Objectives: The most important complication in kidney transplant is acute/chronic rejection. In this study, we investigated the factors affecting kidney rejection and transplant survival.
Materials and Methods: In this survival analysis study, 352 patients (mean follow-up of 12.9 ± 4.4 years) who underwent renal biopsy due to increased creatinine level from 2012 to 2016 were identified by glomerular filtration rate level and rejection. Probable factors affecting renal function and survival rate after transplant rejection were assessed. P < .05 was considered as significant.
Results: Among our study patients, 40.9% developed early and 59.1% developed late acute kidney injury. Graft survival rates at 1 and 5 years were 98.9% and 68.5%, respectively, which was significant when rejection type was considered (P = .002). In addition, patient survival rates at 1 and 5 years were 99.7% and 98.6%, respectively. Graft survival at 5 years was significantly lower among older subjects, those with diabetes, those who received deceased donor organs, and those with late acute kidney injury (P < .002). Patient survival was significantly higher among young patients, those with systemic lupus erythematosus, those who received living donor organs, and those without cytomegalovirus infection (P < .003).
Conclusions: We observed that recipient age, type of donor, underlying disease, infection, and late acute kidney injury had great negative impacts on renal dysfunction and survival. In our center, because of the large number of kidney transplants from deceased donors, the necessity of antithymocyte globulin induction therapy was considered, since this study showed that patients who received rabbit anti-thymocyte globulin induction had better outcomes.

Key words : Acute kidney injury, Glomerular filtration rate, Kidney transplantation

Introduction

Renal transplant is the preferred treatment choice for most patients with advanced renal failure.¹ An average of 75 to 80 kidney transplants per million population are needed annually. At present in our country, those with transplanted kidneys live an average of 12 to 15 years, with 1-year and 3-year survival rates of 92.0% and 85.0%, respectively.²

Factors affecting kidney survival include renal origin, antibody against lymphocytic antigens, HLA antigen mismatch, ischemic kidney duration, age (donor/recipient), underlying disease, infection, especially BK virus and cytomegalovirus (CMV), type of immunosuppression, hypertension, recurrence of primary glomerular disease, and factors related to the transplant center.^3,4

Graft rejection is an unintended event that occurs due to immune system reactions; if the immune system is not suppressed, the donation will certainly be lost.² Kidney transplant rejection includes various types and may be early acute, acute, or chronic. The most important characteristic of rejection is the rapid or gradual increase in creatinine of ≥1.5 mg/dL.⁵ Hyperacute rejection usually occurs promptly or within minutes after transplant, and treatment involves removing the transplanted organ. Today, this event is very rare due to HLA matching tests.⁶

After kidney transplant, patients are often administered a combination of different drugs (usually prednisolone, calcineurin inhibitors [CNIs], and antimetabolites), which are used to prevent rejection. We conducted this study to investigate the prevalence of kidney transplant failure and to determine patient and graft survival after renal transplant and the factors affecting renal transplant failure in renal transplanted recipients of Southern Iran.

Materials and Methods

Study population
We retrospectively evaluated the outcomes of 352 renal transplant recipients who had undergone kidney biopsies from 2012 to 2016 at the Shiraz Nemazee Hospital due to increased serum creatinine levels of >1.5 mg/dL after renal transplant.

Study protocol
Inclusion criteria were kidney transplant recipients with an increase in serum creatinine >1.5 mg/dL who underwent kidney biopsy. Patients were excluded if there were insufficient medical records and lack of access to patients to complete missing data.

From patient medical records, we collected information on recipient age at time of transplant, underlying cause of renal failure, timing from transplant to first kidney biopsy (early acute kidney injury [AKI] was <1 month posttransplant; late AKI was >1 month posttransplant), immunosuppressive drugs, serum BK virus and CMV infections, pathologic findings, induction therapy, serum creatinine levels, urea, and last follow-up visit, which was at least 5 years after transplant; we also collected estimated glomerular filtration rate (eGFR) results at 1, 3, and 5 years posttransplant, which was calculated by the CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) equation.⁷

Immunosuppressive therapy
All patients were treated with CNIs (cyclosporine 6 mg/kg/d or tacrolimus 0.05-0.1 mg/kg/d) or mechanistic target of rapamycin inhibitors (everolimus or sirolimus) combined with mycophenolic acid (1440 mg/d) or mycophenolate mofetil (2 g/d) along with corticosteroids (intravenous methylprednisolone of 500 mg and then prednisolone orally at 1 mg/kg/d). Some patients received 2 mg/kg rabbit antithymocyte globulin (rATG) before and 3 days after transplant. In our center, CNI doses were adjusted to achieve target trough levels (for cyclosporine, 200-300 ng/mL for the first 3 months and then 100-200 ng/mL; for tacrolimus, 8-12 ng/mL for the first 3 months and then 5-8 ng/mL). Oral prednisolone dose was tapered to 20 mg/day after 1 month, 10 mg/day after 2 months, and 5 mg/day after 6 months.

Rejection type
Depending on the histopathology and immunological characteristics, renal transplant rejection can be classified as hyperacute rejection, acute rejection, which is classified into antibody-mediated rejection (AMR) and acute cellular rejection (ACR), chronic allograft nephropathy (CAN), and a mix of acute rejection superimposed on chronic rejection. Because not all patients with renal dysfunction show pathology of rejection in their renal biopsy, we defined early versus late AKI in our study. In addition, 55 patients had simultaneous kidney and pancreas transplant because of underlying disease of diabetes.

Ethical considerations
Our study protocol was in accordance with the Declaration of Helsinki and the local ethics committee. All patients provided written informed consent before inclusion.

Statistical analyses
All statistical analyses were performed using the SPSS software (version 20.0). Data were analyzed with the chi-square or t test, one-way analysis of variance, Kaplan-Meier survey analysis, log-rank test, the Breslow test, and Cox regression analysis. P < .05 was considered as statistically significant.

Results

Our study included 352 renal transplant recipients who developed AKI posttransplant and underwent renal biopsies, which were performed at a range of 6 days to 18 years after transplant. Patient characteristics are shown in Table 1. Mean follow-up of patients was 12.9 ± 4.4 years (minimum to maximum of 5 to 27.0 years). Mean age of patients was 36.3 ± 14.8 years at the time of renal transplant; 61.6% of patients were men. Diabetes mellitus (35.2%) and systemic lupus erythematous were the main causes of renal failure. There were 57 living donors (16.7%), which included 43.9% parent donors and 56.1% sibling donors.

Among the study patients, 23 patients died after transplant (1 to 13 years posttransplant); these patients were older at the time of transplant (54.8 ± 15.7 years of age) compared with patients who survived (35.1 ± 13.8 years of age; P = .001).

In 74.6% of recipients, tacrolimus plus myco-phenolate mofetil was the immunosuppressive regimen. Type of dialysis before transplant was hemodialysis in 88.6% of patients. Among the 352 recipients, 144 (40.9%) developed early AKI and 208 (59.1%) developed late AKI. There was no significant difference between early and late AKI with regard to age, sex, underlying disease, type of dialysis, and immunosuppressive drug (P > .05). Most recipients who received organs from living donors developed late AKI (77.2%) versus 55.6% of recipients who received organs from deceased donors.

The most common pathologic findings in early AKI were ACR (52.7%), AMR (18%), combination of AMR and ACR (7.6%), acute tubular necrosis (5.5%), and acute tubulointerstitial nephritis (2.0%). The reported pathology results in recipients with late AKI included ACR (30%), CAN (24%), AMR (2.4%), acute tubulointerstitial nephritis (7.2%), and acute tubular necrosis (2.8%). Of patients with ACR, 79 (49.7%) and 49 (30.8%) showed class IB and IA of Banff classification, respectively.

The last mean eGFR was 68.0 ± 32.4 mL/min/m2 in the early AKI group and 54.8 ± 33.9 mL/min/m² in the late AKI group (P < .001). Overall, 108 patients (30.7%) had progression of eGFR to ≤30 mL/min/m², with 71.3% of these from the late AKI group (P = .002).

In this study, 45.5% of patients received rATG as induction; these patients showed less probability of developing early AKI versus late AKI (4.5% vs 52.5%; P = .023). Infections with BK virus and CMV were reported in 16.5% and 5.9% of patients, respectively, which mostly occurred during the first month posttransplant. The mean last eGFR result was significantly higher in younger patients, men, those with early AKI, and those who received rATG induction (Table 2; P < .001).

Survival time after graft rejection
The 1-year and 5-year graft survival rates based on eGFR <30 mL/min/m2 was 98.9% and 68.5%, respectively (Figure 1). In addition, patient survival at 1 year was 99.7% and patient survival at 5 years was 98.6% (Figure 2). Graft survival was significantly different based on type of rejection (97.2% at 1 year and 78.2% at 5 years in the early AKI group vs 100.0% at 1 year and 61.6% at 5 years in the late AKI group; P = .002) (Figure 3). Graft survival rates were significantly lower among older patients and those with diabetes (both 1-year and 5-year survival), those who received deceased donor organs (1-year survival), and those with late AKI (5-year survival) (P< .002; Table 3). Results of Cox regression showed only late AKI as a significant risk factor for graft survival (hazard ratio of 1.7; 95.0% CI, 1.1-2.6; P = .013).

Patient survival was significantly lower among older patients, those with diabetes, those who received deceased donor organs (both 1-year and 5-year survival) and those with CMV infection (5-year survival) (P < .003; Table 4). In multivariate analysis, only CMV infection was a significant risk factor for patient survival (hazard ratio of 3.1; 95.0% CI, 1.1-9.1; P = .043).

In addition, 29 patients (8.2%) had GFR <15 mL/min/m² at least 5 years posttransplant. We found that GFR status of ≤15 mL/min/m² was significantly greater in the late AKI group (P < .009; Table 5).

Discussion

This retrospective study analyzed the effects of recipient characteristics, immunosuppressive drugs, type of rejection, and specifically time of renal dysfunction after renal transplant on kidney and patient survival outcomes in renal transplant recipients. We found that late AKI was a determinant of long-term graft survival, as previously reported.^8,9 In our study, the late AKI group had significant lower eGFR at last follow-up than the early AKI group. Although different studies had various definitions of early and late rejection (from 2-12 months), it seems that the most probable causes are related to CNI toxicity, persistent interstitial inflammation and tubulitis, and nonadherence with immunosup-pressive medications for late rejection episodes,¹⁰ leading to CAN and decreased survival of grafts in the long term.

In a multivariant analysis, Koo and colleagues reported no statistical difference between early and late rejection with regard to graft survival. They defined late rejection as that occurring 1 year posttransplant.¹¹ Variable results may be because studies use different definitions for time of renal dysfunction and rejection. Some studies did not report biopsy-proven acute rejection.^8,12 In addition, study populations with regard to deceased or living donors may be different.

The most common pathology finding in our patients was ACR in both the early and late AKI groups. However, the prevalence of AMR was higher in the early AKI group compared with the late AKI group. Although the last eGFR was lower in patients with AMR (61.7 ± 28.5 mL/min/m²) compared with those with ACR (65.7 ± 34.4 mL/min/m²), no statistically significant difference was shown.

With a reported incidence of 30% to 40% in HLA-sensitized patients, AMR is a major risk factor for premature graft loss.^13,14 The presence of donor-specific antibodies, preformed or de novo, is a major contributing factor for development of AMR.¹⁵ There is also strong evidence that antibody-mediated graft injury is a major cause of late graft loss.^16,17 In our study, the prevalence of AMR was lower than shown in other studies, perhaps because most of our pathologies were evaluated before the introduction and use of C4d staining in biopsies as a marker of AMR. We found that late rejection, regardless of type of rejection and renal dysfunction, leads to inferior graft survival, which is consistent with findings in a larger registry study.¹⁸

Our study indicated that rATG induction not only had a positive impact on long-term eGFR but also did not increase the risk of posttransplant viral infections, including BK virus and CMV infections. Patients who received rATG had significantly better last eGFR (65.5 ± 34.3 mL/min/m2) than those without rATG therapy (55.2 ± 32.9 mL/min/m²). Ahn and colleagues showed that rATG is an effective way to prevent rejection in both older and younger recipients.¹⁹ The use of rATG induction to prevent acute rejection episodes among deceased transplant recipients with high immunologic risk factors is highly effective.²⁰ Previous randomized controlled trials have indicated the efficacy of ATG compared with basiliximab in adult kidney transplant recipients.^21-23 Because, in our center at the time of this study, we did not evaluate HLA typing and donor-specific antibodies in all candidates, our recipients were placed in the high-risk immunologic category, with rATG induction seeming to have a promising effect on long-term graft survival.

Our patient survival rates at 1 and 5 years were 99.7% and 98.6%, whereas graft survival rates at 1 and 5 years were 98.9% and 68.5%, respectively. Previous studies have reported their 1-year kidney transplant recipient survival rates as between 90% and 94% and 5-year survival rates that ranged from 68.0% (Doyle and colleagues) to 96% (Rocha and colleagues) among patients without diabetes.^24-29 Long-term outcomes have since increased with better expertise and developments in immunosuppression treatment.³⁰ It is well recognized that survival is influenced by several factors. Higher recipient age, diabetes, deceased donor transplants, and late AKI have been associated with lower graft and patient survival at 1 year and 5 years posttransplant. In our study, Cox regression analysis showed that only late AKI was a significant risk factor for graft loss.

In general, living donor transplant procedures have better outcomes than deceased donor transplant procedures, due to better organ quality and less ischemic time.^24,31 Recipient age may also have a significant correlation with renal transplant outcomes,^32,33 although some studies have found no correlation.^11,26 In the study from Ahn and colleagues, which reported on older renal transplant recipients, the incidence of acute rejection was lower, but these rejections may lead to graft loss more frequently.³⁴ Because systemic lupus erythematosus is more common in younger patients and has a low rate of vasculopathy compared with that shown in patients with diabetes, patients with systemic lupus erythematosus may have better survival. Some studies mentioned that patients with diabetes have poorer graft survival.^29,31

Our multivariate analysis showed that late rejection had a determinantal effect on 5-year graft survival and that CMV infection was a major risk factor for patient survival. In our study, 1-year graft survival was 97.2% and 5-year graft survival was 78.2% for the early AKI group versus 1-year graft survival of 100.0% and 5-year graft survival of 61.6% for the late AKI group. The last eGFR was higher in patients with early transplant rejection, indicating that acute rejection at <1 month had less effect on final survival of the graft; thus, patients with late rejection had worse 5-year graft survival. Another study found a significant relationship between 5-year graft survival and late acute rejection.²⁶ Infection with CMV can vary from 17.8% to 63.2% in renal transplant recipients and remains a significant cause of reduction of eGFR, rejection, reduced graft/patient survival, and mortality. In addition, CMV infection can be a latent problem when ATG is used in induction and acute rejection, as ATG is a risk factor for CMV.³⁵ With a prevalence of 1% to 10%, BK virus infection is a leading cause of kidney allograft failure and is significantly associated with graft survival.^25,36

Conclusions

We found that recipient age more than 30 years, deceased donor status, presence of diabetes mellitus as an underlying disease, CMV infection, and late AKI of the transplanted kidney had strong negative effects on renal function and graft/patient survival after renal transplant. Patients who received rATG induction had better outcomes.

References:

1. Tanagho EA, McAninch JA. Hafezi M, Sobhaniyan K, Sori A, et al, trans-ed. Smith’s General Urology. 16th edition. Tehran: Tabib Teimorzadeh Publisher; 2007.
   CrossRef - PubMed
   
2. Tabatabaei SA, Hashemi SM, Zandi M. Frequency distribution of long time complications of renal transplantation in Alzahra Hospital. J Isfahan Med School. 2011;28(119):1410-1413.
   CrossRef - PubMed
   
3. Courtney AE, McNamee PT, Maxwell AP. The evolution of renal transplantation in clinical practice: for better, for worse? QJM. 2008;101(12):967-978. doi:10.1093/qjmed/hcn136
   CrossRef - PubMed
   
4. Akbarzadeh A, Alikhanpour SH. Quality of life in patients under kidney transplantation. Shahed J. 2018;129(1):100-103.
   CrossRef - PubMed
   
5. Shahsavar F, Entezami A. Basic Immunology for Medical Students. Tehran: Peyvande Mehr Publishing; 2013:7-9.
   CrossRef - PubMed
   
6. Frohn C, Fricke L, Puchta JC, Kirchner H. The effect of HLA-C matching on acute renal transplant rejection. Nephrol Dial Transplant. 2001;16(2):355-360. doi:10.1093/ndt/16.2.355
   CrossRef - PubMed
   
7. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604-612. doi:10.7326/0003-4819-150-9-200905050-00006
   CrossRef - PubMed
   
8. Sijpkens YW, Doxiadis, II, Mallat MJ, et al. Early versus late acute rejection episodes in renal transplantation. Transplantation. 2003;75(2):204-208. doi:10.1097/01.TP.0000041722.34000.21
   CrossRef - PubMed
   
9. Thorogood J, Persijn GG, Schreuder GM, et al. The effect of HLA matching on kidney graft survival in separate posttransplantation intervals. Transplantation. 1990;50(1):146-150. doi:10.1097/00007890-199007000-00027
   CrossRef - PubMed
   
10. Sijpkens YW, Doxiadis, II, van Kemenade FJ, et al. Chronic rejection with or without transplant vasculopathy. Clin Transplant. 2003;17(3):163-170. doi:10.1034/j.1399-0012.2003.00039.x
    CrossRef - PubMed
    
11. Koo EH, Jang HR, Lee JE, et al. The impact of early and late acute rejection on graft survival in renal transplantation. Kidney Res Clin Pract. 2015;34(3):160-164. doi:10.1016/j.krcp.2015.06.003
    CrossRef - PubMed
    
12. Lentine KL, Gheorghian A, Axelrod D, Kalsekar A, L’Italien G, Schnitzler MA. The implications of acute rejection for allograft survival in contemporary U.S. kidney transplantation. Transplantation. 2012;94(4):369-376. doi:10.1097/TP.0b013e318259407f
    CrossRef - PubMed
    
13. Haririan A, Nogueira J, Kukuruga D, et al. Positive cross-match living donor kidney transplantation: longer-term outcomes. Am J Transplant. 2009;9(3):536-542. doi:10.1111/j.1600-6143.2008.02524.x
    CrossRef - PubMed
    
14. Gloor J, Stegall MD. Sensitized renal transplant recipients: current protocols and future directions. Nat Rev Nephrol. 2010;6(5):297-306. doi:10.1038/nrneph.2010.34
    CrossRef - PubMed
    
15. Mohan S, Palanisamy A, Tsapepas D, et al. Donor-specific antibodies adversely affect kidney allograft outcomes. J Am Soc Nephrol. 2012;23(12):2061-2071. doi:10.1681/ASN.2012070664
    CrossRef - PubMed
16. Gaston RS, Cecka JM, Kasiske BL, et al. Evidence for antibody-mediated injury as a major determinant of late kidney allograft failure. Transplantation. 2010;90(1):68-74. doi:10.1097/TP.0b013e3181e065de
    CrossRef - PubMed
    
17. Einecke G, Sis B, Reeve J, et al. Antibody-mediated microcirculation injury is the major cause of late kidney transplant failure. Am J Transplant. 2009;9(11):2520-2531. doi:10.1111/j.1600-6143.2009.02799.x
    CrossRef - PubMed
    
18. Opelz G, Dohler B; Collaborative Transplant Study Report. Influence of time of rejection on long-term graft survival in renal transplantation. Transplantation. 2008;85(5):661-666. doi:10.1097/TP.0b013e3181661695
    CrossRef - PubMed
    
19. Ahn JB, Bae S, Chu NM, et al. The risk of postkidney transplant outcomes by induction choice differs by recipient age. Transplant Direct. 2021;7(7):e715. doi:10.1097/TXD.0000000000001105
    CrossRef - PubMed
    
20. Mohammadi K, Khajeh B, Dashti-Khavidaki S, Shab-Bidar S. Association between cumulative rATG induction doses and kidney graft outcomes and adverse effects in kidney transplant patients: a systematic review and meta-analysis. Expert Opin Biol Ther. 2021;21(9):1265-1279. doi:10.1080/14712598.2021.1960978
    CrossRef - PubMed
    
21. Brennan DC, Daller JA, Lake KD, Cibrik D, Del Castillo D; Thymoglobulin Induction Study Group. Rabbit antithymocyte globulin versus basiliximab in renal transplantation. N Engl J Med. 2006;355(19):1967-1977. doi:10.1056/NEJMoa060068
    CrossRef - PubMed
    
22. Gill J, Sampaio M, Gill JS, et al. Induction immunosuppressive therapy in the elderly kidney transplant recipient in the United States. Clin J Am Soc Nephrol. 2011;6(5):1168-1178. doi:10.2215/CJN.07540810
    CrossRef - PubMed
23. Lentine KL, Schnitzler MA, Xiao H, Brennan DC. Long-term safety and efficacy of antithymocyte globulin induction: use of integrated national registry data to achieve ten-year follow-up of 10-10 study participants. Trials. 2015;16:365. doi:10.1186/s13063-015-0891-y
    CrossRef - PubMed
    
24. de Castro Rodrigues Ferreira F, Cristelli MP, Paula MI, et al. Infectious complications as the leading cause of death after kidney transplantation: analysis of more than 10,000 transplants from a single center. J Nephrol. 2017;30(4):601-606. doi:10.1007/s40620-017-0379-9
    CrossRef - PubMed
    
25. Favi E, Puliatti C, Sivaprakasam R, et al. Incidence, risk factors, and outcome of BK polyomavirus infection after kidney transplantation. World J Clin Cases. 2019;7(3):270-290. doi:10.12998/wjcc.v7.i3.270
    CrossRef - PubMed
    
26. Jalalzadeh M, Mousavinasab N, Peyrovi S, Ghadiani MH. The impact of acute rejection in kidney transplantation on long-term allograft and patient outcome. Nephrourol Mon. 2015;7(1):e24439. doi:10.5812/numonthly.24439
    CrossRef - PubMed
    
27. Doyle SE, Matas AJ, Gillingham K, Rosenberg ME. Predicting clinical outcome in the elderly renal transplant recipient. Kidney Int. 2000;57(5):2144-2150. doi:10.1046/j.1523-1755.2000.00066.x
    CrossRef - PubMed
    
28. Wu MJ, Yu TM, Lin CL, Kao CH. Propensity score-matched analysis of the survival benefit from kidney transplantation in patients with end-stage renal disease. J Clin Med. 2018;7(11):388. doi:10.3390/jcm7110388
    CrossRef - PubMed
    
29. Rocha A, Malheiro J, Martins LS, et al. Kidney transplantation in type 2 diabetic patients: a matched survival analysis. Transplant Proc. 2013;45(6):2141-2146. doi:10.1016/j.transproceed.2012.11.013
    CrossRef - PubMed
30. Davidovits M, Reisman L, Cleper R, et al. Long-term outcomes during 37 years of pediatric kidney transplantation: a cohort study comparing ethnic groups. Pediatr Nephrol. 2021;36(7):1881-1888. doi:10.1007/s00467-020-04908-6
    CrossRef - PubMed
    
31. Damodaran S, Bullock B, Ekwenna O, et al. Risk factors for delayed graft function and their impact on graft outcomes in live donor kidney transplantation. Int Urol Nephrol. 2021;53(3):439-446. doi:10.1007/s11255-020-02687-5
    CrossRef - PubMed
    
32. Briganti EM, Wolfe R, Russ GR, Eris JM, Walker RG, McNeil JJ. Graft loss following renal transplantation in Australia: is there a centre effect? Nephrol Dial Transplant. 2002;17(6):1099-1104. doi:10.1093/ndt/17.6.1099
    CrossRef - PubMed
    
33. Almasi Hashiani A, Rajaeefard A, Hassanzade J, Salahi H. Survival analysis of renal transplantation and its relationship with age and sex. Koomesh. 2010;11(4):302-306.
    CrossRef - PubMed
    
34. Peeters LEJ, Andrews LM, Hesselink DA, de Winter BCM, van Gelder T. Personalized immunosuppression in elderly renal transplant recipients. Pharmacol Res. 2018;130:303-307. doi:10.1016/j.phrs.2018.02.031
    CrossRef - PubMed
    
35. Axelrod D, Chang SH, Lentine KL, et al. The clinical and economic benefit of CMV matching in kidney transplant: a decision analysis. Transplantation. 2021. doi:10.1097/TP.0000000000003887
    CrossRef - PubMed
    
36. Nili F, Mohammadhoseini M, Khatami SM, Seirafi G, Haghzare M. Routine immunohistochemistry study for polyomavirus BK nephropathy in transplanted kidney biopsies, is it recommended? BMC Nephrol. 2021;22(1):226. doi:10.1186/s12882-021-02444-5
    CrossRef - PubMed