EiLSE\/[ER
`SAUNDERS
`
`9-
`
`&
`as‘
`it-aaattrttfisasa
`
`
`
`Andrew Ukleja, MD”, Michelle M. Romano, RD, LD, CNSDb
`“Department of Gastroenterology, Cleveland Clinic Florida, 2950 Cleveland Clinic Boulevard,
`VVeston, FL 33331, USA
`l’lVlayo Clinic College of Medicine, 4-500 San Pablo Road_.jacl<souville, FL 32224. USA
`.».,~.».: »A\ ,1-:t~ ~41./N
`t.«,,r,,
`.. v. x».«,.,« we ..
`“>,\.'«4v»\
`-vi:
`., _ Mt
`,,.«,
`.-a.,rM« . .«.a.,— » W. at. "::(:A
`
`I
`
`
`
`‘
`
`‘:2? arenteral nutrition (PN) has been lifesaving therapy for patients with in-
`5
`testinal failure and one of the greatest discoveries of the twenty-first cen-
`tury. The first patient who received PN was discharged home in 1968
`PN allows delivery of essential nutrients and prolongation of life in patients
`with nonfunctioning intestinal tract. Patients with permanent dysfunction of in-
`testinal tract, such as short-bowel syndrome, severe Crohn’s disease, radiation
`enteritis, or chronic pseudo—obstruction, may require PN for life. Other patients
`may require PN temporarily for such conditions as malignancy, severe acute
`pancreatitis, bone marrow transplantation, perioperative period, and enteric fis-
`tulas. It is estimated that approximately 40,000 patients receive PN every year
`i
`' Several complications of PN have been recognized and some of them can
`
`be life threatening. Deficiencies of biotin, selenium, and essential fatty acid were
`discovered early in PN and eliminated. The complications of PN are divided
`into mechanical,
`i fectious, and metabolic. Mechanical coniplicatioris
`re-
`lated to insertion and care of the central venous catheter (CVC). Septic compli-
`cations are the result of catheter-associated infections. Metabolic complications
`refer to high or low serum levels of any components of PN solution, liver dis-
`ease, and metabolic bone disease. PN complications are associated with in-
`ff}!
`creased mortality and affect the quality of life of PN patients ...>,:.
`
`
` fit ‘=33?
`
`9.
`12¢‘
`ape‘-’ c
`’°“i§%;%§5l“¢‘«'§éfT %’”i3?““‘M“”“t%€t«”1‘::z’i?i'
`
`;
`B}?
`€Z’tf7%W3
`
`Separate from the PN solution, complications can occur centered around the
`vascular access catheter. These catheters cannulate the subclavian, internal jug-
`ular, or antecubital fossa vein. They can be used for short-term (nontunneled)
`or long-term (tunneled, implanted) vascular access. Maintaining a central line is
`the key to providing PN. Obtaining central access, site care, and maintaining
`flow of nutrients through the catheter can provide its own set of complications.
`Early complications include injury to surrounding structures, such as vascular
`perforation, hemothorax, pneumothorax. and air embolism. Complications
`that develop over time include mechanical malfunction, occlusion, or infection.
`
`*Corresponding author. E-mcritl address: ukle§c1o@<:cf.org (A. Ukleia).
`
`0889-8553/O7/$ — see front matter
`dot’:lO.lOl<S/§.gtc.2007.0l .009
`
`© 2007 Elsevier Inc. All rights reserved.
`gostro.thec|inlcs.com
`
`Page 1
`
`
`NPS EX. 2163
`
`CFAD V. NPS
`
`IPR2015—O1093
`
`
`
`Page 1
`
`NPS Ex. 2163
`CFAD v. NPS
`IPR2015-01093
`
`

`
`
`
`l.,§l{l.§:,li’3x éi §Z€..7>Mfia§T\E€fZ‘«~
`
`
`
`Complications that occur when placing the vascular access catheter or device
`can develop immediately, early postprocedure, or delayed postprocedure.
`Placement methods include percutaneous. cut—down, and tunneled. The corn
`plication rate is approximately 70/0 when image guidance is used
`Published
`rates for individual types of complications are highly dependent of patient se-
`lection. Complications that occur early or at the time of procedure are related
`to injury of surrounding vital structures or malpositioning of the catheter tip.
`Pneumothorax, arterial puncture, and line malposition occur in 10/0 to 40/0 of
`central line placements. Between approaches to access, some differences were
`
`reported. Ruesch and colleagues
`conducted a review of the internal jugular
`versus the subclavian approach. The jugular approach was significantly associ-
`ated with arterial puncture than the subclavian approach (3% versus 0.50/0; rel-
`ative risk [RR] 4.70 [950/0 confidence interval
`(CI), 2.05—10.77]). Catheter
`malposition by the jugular approach, however, was significantly less reported.
`There was no difference in the incidence of hemothorax or pneumothorax
`(1.3% Versus 1.50/0; RR 0.76 [0.43—1.33]). A lower incidence of pneumothorax,
`but higher incidence of line malposition, has been reported with peripherally
`inserted central catheter (PICC) lines }‘t},7l. Femoral catheterization was shown
`to have similar risk for mechanical complication when compared with the sub
`
`clavian site in critically ill patients _
`In patients considered for home PN, the
`use of imageguided technology (ultrasound, lluoroscopy) should be strongly
`A
`considered in placing Hickman catheters or ports
`re
` {i};;:tr‘;;:«ilr“"
`Inability to infuse nutrients or fluids or draw blood from a catheter is consid
`ered a mechanical malfunction. Causes include catheter dislodgement or frac-
`ture and catheter or venous thrombosis (the latter is discussed in the next
`section). Catheter breakage from the external segment can be repaired with
`proprietary repair kits. Catheters must be removed or replaced if unable to
`be repaired. Catheter “pinch—ofP’ syndrome has been reported. Pinch-off syn-
`drome occurs when the catheter is compressed between the first rib and the
`clavicle, causing an intermittent mechanical occlusion for both infusion and
`withdrawal, related to postural changes. Fracture of the catheter can occur.
`Findings should be confirmed radiographically. A more lateral replacement
`in the subclavian vein or using an alternative route avoids a recurrent compli~
`cation 0.? 3
` Coth
`
`iiizesltitzaeti %;'~:°l’:=.;sE»~
`<
`3
`
`The Centers for Disease Control and Prevention guidelines for the Prevention
`of Intravascular Catheter-Related Infections estimate that 250,000 cases of
`CVCassociated bloodstream infections (BSI) occur in the United States annu-
`ally
`This translated to nearly 5 cases per 1000 catheter—days. Mortality
`is estimated to be 120/0 to 25°/0 for each infection. CVC infections can occur
`
`at the exit site, tunnel, or pocket, and be infusate-related and catheter~related
`BSI. The most prevalent organisms cultured include coagulasonegative
`
`Page 2
`
`
`Page 2
`
`

`
`CC}?=¢%?§.E<:ATE=fj>€‘wf3 Qt: ~'Aé?,E3§”xf'é”:’£:?rT!,t‘>.l.
`
`staphylococci, Stapflvlocoatzs aureus, and Klebriella pmzumoniac. Patients receiving
`PN may also suffer from significant underlying diseases that may increase their
`risk of infection. The location of the catheter site, patient setting, and type of
`infusion also influence infection risk
`. Patient location in the ICU, where
`
`CVC may be needed for a length of time and are frequently accessed, poses
`an increased risk of infection
`l\4aki and colleagues
`recently reviewed
`200 published prospective studies of infection associated with the various types
`of vascular devices to determine the RR of BSI. Included were catheters com-
`
`monly used for PN support (PICC lines inpatient and outpatient, short-term
`noncuffed and long—term tunneled CVC, central venous ports, medicated
`and nonmedicated CVC). Based on their analysis, PICC rates of BSI for inpa-
`tients were 2.1 per 1000 catheter-days, and for outpatients 1 per 1000 catheter-
`days. For short-term nonrnedicated, nontunneled CVC, rates of infection were
`2.7 per 1000 catheter—days, and for short—term nonmedicated tunneled CVC,
`the rate was 1.7 per 1000 catheter-days. Of the various medicated CVC, the
`lowest rate was with the minocycline—rifampin—irnpregnated catheter at 1.2
`per 1000 catheter—days, and the highest rate was with the silverimpregnated
`catheter at 4.7 per 1000 catheter—days. Cuffed and tunneled CVC had a rate
`of 1.6 per 1000 catheter-days, and central venous ports had a rate of 0.1 per
`1000 catheter»days. Although the rate for PICC was lower than compared
`with short—term CVC, another study showed the infection rate of inpatient
`PICC lines to be 3.5 per 1000 catheter—days
`Many studies looked at infec-
`tion risk with CVC used for multiple purposes (hemodialysis, blood sampling,
`and hemodynamic monitoring’). One study collected prospective data on
`CVCs in 260 patients in the ICU fl
`In 61 of these patients, a single-lumen
`catheter was inserted by the subclavian route, through which only total paren-
`teral nutrition (TPN) infused. The catheters were in place a mean of 4.3 days.
`Fortynine pcrcent of the total catheters were sent for culture. and from those
`210/0 were colonized. Only one of the TPN catheters became colonized during
`the 1-year study period. For home PN patients with suspected line sepsis, a set
`of blood cultures should always be drawn before initiating antibiotic therapy,
`one from each lumen of the central catheter, and one from a peripheral vein.
`If the peripheral venous blood is not obtainable, two samples should be ob—
`tained from each lumen of the device. The device should be removed when
`
`the patient presents with severe sepsis and no other obvious source of infection
`is identified. Strong consideration should be given to removing any device if
`2
`S aureus or a fungus is the source of infection
`ft
`,Bo:«;
`Prevention of catheter-related BSI should include multiple fronts
`includes
`selected guidelines
`from the Centers for Disease Control and
`Prevention [.3 S93. Use of chlorhexidine gluconate for site care compared with po-
`vidone-iodine solution in patients with CVC was studied in a meta—analysis by
`
`Chaiyakunapruk and colleagues
`Eight studies with a total of 4143 cathe-
`ters (in place from 1-10 days) in hospitalized patients met inclusion criteria.
`Chlorhexidine gluconate was found to reduce the risk for catheter-related
`BSI by approximately 50% in this population. Prevention of grampositive
`
`Page 3
`
`
`Page 3
`
`

`
`\J''
`
`a...;‘
`r L“>
`
`IE;{it
`emvwmuwmummm;wwm»zzrmwnm»mamwmwmsm«mm:»;uwwwwmmtimmmvrmmmmmwemmWmmmwmmwamM%:mzw.mm;me>%mt<M1-nwmrfitatwmewmmwwimwmwewemmwtmmwmmxz
`
`Box 1: Selected guidelines for the prevention of intravascular
`catheter-related infections
`
`For CVC, including PlCC, replace gauze dressings every 2 days and transpar-
`ent dressings every 7 days on short-term catheters.
`
`Replace dressing when the catheter is replaced; when the dressing becomes
`damp, loosened, or soiled; or when inspection of the site is necessary.
`gordnplete infusions of lipid-containing fluids within 24 hours of hanging the
`ui
`.
`Complete the infusion of lipid emulsions alone within l2 hours of hanging the
`emulsion.
`
`id-contoinin fluids ever 24 hours. For dex-
`Re lace tube used to administer li
`P
`.
`9
`V
`P
`.
`.
`.
`trose and amino acid SOlUll0i'|, the administration set does not need to be re-
`placed more frequently than every 72 hours.
`
`Use either sterile gauze, transparent, or semi-permeable dressing to cover the
`catheter site.
`
`Select the catheter, insertion technique, and insertion site with the lowest risk of
`complications for the anticipated type and duration of therapy.
`Designate trained personnel for the insertion and maintenance of intravascular
`catheters.
`
`Do not administer transnasal or systemic antimicrobial prophylaxis routinely be-
`fore insertion or during use of an intravascular catheter to prevent catheter col-
`onization or bloodstream infection.
`
`Use totally implantable access devices for patients who require long-term, inter-
`mittent vascular access.
`
`Use an antimicrobial- or antiseptic-im regnated CVC in adults whose catheter
`is expected to remain in place more t on 5 days it, after implementing a com-
`prehensive strategy to reduce rates of BSl, the BSl rate remains above the goal
`set by the institution.
`ln adults, use an upper-extremity instead of a lower-extremity site for catheter
`insertion.
`
`Do not routinely replace CVCs to prevent catheter-related infection.
`
`Do not remove CVC on the basis of fever alone.
`
`catheter-rclatcd infections in oncology patients was reviewed {ltd}. This Co-
`chran databasc review included randomized, controlled trials giving prophylac-
`tic antibiotics bcforc insertion of the tunnclcd CVC, and trials using the
`combination of an antibiotic and heparin to flush the tunnclcd CVC in oncol-
`ogy patients. Eight trials were included with 527 patients. Four reported on
`vancomycin-Lcicoplanin bcforc insertion, four reported on antibiotic flushing
`combined with heparin. The overall effect of an antibiotic bcforc catheter inscr-
`tion decreases the number for granrpositivc tunnclcd CVC infection (odds ra-
`Lio [OR] : 0.55; 95% CI, 0.29—1.0-1}. Flushing with antibiotics and heparin
`proved to bc bcncficial (OR = 0.35; 95% Cl, 0.16 —to 0.77). For intraluminal
`
`Page 4
`
`
`Page 4
`
`

`
`W/in
`
`‘«':‘.(”“i"‘\/Wt. Vjfifii ..
`2?’
`,
`::%‘?aft
` $55 ‘-""5
`"
`mwasmmwkmwmwwmmmamwtvymwesavewewwzzeuezesat»we.mmmawawsammwxmawmmm:~»»m;amism«xaw4mmimmmmm¢ma4:wwamwm«2wwmwmamu»»v:m»mv»;smmsawm»mus>::wwrmma
`
`colonization the baseline infection rate was 150/0, which leads to a number
`needed to treat of 13 (95% Cl, 5-23). The authors concluded both interven-
`tions lead to a positive overall effect but should be considered with care because
`of the small number of studies.
`
`Some studies have indicated that the type. of dressing used for CVCs may
`affect the risk of infection. Gauze and tape, transparent polyurethane film dress-
`ings, or highly moisture—permeable transparent polyurethane film dressings are
`the most common types of dressing used. A recent Cochran database review
`looked at all randomized controlled trials evaluating the effects of dressing
`
`type
`Six studies were included, each one comparing one common type
`of dressing with another. All study data were limited because of small patient
`samples. There was no difference in the incidence of infectious complications
`between any of the dressing types compared. The study results suggest that
`the choice of dressing could be based on patient preference or cost.
`Routine replacement of administration sets has been advocated to reduce in-
`travenous infusion contamination. A Cochran database review was conducted
`
`to identify the optimal interval for the routine replacement of IV administration
`sets when infusate of PN (lipid and nonlipid) solutions are administered to hos-
`
`pitalized patients
`Randomized or quasirandomized controlled trials were
`reviewed and 15 were included, which contained 4783 participants. They con-
`cluded that administration sets that do not contain lipids, blood, or blood prod-
`ucts may be left in place for intervals of up to 96 hours without increasing the
`incidence of infection. There was no evidence to suggest a change in the cur~
`rent recommendation that lipid~containing sets should be changed every 24
`hours.
`
`
`
`Qz‘:t:ius§c>.r2 cmt;i Ei‘2r:;*i7'2i3€;25§S
`€:j},;:iiE':
`The true incidence of catheter-related venous thrombus is not known because it
`
`can be asymptomatic. Risk factors for thrombus include underlying disease, such
`as patients with cancer, and type and location of the catheter. In a study of 50,470
`home parenteral patients, the rate of thrombic catheter dysfunction was 0.23 per
`
`1000 catheter-days [)3
`This resulted in therapy interruption, catheter replace
`ment, unscheduled emergency room visits, or hospitalization. Two deaths
`were attributed to venous thrombus in a study of home PN patients by Scolapio
`and colleagues
`In a study of 102 hospitalized patients receiving PN either
`through a PICC or subclavian or internal jugular placed catheters, PICC lines
`were associated with higher rate of clinically evident thrombophlebitis (P< .01)
`Thrombosis prophylaxis was reported in a systematic review by Klerk and
`
`colleagues 11%,‘?
`, in patients with CVC for PN therapy. In five level 1 studies in-
`cluded in the analysis with a total of 204 patients, no bleeding events occurred.
`Prophylaxis with heparin added to PN was associated with a nonsignificant re-
`duction in the incidence of catheter-related thrombosis. Per the consensus state-
`
`ment by the Home Parenteral and Enteral VVorking Group, when device
`occlusion or dysfunction occurs, mechanical causes should be ruled out and a his—
`tory ofrecent infusions noted to rule out lipid, medication, or mineral precipitates
`
`Page 5
`
`
`Page 5
`
`

`
`fiéiri
`
`
`
`iii}/‘i.
`
`531
`
`iiiflfirixitfritilt
`
`Tissue plasminogen activator is a useful mo-
`as the cause of the dysfunction
`dality for treating catheter obstruction or dysfunction related to thrombus. Se-
`venty percent ethanol may be of value for occlusion related to lipid infusion,
`whereas 0.1-N sodium hydroxide may be of value when the occlusion is caused
`by mineral or drug precipitate. Heparin and saline are equally effective as a flush-
`ing agent. Central vein thrombosis should be treated with anticoagulation ther-
`apy in the absence of contraindication. Prophylactic anticoagulation therapy
`should be considered for those patients who are hypercoagulable or at high
`risk for catheter-related venous thrombosis.
`
`£€i.§§ir’é§i’§.i§€fi{?§°i§%§
`
`7?‘; 3/pa haters’: in
`The most common cause of hyperglycemia is excess of dextrose infusion (Box
`After the start of PN, a transient elevation of serum glucose can be observed
`followed by normalization of serum glucose after secretion of endogenous
`insulin is adjusted to the rate of dextrose infusion. Tolerance to dextrose
`depends on PN infusion rate and underlying medical condition. Patients at
`risk include the critically ill and those with sepsis, diabetes, acute pancreatitis,
`and corticosteroids use. Uncontrolled hyperglycemia from dextrose overfeed-
`ing can lead to immune system dysfunction and increased susceptibility to in-
`fection
`Suboptimal glucose control has been associated with higher rates of
`nosocomial and wound infections in critically ill and surgical patients
`A
`500/0 reduction in mortality and 460/0 reduction in septicemia among patients
`in surgical ICU liax/e been reported with intensive insulin therapy to maintain
`blood glucose levels between 80 and 110 mg/dL i;'Zi*>’l. Dextrose infusion at rate
`4 to 5 rng/kg/rnin seems to be optimal in stressed patients to avoid hyperglyce-
`r
`mia and related complications “’f~7/E. Hyperglycemia should be corrected with
`insulin intravenous drip rather than with sliding scale in the critical care setting.
`
`Box 2: Metabolic complications of PN
`
`Hyperglycemia
`
`Hyperlipidemia
`Hypercopnio
`Acid-base disturbance
`
`Electrolyte abnormalities
`
`Refeeding syndrome
`Manganese toxicity
`
`Hepatobiliary disorders
`Bone disease
`
`Hypoglycemia
`
`E
`
`7'
`
`Page 6
`
`
`Page 6
`
`

`
`£¥C.>fi.ft§>l,lCAl'§QNf3 C)?‘ i>.«‘itlL., mzzwsmammmmawaaenmmum~amwmum:sm«w.asaawmap2;m.;mnmvmamummmasmzmmmmmmwarmwmnwmmamwur-zumwwmwmmammnmm-mamzmmimammmwmmwwmama
`
`Dextrose load in PN should be increased slowly to meet caloric goal. whereas
`insulin dosage adjustments are made in PN. Excess of caloric load stimulates
`glucose conversion to fat leading to hypertriglyceridemia and hepatic steatosis.
`Sudden interruption of PN or excess of insulin in PN solution can result in
`hypoglycemia. Reactive hypoglycemia occurs 15 to 60 minutes after PN cessa-
`tion as a result of prolonged elevated level of endogenous insulin in response to
`caloric load of PN
`Patients at risk are those with renal and liver disease,
`severe malnutrition, sepsis, starvation, hyperthyroidism, and infants. Infusion
`of 100/0 dextrose immediately after discontinuation of PN or gradual tapering
`of PN over 1 to 2 hours before complete discontinuation can prevent hypogly-
`
`cemia if oral intake is not resumed
`L. A few studies, however, have shown
`no clinically significant hypoglycemia after tapered or abrupt cessation of PN
`Insulin dose in PN should be adjusted accordingly when underlying
`hyperglycemia is resolving. Monitoring of serum glucose levels is essential after
`Fix ,.;
`discontinuation of PN ta)‘:-‘
`
`
`.
`rivg.
`Excess of lipids or dextrose in the PN solution, or impaired lipid clearance, is
`responsible for PN-induced hyperlipidemia. Obesity, diabetes, sepsis, pancrea-
`titis, and liver disease predispose to hypertriglyceridemia because of decreased
`lipids clearance.
`Composition of lipids (phospholipid/triglyceride ratio) in PN affects the rate
`
`of lipid clearanc
`Hypertriglyceridemia from PN can precipitate acute pan-
`
`creatitis, especially when serum triglyceride levels are above 1000 mg/dL {E41
`In any case of hypertriglyceridemia in PN patients, dextrose overfeeding has
`to be considered. Dextrose load should be reduced first, followed by lipids
`reduction if hyperlipidernia is not corrected. Lipid infusion should not exceed
`
`0.12 g/kg/h in critically ill patients or those with impaired lipid clearance
`Continuous infusion of lipids over 24 hours compared with cyclic lipid infusion
`has been associated with improved lipid oxidation and serum fatty acid profile,
`and less deleterious effects on reticuloendothelial system function
`Caution
`has to be used when a PN patient is also receiving propofol, a source of extra
`
`lipid calories
`Propofol, a lipid-based drug, provides 1.1 kcal/mL of infu-
`sion. Lipid amount has to be adjusted in PN to avoid lipid overload when a pa-
`tient is given propofol infusion. Serum triglyceride levels should be monitored
`during PN therapy. Baseline serum triglyceride levels should be measured
`before initiation of PN and after lipid caloric goal
`is achieved. Daily lipid
`infusion should be discontinued when serum triglyceride concentration ex-
`ceeds 400 mg/dL. Lipid emulsion two to three times weekly should be contin-
`
`ued, however, to prevent essential fatty acid deficiency ?
`
`
`Overfeeding of total calories and dextrose can result in excess of carbon diox-
`ide production during carbohydrates metabolism. This may occur within hours
`of overfeeding, particularly in severely malnourished patients. Increased car-
`bon dioxide production stimulates minute ventilation and increases respiratory
`
`Page 7
`
`
`Page 7
`
`

`
`Elli‘ 37:}/it 33 ?—Z<C}:‘V’2/«‘«,é"».5<.l‘
`K<A}?wen:7;Js%%r¥iSV9!'£*£?-4k;9\K@§.:‘z"w»:»¥V>§w%\v\viX.W¢t$1$i./anwiééskhfzl?6nfiA3EHfl Y%zw&‘§ifi%mVQVM€Wl«&V;%m&5M 1$%@553»;-J‘0%"¥«€<‘§ak\¥Asfi6>‘&‘$$-°aw$#ydVfi“$&fixVfz.»?W<%%€s€iK#w%~%fixi%¢¢S\V7:%»M#:)mcW4$&fi~
`
`
`
`work load resulting in difficulty weaning off from mechanical ventilation 1,
`Reduction of caloric load helps correct hypercapnia.
`32
`
`’yr»s::*§;/'e:::»z*<2»%:3
`Rapid nutritional repletion in severely malnourished individuals can result in
`severe fluid and electrolyte disturbances including hypernatremia, hypophos-
`phatetnia, hypokalemia, and hypornagnesemia
`Infusion of dextrose stim~
`ulates insulin secretion, which is responsible for shifting‘ of phosphorus and
`potassium intracellularly. Severe hypophosphatemia can cause weakness, coir
`
`vulsions, respiratory failure. and cardiac decompensation leading to death
`In patients at risk, PN should be advanced gradually to achieve caloric goal
`over 3 to 5 days. The most important steps are to identify patients at risk
`for developing refeeding syndrome, provide nutrition support cautiously,
`and correct and supplement electrolyte and vitamin deficiencies to avoid re—
`feeding syndrome.
`
`
`' it / g
`
`Selenium deficiency has been reported in patients receiving long-term Pl\t :L
`J
`The recommended dietary allowance for selenium is 0.87 ttg/kg, of which 80%
`is absorbed. In a study of adult long—term TPN patients who received 40 to
`60 ttg of selenium daily in their PN solution, 750/0 of patients had low serum
`selenium levels
`The findings were suggestive of impaired renal homeosta-
`sis of selenium conservation. Serum selenium levels should be monitored peri»
`odically in patients receiving longterm PN.
`
`ii’ aczi 6
`Renal disorders associated with PN include hyperoxaluria, hypercalciuria, and
`tubular renal defects. Increased creatinine clearance resulting from glomerular
`hyperfiltration has been reported with short-term TPN infusion. TPN-
`associated nephropathy characterized by decline in creatinine clearance and
`impaired tubular function has been reported in adults and children receiving
`long-term PN
`Progression to chronic renal failure has not been documented
`in long~term PN patients. The renal dysfunction is multifactorial and related to
`amino acid load, use of nephrotoxic drugs, and possibly previous BSI
`The
`amount of parenteral amino acids and heavy metal contaminants has not been
`associated with decline in renal function except for chromium contaminant of
`PN in children
`Hypercalciuria, most likely related to vitamin C content,
`amino acid load, and cyclic PN, has been reported in adults on PN
`
`
`
`Intestinal morphologic
`PN is provided and enteral nutrition is withheld
`
`and functional changes occur to a lesser degree in humans for whom PN is
`the only nutritional source
`Factors contributing to intestinal atrophy in-
`clude lack of stimulation from luminal nutrients; lack of fuel source (gluta-
`mine); and impaired hormonal response. The loss of mucosal structure may
`
`Page 8
`
`
`Page 8
`
`

`
`’ ' "<7.,{'""%%’v‘*.§-"’i..§”iA"?’§{“Ni} C V’ 3%
`
`
`mwammmwwmmmaw«mewwe»amwzgasacammmswcwwwuzxm-may:mm.c;n:n-e;~mw«~:xe:.ezawm:«a»zzs»wmxaawtaxez-earwarmmmmwVmzmnwymumyeezwmwwmmmmmmwm/:a.wmzm«mwa¢w
`
`be sufficient to increase intestinal permeability and bacterial translocation, the
`clinical significance of which remains to be determined. The use of concomitant
`oral feeding or enteral nutrition is important in restoring and prevention of in-
`testinal changes associated with PN. Glutamine and arginine supplementation
`may be beneficial in this setting
`1“
`J
`.
`\Y';Tn§I£?‘i
`(Z?"t'¥:1§
`Abnormal gastric motility has been reported in patients with PN therapy ‘
`Delayed gastric emptying has been documented in normal individuals receiv-
`ing long-chain triglyceride-based lipid emulsion
`Hyperglycemia from dex-
`trose load in PN can also induce gastroparesis. The rate of gastric emptying
`correlated with the increase in blood glucose induced by the parenteral nutrient
`
`load
`Early satiety and the intolerance to oral intake, displayed by some
`individuals receiving oral and high-caloric PN, can be explained by the previ-
`ously mentioned mechanisms.
`
`%l?%§§.f3a.f?'i:§
`
`Numerous hepatobiliary complications have been associated with PN use in
`adults and pediatric patients (iitix Ii). The first case of PN-associated liver dis-
`ease was reported in 1971 in an infant who presented with severe cholestasis
`Cholestasis has been reported more frequently in infants than in adults.
`Contrary, hepatic steatosis is predominantly seen in adults receiving PN. It is
`clinically characterized by elevated serum arninotransferases and hepatomeg-
`aly. Ultrasound or CT scan can confirm steatosis.
`Liver function test abnormalities have been reported in 200/0 to 90°/0 of PN
`patients
`Asymptomatic increase in serum aminotransferases (>15 times
`normal) is commonly observed within the first 2 to 3 weeks from initiation
`of PN therapy
`Aminotransferases elevation is typically followed by in-
`crease in alkaline phosphatase and bilirubin. These mild biochemical abnormal-
`ities resolve with discontinuation of TPN therapy. The elevation of serum
`
`bilirubin is more often seen in children, particularly in preterm infants
`:
`Immature bile acid transport and metabolism early in life may be partially
`responsible for an increased susceptibility to liver injury. The liver function
`
`Box 3: PN-associated liver disorders
`
`Steczfosis
`
`Stecitohepofifis
`Fibrosis
`
`Cirrhosis
`
`Gallstones or biliary sludge
`Cholestcxsis
`
`Cholecysfitis
`
`Page 9
`
`
`Page 9
`
`

`
`
`
`3'1?
`mwwmmwms-wmmawmmmmmw
`
`
`":%‘€§f'},«’*Aiv‘»it“v:‘f,.“.-“
`liar
`’
`'2.
`«mm:amw::.m4=»».we>atwmsmm.wmvmnmmmwmmmmwmwmsmcsxwmmammmramwwmwrzmzamwmamamasmwmwwammmwammamxatmm
`
`test abnormalities poorly correlate with histologic findings on the liver biopsy.
`Macrovesicular and microvesicular steatosis is the most common finding ob-
`served on biopsy specimen. The other histopathologic changes include steato-
`hepatitis,
`intrahepatic cholestasis, and phospholipidosis. More advanced
`changes can be found including steatonecrosis, fibrosis, and deposits of lipofus-
`cin within Kupffer’s cells. The progression of liver disease from mild steatosis
`to fibrosis and micronodular cirrhosis over a 5-year period in the patient with
`Crohn’s disease and short-bowel syndrome receiving PN, documented by se-
`
`quential liver biopsies, has been reported by Craig and colleagues
`:. Hepatic
`
`steatosis and steatohepatitis can be reversed by PN caloric manipulations {iii
`End-stage liver disease has been reported in 15% of adults receiving PN and
`more often in neonates
`Pathogenesis of PN—associated liver disease is
`multifactorial.
`
`t
`
`of ilicitczl
`
`‘::§’.3iOE'l€3f» and C3e:r.%z'ose
`
`
`
`Overfeeding is associated with increased lipogenesis in the liver and impaired
`fat mobilization and use leading to hepatic steatosis. Caloric overload and an
`imbalanced source of energy in PN play a pivotal role in steatosis. lncreased
`insulin/glucagon ratio in portal circulation and resultant hyperinsulinemia
`from excess glucose in PN are responsible for glucose conversion to fat in
`the liver
`. 2. Replacement of dextrose by fat as a calorie source resulted in re-
`duced severity of liver function test elevation. In a prospective study, Meguid
`
`and colleagues
`showed the difference in liver function test elevation with
`glucose-based PN versus one third of calories from glucose replaced by lipids.
`Reduction in hepatic steatosis has been reported in 530/0 of patients who re-
`ceived only dextrose infusion, compared with 170/0 of those who received
`mixed dextrose and lipid emulsions (70:30 ratio of nonprotein calories) ftiffi.
`A balanced PN regimen should be used to provide 50% to 600/u of total calories
`from dextrose and 250/0 to 30% of calories from lipids. Reducing the carbohy-
`drate load in PN should be considered if elevated liver function tests are found
`
`to prevent steatosis.
`
`T
`
`*
`
`as
`
`
`
`Liver dysfunction has been reported with high dosages of lipid emulsions and
`essential fatty acid deficiency. Lipid emulsions may have a protective effect on
`the liver by allowing reduction of nonprotein calories in PN and providing es-
`sential
`fatty acids important for the production of hepatic phospholipids
`(protective from steatosis). Hepatic steatosis has been reported in patients
`with lipid—free PN, containing :1 mixture of dextrose and amino acids only,
`
`and resolution of steatosis was observed after lipid supplementation
`It was suggested, however, that essential fatty acid deficiency could be respon-
`sible for steatosis in those studies. It is recommended to provide a minimum 20/0
`to 4% of calories as linoleic fatty acid to avoid essential fatty acid deficiency.
`The role of essential fatty acid deficiency in the development of steatosis, how-
`ever, remains unclear. Excess of lipid infusion should be avoided. Fat overload
`syndrome has been reported with excess of lipid infusion at dosages greater
`
`Page 10
`
`
`Page 10
`
`

`
`<:.i'L*:‘~:‘a:??‘* %C}«¢x"i"E{_‘3%*‘~i:3 (Di:
`
`
`wmzmm'uarwax.m§mwm99v1»&tm'.wma&$M€&m5mmmm¥Azw#3:¥.&3w&m:za$7¢»>:ztv:i5‘e.2‘x¢Aa:9wiwwwmmmwmfiammzz42auxtM)v¢Mw2n¢»:vw»;ma«¢;w%'«x*zw2vstxam;;N«4n<mauwmv/u r = .
`
`The association between lipid emulsions and cholestasis has
`than 4 g/kg/d
`been reported in both adults and children §£“§9,?’
`Cholestasis is characterized
`
`by ballooning of hepatocytes, Kupffer cell hyperplasia, and bile duct plugging
`on liver biopsy. Lipid emulsions induce dose-dependent inhibition of choles-
`terol uptake by hepatocytes and reduce cholesterol availability for bile forma-
`tion leading to decreased bile volume and reduced bile flow
`In the United
`States, lipid emulsions are made of long-chain triglycerides derived from soy»
`bean or soybean—safflower oil. Medium—chain triglycerides in comparison
`with long—chain triglycerides undergo faster oxidation and the medium-
`chain—long-chain triglyceride mixture may be better tolerated and may be
` Such a mixture is not available,
`less likely to cause hepatic dysfunction [L
`however, in the United States.
`
`fltztitis E,3<C?Z":S‘S
`
`The development of PN~associated cholestasis (PNAC) in children has been
`linked to excess amounts of amino acids and their toxicity. The amino acids
`may have a direct effect on the canalicular membrane, with a tendency to re-
`duce bile flow and bile salt secretion. The alteration in canalicular How and
`
`membrane permeability leads to accumulation of hepatotoxic bile acids and im-
`
`paired bile acid transport
`It seems prudent not to exceed limits of amino
`acids in PN to avoid hepatotoxicity. A definite relationship between plasma
`amino acid concentrations and liver dysfunction has not been established.
`
`
`Methionine is an essential sulfur—containing amino acid that was found to be
`
`a contributing factor to development of cholestasis and steatosis {Z
`Higher
`blood methionine levels were reported in infants receiving PN. In premature
`
`infants, a low cystathion