Postoperative outcomes of lung transplant recipients with preformed donor-specific antibodies

Abstract

OBJECTIVES

Few studies have evaluated the outcomes of lung transplantation (LTx) in recipients with preformed donor-specific antibodies (DSAs). This study investigated the postoperative changes in preformed DSAs based on prospectively collected data of DSAs, and the influences of preformed DSAs on postoperative outcomes among LTx recipients.

METHODS

Between July 2010 and December 2019, 216 recipients underwent LTx (81 living-donor lobar lung transplants and 135 deceased-donor lung transplants). We reviewed 8 cases with preformed DSAs to determine postoperative changes in DSAs and compared postoperative outcomes between recipients with and without DSAs.

RESULTS

The preoperative mean fluorescence intensity of preformed DSAs ranged from 1141 to 14 695. Two recipients experienced antibody-mediated rejection within 2 weeks after LTx. DSAs disappeared in 7 recipients; however, 1 recipient experienced the relapse of DSAs and died from chronic lung allograft syndrome (CLAD), whereas 1 recipient had persisting DSAs within the study period and died from CLAD. Neither overall survival (OS) nor CLAD-free survival was significantly different between recipients with and without DSAs (P = 0.26 and P = 0.17, respectively). However, both OS and CLAD-free survival were significantly lower in recipients with DSAs against HLA class II than in those without these antibodies {5-year OS: 25.0% [95% confidence interval (CI): 0.9–66.5%] vs 72.1% (95% CI: 63.8–78.9%), P = 0.030 and 5-year CLAD-free survival: 26.7% (95% CI: 1.0–68.6%) vs 73.7% (95% CI: 66.5–79.5%), P = 0.002}.

CONCLUSIONS

Prognosis in recipients experiencing the relapse of preformed DSAs and those with persisting DSAs may be poor. The recipients with anti-HLA class II preformed DSAs had a significantly worse prognosis.

INTRODUCTION

Lung transplantation (LTx) is the final treatment option for patients with end-stage lung diseases [1, 2]. Since the Toronto group reported the first case of successful LTx in 1983 [3], the number of LTx cases has increased, currently reaching >69 000 [4]. Most LTx surgeries were performed in European and North American countries, and the number of such surgeries is increasing in these countries and in Asia [4].

In general, ABO-incompatible LTx or LTx in recipients whose preoperative complement-dependent cytotoxicity (CDC) crossmatch is positive has been rarely performed. In contrast, the number of LTx in recipients with preformed donor-specific antibodies (DSAs) is increasing, and few studies have evaluated the outcomes of LTx in these patients. Furthermore, the results are conflicting, given that 2 studies reported that the outcomes were worse in these recipients than in recipients without preformed DSAs [5, 6], whereas 3 studies found that the outcomes were similar between these 2 groups [7–9].

Since we experienced the first case of antibody-mediated rejection (AMR) after living-donor lobar lung transplantation (LDLLT) in 2010 [10, 11], preoperative and postoperative anti-HLA antibodies have been screened periodically. In this study, the postoperative changes in DSAs in recipients with preformed DSAs were reviewed based on prospectively collected data of DSAs, and the influences of preformed DSAs on the postoperative outcomes of LTx recipients were retrospectively investigated.

PATIENTS AND METHODS

Between July 2010 and December 2019, 216 LTx [81 LDLLTs and 135 deceased-donor lung transplantations (DDLTs)] were performed in Kyoto University Hospital. In this population, 8 recipients had preformed DSAs, and the remaining 208 recipients had no preformed DSAs. The cases with preformed DSAs were reviewed to determine the postoperative changes in DSAs and the frequency of AMR or chronic lung allograft dysfunction (CLAD). After that, the postoperative outcomes between recipients with or without preformed DSAs were compared. Since anti-HLA class II preformed DSAs were reported to have detrimental effects on postoperative outcomes [5, 6], the postoperative outcomes between recipients with or without these antibodies were also compared.

The observation period was defined as the interval between the date of LTx and the date of the last follow-up or death. The observation period of CLAD-free survival was defined as the date of LTx to the date of the last follow-up, CLAD, or death. Follow-up was censored in April 2020. This study was approved by the Kyoto University Institutional Review Board (YC1378). The requirement of patient informed consent was waived because of the retrospective nature of the study.

Anti-HLA antibody monitoring and immunosuppressive agents

Anti-HLA antibodies were prospectively screened periodically since July 2010 using the LABScreen Mixed (One Lambda, CA). Antibody screening was routinely performed before LTx, and at 1 week, 1 month, 3 months, 6 months and 1 year after LTx. After that, screening was performed annually. Anti-HLA antibodies were also measured when LTx recipients presented symptoms or abnormal findings. Anti-HLA antibodies were considered positive when the normalized mean fluorescence intensity (MFI) was >1000. Antibody specificity was determined using LABScreen Single Antigen (One Lambda).

The immunosuppressive agents used postoperatively were a calcineurin inhibitor (cyclosporine or tacrolimus), mycophenolate mofetil or azathioprine and prednisolone, as previously reported [12, 13]. CLAD was diagnosed on the basis of the definition in the consensus report published in 2019 [14]. Since a transbronchial lung biopsy is not routinely performed in our institution, the diagnosis of acute rejection is determined clinically based on the deterioration of respiratory condition and abnormal findings on a chest X-ray or computed tomography [13, 15]. The definition of AMR was based on the consensus report from the International Society for Heart and Lung Transplantation [16]. In this study, all suspected AMRs were clinically possible AMRs, which were diagnosed on the basis of the presence of DSAs and allograft dysfunction, following the exclusion of other causes, such as infection or congestive heart failure.

Eligibility criteria for LDLLT and DDLT in recipients with DSAs

In our institution, LDLLT is not recommended for potential LDLLT recipients with preformed DSAs; however, LDLLT is considered when MFI is <5000, and both CDC crossmatch and flow cytometry crossmatch are negative. In contrast, as a rule of DDLTs in Japan, lungs from brain-dead donors are allocated to recipients whose preoperative CDC crossmatch to donors is negative, and when the allocation is done, HLA typing of donors is performed at A, B and DR loci. When a potential DDLT recipient is preoperatively known to have preformed DSAs, LTx can be considered when the MFI is <5000. In LTx donors, HLA typing is performed in loci A, B, C, DP, DQ and DR in our institution, and anti-HLA antibodies in LTx recipients, including DSAs, are measured periodically (Fig. 1).

Figure 1:

Flow chart of patient recruitment, allocation and clinical management of lung transplant recipients with and without preformed donor-specific antibodies. AMR: antibody-mediated rejection; ATG: anti-thymoglobulin; CDC: complement-dependent cytotoxicity; CLAD: chronic lung allograft dysfunction; DDLT: deceased-donor lung transplantation; DSA: donor-specific antibody; IVIg: intravenous immunoglobulin; LDLLT: living-donor lobar lung transplantation; MFI: mean fluorescence intensity; PE: plasma exchange.

Statistical analyses

Descriptive statistics were obtained using EZR software version 1.33, a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria) [17]. Continuous variables were presented as medians with range, and categorical variables were expressed as percentages. The Mann–Whitney U-test and Fisher’s exact test were used for between-group analysis. Actuarial survival rates were calculated using the Kaplan–Meier method, and the groups were compared using a log-rank test. No tests were corrected for multiple testing. Statistical significance was defined as P < 0.05.

RESULTS

Patient characteristics

The population of LTx recipients included 111 males and 105 females. The median age was 43 years (range, 6–64 years), and the median body mass index was 18.3 kg/m² (range, 9.9–30.5 kg/m²). The most common indications for LTx were interstitial pneumonia (n = 95, 44.0%), followed by pulmonary complications after haematopoietic stem cell transplantation (n = 46, 21.3%). The median observation period was 1158 days (range, 14–3578 days).

Among the 8 recipients with preformed DSAs, age varied from 8 to 61 years, and body mass index ranged from 14.3 to 26.6 kg/m². This group included 2 males and 6 females, and their indications for LTx were comprised of interstitial pneumonia (n = 4), lymphangiomyomatosis (n = 2), idiopathic pulmonary arterial hypertension (n = 1) and cystic fibrosis (n = 1). This group underwent bilateral DDLT (3 patients), single DDLT (3 patients) or bilateral LDLLT (2 patients).

There were no significant differences in patient characteristics, including age, sex, total ischaemic time and operative methods, between recipients with and without preformed DSAs (Table 1).

Table 1:

Patient characteristics and comparison between lung transplant recipients with and without preformed donor-specific antibodies

Variables	Overall (n = 216)	Without DSAs (n = 208)	With DSAs (n = 8)	P-value
Age (years), median (range)	43 (6–64)	43 (6–64)	46.5 (8–61)	0.48
Sex (%), n (%)
Male	111 (51.4)	109 (52.4)	2 (25.0)	0.16
Female	105 (48.6)	99 (47.6)	6 (75.0)
Body mass index (kg/m2), median (range)	18.3 (9.9–30.5)	18.2 (9.9–30.5)	19.7 (14.3–26.6)	0.58
Indications for LTx, n (%)
Interstitial pneumonia	95 (44.0)	92 (44.2)	3 (37.5)	0.27
IPAH	19 (8.8)	18 (8.7)	1 (12.5)
LAM	12 (5.6)	10 (4.8)	2 (25.0)
COPD	13 (6.0)	13 (6.2)	0 (0.0)
Others	31 (14.4)	30 (14.4)	1 (12.5)
Pulmonary complications after HSCT	46 (21.3)	45 (21.6)	1 (12.5)
Total ischaemic time (min), median (range)	395.5 (84–780)	395.5 (84–780)	400 (177–606)	0.83
Operative method (%), n (%)
Bilateral DDLT	67 (31.0)	64 (30.8)	3 (37.5)	1.0
Bilateral LDLLT	60 (27.8)	58 (27.9)	2 (25.0)
LDLLT with additional procedures	11 (5.1)	11 (5.3)	0 (0.0)
Single DDLT	68 (31.5)	65 (31.2)	3 (37.5)
Single LDLLT	10 (4.6)	10 (4.8)	0 (0.0)

COPD: chronic obstructive pulmonary disease; DDLT: deceased-donor lung transplantation; DSAs: donor-specific antibodies; HSCT: haematopoietic stem cell transplantation; IPAH: idiopathic pulmonary arterial hypertension; IQR: interquartile range; LAM: lymphangiomyomatosis; LDLLT: living-donor lobar lung transplantation; LTx: lung transplantation.

Preoperative and postoperative changes in preformed DSA and frequency of postoperative AMR among recipients with preformed DSAs

The number of times anti-HLA antibodies were measured preoperatively varied from 1 to 5. All recipients had preformed DSAs when receiving LTx. Preformed DSAs were detected 5 times in 1 recipient (case 6) and once in 7 recipients (Table 2). In 4 out of these 7 recipients, having preformed DSAs was postoperatively detected from the immediate pretransplant serum. Two recipients had anti-HLA class I preformed DSAs, whereas the remaining 6 recipients had anti-HLA class II preformed DSAs. The MFI of preformed DSAs in the last test varied from 1141 to 14 695. One recipient (case 6) received plasma exchange (PE) preoperatively because she was known to have a preformed DSA before LTx (Fig. 2). Two LDLLT recipients with DSAs did not receive preoperative treatments because both CDC crossmatch and flow cytometry crossmatch were negative. The remaining 5 DDLT recipients did not receive preoperative treatment either because they were found to have DSAs a few days after LTx.

Table 2:

Preoperative characteristics of recipients with preformed donor-specific antibodies

No.	Age (years)	Sex	BMI (kg/m2)	Indication for LTx	Type of transplant	Number of times that anti-HLA antibodies were measured preoperatively	Number of times that DSAs were positive	MFI of preformed DSAs before LTx in the last test	Type of DSAs	Locus of preformed DSA	Additional preoperative management
1	52	Female	20.0	Interstitial pneumonia	LDLLT (B)	1	1	1633	Class II	DR14	None
2	41	Female	16.4	Lymphangiomyomatosis	DDLT (S)	1	1	6095	Class I	B54	None
3	57	Female	14.3	Pulmonary complications after haematopoietic stem-cell transplantation	LDLLT (B)	1	1	1720	Class II	DR13	None
4	8	Male	15.2	Cystic fibrosis	DDLT (B)	2	1	3960	Class II	DQ8	None
5	54	Male	26.6	Interstitial pneumonia	DDLT (S)	2	1	14 695	Class II	DQ7	None
6	41	Female	22.2	Lymphangiomyomatosis	DDLT (B)	5	5	3334	Class I	B35	Plasma exchange
7	61	Female	19.4	Interstitial pneumonia	DDLT (S)	3	1	1141	Class II	DQ4	None
8	38	Female	20.4	Idiopathic pulmonary arterial hypertension	DDLT (B)	2	1	1345	Class II	DR4	None

BMI: body mass index; DDLT: deceased-donor lung transplantation; DSAs: donor-specific antibodies; LDLLT: living-donor lobar lung transplantation; LTx: lung transplantation; MFI: mean fluorescence intensity.

Figure 2:

A representative case involving a 41-year-old female recipient (case 6) undergoing bilateral deceased-donor lung transplantation because of lymphangiomyomatosis. Anti-HLA antibodies were measured periodically before lung transplantation because she had multiple anti-HLA antibodies. Since the complement-dependent cytotoxicity crossmatch was negative despite the presence of anti-HLA class I donor-specific antibodies (B35), she underwent bilateral deceased-donor lung transplantation after receiving preoperative plasma exchange. The postoperative course was uneventful, and she was discharged without developing antibody-mediated rejection. Preformed donor-specific antibodies disappeared 1 month after lung transplantation, and the patient had no chronic lung allograft dysfunction after lung transplantation during the 26-month follow-up. LTx: lung transplantation; PE: plasma exchange.

AMR requiring treatment within 1 month after LTx occurred in 2 recipients (cases 5 and 8). One recipient (case 5) was treated with steroid pulse and intravenous immunoglobulin (IVIg), and the other recipient (case 8) was treated with PE, anti-thymoglobulin, steroid pulse and IVIg because her allograft dysfunction was severe. Additional treatments were not performed perioperatively for the remaining 6 recipients because they did not develop perioperative AMR.

Postoperatively, DSAs disappeared within the observation period in 7 recipients; however, 1 of these 7 patients (case 5) experienced the relapse of the preformed DSA and eventually developed AMR and CLAD (Fig. 3). In 1 recipient (case 4), his DSA did not disappear during the observation period (Table 3). During this period, CLAD occurred in 4 recipients, 3 of whom died from CLAD. The comparison of recipients with preformed DSAs who developed CLAD with those who did not develop CLAD is shown in Supplementary Material, Table S1.

Figure 3:

A representative case involving a 54-year-old male recipient (case 5) undergoing right single deceased-donor lung transplantation because of interstitial pneumonia. It was detected postoperatively that he had preformed donor-specific antibodies from the immediate pretransplant serum. The patient developed acute antibody-mediated rejection on postoperative day 6 and was treated with methylprednisolone pulse therapy for 3 days and intravenous immunoglobulin for 2 days. After treatment, he recovered from antibody-mediated rejection and was discharged and preformed donor-specific antibodies disappeared (Fig. 2). However, preformed donor-specific antibodies reappeared 4 months after lung transplantation. Although being asymptomatic, the patient was treated with intravenous immunoglobulin monthly for 6 months; however, donor-specific antibodies persisted. He developed chronic lung allograft dysfunction associated with antibody-mediated rejection at 18 months after lung transplantation and died from chronic lung allograft dysfunction at 25 months after lung transplantation, despite intensive treatment. AMR: antibody-mediated rejection; CLAD: chronic lung allograft dysfunction; IVIg: intravenous immunoglobulin; LTx: lung transplantation; mPSL: methylprednisolone; PE: plasma exchange.

Table 3:

Postoperative outcomes of recipients with preformed donor-specific antibodies

No.	Allograft ischaemic time (min)	PGD within 72 hours	Immunosuppressant agents	AMR within 1 month after LTx	Management of AMR	Postoperative status of preformed DSAs	Management of persistent or relapse of preformed DSA	Prognosis
1	196	0	FK + MMF + PSL	No	No	Disappeared at 1 month after LTx	No	Alive at 113 months after LTx
2	412	1	FK + MMF + PSL	No	No	Disappeared at 6 months after LTx	No	Alive at 108 months after LTx
3	177	0	FK + MMF + PSL	No	No	Disappeared at 1 month after LTx	No	Dead at 17 months after LTx due to CLAD
4	390	0	CyA + MMF + PSL	No	No	Not disappeared	mPSL pulse, IVIg, ATG, rituximab for AMR	Dead at 12 months after LTx due to CLAD
5	242	3	CyA + MMF + PSL	On postoperative day 6	mPSL pulse, IVIg	Disappeared but relapsed	mPSL pulse, plasma exchange, IVIg, rituximab for AMR	Dead at 25 months after LTx due to CLAD
6	522	1	FK + MMF + PSL	No	No	Disappeared at 1 month after LTx	No	Alive at 26 months after LTx
7	410	0	FK + MMF + PSL	No	No	Disappeared at 3 months after LTx	No	Alive at 9 months after LTx
8	606	3	FK + MMF + PSL	On postoperative day 8	Plasma exchange, ATG, mPSL pulse, IVIg	Disappeared at 5 months after LTx	No	Alive at 5 months after LTx

AMR: antibody-mediated rejection; ATG: anti-thymoglobulin; CLAD: chronic lung allograft dysfunction; CyA: cyclosporine; DSA: donor-specific antibody; FK: tacrolimus; IVIg: intravenous immunoglobulin; LTx: lung transplantation; MMF: mycophenolate mofetil; mPSL: methylprednisolone; PGD: primary graft dysfunction; PSL: prednisolone.

Comparison of the outcomes between recipients with and without preformed DSAs

The actuarial 5-year overall survival (OS) was non-significantly lower in recipients with preformed DSAs than in recipients without preformed DSAs {50.0% [95% confidence interval (CI), 11.1–80.4%] vs 71.9% (95% CI, 63.5–78.6%); P = 0.26} (Fig. 4A). Moreover, the 5-year CLAD-free survival rate was non-significantly lower in recipients with preformed DSAs than in patients without preformed DSAs [25.7% (95% CI, 1.3–65.6%) vs 61.9% (95% CI, 53.0–69.5%); P = 0.17] (Fig. 4B).

Figure 4:

(A) Actuarial 5-year overall survival rate was non-significantly lower in lung transplant recipients with preformed donor-specific antibodies than in recipients without preformed donor-specific antibodies [50.0% (95% confidence interval, 11.1–80.4%) vs 71.9% (95% confidence interval, 63.5–78.6%); P = 0.26]. (B) The actuarial 5-year chronic lung allograft syndrome-free survival rate was non-significantly lower in recipients with preformed donor-specific antibodies than in recipients without donor-specific antibodies [25.7% (95% confidence interval, 1.3–65.6%) vs 61.9% (95% confidence interval, 53.0–69.5%); P = 0.17). DSA: donor-specific antibody; CLAD: chronic lung allograft syndrome; LTx: lung transplantation.

We also compared the outcomes between recipients with (n = 6) and without (n = 210) anti-HLA class II preformed DSAs. The 5-year OS was significantly lower in recipients with preformed DSAs to these antigens than in patients without antibodies against these antigens [25.0% (95% CI: 0.9–66.5%) vs 72.1% (95% CI, 63.8–78.9%); P = 0.030] (Fig. 5A). Similarly, 3-year CLAD-free survival was significantly lower in recipients with anti-HLA class II preformed DSAs than in patients without these antibodies [26.7% (95% CI, 1.0–68.6%) vs 73.7% (95% CI, 66.5–79.5%); P = 0.002] (Fig. 5B). The median CLAD-free survival time in recipients with or without anti-HLA class II preformed DSAs was 19 and 89 months, respectively.

Figure 5:

(A) Actuarial 5-year overall survival rate was significantly lower in lung transplant recipients with anti-HLA class II preformed donor-specific antibodies than in patients without these antibodies [25.0% (95% confidence interval, 0.9–66.5%) vs 72.1% (95% confidence interval, 63.8–78.9%); P = 0.030]. (B) The 3-year chronic lung allograft syndrome -free survival rate was significantly lower in recipients with anti-HLA class II preformed donor-specific antibodies than in patients without these antibodies [26.7% (95% confidence interval, 1.0–68.6%) vs 73.7% (95% confidence interval, 66.5–79.5%); P = 0.002]. CLAD: chronic lung allograft syndrome; LTx: lung transplantation.

DISCUSSION

In our study, the results showed that 25% of LTx recipients with preformed DSAs developed AMR within 2 weeks after LTx, and CLAD occurred in 50% of recipients with preformed DSAs. Preformed DSAs disappeared in 87.5% of our cohort. However, 1 patient experienced the relapse of DSAs, developed CLAD associated with AMR and died from CLAD. Notably, both OS and CLAD-free survival in recipients with anti-HLA class II preformed DSAs were significantly lower than those in recipients without antibodies against these antigens.

Although there was no perioperative death in our cohort, AMR occurred in 2 recipients within 2 weeks after LTx, and CLAD occurred in 4 patients. Since primary graft dysfunction grade 3 within 72 hours after LTx was detected in these 2 recipients (Table 3), it was hypothesized that the expression of HLA antigens in lung grafts was increased because of ischaemia–reperfusion injury, which could cause AMR. AMR in these 2 recipients was fortunately improved using multimodal therapy; however, the use of rituximab, PE and IVIg varies between institutions because an effective treatment has not been established [13, 16, 18, 19]. It is desired to determine the effective treatments for AMR as early as possible.

Little is known about changes in preformed DSAs after transplantation and the influences of these changes on prognosis. The reported postoperative clearance rate of preformed DSAs after LTx was 33.3–71.4% [7–9]; however, the time and frequency of measurements of anti-HLA antibodies differ across institutions, and no studies have made long-term measurements of these antibodies. In turn, anti-HLA antibodies are measured periodically after LTx in our department. In this study, DSAs disappeared in 7 recipients (87.5%) after LTx within the study period; however, 1 of these recipients experienced the relapse of the preformed DSA. This recipient and the patient with persisting DSAs developed CLAD associated with AMR and died from CLAD, suggesting that the prognosis of recipients with preformed DSAs can be a cause for concern, and detecting such cases by periodical measurement of anti-HLA antibodies postoperatively is crucial.

Our results showed that the prognosis of recipients with preformed DSAs seemed to be worse than that of recipients without these antibodies, although the difference was not statistically significant. Moreover, the prognosis of recipients with anti-HLA class II preformed DSAs was significantly worse than that of patients without these antibodies, which agrees with previous studies [5, 6]. The reasons for these results remain unclear. Previous studies found that de novo DSAs against HLA class II antigens were a prognostic factor [20, 21], and HLA class II expression was higher in the lung tissue of patients with infections or acute AMR than in normal lung tissue [22–26]. These results may explain the dysfunction of transplanted lungs in recipients with anti-HLA class II preformed DSAs; nonetheless, further studies are required to confirm these findings.

Limitations

The present study has limitations. First, this study had a retrospective, nonrandomized, single-center design, although data on anti-HLA antibodies were collected prospectively. In addition, the number of recipients with preformed DSAs was relatively small. Therefore, multicentre studies with larger cohorts are necessary. Second, there remains uncertainty regarding the clinical relevance of preformed DSAs because there are no common immunological risk assessment strategies, such as variable MFI thresholds. Third, the number of recipients undergoing LDLLT was relatively large. Therefore, the backgrounds and characteristics of the study participants may differ from those of other studies. However, since no studies have measured DSAs both before and after LTx in recipients undergoing LDLLT, a strength of this study was the large amount of data on LDLLT recipients. In contrast, in countries in which the number of deceased donors is small, as in Japan, LDLLT is a useful option [27, 28]; however, increasing the number of deceased donors is essential.

CONCLUSION

We performed LTx for 8 recipients with preformed DSAs. Despite the relatively small number of recipients, prognosis in recipients experiencing the relapse of preformed DSAs and those with persisting DSAs might be poor. Furthermore, the recipients with anti-HLA class II preformed DSAs had a significantly worse prognosis.

SUPPLEMENTARY MATERIAL

Supplementary material is available at ICVTS online.

ABBREVIATIONS

AMR

Antibody-mediated rejection

CDC

Complement-dependent cytotoxicity

CI

Confidence interval

CLAD

Chronic lung allograft dysfunction

DDLT

Deceased-donor lung transplantation

DSA

Donor-specific antibody

IVIg

Intravenous immunoglobulin

LDLLT

Living-donor lobar lung transplantation

LTx

Lung transplantation

MFI

Mean fluorescence intensity

PE

Plasma exchange