Nutritional Risks in Patients Undergoing Total Pancreatectomy Islet AutoTransplantation in the POST Consortium

Abstract

Background and Aims

Total pancreatectomy with islet autotransplantation (TPIAT) can relieve pain for individuals with acute recurrent or chronic pancreatitis. However, TPIAT may increase the risk of poor nutritional status with complete exocrine pancreatic insufficiency, partial duodenectomy, and intestinal reconstruction. Our study’s objective was to evaluate nutritional status, anthropometrics, and vitamin levels before and after TPIAT.

Methods

The multicenter Prospective Observational Study of TPIAT (POST) collects measures including vitamins A, D, and E levels, pancreatic enzyme dose, and multivitamin (MVI) administration before and 1-year after TPIAT. Using these data, we studied nutritional and vitamin status before and after TPIAT.

Results

348 TPIAT recipients were included (68% adult, 37% male, 93% Caucasian). In paired analyses at 1-year follow-up, vitamin A was low in 23% (vs 9% pre-TPIAT, p < 0.001); vitamin E was low in 11% (vs 5% pre-TPIAT, p = 0.066), and 19% had vitamin D deficiency (vs 12% pre-TPIAT, p = 0.035). Taking a fat-soluble multivitamin (pancreatic MVI) was associated with lower risk for vitamin D deficiency (p = 0.002). Adults were less likely to be on a pancreatic MVI at follow-up (34% vs 66% respectively, p < 0.001). Enzyme dosing was adequate. More adults versus children were overweight or underweight pre- and post-TPIAT. Underweight status was associated with vitamin A (p = 0.014) and E (p = 0.02) deficiency at follow-up.

Conclusions

Prevalence of fat-soluble vitamin deficiencies increased after TPIAT, especially if underweight. We strongly advocate that all TPIAT recipients have close post-operative nutritional monitoring, including vitamin levels. Pancreatic MVIs should be given to minimize risk of developing deficiencies.

Introduction

Individuals with acute recurrent pancreatitis (ARP) and chronic pancreatitis (CP) have multifactorial risk factors for poor nutritional status and fat-soluble vitamin (FSV) deficiencies. Key amongst those factors is the development of exocrine pancreatic insufficiency (EPI) and the resultant malabsorption it produces. In those with CP, the rate of vitamin E deficiency was up to 8 times higher and vitamin A deficiency was 2–4 times higher when compared to normative population data for children¹ or a control cohort for adults.² Vitamin D deficiency rates were similar between these groups and reflective of increased rates in the population as a whole.³ Increased rates of FSV deficiencies can be seen even when individuals are treated with appropriate doses of pancreatic enzyme replacement therapy (PERT).^4,5

Total pancreatectomy with islet autotransplantation (TPIAT) can be considered for selected individuals with ARP or CP for whom maximal medical, endoscopic, and/or other surgical procedures have failed to improve chronic pain and impaired quality of life. While pain relief and improved quality of life have been demonstrated,^6–9 few studies have addressed nutritional status after TPIAT. TPIAT may increase the risk of poor nutritional status due to complete iatrogenic EPI resulting from the pancreatectomy. Because TPIAT also involves partial duodenectomy and intestinal reconstruction, FSV deficiencies may also be more likely to occur if sites of absorption are largely bypassed as has been seen in certain bariatric surgeries.^10,11 Therefore, lifelong PERT and FSV supplementation are needed.

Increasing prevalence of FSV deficiency after TPIAT has been demonstrated in a single-center cohort of pediatric TPIAT recipients,¹ however, no official consensus has been reached regarding FSV monitoring. In addition, standard guidelines regarding vitamin supplementation after TPIAT do not exist and current practice is highly variable.

Using multicenter data from the Prospective Observational Study of TPIAT (POST), we studied nutritional outcomes and FSV status before and after TPIAT, with the goal to determine the frequency of, and risk factors for, nutritional deficiencies before and after TPIAT. We structured our analyses with two aims: 1) to compare nutritional status and FSV deficiencies in adults vs children before and at 6 months and 1 year after TPIAT; and 2) to determine the trajectories in nutritional status and FSV before and after TPIAT.

Methods

Study Design

The current study used available data from POST, a national multi-center registry study of pediatric and adult TPIAT recipients from 13 collaborating centers. POST prospectively collected data on pancreatitis and surgical related outcomes before TPIAT and after TPIAT as previously described.¹² All participating centers followed their own center’s ethical committee guidelines and received approval from their institutional review board, with informed consent or parental consent plus patient assent as age-appropriate. For the current analyses, we included data from children (< 18 years old) and adult (≥ 18 years old) participants enrolled from 1/2017 to 8/2021 detailing age at transplant, sex, ethnicity, pancreatitis history, genetic risk factors for pancreatitis (when tested), anthropometric measurements, clinical diagnosis of EPI before surgery, FSV levels, PERT usage and dosing, and use of vitamin supplementation before TPIAT and at 6 months and 1 year after TPIAT. Children and adults were compared to each other first for better generalizability of results between pediatric vs adult care centers, but then also compared to themselves longitudinally (in paired analyses) pre- to post-TPIAT to meet our aims as stated above.

Categorization of Fat-soluble Vitamins

POST collects clinically available laboratory data for vitamin A, E, and 25(OH) D levels. Vitamin K is not routinely assessed in clinical care given variability of assay from intake, gut microbiome, and other factors; INR can be a useful functional assay, but is also non-specific. At each time point, FSV levels were classified as low or normal / high based on each institution’s laboratory-specific reference ranges (as absolute values may differ between laboratories), or as unknown.

Vitamin D status was additionally classified as insufficient (< 30 ng/mL) or deficient (< 20 ng/mL) as per the Institute of Medicine definition.¹³

Vitamin supplementation history was collected, and defined as being in one of four categories for analysis purposes: (1) taking a pancreatic-specific multivitamin (MVI) preparation, which contains water-soluble formations of the FSV specifically designed for improved absorption in the setting of pancreatic insufficiency and fat malabsorption; (2) taking a standard MVI; (3) taking any type of individual vitamin supplements without an MVI; or (4) taking no vitamin supplements. Participants were assigned to the first category they fit into according to the ordering (1)-(4); for example, a participant taking a pancreatic MVI and a standard MVI would be assigned to category (1).

Categorization of Nutritional Status

Body mass index (BMI) was used as a proxy for nutritional status. Adults were grouped according to absolute values into underweight (BMI < 18.5 kg/m2), normal weight (BMI ≥ 18.5 to < 25 kg/m2), or overweight (BMI ≥ 25 kg/m2) categories. Children were grouped according to percentiles into underweight (BMI < 5th%), normal weight (BMI ≥ 5th% to < 85th%), or overweight (≥ 85th%) categories.

Surgical Technique of TPIAT

Individuals were enrolled in POST based on each center’s assessment of eligibility for TPIAT. The surgical technique for TPIAT has been well-described,^9,14 including removal of the pancreas, spleen, gallbladder, appendix, and duodenum (with or without pylorus preservation). The gastrointestinal tract is then reconstructed with a Roux-en-Y hepatico- or choledocho-jejunostomy and duodenojejunostomy or gastrojejunostomy.¹⁵ Isolated islets are generally infused into the portal vein to the liver, with some patients having islets infused elsewhere.

Statistical Analyses

Fisher’s exact test was used to compare categorical variables between groups and Welch’s t-test was used to compare numeric variables between groups. For analyses comparing groups at a single time point, individuals with data available at that time point were included. For comparing distributions at baseline versus 6 months and baseline versus 12 months, McNemar’s test and the paired t-test were used for categorical and numeric variables, respectively, to account for paired observations (observations from the same participant at two time points). For paired analyses, only individuals with data available at both time points were included. P-values less than 0.05 were considered significant.

Because of the ‘living’ nature of the POST database, not all participants have completed follow up at each time point and not all participants may have a full complement of biochemical assessments at each time point given variations in clinical practice. Total ‘n’ for each assessment is detailed in the results tables.

Results

Demographics

A total of 348 TPIAT recipients were included, with 235 (68%) adults and 113 children (Table 1). Overall, 37% of recipients were male and 93% were Caucasian.

Table 1.

Demographic Characteristics of the Entire Cohort

	Overall n = 348		Adults n = 235		Children n = 113		p-value
Age	27 (15, 42)		38 (27, 49)		12 (8, 14)		NA
Sex							0.2
Male	129	37%	81	34%	48	42%
Female	219	63%	154	66%	65	58%
Ethnicity							0.4
White/Caucasian	322	93%	216	92%	106	94%
Asian	5	1%	4	2%	1	1%
Black/African American	8	2%	7	3%	1	1%
Native Hawaiian/Pacific Islander	2	1%	2	1%	0	0%
American Indian/Alaska Native	1	0.3%	0	0%	1	1%
Mixed race	9	3%	6	3%	3	3%
Don’t know/declined to answer	1	0.3%	0	0%	1	1%
Chronic pancreatitis prior to TPIAT							0.7
Recurrent AP only	41/339	12%	26/226	12%	15/113	13%
Chronic pancreatitis	298/339	88%	200/226	88%	98/113	87%
Duration of pancreatitis (years)*	4.6	(2.4, 9.5)	5.8	(3.3, 12.7)	2.8	(1.4, 5.7)	< 0.001
Exocrine pancreatic insufficiency**	126/346	39%	83/233	36%	43/113	38%	0.7
Fecal elastase < 100ug/g	54/160	34%	30/85	35%	24/75	32%	0.7
Genetic risk factors
Any genetic tests done	274/339	81%	161/226	71%	113/113	100%	< 0.001
Genetic risk factor***	203/339	60%	107/226	47%	96/113	85%	< 0.001
PRSS1***	74/339	22%	25/226	11%	49/113	43%	< 0.001
SPINK1***	62/339	18%	35/226	15%	27/113	24%	0.073
CFTR***	104/339	31%	65/226	29%	39/113	35%	0.318

Results displayed as n (%) or median (IQR)

Bold values indicate p < 0.05

^*n = 339, 226, and 113 for overall, adults, and children

^**Exocrine pancreatic insufficiency or dysfunction diagnosed by the treating physician

^***Denominators include those not tested for genetic risk factors: 78 adults and 8 children were not tested for PRSS1; 80 adults and 14 children were not tested for SPINK1; 80 adults and 13 children were not tested for CFTR

The groups were largely similar except for a few key differences. Adults had a two-fold longer duration of diagnosed disease with median 5.8 years vs 2.8 years for children (p < 0.001). Adults were less likely to have genetic risk factors assessed (p < 0.001) and less likely to have a genetic risk factor present (p < 0.001) compared to children. The majority of children (85%) had underlying genetic risk factor(s), split between PRSS1, SPINK1, and CFTR.

Assessment of Nutritional Status in Children vs Adults before and after TPIAT

For each outcome of interest, we compared the nutritional outcomes in children and adults at each time point (pre-TPIAT, 6 months after TPIAT, and 1 year after TPIAT) using all available data.

BMI in Children and Adults

Before TPIAT, adults were more often overweight compared to children (52% vs 36% respectively) or underweight (4% vs 2% respectively), and fewer adults than children were normal weight (43% vs 62% respectively) (p = 0.004, Fig. 1).

Fig. 1.

BMI in children and adults before and after TPIAT. P-values are for a test of the difference in weight distribution between children vs adults at each time. Adults were more often underweight or overweight at baseline and at 1 year after TPIAT vs children. Adults (≥ 18 years old), children (< 18 years old). Black columns: Underweight (adults BMI < 18.5 kg/m², children BMI %ile < 5th). White columns: Normal weight (adults BMI ≥ 18.5 to < 25 kg/m², children BMI %ile ≥ 5th to < 85th). Gray columns: Overweight (adults BMI ≥ 25 kg/m², children BMI %ile ≥ 85th). BMI = body mass index. TPIAT = total pancreatectomy and islet autotransplantation

BMI status remained significantly different (p = 0.004) at 1-year follow-up (Fig. 1), with again more adults than children overweight (35% vs 24% respectively) or underweight (10% vs 2% respectively), and again fewer adults than children normal weight (54% vs 73% respectively).

Vitamin Deficiencies in Children and Adults

Prevalence of vitamin A deficiency did not differ significantly between children and adults at baseline, but prevalence was significantly higher in adults compared to children by 6 months and 1 year (p < 0.001) (Table 2). Prevalence of vitamin D deficiency was significantly higher among adults at baseline (p = 0.011) and 6-month follow-up (p = 0.043), but did not differ significantly between the two groups at 1-year follow-up due to the increased frequency of vitamin D deficiency in children. Vitamin E deficiency did not differ between groups.

Table 2.

Fat-soluble vitamin deficiencies in children vs adults at each time point

	adults		children		p-value
Pre-TPIAT
Low vitamin A	19/201	10%	6/101	6%	0.4
Low vitamin E	14/204	7%	6/99	6%	> 0.9
Vitamin D insufficiency (< 30 ng/mL)*	101/213	47%	48/104	46%	> 0.9
Vitamin D deficiency (< 20 ng/mL)	34/213	16%	6/104	6%	0.011
6 months after TPIAT
Low vitamin A	47/152	31%	9/98	9%	< 0.001
Low vitamin E	20/152	13%	6/96	6%	0.093
Vitamin D insufficiency (< 30 ng/mL)*	69/166	42%	37/99	37%	0.5
Vitamin D deficiency (< 20 ng/mL)	29/166	17%	8/99	8%	0.043
1 year after TPIAT
Low vitamin A	47/140	34%	8/76	11%	< 0.001
Low vitamin E	19/139	14%	5/77	7%	0.12
Vitamin D insufficiency (< 30 ng/mL)*	65/144	45%	42/79	53%	0.3
Vitamin D deficiency (< 20 ng/mL)	27/144	19%	15/79	19%	> 0.9

Bold values indicate p < 0.05

TPIAT total pancreatectomy and islet autotransplantation; adults (≥ 18 years old), children < 18 years old

^*includes insufficiency and deficiency together, thus all with vitamin D values below 30 ng/mL

Associations with FSV at 1 year after TPIAT

At the 1-year follow-up, those who were underweight were more likely to be vitamin A (p = 0.014) or vitamin E (p = 0.02) deficient (Table 3). TPIAT recipients on a pancreatic MVI had decreased rates of vitamin D deficiency compared to those on standard MVI, individual vitamins, or no supplementation (p = 0.002) (Table 3). Those on a pancreatic MVI also tended to have lower rates of vitamin A and E deficiency when compared to patients on no vitamins or a standard MVI, but tests for differences among all 4 groups were not statistically significant (Table 3).

Table 3.

Prevalence of fat-soluble vitamin deficiencies at 1 year after TPIAT by weight status and vitamin supplementation

	Underweight	Normal weight	Overweight	p-value for BMI	PancreaticMVI	Standard MVI	Individual vitamins only	No vitamins	p-value for MVI
Low vitamin A	8/15 (53%)	36/135 (27%)	11/65 (17%)	0.014	18/99 (18%)	21/65 (32%)	5/20 (25%)	10/29 (34%)	0.12
Low vitamin E	5/14 (36%)	14/135 (10%)	5/66 (8%)	0.020	8/100 (8%)	10/64 (16%)	1/20 (5%)	5/29 (17%)	0.3
Vitamin D insufficiency (< 30 ng/ mL)*	8/15 (53%)	64/139 (46%)	35/68 (51%)	0.7	43/101 (43%)	33/67 (49%)	9/21(43%)	18/28 (64%)	0.2
Vitamin D deficiency (< 20 ng/ mL)	4/15 (27%)	27/135 (19%)	11/68 (16%)	0.6	12/101 (12%)	11/67 (16%)	6/21 (29%)	12/28 (43%)	0.002

Overweight (adults BMI ≥ 25 kg/m2, children BMI %ile ≥ 85th)

Normal weight (adults BMI ≥ 18.5 to < 25, children BMI %ile ≥ 5th to < 85th)

Underweight (adults BMI < 18.5 kg/m2, children BMI %ile < 5th)

Bold values indicate p < 0.05

MVI multivitamin

^*Includes insufficiency and deficiency together, thus all with vitamin D values below 30 ng/mL

Participants who had a pylorus-resecting surgery were more often vitamin A deficient (43%, vs 20% with pylorus-preserving surgery, p = 0.002) and underweight (13%, vs 4.5% in the pylorus-preserved group, p = 0.036). When limited to a sub-analysis of adults only, vitamin A deficiency remained significantly greater with pylorus resection (p = 0.025). Vitamin A deficiency occurred most frequently in those with Roux-en-Y gastrojejunostomy (49%) and least often with Roux-en-Y duodenojejunostomy (16%, p < 0.001) with similar findings when considering adults only (p = 0.009). There were no other statistically significant associations of surgical approach with FSV deficiency, although vitamin D deficiency tended to occur least often with Roux-en-Y duodenojejunostomy (p = 0.081).

Vitamin Supplementation

Similar usage rates of standard vs pancreatic MVI were observed at baseline before TPIAT in children and adults (Supplemental Table 1).

At 1-year follow-up, children were more likely to be using a pancreatic MVI (66% vs adults 34%; p < 0.001) while adults were more likely to be using a standard MVI (42% vs children 23%; p = 0.002). Adults were more likely to be on additional vitamin D (40% versus 24%, p= 0.013) than children.

PERT Usage

Adults trended towards higher PERT utilization before surgery for treatment of EPI or for treatment of pancreatitis, with 70% of adults vs 59% of children on PERT before TPIAT (p = 0.052). PERT dosage (as lipase units/kg/meal) was otherwise similar between the groups at baseline and follow-up, although doses were approximately 50% higher at follow-up. At 1 year after TPIAT, adults were on a median of 1525 (IQR 1074, 1984) lipase units/kg/meal while children were on a median of 1450 (IQR 1189, 1924) lipase units/kg/meal. High self-reported PERT utilization was seen in both adults (97%) and children (100%) at follow-up (Supplemental Table 1).

Changes in Nutritional Status after TPIAT, Compared to before TPIAT, in Children and Adults

For each outcome of interest, we compared the nutritional outcomes in children and adults separately post-TPIAT compared to before TPIAT. For these analyses, we included only those recipients with paired data for measures before TPIAT and after TPIAT.

Trend in BMI after TPIAT

BMI Trajectory in Children

Underweight was rare in children, affecting only 2% at baseline, 4% at the 6-month follow-up, and no children at 1 year (Table 4). The fraction with normal weight steadily improved from baseline (62%) to 6 months (70%) to 1 year (76%), while the fraction overweight steadily decreased from baseline (36%) to 6 months (26%) to 1 year (24%). Trends versus baseline were significant at 6 months (p = 0.015) but not at 1 year (p = 0.099).

Table 4.

Trajectory from pre-TPIAT to post-TPIAT analyzing paired pre-TPIAT and post-TPIAT (6 month or 1 year) results

	Pre-TPIAT		6 months after TPIAT		p-valuea	12 months after TPIAT		p-valueb
OVERALL COHORT
BMI status					< 0.001 **			< 0.001
Underweight	11/306	4%	19/306	6%		18/268	7%
Normal weight	150/306	49%	191/306	62%		165/268	62%
Overweight	145/306	47%	96/306	31%		85/268	32%
Fat-soluble vitamin status
Low vitamin A	20/232	9%	52/232	22%	< 0.001	45/194	23%	< 0.001
Low vitamin E	11/230	5%	23/230	10%	0.038	22/193	11%	0.066
Vitamin D insufficiency (< 30 ng/mL)*	118/251	47%	96/251	38%	0.020	100/209	48%	0.6
Vitamin D deficiency (< 20 ng/mL)	29/251	12%	33/251	13%	0.6	39/209	19%	0.035
CHILDREN (< 18 YEARS)
BMI status					0.015 **			0.099**
Underweight	2/108	2%	4/108	4%		0/90	0%
Normal weight	67/108	62%	76/108	70%		68/90	76%
Overweight	39/108	36%	28/108	26%		22/90	24%
Fat-soluble vitamin status
Low vitamin A	6/91	7%	9/91	10%	0.6	7/70	10%	> 0.9
Low vitamin E	4/87	5%	5/87	6%	> 0.9	5/69	7%	0.7
Vitamin D insufficiency (< 30 ng/mL)*	44/94	47%	34/94	36%	0.11	40/75	53%	0.3
Vitamin D deficiency (< 20 ng/mL)	6/94	6%	7/94	7%	> 0.9	15/75	20%	0.016
ADULTS (> / = 18 YEARS)
BMI status					< 0.001			< 0.001
Underweight	9/198	5%	15/198	8%		18/178	10%
Normal weight	83/198	42%	115/198	58%		97/178	54%
Overweight	106/198	54%	68/198	34%		63/178	35%
Fat-soluble vitamin status
Low vitamin A	14/141	10%	43/141	30%	< 0.001	38/124	31%	< 0.001
Low vitamin E	7/143	5%	18/143	13%	0.029	17/124	14%	0.1
Vitamin D insufficiency (< 30 ng/mL)*	74/157	47%	62/157	39%	0.12	60/134	45%	> 0.9
Vitamin D deficiency (< 20 ng/mL)	23/157	15%	26/157	17%	0.7	24/134	18%	0.5

Adults (≥ 18 years old), children < 18 years old

Overweight (adults BMI ≥ 25 kg/m², children BMI %ile ≥ 85th)

Normal weight (adults BMI ≥ 18.5 to < 25 kg/m², children BMI %ile ≥ 5th to < 85th)

Underweight (adults BMI < 18.5 kg/m², children BMI %ile < 5th)

Bold values indicate p < 0.05

TPIAT total pancreatectomy and islet autotransplantation

BMI body mass index

^*Includes insufficiency and deficiency together, thus all with vitamin D values below 30 ng/mL

^*a: p-value comparing pre-TPIAT to 6 months

^*b: p-value comparing pre-TPIAT to 1 year

^**p-value is for normal or underweight versus overweight

BMI Trajectory in Adults

The prevalence of underweight status increased after TPIAT in adults. At 1 year, twice as many were underweight compared to baseline (10% vs 5%, Table 4). Over half of adults were overweight at baseline (54%), with fewer adults overweight at 6 months (34%), with a stable rate at 1 year (35%). As a result, the prevalence of normal weight status in adults increased after TPIAT (from 42 to 54%). Trends at 6 months and 1 year versus baseline were significant (both p < 0.001). For both children and adults, change in weight category from pre-TPIAT is displayed in Supplemental Table 2.

Trend in FSV Deficiencies

FSV Levels in Children

Prevalence of vitamin A and E deficiencies did not change significantly from baseline in children (Table 4). Changes in the prevalence of vitamin D insufficiency were not statistically significant from baseline to 6-month and to 1-year follow-up but vitamin D insufficiency was generally common (36–53% at each time point). Prevalence of vitamin D deficiency significantly increased from baseline to 1-year follow-up (6% at baseline vs 20% at 1 year, p = 0.016).

FSV Levels in Adults

Prevalence of vitamin A deficiency was 3 times more common at 6 months (30%) and 1 year (31%) compared to pre-TPIAT baseline (10%) (p < 0.001) (Table 4). Prevalence of vitamin E deficiency also increased from baseline to 6 months (5% vs 13%, p = 0.029), but did not change significantly from baseline to 1 year. Prevalence of vitamin D insufficiency and vitamin D deficiency were high and did not differ post-TPIAT compared to baseline (39–47% prevalence for insufficiency, 15–18% prevalence for deficiency). For both children and adults, change in fat-soluble vitamin status from pre-TPIAT is displayed in Supplemental Table 3.

Discussion

This analysis from POST is the first multicenter study to evaluate nutritional status in children and adults undergoing TPIAT, including their outcomes after surgical intervention. Adults were more likely to be overweight or underweight, both before and after TPIAT, and had a higher rate of vitamin A deficiency after surgery compared to children. Fat-soluble vitamin deficiencies were overall common and increased after TPIAT, despite adequate enzyme replacement, likely due to the complete pancreatic exocrine insufficiency and alterations in gastrointestinal anatomy based on surgical technique. Taking a pancreatic-specific multivitamin reduced the risk for FSV deficiencies, particularly vitamin D deficiency. Overall, these findings support the importance of close nutritional monitoring after TPIAT, including regular biochemical assessments of FSV, appropriate pancreatic enzyme supplementation, and the need for adequate administration of fat-soluble multivitamins designed for use in pancreatic insufficiency. Our general recommendations for monitoring and supplementation are included (Table 5) and guided by previous recommendations and research for people with cystic fibrosis.^16–18

Table 5.

TPIAT nutritional monitoring and vitamin recommendations

Component	Timeframe
Anthropometrics: weight, height, BMI at minimum	Baseline; 3, 6, 12 months and then annually post-op; sooner if concerns
Fat soluble vitamins: dosing, adherence, serum levels	Baseline; 3, 6, 12 months and then annually post-op; plus every 2–3 months as deficiencies are identified and treated
Pancreatic enzymes: regimen, dosing, adherence	3, 6, 12 months and then annually post-op; sooner if concerns
Individual vitamins	Daily dosing recommendations
A*	1–3 years: 1500 mcg; 4–8 years: 1500 – 3000 mcg; > 8 years: 3000 mcg
E*	1–3 years: 80 – 150 mg; 4–8 years: 100 – 200 mg; > 8 years: 200 – 400 mg
D**	1–10 years: 20 – 25 mcg (800 – 1000 units); > 10 years: 20 – 50 mcg (800 – 2000 units)
K*	1–3 years: 0.3 – 0.5 mg; 4–8 years: 0.3 – 0.5 mg; > 8 years: 0.3 – 0.5 mg

BMI body mass index

mcg micrograms

mg milligrams

^*See Borowitz D, Baker RD, Stallings V. J Pediatr Gastroenterol Nutr 2002; 35: 246 Yankaskas JR, Marshall BC, Sufian B, et al. Chest 2004; 125 (1 Suppl):1S

^**See Tangpricha V, Kelly A, Stephenson K, et al. J Clin Endocrinol Metab 2012; 97:1082

The prevalence of fat-soluble vitamin deficiencies increased after TPIAT. At 1 year after TPIAT, nearly half of all patients were insufficient or deficient in vitamin D, with about 1 in 5 having a deficiency (vitamin D < 20 mcg/dL). Vitamin A deficiency was present in 34% of adults and 11% of children, and vitamin E deficiency in 14% of adults and 7% of children. Rates of vitamin A deficiency in adults tripled while rates of vitamin D deficiency in children more than tripled. The increasing risk for fat-soluble vitamin deficiency after TPIAT is likely attributable to the complete exocrine pancreatic insufficiency induced by the pancreatectomy, with possible other contributing factors including altered gastrointestinal anatomy. Children were less likely to have low vitamin A and trended towards having low vitamin E less often. This might be attributable to different surgical approaches in children,¹⁵ with pylorus resection and gastrojejunostomy performed more often in adults, vs pylorus-sparing duodenojejunostomy in most children. Pyloric resection and more rapid intestinal transit due to dumping syndrome may decrease absorption.¹⁹ It could also be due to closer nutritional monitoring and more aggressive supplementation with pancreatic-specific multivitamin preparations by pediatric practitioners.

Pancreatic-specific multivitamin preparations, which contain water-soluble formulations of the fat-soluble vitamins, are specifically designed for improved absorption of these micronutrients in patients with EPI. Several companies currently have preparations available; dosing will depend on brand, formulation, and patient age. Only one-third of adults and two-thirds of children were on pancreatic-specific multivitamins at follow-up, with the remainder on standard multivitamins or no vitamin supplementation at all. Those taking pancreatic-specific multivitamin preparations were least likely to have vitamin D deficiency, and also trended toward lower rates of vitamin A and E deficiency, particularly compared to standard MVI users. It is unclear why more patients were not treated with pancreatic-specific MVI preparations, but this could be related to high cost, varying insurance coverage, and/or lack of provider awareness regarding the advantages of pancreatic-specific MVIs for those with EPI.

Fat-soluble vitamin deficiencies were common despite the median PERT dosing (as lipase units/kg/meal) in the middle range of recommended, and nearly universal self-reported PERT usage or adherence. PERT will never be a perfect replacement for the natural enzymatic output of the pancreas that can respond more precisely to the macronutrient content of food entering the duodenum; therefore, some degree of malabsorption will always occur with PERT, increasing the risk for fat-soluble vitamin deficiencies. At baseline, even though over one-third of the overall cohort had severe EPI, < 10% of children and adults were on a pancreatic multivitamin; correction of deficiencies prior to TPIAT should be pursued as this will be more difficult post-operatively in the setting of now complete EPI.

In assessing nutritional status, at baseline almost half the cohort was overweight/obese, similar to other studies in CP.²⁰ BMI was used as a proxy of nutritional status, with recognition that this is an imperfect assessment, as are other biochemical markers. Triceps skin fold measurements and mid upper arm circumference were not routinely captured or monitored. Improvement in BMI (particularly in adults) was seen over time with fewer overweight and more normal weight individuals. While a small increase in the number of underweight individuals was seen, most patients are not grossly underweight before or after TPIAT. However, 10% of adults were underweight 1 year after TPIAT, and these individuals were much more likely to have low vitamin A and vitamin E.

This study has some limitations, including its ongoing prospective and observational nature, with these results reflective of interim analyses. Not all participants had completed follow-up, nor had full complement of vitamin levels, at each follow-up time point. However, given the large amount of data already available with about two-thirds of the cohort completing follow-up, we anticipate that these results will be reasonably reflective of the entire study population. While all available data were used for comparing adults and children, analyses of change from baseline to after TPIAT was restricted to those with paired pre and post assessments. POST is capturing data from TPIAT recipients at 13 major tertiary referral centers, with inherent differences in practice and potential center biases. Further analyses regarding interactions between sex and outcomes were not conducted, and could be considered for future studies. BMI as above was used as a proxy of nutritional status, which does not always correlate with lean body mass; other measures to study lean body mass should be included in future research. Children and adults were compared in some analyses without adjusting for variations in surgical technique, BMI, underlying etiology of pancreatitis, among others. PERT and vitamin supplementation data were also reliant on participant self-reported information, and may overestimate usage or adherence. This study does not capture the season of measurement for vitamin D assays, nor does it contrast vitamin E measurements against lipid profiles. Centralized analysis of fat-soluble vitamin levels was beyond the scope of the current study, and thus the current analyses relied upon categorical characterization of FSV as normal or low. Future research could assess clinical manifestations of deficiencies that we were unable to include in this short-term study.

Conclusions

The now complete EPI and altered intestinal anatomy after TPIAT increases the risk for poor nutritional status and fat-soluble vitamin deficiencies. This was seen in our large multicenter cohort even with adequate and near-universal usage of pancreatic enzyme replacement therapy. Risk was further exacerbated by surgical techniques including pylorus resection and/or Roux-en-Y gastrojejunostomy, decreased utilization of pancreatic-specific multivitamin preparations after TPIAT, and for those that were underweight at follow-up.

Adults were more likely to be overweight or underweight, both before and after TPIAT, and had a higher rate of vitamin A deficiency after surgery compared to children.

We would strongly advocate that all TPIAT recipients be monitored regularly for fat-soluble vitamin deficiencies after TPIAT and be administered a multivitamin with FSV formulations designed for use in pancreatic insufficiency to minimize risk of developing deficiencies.

Abbreviations

ARP

Acute recurrent pancreatitis

BMI

Body mass index

CP

Chronic pancreatitis

EPI

Exocrine pancreatic insufficiency

FSV

Fat-soluble vitamin(s)

MVI

Multivitamin

PERT

Pancreatic enzyme replacement therapy

POST

Prospective Observational Study of TPIAT

TPIAT

Total pancreatectomy with islet autotransplantation

Data Transparency Statement

Data and study materials will not be made available to other researchers.