Recipient functional status impacts on short and long-term intestinal transplant outcomes in United States adults

Abstract

BACKGROUND

Recipient functional status prior to transplantation has been found to impact post-transplant outcomes in heart, liver and kidney transplants. However, information on how functional status, before and after transplant impacts post-transplant survival outcomes is lacking.

AIM

To investigate the impact of recipient functional status on short and long term intestinal transplant outcomes in United States adults.

METHODS

We conducted a retrospective cohort study on 1254 adults who underwent first-time intestinal transplantation from 2005 to 2022. The primary outcome was mortality. Using the Karnofsky Performance Status, functional impairment was categorized as severe, moderate and normal. Analyses were conducted using Kaplan-Meier curves and multivariable Cox regression.

RESULTS

The median age was 41 years, majority (53.4%) were women. Severe impairment was present in 28.3% of recipients. The median survival time was 906.6 days. The median survival time was 1331 and 560 days for patients with normal and severe functional impairment respectively. Recipients with severe impairment had a 56% higher risk of mortality at one year [Hazard ratio (HR) = 1.56; 95%CI: 1.23–1.98; P < 0.001] and 58% at five years (HR = 1.58; 95%CI: 1.24–2.00; P < 0.001) compared to patients with no functional impairment. Recipients with worse functional status after transplant also had poor survival outcomes.

CONCLUSION

Pre- and post-transplant recipient functional status is an important prognostic indicator for short- and long-term intestinal transplant outcomes.

Key Words: Intestinal transplantation, Functional status, Karnofsky performance status, Post-transplant outcomes, Survival analysis, Transplant outcome

Core Tip: Pre- and post-transplant assessment of recipient functional status using the Karnofsky Performance Status scale is crucial for predicting short- and long-term outcomes in intestinal transplantation. Severe impairment significantly increases the risk of mortality at one- and five-years post-transplant, emphasizing the importance of comprehensive pre-transplant evaluation and potential interventions to optimize functional status in recipients undergoing intestinal transplantation in the United States.

INTRODUCTION

Intestinal transplantation is the mainstay treatment option for eligible patients with severe gastrointestinal diseases, including inflammatory bowel diseases, acute mesenteric infarction, and refractory end-stage intestinal failure[1]. Over the last three years, there have been a rise in the number of intestinal transplantations in the United States, with an average of 90 transplants a year since 2019[2]. While there have been significant advancements in surgical techniques, immunosuppression, and patient management, as high as 44% of intestinal transplant recipients still die within 5-10 years of transplant in the United States[3]. The process of identifying appropriate candidates for intestinal transplants is multifaceted and focuses on classifying potential recipients based on their present clinical condition, nutritional support requirements, and overall intestinal function[4]. Thus, identifying reliable prognostic indicators for post-transplant outcomes is paramount in optimizing patient selection, pre-transplant management, and patient counseling[5]. Recipient functional status (RFS) has emerged as an influential factor impacting post-transplant outcomes across various solid organ transplantations, such as heart, liver, and kidney transplants[6-9]. The Karnofsky Performance Status (KPS) scale is a widely accepted and validated measure of RFS in clinical practice and research[10]. It has been employed as a prognostic tool in liver, kidney, and lung transplantation[11-13]. However, there is limited research on the impact of RFS on short-term and long-term outcomes in intestinal transplantation.

In this study, our objective was to investigate the association of RFS, assessed using the KPS scale, with one and five-year intestinal transplant outcomes among adult recipients in the United States. We hypothesize that lower functional status prior to transplantation is associated with poorer one and five-year post-transplant outcomes. Further, we hypothesize that reduced post-transplant KPS is associated with poorer post-transplant survival[14]. To evaluate these hypotheses, we performed a retrospective cohort study of all adult patients who underwent a first-time intestinal transplantation in the United States between 2005 and 2022 using the United Network for Organ Sharing (UNOS) database. The UNOS records information about all transplants performed in the US, since its establishment in 1988[15]. By examining the relationship of RFS with intestinal transplant outcomes, our study aims to offer valuable insights for clinicians in prognostic counseling, patient selection, and pre-transplant management. Ultimately, this study will contribute to the ongoing efforts to improve outcomes for intestinal transplant recipients.

MATERIALS AND METHODS

Study design and population

This retrospective cohort study utilized the UNOS dataset, which contains information on all intestinal transplantations performed in the United States from 1988 to date[15]. We obtained the fully anonymized data from the UNOS database via a request submitted to the Organ Procurement and Transplant Network. UNOS is provided with the patient data at the point of patient registration (listing) and updated it during transplantation. The database is updated annually with post-transplant outcomes information. Since the UNOS database is a public resource and was already deidentified before we obtained it, the data was deemed exempt from Institutional Review Board supervision by the Yale institutional review board. The datasets utilized in this study can be found on the UNOS website at https://unos.org/[15]. Patients were followed up to the time of death or loss to follow -up, whichever occurred first. We included adult patients (≥ 18 years) who underwent a first-time intestinal transplantation between January 1, 2005 and December 31, 2022.

Exclusion criteria: (1) Pediatric transplants (< 18 years); (2) Multi-organ transplants; and (3) Missing data on functional status prior to transplant.

Exposure variable

The primary exposure of interest was RFS at transplant, assessed using the KPS scale (Table 1). The KPS quantifies functional status using an 11-point scale, ranging from 100% (normal, without complaints or signs of disease) to 0% (deceased), with increments of 10% and three corresponding categories[10,16]. We utilized the three-tiered version of the scale due to its higher inter-rater reliability scores[9,10]. As adopted by Dolgin et al[7] the three categories of functional status were based on the level of functional impairment/disability: (1) None/normal (function) (80%-100%); (2) Moderate (impairment in function) (50%-70%); and (3) Severe (functional impairment/disability) (10%-40%).

Table 1 Karnofsky Performance Status rating and categories for patient functionality.

Functional level	Score range (%)	Patient capabilities and care requirements
High function	80-100	100%: Fully active, no symptoms of disease
		90%: Minor symptoms, fully active
		80%: Some symptoms, normal activity with effort
Moderate function	50-70	70%: Manages self-care, limited activity
		60%: Occasionally needs help but manages most personal needs
		50%: Requires regular assistance and medical care
Low function	0-40	40%: Disabled, requires special assistance
		30%: Severely disabled, frequent hospital care
		20%: Very ill, active medical intervention needed
		10%: Terminally ill, rapid disease progression
		0%: Deceased

Outcome measures

The primary outcomes of interest were one and five-year all-cause mortality. We examined the hazard ratios of post-transplant outcomes according to pre-transplant RFS. We also assessed the KPS at follow-up to examine post-transplant changes in KPS.

Statistical analysis

Descriptive statistics were used to summarize baseline recipient, donor, and clinical characteristics (covariates) across the functional status categories. Recipient clinical characteristics at transplant considered for inclusion in the analysis were based on clinical plausibility or prior literature on organ transplantation. Specifically, the covariates assessed were recipient age, sex, race/ethnicity, donor age, and clinical factors including smoking history, illicit drug use, cancer history, hypertension or diabetes, recipient diagnosis, the presence of any form of life support such as invasive ventilation (mechanical or extracorporeal membrane oxygenation) or non-invasive artificial ventilation (such as Bilevel positive airway pressure or continuous positive airway pressure) , creatinine levels (an indicator of kidney failure)[17], bilirubin levels (an indicator of liver disease)[18], ischemic time (the duration between graft harvest and reperfusion in the recipient)[19], weight mismatch (donor-recipient weight ratio < 0.8), Cytomegalovirus (CMV) mismatch (presence of positive CMV IgG serology in the donor and negative CMV serology in the recipient), ABO blood group incompatibility (differing ABO blood groups between donor and recipient), and Human Leukocyte Antigen mismatch (≥ 4 mismatched antigens)[20]. Further, we conducted a subgroup analysis according to different time periods, categorized as 2003-2014 and 2014-2022. Univariate cox regression analysis was performed to evaluate the effect of each covariate on one and five-year mortality. Those covariates that were associated with one or five-year mortality in univariate analysis (P < 0.2) were included in a multivariable cox regression model.

We utilized multivariable Cox regression models to estimate the adjusted hazard ratios in one and five-year post-transplant outcomes according to pre-transplant RFS. Models were adjusted for covariates on recipient and donor characteristics as well as clinical information (listed above).

We additionally conducted a predictive analysis of the prognostic factors accounting for poor KPS after transplant using multivariable cox regression analysis. A landmark time of 30 days was used for the univariable and multivariable cox regression analysis.

To assess the effect of change in KPS on survival outcomes, survival probabilities, based on changes in KPS, were calculated separately for the three groups of baseline KPS using Kaplan-Meier method. KPS scores from the initial three months post-transplant were excluded to avoid the recovery phase, focusing on the 3–12 month period for patient stratification. The criteria for KPS improvement varied across groups. High KPS patients (80%–100%) were divided based on a KPS score decrease of at least 10% or any improvement. Intermediate KPS patients (50%–70%) were grouped into no improvement (≤ 0%), an improvement of 10%–20%, or ≥ 30%. Low KPS patients (≤ 40%) were categorized into no improvement (≤ 0%), improvements of 10%–40%, 50%–70%, or ≥ 80%. This stratification aimed to reflect potential variances in KPS changes among the groups. We estimated the post-transplant survival probability at specific times, from one to five years, and median survival times across the different categories. This stratification has been previously used in assessing KPS changes in post liver transplants by Thuluvath et al[14].

All statistical comparisons of continuous data were made using one-way analysis of variance, and the chi-square test was used for categorical data. All data were presented as mean ± SD or number with percentage. Hazard ratios were presented with 95%CI. All statistical analyses were performed using SAS software, version 9.4 (SAS Institute, Cary, NC, United States). A two-sided P-value of less than 0.05 was considered statistically significant.

Sensitivity analysis using the E-value

To evaluate the robustness of our study findings and to quantify the potential impact of unmeasured confounding, we conducted a sensitivity analysis using the E-value method. The E-value is a statistical tool that provides a measure of the minimum strength of association, on the hazard ratio scale, that an unmeasured confounder would need to have with both the treatment and the outcome, above and beyond the measured confounders, to fully explain away a specific treatment–outcome association[19]. It provides an evidence-based approach to discuss the possible impact of unmeasured confounding on the observed study results. For our analysis, we computed the E-values for both the observed hazard ratios and their lower confidence limits. A larger E-value indicates that greater unmeasured confounding would be needed to explain away the treatment–outcome association, providing stronger evidence for a causal association. We computed the E-values using the formulas provided by VanderWeele and Ding[21].

RESULTS

A total of 2379 first time intestinal transplants were performed between January 1, 2005 and December 31, 2022. Of these, 1254 intestinal transplants were included in the study. The recipients ranged in age from 18 to 73 years (mean 40.9 years). Overall, there were 53.4% females, 11.4% Blacks, 78.5% Whites and 8.1% Hispanics. In terms of the RFS, 28.3% were classified as near death or severely disabled using the KPS, 51.7% as requiring some assistance, and 20% as requiring no assistance. In the majority of recipients, the indication for transplant was short gut syndrome (62.1%). The median survival time was 906.6 days (mean 1526 days). The median survival time was 1331 days for recipients with normal KPS, 1078 days for those with moderate impairment and 560 days for those with severe impairment. Table 2 presents the baseline characteristics of the patients in each functional status group, considering all intestinal transplants without any time periods. No significant differences were observed among the three groups for recipient age (P = 0.22), and race (P = 0.79). However, significant differences were found between the functional status groups for recipient gender (P = 0.034), patients on life support at the time of transplant (P value = 0.004), patients in the intensive care unit (ICU) (P < 0.001), and total ischemic time (P = 0.01).

Table 2 Comparison of patient outcomes by functional status categories, n (%)/mean ± SD.

	Functional status at transplant
	Requires no assistance (n = 251)	Requires some assistance (n = 648)	Near death or severely disabled (n = 355)	Total (n = 1254)	P value
Gender					0.0341
Female	130 (51.8)	368 (56.8)	172 (48.5)	670 (53.4)
Male	121 (48.2)	280 (43.2)	183 (51.5)	584 (46.6)
Age (Year)					0.222
n	251	648	355	1254
Mean (SD)	40.9 (14.18)	42.7 (13.04)	42.1 (12.79)	42.1 (13.21)
Median	41.0	44.0	42.0	43.0
Range	18.0-73.0	18.0-72.0	18.0-70.0	18.0-73.0
Race					0.791
Hispanic	22 (8.8)	49 (7.6)	31 (8.7)	102 (8.1)
Non-Hispanic Asian	3 (1.2)	11 (1.7)	5 (1.4)	19 (1.5)
Non-Hispanic Black	29 (11.6)	71 (11.0)	43 (12.1)	143 (11.4)
Non-Hispanic White	194 (77.3)	515 (79.5)	275 (77.5)	984 (78.5)
Other/Multiracial	3 (1.2)	2 (0.3)	1 (0.3)	6 (0.5)
Patients on life support					0.00402
No	242 (96.4)	602 (92.9)	317 (89.3)	1161 (92.6)
Yes	9 (3.6)	46 (7.1)	38 (10.7)	93 (7.4)
Sex match					0.161
No	125 (49.8)	288 (44.4)	177 (49.9)	590 (47.0)
Yes	126 (50.2)	360 (55.6)	178 (50.1)	664 (53.0)
Recipient total bilirubin at transplantation (mg/dL)					0.00062
n	249	644	354	1247
Mean (SD)	2.2 (5.36)	2.3 (4.57)	4.6 (8.44)	2.9 (6.15)
Median	0.8	0.8	1.1	0.9
Range	0.1-50.0	0.1-43.0	0.1-48.2	0.1-50.0
Serum creatinine at time of transplantation (mg/dL)					0.361
n	251	645	353	1249
Mean (SD)	1.0 (0.72)	1.1 (0.96)	1.2 (1.04)	1.1 (0.94)
Median	0.9	0.9	0.9	0.9
Range	0.3-7.6	0.3-10.1	0.1-9.0	0.1-10.1
Total ischemic time (sec)					0.0102
n	222	619	347	1188
Mean (SD)	6.9 (1.88)	7.5 (2.32)	7.5 (2.28)	7.4 (2.24)
Median	7.0	7.3	7.4	7.2
Range	1.0-11.8	0.4-31.5	0.9-31.7	0.4-31.7
Recipient's primary diagnosis					0.010411
Short gut syndrome	154 (19.8)	409 (52.5)	216 (27.7)	779 (62.1)
Functional bowel problem	27 (28.1)	51 (53.1)	18 (18.8)	96 (7.7)
Retransplant	4 (6.2)	39 (60.9)	21 (32.8)	64 (5.1)
Others	66 (21.0)	149 (47.3)	100 (31.7)	315 (25.1)
Donor age (years)					0.362
n	251	648	355	1254
Mean (SD)	22.4 (10.79)	21.1 (10.54)	21.3 (11.47)	21.4 (10.86)
Median	20.0	19.0	20.0	19.0
Range	4.0-56.0	2.0-52.0	1.0-59.0	1.0-59.0
Race match					0.431
No	119 (47.4)	278 (42.9)	152 (42.8)	549 (43.8)
Yes	132 (52.6)	370 (57.1)	203 (57.2)	705 (56.2)
Weight mismatch					0.00691
NA	4 (1.6)	7 (1.1)	0 (0.0)	11 (0.9)
No mismatch	170 (67.7)	397 (61.3)	201 (56.6)	768 (61.2)
Weight mismatch	77 (30.7)	244 (37.7)	154 (43.4)	475 (37.9)
History of diabetes					0.0821
Diabetes	21 (8.4)	68 (10.5)	49 (13.9)	138 (11.0)
Not diabetes	230 (91.6)	580 (89.5)	303 (86.1)	1113 (89.0)
Liver function					< 0.00011
Liver dysfunction	62 (24.9)	148 (23.0)	126 (35.6)	336 (26.9)
No dysfunction	187 (75.1)	496 (77.0)	228 (64.4)	911 (73.1)
Renal function					0.0381
No dysfunction	244 (97.2)	617 (95.7)	328 (92.9)	1189 (95.2)
Renal dysfunction	7 (2.8)	28 (4.3)	25 (7.1)	60 (4.8)
HLA mismatch					0.201
High HLA mismatch	208 (87.0)	522 (82.6)	293 (85.7)	1023 (84.3)
Low HLA mismatch	31 (13.0)	110 (17.4)	49 (14.3)	190 (15.7)
Recipient's weight (kg)					0.0572
n	247	641	355	1243
Mean (SD)	68.0 (17.35)	66.8 (17.16)	69.2 (17.83)	67.7 (17.41)
Median	65.8	64.0	66.8	65.1
Range	34.0-130.0	36.3-140.1	27.4-122.0	27.4-140.1
Calculated body mass index of recipients (kg/m2)					0.282
n	246	635	355	1236
Mean (SD)	23.9 (5.25)	23.4 (4.83)	23.9 (5.28)	23.7 (5.05)
Median	23.2	22.8	23.0	23.0
Range	15.0-40.6	15.1-41.9	15.1-44.2	15.0-44.2

¹Chi-Square P-value.

²Kruskal-Wallis P-value.

HLA: Human leukocyte antigen.

In the unadjusted analysis at one year post-transplant (Table 3), compared to patients with no functional limitations, the hazard risk of death for patients who required some assistance was 1.03 times those with mild/no functional limitation (95%CI: 0.84-1.28; P = 0.77), and for those near death or severely disabled, it was 1.24 times compared to those with mild/no functional limitation (95%CI: 0.99–1.57; P < 0.001). At five years post-transplant, compared to recipients with no functional limitations, the hazard risk of death for patients who required some assistance was 1.03 times (95%CI: 0.83–1.27; P = 0.35), and for those near death or severely disabled, it was 1.27 times the risk of those with mild/no functional limitation (95%CI: 1.01–2.32; P < 0.042). This data aligns with the Kaplan-Meier survival curves depicted in Figures 1 and 2, illustrating that recipients with severe disability consistently demonstrated the lowest survival rates at both one-year and five-year follow-up intervals. In the adjusted analysis at one year post-transplant (Table 3), compared to patients with no functional limitations, the adjusted hazard risk of death for patients who required some assistance was 1.10 times those with mild/no functional limitation (95%CI: 0.90–1.37; P = 0.36), and for those near death or severely disabled, it was 1.56 times that of recipients with mild/no functional limitation at transplant (95%CI: 1.23–1.98; P < 0.001). At five years post-transplant, the adjusted risk of death for patients who required some assistance was 1.09 times (95%CI: 0.89–1.36; P = 0.38) and for those near death or severely disabled, it was 1.58 times (95%CI: 1.24–2.00; P < 0.001) compared to those with mild/no functional limitation.

Figure 1 Survival curves by functional status at one year. Kaplan-Meier survival curves illustrating the survival probability of up to one year of follow up for recipients categorized by their Karnofsky Performance Status functional status at the time of transplant. The categories are color-coded as follows: Red for "Requires No Assistance". Green for "Requires Some Assistance" and Blue for "Near Death or Severely Disabled". Numbers at risk at various time points are displayed below the x-axis for each category.

Figure 2 Survival curves by functional status at five years. Kaplan-Meier survival curves illustrating the survival probability of up to five years of follow up for recipients categorized by their Karnofsky Performance Status functional status at the time of transplant. The categories are color-coded as follows: Red for "Requires No Assistance". Green for "Requires Some Assistance", and Blue for "Near Death or Severely Disabled". Numbers at risk at various time points are displayed below the x-axis for each category.

Table 3 Hazard ratios for functional status in transplant outcomes.

Variable	Timepoints	Unadjusted HR			Adjusted HR1
Functional status		HR	95%CI	P value	HR	95%CI	P value
None (Ref.)	1 year	1.00			1.00
Moderate		1.03	0.84-1.28	0.77	1.10	0.90–1.37	0.36
Severe		1.24	0.99–1.57	0.066	1.56	1.23–1.98	0.00029
None (Ref.)	5 years	1.00			1.00
Moderate		1.03	0.83–1.27	0.78	1.09	0.89–1.36	0.38
Severe		1.27	1.00–1.61	0.042	1.58	1.24–2.00	0.00019

¹Adjusted hazard ratio is adjusted for recipient creatinine levels, era, weight mismatch between donor and recipient and age of donor.

HR: Hazard ratio.

When patients were stratified by time period of transplant, the adjusted hazard ratios at one and five years, for severe impairment vs mild impairment remained significant in both the 2005- 2014 [ (one year- a (hazard ratio) HR: 1.64; 95%CI: 1.19–2.24; P = 0.0021); (five years- aHR: 1.70; 95%CI: 1.24–2.32; P < 0.0001)] and 2014-2022 groups [ (one year- aHR: 1.53; 95%CI: 1.03–2.26; P = 0.034); (five years- aHR: 1.53; 95%CI: 1.04–2.26; P = 0.035)] These results are displayed in Tables 4 and 5.

Table 4 Hazard ratios for functional status in transplant outcomes for era 2005-2014.

Variables	Timepoints	Unadjusted HR			Adjusted HR1
Functional status		HR	95%CI	P value	HR	95%CI	P value
None (Ref.)	1 year	1.00			1.00
Moderate		1.11	0.86-1.44	0.4247	1.09	0.84–1.41	0.52
Severe		1.63	1.20–2.23	0.002	1.64	1.19–2.24	0.002
None (Ref.)	5 years	1.00			1.00
Moderate		1.10	0.85–1.42	0.487	1.08	0.83–1.39	0.57
Severe		1.69	1.24–2.30	< 0.0001	1.70	1.24–2.32	< 0.0001

¹Adjusted hazard ratio is adjusted for recipient creatinine levels, race matching, weight mismatch between donor and recipient and age of donor.

HR: Hazard ratio.

Table 5 Hazard ratios for functional status in transplant outcomes for era 2014-2022.

Variables	Timepoints	Unadjusted HR			Adjusted HR1
Functional status		HR	95%CI	P value	HR	95%CI	P value
None (Ref.)	1 year	1.00			1.00
Moderate		1.12	0.77–1.64	0.56	1.14	0.78–1.68	0.49
Severe		1.48	1.00–2.18	0.049	1.53	1.03–2.26	0.034
None (Ref.)	5 years	1.00			1.00
Moderate		1.11	0.76–1.62	0.59	1.13	0.77–1.66	0.54
Severe		1.48	1.00–2.19	0.046	1.53	1.04–2.26	0.035

¹Adjusted hazard ratio is adjusted for recipient creatinine levels, race matching, weight mismatch between donor and recipient and age of donor.

HR: Hazard ratio.

Among recipients with severe functional impairment, the predictive prognostic factors associated with poor outcomes are displayed in Table 6. Notably, older age, Black race, ICU stay, and baseline liver dysfunction were associated with worse survival outcomes, among recipients with severe impairment (Table 6).

Table 6 Prognostic factors for survival in patients with decreased Karnofsky Performance Status undergoing intestinal transplantation.

Variable	HR	95%CI	P value
Age	1.02	1.00-1.03	0.001
Race
Hispanic	1.21	0.66-2.23	0.53
Non-Hispanic Asian	1.20	0.287-4.98	0.81
Non-Hispanic Black	1.98	1.27-3.08	0.003
Other/Multiracial	2.77	0.38-20.46	0.32
Non-Hispanic White (Ref.)	1.00
Female gender	0.85	0.62-1.17	0.32
Donor age	1.02	1.01-1.04	0.008
Weight matched	0.78	0.55-1.09	0.008
ICU stay	2.13	1.36-3.38	0.001
Total bilirubin	1.02	1.01-1.03	0.04
Total ischemic time	1.05	0.99-1.10	0.09

ICU: Intensive care unit; HR: Hazard ratio.

As shown in Figure 3, majority of the transplant recipients had an improvement in KPS score after intestinal transplant. For patients with a low baseline KPS, those with no improvement in KPS scores post-transplant experienced a markedly lower one-year survival rate of 39%. This is in stark contrast to those who demonstrated improved KPS scores, where the survival rates soared between 98% to 100%. Similarly, in the intermediate KPS group, patients who did not show any improvement had a one-year survival of 81%, compared to those with improved KPS scores, who exhibited survival rates of 98%.

Figure 3 Distribution Karnofsky Performance Status score difference between the time of transplant and at follow-up. Distribution of the difference in Karnofsky Performance Status (KPS) scores between the time of transplant and follow-up at 3-12 months. The x-axis represents the difference in KPS scores, with negative values indicating a decrease and positive values indicating an improvement. The y-axis represents the percentage of individuals. Yellow bars indicate a decrease in KPS, while blue bars indicate an improvement. KPS: Karnofsky Performance Status.

Even among patients with a high baseline KPS, the survival outcomes varied based on KPS changes. Those with a decrease in KPS experienced a one-year survival rate of 88%, while those with stable or improved KPS maintained a notably higher survival rate of 100%. These results are shown in Table 7 and Figure 4.

Figure 4 Post transplant survival probability curves of intestinal transplant recipients by Changes in Karnofsky Performance Status after transplant. A: Among patients with low Karnofsky Performance Status (KPS). Survival curves illustrating post-transplant survival probability for recipients with low KPS at the time of transplant. The categories are based on the improvement in KPS: Red for "Improvement by ≥ 50%", Green for "Improvement by 10%-40%" and Blue for "No improvement". Numbers at risk at various time points are displayed below the x-axis for each category; B: Among patients with intermediate KPS. Survival curves illustrating post-transplant survival probability for recipients with intermediate KPS at the time of transplant. The categories are based on the improvement in KPS: Red for "Improvement by ≥ 30%, "Green for "Improvement by 10%-20%" and Blue for "No improvement". Numbers at risk at various time points are displayed below the x-axis for each category; C: Among patients with high KPS. Survival curves illustrating post-transplant survival probability for recipients with high KPS at the time of transplant. The categories are based on the change in KPS: Red for "Decrease by at least 10%" and Blue for "Improvement by ≥ 0%". Numbers at risk at various time points are displayed below the x-axis for each category.

Table 7 Post transplant survival probabilities of intestinal transplant recipients by changes in Karnofsky Performance Status after transplant, 95%CI.

Time since transplant
Changes in KPS	6 months	1 year	2 years	3 years	5 years
Low KPS at transplant
Improvement by ≥ 50% (%)	100 (100-100)	97 (92-100)	60 (46-79)	51 (37-71)	31 (19-51)
Improvement by 10%–40% (%)	98 (96-100)	80 (71-90)	57 (46-70)	38 (28-52)	17 (9.9-29)
No improvement (%)	39 (26-60)	21 (11-41)	12 (4.8-30)	12 (4.8-30)	6.1 (1.6-23)
Intermediate KPS at transplant
Improvement by ≥ 30% (%)	98 (95-100)	95 (89-100)	75 (64-87)	61 (49-75)	48 (37-63)
Improvement by 10%–20% (%)	98 (95-100)	87 (80-94)	71 (63-81)	60 (51-70)	46 (37-57)
No improvement (%)	81 (74-89)	62 (53-72)	36 (28-47)	31 (24-42)	21 (14-30)
High KPS at transplant
Decrease by at least 10% (%)	81 (68-96)	48 (34-70)	26 (14-47)	16 (7.2-36)	13 (5.2-32)
Improvement by ≥ 0% (%)	100 (100-100)	96 (92-100)	76 (66-88)	65 (54-79)	47 (36-62)

KPS: Karnofsky Performance Status.

Median survival times further substantiated these observations (Table 8). Low KPS patients at transplant with significant improvement (≥ 50%) had a median survival of 1146 days, whereas those with minimal or no improvement had markedly shorter median survival times of 177 days. Intermediate KPS patients showing significant improvement had median survivals exceeding 1700 days, whereas those with no improvement had much shorter survival times of 533 days. High KPS patients who experienced a decrease in performance had the shortest median survival times compared to those with stable or improved KPS.

Table 8 Post transplant median survival times of intestinal transplant recipients by changes in Karnofsky Performance Status after transplant, 95%CI.

Characteristic	Median survival
Low KPS at transplant
Improvement by = 50%	1146 (641-1971)
Improvement by 10%–40%	828 (651-1128)
No improvement	177 (136-243)
Intermediate KPS at transplant
Improvement by = 30%	1783 (945-2369)
Improvement by 10%–20%	1712 (1226-2132)
No improvement	533 (415-641)
High KPS at transplant
Decrease by at least 10%	304 (251-641)
Improvement by = 0%	1703 (1331-2556)

KPS: Karnofsky Performance Status.

In the sensitivity analysis using the E-value method (Table 9), the E-value for the point estimate of the hazard ratio for moderate functional impairment at 1 year was 1.43, and for the lower confidence interval, it was 1. For severe functional impairment at one year, the E-value for the point estimate was 2.49, and for the lower confidence interval, it was 1.76. This suggests that an unmeasured confounder would need to be associated with both functional impairment and one year mortality by a risk ratio of at least 1.43-fold above and beyond the measured confounders to explain away the observed association for moderate impairment, and by a risk ratio of at least 2.49-fold for severe impairment.

Table 9 E-values for the hazard ratios of functional status in transplant outcomes at one year and five years.

Functional Status	Timepoints	Adjusted HR1	E-value (point estimate)2	E-value (lower 95%CI)
Moderate	1 year	1.10	1.43	1
Severe		1.56	2.49	1.76
Moderate	5 years	1.09	1.40	1
Severe		1.58	2.54	1.79

¹Adjusted hazard ratio is adjusted for recipient creatinine levels, era, weight mismatch between donor and recipient and age of donor.

²The E-values indicate the minimum strength of association that an unmeasured confounder would need to have with both functional status and transplant outcomes, above and beyond the measured confounders, to fully explain away the observed associations.

HR: Hazard ratio.

At the five year mark, the E-value for the point estimate of the hazard ratio for moderate functional impairment was 1.40, and for the lower confidence interval, it was 1. For severe functional impairment at five years, the E-value for the point estimate was 2.54, and for the lower confidence interval, it was 1.79. This indicates that an unmeasured confounder would need to be associated with both functional impairment and five year mortality by a risk ratio of at least 1.4-times the measured confounders to explain away the observed association for moderate impairment, and by a risk ratio of at least 2.54-fold for severe impairment.

DISCUSSION

In this study, we aimed to delve deeper into the significance of pre-transplant functional status, as measured by the KPS and categorized into three groups as a prognostic indicator for post-operative mortality in intestinal transplant patients. Our findings revealed that, recipients with severe impairments had a 56% higher risk of mortality at one year (95%CI: 1.23–1.98; P < 0.001) and 58% higher risk of mortality at five years, when compared to recipients with no impairments (95%CI: 1.24–2.00; P < 0.001).

Additionally, our study showed that recipients with improved KPS after transplant have better outcomes, with up to six times longer median survival times, compared with recipients with worsening KPS after transplant. This finding that lower KPS categories pretransplant, and worsening KPS after transplant are linked to increased post-operative mortality in intestinal transplant patients, is consistent with findings with heart, liver, and kidney transplant recipients[8,12,22,23]. Grimm et al[22] conducted a comparable analysis of the UNOS database, demonstrating that a lower preoperative KPS score in lung transplant patients was linked to a higher mortality rate at one year[22]. Similarly, Shaw and Pickett[24] demonstrated that functional status serves as a potent predictor of survival at 30 days and one year following heart transplantation[24]. In the context of liver transplantation, Thuluvath et al[12] analyzed transplant patients in the UNOS database and found that both graft and patient survival rates were notably lower for liver transplant recipients with lower KPS scores[12]. Thuluvath et al[14] also found that, recipients with decreased KPS after liver transplantation had significantly worse survival outcomes. In kidney transplants, Bui et al[23] also discovered that deceased donor kidney transplant recipients with higher KPS scores had a greater likelihood of survival, even when receiving lower-quality organs, as determined by the kidney donor profile index[23]. Our investigation supports these findings and expands their applicability to intestinal transplantation.

The significant relationship between pretransplant functional status and posttransplant survival can be attributed to various factors. Patients requiring complete assistance before surgery are more likely to be weakened and possess less physiological reserve than those who do not need assistance[25]. Our comparison of baseline characteristics showed that individuals with lower functional status had a significantly higher likelihood of being on life support before transplantation. Moreover, approximately 13% of the patients with lower functional status also needed intensive care unit stay before undergoing intestinal transplantation.

Gerlach et al[26] in a multicenter study found that early identification and transplantation of patients with Crohn's disease and intestinal failure was associated with an increase in post-transplant KPS, patient and graft survival. They utilized a modified American Gastroenterology Association scoring system for Intestinal transplantation that enhances the identification and early referral of patients with severe and transplant-eligible CD[26]. Our findings also show that transplantation results in an improvement in post-transplant KPS and survival outcomes in the majority of eligible patients, highlighting the need for appropriate selection of candidates to maximize outcome benefits.

Although the concentration of these high-risk variables in the group with lower functional status undoubtedly played a role in the observed association with poorer survival outcomes, multivariable analysis indicated that the influence of functional status on outcomes persisted even after accounting for these factors, implying a likely independent relationship. Furthermore, the sensitivity analyses suggest that our findings are relatively robust to potential unmeasured confounding. The E-values indicate that a substantial level of confounding would be needed to explain away the observed associations between functional impairment and post-transplant mortality, particularly for severe impairment at the 1-year mark.

The E-value for the point estimate of severe functional impairment and mortality at 1 year was larger than the risk ratios reported in two previous studies by O^,Súilleabháin et al[27] and Sáenz et al[28] that estimated the risk of reduced functional status on mortality to be 1.3 and 2.16, respectively. This suggests that even if an unmeasured confounder associated with both functional impairment and mortality was present, it would need to have a stronger association than those reported in these previous studies to fully account for our findings.

These results highlight the potential importance of assessing functional status in patients undergoing intestinal transplant and suggest that interventions to improve functional status could potentially improve post-transplant outcomes. However, further research is needed to investigate the mechanisms underlying these associations and to develop effective interventions. Future studies with the ability to measure additional potential confounders could help to further clarify these associations.

Our observations, combined with the ease of calculating the KPS, highlight the potential value of preoperative functional status as a risk assessment tool for candidates undergoing intestinal transplantation.

As the demand for intestinal transplants continues to rise, it becomes increasingly important to identify factors that can be optimized to improve patient outcomes. Our results are meant to reiterate the potential for preoperative intervention and optimization in patients with low functional status, particularly when the causes of declining KPS scores can be identified and addressed. This highlights the need for a multidisciplinary approach that involves surgeons, transplant teams, physical therapists, and nutritionists to develop tailored pre-transplant rehabilitation programs.

The impact of functional status on post-transplant outcomes extends beyond survival rates. It also has implications for the patient's quality of life, recovery time, and long-term morbidity[25,29,30]. By improving functional status, transplant centers can not only improve survival rates but also enhance the overall well-being of intestinal transplant patients[25,29,30]. Moreover, understanding a patient's pre-transplant functional status can help transplant teams anticipate postoperative care requirements, facilitate communication among the multidisciplinary team, and coordinate resources for postoperative critical care, rehabilitation, and potential longer-term occupational therapy[31-35].

Our study also contributes to the growing body of evidence supporting the potential benefits of prehabilitation in transplant patients. Prehabilitation, which involves physical therapy, nutritional support, and other interventions designed to improve functional status before surgery, has been shown to be effective in improving postoperative recovery and outcomes in major abdominal surgery[36,37]. Further research on prehabilitation specific to intestinal transplant patients is warranted to determine its effectiveness in this population.

Our study has several limitations that should be considered when interpreting the results. First, as this is an observational study, our findings may be subject to unmeasured confounding. However, we adjusted for a comprehensive list of potential confounders in our analyses, including donor age, sex, ethnicity, ischemic time, as well as recipient age, sex, and ethnicity. Thus, although we acknowledge the potential for residual confounding, we have made every effort to minimize its impact on our findings. Second, the extended time frame over which data was collected could introduce issues such as changes in allocation policies, eligibility criteria, and follow-up periods, all of which may affect the study outcomes. Third, one potential limitation of our study is the selection bias inherent in the database, as the functional status pre-transplant KPS can directly impact the likelihood of a patient being selected for transplantation as well as the post-transplant outcomes. Such bias might obscure the independent effect of the KPS on post-transplant outcomes, which is the primary interest of our study. However, it is important to note that this bias is inherent in all observational studies using secondary data and is not unique to our study. To address this potential limitation, we conducted a sensitivity analysis using the E-value method to assess the robustness of our results to potential unmeasured confounding. The E-value provides an estimate of the minimum strength of association that an unmeasured confounder would need to have with both the exposure (KPS) and the outcome (post-transplant survival), above and beyond the measured confounders, to fully explain away the observed associations. Our E-value analysis demonstrated that the associations we found between pre-transplant functional status and post-transplant mortality are robust to potential unmeasured confounding, at least for some categories of functional status. Despite these limitations, our study possesses several notable strengths that lend credibility to our conclusions. First, we used a dataset comprising the most extensive cohort of intestinal transplant patients, which allowed for increased statistical power and precision. Second, our study is the first to examine the impact of RFS on short and long-term intestinal transplant outcomes in United States adults, which represents an innovative and critical area of research with potential implications for clinical decision-making and patient outcome improvement. By utilizing data from the UNOS database between 2005 and 2022, our study offers a comprehensive understanding of the influence of RFS on survival outcomes in adult intestinal transplant patients and the factors associated with them.

CONCLUSION

Our study highlights the importance of pre- and post-transplant functional status in determining post-operative outcomes in intestinal transplant patients, as evidenced by the increased risk of mortality at one and five years in severe functionally impaired patients compared to those with mild/no functional impairment, as well in those with poor post transplant survival outcomes and emphasizes the need for preoperative optimization and intervention to improve patients' functional status. This comprehensive approach to patient care has the potential to significantly enhance not only survival rates but also the overall quality of life for intestinal transplant recipients. Future studies should continue to investigate the impact of functional status on outcomes and explore the potential of rehabilitation in the context of intestinal transplantation, with the aim of informing the development of more effective allocation policies and pre-transplant rehabilitation strategies.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge the UNOS for providing the data used in this study.