Iodine Toxicity Treated With Hemodialysis and ContinuousVenovenous Hemodiafiltration

Sharan Kanakiriya, MD, Ives De Chazal, MD, Karl A. Nath, MD, Eric N. Haugen, MD,Robert C. Albright, DO, and Luis A. Juncos, MD

● Continuous mediastinal irrigation with povidone-iodine is used commonly for treating severe postoperativemediastinitis. However, concurrent iodine toxicity has been reported, particularly in patients with renal dysfunction(likely because absorbed iodine is renally excreted). The authors were consulted on a 45-year-old patient withmediastinitis who had renal and hepatic dysfunction while being treated with mediastinal irrigation of povidone-iodine. The povidone-iodine irrigation was discontinued because he had toxic plasma iodine levels. Despite this, hiscondition worsened, and the iodine levels remained elevated. Thus, hemodialysis (HD) was initiated using high-fluxmembranes followed by continuous venovenous hemodiafiltration (CVVHDF; 2 L/h of hemofiltration and 2 L/h ofHD). Plasma and effluent iodine levels were measured repeatedly to determine iodine clearance by these 2modalities (HD, 120 mL/min; CVVHDF, 37 mL and 44 mL/min on days 1 and 2, respectively). Hepatic and renalfunctions improved with decreasing plasma iodine levels. Based on this experience and after reviewing theliterature the authors conclude that: (1) iodine irrigation can increase blood iodine levels significantly, especially inthe setting of renal failure, and lead to increased morbidity and mortality; (2) plasma iodine levels should bemonitored in patients with renal insufficiency; and (3) HD and CVVHDF are effective at clearing iodine. The authorssuggest that patients that are at high risk or already developing signs of iodine toxicity should have the iodineirrigation discontinued and may benefit from renal replacement therapy (RRT). Alternatively, concomitant RRTduring iodine irrigation may be attempted to maintain the systemic iodine levels at nontoxic levels. Am J Kidney Dis41:702-708.© 2003 by the National Kidney Foundation, Inc.

INDEX WORDS: Hemodialysis (HD); hemodiafiltration; iodine; mediastinum; sternotomy.

POVIDONE-IODINE (betadine) is a widelyused antiseptic introduced by Shelanski and

Shelanski1 in 1956. It has microbicidal activityagainst a wide variety of organisms includingbacteria, fungi, protozoa and viruses1-4 and re-tains its antimicrobial properties even at lowconcentrations.2,5 It is used in a variety of clinicalsituations, one of which has been in the treatmentof postoperative mediastinitis, the latter origi-nally described by Thurer et al.6 In this situationdirect irrigation of the mediastinum with concur-rent drainage via chest tubes is performed. Sev-eral antibiotic and antiseptic irrigants have beentried, but because of the reported systemic andlocal toxicities, povidone-iodine became a pre-ferred irrigant for treating this complication and

has been considered relatively safe. However,there are isolated reports of iodine toxicity thathave led to serious morbidity or death in patientstreated with iodine-containing solutions for burns,surgical wounds, decubital ulcers, and mediasti-nitis. The majority of these cases occurred inpatients with renal insufficiency and may berelated to the fact that iodine is excreted primar-ily by kidneys. In this report, we describe apatient with renal insufficiency who had featuresof iodine toxicity after being treated with povi-done-iodine by closed mediastinal irrigationmethod. Despite discontinuing the povidone-iodine irrigation, he had persistently high bloodiodine levels (likely caused by his renal failure)and signs of iodine toxicity. Therefore, we treatedhim with hemodialysis (HD) followed by continu-ous venovenous hemodiafiltration (CVVHDF)and determined the iodine clearances with thesemodalities.

CASE REPORT

A 45-year-old white man was admitted with an acuteanterior wall myocardial infarction. His past medical historywas significant for diabetes for 15 years, hypertension, andchronic renal insufficiency secondary to diabetes and hyper-tension. Insulin, lisinopril, glipizide, amlodipine, hydrochlo-rothiazide, atenolol, and aspirin were his medications onadmission. His laboratory examination results showed el-

From the Department of Medicine, Division of Nephrol-ogy and Division of Critical Care Medicine. Mayo Clinicand Foundation, Rochester, MN.

Received July 22, 2002; accepted in revised form Decem-ber 2, 2002.

L.A.J. is supported in part by a Clinician-Scientist Awardfrom the NKF and NIH DK02943.

Address reprint requests Luis Juncos, MD, Mayo Clinic,Stabile 7, 200 First Street SW, Rochester, MN 55905. E-mail:juncos.luis@mayo.edu

© 2003 by the National Kidney Foundation, Inc.0272-6386/03/4103-0021$30.00/0doi:10.1053/ajkd.2003.50134

American Journal of Kidney Diseases, Vol 41, No 3 (March), 2003: pp 702-708702

evated cardiac enzymes, increased blood urea nitrogen (BUN)and creatinine (49 mg/dL or 17.5 mmol/L; and 1.8 mg/dL or159.1 �mol/L), and an aminoaspartate transaminase (AST)level of 52 U/L. Coronary angiogram showed severe triplevessel disease. He underwent triple vessel bypass surgerythe next day. Because he was profoundly hypotensive, abiventricular assist device and an intra-aortic balloon coun-terpulsation device were placed in the operating room, andhe was transferred to the intensive care unit on inotropes. Hewas found to be febrile on the fourth postoperative day, andhis blood cultures grew Gram-positive cocci and Gram-negative rods for which broad-spectrum antibiotics werestarted (vancomycin, cefepime, and metronidazole).

On the seventh postoperative day a foul-smelling dis-charge was noticed from his mediastinal wound. He wastaken to the operating room where the wound was excised,generously debrided, and irrigated with mycifradin, cefazo-lin and povidone-iodine solutions. The intra-aortic balloonand ventricular assist device were removed, and closedcontinuous mediastinal irrigation using 0.5% povidone-iodine at 100 mL/h was commenced. Cultures from thedischarge grew Enterobacter cloacae and Enterococcus feca-lis. He was started on Meropenem after discontinuing theother antibiotics.

His renal function deteriorated over the subsequent days,and his BUN and serum creatinine levels were 94 mg/dL(33.6 mmol/L) and 2.4 mg/dL (212.2 �mol/L), respectivelyon the 13th postoperative day. He remained nonoliguric onhigh-dose diuretics with a urine output of approximately1,000 mL/d. His blood iodine level, which was checkedbecause of the worsening renal function, was found to beelevated at 1,190 �g/dL. His liver function, which wasslightly abnormal in the immediate postoperative period,deteriorated further, and its etiology was unclear. His liverfunction tests on the 13th postoperative day were as follows:AST, 355 U/L; alkaline phosphatase, 3,203 U/L; total biliru-bin, 29.8 mg/dL (509.6 �mol/L); and direct bilirubin, 25.5mg/dL (436.1 �mol/L).

Ultrasonographic studies of the hepatobiliary system wereunremarkable. Because of the concern of iodine toxicitycausing liver deterioration, the povidone-iodine irrigationwas stopped, and the plasma iodine levels were closelymonitored. BUN, creatinine, bilirubin, and liver enzymelevels continued to increase, and his urine output remainedbetween 800 and 1,000 mL/d on high doses of furosemide.However, despite stopping the irrigation, his iodine levelsremained elevated, his liver function continued to worsen,and his acute renal failure (ARF) persisted. Thus, the patientwas started on renal replacement therapy. Because we wantedto ensure adequate iodine clearance early on, we initiallystarted him on dialysis (standard HD) for 3 hours using ahigh-flux F80 dialyzer without ultrafiltration with bloodflow and dialysate flow rates of 250 mL/min and 550mL/min, respectively. The effluent was collected, and iodinelevels were determined in plasma and the effluent after 90minutes of dialysis and again at the end of dialysis. At thecompletion of HD, we initiated him on CVVHDF to achievecontinued clearance of uremia and iodine. The blood flowand dialysate flow rates were 150 mL/min and 2 L/h,respectively, and the HF rate was 2 L/h. Plasma and effluentiodine levels (mass spectrophotometry) were checked every

6 hours, and liver function tests were monitored daily. Theaverage of effluent (U) and plasma (P) iodine values wereused to calculate its clearances on HD and CVVHDF usingthe formula UV/P.

With the initiation of renal replacement therapy, his plasmaiodine level decreased from 1,010 �g/dL to 117 �g/dLwithin 48 hours (Fig 1). The clearance of iodine achievedwith HD was 120 mL/min and that with CVVHDF was 37mL/min on day 1 and 44 mL/min on day 2 (Table 1). Hisliver function started to improve (Fig 2A-C), and there was adramatic overall clinical improvement. Continuous renalreplacement therapy (CRRT) was continued for 5 days andstopped. He was extubated after 2 days, transferred out ofthe intensive care unit a week later, and finally dischargedhome after rehabilitation. His liver function test results werenormal, and his creatinine level was 1.8 at the time ofdischarge.

DISCUSSION

Poststernotomy mediastinitis is a rare compli-cation of thoracotomy with a reported incidenceof 0.4 to 5%.7 Early mediastinitis can be treatedeffectively with thorough wound debridementand mediastinal irrigation.8 Bryant et al9 de-scribed a method that involves mediastinal reex-ploration, placement of mediastinal irrigationtubes, and primary closure of the sternum. Sev-eral antimicrobial irrigants have been used toirrigate the mediastinum during this procedureincluding neomycin, bacitracin, polymyxin, kana-mycin, and cephalothin, but there were increas-ing reports of toxicities and microbial resistance.

Fig 1. Decrease in plasma iodine levels (�g/dL)after initiation of RRT.

DIALYTIC THERAPY OF IODINE TOXICITY 703

In 1974, Thurer et al6 reported the successful useof povidone-iodine as the irrigant. Because itwas thought to be effective and safe, it becamewidely used for treating poststernotomy medias-tinitis. However, its efficacy has not been estab-lished firmly, thus, it currently is not consideredthe standard of care for postoperative mediastini-tis. Despite this, povidone-iodine irrigation ofthe mediastinum still is used commonly for se-vere postoperative mediastinitis because of itsperceived efficacy and low toxicity profile.

Polyvinylpyrollidone-iodine is a water-solublecompound that results from the combination ofmolecular iodine and a hydrophilic polymer,polyvinylpyrollidone. This polymer has no intrin-sic antibacterial activity2 but delivers free io-dine10 directly to bacterial cell surface by virtueof its affinity to cell membranes, which appearsto be the crucial step of its antibacterial ac-tion.11,12 It is used widely as an antiseptic and isavailable as a solution, ointment, and scrub. A10% solution contains 0.0001% or 1 part permillion (ppm) of free iodine.13 The free-iodinelevel depends on the concentration of the solu-tion and follows a bell-shaped curve. The levelobtained from 1% and 10% solutions is 1 ppm,and the maximum level of 24 ppm is obtained

from a 0.7% solution.13 So even dilute solutionsretain their bactericidal power.2,5,13-16

Povidone-iodine administration through anyroute results in systemic absorption of iodine.13

The amount of iodine absorbed depends on theconcentration of the solution and the route ofadministration, with more absorbed through de-nuded skin and mucosal surfaces than throughintact skin.17-23 Systemically absorbed iodinebound to albumin follows first-order eliminationpharmacokinetics and is excreted by the kid-neys2,23,24; the clearance is directly related toglomerular filtration rate.2,10,25 In first-order elimi-nation, the time required to reach a steady serumconcentration depends on elimination half-life,and 90% of this concentration is reached in 3.3half-lives. The magnitude of this concentration isdirectly proportional to the absorption rate andinversely proportional to the clearance, which isdependent on renal function.24

Iodine plays an important role in 2 importantbiologic functions. First, iodine in thyroid hor-mones regulates basal intracellular metabolism,and, secondly, as cosubstrate for neutrophiliclysosomal myeloperoxidase, inorganic iodide un-dergoes covalent linkage to bacterial cell wallsurface glycoproteins, thereby potentiating thebactericidal capacity of polymorphonuclear leu-kocytes.26 However, excessive plasma levels leadto systemic toxicity. Iodine toxicity during medi-astinal irrigation with iodine-containing solu-tions can occur because of either absorption oflarge quantities of iodine in the mediastinum orbecause of inadequate clearance as a result ofrenal dysfunction.27-31 Urinary iodine excretionin patients with renal failure has been shown tobe minimal—150 �g/d compared with that inpatients with normal renal function—912,000�g/d. Treatment of burns, surgical wounds, orulcers with povidone iodine also can result iniodine toxicity.10,17,26,32-34 The signs and symp-toms reportedly caused by iodine toxicity in-clude hypernatremia, metabolic acidosis, hepaticdysfunction, renal failure, and thyroid abnormali-ties (Tables 2 and 3).

Iodine-induced hypernatremia was reported ina patient with severe burns. The hypernatremiawas thought to be caused by free water losses asa result of the strong osmotic gradient of povi-done-iodine with subsequent intravascular dehy-dration,33 but the severe burns also may have

Table 1. Iodine Clearances by HD and CVVHDF

Hemodialysis using F-80 membranesU (effluent iodine) � 140 �g/dLV (effluent volume) � 550 mL/minP (plasma iodine) � 640 �g/dLIodine clearance (Ic)

�140 �g/dL � 550 mL/min

640 �g/dL� 120.31 mL/min

CVVHDF (Dialysate flow rate of 2 L/h with HF rateof 2 L/h)

Day 1U (effluent iodine) � 140 �g/dLV (effluent volume) � 96,000/1,440 � 66.67 mL/minP (plasma iodine) � 250 �g/dLIodine clearance (Ic)

�140 �g/dL � 66.67 mL/min

250 �g/mL� 37.34 mL/min

Day 2U (effluent iodine) � 58.5 �g/dLV (effluent volume) � 66.67 mL/minP (plasma iodine) � 87.4 �g/dLIodine clearance (Ic)

�58.5 �g/dL � 66.67 mL/min

87.4 �g/dL� 44.62 mL/min

KANAKIRIYA ET AL704

contributed to the losses. Metabolic acidosis isbelieved to result from the combination of freeiodine and sodium bicarbonate with constantelimination of carbon dioxide as shown in thefollowing equation: 6NaHCO3 � 3123 5NaI �NaIO3 � 6CO2 � 3H2O.35 Hyperchloremia hasbeen reported in some cases, which probablyrepresents spurious elevations caused by interfer-ence of the assay by iodine.25 Serum chloridedeterminations made by ion autoanalyzers resultin false elevations in serum chloride levels evenat low iodide levels. In the presence of elevatediodine concentrations, the silver halide precipita-tion assay should be used for an accurate measure-ment of serum chloride concentration.36

Renal failure and hepatic dysfunction havebeen reported in patients with high plasma iodinelevels.17,26 A specific renal lesion in the humanhas not been identified. Acute tubular necrosiswas found on results of autopsy performed in 2patients, and pronounced vacuolar degenerationof proximal convoluted tubules was found inanother who reportedly died with high serumiodine levels.17,32 Studies in animals have foundsome functional and microscopic hepatic andrenal damage. Fatty changes in liver and necrosisof convoluted tubules with numerous hemoglo-bin casts in kidneys were the histologic changesnoticed in mice after administration of iodates.37

In humans with fatal iodine poisoning there weregross changes in the liver and kidneys, consistingof congestion and cloudy swelling.17 The bio-chemical basis for the clinically recognized tox-icity of iodine in such tissues as the liver andkidney is unknown. In this respect, we offer thefollowing speculation with regard to a possiblemechanism not previously incriminated in thepathogenesis of clinically observed iodine-in-duced organ injury. This is based on the fact thatiodide is a member of the halide family, whichalso includes chloride and bromide. Under the

Table 2. Features of Iodine Toxicity

Fever, rash, nausea, diarrhea (levels � 30 �g/dL)High anion-gap acidosisAcute respiratory distressHypo/hyperthyroidismHypernatremiaHyperosmolalityConfusionSeizuresHepatic dysfunctionRenal failureDeath (reported at levels � 7,000 �g/dL)

Table 3. Patients With Elevated Iodine Levels After Treatment With Povidone-Iodine for Burns, Ulcers, andMediastinal Irrigation

StudyIodine Level

(�g/dL) Side Effects Condition RRT Outcome

Lavelle et al17 41,300 Metabolic acidosis, renal/respiratoryfailure, liver dysfunction, confusion,agitation

Burns HD Died

30,000 Metabolic acidosis, renal/respiratoryfailure, liver dysfunction, confusion,agitation

Burns HD Died

10,000 Metabolic acidosis, renal/respiratory failure Burns HD DiedPietsch and Meakins32 48,000 Metabolic acidosis, renal/respiratory failure Burns HD Died

17,600 Metabolic acidosis, renal/respiratory failure Burns HD DiedScoggin et al33 Metabolic acidosis, hypernatremia Burns None DiedAiba et al34 20,600 Metabolic acidosis, renal/respiratory failure

liver dysfunction, confusion, agitationBurns None Died

Glick et al28 9,375 Metabolic acidosis, confusion, agitation Mediastinitis None DiedD’Auria26 7,000 Metabolic acidosis, hepatic dysfunction Hip wound None DiedCampistol et al29 — Acute renal failure Mediastinitis None DischargedAlvarez41 4,500 Diarrhea, neutropenia, confusion None DischargedDe la Cruz et al36 2,700 Metabolic acidosis, renal failure Decubitus

ulcerNone Discharged

Zec et al27 1,200 Seizures, renal failure Mediastinitis None DischargedRyan et al30 Metabolic acidosis, renal failure Mediastinitis HD DischargedCurrent case 1,200 Hepatic and renal failure Mediastinitis HD and Discharged

CRRT

DIALYTIC THERAPY OF IODINE TOXICITY 705

catalytic effect of myeloperoxidase, an enzymereleased in large amounts by activated leuko-cytes, and in the presence of hydrogen peroxide,halides generate hypohalous acids, the latter rep-resenting highly potent oxidants38; the latter oxi-dants account, at least in part, for the killing ofmicroorganisms by activated leucocytes and fortissue injury incurred by the influx and activationof leukocytes during inflammation.38 On a ki-netic basis, iodide interacts much more effec-tively than either chloride or bromide withmyeloperoxidase/hydrogen peroxide.4 However,because of the markedly greater amounts ofchloride (approximately 100 mmol/L) as com-pared with bromide (50 �mol/L) or iodide (0.5�mol/L) normally present in extracellular fluid,chloride usually is favored over iodide or bro-mide as the substrate for myeloperoxidase.4 Thispredilection for chloride over iodide by myeloper-oxidase may not be true in settings in which largeamounts of iodide are present, as occurred in ourpatient. We thus speculate that in our patient,ischemia or sepsis recruits and activates leuko-cytes in the liver and kidney; in turn, myeloper-oxidase, released by activated leukocytes, uti-lizes iodide, present in copious amounts inextracellular fluid, to generate the correspondinghypohalous acid. Oxidant-dependent pathwaysof tissue injury thus may be driven by largeconcentrations of iodine in plasma and tissues.

Our patient had both ARF and liver dysfunc-tion. He had underlying renal insufficiency, whichgradually got worse with the initiation of povi-done-iodine irrigation. Although he was not oli-guric, it was quite evident from his BUN, creati-nine, and plasma iodine levels (relativelyunchanged over 36 hrs; 1,190 �g/dL B 1,100�g/dL) that significant renal clearance did notoccur. The most likely potential causes of hisARF were sepsis, hypotension, contrast, or io-dine. We cannot determine with certainty theprimary factor, but the one that correlates thebest is the iodine irrigation. His sepsis syndromewas improving even before changes in his urineoutput or creatinine level became apparent. Like-wise, he had not had significant hypotension forseveral days, and his renal function was stablefor 10 days after his cardiac catheterization,making contrast nephropathy less likely. Thus, itseems likely that the iodine irrigation may havecontributed to his ARF. In this respect, it is

important to note that although it is rare, ARFduring mediastinal irrigation with povidone-iodine has been reported previously.29,30 In addi-tion, it was recently also seen after irrigation of arecurrent lymphocele.39 However, iodine-in-duced liver dysfunction is well established andwas the primary concern in our patient. His liverfunction deteriorated with iodine irrigation, andit continued to do so despite discontinuation ofiodine for 36 hours (likely secondary to thepersistently high iodine levels). Imaging studiesof the hepatobiliary system and further bloodtests failed to find any obvious cause for hepaticdysfunction. After initiation of RRT with theconsequent decreases in plasma iodine levels,liver function strikingly and quickly improved(Figs 1 and 2A-C). Renal recovery lagged sev-eral days behind.

The efficacy of HD for removal of iodine hasbeen examined previously, particularly with re-

Fig 2. Decrease in serum AST (A), ALT (B), andalkaline phosphatase levels (C) after initiation of RRT.

KANAKIRIYA ET AL706

gards to contrast media–derived iodine. For in-stance, Matzkies et al40 compared the clearanceof iopromide by either cuprophan or polysul-phone filters. They reported an iodine clearancerate of 87 mL/min using cuprophane membranesand 147 mL/min using F-6 polysulfone mem-branes. In addition, they also found that addi-tional ultrafiltration resulted in further increasein plasma clearance of iodine. Our HD data werecomparable with their findings, albeit somewhatlower: we estimated iodine clearance of HDusing F-80 membranes to be 120 mL/min. Inaddition, the current report extends the previousones to show that CVVHDF is quite efficient atremoving iodine as well. The fact that iodine iscleared by this modality is not surprising (ahigh-flux filter is used), but the magnitude of theremoval (considering the dialysis flow rate) is.This suggests that the high hemofiltration rateswere likely contributing significantly to iodineclearance. The CVVHDF clearance rates were44 mL/min on the first and 37 mL/min on thesecond day. The small differences in clearancerates likely represent variability in the effluentiodine measurements and not significant changesin actual clearance.

Mediastinal irrigation with iodine-containingsolutions can cause accumulation of largeamounts of iodine in blood and cause significantmorbidity and mortality, especially in the settingof renal failure. Thus, we suggest that patientsreceiving this therapy, particularly those withrenal insufficiency, should be monitored closelyfor signs and symptoms of toxicity, and plasmaiodine levels should be checked. In the case ofiodine toxicity, discontinuing the iodine irriga-tion combined with HD or CVVHDF can be usedto decrease the iodine concentration to nontoxiclevels. Alternatively, we speculate that concomi-tant RRT during iodine irrigation may be at-tempted to maintain the systemic iodine levels atnontoxic levels in patients that are at high risk oralready developing signs of iodine toxicity.

REFERENCES

1. Shelanski HA, Shelanski MV: PVP-iodine: History,toxicity, and therapeutic uses. J Int Coll Surg 25:727-734,1956

2. Zamora JL: Chemical and microbiologic characteris-tics and toxicity of povidone-iodine solutions. Am J Surg151:400-406, 1986

3. Rodeheaver G, Bellamy W, Kody M, et al: Bactericidal

activity and toxicity of iodine-containing solutions in wounds.Arch Surg 117:181-186, 1982

4. Furtmuller PG, Burner U, Obinger C: Reaction ofmyeloperoxidase compound I with chloride, bromide, io-dide, and thiocyanate. Biochemistry 37:17923-17930, 1998

5. Gerschenfeld L: Povidone-iodine as a topical antisep-tic. Am J Surg 44:938-939, 1957

6. Thurer RJ, Bognolo D, Vargas A, et al: The manage-ment of mediastinal infection following cardiac surgery. Anexperience utilizing continuous irrigation with povidone-iodine. J Thorac Cardiovasc Surg 68:962-968, 1974

7. El Oakley RM, Wright JE: Postoperative mediastinitis:Classification and management. Ann Thorac Surg 61:1030-1036, 1996

8. Satta J, Lahtinen J, Raisanen L, et al: Options for themanagement of poststernotomy mediastinitis. Scand Cardio-vasc J 32:29-32, 1998

9. Bryant LR, Spencer FC, Trinkle JK: Treatment ofmedian sternotomy infection by mediastinal irrigation withan antibiotic solution. Ann Surg 169:914-920, 1969

10. Hunt JL, Sato R, Heck EL, et al: A critical evaluationof povidone-iodine absorption in thermally injured patients.J Trauma 20:127-129, 1980

11. Vratsanos SM: On the structure and function of thepolyvinylpyrollidone-iodine complex, in Degenes G (ed):Proceedings of the International Symposium on Povidone.Lexington, KY, University of Kentucky, 1983, pp 289-301

12. DiGenis GA, Rokem S, Eckert D, et al: Studies on theassociation of 14 C povidone and 131 I iodine complex withred blood cells and bacterial membranes, in Degenes G (ed):Proceedings of the International Symposium on Povidone.Lexington, KY, University of Kentucky, 1983, pp 302-311

13. Horn D, Dltter W: Physical-chemical fundamentalsof microbicidal action of povidone-iodine, in Degenes G(ed): Proceedings of the International Symposium on Povi-done. Lexington, KY, University of Kentucky, 1983, pp120-140

14. Scherr DD, Dodd TA: In vitro bacteriological evalua-tion of the effectiveness of antimicrobial irrigating solutions.J Bone Joint Surg Am 58:119-122, 1976

15. Ahrenholz DH, Simmons RL: Povidone-iodine inperitonitis. I. Adverse effects of local instillation in experi-mental E coli peritonitis. J Surg Res 26:458-463, 1979

16. Berkelman RL, Holland BW, Anderson RL: In-creased bactericidal activity of dilute preparations of povi-done-iodine solutions. J Clin Microbiol 15:635-639, 1982

17. Lavelle KJ, Doedens DJ, Kleit SA, et al: Iodineabsorption in burn patients treated topically with povidone-iodine. Clin Pharmacol Ther 17:355-362, 1975

18. Arango A, Lester JL 3rd, Martinez OV, et al: Bacterio-logic and systemic effects of intraoperative segmental bowelpreparation with povidone iodine. Arch Surg 114:154-157,1979

19. King IR, Diddle AW: Protein-bound iodine and T4tests after vaginal application of povidone-iodine. Am JObstet Gynecol 108:1175-1177, 1970

20. Knolle P, Globel B, Globel H, et al: Release of iodidefrom povidone-iodine (PVP-I) in PVP-I preparations. Areview of iodide tolerances and a comparative clinical phar-macological study, in Degenes G (ed): Proceedings of the

DIALYTIC THERAPY OF IODINE TOXICITY 707

International Symposium on Povidone. Lexington, KY, Uni-versity of Kentucky, 1983, pp 342-369

21. Quagliana JM: Effect of topical povidone-iodine (be-tadine) on serum protein-bound iodine. J Clin EndocrinolMetab 23:395-396, 1963

22. Higgins HP, Hawks GH, O’Sullivan M, et al: Theeffect of povidone-iodine (betadine) on serum protein-bound iodine when used as a surgical preparation on intactskin. Cancer Med Assoc 90:1298-1300, 1964

23. Ninnemann JL, Stein MD: Suppressor cell inductionby povidone-iodine: In vitro demonstration of a conse-quence of clinical burn treatment with betadine. J Immunol126:1905-1908, 1981

24. Zamora JL: Iodine toxicity. Ann Thorac Surg 41:462-463, 1986

25. Aronoff GR, Friedman SJ, Doedens DJ, et al: In-creased serum iodide concentration from iodine absorptionthrough wounds treated topically with povidone-iodine. Am JMed Sci 279:173-176, 1980

26. D’Auria J, Lipson S, Garfield JM: Fatal iodine toxic-ity following surgical debridement of a hip wound: casereport. J Trauma 30:353-355, 1990

27. Zec N, Donovan JW, Aufiero TX, et al: Seizures in apatient treated with continuous povidone-iodine mediastinalirrigation. N Engl J Med 326:1784, 1992

28. Glick PL, Guglielmo BJ, Tranbaugh RF, et al: Iodinetoxicity in a patient treated by continuous povidone-iodinemediastinal irrigation. Ann Thorac Surg 39:478-480, 1985

29. Campistol JM, Abad C, Nogue S, et al: Acute renalfailure in a patient treated by continuous povidone-iodinemediastinal irrigation. J Cardiovasc Surg (Torino) 29:410-412, 1988

30. Ryan M, Al-Sammak Z, Phelan D: Povidone-iodinemediastinal irrigation: A cause of acute renal failure. J Car-diothorac Vasc Anesth 13:729-731, 1999

31. Glick PL, Guglielmo BJ, Winter ME, et al: Iodine

toxicity secondary to continuous povidone-iodine mediasti-nal irrigation in dogs. J Surg Res 49:428-434, 1990

32. Pietsch J, Meakins JL: Complications of povidone-iodine absorption in topically treated burn patients. Lancet1:280-282, 1976

33. Scoggin C, McClellan JR, Cary JM: Hypernatraemiaand acidosis in association with topical treatment of burns.Lancet 1:959, 1977

34. Aiba M, Ninomiya J, Furuya K, et al: Induction of acritical elevation of povidone-iodine absorption in the treat-ment of a burn patient: Report of a case. Surg Today29:157-159, 1999

35. Cohn BNE: Absorption of compound solution ofiodine from the gastrointestinal tract. Arch Intern Med49:950-956, 1932

36. Dela Cruz F, Brown DH, Leikin JB, et al: Iodineabsorption after topical administration. West J Med 146:43-45, 1987

37. Webster SH, Rice ME, Highman B, et al: Toxicologyof potassium and sodium iodates: Acute toxicity in mice.J Pharmacol Exp Ther 120:171-178, 1957

38. Halliwell B, Gutteridge JMC: The chemistry of freeradicals and related reactive species, in Halliwell B, Gut-teridge JMC (eds): Free Radicals in Biology and Medicine(ed 3). Oxford, England, Oxford University Press, 1999, pp36-104

39. Manfro RC, Comerlato L, Berdichevski RH, et al:Nephrotoxic acute renal failure in a renal transplant patientwith recurrent lymphocele treated with povidone-iodineirrigation. Am J Kidney Dis 40:655-657, 2002

40. Matzkies FK, Reinecke H, Tombach B, et al: Influ-ence of dialysis procedure, membrane surface and mem-brane material on iopromide elimination in patients withreduced kidney function. Am J Nephrol 20:300-304, 2000

41. Alvarez E: Neutropenia in a burned patient beingtreated topically with povidone-iodine foam. Plast ReconstrSurg 63:839-840, 1979

KANAKIRIYA ET AL708