A National Survey of Transplant Surgeons and Nephrologists on Implementing Apolipoprotein L1 ( APOL1) Genetic Testing Into Clinical Practice

Abstract

Introduction:

There is debate over whether Apolipoprotein L1 (APOL1) gene risk variants contribute to African American (AA) live donors' (LD) increased risk of kidney failure. Little is known about factors influencing physicians' integration of APOL1 genetic testing of AA LDs into donor evaluation.

Design:

We conducted a cross-sectional survey, informed by Roger's Diffusion of Innovations theory, among nephrology and surgeon members of the American Society of Nephrology, American Society of Transplantation, and American Society of Transplant Surgeons about their practices of and attitudes about APOL1 genetic testing of AA potential LDs. Descriptive statistics and bivariate analyses were performed.

Results:

Of 383 completed surveys, most physicians believed that APOL1 testing can help AA LDs make more informed donation decisions (87%), and the addition of APOL1 testing offers better clinical information about AA LD's eligibility for donation than existing evaluation approaches (74%). Among respondents who evaluate LDs (n = 345), 63% would definitely or probably begin or continue using APOL1 testing in the next year, however, few use APOL1 testing routinely (4%) or on a case-by-case basis (14%). Most did not know the right clinical scenario to order APOL1 testing (59%), but would use educational materials to counsel AA LDs about APOL1 testing (97%).

Discussion:

Although physicians were highly supportive of APOL1 genetic testing for AA LDs, few physicians use APOL1 testing. As more physicians intend to use APOL1 testing, an ethical framework and clinical decision support are needed presently to assist clinicians in clarifying the proper indication of APOL1 genetic testing.

Introduction

African Americans (AAs)/Blacks have disproportionately greater prevalence of chronic kidney disease (CKD) and end-stage renal disease (ESRD) than other racial/ethnic groups.¹ Such disparities extend to AA living kidney donors (donors). The incidence of ESRD and CKD, and their risk factors, is significantly higher in AA donors than in European American donors.^2-4 The estimated risk of ESRD at 15 years post-donation is higher in AA donors than in European American donors: 74.7 versus 22.7 per 10 000 living donors.

Genetic factors may contribute to these disparities.^5-7Apolipoprotein L1 (APOL1) gene risk variants are significantly associated with CKD and ESRD in individuals of African ancestry (such as AAs),⁸ but not other populations.^9,10 The Dallas Heart Study found that, among AAs without diabetes, significantly more participants with 2 APOL1 risk variants (6.7%) had CKD (GFR <60 mL/min per 1.73 m²) than those with 0 to 1 risk variants (1.7%).⁷ In several large population-based studies, 13% to 15% of AAs without kidney disease have 2 APOL1 risk variants and 39% have one risk variant.^7,11 African Americans with 2 gene risk variants have a 10-fold odds of FSGS-, and 7-fold odds of hypertension-attributed ESRD.⁹

Evidence suggesting the contribution of APOL1 risk variants to donor outcomes comes from deceased donors (DD). The APOL1 risk variants in AA DDs significantly increased the risk of graft failure in kidney transplant recipients, regardless of recipients’ race/ethnicity¹² or possessing 2 risk variants.¹³ Kidneys from DDs with 2 APOL1 risk variants had significantly worse graft survival (2-fold hazard ratio) than kidneys from DDs with no or one variant.^12,14 The APOL1 risk variant was found in 13% of AA DDs,¹⁵ and virtually all DDs with APOL1 risk variants were of African ancestry.¹² Case reports of living donors developing ESRD suggest APOL1 risk variants as contributing factors.^16,17

These findings have generated heightened ethical concern and uncertainty in the transplant community over whether to use genetic testing to protect donor safety, improve donor informed consent, and reduce disparities.^18,19 The American Society of Transplantation’s (AST) 2015 consensus conference recommended against routine APOL1 testing for AA donors because CKD/ESRD rarely arises in individuals with 2 gene variants; the risk of CKD/ESRD was very low among living donors; and utilizing APOL1 testing may potentially magnify disparities for AAs in access to kidney transplantation.²⁰ However, the Kidney Disease: Improving Global Outcomes guidelines subsequently recommended “considering APOL1 genotyping” in AA donors.^21(pS76) The National Institutes of Health has since funded the APOL1 Long-term Kidney Transplantation Outcomes Network (APOLLO) in 2017 to assess APOL1 risk variants’ effect on kidney function in kidney transplant recipients receiving kidneys from AA deceased or living donors.

Without professional consensus and definitive outcomes data, transplant physicians must rely on their clinical judgment to determine whether to use APOL1 testing, which could affect donors’ decision making. As proponents posit, APOL1 testing could risk stratify donors; the presence of 2 risk variants should comprise a relative contraindication to donation, requiring careful counseling and consent.^6,22 Opponents contend that APOL1 testing is premature given its lack of specificity in predicting disease,²² and lack of long-term population-based studies of APOL1 in donors.²³

With the advent of commercial testing for APOL1 risk variants, more transplant centers use APOL1 testing.⁶ The AST surveyed 83 AST/American Society of Nephrology (ASN) Transplant Nephrology Fellowship Directors and transplant programs, of which 13% used APOL1 genotyping.²⁰ Little is known about physicians’ attitudes about APOL1 testing and barriers and facilitators to the adoption of APOL1 testing into clinical practice.

Our national survey aimed to assess transplant and nephrology physicians’ normative attitudes, behavioral intentions, and practices of APOL1 testing. Attitudes and intentions are important to evaluate because, as precursors to actual behaviors,²⁴ they inform clinical judgment affecting care that donors receive now and in the future. Assessing attitudes before APOLLO study results emerge can illuminate factors that may influence the adoption of APOL1 genetic testing into clinical practice.

Design/Methods

Design

A cross-sectional survey assessed physicians’ attitudes and use of APOL1 testing. Northwestern University’s Institutional Review Board approved this study.

Sample Population and Recruitment

United States-based transplant surgeons, transplant nephrologists, and general nephrologists (MD or DO) who were members of the American Society of Transplant Surgeons (ASTS), the ASN, or AST were eligible to participate. Eligible participants were recruited differently depending on how each Society shared their membership lists. The ASTS (900 US physician members) provided members’ e-mail addresses. Members were recruited via e-mail with a URL link to the online survey, with 5 reminder e-mails sent every 1 to 4 weeks.

The ASN (8104 US physician members) provided single use access to members’ mailing addresses. General nephrologists’ perspectives are important because they have expertise in evaluating risk of kidney disease progression, follow many donors post-donation long-term, and provide insight into the risk of kidney disease progression for donors outside the transplant context. Given limited research resources, a random sample stratified by state of 4005 members was recruited once via postal mail and given a URL link to the online survey. The statistician (J.L.) provided a computer-generated random number list with different sample sizes for different strata (ie, states with members <50 were oversampled) to generate the sample.

The AST declined paid access to members’ contact information. Therefore, physician members of AST’s Kidney Pancreas, Transplant Diagnostics, Live Donor, and Psychosocial Communities of Practice were invited to participate via e-mail to each one’s listserv with a URL link to the online survey. Communities of Practice included 147, 146, 212, and 18 US physician members, respectively. Four e-mails to each listserv were sent every 1 to 4 weeks. The online survey, mailed survey, and recruitment letter included a fact sheet describing the current science on APOL1 so that potential participants gained a knowledge base with which to respond to the survey.

To increase response rates, multiple recruitment attempts were used. Participants had an option to enter a drawing to win a $10 Starbucks gift card or an Android or Apple watch.²⁵ Survey initiation indicated consent to participate in the study.

Theoretical Framework

Roger’s theory of Diffusion of Innovations²⁶ guided survey question development, as used in another study of adopting genetic testing.²⁷ Roger’s theory explains how a new behavior (ie, APOL1 testing) spreads through a social system, that is, the transplant field, and gets adopted into practice.²⁶ Physicians’ adoption of APOL1 testing will depend on perceptions of APOL1 testing as new or innovative. Physicians will vary in their stage of adoption from awareness of the need for the innovation, to deciding to adopt testing, to initial use of APOL1 testing, and to continued use of APOL1 testing. Diffusion of Innovations theory, as it applies to the transplant context, postulates that physicians are more likely to adopt the innovation if they perceive it as: (1) advantageous in conferring benefits to potential donors, based on knowledge of the existence of, indications for, and intended use of APOL1 testing (ie, for decision-making), (2) compatible with organizational and individual norms, attitudes, and needs, (3) simple and readily available to use, (4) usable to try out before committing to adoption, (5) observably conferring benefits to potential donors, that is, awareness of cases of APOL1 testing, and (6) open to modification by providers, that is, integrating genetic services into care.²⁶

Data Collection

The survey was developed by experts in research methods (E.J.G., R.R.S., and C.W.) and transplant and nephrology (J.F. and E.J.G.) to enhance its content validity. The survey was pilot-tested among 8 transplant surgeons, transplant and general nephrologists at Northwestern Medicine® to refine question clarity, response options, and order. The survey included 56 closed-ended and 4 open-ended questions. Topics covered: current clinical practice of APOL1 testing (n = 10), hypothetical clinical decisions about donors with APOL1 risk variants (n = 6), normative expectations about APOL1 testing (n = 3), attitudes about APOL1 testing (n = 11), barriers and facilitators to adoption and use of APOL1 testing (n = 7), thresholds for adopting APOL1 testing into clinical practice (n = 2), demographics (n = 3), and clinical practice characteristics (n = 14). A copy of the survey is freely available upon request.

Data Analysis

Descriptive statistics (eg, frequencies, percentages, means, standard deviations) assessed the distribution of survey responses across the sample. X² or Fisher exact test statistics, as appropriate, were used to assess the relationship between attitudinal variables and physician type; and between physicians’ experience with cases of APOL1 genetic test results and donor decision-making. Likert scale questions were treated as categorical variables. Transplant surgeons were compared to all nephrologists, and transplant physicians (surgeons, nephrologists) were compared to general nephrologists. All tests were 2-sided, performed using SPSS version 24 (Armonk, New York), and P value <.05 was considered statistically significant.

Results

Participant Characteristics

Of 11 906 members, 5177 were eligible to participate or were sampled from the list (ASN only) (Figure 1). Physicians were ineligible (n = 749) due to bounced back e-mails, returned mail, not being US physicians, or self-identified as ineligible given retirement or clinical inactivity. Of the eligible participants, 466 members started the survey, but 83 were excluded due to not specifying clinical specialty, nonphysician status, and duplicate entries. A total of 383 surveys were included in analysis. Table 1 presents participants’ demographic and clinical practice data. Participants were mostly male (68%), white (59%), general nephrologists (49%), transplant nephrologists (27%), and transplant surgeons (24%).

Figure 1.

Participation rate.

Table 1.

Participants’ Demographic and Transplant Center Characteristics.^a

Characteristic	Nb	%
Sex	379
Male	262	68
Female	117	31
Age, years, mean (SD), range	380, 48 (12), 25-83
20-29	6	1
30-39	89	23
40-49	124	33
50-59	87	23
60-69	56	15
70+	18	5
Race/ethnicity	371
White	220	59
Asian/Pacific Islander	95	26
Black/African American	25	7
Hispanic	22	6
Other/multiracial	9	2
Clinical specialty	383
General nephrology	187	49
Transplant nephrology	105	27
Transplant surgery	91	24
Years in clinical practice, mean (SD), range	372, 16 (11), 0-52
0-10	164	44
11-20	93	25
21-30	74	20
31+	41	11
Influence over the living donor evaluation and testing processc	367
No influence	74	20
A little influence	78	21
Some influence	96	26
A lot of influence	105	29
Complete influence	14	4
Membership in professional societiesd	383
American Society of Nephrology	281	74
American Society of Transplantation	163	43
American Society of Transplant Surgeons	92	24
Number of patients on dialysis,e1 mean (SD), range	193, 69 (102), 0-1000
0-20	69	36
21-50	52	27
51-100	38	20
101+	34	17
Percentage of nephrology practice’s dialysis patients who are African Americane2	260
0%-5%	32	12
6%-10%	27	10
11%-15%	15	6
16%-20%	36	14
21%-30%	40	16
31%+	110	42
Type of practice	376
Academic practice	272	72
Private practice	104	28
Practice affiliated with a transplant center	377
Yes	317	84
No	60	16
Kidney transplant volume in 2015c	189
0-50 kidney transplants	36	19
51-100 kidney transplants	58	31
101-200 kidney transplants	52	27
201+ kidney transplants	43	23
Living kidney donor transplant volume in 2015c	303
0-10 living donor kidney transplants	49	16
11-30 living donor kidney transplants	108	36
31-50 living donor kidney transplants	53	18
51-75 living donor kidney transplants	41	13
76+ living donor kidney transplants	52	17
Percentage of center’s living kidney donors who are African Americanc	291
0%-5%	77	27
6%-10%	71	24
11%-15%	45	15
16%-20%	45	15
21%-30%	31	11
31%+	22	8
Respondents by UNOS region	308
Region 1	25	8
Region 2	39	13
Region 3	28	9
Region 4	17	6
Region 5	31	10
Region 6	10	3
Region 7	41	13
Region 8	30	10
Region 9	26	8
Region 10	23	8
Region 11	38	12

Abbreviations: ASN, American Society of Nephrology; AST, American Society of Transplantation; ASTS, American Society of Transplant Surgeon.

^aN = 383.

^bN varies by question as some participants did not respond resulting in missing responses.

^cOnly transplant physicians (transplant surgeons + transplant nephrologists) were asked this question.

^dParticipants could select more than 1 society membership. The ASN includes ASN + AST n = 99. The ASTS includes ASTS + AST n = 55, and ASN + AST + ASTS n = 3.

^e1Only nephrologists (general nephrologists + transplant nephrologists) were asked this question. Of n = 292 asked, n = 193 responded

^e2Only nephrologists (general nephrologists + transplant nephrologists) were asked this question. Of n = 292 asked, n = 260 responded.

Clinical Practice of Using APOL1 Testing

Most respondents were aware of APOL1 testing as an option for AA potential donors (78%). Among respondents who evaluate donors (n = 345), few currently use APOL1 testing for AA potential donors routinely (4%), or on a case-by-case basis (14%), or did not know (16%).

Experience With APOL1 Testing

Forty-nine (13%) of 380 respondents knew of or had participated in the care of an AA potential donor for whom APOL1 testing was used. Among 48 of these respondents, more genetic test results were positive (i.e., 2 risk variants were found) (n = 26, 54%) than negative (n = 22, 46%). Among 47 cases, the results strongly or slightly dissuaded donors from donating (40%), strongly or slightly encouraged donors to donate (19%), or had no effect (40%). Bivariate analysis revealed a significant relationship between test results and donors’ decision making (P < .001). Positive test results were related to donors being strongly or slightly dissuaded from donating (79.2%), while negative results were related to donors being strongly or slightly encouraged to donate (33.3%). In a few cases, positive results had no effect on donors’ decision (23.1%), and negative results had no effect on donors’ decision (61.9%).

Figure 2 identifies clinicians who provided counseling/education to AA potential donors before and after APOL1 genetic testing. On both occasions, leading clinicians were the respondent, another physician, and nurse or nurse practitioner. Leading topics discussed before APOL1 testing included: medical (51/63, 81%), donor candidacy (48/63, 76%), family/social (29/63, 46%), insurance (22/63, 35%), and financial (18/63, 29%) implications of positive test results.

Figure 2.

Clinicians providing counseling before and after using Apolipoprotein L1 (APOL1) genetic testing.

Likelihood of Using APOL1 Genetic Testing

Among respondents who evaluate donors, 63% would definitely or probably begin or continue using APOL1 testing in the next year, while 21% would definitely or probably not, and 16% did not know. More nephrologists than surgeons/transplant physicians would definitely or probably begin or continue using APOL1 testing in the next year (P < .002, P < .006). Of the general nephrologists surveyed, most (83%) evaluate donors, and of these, 63% would definitely or probably begin or continue using APOL1 testing in the next year.

Hypothetical Clinical Decisions About Donors With APOL1 Risk Variants

In a hypothetical scenario in which an AA potential donor had 2 APOL1 risk variants, half of respondents involved in evaluating donors reported they would not proceed with the living donation or would strongly recommend against it (50%), the rest would either proceed with donation or discuss the pros and cons and let the potential donor decide (40%), or were unsure (10%). When the scenario changed to the donor having one APOL1 risk variant, few would not proceed with the donation or would strongly recommend against it (9%), most would recommend proceeding with donation or discuss the pros and cons and let the potential donor decide (80%), or were unsure (12%). Respondents reported that the decision whether AA donors with 2 APOL1 risk variants should donate resided in the multidisciplinary team (51%), donor (30%), transplant nephrologist (12%), or others.

Ethical Perspectives and Attitudes About APOL1 Testing

Potential donor considerations.

Most respondents believed that all AA potential donors should be informed about the option of APOL1 testing (69%) than were against it (4%) or were unsure (27%); significantly more nephrologists agreed than surgeons (P < .001) (Table 2). Most strongly or slightly agreed that APOL1 testing can help donors make more informed donation decisions (87%) and is beneficial for AA donors (74%), but it would reduce the number of AA donors (73%). Significantly more nephrologists expressed agreement with these attitudes than surgeons/transplant physicians (Table 3).

Table 2.

Ethical Perspectives about APOL1 Testing.

		Yes	No	Unsure
	Na	%	%	%	P b
All African American potential living donors should be informed about the option of APOL1 testing.	379	69	4	27
Nephrologist	289	75	3	22	<.001
Surgeon	90	51	6	43
General nephrologist	185	81	3	16	0.001
Transplant physician	194	57	5	38
Potential recipients should be informed about their African American potential living donors’ APOL1 test results.	381	43	27	30
Nephrologist	290	47	24	29	0.002
Surgeon	91	28	38	34
General nephrologist	186	58	16	26	<.001
Transplant physician	195	28	38	34
African American living donors who have already donated should be tested for APOL1 risk variants.	382	28	29	43
Nephrologist	291	30	31	39	0.04
Surgeon	91	21	25	54
General nephrologist	186	34	29	37	0.02
Transplant physician	196	22	30	48

Abbreviation: APOL1, Apolipoprotein L1

^aRefers to the number of respondents to each question.

^bFor each item, the top P value compares the distribution of responses between all nephrologists and surgeons. The bottom P value compares the distribution of responses between general nephrologists and transplant physicians (transplant nephrologists and surgeons).

Table 3.

Attitudes about APOL1 Genetic Testing.

	Na	Strongly Agree	Slightly Agree	Neutral	Slightly Disagree	Strongly Disagree	P b
		%	%	%	%	%
The transplant community should pay more attention to improving potential living kidney donors’ understanding of disparities in risk of kidney failure post-donation.	383	62	28	8	1	1
Nephrologist	292	67	26	6	1	0	<.002
Surgeon	91	49	31	17	1	2
General nephrologist	187	64	29	5	1	1	0.33
Transplant physician	196	61	25	11	2	1
APOL1 testing results could help African American potential living kidney donors make more informed decisions about donating.	380	50	37	10	3	0
Nephrologist	291	57	35	6	2	0	<.001
Surgeon	89	29	42	22	5	2
General nephrologist	186	62	32	5	1	0	<.001
Transplant physician	194	39	41	14	5	1
I believe that having 2 APOL1 risk variants will further increase the risk of getting end-stage renal disease among African American potential living kidney donors undergoing unilateral nephrectomy.	381	46	37	16	1	0
Nephrologist	291	52	34	13	1	0	<.001
Surgeon	90	27	45	27	1	0
General nephrologist	186	54	33	13	0	0	<.008
Transplant physician	195	39	41	18	2	0
I believe that APOL1 testing is beneficial for African American potential living kidney donors.	383	42	32	20	5	1
Nephrologist	292	49	32	13	5	1	<.001
Surgeon	91	21	30	44	4	1
General nephrologist	187	53	31	11	4	1	<.001
Transplant physician	196	33	33	28	6	0
Clinicians in my institution/clinical practice would conceptually support the idea of APOL1 testing.	382	35	39	22	3	1
Nephrologist	291	41	36	19	3	1	<.001
Surgeon	91	15	47	30	7	1
General nephrologist	187	43	33	22	2	0	<.001
Transplant physician	195	27	45	21	6	1
Compared to existing evaluation approaches for living donors, the addition of APOL1 testing will offer better clinical information about African American potential living donors’ eligibility for donation.	383	37	37	21	5	0
Nephrologist	292	42.8	36	16.8	4	0	<.001
Surgeon	91	20	38	34	6	2
General nephrologist	187	46	36	14	4	0	<.002
Transplant physician	196	30	37	27	5	1
APOL1 testing will reduce the number of African American potential living kidney donors.	380	24	49	22	4	1
Nephrologist	291	26	53	17	3	1	<.001
Surgeon	89	18	36	38	7	1
General nephrologist	186	23	57	15	3	2	<.006
Transplant physician	194	24	41	29	5	1
I know how to interpret APOL1 test results.	381	13	31	22	21	13
Nephrologist	290	15	37	21	18	9	<.001
Surgeon	91	7	12	26	32	23
General nephrologist	187	12	34	22	20	12	<.87
Transplant physician	194	13	29	23	22	13
I know enough about APOL1 testing to know the right clinical scenario in which to order APOL1 testing for African American potential living donors.	382	12	29	26	21	12
Nephrologist	291	12	35	25	19	9	<.001
Surgeon	91	10	11	29	28	22
General nephrologist	187	12	34	24	20	10	<.28
Transplant physician	195	11	25	27	23	14
My institution/clinical practice would financially support APOL1 testing.	379	16	22	46	9	7
Nephrologist	289	15	22	46	9	8	0.93
Surgeon	90	18	23	44	9	6
General nephrologist	186	12	24	50	8	6	0.21
Transplant physician	193	19	21	42	10	8
I feel prepared to counsel potential living kidney donors about the risks and benefits of APOL1 genetic testing.	381	12	26	22	23	17
Nephrologist	291	14	31	20	22	13	<.001
Surgeon	90	6	10	28	28	28
General nephrologist	186	12	30	19	23	16	0.42
Transplant physician	195	12	22	25	23	18
African American potential living donors with 2 APOL1 risk variants should not donate in order to reduce their own risk of getting kidney failure.	379	30	36	21	10	3
Nephrologist	288	31	37	18	11	3	0.03
Surgeon	91	24	32	33	7	4
General nephrologist	184	31	36	20	10	3	0.89
Transplant physician	195	28	36	23	9	4
If APOL1 were clearly linked to kidney graft survival, potential kidney recipients should accept kidneys from African American living donors with 2 APOL1 variants, rather than stay on dialysis.	379	23	27	24	16	10
Nephrologist	289	21	29	24	16	10	0.64
Surgeon	90	28	22	24	16	10
General nephrologist	186	22	28	26	15	9	0.84
Transplant physician	193	23	27	22	17	11
African American potential living donors with 2 APOL1 variants should not donate in order to prevent the potential recipient from getting a kidney with potentially shorter graft survival.	379	10	22	26	30	12
Nephrologist	290	11	25	24	30	10	0.06
Surgeon	89	9	14	29	30	18
General nephrologist	186	12	25	26	27	10	0.38
Transplant physician	193	9	20	25	33	13

Abbreviation: APOL1, Apolipoprotein L1

Note. Bolded p-values emphasize that they are statistically significant.

^aRefers to the number of respondents to each question.

Most reported that the addition of APOL1 testing to donor evaluation would offer better clinical information about AA donor’s eligibility for donation than existing evaluation approaches (74%). Significantly more nephrologists expressed agreement with these attitudes than surgeons/transplant physicians (Table 2).

More agreed (66%) than disagreed (13%) or were neutral (21%) that AA potential donors with 2 APOL1 risk variants should not donate to reduce their own risk of kidney failure; significantly more nephrologists agreed than surgeons (P = .03) (Table 3). One-third agreed that AA potential donors with 2 APOL1 risk variants should not donate to prevent potential recipients from getting a kidney with potentially shorter graft survival (32%) (Table 3). More respondents were unsure (43%) than opposed (29%) or supported (28%) AA donors who have already donated getting tested for APOL1 risk variants; significantly more surgeons/transplant physicians expressed uncertainty than nephrologists (P = .04, P = .02) (Table 3).

Potential recipient considerations.

Half reported that, if APOL1 risk variants were clearly linked to kidney graft survival, potential kidney recipients should accept kidneys from AA donors with 2 APOL1 risk variants rather than stay on dialysis (50%) (Table 3). More respondents believed that potential recipients should be informed about their donors’ APOL1 test results (43%) than were against it (27%) or unsure (30%); significantly more nephrologists expressed agreement than surgeons/transplant physicians (P = .002, P < .001) (Table 2).

Factors Affecting Adoption of APOL1 Testing

Leading barriers to using APOL1 testing were lack of professional guidelines on APOL1 testing (52%), prospective population data on donors (47%), randomized controlled trials on living donor outcomes associated with APOL1 risk variants (46%), and insurance coverage for APOL1 testing (42%) (Figure 3). Of 382 respondents, most (93%) would feel more comfortable offering APOL1 testing to AA potential donors if professional guidelines on APOL1 testing were available.

Figure 3.

Barriers to using Apolipoprotein L1 (APOL1) genetic testing.

More were neutral or did not know enough about APOL1 testing to know the right clinical scenario to order APOL1 testing (59%). Significantly more nephrologists than surgeons reported knowing the clinical scenario (P < .001) (Table 3). Of 382 respondents, most (97%) would use educational materials to counsel AA donors about APOL1 testing.

To adopt APOL1 testing into clinical practice, most of the 367 (56%) respondents minimally needed a 2-fold increase in relative risk of kidney failure attributed to APOL1 risk variants. To integrate APOL1 testing into routine donor evaluation, most of the 362 (61%) respondents required a minimum of 5% or 10% of the AA population to have 2 APOL1 risk variants.

Discussion

This is the first theoretically driven study assessing transplant and nephrology physicians’ attitudes toward and practices of integrating APOL1 genetic testing into clinical practice for AA potential donors. A key finding was the discrepancy between physicians’ attitudes and practices. Although physicians were highly supportive of APOL1 genetic testing for AA donors, few physicians use APOL1 testing routinely (4%) or on a case-by-case basis (14%). These rates correspond with innovators (3%) and early adopters (14%), respectively, who, according to Diffusion of Innovation theory, are more apt to adopt the innovation. Most physicians identified barriers to adopting APOL1 testing that appear to supersede their ethical norms in favor of APOL1 testing.

Unlike AST’s findings in which 13% of physicians use APOL1 testing, we found high levels of support for APOL1 testing, as most (63%) physicians plan to begin or continue using APOL1 testing in the next year. According to Roger’s theory, physicians are more likely to adopt APOL1 testing if they perceive it as advantageous. Most physicians reported that APOL1 testing conferred ethical and clinical benefits: improving donors’ informed consent, reducing donors’ risk of kidney failure, and providing physicians with more clinically relevant information for donor evaluation than is currently available. Nephrologists’ stronger perceptions of APOL1 testing than surgeons may derive from their long-term patient care and being more likely to see transplantations’ impact on patients and families.

Despite their support for APOL1, few physicians currently use APOL1 testing due to various barriers, including absence of professional guidelines, prospective data, and randomized controlled trials on APOL1 in AA donors, which coincide with theoretical requirements for innovations to be compatible with organizational needs. Although genetics guidelines address ethical issues in other clinical conditions,²⁸ available consensus guidelines on APOL1 are limited in scope due to lack of evidence providing insight into long-term outcomes.^20,21 Consequently, physicians are unclear as to how to use APOL1 test results in the context of living donation.

Perceptions about APOL1 may also affect physicians’ use of APOL1 testing. The majority of physicians (56%) believed the relative risk of kidney failure attributed to APOL1 risk variants would need to increase by at least 2 times to adopt APOL1 testing into clinical practice. Even though 13% of the AA general population has 2 APOL1 risk variants, most (61%) reported that a minimum of 5% or 10% of the AA population would need to have 2 risk variants to warrant integrating APOL1 testing into routine donor evaluation. Thus, one might expect higher rates of APOL1 testing.

Physicians shared an ethical concern that APOL1 testing would reduce the number of potential donors. Some physicians (13%) knew of or had participated in the care of an AA potential donor for whom APOL1 genetic testing was used. Positive test results dissuaded some donors from donating, as expected. Negative test results had little effect on donors’ decisions as most pursued donation.

Despite AST’s recommendations against APOL1 testing, some physicians are still using, and most intend to use APOL1 testing. Thus, education is necessary to guide physicians’ clinical practice. Almost all physicians reported they would use educational materials to counsel AA donors about APOL1 testing (97%). We are currently developing culturally targeted educational materials about APOL1 genetic testing for AA donors. Future research should assess whether educational materials increase physicians’ knowledge of APOL1 testing, and donor comprehension of APOL1 testing for informed consent.

The inconsistency between physicians’ attitudes and practices may highlight a moral tension between promoting donors’ informed decision-making and promoting living donation to help potential recipients, which can ameliorate disparities in AAs’ access to transplantation. Our finding that more physicians intend to use APOL1 testing than currently do suggests that other motivators affect adopting APOL1 testing, for example, ethical concerns about protecting donors from risk.

If APOL1 genetic testing becomes a more routine part of donors’ clinical care, addressing these moral tensions may become difficult to do and contribute to uneven care for potential donors. Therefore, broader professional debate about evaluating donors that addresses the ethical considerations of APOL1 testing is presently needed to proactively protect potential donors and potential recipients.

Our study’s strengths include a theoretically driven approach. Study limitations include a low participation rate. However, low participation rates may be less disconcerting in physician surveys than general public surveys because physicians are usually more homogenous in knowledge, attitudes, and practices.²⁹ Findings may not be generalizable to all AST, ASTS, and ASN physician members. Nonresponse bias may have affected the results: physicians with greater knowledge of APOL1 testing may have been more inclined to respond to the survey than those who did not. If applicable, this bias would partly explain differences between our and AST’s findings. The survey may have enabled cognitive biases through anchoring effects. A limitation may be physician’s lack of familiarity with and/or use of APOL1 genetic testing. Most general nephrologists surveyed evaluate donors, though the few (17%) who were not involved in evaluating donors might have affected the results. Providing a fact sheet may have skewed participants’ familiarity with APOL1 testing. Participants’ responses to questions may have differed according to their donor program’s volume, influence over evaluating donors, and patients’ and donors’ racial demographics.

APOL1 testing is at the early stages of diffusion. Although available circumstantial data are compelling, it is unclear how much APOL1 testing may provide useful prognostic information for evaluating AA potential donors.¹⁵ Genotyping may potentially personalize medical care through genetic assessment of ESRD risk for potential donors.³⁰ Physicians’ current and intended use of APOL1 testing highlights the need for an ethical framework and training on APOL1 testing. Professional guidelines and clinical decision support may assist physicians evaluating AA donors to clarify the proper use of APOL1 genetic testing.