Reduced Bone Mineral Density in the First Year After Total Pancreatectomy with Islet Autotransplantation (TPIAT)

Abstract

Background/Objectives:

The effect of total pancreatectomy with islet autotransplantation (TPIAT) on bone mineral density (BMD) in patients with CP is unknown. We aimed to assess bone health in patients with CP after TPIAT.

Methods:

We measured BMD, BMD Z-score, and bone mineral content (BMC) for total body, lumbar spine, right and left hip in 78 patients before and after TPIAT using dual-energy X-ray absorptiometry (DXA, n=78 pre-TPIAT, n=65 paired pre- and 12 months post-TPIAT, n=33 paired 12 and 18 months post-TPIAT), and tested for association with clinical history including age, smoking status, and medications using paired and two-sample t-tests, linear regression, and Fisher’s exact test. Laboratory measures related to bone health were also assessed.

Results:

In the patients with pre-TPIAT DXA, 12% had low BMD (Z-score ≤ −2). BMD, BMD Z-score, and BMC all decreased from pre- to 12 months post-TPIAT. BMD declined by 1.7%–4.1% with the greatest change at the hips. Adjusted for change in lean and fat body mass, DXA changes remained significant for total body and hip. Serum carboxy-terminal collagen crosslinks telopeptide and alkaline phosphatase increased at 12 months post-TPIAT, suggesting possible increased bone remodeling. BMD, BMD Z-score, and BMC did not change between 12 months and 18 months in any of the four regions (p>0.6).

Conclusions:

TPIAT is associated with decreases in BMD in the body, lumbar, and hip regions of patients with CP in the first year after TPIAT but these appear to stabilize between 12 and 18 months after TPIAT.

Introduction

Chronic pancreatitis (CP) causes irreversible damage to the pancreas, resulting in loss of endocrine and exocrine function with serious long-term sequelae.[1,2] CP associated exocrine pancreatic insufficiency (EPI) results in malabsorption, weight loss, and 25-hydroxyvitamin D (25(OH)D) deficiencies.[1,2] This in turn impacts bone health and increases risk of fractures.[2–4]

Bone mineral density (BMD) and bone mineral content (BMC) are measures of bone health that can be assessed using dual-energy X-ray absorptiometry (DXA).[5,6] Multiple studies have found reduced BMD and BMC and increased fracture risk in patients with CP.[3,4,7–10] A systematic review and meta-analysis by Duggan et al in 2014 found that in 513 patients with CP, 65% had osteopenia or osteoporosis.[8] Adjusted for age and sex, patients with CP have 2.4-fold higher odds of low-trauma fractures compared to healthy controls.[4] The pathophysiology of impaired bone health is likely multifactorial, with malnutrition, 25(OH)D deficiency, and chronic inflammation all proposed as contributing factors.[2,7,11]

In severe cases of CP, total pancreatectomy with islet autotransplantation (TPIAT) can be performed to mitigate chronic abdominal pain while preventing a severe form of pancreatogenic diabetes.[6,12] In TPIAT, the pancreas is completely resected, along with a portion of the duodenum, and Roux-en-Y duodenojejunostomy is usually performed.[13] The resulting EPI and altered gastrointestinal anatomy may exacerbate malabsorption and 25(OH)D deficiencies that present risk for low BMD.[14],[15] On the other hand, chronic inflammation might be lessened.[12] The effect of TPIAT on bone health over time has not been studied.

The aim of the current study was to determine if bone health was adversely affected by TPIAT in patients with CP. To do this, we compared DXA and laboratory measures of bone health and turnover in patients with CP before and after TPIAT.

Methods

Subjects

This study included 90 adult patients (age ≥ 18) who underwent TPIAT for chronic pancreatitis at the University of Minnesota between September 2010 and December 2013 who participated in a previously published randomized, placebo-controlled trial examining the effect of sitagliptin on metabolic outcomes (NCT01186562) and had at least one DXA scan completed during study participation.[16] The clinical trial methods and inclusion and exclusion criteria were described previously.[16] Patients who screened eligible for the parent study received either 100 mg of sitagliptin or placebo daily for 12 months following the operation. No drug was administered between 12 and 18 months after TPIAT. The present analyses included all patients with data on bone health regardless of treatment arm.[16] The protocol was reviewed and approved by the University of Minnesota institutional review board and informed consent was obtained from all participants.

DXA was included in the main study protocol as funding allowed. Participants not seen for follow up – either due to failure to meet eligibility criteria, or a missed visit – did not have follow up DXA scans. We used all available data from DXA in the 90 participants to construct three datasets: (1) pre-TPIAT DXA measures (n=78), (2) change from pre to 12 months post TPIAT in DXA measures (n=65), and (3) as an exploratory analysis, change from 12 to 18 months in DXA measures (n=33).

TPIAT Procedure and Postoperative Management

Patients underwent total pancreatectomy as previously described, typically with pylorus sparing partial duodenectomy, splenectomy, cholecystectomy, and Roux-en-Y duodenojejunostomy.[16] Islets were isolated by enzymatic and mechanical digestion and usually transfused through the portal vein of the liver.[16] At our institution, we recommend a daily multivitamin after TPIAT for all patients, preferably formulated for EPI. Patients in this study primarily took ADEKs, AquADEKs, Source EF, or MVW vitamin formulations, which provide fat soluble vitamins in a water-miscible form for better absorption. Patients completed laboratory and DXA assessments before the TPIAT and at 12 and 18 month (+/− 4 weeks) intervals postoperatively.

Data collection

Patient characteristics including age, sex, race, ethnicity, alcohol abuse, and smoking history were collected as well as disease characteristics, including CP etiology, duration and prior interventions. At each visit, height (cm), weight (kg), and BMI (kg/m²) were recorded. Medications were reviewed for use of estrogen replacement, proton pump inhibitors, psychiatric medications, bisphosphonates, pancreatic enzyme replacement therapy (PERT, recorded as lipase units/meal/kg), 25(OH)D and calcium supplementation. Menopausal status was recorded. Total 25(OH)D and alkaline phosphate levels were routinely measured at each visit.

Bone mineral density and content assessment

BMD and bone mineral content (BMC) was measured by DXA in the total body, lumbar spine, and right and left hip (GE Healthcare Lunar Prodigy; General Electric Medical Systems, Madison, WI) and analyzed using General Electric Medical Systems software (platform version 13.6). Standard positioning and protocols were followed. Z-scores for BMD were calculated based on sex and age normative data. Because our population was mostly under 50 years and/or premenopausal, we used Z-scores and did not analyze T-scores, which are applicable only if >50 years or post-menopausal.

CTX and P1NP analysis

In an exploratory subset (n=65), serum carboxy-terminal collagen crosslinks telopeptide (CTX) and procollagen type 1 intact N terminal propeptide (P1NP) were measured pre- and post-TPIAT as markers of bone turnover from stored serum samples using ELISA (CTX catalog #HU9654 and P1NP catalog #HU9656; Biotang, Lexington, MA, USA).

Statistical analysis

Low BMD, defined as a binary outcome, was deemed present if any of the 4 measured areas (lumbar, left hip, right hip, and body) had Z-score ≤ −2.0, per standard criteria established by the American College of Radiology.[17] For a measure of bone mineral density on a continuous scale, we used the lowest of the four Z-scores. Association of the binary outcome with pre-TPIAT characteristics was tested with Fisher’s exact test (categorical characteristics) or a t-test (continuous characteristics); association of the continuous outcome was tested with a t-test or ANOVA (categorical characteristics) or Pearson’s correlation (r; continuous characteristics).

Unadjusted tests of changes from pre-TPIAT to 12 months post-TPIAT used paired t-tests. Adjusted tests used linear regression with dependent variable the change in the outcome and adjusters change in weight and fat and lean body mass; the change in the outcome adjusted for no change in weight or fat and lean body mass was this regression’s estimated intercept. Changes in bone health-outcomes were tested for association with categorical characteristics (including 12-month opiate use and insulin independence) using t-tests or one-way ANOVA, and with continuous characteristics (serum measures at 12 months) using simple linear regression (equivalent to testing Pearson’s r). To examine the change in BMD and BMC from 12 to 18 months after TPIAT, we used a mixed linear model (which generalizes repeated-measures ANOVA to allow missing data) with person as the random effect and time (baseline, 12 months, 18 months) as the fixed effect. We tested the change from 12 to 18 months using a contrast within the time effect.

Results

Baseline subject and disease characteristics

Table 1 summarizes patient and disease characteristics for the 90 adults included. Participants had a mean age of 39.6 (SD 10.9) years. Seventy-one (78.9%) were female and most self-identified as Caucasian (92.2%) and non-Hispanic (95.6%).

Table 1.

Characteristics of participants (n=90).

Characteristic
Age at TPIAT (years), mean (SD)	39.60 (10.9)
Duration of Pancreatitis (years), mean (SD)	6.33 (5.79)
Etiology, n (%)
Idiopathic	37 (41.1)
Hereditary	15 (16.7)
Other	38 (42.2)
BMI (kg/m2), mean (SD)
At TPIAT	25.4 (5.89)
12 Months Post TPIAT	23.8 (4.73)
Sex, n (%)
Female	71 (78.9)
Male	19 (21.1)
Race, n (%)
Asian	1 (1.11)
Black	2 (2.22)
Caucasian	83 (92.2)
Native American/Alaska Native	2 (2.22)
Unknown	2 (2.22)
Ethnicity, n (%)
Hispanic	4 (4.44)
Non-Hispanic	86 (92.2)
Alcohol Abuse	6 (6.7)
ERCP Before Transplant, n (%)	83 (92.2)
Pancreatic Surgery Before Transplant, n (%)	10 (11.1)

BMI = body mass index; ERCP= endoscopic retrograde cholangiography

Factors associated with low BMD before TPIAT

Seventy-eight participants had pre-TPIAT DXA data. Low BMD was present before TPIAT in 11 (14.1%) participants. Patients with low BMD tended to have higher BMI, with average 28.7 (SE 1.8) compared to 24.7 (SE 0.7) in patients with normal BMD (p=0.04). Lean body mass by DXA was no different in those with or without low BMD (p=0.47) but fat mass was higher in those with low BMD (p=0.03). We also considered BMD as a continuous measure defined as the minimum Z score for the four body areas (lumbar, left hip, right hip, or body). Total 25(OH)D level was the only characteristic associated with this minimum Z score, with higher Z score associated with higher total 25(OH)D levels (r=0.26, p=0.03).

Low BMD or lowest BMD Z-score were not associated with age, sex, race/ethnicity, cigarette smoking, menopause status, etiology, prior pancreatic surgery, family history of osteoporosis, personal history of fractures, or medication use. Duration of pancreatitis before TPIAT was also not associated with low bone density before TPIAT. Laboratory measures for alkaline phosphatase, serum CTX, and serum P1NP levels were not associated with low BMD or BMD Z-score before TPIAT.

BMD and BMC Changes 12 months after TPIAT

Sixty-five subjects had paired pre- and 12 months post-TPIAT DXA scans (Table 2). Both BMD and BMC decreased significantly for total body, lumbar spine, and left and right hip after TPIAT (p<0.001 for total body, left hip, right hip, and p=0.003 and 0.015 for lumbar BMD and BMC respectively). Overall BMD declined by 1.7%–4.2% in this 12-month interval: 1.7% for total body, 1.8% for lumbar, 3.2% for right hip, and 4.2% for left hip. The Z-score decreased significantly in the right and left hip (p<0.001) with a marginally significant decrease for total body (p=0.02). Of ten patients who had low BMD before TPIAT, six were still low at 12 months while four were not (these were all borderline scores). Of fifty-five people with normal BMD before TPIAT, one was low at 12 months, while fifty-four were not (p=0.37).

Table 2.

DXA measures in patients (n=65) pre-TPIAT and 12 months post-TPIAT.

DXA Measure	Pre-TPIAT	Post-TPIAT	Average of the change (SE)	p-value
Total Body
BMD (g/cm2)	1.15	1.13	−0.02 (0.004)	<0.0001
BMC (g)	2604.85	2475.08	−129.77 (22.92)	<0.0001
Z-score BMD (SDS)	0.09	−0.02	−0.11 (0.046)	0.021
Lumbar Spine
BMD (g/cm2)	1.14	1.12	−0.02 (0.007)	0.003
BMC (g)	61.08	59.77	−1.31 (0.53)	0.02
Z-score BMD (SDS)	−0.33	−0.37	−0.04 (0.06)	0.50
Right Hip
BMD (g/cm2)	0.95	0.92	−0.03 (0.004)	<0.0001
BMC (g)	29.75	28.76	−0.99 (0.16)	<0.0001
Z-score BMD (SDS)	−0.37	−0.53	−0.16 (0.04)	<0.0001
Left Hip
BMD (g/cm2)	0.96	0.92	−0.04 (0.004)	<0.0001
BMC (g)	30.06	28.96	−1.10 (0.14)	<0.0001
Z-score BMD (SDS)	−0.35	−0.51	−0.16 (0.03)	<0.0001

BMD= bone mineral density; BMC= bone mineral content; SDS= standard deviation score for age and sex

When these analyses were adjusted for change in weight, fat, and lean body mass over 12 months post-TPIAT, the DXA changes remained significant for total body and right and left hip. We then assessed whether the observed decreases in BMD, BMC, and Z-score were associated with age, sex, smoking history, etiology, menopausal status, or medications known to impact bone health (Supplemental Table 1). Smoking was associated with greater magnitude decreases in lumbar BMC (p=0.034) and lumbar Z-score (p=0.024), with the effect being greatest in current, then former, then never smokers. SSRI use was associated with a greater decrease in lumbar BMC (p=0.006), lumbar BMD (p=0.035), and lumbar Z-score (p=0.035). We also observed an association of antipsychotic medications with lumbar BMD (p=0.024) and left hip BMD (p=0.019), in this case being on treatment was associated with a small increase in bone measures. Use of other psychiatric medications (other than SSRI or antipsychotics) was also associated with increases in lumbar Z-score (p=0.039) and left hip Z-score (p=0.029). Notably, sitagliptin treatment and lipase units/meal/kg were not associated with changes in bone density from baseline to 12 months post TPIAT.

Lastly, we examined the effect of opiate use and insulin independence at 12 months on the change from pre-surgery to 1 year in DXA measures. Those patients still using opioids at 12 months post-TPIAT had greater decrease in body BMC (p=0.007) post TPIAT. However, opioid use was not significantly associated with changes in BMD or BMD z-score at any other site. Insulin use at 12 months was not significantly associated with any DXA measures.

Markers of bone health after TPIAT

Serum CTX (p=0.01) and alkaline phosphatase (p=0.002) both increased significantly from before to 12 months after TPIAT (Table 3). Changes in serum P1NP and 25(OH)D were not significant. Higher serum CTX at 12 months was associated with increased body BMD (estimate 1.03 × 10⁻⁵, SE 4.7 × 10⁻⁶, p=0.03) while higher alkaline phosphatase at 12 months was associated with decreases in lumbar BMD (Estimate −5.1 × 10⁻⁴, SE 1.4 × 10⁻⁴, p=0.001), right hip BMD (Estimate −2.1 × 10⁻⁴, SE 8.4 × 10⁻⁵, p=0.02) and left hip BMD (Estimate −2.5 × 10⁻⁴, SE 7.5 × 10⁻⁵, p=0.001).

Table 3.

Laboratory studies of bone turnover and health in patients (n=65) pre-TPIAT and 12 months post-TPIAT.

Measure	Pre-TPIAT	Post-TPIAT	Average of the change (SE)	P-value
Serum CTX (pg/mL), n=47	1330.46	1666.90	336.44 (125.87)	0.01
Serum P1NP (ng/mL), n=47	4.27	4.07	−0.20 (2.23)	0.93
25(OH)D (ug/L), n=75	28.41	30.17	1.76 (1.33)	0.19
Alk Phos (U/L), n=77	83.17	100.34	17.17 (5.28)	0.002

CTX= carboxy-terminal collagen crosslinks; P1NP= Procollagen type I propeptide; Alk Phos=alkaline phosphatase; 25(OH)D=25-hydroxyvitamin D

BMD and BMC Changes 18 months after TPIAT

Thirty-three patients had DXAs at both 12 months and 18 months post-TPIAT. BMD, BMD Z-score, and BMC did not change significantly between 12 to 18 months in any of the four regions (p>0.6) (Table 4).

Table 4.

DXA measures in patients (n=33) at 12 and 18 months post-TPIAT.

DXA Measure	12 months	18 months	Average of the change (SE)	p-value
Total Body
BMD (g/cm2)	1.13	1.12	−0.011 (0.005)	0.91
BMC (g)	2435	2315.2	−119.8 (29.6)	0.69
Z-score BMD (SDS)	0.02	0.08	0.06 (0.05)	0.97
Lumbar Spine
BMD (g/cm2)	1.12	1.13	0.01 (0.01)	0.98
BMC (g)	59.1	56.8	−2.26 (0.85)	0.98
Z-score BMD (SDS)	−0.31	−0.06	0.25 (0.08)	0.87
Right Hip
BMD (g/cm2)	0.92	0.91	−0.01 (0.006)	0.95
BMC (g)	28.3	27.26	−1.02 (0.23)	0.80
Z-score BMD (SDS)	−0.52	−0.45	0.07 (0.04)	0.99
Left Hip
BMD (g/cm2)	0.92	0.91	−0.01 (0.006)	0.99
BMC (g)	28.6	27.4	−1.15 (0.21)	0.99
Z-score BMD (SDS)	−0.5	−0.47	0.03 (0.04)	0.98

Abbreviations: BMD= bone mineral density; BMC= bone mineral content; SDS= standard deviation score for age and sex

Discussion

Emerging data have established that CP is a risk factor for impaired bone health, osteopenia, and bone fractures in adult patients.[3,4,7] TPIAT is reserved for patients with chronic pancreatitis that does not respond to medical and endoscopic therapies.[6,12] Because TPIAT results in complete EPI and alters continuity of the gastrointestinal tract (e.g., Roux-en-Y duodenojejunostomy), TPIAT may introduce new nutritional risks that could adversely impact bone health in those with CP.[14] On the other hand, removing the source of the pancreatitis may decrease chronic inflammation and improve overall health and activity with positive impacts on BMD.[12] For the first time, we evaluated bone health in adult patients before and after undergoing TPIAT. BMD, BMC, and Z-scores decreased significantly on average in the first year after TPIAT. These changes remained significant after adjusting for change in weight, fat, and lead body mass, suggesting that changes in bone health are not solely explained by anthropometric changes related to TPIAT surgery. Serum CTX and alkaline phosphatase levels increased during this period, consistent with bone remodeling. However, reassuringly, DXA measures appeared to stabilize from 12 to 18 months after TPIAT, suggesting adverse effects on bone health may be confined to the early post-operative period.

Multiple studies have found a high prevalence of osteopenia and osteoporosis in patients with CP, with some suggesting either is present in 5% to 60% of patients.[7,8] Our study found that about 14% of our patients had low BMD before TPIAT, lower than some studies. However, patients who undergo TPIAT are often clinically distinct from the general CP population. TPIAT patients are often younger (mean age 39.6 years in our cohort) and more commonly have genetic or idiopathic CP, with very few having alcohol-related CP due to medical and psychosocial complications of performing TPIAT in the setting of alcohol use disorder.[13] Interestingly, we found a correlation between increased BMI and low BMD in CP pre-TPIAT. This is surprising as malnourished, thin patients are typically thought to be higher risk for osteoporosis.[18] However, BMI may not accurately capture nutrition status, and, in our cohort, the higher BMI in the low BMD group appears to be entirely attributable to increased fat (but not muscle) mass. We have previously reported overweight or obese sarcopenic patients in our population.[19–21] Our cohort also had relatively few underweight (~9%) patients. Low 25(OH)D levels was weakly associated with low BMD Z-scores before TPIAT but assessing 25(OH)D status based on one time point is limited, as a single value may be impacted by supplementation and season and not reflect long-term 25(OH)D exposure.

To our knowledge, no study has described BMD over time after TPIAT. In our CP patients, BMD declined by 1.7%–4.2% at the body, lumbar, and hip regions over 12 months after TPIAT, with the greatest change at the hips. This change in BMD is larger than typically expected with aging. Although data vary depending on the population’s age and sex, one study reported a rate of 0.5% annual loss in BMD in the lumbar spine and 0.35% in the hip in premenopausal women.[22] The changes we found in BMD occurred alongside increases in serum CTX and alkaline phosphatase levels, suggesting that TPIAT may lead to increased bone resorption in patients with CP early after TPIAT. We hypothesize that these findings in the first year after TPIAT are explained by nutritional deficiencies exacerbated by anatomical changes created by the surgery. However, weight loss alone does not account for these changes, since differences remained significant after adjusting for changes in weight and fat and lean body mass. Patients have complete EPI after surgery and may require titration of pancreatic enzyme therapy as their oral diet increases in the first months after surgery. Patients undergoing TPIAT also undergo duodenectomy and Roux-en-Y duodenojejunostomy. Similar findings of increased bone mass and bone turnover have been observed up to 3 years following gastric bypass surgery.[23],[24] Besides 25(OH)D and calcium malabsorption, changes in inflammatory compounds and gastrointestinal hormones have been hypothesized to increase bone loss.[25] Patients with CP and intestinal surgery are also at risk for small intestinal bacterial overgrowth, which may further exacerbate malabsorption.[26] Finally, these changes may simply be the result of undergoing a complex operation. Bone loss has been shown to occur for up to one year in patients who experience a critical illness event related to nutritional issues, inflammation, and impaired mobility.[27] Aggressive rehabilitation and nutrition could be important in mitigating these potential adverse effects of surgery.

Reassuringly, in patients who were followed with DXA from 12 to 18 months, no further changes in BMD, BMC, or Z-scores were observed, suggesting that the insult to bone health may stabilize following the immediate post-operative period. Following recovery from major surgery, patients may become more physically active, benefitting their bone health.[28] Recurrent pancreatic inflammation resolves, and decreased inflammation may reduce bone loss, as some studies have shown that inflammation mediates bone resorption in CP patients.[7] Also, medical therapy requires adjustments in the post-operative period including titration of pancreatic enzyme replacement therapy and introduction of multivitamin supplementation.[13,29] At our institution, once patients are on a normal oral diet, we target 1,000 lipase units/kg/meal (range 500–2000) and introduce a pancreatic multivitamin, with multivitamin and added 25(OH)D supplementation adjusted as needed based on fat soluble vitamin levels obtained at 3, 6, and 12 months after TPIAT.[30] Nutritional labs are assessed yearly. Our findings support the need for careful monitoring of nutrition after TPIAT. Longitudinal follow up is needed to assess long-term effects of TPIAT bone health after TPIAT.

Bone loss was more severe in current or former smokers, or those who were treated with SSRIs. This is consistent with published literature: smoking is associated with worse DXA scans in patients with CP.[9] The effect of SSRIs on bone health is more controversial, with some studies suggesting a role in bone loss and others showing no effect.[31] Although the literature suggests higher risk of fractures and osteoporosis in chronic opioid users and those with opioid use disorder, we observed an association of opioids only with lower total body BMC.[32,33] It is likely that we did not find a larger impact of opioid use in our participants because all were opioid exposed in the pre-operative and early post-operative periods. Finally, we found that insulin dependent diabetes was not associated with bone loss 12 months after TPIAT. Although diabetes is a known risk for osteoporosis and fractures in the literature,[34,35] for our participants the duration of diabetes was short (≤1 year). Longer term follow up may be needed in TPIAT recipients to determine the impact, if any, of diabetes or persistent opioid use after surgery. Because patients in our series are younger than other osteopathy reports in CP, it will be particularly important to assess for osteopenia or osteoporosis risk as these patients age, and ensure ongoing monitoring of nutritional measures including vitamin D.[36]

Because our study data were collected in the context of a clinical trial, the DXA equipment remained the same, and we have detailed data for labs, patient history, and medications, allowing us to reduce confounding. However, we were not able to collect all potentially relevant variables, such as vitamin K levels. Some studies have suggested that Vitamin K may play an important role in bone density loss in patients with CP, particularly for males.[37] While all of our patients were advised to take vitamins formulated for pancreatic insufficiency, we lack precise information regarding the dosages to include vitamin D or K doses in our analyses. Our analyses were also limited by variable numbers of DXA studies at each time point, related to protocol changes and, rarely, missed patient visits. In particular, we have fewer participants with 18-month DXA data. However, even with this smaller sample size, there does not appear to be a decrease in BMD from 12 to 18-months. We do not have CTX1 or PN1P results at 18 months to evaluate bone turnover markers in the 12 to 18-month postoperative period. Additionally, we did not collect a reliable measure of exocrine insufficiency, such as stool elastase, in these patients before TPIAT. As a result of this and the fact that none of the patients had diabetes before surgery, we are unable to draw conclusions about the impact of diabetes and exocrine insufficiency on bone health before TPIAT. Finally, we have no data beyond 18 months post-TPIAT to assess longer-term bone health. Nutritional deficiencies may take time to develop, which could worsen long-term bone loss. One small prospective study from 1991 showed decreased BMC at the radial site 5 years after total pancreatectomy.[38] Future follow up on long-term bone health after TPIAT is needed.

In conclusion, TPIAT is associated with decreases in bone density in the body, lumbar, and hip regions of patients with CP in the first year after TPIAT. These changes appear to stabilize between 12 and 18 months after TPIAT. Smoking and SSRI use are associated with greater losses. DXA changes occurred with increases in serum CTX and alkaline phosphatase levels, suggesting that TPIAT may lead to increased bone resorption in patients with CP early after TPIAT. Further studies are needed to determine the long-term effect of TPIAT on bone health.