Predicting kidney allograft survival with explainable machine learning.

INTRODUCTION: Despite significant progress over the last decades in the survival of kidney allografts, several risk factors remain contributing to worsening kidney function or even loss of transplants. We aimed to evaluate a new machine learning method to identify these variables which may predict the early graft loss in kidney transplant patients and to assess their usefulness for improving clinical decisions. MATERIAL AND METHODS: A retrospective cohort study was carried out with 627 kidney transplant patients followed at least three months. All these data were pre-processed, and their selected features were used to develop an automatically working a machine learning algorithm; this algorithm was then applied for training and parameterization of the model; and finally, the tested model was then used for the analysis of patients' features that were the most impactful for the prediction of clinical outcomes. Our models were evaluated using the Area Under the Curve (AUC), and the SHapley Additive exPlanations (SHAP) algorithm was used to interpret its predictions. RESULTS: The final selected model achieved a precision of 0.81, a sensitivity of 0.61, a specificity of 0.89, and an AUC value of 0.84. In our model, serum creatinine levels of kidney transplant patients, evaluated at the hospital discharge, proved to be the most important factor in the decision-making for the allograft loss. Patients with a weight equivalent to a BMI closer to the normal range prior to a kidney transplant are less likely to experience graft loss compared to patients with a BMI below the normal range. The age of patients at transplantation and Polyomavirus (BKPyV) infection had significant impact on clinical outcomes in our model. CONCLUSIONS: Our algorithm suggests that the main characteristics that impacted early allograft loss were serum creatinine levels at the hospital discharge, as well as the pre-transplant values such as body weight, age of patients, and their BKPyV infection. We propose that machine learning tools can be developed to effectively assist medical decision-making in kidney transplantation.

Impact of COVID-19 vaccination on clinical outcomes in kidney transplant patients.

INTRODUCTION: The global health crisis caused by the COVID-19 pandemic has resulted in severe mortality and morbidity. Immunosuppressed patients, such as kidney transplant recipients, are particularly susceptible to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections. OBJECTIVE: The aim of this cohort study was to evaluate the impact of COVID-19 vaccination on clinical outcomes in patients with kidney transplants. MATERIALS AND METHODS: In this retrospective study, 254 patients with kidney transplants were vaccinated against SARS-CoV-2 and a fraction of these contracted COVID-19. The diagnosis of COVID-19 was carried out by reverse transcriptase-polymerase chain reaction testing, and the patients received treatment involving immunosuppressive and COVID-19-specific protocols. RESULTS: SARS-CoV-2 infection was diagnosed in 38 (14.96%) patients before the COVID-19 vaccine was administered. After vaccination, an additional 29 (11.42%) patients were diagnosed with COVID-19. Risk factors for hospitalization included age, body mass index (BMI), comorbidities, and time elapsed since renal transplantation (p = 0.025, 0.038, 0.012, and 0.046, respectively). COVID-19 vaccination resulted in a significant decrease in the rate of hospital-acquired SARS-CoV-2 infection from 63.16% to 34.48% (p = 0.020). The proportion of patients from this cohort placed in intensive care units decreased from 23.68% to zero. Allograft rejections exhibited a decreasing trend from 13.16% to 6.90% (p = 0.690). This patient cohort displayed 15.79% mortality prior to COVID-19 vaccination that was reduced to nil after immunization. CONCLUSION: COVID-19 vaccination significantly reduced COVID-19 severity and mortality in this cohort of patients with kidney transplants. The risk factors for hospitalization were determined to be age, BMI, comorbidities, and time since renal transplantation. COVID-19 vaccination resulted in a clinical outcome of reduced hospitalization and a decrease in clinical complications. The COVID-19 vaccination-derived adverse effects in this cohort were found to be comparable to those in the immunocompetent population.

Allogeneic hematopoietic stem cell transplantation: A single-center experience, 2017 to 2021.

INTRODUCTION: Hematopoietic stem cell transplantation (HSCT) remains a critical treatment for advanced or high-risk hematological malignancies, posing challenges such as finding suitable donors and managing of graft-versus-host disease (GvHD). This study estimates 3-year overall survival in patients who underwent HSCT at our referral service in the state of Minas Gerais, Brazil. MATERIAL AND METHODS: This retrospective observational cohort study involved 41 patients who received HSCT between 2017 and 2021 at the Felício Rocho Hospital. Recipients received HSCT from either haploidentical donor (Haplo), matched unrelated donor (MUD), or HLA-matched sibling donor (MSD). The study evaluated parameters that included 3-year overall survival (OS), treatment-related mortality (TRM), GvHD incidence, post-transplant relapse rate, and engraftment. ANOVA, Kruskal-Wallis, and chi-square tests were used for statistical analysis. Survival curves were calculated using the Kaplan-Meier method and the Log-rank test compared the curves. RESULTS: Our study found that the engraftment time differed among groups: Haplo recipients engrafted earlier within a median of 16 days (ranging between 10 and 20 days) than MSD recipients with 18 days (ranging between 11 and 28 days), and MUD recipients with 19 days (ranging between 11 and 24 days; p = 0.019). Mild acute GvHD (grade I-II) was observed in 13 patients, progressing to chronic GvHD in 5 patients. Three-year OS rates were as follows: MSD group - 67.7%, Haplo group - 42.2%, and MUD group - 44.4% (MSD vs Haplo, p = 0.039). Three-year cumulative treatment-related mortality (TRM) rates were 17.8% for MSD group, 22.9% for Haplo group, and 22.1% for MUD group (pairwise comparisons p > 0.05). Infection-related mortality was reported in eight patients, while relapse rates at 3 years were similar across MSD, Haplo, and MUD groups (p = 0.891). Donor age influenced OS rates, showing better outcomes with donors under 45 years old, and significant differences were found in pairwise comparisons (p = 0.015). CONCLUSION: Donor type and donor age significantly impacted HSCT outcomes in our analysis, thus emphasizing the importance of rigorous donor selection in risk stratification and suggesting potential benefits for younger donors.

Remarkable 107-year-old kidney with a 49-year of long-term allograph survival through continuous azathioprine monotherapy.

BACKGROUND: The main goal of kidney allograft transplantation is to improve survival in patients with end-stage kidney failure. Herein, we report a 49-year long-term allograft survival with non-identical human leukocyte antigens (HLA). The purpose of this study was to report the successful clinical outcome of 49 years of transplant survival in a 79-year-old patient with a 107-year-old kidney undergoing continued immunosuppressive monotherapy. MATERIAL AND METHODS: The patient was evaluated clinically and immunologically with HLA typing and anti-HLA antibodies before transplantation. Post-transplant, the patient's clinical and immunological survival were monitored for 49 years. The state of the chimerism was assessed using the polymerase chain reaction to amplify 24 short tandem repeats using a DNA thermocycler and DNA analyzer. RESULTS: The patient and donor were haploidentical and the patient was treated with azathioprine monotherapy. Donor-specific antibodies were detected only for the HLA-DPB1* 03:01 mismatch. This patient developed multiple skin tumors 26 years after transplant, which were successfully treated with topical therapy or surgical removal. The patient developed an intestinal adenocarcinoma 43 years after kidney transplantation, which was surgically removal; six years later, adenocarcinoma was diagnosed in a finger, followed by axillar and hepatic metastases. After 49 years of graft survival of a kidney of 107 years old in a patient with 79 years of age, the patient's health worsened with severe dehydration, anemia, and bacterial infection. The patient was hospitalized with a serum creatinine level of 3.45 mg/dL, urea level of 188 mg/dL, and estimated glomerular filtration rate of 22 mL/1.72 m2; septicemia developed and was treated with antibiotics. The patient had poor clinical progress, was intubated, and later died due to septic shock. CONCLUSIONS: To the best of our knowledge, this is the first case of a 107-year-old kidney, transplanted into a recipient who was treated with azathioprine monotherapy for 49 years.

Impact of early blood transfusion after kidney transplantation on the clinical outcomes and allograft survival.

INTRODUCTION: Anemia in chronic kidney disease is of great concern regarding blood transfusions and the possibility of allosensitization for future kidney transplants and the occurrence of rejection and allograft loss in the post-transplant period. The aim of this study was to evaluate the effect of early blood transfusion on the occurrence of rejections, allograft function and survival in the first year after transplantation. MATERIAL AND METHODS: This retrospective study was carried out with 445 patients submitted to kidney transplant allocated to two groups. The first group received early blood transfusions after transplant (n = 125, 28.09%), and the second group did not receive blood transfusions (n = 320, 71.91%). The patient outcomes were evaluated during a 1-year follow-up. RESULTS: 14 patients given blood transfusion (11.2%) lost their allograft in the first year in comparison with 8 (2.5%) without transfusion (p < 0.001). There were 9 deaths in each group, which corresponded to 7.2% of the patients who received blood transfusions and 2.81% of those who did not (p < 0.035). Patient hospitalization lasted 15 days in transfusion group and 8.5 days in non-transfusion group (p < 0.001). Creatinine levels were higher in the patients who received blood transfusion than in those without transfusion in the first and third months after transplantation (p = 0.012 and 0.038, respectively). During the first year, the patients who received blood products experienced more antibody-mediated rejection (ABMR) (13.60%) than patients who did not (4.38%) (p < 0.001). Those who received blood transfusions also developed de novo DSA in higher proportion than those without transfusion against both class I and class II HLA (p < 0.001). CONCLUSION: This study showed that blood transfusions in the first month after transplantation had a negative impact on kidney function, graft survival, and contributed to the development of de novo DSA, an increased risk of ABMR and infections.

Effects of Bacterial Urinary Tract Infection on Clinical Outcome and Survival of Kidney Transplant Patients.

BACKGROUND: There is no consensus on whether the development of urinary tract infections (UTIs) leads to high mortality or graft loss in kidney transplant patients. A high incidence of multidrug resistant (MDR) infections was observed worldwide and is associated with these complications. The aim of this study was to analyze the effects of UTIs on the clinical outcome and survival in kidney transplant patients. METHODS: This retrospective study evaluated 601 kidney transplant patients who were categorized as follows: group 1 (G1) patients without a UTI, group 2 (G2) patients with a UTI, and group 3 (G3) recipients with a recurrent UTI. Patients were followed up for at least 1 year after transplantation. Graft survival, risk of graft loss, and risk of developing a UTI were analyzed by the Kaplan-Meier method, Cox regression, and logistic regression methods, respectively. Differences with P < .05 were considered statistically significant. RESULTS: The proportion of rejection episodes was higher in G3 (32.35%) than in G1 (20.89%) and G2 (21.88%) (P < .001). The graft survival after the 10-year follow-up was better in G1 (73.29%) than in G3 (61.62%) (P = .019). UTI recurrent episodes increased the risk of graft loss >2.5-fold. Women and those who received a kidney from a deceased donor (DD) were at risk of at least 1 UTI event during follow-up. A greater proportion of MDR infections was observed in G3 than in G2 (P < .001). CONCLUSIONS: The risk factors for developing a UTI were female sex, receiving a DD kidney, susceptibility to other infections, episodes of rejection, and delayed graft function. Moreover, a UTI, especially a recurrent UTI, was an important risk factor for allograft loss.

Malignancy Diseases in Kidney Transplantation, Clinical Outcomes, Patient, and Allograft Survival: A Case-Control Study.

BACKGROUND: Malignancy is a well-known complication in patients after kidney transplantation (KT), but its effect on posttransplant outcomes, allograft, and patient survival remains unexplored. The aim of this study is to report the impact of the comorbidity on clinical outcome, function, and failure of an allograft kidney. METHODS: This case-control study included 101 KT patients. Twenty-six patients who developed cancer (CA) were assigned to the case group and 75 to the control group. Statistical analysis was performed using logistic regression models, and graft survival was analyzed using the Kaplan-Meier curve. RESULTS: Non-melanoma skin CA was the most common malignancy, accounting for almost 60% of cases, followed by stomach CA, prostate CA, and lymphoproliferative diseases (7.70% each). Difference in graft and patient survival was not significant between the two groups (P > .05). A tumor in nonfunctioning in the first nonfunctioning KT was identified in 1 KT patient with a second allograft and by anatomopathological was detect Fuhrman grade II renal cell carcinoma. This KT patient was in good clinical condition with serum creatinine level of 1.5 mg/dL. CONCLUSIONS: No association was observed between CA development and risk factors, including family history and smoking habit, and no differences in allograft and patient survival were found. Nevertheless, in our data, CA in KT patients occurred early after transplantation. Renal cell carcinoma in allograft failure was identified in a patient; that suggested that nephrectomy of kidney failure must be performed to avoid patient allosensitization and neoplasia. Thus, we suggest continuous screening of malignancy diseases for KT patients.

Five novel HLA-A, -B, and -C alleles identified in Brazilian individuals by next-generation sequencing.

The novel HLA-A*11:379, B*45:01:11, B*15:571, B*57:137, C*07:893 alleles were identified in Brazilian individuals.

The novel HLA-DRB1*03:178, -DRB1*03:179, and -DRB1*11:276 alleles identified in a healthy Brazilian individuals.

The novel HLA-DRB1*03:178, -DRB1*03:179, -DRB1*11:276 alleles were identified in healthy Brazilian individuals.

Impact of the immunotherapy induction on allograft outcome and survival in kidney transplant patients with donor-specific antibodies to HLA-DQB1.

BACKGROUND: The presence of donor-specific antibodies (DSAs) against HLA-DQB1 is considered a significant barrier to good outcome and allograft survival in kidney transplantation (KT). This study aimed to assess the impact of induction immunotherapy on the outcome and allograft survival in KT patients with HLA-DQB1-DSA. METHODOLOGY: Thirty-two patients who had undergone KT and found to be positive for HLA-DQB1-DSA were monitored at least one to 10 years. They were allocated into two groups of patients: G1 received induction immunotherapy (n = 14 patients; 43.75%), and G2 did not (n = 18 patients; 56.25%). RESULTS: In G1, 6 (42.86%) patients experienced rejection episodes (RE), 2 (14.29%) due to antibody-mediated rejection (ABMR) and 4 (28.57%) due to T-cell-mediated rejection (TCMR). In G2, 13 (72.22%) patients experienced RE, 3 (16.67%) due to ABMR, and 10 (55.56%) due to TCMR. Graft loss occurred in 4 patients from G1, 2 (14.29%) due to ABMR and 2 (14.29%) due to non-immunological causes. In G2, 9 (50.00%) patients lost their grafts, 2 (11.11%) due to TCMR, 2 (11.11%) due to ABMR, and 5 (27.78%) due to non-immunological causes. The graft survival rate was 64.29% in G1 and 45.83% in G2. Glomerulitis and peritubular capillaritis were observed in 3 and C4d-positive patients with/or without induction who lost their grafts by ABMR by HLA-DQ DSA. Two patients from G2 lost their graft by TCMR due to interstitial lymphocytic infiltrate (i1), foci of mild tubulitis (t2), interstitial edema, moderate interstitial fibrosis and tubular atrophy. Better graft survival rates were shown in patients from G1 who received induction immunotherapy. CONCLUSION: Our study suggests that patients with an immunological profile of HLA-DQ+ DSA+ treated by immunotherapy induction have a decreased risk of ABMR and increased allograft survival, and the presence of anti-HLA-DQB1 DSA+ detected before and after KT were associated with ABMR episodes and failure.

Characterization of 15 novel HLA alleles by next generation sequencing in Brazilian individuals.

Fifteen novel HLA alleles in the HLA-A, -B, -C, DRB1, and DQB1 loci were described.

Two novel HLA-DRB1 alleles, DRB1*11:261 and DRB1*13:286 identified by sequencing in Brazilian individuals.

Two novel HLA alleles DRB1*11:261 and DRB1*13:286 have nonsynonymous mutations in exon 2.

Effect of Glomerulopathy Recurrence in the Outcome and Graft Survival of Kidney Transplanted Patients.

BACKGROUND: Recurrent glomerulopathy (GP) after kidney transplantation is a complication of kidney transplantation that could negatively affect kidney function and graft survival. This study aimed to evaluate the outcome, graft survival, and GP recurrence and its predictive factors in kidney-transplanted patients. METHODS: Patients were divided into 2 groups: G1 (with GP; n = 95) and G2 (with other causes of end-stage renal disease; n = 373). Graft survival analyses were performed using the Kaplan-Meier for living donor (LD) and deceased donor (DD). Cox proportional hazards regression were used to investigate the predictors for graft loss and for GP recurrence. RESULTS: Disease recurrence was observed in 9 patients who received a kidney from an LD, of which 4 lost their grafts. In patients who received a kidney from a DD, recurrence was also observed in 9 patients, of which 3 lost their grafts. No statistically significant differences in graft survival between G1 and G2 in relation to LD and DD were noted (P = .299 and .434, respectively). However, differences in graft survival were found when GP subtypes and GP recurrence were analyzed. The predictors of graft loss were delayed graft function (hazard ratio [HR] = 2.226, P = .002), rejection episodes (HR = 1.904, P = .017), and recurrence or transplant GP (HR = 3.243, P = .006). The predictors of disease recurrence or transplant GP were age (HR = 0.945, P = .028) and cold ischemia time (HR = 1.117, P = .003). CONCLUSION: Kidney transplantation could be a reasonable treatment for GP with end-stage renal disease. Despite the disease recurrence, which is a significant cause of graft loss in transplant recipients, graft survival remains satisfactory.

Next-generation sequencing of HLA: validation and identification of new polymorphisms in a Brazilian population.

BACKGROUND: Next-generation sequencing (NGS) is the most modern sequencing technique that has revolutionized HLA typing, providing high-resolution results with low ambiguity rates. This study aimed to show the experiences and challenges of an HLA laboratory in the validation process of the NGS methodology for HLA typing and show the use of this method for the study of HLA genetic diversity. METHODS: We used 115 samples that comprised a comprehensive testing panel for validation of the NGS methodology using the AllType kit (One Lambda, Canoga Park, California) on the Ion Torrent S5 NGS platform. All quality metrics were analyzed. During validation, two new HLA sequences were identified and named by the HLA Nomenclature Committee. RESULTS: A total of 1380 alleles from the HLA-A, -B, -C, -DRB1, -DQB1, and -DPB1 loci were examined by NGS. This validation panel provided a wide range of HLA sequence variations, including non-CWD HLA alleles, new variants, and homozygous alleles. The concordance rate with Sanger sequencing-based typing was 100.0% for HLA-A, -B, -C, -DRB1, -DQB1, and 99.93% for HLA-DPB1. The newly identified HLA alleles were HLA-B*14:69N and HLA-DQB1*02:145. CONCLUSION: We have successfully validated NGS HLA typing despite numerous challenges, contributing to the identification of novel alleles that impact on HLA matching and antibody evaluation in organ and tissue transplantation.

Effects of immunotherapy induction on outcome and graft survival of kidney-transplanted patients with different immunological risk of rejection.

BACKGROUND: In kidney transplantation, immunotherapy with thymoglobulin (rATG) has been used to down-regulate the patient immune system. rATG is a powerful immunobiologic drug used to deplete lymphocytes to prevent early acute rejection. The aim of this research was to evaluate the effects of immunotherapy by rATG on graft suvival during a 9-year period in kidney-transplanted patients with different immunological profiles. METHODS: A sample of 469 patients were allocated into four groups (G) based on immunological risk of rejection: G1, low risk, not sensitized recipients, solid-phase immunoassay with single antigen beads (SPI-SAB) < 10%; G2, medium risk I, sensitized recipients, SPI-SAB ≥ 10 < 50%; G3, medium risk II sensitized (SPI-SAB ≥50%); and G4, high risk, sensitized recipients, SPI-SAB- donor-specific antibody positive (DSA+). Only patients from G3 and G4 received immunotherapy. RESULTS: Of 255 patients who received a kidney from a living donor (LD), 42 (16.47%) from all groups (G) had T-cell-mediated rejection (TCMR) and four (G1) lost their grafts, 8 (3.14%) had antibody-mediated rejection (AMR), and two lost their graft in G1 and G4. Of 214 patients who received a kidney from deceased donors (DD), 37 (17.29%) had TCMR with one lost graft in G1. AMR was shown in 13 (6.07%) patients, with three losses observed in G2. Statistical differences between the groups in the 9-year graft survival rate were found only in the comparison of G1 versus G2 (P = 0.005) and G2 versus G4 (P = 0.047) for DD. For LD, no statistical differences were found. CONCLUSION: This clinical retrospective study shows that immunotherapy induction was associated with improvement of outcomes, graft function, and survival in patients treated with immunotherapy in comparison with patients who did not received induction therapy. These findings strongly suggest that immunotherapy should be used for all patients transplanted with kidneys from deceased donors.

HLA-A, -B, -DRB1, -DQA1, and -DQB1 profile in a population from southern Brazil.

The aim of this study was to determine the allele and haplotype frequencies of HLA-A, -B, -DRB1, and -DQB1 in a self-declared White population from the north and northwestern state of Paraná, southern Brazil, and compare the data with populations worldwide. The genotyping was performed with a group of 641 individuals, based on PCR-SSO and -SSP methods, and allele and haplotype frequencies were estimated. Comparisons with European, African, Asian, and Amerindian populations were performed. The most frequent allelic groups, alleles and haplotypes were: HLA-A*02, HLA-B*35, HLA-DRB1*07:01, HLA-DQB1*03:01, and HLA-A*01/B*08/DRB1*03:01. The results reinforced a predominance of a European composition in the self-declared White population from the north and northwestern Paraná.

The distribution of HLA haplotypes in the ethnic groups that make up the Brazilian Bone Marrow Volunteer Donor Registry (REDOME).

The Registries of Bone Marrow Donors around the world include more than 30 million volunteer donors from 57 different countries, and were responsible for over 17,000 hematopoietic stem cell transplants in 2016. The Brazilian Bone Marrow Volunteer Donor Registry (REDOME) was established in 1993 and is the third largest registry in the world with more than 4.3 million donors. We characterized HLA allele and haplotypes frequencies from REDOME comparing them with the donor self-reported race group classification. Five-locus haplotype frequencies (A~C~B~DRB1~DQB1) were estimated for each of the six race groups, resolving phase and allelic ambiguity using the expectation-maximization (EM) algorithm. The top 100 haplotypes in the race groups were separated into eight clusters of haplotypes, based on haplotype similarity, using CLUTO. We present HLA allele and haplotype frequency data from six race groups from 2,938,259 individuals from REDOME. The most frequent haplotype was the same for all groups: A*01:01g~C*07:01g~B*08:01g~DRB1*03:01g~DQB1*02:01g. Some frequent haplotypes such as A*02:01g~C*16:01g~B*44:03~DRB1*07:01g~DQB1*02:01g was not found in people with Preta (Sub-Saharan African descent). A cluster including Branca (European) and Parda or non-informed (admixed) could be distinguished from both Preta (SubSaharan) and Indígena (Amerindian) groups, and from the Amarela (Asian) ones, which clustered with their original population. These results have implications on cross-population matching and can help in donor searches and population-based recruitment strategies.