Abstract
Among patients with ESRD, cancer risk is affected by kidney dysfunction and by immunosuppression after transplant. Assessing patterns across periods of dialysis and kidney transplantation may inform cancer etiology. We evaluated 202,195 kidney transplant candidates and recipients from a linkage between the Scientific Registry of Transplant Recipients and cancer registries, and compared incidence in kidney function intervals (time with a transplant) with incidence in nonfunction intervals (waitlist or time after transplant failure), adjusting for demographic factors. Incidence of infection-related and immune-related cancer was higher during kidney function intervals than during nonfunction intervals. Incidence was most elevated for Kaposi sarcoma (hazard ratio [HR], 9.1; 95% confidence interval (95% CI), 4.7 to 18), non-Hodgkin’s lymphoma (HR, 3.2; 95% CI, 2.8 to 3.7), Hodgkin’s lymphoma (HR, 3.0; 95% CI, 1.7 to 5.3), lip cancer (HR, 3.4; 95% CI, 2.0 to 6.0), and nonepithelial skin cancers (HR, 3.8; 95% CI, 2.5 to 5.8). Conversely, ESRD-related cancer incidence was lower during kidney function intervals (kidney cancer: HR, 0.8; 95% CI, 0.7 to 0.8 and thyroid cancer: HR, 0.7; 95% CI, 0.6 to 0.8). With each successive interval, incidence changed in alternating directions for non-Hodgkin’s lymphoma, melanoma, and lung, pancreatic, and nonepithelial skin cancers (higher during function intervals), and kidney and thyroid cancers (higher during nonfunction intervals). For many cancers, incidence remained higher than in the general population across all intervals. These data indicate strong short-term effects of kidney dysfunction and immunosuppression on cancer incidence in patients with ESRD, suggesting a need for persistent cancer screening and prevention.
Keywords: cancer, epidemiology and outcomes, kidney transplantation, dialysis
Kidney transplant recipients differ from other organ recipients because dialysis allows for survival after a failed transplant and during the wait for another kidney. As a result, recipients may undergo successive intervals of kidney function and nonfunction. These intervals are characterized by abrupt changes in the use of immunosuppressant medications (needed during kidney transplant intervals to prevent graft rejection) and dialysis (needed during intervals of kidney nonfunction).
Compared with the general population, kidney recipients have an elevated risk for cancer, especially virus-related cancers and skin cancers, largely attributable to the effects of immunosuppression.1 Dialysis patients experience elevated risk for kidney, thyroid, and bladder cancer, possibly related to side effects of kidney failure, including metabolic changes or inflammation.2
Prior studies of patient with ESRD show differences in cancer risk between the pretransplant dialysis period and after kidney transplantation, but have not identified whether cancer risk changes again if the transplant fails.3,4 Cancer incidence across intervals of kidney function and failure has been assessed only in one Australian study.5 However, due to sample size limitations, this study did not provide precise estimates for many individual cancers, and except for melanoma and lip cancer,6,7 did not address whether cancer incidence changed over multiple intervals.
Understanding how cancer risk changes as individuals with ESRD move between periods of dialysis (kidney nonfunction intervals) and transplant periods (kidney function intervals) can provide insight into cancer etiology. We examined this question using data from the United States transplant registry and linked cancer registries.
Results
Study Population
We included 202,195 unique kidney transplant candidates and/or recipients in our study. Most contributed time to the waitlist (90%, n=182,167), and more than half contributed to the first transplant interval (56%, n=113,038; Table 1). Substantially fewer individuals contributed time to subsequent intervals. The follow-up period during kidney function intervals was longer than for kidney nonfunction intervals (median 1530 days versus 661 days, respectively). Compared with individuals with nonfunctioning kidneys, those with transplants were more frequently white (Table 1). Median age decreased across successive intervals.
Table 1.
Characteristics of ESRD patients by intervals of kidney function and nonfunction
| Interval | Waitlist | First Transplant | First Graft Failure | Second or Higher Transplant | Second or Higher Graft Failure |
|---|---|---|---|---|---|
| Total N | 182,167 | 113,038 | 25,448 | 9401 | 2430 |
| Follow-up in days, median (IQR) | 624 (245–1269) | 1556 (676–2795) | 946 (398–1880) | 1239 (510–2292) | 1005 (436–1954) |
| Female sex, n (%) | 73,745 (40.5) | 45,481 (40.2) | 10,738 (42.2) | 3964 (42.2) | 978 (40.3) |
| Race/ethnicity, n (%) | |||||
| White, nonHispanic | 79,946 (43.9) | 58,117 (51.4) | 11,592 (45.6) | 5460 (58.1) | 1335 (54.9) |
| Black, nonHispanic | 49,352 (27.1) | 25,829 (22.9) | 8110 (31.9) | 1958 (20.8) | 639 (26.3) |
| Hispanic and other races | 52,869 (29.0) | 29,092 (25.7) | 5746 (22.6) | 1983 (21.1) | 456 (18.8) |
| Age at interval start in years, median (IQR) | 49 (38–59) | 47 (34–57) | 45 (33–55) | 40 (30–50) | 39 (29–49) |
IQR, interquartile range.
Cancer Incidence
When cancers were grouped, incidence rates for infection-related and immune-related cancers were higher during kidney function intervals than during nonfunction intervals (Figure 1). By contrast, incidence of ESRD-related cancers was higher during kidney nonfunction intervals. We did not observe an alternating incidence pattern for the group of other included cancers (colorectal, prostate, breast, esophageal, pancreas, and uterine cancers, along with myeloma and leukemia).
Figure 1.
Incidence rates for grouped cancers, by intervals of kidney function and nonfunction. Vertical error bars represent 95% CI for each incidence rate estimate. The x-axis is labeled with the kidney function or nonfunction interval corresponding with each incidence rate estimate. Definitions for the abbreviated kidney function and nonfunction intervals are: 1 Tx, First transplant; 1 GF, Graft failure following first transplant; 2+ Tx, Second or higher transplant; 2+ GF, Graft failure following second or higher transplant; WL, Waitlist.
When considered individually, incidence of numerous cancers differed between kidney function and nonfunction intervals. Among the infection-related cancers, Kaposi sarcoma (KS), non-Hodgkin’s lymphoma (NHL), and Hodgkin’s lymphoma had much higher incidence in kidney function intervals relative to nonfunction intervals (adjusted hazard ratios [HRs] of 9.1; [95% CI, 4.7 to 18], 3.2; [95% CI, 2.8 to 3.7], 3.0; [95% CI, 1.7 to 5.3], respectively, Table 2). Incidence of anal and genital cancers was also higher during kidney function intervals, whereas incidence of liver cancer was higher during nonfunction intervals, although the magnitude of the adjusted HRs was smaller.
Table 2.
Cancer incidence rates across intervals of kidney function and nonfunction
| Cancer Type | Total Cases | Incidence Rate per 100,000 Person-Years (95% CI), by Interval | Adjusted HRd (95% CI) | P Value | ||||
|---|---|---|---|---|---|---|---|---|
| Waitlist | First Transplant | First Graft Failure | Second or Higher Transplant | Second or Higher Graft Failure | ||||
| Infection-related | ||||||||
| Kaposi sarcoma | 96 | 1.7 (0.75 to 3.4) | 14 (11 to 18) | 1.1 (0.03 to 6.0) | 5 (0.6 to 18) | 11 (0.28 to 61) | 9.1 (4.7 to 18) | <0.001 |
| Non-Hodgkin’s lymphoma | 1171 | 45 (39 to 52) | 150 (140 to 160) | 26 (17 to 39) | 100 (75 to 140) | 0 (0 to 41) | 3.2 (2.8 to 3.7) | <0.001 |
| Hodgkin’s lymphoma | 81 | 2.6 (1.3 to 4.5) | 11 (8.1 to 14) | 2.2 (0.26 to 7.8) | 9.9 (2.7 to 25) | 11 (0.28 to 61) | 3 (1.7 to 5.3) | <0.001 |
| Liver | 196 | 25 (21 to 30) | 11 (8.9 to 15) | 9.7 (4.4 to 18) | 5 (0.6 to 18) | 11 (0.28 to 61) | 0.59 (0.44 to 0.80) | 0.001 |
| Stomach | 227 | 20 (16 to 25) | 20 (16 to 24) | 15 (8.3 to 25) | 9.9 (2.7 to 25) | 0 (0 to 41) | 1 (0.80 to 1.4) | 0.78 |
| Oropharynxa | 100 | 8.7 (6.2 to 12) | 8.9 (6.6 to 12) | 4.3 (1.2 to 11) | 9.9 (2.7 to 25) | 0 (0 to 41) | 1.2 (0.79 to 1.8) | 0.42 |
| Anus | 98 | 5.9 (3.9 to 8.5) | 10 (7.7 to 13) | 5.4 (1.8 to 13) | 17 (7 to 36) | 0 (0 to 41) | 1.8 (1.1 to 2.7) | 0.01 |
| Cervix | 74 | 14 (9.6 to 21) | 15 (11 to 21) | 10 (2.8 to 26) | 23 (6.3 to 59) | 28 (0.71 to 160) | 1.2 (0.74 to 1.9) | 0.5 |
| Other genital sitesb | 123 | 6.7 (4.6 to 9.5) | 11 (8.7 to 14) | 12 (5.9 to 21) | 27 (14 to 49) | 44 (12 to 110) | 1.5 (1 to 2.2) | 0.03 |
| Immune-related | ||||||||
| Lung | 1425 | 110 (97 to 120) | 140 (130 to 150) | 81 (64 to 100) | 100 (75 to 140) | 55 (18 to 130) | 1.3 (1.1 to 1.4) | <0.001 |
| Melanoma | 447 | 23 (19 to 28) | 54 (48 to 60) | 9.7 (4.4 to 18) | 37 (21 to 61) | 11 (0.28 to 61) | 1.9 (1.6 to 2.4) | <0.001 |
| Lip | 109 | 2.4 (1.2 to 4.3) | 15 (12 to 19) | 2.2 (0.26 to 7.8) | 12 (4.0 to 29) | 22 (2.7 to 79) | 3.4 (2 to 6) | <0.001 |
| Nonepithelial skinc | 172 | 4.3 (2.6 to 6.7) | 22 (19 to 27) | 7.6 (3.0 to 16) | 35 (19 to 58) | 0 (0 to 41) | 3.8 (2.5 to 5.8) | <0.001 |
| ESRD-related | ||||||||
| Kidney | 2128 | 200 (190 to 220) | 140 (130 to 150) | 300 (270 to 340) | 200 (160 to 250) | 350 (240 to 500) | 0.77 (0.7 to 0.84) | <0.001 |
| Urinary tract | 414 | 31 (27 to 37) | 39 (34 to 45) | 22 (13 to 33) | 35 (19 to 58) | 55 (18 to 130) | 1.2 (0.95 to 1.4) | 0.13 |
| Thyroid | 497 | 46 (40 to 52) | 34 (29 to 39) | 77 (60 to 97) | 25 (12 to 46) | 66 (24 to 140) | 0.67 (0.56 to 0.81) | <0.001 |
| Other | ||||||||
| Colorectum | 860 | 85 (76 to 93) | 69 (63 to 76) | 47 (34 to 64) | 42 (25 to 67) | 33 (6.8 to 96) | 0.88 (0.76 to 1) | 0.07 |
| Prostate | 1517 | 230 (220 to 250) | 230 (210 to 240) | 130 (98 to 160) | 140 (95 to 200) | 140 (62 to 280) | 1.2 (1.1 to 1.3) | 0.003 |
| Breast | 918 | 210 (190 to 240) | 170 (160 to 190) | 160 (130 to 210) | 100 (62 to 160) | 250 (120 to 480) | 0.81 (0.71 to 0.93) | 0.002 |
| Esophagus | 113 | 8 (5.6 to 11) | 11 (8.1 to 14) | 13 (6.7 to 23) | 2.5 (0.06 to 14) | 11 (0.28 to 61) | 1.1 (0.76 to 1.6) | 0.61 |
| Pancreas | 222 | 17 (13 to 21) | 22 (18 to 26) | 11 (5.2 to 20) | 15 (5.4 to 32) | 0 (0 to 41) | 1.5 (1.1 to 2) | 0.004 |
| Uterus | 178 | 37 (29 to 46) | 36 (28 to 44) | 38 (21 to 62) | 40 (16 to 83) | 0 (0 to 100) | 0.96 (0.71 to 1.3) | 0.77 |
| Myeloma | 210 | 20 (16 to 24) | 16 (13 to 20) | 17 (9.9 to 28) | 15 (5.4 to 32) | 22 (2.7 to 79) | 0.88 (0.66 to 1.2) | 0.36 |
| Leukemia | 240 | 17 (14 to 21) | 22 (18 to 26) | 25 (16 to 37) | 20 (8.5 to 39) | 11 (0.28 to 61) | 1.2 (0.9 to 1.5) | 0.25 |
Oropharynx cancer includes cancers of the base of tongue, tonsils, and other oropharynx sites.
Other genital cancers include cancers of the vagina, vulva, and penis.
Nonepithelial skin is defined as skin cancers excluding melanoma, Kaposi sarcoma, and squamous and basal cell carcinomas.
The HR compares all kidney function intervals (i.e., transplant intervals) to all kidney nonfunction intervals (waitlist and graft failure intervals), adjusting for sex, race/ethnicity, calendar year, and age (which is used as the time scale in the Cox model). Cervical, breast, and uterine cancers were evaluated only among women. Prostate cancer was evaluated only among men.
For all immune-related cancers, incidence was significantly higher during kidney function intervals (Table 2). The strongest associations were seen for lip and nonepithelial skin cancers (adjusted HRs of 3.4; [95% CI, 2.0 to 6.0] and 3.8; [95% CI, 2.5 to 5.8], respectively). Among the nonepithelial skin cancers, the association was especially strong for Merkel cell carcinoma (adjusted HR of 4.8 [95% CI=2.6-8.8]), which comprised 55% of these cases. However, higher incidence during kidney function intervals was also observed for other nonepithelial skin cancers (adjusted HR of 3.1; [95% CI=1.7-5.4]).
In the group of ESRD-related cancers, lower incidence during kidney function intervals was seen for kidney and thyroid cancers (adjusted HRs of 0.77; [95% CI, 0.70 to 0.84] and 0.67; [95% CI, 0.56 to 0.81], respectively, Table 2). By contrast, urinary tract cancer incidence did not differ.
Of the other cancers examined, few showed clear patterns across intervals. However, prostate and pancreatic cancer incidence rates were significantly higher during kidney function intervals, while breast cancer incidence was significantly lower during function intervals (Table 2).
Supplemental Table 1 presents results of our tests for alternating slopes between successive intervals. Of the cancers with higher incidence during kidney function intervals, NHL, anal cancer, lung cancer, melanoma, nonepithelial skin cancers, and pancreatic cancer consistently increased in incidence with each transition to a kidney function interval and decreased with each transition to a nonfunction interval, as demonstrated by the adjusted HRs comparing successive intervals (Figure 2). The two types of transitions were significant for NHL, lung cancer, melanoma, nonepithelial skin cancers, and pancreatic cancer (Supplemental Table 1). Among cancers with lower incidence during kidney function intervals, kidney cancer, thyroid cancer, and myeloma consistently decreased in incidence with each transition to a kidney function interval and increased in incidence with each transition to a nonfunction interval (Figure 2). Both transitions were significant for kidney and thyroid cancer (Supplemental Table 1).
Figure 2.
Hazard ratios comparing successive intervals for cancers with distinct alternating incidence patterns. Each point represents the HR comparing each interval to the immediately preceding interval. No HR estimate is given for the last interval when no cancers were observed in that interval. Vertical error bars represent 95% CI. Cancers are grouped as infection-related cancers (Panel A), immune-related cancers (Panel B), ESRD-related cancers (Panel C), and other cancers (Panel D). Within each panel, results are shown for cancer types that demonstrated alternating patterns of cancer incidence across intervals (i.e., based on the point estimates for the HRs, regardless of their statistical significance). Additional information regarding the formal test for alternating slopes is presented in Supplemental Table 1. Definitions for the abbreviated interval comparisons are: 1 Tx, First transplant compared with waitlist interval; 1 GF, Graft failure following first transplant compared with first transplant interval; 2+ Tx, Second or higher transplant compared with first transplant interval; 2+ GF, Graft failure following second or higher transplant compared with second or higher transplant; NHL, non-Hodgkin's lymphoma; WL, Waitlist, reference group.
For kidney cancer, incidence of localized cancer was higher during kidney nonfunction intervals than function intervals, but incidence of regional/distant cancers showed no clear pattern (Figure 3). For other cancers, the alternating incidence patterns did not appear to differ for local versus regional/distant stage cancers.
Figure 3.
Cancer incidence rates by stage, for cancers with alternating incidence patterns. The x-axis is labeled with the kidney function or nonfunction interval corresponding with each incidence rate estimate. Definitions for the abbreviated kidney function and nonfunction intervals are: 1 Tx, First transplant; 1 GF, Graft failure following first transplant; 2+ Tx, Second or higher transplant; 2+ GF, Graft failure following second or higher transplant; NHL, non-Hodgkin's lymphoma; WL, Waitlist.
Standardized Incidence Ratios
Within every interval, all of the ESRD-related cancers occurred at higher incidence than in the general population, although the standardized incidence ratios (SIRs) for kidney and thyroid cancer were noticeably higher in nonfunction intervals (Table 3). For kidney cancer, SIRs for both localized cancers and distant/regional stage cancers were significantly elevated in all intervals (not shown). Among the infection-related and immune-related cancers, SIRs for KS, lung cancer, and lip cancer were elevated across all intervals and were noticeably higher during function intervals (Table 3). SIRs for NHL, anal cancer, melanoma, and nonepithelial skin cancer were significantly elevated during function intervals, but not elevated across all kidney nonfunction intervals. For myeloma, leukemia, and genital cancers, SIRs were increased across all intervals, with no distinct contrast between function and nonfunction intervals.
Table 3.
Standardized incidence ratios across intervals of kidney function and nonfunction
| Cancer Type | Standardized Incidence Ratio (95% CI), by Interval | ||||
|---|---|---|---|---|---|
| Waitlist | First Transplant | First Graft Failure | Second or Higher Transplant | Second or Higher Graft Failure | |
| Infection-related | |||||
| Kaposi sarcoma | 6.4 (2.8 to 13) | 55 (44 to 68) | 5.9 (0.15 to 33) | 33 (4 to 120) | 100 (2.5 to 560) |
| Non-Hodgkin’s lymphoma | 1.7 (1.5 to 2) | 5.9 (5.5 to 6.3) | 1.3 (0.84 to 2) | 5.8 (4.1 to 7.8) | 0 (0 to 2.7) |
| Hodgkin’s lymphoma | 0.87 (0.45 to 1.5) | 3.4 (2.6 to 4.3) | 0.71 (0.09 to 2.6) | 3.1 (0.86 to 8.1) | 3.3 (0.08 to 19) |
| Liver | 1.8 (1.5 to 2.2) | 1 (0.79 to 1.3) | 0.99 (0.45 to 1.9) | 0.64 (0.08 to 2.3) | 1.6 (0.04 to 9.1) |
| Stomach | 1.4 (1.1 to 1.7) | 1.6 (1.3 to 1.9) | 1.5 (0.84 to 2.6) | 1.4 (0.37 to 3.5) | 0 (0 to 6.7) |
| Oropharynxa | 1.2 (0.88 to 1.7) | 1.3 (0.97 to 1.7) | 0.78 (0.21 to 2.0) | 1.9 (0.52 to 4.9) | 0 (0 to 9.5) |
| Anus | 2.6 (1.7 to 3.8) | 4.8 (3.7 to 6.2) | 3 (0.96 to 6.9) | 10 (4.1 to 21) | 0 (0 to 28) |
| Cervix | 0.89 (0.59 to 1.3) | 1.1 (0.75 to 1.5) | 0.71 (0.19 to 1.8) | 1.8 (0.48 to 4.6) | 2.3 (0.06 to 13) |
| Other genitalb | 3 (2 to 4.2) | 5.1 (4 to 6.5) | 6.8 (3.4 to 12) | 18 (8.9 to 32) | 36 (9.9 to 93) |
| Immune-related | |||||
| Lung | 1.2 (1.1 to 1.3) | 1.6 (1.5 to 1.7) | 1.3 (1 to 1.6) | 2.1 (1.5 to 2.9) | 1.4 (0.47 to 3.3) |
| Melanoma | 1.5 (1.2 to 1.8) | 2.8 (2.5 to 3.2) | 0.83 (0.38 to 1.6) | 2.4 (1.3 to 4) | 0.98 (0.03 to 5.5) |
| Lip | 3.5 (1.7 to 6.2) | 18 (15 to 22) | 4.3 (0.53 to 16) | 24 (7.7 to 56) | 50 (6.1 to 180) |
| Nonepithelial skinc | 2.5 (1.5 to 3.9) | 13 (11 to 15) | 5.5 (2.2 to 11) | 25 (14 to 44) | 0 (0 to 37) |
| ESRD-related | |||||
| Kidney | 9 (8.4 to 9.6) | 6.4 (5.9 to 6.8) | 18 (16 to 20) | 13 (10 to 16) | 28 (19 to 39) |
| Urinary tract | 1.6 (1.3 to 1.9) | 1.9 (1.6 to 2.2) | 1.8 (1.1 to 2.8) | 3 (1.6 to 5.3) | 7.5 (2.4 to 17) |
| Thyroid | 4 (3.5 to 4.6) | 2.9 (2.5 to 3.4) | 7.4 (5.8 to 9.4) | 2.1 (1 to 3.8) | 6.7 (2.4 to 15) |
| Other | |||||
| Colorectum | 1.2 (1.1 to 1.3) | 1.1 (0.96 to 1.2) | 0.95 (0.69 to 1.3) | 1 (0.57 to 1.6) | 1 (0.21 to 3) |
| Prostate | 0.85 (0.78 to 0.92) | 0.92 (0.85 to 0.98) | 0.65 (0.5 to 0.83) | 0.95 (0.65 to 1.3) | 1.3 (0.55 to 2.5) |
| Breast | 1.2 (1 to 1.3) | 0.95 (0.86 to 1.0) | 1.1 (0.86 to 1.4) | 0.76 (0.45 to 1.2) | 2.4 (1.1 to 4.5) |
| Esophagus | 0.96 (0.68 to 1.3) | 1.3 (1 to 1.7) | 2.2 (1.1 to 3.8) | 0.53 (0.01 to 2.9) | 2.9 (0.07 to 16) |
| Pancreas | 1.1 (0.86 to 1.4) | 1.5 (1.3 to 1.8) | 0.98 (0.47 to 1.8) | 1.7 (0.62 to 3.7) | 0 (0 to 5.8) |
| Uterus | 0.93 (0.72 to 1.2) | 0.94 (0.75 to 1.2) | 1.4 (0.77 to 2.3) | 1.6 (0.64 to 3.3) | 0 (0 to 5.8) |
| Myeloma | 1.8 (1.5 to 2.2) | 1.8 (1.4 to 2.1) | 2.2 (1.3 to 3.6) | 2.6 (0.97 to 5.7) | 4.4 (0.54 to 16) |
| Leukemia | 1.4 (1.1 to 1.8) | 1.8 (1.5 to 2.1) | 2.7 (1.7 to 4.1) | 2.3 (1 to 4.6) | 1.6 (0.04 to 9) |
Oropharynx cancer includes cancers of the base of tongue, tonsils, and other oropharynx sites.
Other genital cancers include cancers of the vagina, vulva, and penis.
Nonepithelial skin is defined as skin cancers excluding melanoma, Kaposi sarcoma, and squamous and basal cell carcinomas.
Discussion
In this population-based study of more than 200,000 patients with ESRD, we observed distinct patterns of incidence for several cancer types across intervals of kidney function and nonfunction. Overall, risk of infection-related and immune-related cancers was markedly higher during function intervals, when recipients had a transplant and were receiving immunosuppressant medications. By contrast, risk of ESRD-related cancers was higher during dialysis intervals following kidney failure. Some of our findings confirm those reported in a smaller Australian study that included only 7809 patients.5 Given our study’s much larger size, we were able to evaluate a number of specific cancer types not assessed previously, and we extended the previous findings by assessing cancer incidence over multiple consecutive intervals of kidney function and nonfunction. Remarkably, for a subset of cancers, consistent increases and decreases were seen with each interval transition, indicating a tight relationship between risk and immunosuppression or poor kidney function.
Overall, the higher risk of infection-related cancers during transplant intervals likely reflects the impact of the immunosuppression administered during these intervals. In particular, the incidence of KS and NHL differed greatly between transplant and dialysis intervals. NHL, which is part of a spectrum of post-transplant lymphoproliferative disorders, rises steeply in incidence soon after transplantation as a result of the intense immunosuppression administered during that time.8,9 We observed that NHL incidence changed markedly with each transition from a dialysis to a transplant interval. While fewer cases of KS were observed, the changes in incidence were of similar magnitude, and the SIRs during transplant intervals were much higher than for any other cancer type. Our findings extend upon those of the Australian linkage study, which identified a significantly higher NHL incidence and a (nonsignificant) higher KS incidence during transplant time, but did not evaluate transitions across multiple successive intervals. Both NHL and KS are linked with poor immune status in HIV populations,10,11 and the incidence of both declines rapidly in HIV patients after the initiation of antiretroviral therapy that leads to immune reconstitution.11,12
As we demonstrate for the first time, the human papillomavirus-related cancers (oropharyngeal, anal, cervical, and other genital cancers) had incidence patterns less distinctly associated with transplant intervals. After a human papillomavirus infection, immunosuppression is thought to increase the likelihood that the infection becomes persistent instead of being cleared.13,14 Years may pass between an initial persistent infection and the development of cancer. We believe that this long lag time, and the importance of immunosuppression especially at early time points in this process, prevents the incidence of these cancers from varying strongly across the intervals of immunosuppression. For example, the time between development of cervical intraepithelial neoplasia and progression to invasive cervical cancer can span decades.15 Likewise, in the HIV population, immunosuppression appears to influence risk of developing anal cancer over a 6–7-year period.16
Incidence of all immune-related cancers was higher during transplant intervals. Lung cancer, melanoma, and nonepithelial skin cancer risks changed distinctly across intervals. Inflammation and lung infections appear to contribute to the development of lung cancer.17,18 Our findings for lung cancer are consistent with an effect of immunosuppression, but they differ from those reported in the Australian study, which found higher lung cancer incidence during dialysis intervals, but not at a level that reached significance.5 The importance of the immune system for the control of melanoma is highlighted by the excess of nevi observed in immunosuppressed populations, and the responsiveness of melanoma to immunotherapy.19,20 The nonepithelial skin cancers in our study were largely cases of Merkel cell carcinoma, an uncommon cancer related to Merkel cell polyomavirus,21 but this category also included rarer subtypes of skin cancer not linked to infection.
A strongly contrasting pattern was seen for two ESRD-related cancers, kidney cancer and thyroid cancer, which manifested higher incidence during dialysis intervals. Dialysis and kidney dysfunction may increase kidney cancer risk for a number of reasons, including by inducing inflammation and allowing the accumulation of uremic toxins in nonfunctioning kidneys. Perhaps the most important risk factor is acquired polycystic kidney disease, a condition that develops in people with CKD and decreases in prevalence following kidney transplantation.22 Acquired polycystic kidney disease is associated with an elevated risk of kidney cancer,23,24 particularly localized stage cancer,22 which may explain why the changes in kidney cancer risk in our study were limited to localized cancers. It has been hypothesized that the association of thyroid cancer with ESRD could be a result of metabolic dysfunction arising with kidney failure.5,25 An association with thyroid cancer was also observed in the Australian study and other studies of ESRD populations.2,3,5 The pattern of alternating risk observed in our study does not seem to be attributable to increased surveillance for thyroid cancer, as risk during nonfunction intervals was elevated for both regional/distant and localized stage cancers.
Most other cancers did not show distinct alternating patterns of incidence, with the exception of pancreatic cancer. Risk of pancreatic cancer was consistently higher during transplant intervals, though a biologic rationale for this finding is not readily available. Pancreatic cancer incidence was higher in our kidney recipients than in the general population, but elevated risk has not been observed consistently in other immunosuppressed populations.1,26
Of note, the alternating patterns that we observed for a number of specific cancers are consistent with short-term effects of kidney damage and immunosuppression. The declines in incidence following recovery of kidney function or cessation of immunosuppression also indicate that these effects are at least partly transient. Together, these findings suggest that these biologic disturbances influence the later steps of carcinogenesis, perhaps promoting tumor cell proliferation resulting in a rapid progression to malignancy.
Nonetheless, another noteworthy finding of our study is that, for many cancers, risk was consistently higher than in the general population across all intervals. Elevated SIRs were observed in all intervals for all ESRD-related cancers, suggesting that some long-term effects of kidney damage and dialysis may linger after transplantation. The SIRs for kidney and thyroid cancer in nonfunction intervals were higher than those reported in prior studies of dialysis populations, while the incidence rates for these cancers were lower in our study than in the prior studies.2,27,28 These contrasts likely reflect the younger average age of our population, and the fact that cancer risk increases with age in the general population. Risk was also elevated for most infection-related and immune-related cancers in all of the intervals, likely reflecting some immune dysfunction present even during dialysis, due to the immunosuppressive and inflammatory effects of uremia.29 This may also reflect that some patients may still be maintained on low levels of immunosuppressants after graft failure.
Our population of ESRD patients provides an ideal setting for examining the effects of immunosuppression and kidney damage, because the time intervals during which these exposures occurred were clearly demarcated by well documented clinical events (i.e., transplantation and graft failure). Individuals during all of these intervals were likely to have a similar distribution of risk behaviors (e.g., smoking) and comorbidities. Although these characteristics were not measured in our study, these underlying similarities facilitated comparisons of cancer risk. Our study benefited from linkage to population-based cancer registries, which provide reliable ascertainment of cancer diagnoses. We also evaluated the largest population of kidney transplant candidates and recipients currently available, a population more than 20 times larger than the prior Australian linkage study. As a result, we were able to examine a greater number of cancer types over multiple intervals.
Despite these advantages, several limitations remain. For the kidney function intervals, we cannot disentangle the effects of immunosuppression from direct effects of immunosuppressive medications. Likewise, for the nonfunction intervals, we cannot separate the effects of poor kidney function from the effects of dialysis. Also, while individuals in all time intervals were from the same ESRD population, selection related to survival or receipt of a transplant may have influenced our results. Sicker patients may have never received a transplant, while healthier patients would have been more likely to survive to receive multiple kidney transplants. This likely explains why the median age was lower for each subsequent interval. Despite this limitation, any selection effects would not be expected to produce the dramatic alternating incidence patterns observed in our study. Finally, we were not able to examine incidence of squamous cell or basal cell skin cancers, even though they are common and important cancers in the transplant population, because these cancers are not ascertained by cancer registries.
In conclusion, we found that for ESRD-related cancers, particularly kidney cancer and thyroid cancer, risk was higher during intervals of dialysis and absent kidney function. This heightened risk decreased after the receipt of kidney transplant. By contrast, risks for infection-related and immune-related cancers, particularly NHL, melanoma, lung cancer, and nonepithelial skin cancer, were higher during transplant intervals, and decreased after transplant failure and the cessation or reduction of immunosuppressant use. The time course of these effects suggests that immunosuppression and kidney damage influence the later steps of development of these malignancies. While incidence changed dramatically across transplant and dialysis intervals, risk for several cancers was elevated in all intervals when compared with the general population, highlighting the excess burden of cancer in this population that persists over time. Our findings underscore the need for targeted cancer prevention and screening strategies in the ESRD population, and demonstrate that these needs may change across kidney function and nonfunction intervals.
Concise Methods
Study Design
The Transplant Cancer Match Study links 15 population-based cancer registries to the Scientific Registry of Transplant Recipients (SRTR), a registry of all American solid organ transplant candidates and recipients (including 517,686 kidney candidates or recipients).1 For this study, we included kidney candidates and recipients who resided in areas covered by participating cancer registries and who had a follow-up time as described below (39% of all American kidney candidates/recipients).
Intervals of kidney nonfunction were defined as time from first kidney wait-listing to transplant, or from failure of a kidney transplant to subsequent transplant, if one occurred. Intervals of kidney function were defined as any period after receipt of a kidney transplant and before kidney graft failure. As only a small number of people received more than two kidney transplants, intervals of kidney function corresponding to a second or later kidney transplant were combined. Similarly, all kidney nonfunction intervals corresponding to failure of second or later transplants were combined. This grouping yielded five intervals: waitlist, first transplant, first graft failure, second or higher transplant, and second or higher graft failure. Each person could contribute to each interval.
We focused on infection-related cancers (KS, lymphomas, and cervical, anal, vaginal/vulvar, penile, oropharyngeal, liver, and stomach cancers),30 ESRD-related cancers (i.e., cancers increased in dialysis patients: kidney, urinary tract, and thyroid cancers),5,27 immune-related cancers (i.e., cancers unrelated to known infections but increased in immunosuppressed populations: melanoma and lung, lip, and nonepithelial skin cancers), and other common cancers with counts greater than 100 across all intervals (e.g., colorectal, breast, and prostate cancers).26 Associations were considered with cancer groups (infection-related, ESRD-related, immune-related, and other) and individual cancers.
Statistical Analyses
At-risk time for each person started at the latest of: start of cancer registry coverage or entry onto the kidney transplant waitlist (or kidney transplant, if it was performed pre-emptively). The follow-up period ended at the earliest of: death, nonkidney transplant (including multiorgan kidney transplant), loss-to-follow-up by the SRTR, or end of cancer registry coverage. Nonkidney transplants were excluded because their immunosuppressant use may not clearly correspond with kidney function.
Incidence rates were calculated for each cancer within each of the five intervals. Cox regression models were used to estimate HRs comparing cancer risk in all kidney function time to all nonfunction time. Age was used as the time scale. Multivariable regression models additionally adjusted for sex, race/ethnicity, and attained calendar year. To capture how risk changed over time, we also assessed models that compared each interval to the immediately preceding interval.
We hypothesized that cancers tightly related to immunosuppression (or kidney failure) would exhibit clear changes in incidence with each successive change in kidney function. For each cancer type, we used an alternating slopes test that evaluated whether the combined relative changes from kidney function to nonfunction intervals were statistically significant, and whether the combined relative changes from kidney nonfunction to function intervals were statistically significant.31 If both slope tests indicated statistical significance, then the cancer was considered tightly linked to immunosuppression or kidney failure. We also calculated SIRs, which describe cancer incidence relative to the expected incidence, based on general population rates obtained from the cancer registries participating in the Transplant Cancer Match Study.
Because the effects of immunosuppression and kidney failure on cancer development are unlikely to be immediately apparent, all analyses were done with 3-month lags, in which incidence and association measures were calculated for cancers occurring 3 months after the time in each interval. We also examined incidence patterns separately for local and regional/distant stage cancers, to assess whether patterns might be due to differential surveillance (which would mainly lead to changes in the incidence of local stage cancer) or biologic factors broadly affecting cancer aggressiveness.
Disclosures
E.L.Y. and E.A.E. developed the concept and design of the study. E.L.Y. performed all data analyses and wrote the initial draft of the manuscript. All authors were involved in the evaluation of study methods, interpretation of the results, and revisions to the manuscript. All authors approved the final version of the manuscript. E.L.Y. had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Supplementary Material
Acknowledgments
The authors gratefully acknowledge the support and assistance provided by individuals at the Health Resources and Services Administration, the Scientific Registry of Transplant Recipients (SRTR), and the following cancer registries: the states of California, Colorado, Connecticut, Georgia, Hawaii, Iowa, Illinois, Michigan, New Jersey, New York, North Carolina, Texas, and Utah, and the Seattle-Puget Sound area of Washington. We also thank analysts at Information Management Services for programming support.
The views expressed in this paper are those of the authors and should not be interpreted to reflect the views or policies of the National Cancer Institute, Health Resources and Services Administration, SRTR, cancer registries, or their contractors.
Some results from this manuscript were presented at the World Transplant Congress, July 26–31, 2014, in San Francisco, California.
This research was supported by the Intramural Research Program of the National Cancer Institute.
During the initial period when registry linkages were performed, the SRTR was managed by Arbor Research Collaborative for Health in Ann Arbor, Michigan (contract HHSH234200537009C), beginning in September 2010, the SRTR was managed by Minneapolis Medical Research Foundation in Minneapolis, Minnesota (HHSH250201000018C).
The following cancer registries were supported by the National Program of Cancer Registries of the Centers for Disease Control and Prevention: California (agreement 1U58 DP000807-01), Colorado (U58 DP000848-04), Georgia (5U58DP003875-01), Illinois (5U58DP003883-03), Michigan (5U58DP003921-03), New Jersey (5U58/DP003931-02), New York (U58DP003879), North Carolina (U58DP000832), and Texas (5U58DP000824-04).
The following cancer registries were supported by the SEER Program of the National Cancer Institute: California (contracts HHSN261201000036C, HHSN261201000035C, and HHSN261201000034C), Connecticut (HHSN261201000024C), Hawaii (HHSN261201000037C, N01-PC-35137, and N01-PC-35139), Iowa (HSN261201000032C and N01-PC-35143), New Jersey (HHSN261201300021I, N01-PC-2013-00021), Seattle-Puget Sound (N01-PC-35142), and Utah (HHSN261201000026C).
Additional support was provided by the states of California, Colorado, Connecticut, Illinois, Iowa, New Jersey, New York (including the Cancer Surveillance Initiative), Texas, and Washington, as well as the Fred Hutchinson Cancer Research Center in Seattle, Washington.
Footnotes
Published online ahead of print. Publication date available at www.jasn.org.
See related editorial, “Cancer in ESRD: Clear on the Epidemiology, Hazy on the Mechanisms” on pages 1272–1275.
This article contains supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2015040373/-/DCSupplemental.
References
- 1.Engels EA, Pfeiffer RM, Fraumeni JF Jr, Kasiske BL, Israni AK, Snyder JJ, Wolfe RA, Goodrich NP, Bayakly AR, Clarke CA, Copeland G, Finch JL, Fleissner ML, Goodman MT, Kahn A, Koch L, Lynch CF, Madeleine MM, Pawlish K, Rao C, Williams MA, Castenson D, Curry M, Parsons R, Fant G, Lin M: Spectrum of cancer risk among US solid organ transplant recipients. JAMA 306: 1891–1901, 2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Maisonneuve P, Agodoa L, Gellert R, Stewart JH, Buccianti G, Lowenfels AB, Wolfe RA, Jones E, Disney AP, Briggs D, McCredie M, Boyle P: Cancer in patients on dialysis for end-stage renal disease: an international collaborative study. Lancet 354: 93–99, 1999 [DOI] [PubMed] [Google Scholar]
- 3.Vajdic CM, McDonald SP, McCredie MR, van Leeuwen MT, Stewart JH, Law M, Chapman JR, Webster AC, Kaldor JM, Grulich AE: Cancer incidence before and after kidney transplantation. JAMA 296: 2823–2831, 2006 [DOI] [PubMed] [Google Scholar]
- 4.Fischereder M: Cancer in patients on dialysis and after renal transplantation. Nephrol Dial Transplant 23: 2457–2460, 2008 [DOI] [PubMed] [Google Scholar]
- 5.van Leeuwen MT, Webster AC, McCredie MR, Stewart JH, McDonald SP, Amin J, Kaldor JM, Chapman JR, Vajdic CM, Grulich AE: Effect of reduced immunosuppression after kidney transplant failure on risk of cancer: population based retrospective cohort study. BMJ 340: c570, 2010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.van Leeuwen MT, Grulich AE, McDonald SP, McCredie MR, Amin J, Stewart JH, Webster AC, Chapman JR, Vajdic CM: Immunosuppression and other risk factors for lip cancer after kidney transplantation. Cancer Epidemiol Biomarkers Prev 18: 561–569, 2009 [DOI] [PubMed] [Google Scholar]
- 7.Vajdic CM, van Leeuwen MT, Webster AC, McCredie MR, Stewart JH, Chapman JR, Amin J, McDonald SP, Grulich AE: Cutaneous melanoma is related to immune suppression in kidney transplant recipients. Cancer Epidemiol Biomarkers Prev 18: 2297–2303, 2009 [DOI] [PubMed] [Google Scholar]
- 8.Gibson TM, Engels EA, Clarke CA, Lynch CF, Weisenburger DD, Morton LM: Risk of diffuse large B-cell lymphoma after solid organ transplantation in the United States. Am J Hematol 89: 714–720, 2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.van Leeuwen MT, Grulich AE, Webster AC, McCredie MR, Stewart JH, McDonald SP, Amin J, Kaldor JM, Chapman JR, Vajdic CM: Immunosuppression and other risk factors for early and late non-Hodgkin lymphoma after kidney transplantation. Blood 114: 630–637, 2009 [DOI] [PubMed] [Google Scholar]
- 10.Martró E, Esteve A, Schulz TF, Sheldon J, Gambús G, Muñoz R, Whitby D, Casabona J; Euro-Shaks study group : Risk factors for human Herpesvirus 8 infection and AIDS-associated Kaposi’s sarcoma among men who have sex with men in a European multicentre study. Int J Cancer 120: 1129–1135, 2007 [DOI] [PubMed] [Google Scholar]
- 11.Yanik EL, Napravnik S, Cole SR, Achenbach CJ, Gopal S, Olshan A, Dittmer DP, Kitahata MM, Mugavero MJ, Saag M, Moore RD, Mayer K, Mathews WC, Hunt PW, Rodriguez B, Eron JJ: Incidence and timing of cancer in HIV-infected individuals following initiation of combination antiretroviral therapy. Clin Infect Dis 57: 756–764, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lacombe JM, Boue F, Grabar S, Viget N, Gazaignes S, Lascaux-Cametz AS, Pacanowski J, Partisani M, Launay O, Matheron S, Rosenthal E, Rouveix E, Tattevin P, de Truchis P, Costagliola D, Goedert JJ: Risk of Kaposi sarcoma during the first months on combination antiretroviral therapy. AIDS 27: 635–643, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Massad LS, Ahdieh L, Benning L, Minkoff H, Greenblatt RM, Watts H, Miotti P, Anastos K, Moxley M, Muderspach LI, Melnick S: Evolution of cervical abnormalities among women with HIV-1: evidence from surveillance cytology in the women’s interagency HIV study. J Acquir Immune Defic Syndr 27: 432–442, 2001 [DOI] [PubMed] [Google Scholar]
- 14.Chin-Hong PV, Palefsky JM: Natural history and clinical management of anal human papillomavirus disease in men and women infected with human immunodeficiency virus. Clin Infect Dis 35: 1127–1134, 2002 [DOI] [PubMed] [Google Scholar]
- 15.McCredie MR, Sharples KJ, Paul C, Baranyai J, Medley G, Jones RW, Skegg DCG: Natural history of cervical neoplasia and risk of invasive cancer in women with cervical intraepithelial neoplasia 3: a retrospective cohort study. Lancet Oncol 9: 425–434, 2008 [DOI] [PubMed] [Google Scholar]
- 16.Bertisch B, Franceschi S, Lise M, Vernazza P, Keiser O, Schöni-Affolter F, Bouchardy C, Dehler S, Levi F, Jundt G, Ess S, Pawlita M, Kovari H, Wandeler G, Calmy A, Cavassini M, Stöckle M, Clifford G; Swiss HIV Cohort Study Investigators : Risk factors for anal cancer in persons infected with HIV: a nested case-control study in the Swiss HIV Cohort Study. Am J Epidemiol 178: 877–884, 2013 [DOI] [PubMed] [Google Scholar]
- 17.Shebl FM, Engels EA, Goedert JJ, Chaturvedi AK: Pulmonary infections and risk of lung cancer among persons with AIDS. J Acquir Immune Defic Syndr 55: 375–379, 2010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Shiels MS, Pfeiffer RM, Hildesheim A, Engels EA, Kemp TJ, Park JH, Katki HA, Koshiol J, Shelton G, Caporaso NE, Pinto LA, Chaturvedi AK: Circulating inflammation markers and prospective risk for lung cancer. J Natl Cancer Inst 105: 1871–1880, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Grob JJ, Bastuji-Garin S, Vaillant L, Roujeau JC, Bernard P, Sassolas B, Guillaume JC: Excess of nevi related to immunodeficiency: a study in HIV-infected patients and renal transplant recipients. J Invest Dermatol 107: 694–697, 1996 [DOI] [PubMed] [Google Scholar]
- 20.Topalian SL, Sznol M, McDermott DF, Kluger HM, Carvajal RD, Sharfman WH, Brahmer JR, Lawrence DP, Atkins MB, Powderly JD, Leming PD, Lipson EJ, Puzanov I, Smith DC, Taube JM, Wigginton JM, Kollia GD, Gupta A, Pardoll DM, Sosman JA, Hodi FS: Survival, durable tumor remission, and long-term safety in patients with advanced melanoma receiving nivolumab. J Clin Oncol 32: 1020–1030, 2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Clarke CA, Robbins HA, Tatalovich Z, Lynch CF, Pawlish KS, Finch JL, Hernandez BY, Fraumeni JF Jr, Madeleine MM, Engels EA: Risk of merkel cell carcinoma after solid organ transplantation. J Natl Cancer Inst 107: 1–9, 2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Bonsib SM: Renal cystic diseases and renal neoplasms: a mini-review. Clin J Am Soc Nephrol 4: 1998–2007, 2009 [DOI] [PubMed] [Google Scholar]
- 23.Ishikawa I: Development of adenocarcinoma and acquired cystic disease of the kidney in hemodialysis patients. Princess Takamatsu Symp 18: 77–86, 1987 [PubMed] [Google Scholar]
- 24.Gulanikar AC, Daily PP, Kilambi NK, Hamrick-Turner JE, Butkus DE: Prospective pretransplant ultrasound screening in 206 patients for acquired renal cysts and renal cell carcinoma. Transplantation 66: 1669–1672, 1998 [DOI] [PubMed] [Google Scholar]
- 25.Vamvakas S, Bahner U, Heidland A: Cancer in end-stage renal disease: potential factors involved -editorial-. Am J Nephrol 18: 89–95, 1998 [DOI] [PubMed] [Google Scholar]
- 26.Grulich AE, van Leeuwen MT, Falster MO, Vajdic CM: Incidence of cancers in people with HIV/AIDS compared with immunosuppressed transplant recipients: a meta-analysis. Lancet 370: 59–67, 2007 [DOI] [PubMed] [Google Scholar]
- 27.Shebl FM, Warren JL, Eggers PW, Engels EA: Cancer risk among elderly persons with end-stage renal disease: a population-based case-control study. BMC Nephrol 13: 65, 2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Butler AM, Olshan AF, Kshirsagar AV, Edwards JK, Nielsen ME, Wheeler SB, Brookhart MA: Cancer incidence among US Medicare ESRD patients receiving hemodialysis, 1996-2009. Am J Kidney Dis 65: 763–772, 2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Betjes MG: Immune cell dysfunction and inflammation in end-stage renal disease. Nat Rev Nephrol 9: 255–265, 2013 [DOI] [PubMed] [Google Scholar]
- 30.Bouvard V, Baan R, Straif K, Grosse Y, Secretan B, El Ghissassi F, Benbrahim-Tallaa L, Guha N, Freeman C, Galichet L, Cogliano V; WHO International Agency for Research on Cancer Monograph Working Group : A review of human carcinogens--Part B: biological agents. Lancet Oncol 10: 321–322, 2009 [DOI] [PubMed] [Google Scholar]
- 31.Follmann D: A simple multivariate test for one-sided alternatives. J Am Stat Assoc 91: 854–861, 1996 [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.