances and chronic inflammatory status is postulated to improve cancer outcomes following BS [6]. Several malignancies including breast, endometrial, colorectal, prostate, hepatocellular, and hematologic cancers have shown to be positively impacted with any modality of BS [7–12].

Malignancy following solid organ transplantation is a known complication attributed to the effect of immunosuppressive medications. In fact, reports have demonstrated a 2- to 4-fold increased risk of overall malignancies with an excess absolute risk of .7% per year secondary to transplantation [13]. Kaposi’s sarcoma, non-Hodgkin’s lymphoma, kidney, skin, ovarian, and colon cancers are among some of the most common cancers in patients undergoing renal transplant [14–16]. Renal transplant recipients (RTR) who develop obesity postoperatively may be at an increased risk of cancer. BS prior to transplantation has been shown to be effective in improving candidacy, while mitigating post-transplant weight gain and the resultant sequelae of poor outcomes [17–19]. Current research seeks to elucidate the feasibility of BS in the posttransplantation setting for patients who develop severe obesity postoperatively. Small scale studies have indicated that BS is safe and effective in achieving meaningful weight loss [20,21]. However, the effect of BS in reducing cancer risk in RTR with severe obesity is not well known.

Our study primarily evaluates the effect of post-transplantation BS on the incidence of cancers in RTR with severe to morbid obesity. Our secondary outcome was to explore weight loss in RTR up to 5 years after transplantation.

Methods

A retrospective study was designed using TriNetX database, leveraging the US Collaborative Network, comprising data from 48 Healthcare Organizations. This dataset, based off electronic medical records, encompassed a wide range of information, including demographic details, diagnoses (recorded using the International Classification of Diseases, 10th Revision, Clinical Modification, ICD-10-CM codes), procedures (coded under the International Classification of Diseases, 10th Revision, Procedure Coding System, ICD-10-PCS, or Current Procedural Terminology, CPT), medications (coded as per the Veterans Affairs National Formulary), laboratory tests (coded with Logical Observation Identifiers Names and Codes, LOINC), genomics data (coded using the Human Genome Variation Society, HGVS), and healthcare utilization information.

Two cohorts, consisting of adult RTR with severe obesity (defined as BMI > 35kg/m²) which did or did not undergo post-transplant BS, were established. Renal transplantation was confirmed based on CPT codes (1008098, 1008109, 50360, 50365, 00868) and ICD-10 codes (Z94.0, 0TY0, 0TY00Z0, 0TY00Z1, 0TY00Z2, 0TY1, 0TY10Z0, 0TY10Z1, and 0TY10Z2). The presence of BS was confirmed based on CPT codes (43775, 43644, 43645, and 43846). Patients with a history of previous transplantation, concurrent multiorgan transplantation, pre-transplant BS, or previous history of cancer were excluded.

All cancers of interest were categorized as gastrointestinal, cutaneous, reproductive, and transplant-related cancers (Table 1). As a unique International Classification of Diseases (ICD)-10 codes for native renal cell carcinoma (RCC) and allograft RCC do not exist, and the incidence of native RCC is reported to be insignificant compared to the allograft RCC incidence, allograft and native kidney RCC were considered for analysis all together. Transplant-related cancers combined all-type RCC along with Post-transplant Lymphoproliferative Disorder and Non-Hodgkin’s lymphoma as these malignancies are associated with the immunosuppressive status of transplant recipients. During the outcome search, all subcodes and corresponding ICD9 codes were included to identify as many feasible records as possible. The earliest catch for each patient was included in the resulting dataset, which was then enriched with patient metadata and additional clinical data. Patient metadata, including sex, race, ethnicity, and age at transplant for both cohorts were incorporated into the dataset. BMI was calculated using the patient’s height (inches) and weight (pounds) according to the formula (BMI = weight/[height^2] * 703) to convert to appropriate units of kg/m². In the RTR-BS cohort, weight and height documented up to 180 days prior to BS were used. In the RTR-only cohort, weight and height documented 1–2 years following transplantation were used. Of the RTR-BS, weight loss was further evaluated between those who underwent Sleeve Gastrectomy versus Roux-en-Y Gastric Bypass (RYGB) at 1 year, 3 years and 5 years following BS.

Table 1.

Composite of Cancers created based on ICD-10-CM Codes

ICD-10-CM Based Cancer Composites
C00-D49	Over all Malignancies	Over all Malignancies
C15-C26	Gastrointestinal Malignancies	GI Malignancies
C46	Kaposi’s Sarcoma	Cutaneous Malignancies
L57.0	Actinic keratosis
C43	Malignant melanoma of skin
C44	Basel and squamous cell carcinoma of skin
C50	Malignant neoplasm of breast	Reproductive Malignancies
C61	Prostate Cancer
C51-C58	Female Reproductive Organ Cancers
C64	Malignant neoplasm of kidney, except renal pelvis	Transplant Related Malignancies
C85.9	Non-Hodgkin lymphoma, unspecified
D47.Z1	Posttransplant lymphoprpliferative disorder

Cohorts were balanced using a Propensity Score Matching and Inverse Probability Treatment Weighting approach, matching patients based on sex, race, ethnicity, age, and BMI > 35 kg/m². The statistical significance between the two cohorts was calculated using the Chi-Square contingency test from the Python package SciPy (version 1.10.1), with a P value of .05. The null hypothesis stated that there was no significant difference between the observed and expected frequencies, and any difference was due to chance.

A total of 53,989 RTR were identified from 2000 to 2023 in the United States. There were 490 RTR with documented BS and 53,499 RTR without a history of BS. Of these, 235 and 1135 patients were excluded from RTR-BS and RTR-only cohorts, respectively, due to previous history of kidney transplantation, concurrent multiorgan transplantation, and history of BS prior to transplantation. Due to a 100-fold difference between RTR-only cohort versus RTR-BS, random selection of 9956 RTR-only was performed. Then, 9286 and 102 patients were excluded in RTR-only and RTR-BS, respectively, based on age less than 18 years and a BMI under 35 kg/m² at the time of BS. Of the RTR-only cohort, 23 were excluded due to lack of adequate records at least 2 years following transplantation. A final cohort of 153 RTR-BS was propensity matched to the 656 RTR-only cohort for outcome analysis (Fig 1).

Figure 1.

Flowchart of Inclusion and exclusion.

Results

The median time to BS following transplantation was 3.2 years. The average age was 49.4 and 50.3 years in RTR-BS and RTR-only cohorts, respectively (P = .5). Female RTR (67%, n = 103) predominately underwent BS compared to their male counterparts (33%, n = 50). There was no significant difference in gender distribution and race/ethnicity distribution between both cohorts (Table 2). The average BMI was 42.96 kg/m² and 42.58 kg/m² in RTR-BS and RTR-only cohorts, respectively (P = .5).

Table 2.

Propensity Matched Patient Demographics

Total Matched n = 153	RTR-BS (%)	RTR-only (%)	P-value
Age (Yrs ± SD)	49.4 ± 10.9	50.3 ± 11.8	.5
Ethnicity
Hispanic or Latino	10 (6.5)	6 (3.9)	.44
Not Hispanic or Latino	131 (85.6)	134 (87.6)	.74
Unknown	12 (7.8)	13 (8.5)	1
Race
Asian	1 (0.65)	1 (0.65)	1
Black/African American	58 (37.9)	56 (36.6)	.9
White	78 (51)	80 (52.3)	.9
Other	16 (10.5)	16 (10.5)	1
Sex
Female	103 (67.3)	101 (66)	.9
Male	50 (32.7)	52 (34)	.9
BMI (kg/m2) ± SD	42.9 ± 5.30	42.6 ± 4.57	.5

BMI, body mass index.

Overall, all cancer incidence was significantly higher in the RTR-only cohort (54.9%, n = 84) compared to RTR-BS (38.5%, n = 59) with an odds ratio (OR) of .52 (P = .005; confidence interval [CI] .32, .83). Incidence of transplant-related cancers was also significantly higher in the RTR-only cohort (13.1%, n = 20) compared to the RTR-BS (5.8%, n = 8) with an OR of .42 (P = .05; CI .136, .99). No significant difference in cutaneous cancers between the RTR-only (15.7%, n = 24) and RTR-BS (11.1%, n = 17) cohorts was noted with an OR of .7 (P = .31; CI .32, 1.37). No significant difference in gastrointestinal malignancies between RTR-only (3.9%, n = 6) and RTR-BS (4.6%, n = 7) was noted with an OR of 1.2 (P = 1; CI .33, 4.34). No significant difference in reproductive malignancies between RTR-only (5.8%, n = 9) and RTR-BS (5.2%, n = 8) was noted with an OR of .8 (P = 1; CI .29, 2.66) (Table 3). Sub-analysis of RTR-BS cohort was conducted to evaluate the differences in outcomes based on type of surgery performed. Of these, 113 and 40 received RYGB and SG, respectively. Propensity matching was performed prior to analysis. No significant difference in overall cancer incidence, transplant-related cancer incidence, cutaneous cancer incidence, gastrointestinal cancer incidence, and reproductive cancer incidence between SG and RYGB was demonstrated (Table 4). No differences in cancer incidence were observed based on gender among each cohort.

Table 3.

Cancer Incidence in RTR-BS and RTR-only^*

Cancer Groups	RTR/BS (%)	RTR (%)	OR (CI)	P Value
All	59 (38.56)	84 (54.9)	.52 (0.32, 0.83)	.005
Transplant	9 (5.88)	20 (13.7)	0.42 (.136, 0.99)	.05
Renal Neoplasms	3 (1.0)	18 (5.9)	0.15 (.02–0.5)	.002
Post-transplant lymphoproliferative disorder	5 (1.63)	3 (1.0)	1.69 (.32–11.1)	.72
Non-Hodgkin lymphoma	3 (1.0)	2 (.65)	1.51 (.17–18.3)	1
Cutaneous	17 (11.1)	24 (15.68)	.67 (.32, 1.37)	.31
Actinic keratosis	12 (3.92)	6 (1.96)	2.08 (.69–06.90)	.22
Skin Neoplasms (BCC + SCC)	10 (3.27)	14 (4.58)	0.70 (.27 –1.75)	.52
Melanoma	1 (0.32)	3 (1.0)	0.33 (.01–4.16)	.61
Gastrointestinal	7 (4.57)	6 (3.9)	1.2 (.33, 4.34)	1
Stomach	1 (0.32)	0	N/A	1
Liver	3 (1.0)	0	N/A	.25
Colon	3 (1.0)	7 (2.2)	0.42 (.07–1.90)	.33
Pancreas	0	2 (.65)	0	.48
Rectum	0	1 (.32)	0	1
Reproductive	8 (5.23)	9 (5.8)	0.88 (.29, 2.66)	1
Cervix Uteri	0	2 (.65)	0	.48
Corpus Uteri	0	2 (.65)	0	.48
Prostate	3 (1.0)	6 (1.96)	0.49 (.08–2.35)	.50
Vulva	1 (0.32)	0	N/A	1
Breast	4 (1.31)	4 (1.31)	1 (.18–5.48)	1
Uterus	1 (.32)	0	N/A	1

RTR-BS, Renal Transplant Recipients with Bariatric Surgery; RTR, Renal Transplant Recipients; OR, Odds ratio.

^*We document the number of patients diagnosed with each cancer, however, a singular patient could have been diagnosed with multiple cancers. Therefore, the cancer incidence within each category is not equal to the sum of specific cancer incidence.

Table 4.

Incidence of cancer between renal transplant recipients who underwent SG versus Gastric Bypass

Cancer Groups	SG (%)	RYGB (%)	OR (CI)	P-Value
All	15 (36.6)	18 (43.9)	.74 (.28, 1.96)	.65
Transplant	2 (4.9)	1 (2.4)	2.03 (.1, 123.9)	1
Cutaneous	5 (12.2)	7 (17.1)	.68 (.15, 2.76)	.75
Gastrointestinal	2 (4.9)	2 (4.9)	1.0 (.07, 14.44)	1
Reproductive	3 (7.3)	4 (9.7)	.73 (.10, 4.66)	1

SG, Sleeve Gastrectomy; RYGB, Roux en-Y Gastric Bypass; OR, Odds ratio.

On evaluation of weight loss outcomes, in the RTR-BS cohort, there was an average loss in BMI of 8.22 kg/m² at 6 months (n = 150), 10.45 kg/m² at 1 year (n = 128), 10.37 kg/m² at 3 years (n = 77), 10.83 kg/m² at 5 years (n = 40), which was significantly higher at all timepoints compared to RTR-only cohort. There was an average percent Excess Weight Loss (%EWL) and an average percent total body weight loss %TBWL of 47.1% and 19.0% at 6 months, 59.1% and 24.15% at 1 year, 55.5% and 23.5% at 3 years, 57.8% and 24.4% at 5 years (Fig. 2). Average absolute loss in BMI was significantly higher in SG patients compared to RYGB patients at all timepoints (P = .006, .02, .02, .04 at 6-mo, 1-yr, 3-yr, and 5yr respectively). SG patients had an average %EWL and %TBWL of 52.35% and 21.8% at 6 months, 66.41% and 27.9% at 1 year, 73.19% and 30.9% at 3 years, 73.38% and 31.5% at 5 years. RYGB patients had an average %EWL and %TBWL of 45.28% and 18.1% at 6 months, 56.42% and 22.7% at 1 year, 49.33% and 20.9% at 3 years, 49.38% and 20.6% at 5 years (Figure 2 & 3). The %EWL was significantly higher in SG patients compared to RYGB patients at 6 months and 3 years but not 1 year and 5 years (P = .04, .08, .008, .07 at 6-mo, 1-yr, 3-yr, and 5-yr, respectively). Percent TBWL was significantly higher in SG patients compared to RYGB patients at all timepoints (P = .008, .02, .01, .04 at 6-mo, 1-yr, 3-yr, and 5-yr, respectively).

Figure 2.

Average % Excess Weight Loss in Renal Transplant Recipients with Bariatric Surgery (RTR-BS) and Renal Transplant Recipients without Bariatric Surgery (RTR-only). RTR-only patients have significantly lower %EWL compared to RTR-BS. SG patients have significantly higher %EWL than RYGB at 6 months and 3 months but not 1 year and 5 years. RTR, Renal Transplant Recipients; BS, Bariatric Surgery; SG, Sleeve Gastrectomy; RYGB, Roux en-Y Gastric Bypass; EWL, Excess Weight Loss.

Figure 3.

Average % Excess Weight Loss in Renal Transplant Recipients after Sleeve Gastrectomy (SG) and Roux-En Y Gastric Bypass (RYGB). SG patients have significantly higher %EWL than RYGB at 6 months and 3 months but not 1 year and 5 years. SG, Sleeve Gastrectomy; RYGB, Roux en-Y Gastric Bypass; EWL, Excess Weight Loss.

Discussion

Our study evaluated the effect of BS in curtailing the risk of most common obesity and transplant-related cancers in adult RTR with severe to morbid obesity. A significant reduction in overall and transplant-related cancers was demonstrated up to 5 years following BS but the same was not shown from this database for cutaneous, gastrointestinal, and reproductive cancers. Scores of studies since the popularization of BS have shown the cancer-abating benefits of BS. A population-based study of 3 Nordic countries recently showed a significant reduction in esophageal adenocarcinoma following BS [22]. Researchers in Cleveland, Ohio in 2023 have also shown a significantly lower incidence of obesity-related cancer (breast, colon, liver, endometrial, and ovarian) following BS in patients with severe obesity [23]. All cancer risk along with hormonal cancer risk was demonstrated to be significantly reduced with BS as well [24,25]. Despite our low sample size of patients who underwent post-transplantation BS, our data demonstrated a reduced cancer risk in congruence with other studies. Counter to previously published data demonstrating the reduction in incidence of reproductive, obesity-related, and endocrine-related cancers following BS is not symmetrically reflected in our data. We surmise that this may be due to not only the short-term follow up but also due to the small number of RTR who undergo posttransplantation BS for weight loss. As BS becomes more popular in this specific patient population, it is reasonable to hypothesize that the incidence of these cancers will follow along the trends seen in the general population with obesity, which are treated with BS. Further speculation based on data showing increased risk of gastroesophageal reflux disease following SG compared to RYGB with reduced risk of other diseases would suggest that the metabolic sequelae following BS is partially influenced by the variant of surgery [25]. With the growing advances in BS and new patient populations reaping the benefits of effective weight loss following surgery, it is imperative to understand the implications of BS in RTR with obesity as well. Transplant physicians in the pre-transplant and posttransplant setting need to consider the risk of posttransplant obesity with an added emphasis on posttransplant obesity mediated cancer risk. Furthermore, recommendations regarding the potential need for BS based on the patient’s weight gain, especially if other avenues of weight loss have already been thoroughly attempted must be part of follow up. With two independent risk factors for cancer at play, this is a vulnerable population that could benefit in many ways. The disease burden of cancer is well established, especially in the United States [26–29]. Certainly, RTR are similarly encumbered by their increased cancer risk and the role of BS in assuaging this increased health care cost in the way of reducing cancer risk needs to be further explored.

Weight loss in terms of absolute weight loss in pounds, average change in BMI, percent excess weight loss, and total body weight loss are benchmarks employed by bariatric surgeons to evaluate the efficacy of weightloss surgery in patients with obesity. Our study demonstrated that post-transplantation BS had an overall significant reduction in BMI and higher %EWL and %TBWL. This is consistent with studies showing effective long-lasting weight loss following BS [30,31]. Our sub-analysis further showed a much higher reduction in BMI along with higher %EWL and %TBWL in patients who underwent SG versus RYGB. This, however, is not consistently reflected in studies evaluating the differences between SG and RYGB. RYGB patients, on average, tend to have significantly higher reduction in weight [32,33]. While our data were propensity matched based on sample size, the paradoxical results may not completely represent the actual differences in weight loss between SG and RYGB due to dwindling sample size at each time point as well as discordant electronic health records between the two sub-cohorts.

As this is a retrospective study based on an electronic medical records database, there is an inherent limitation due to a reliance on appropriate documentation that was used for analysis. The database also does not provide information regarding pathology or imaging studies, which are the gold standard for cancer diagnosis, depending on the type of cancer. As such, ICD codes were used as a proxy for adequate cancer diagnosis, which may have either overestimated or underestimated certain cancers. The majority of skin squamous cell carcinoma or basal cell carcinomas are diagnosed clinically but not always documented as malignant neoplasms, for example. Screening and diagnostic procedures may be better indicators of cancer incidence; however malignant diagnosis does not directly correlate to procedures undergone during cancer work-up.

It is evident that most cancers are insidious in nature and take years before they are clinically obvious, especially if protocols do not exist for routine screening in this patient population for all cancers. Skin, breast, and colorectal cancers have well-established guidelines for screening, but the rest of the cancers are clinically evaluated based on symptoms [34–41]. Screening guidelines, which mostly parallel those that apply to the general population, recommend annual skin exams for skin cancer, mammography every 1–2 years for breast cancer, and colonoscopy every 10 years for colorectal cancers following transplantation. No consensus exists for screening guidelines for the remainder of cancers, but annual comprehensive history and physical is recommended to identify the need for workup for other cancers [42]. While the screening guidelines depending on the cancer type may vary, the need for continued screening beyond 5 years of BS is under no dispute. Furthermore, data from Australia demonstrated that the mean time to diagnosis for all cancers is 9.4 years (SD 6.5 years). Small intestinal cancer had the lowest mean time to diagnosis at .5 years (SD .06 years) and liver cancer had the highest mean time to diagnosis at 15.2 years (SD 6.1 years) [42]. While significant weight loss was noted in post-transplant BS patients, the impact on cancer risk may not be adequately surmised within 5 years of intervention. As post-transplantation BS continues to be relatively new in the algorithm of weight-loss management in renal transplant patients, this database is the best available comprehensive information for 5 years following BS. While our data exploration included data at 10 and 20 years, key information regarding weight and cancer incidence was not reliable enough for our team to safely report the data. The preliminary significances noted in our study in overall cancer and transplant-related cancer risk may not be the only cancer groups that are positively impacted if the patient population is monitored beyond 5 years following BS.

Both cancer risk and weight loss following BS have been positively impacted based on plethora of data from other previous studies. We demonstrated the same consistent effect in RTR with severe to morbid obesity who are treated with BS following transplantation.

Conclusion

Obesity is a common issue in RTR, influenced by immunosuppressive medications and lifestyle changes posttransplantation. Managing obesity in this population is vital to ensure optimal transplant outcomes and long-term health. BS, as a means of weight management, offers promise in not only reducing cancer risk but also achieving effective weight loss up to 5 years following intervention in RTR with severe to morbid obesity. As such, it is imperative to include BS as an option for the management of transplant patients suffering from obesity. Continued evaluation is necessary to truly understand the effect of BS for all cancers both in the short-and long-term periods. If substantiated, this approach may improve the risk of cancer and, subsequently, quality of life for RTR with severe to morbid obesity.