Peripelyic Extravasation During Intravenous Urography, Evidence for an Additional Route for Backflow After Ureteral Obstruction

1961

Co.

PERI PELVIC EXTRA.VASATIO::'\ DURlNG I~\iTHAVENOl}S 1:noGH\.PilY, EVIDEi'fCE FOR AN ADDITIONAL ROUTE FOR BACKFLOW ;\FTER rRETEH.AL OBSTRUCTION FRANK HINMAN, JR. Frn111 the Dwpartrnenl of Surqeru, Diui.sion of Croloqy, lini1,ersity of Cu/1:fomiu Schon!

or Jlo/icine,

San Fra11cisco 22, Cal. Extnixasation of urine about the reual pclYis

,tml npper ureter dnrin1-s i11trnvcuous urography, which wr have studied in :S patients during ltrC'ternl obstrnction, doeR not indicate rupture of th;-: peh·is hut proves to be mc·rely an extreme form of pyelcminous baekflow The prrsent rqiort deRcribes our attempt to ckmormtratc the quasiphysiologic 11at,ure of thi, rdatin,1:1' infrequent phenomenon by means of obsen·ations on human being~ anrl by injeded rnrrelatC'd 1,·ith deseriptions of similnr phenomrna made by investigatorn. This may contrilmte to the 1mdcrstancling of ~u.eh related problems m, periurl'teral fibrosis, perinretl'ra.l ''f'uema,"' the ancl lornl effect::; of ureternl obRtruction, and "physiologic'' extnwa,mtion in l'ontrast to actual pelvic rupture. It, will be shomt that under c:onditions of complrk 1.tn:teral obstruction, hackflow may start \Ya)' of the forniccR, paRs along the calyceal ,rnll, ancl subsequrntly 1.Yill rencli the pcripcll-i(' tissrn's (fig. CLIN!C.\L OBSl'~RYATl.ONS ON P8BlPF:LYJC EXTHA Y ASXTIO>i

Recent experienct' with 5 tiatil:nts kcl to the iollo\1·ing ohservaticm:,. A. The "edema" fluid cornrnonl~ found periurctera.lly at operation for ureternl obstruction is not neecssarily lymphatic in origin. An azo U)yriclium), aclminister('d before e;-; plomi,ion uf thr flank in a worrn1n with un accidentally ligated Hrnter, wa::, seen in higb concentration in the; riuid abmtt the p,-J,,is and ureter at operation. Thi~ 1Yas uot true lymphatic fluid, t3il)(:e the concentration of the fhowPd tbat it urose rather frorn the rn1ml pelvi,s. its diffuse distribu Uon (in contrast t,o its being confined to the dmnnr:ls s0en in lympbat1c biu·kf10I1 at 1.1rogand its association with "exi,ra,-asatecl" 8,L annual meeting of AmNicn.1.1 Association 8urgeons, De[J,rbnrn. :'vlich.,

:385

contrast nwdium which \\'as sc,'n on 11rogrnph_1 in this patient (see latrr) leads tis to co1wludc that it originated from the renal 1wh·is, passmg by way of the fornices out tlt" hdns. B. The, contrasl rrwdium distri l>t,ted nbout i:.]H· renal pch·is on intnwenous ureteml obstrnction, whid1 of peripeh·ic extnwasation, les tHOl'l' ,lirrd. clinical evidence of comnrnniuition lJetm-'t'll th'

fornfres and the pcripel ,·ic space. Five l'ascs of pcripeh-il' cxt,nL\-as:1t10n lia vc been seen arn:l arc cll'scribcd 1). Case i. N. T., a 1,oman, sufl\,,n·,! left flank pain mi _:\fay 4, I ~hi\J, five ,1ft.cr total hysten:ctomy. lntrav<'llOlt,; m·ogrnms taken on the lcigltth sh01rnd rstrn \·n~:ition abont Lhe left renal pelvis (fig. 2, .4 and lfi. Rdrogn,rle pyrlograms, taken after a rntJwt<.·1 harl bet:n nmnipulakd through ,,. ck·1·i;itC'd ,Lt·· notic segment of thP ureter 2."i cm. :1bcl'1'<' 1hc: bladder, showed no peh-ic leakage 2, C) At neplirostomy (elected hc·,·,rns<·' t>f the: <'Xtr::, vasation aml the inability to reliPYL' the low,:1 nreteral obstruction rnmpktc>ly) tlw 1.i~s1 1,· 1vithin Gerota's fasc:ia wa, 1·er>· 1wt :1ml cle.eply Rt11inPd \Yith the pyriclium thnt thl' tient had taken on,11:v prior t" <,pc-r:,tion. Th,· pelvis itsrlf shomxl no abnonurdities Afte,r 2 montlrn, ren'ak
n,mo,·ed. The urine is no,y skrile. Comment: Peripl:h·ic cxtrn.va,,ntion \1as s,•1·1, on roentgenologic studies aml 11,i,; rontirnwd ln the finding; of azo dye irrn1de Gl'l'Ota · fas1·w. Had we realized the, n:itur(' of 11,e roentgcnologic changes from thv :1!101-:m·P of ll'ais · a,ge on the rl'trograde \\'t' 1nnild hn,:e :wuidnl 1wphrostomy Case 2. J. T., :i ,knti·4. 11·ho hac! history of colicky rigbt flan!, fur 20 w,11,-, was admitted on :\Iay 13, 19i\9, after :2-l Lo1tr~ ui sharp riglit lower vomiting;

386

FRANK HINMAN, JR.

Comment: Since we saw this patient with extravasation soon after the patient in case 1, we merely extracted the stone and provided no perirenal drainage. Case 3. N. D., a 48-year-old man, was awakened by pain in the left lower quadrant which radiated into the left flank. He had no urinary symptoms. He was admitted to the hospital on May 2, 1955. An intravenous urogram taken at this time showed that function on the left was delayed and there appeared to be extravasation of medium around the renal pelvis and upper third of the left ureter (fig. 4, A). The following day, the patient's temperature had risen to 38.4°C. Cystoscopy showed some delay in the appearance of indigo carmine on the left. Fifteen cubic centimeters of clear urine were aspirated from the left renal pelvis. Films made after the pelvis was filled showed no extravasation (fig. 4, B). The catheter remained in place overnight. Intravenous urograms taken after 8 days, and taken again 2 years later, were normal. No stone was recovered. Comment: Although the extravasation was not interpreted as physiologic at the time, no operative procedure was done since the retrograde pyelogram showed no leakage from the renal pelvis. The patient's subsequent course demonstrated this to be the correct approach. Case 4. L. L., a 64-year-old man, had an attack of severe, steady, right flank pain with

Intravenous urograms showed extravasation of dye about the right renal pelvis (fig. 3, A). The column of dye was obstructed about 12 cm. below the ureteropelvic junction. Little calyceal dilatation was present. A stone manipulation was done, during which a 3.9 mg. calcium oxalate calculus was extracted. An immediate bulb ureterogram failed to produce extravasation (fig. 3, B). The patient was discharged 2 days later.

Fm. 1. Route of backflow in peripelvic extravasation. TABLE

1. Five cases of peripelvic extravasation during intravenous urography I

Case Number Pt. --- ---

HT

1

Cause

Partial ureteral ligation

I

Peripelvic extravasation

I ? NormaJ

I 2 I

U reteral calcul us

Peri pel vie extravasation

KD

? Ureteral cal-

Peripelvic extravasation

culus 4

I LL

I

? Ureteral cal-

Peripelvic extravasation

culus

Pyelogram

JR

U reteral calculus

I

Result

Drainage; nephros1,omy

Ureter patent in 2 months; patient well

Norrnal

Stone m anipulation

Discharged in 2 days

Normal

Indwelli ng ureteral catheter

Immediate recovery

Not done

Drainage of flank, pyelotomy

Drainage 3 weeks; recovered

I

I

I

I I

5

Oper ation

I

JI I

3

Retrograde

Intravenous Pyelogram

Peripelvic extravasation

I

I Not done I I

I Ureterol ithotom y I

Discharged in 6 days

PERIPELVIC EXTRAVASATION DURING INTRAVENOUS UROGRAPHY

387

Fm. 2. Case 1. Partial ligation, lower ureter. A, 5 minute intravenous urogram shows absent psoas shadow and beginning peripelvic accumulation of contrast medium. B, 20 minute urogram. Contrast medium now fills entire peripelvic space. C, retrograde pyelogram later same day shows no direct pelvic leakage.

Fm. 3. Case 2. A, ureter al calculus. A 20 minute intravenous urogram shows diffuse peripelvic distribution of contrast medium. B, bulb pyelogram which shows that no direct pelvic leakage has occurred. radiation into the groin on March 5, 1955, which was associated with tenderness at the right costovertebral angle. Intravenous urograms, which were taken immediately, showed no calculi, but demonstrated extravasation of dye about the right renal pelvis and along the course of the ureter (fig. 5). Immediate exploration revealed wet retroperitoneal tissue, the fluid spreading all the way to the pelvic brim. No perforation was

found in the pelvis, nor was a stone located in the ureter. The pyelotomy was left open and the retroperitoneum drained. His postoperative course was marked by prolonged urinary drainage but he subsequently completely recovered. Comment: Operation was done on the assumption that pelvic rupture had occurred, but none was found. In retrospect, nonoperative treatment would have been adequate, especially if a retro-

388

FRANK HINMAX, JR.

Fm. 4. Case 3. A, probable ureteral calculus. A 15 minute intravenous urogram shows peripelvic extrasasation. B, retrograde pyelogram. No pelvic defect seen.

FrG. 5. Case 4. Probable ureteral calculus. A 20 minute intravenous urogram shows marked peripelvic and periureteral spread of contrast medium.

Fm. 6. Case 5. Ureteral calculus. A l hour intravenous urogram shows obstructing stone and distribution of contrast medium about upper margin of pelvis.

grade pyelogram had been done, which would have shown that pelvic rupture did not exist. Case 5. J. R., a 44-year-old psychiatrist, had an attack of acute left renal colic with radiation of pain to the testis on April 26, 1960. A plain film showed a 8 mm. stone in the region of the left renal pelvis, but intravenous urograms the following clay showed that the stone had moved 8 cm. down the ureter. Extravasatcd contrast medium could be seen about the superior border of the pelvis (fig. 6). Because the stone had not moved by the following day, it was removed operatively. Considerable "edema" was encountered inside Gerota's fascia at operation. Comment: Ureteral obstruction resulted in a

localized extravasation of contrast medium, the extent of which was demonstrated at operation by the distribution of fluid (urine) in the periureteral spaces. C. Abnormal fibrosis about the upper ureter and renal pelvis may be explained by peripelvic extravasation. An illustrative case follows: J. S., a 31-year-old man, had had previous attacks of right renal colic with hematuria, but no stones were recovered. Ten clays before admission he noted hematuria and dysuria. On December 13, 1959, eight hours before admission, he had an attack of right flank pain with radiation into the testis, and he vomited. An intravenous urogrRm mRde on entry showed

PFRIPl
FIG. 7. Cross-section of calyx [llld margin nf pelvis showing; spread nf Jnctia ink i,o perifornin1I 1·c-g;H)I,. ('..l-\lh'l,old iuf:mt).

stone :tt tlic right urctl'.ropelviG ,iunctiou of tl1e 10\n'r segnwut of a double kidrn>.1-. No cxtrnvasation \\'D-S ~e(:n. On t.lw t.hircl d,i:1·, since thr stone had not moved. a urett>rolithotomy was done through a Folt,_1- mn.,c k-splith11g i11cisio11. l\Iinimal rrtro peritoneal "cdc;rna'' was found, but the ureter was i111·oln·cl in a clense fibrous reaction which bound it finnlv to thr psoas as if a previ-ous operation had been performr;cl on the ureter. The stone 11·as extracted through a sbort incision in tiw uretel'. rrine clrainrd for 7 cb_rn, and recm·rry was complete. Comment· It seems reawnable to explain tlwsc fimlingc on the basis of the effects of the peripeh·ie ueinar~· extrnvasatimi th[lt had occurred 1i-ith tlw pati<'ut'~ JJIC\"ions attac:ks of mekml obftrnc:tion. An i11stance of exten~in: periureternl fibrosis obstructing the upper ureter in association 11·ith a peh-ic stone has romc to the author's attention. Lt \Yill be n'portcd Dr. T. L. Griffith. It is possible that this cas(, rcprc,sents a morr extreme late result of peripdvic extravasation, perhaps assoei,1tecl witb infortrcl urine;. EXPEHDlE'fl'."cL OBSJ·:R\'A'J'TON,i ON Pf;RTPELYJC

8XTHXVASc\.TT01'

Fresh kidneys from adult anrl infant patients were obtained t1.t autopsy, and were drarerl of perin:nal fat. ~.'\,_ blunt mounted on a .5 cc syringe. ,,as applied to the: junc-turE' of the renal

peh-is with the hilar conm•ctin' tissue and 2 1,1i J. cc· of India ink was i11jcckd umlt>r pn-'ssurc. ::\[icroscopic sections shower! tlmt the, i1,l.; sprPad along th<' wall of the peh-is, J)('rme11ted tk connectin, tissnr s(q.mrating the from tlw renal 1mrenchyma, and finally distrilmted it~eir around the papilla in the fornical tissue (fig. ; , This distribution of is interpreted as (Ti
The rrlative importam'r-, of the 1·r,nor1s ,,n,\ J~·mphatic drainage of tlw kidne:· in <'arrying ,,ff the urine which rscapes by backflow aftt'l' ,,J ,_ struction to the ml't,l,r is being n,-r·xmni1wdi· Hern· the urine reaches tlwsc 1-asculnr strncturc, had been clc,tennim:d _:; ' Onr nh ::;erTations that pyridium-stainr,cl urnw wil: infiltrate the p1,rirrnal com1ectivl' tissue afo,, Jo1Yer ureteral obstruction, and tliat frank p1•1·1 peh·ic and periurckral cxtrn,vasation uf op,1.qll1' urine can he seen during; intrn.1-c'nou~ J Good1Yin, W. E. and Kanfm,rn, J .I.· Rell1J; lymphatics. Il. Preliminary exp1criments ..J. !1 rol

76: 702, HJ5(i

'i\Iurphy, J . .J., Myint, l\l. K Rnttner, W. H , Klaus, R. and Shallow, J .. The pha t,l<' sydl ,,m of the kidney . .J. l~rol., 80: 1, 'J Fuchs, F. · Tbc Fllrn of vllut,er 'fhrnng:l, ch,, Kiclnei- \rc1Y York: l\lanlrnttan Priutin~: 1[)44

.

arath, P. A.: Renal Pd vis ancl ·ureter York: nrune [lflll Stratton. 1%1 1 :\

390

FRANK HINMAN, JR.

filling of tubules (pyelotubular backflow)

--->

increased intrapelvic pressure penetration of fornix (fornical backflowJ

/

/'

-----+

~

reabsorption absorption by lymphatics (pyelolymphatic backflow) rupture into veins (pyelovenous backflow) dissection in sinus renalis (peripelvic extravasation)

t

absorption by lymphatics FIG. 8. Schema of ultimate routes of urinary backflow after ureteral obstruction under similar circumstances, led us to re-examine the function of the renal sinus during "pyelovenus" backflow. Backfiow (the relation of peripelvic extravasation to routes of backfiow) (fig. 8). In order to understand peripelvic extravasation, it is necessary to relate it to other forms of backflow. Three major types of backflow can be described, which are actually different phases of the same process, since all are dependent on penetration of the fornix of a calyx by urine (fig. 1). Fornical backflow (pyelosinous transflow): a. "Pyelolymphatic" backflow (sinolymphatic absorption). b. Pyelovenous backflow (sinovenous ingression), to which we would add the form described in the present series: c. Peripelvic extravasation. Notice that "pyelovenous backflow," a term used by Hinman, Sr. in 1924 5 and used generally for all types of backflow, forms just one type in the classification, since it actually is part of the fornical backflow in which the urine escapes by way of the lymphatics, the veins, or, as we shall see, by way of the pericalyceal and perirenal tissues. (We will not discuss a separate type of backflow, pyelotubular backflow, which has a characteristic roentgenographic appearance in retro-· grade pyelography (tufts are seen as a result of filling of the collecting tubules by way of their openings on the papilla). Besides, its appearance on intravenous urograms cannot be distinguished from the more common occurrence of concentration of the dye in the collecting tubules during excretion.) Fornical backflow, a necessary precursor to pyelolymphatic and pyelotubular backflow is, as would be expected, a more common finding on 5 Hinman, F. and Lee-Brown, R. K.: Pyelovenous back flow. J.A.M.A., 82: 607, 1924.

urograms (including those made not only by retrograde injection but also by the intravenous method) than other forms of backflow. The hornshaped outflow (N arath) is readily recognized roentgenographically The fact that such backflow occurs on intravenous urography during abdominal compression or with some other obstruction to the ureter is less well appreciated. Fuchs in 1931 6 first described the phenomenon which occurred during an attack of renal colic, the cause of which was not determined. Since then, well over one hundred cases have been described. The over-all incidence of backflow with compression in intravenous urography is about 2.2 per cent (table 2). Olsson12 described and illustrated the sequential changes of fornical backflow in detail. a. Pyelolymphatic backflow may occur with either retrograde or intravenous pyelography. The wavy outlines of the lymphatic channels coursing toward the midline are not uncommonly seen after forceful retrograde injection of opaque solutions, arising from an irregular break which may be seen in one or more fornices. Reabsorption by this route may be complete in a few minutes. b. Pyelovenous backflow: The urine which contains opaque medium may tear the veins away from the renal parenchyma by distending the tissues just outside the fornix, and cause the dye to spread into a venous plexus. This is a common occurrence in experiments on excised kidneys. It is much less often observed under clinical conditions, probably because the dye is carried away too rapidly to register on the film or because the insult is less traumatic. c. Peripelvic extravasation: ·yve can now ask 6 Fuchs, F.: Die Hydromechanik der Niere. Anatomische und experimentelle Grundlagen, biologische und klinische Bedeutung. Ztschr. urol. Chir., 33: 1, 1933.

PERlPELVJC EXTRAVAS_~T[Oc\f DUR[i\'(; l:\"TRAVENCH1 S UIWG!tAYHY

fABLI,

2. Ir,trauenons 11rourams showinq only .fornical rupwrc Causi-'

~-\uthor

Number of Cases UretcraJ stone , --------

Fuchl, 5 Hemlrioek; F1tchs' Wulff 0 . Lindbom 1 ". Lindborn 10 Coliezn. Olsso11 12 . Bo:yarsk~~ flIHl associa1Jes 1 ::i Total

case;;

1931 193-1 JD36

(eolic)

l 1

HJ3(j 1943 1943

(1 . 2c;c)

X X

15 (l . G';;,)

X X

8G (2 .5(?c) 17 (2 .9%) 124 (2. 2r;70

thl: question: Ho\\ is peripelvi~ extrnvasation related to other formA oJ' backf!ow" Our study of crosR--Rections of normal human kidnrys ,vhich we injected with India ink from the rim of the hilus renalis suggests the probable rnute for csc'llflC of conk11ts after they tn1.verse. the fornicaL barrier, that is, by way of the sin11s mid hilus. * Howt•ver, earlier ,vorkcrs 7

Henclriock, A. lntrnvenijse Urogrnphie und t"bertritt Zrntrnlbl. Chir., 61: 1822,

8

Fnchs_ F .. Xiereulwcke11exh·avasat.e. l':1.sclu.

urn!., 30: ±2:3. 1936.

"Wulff H. 13.: Die Zuverhissigkeit dcr Riint g,mcliagnostik~liesondcrs hinsichtlich des Wertes der Crographie--unc! die Prognose bei Nieren-und Harulcilerntrinen. Al'ta rndiol., suppl. 32, HJ3(i 1 " Liudhom. A.· Fornix lrnckflow in excretion 11rogrnpli_1·; its signi ficnnce in the differential tliagnosis of l..ubercnlosis oJ' the kidney. Acta rn,!iol 24: -±11, 194:3. 1L R.: Les t,igneH racliologiqncs de staAe et de snrprnssion urdero-retmle an conrs de l'urogn.1.phie intra-veinense. .J de radiol. et d'clectrol.. 28: :311 Hl47. '"Olssor1s 0.: Stndies on baek-f!o\\· in excretion mogrnpb>-·· Acta n1diol. suppl. 70, HJ4k. 13 Boynrsk\', S.. A. n.nd Baylin, Cl. J.: E'(trnvasnt.ion .in nrograpby. Urol. .mternaL, 1: l!ll, 1955 ' 1 Dorlnnd's Illrn,t-rnfrcl l\fodical Dictiorrnn-, 2:3rd c,cl. Philadelphia. W. B. Saunders Co., 1957' "Harpn's Latin Dictionary Xew York Arnericm1 lfook Co., HJ07. * Tlie rnnal hilus (pl. hili, pron. hI-1\ (hilus rnrrnlis) is t.hc> "opening; in the> concave edge of thH t hnrngh ,vhich 1,he vessels and nerves (Dorlanc!Ji-,_ and ,ms called the port.a the earlier anatomists. H·ilwn (pl. hila) (a. uf the La.tin nihilnrn, "a litt.le thing, a Latin Dictiorrnry 15 ), is an older form iusist,ed upon hy some authorn, 1wtabl_\' Xarai.rt, lrnt is placed as an altern:1tive by

X X

)

believed that this route was cJosed tu urnH stmc:tural reasons. in Utt: scription, conclndcd that the connectiv<-: ,Yithin the sinus renalis obliterakcl a.ny space, and otlwr invesbguton, l,a ve adnal diaphragm c·.losing tlie hili1s Karath 17 describes these relationships "Close to the hilum the adl1crcncc of the !'Ii,•· sule becomes more and more and here 1 is impossible to the ca.psllle off wit.hoil!tearing into the varenrbymn. The capsule spread in lm1shlike manner tex and over to the extrnn:na.l part or the ureter and onto the blood with their advrntitia. In this \Yay sealed off, the citpsnlc forms a ,c1rnun1 i the pedide. So the capsule rnprcscnts n closed in whid1 the whole complicated ,;t,rncture tJ,, reual pehis, the sinus renalis \\-ith its w,ss,·],; nm 1 the pa.nmchyma is embedded.'' '' HcnY firmly the fibrous caps ult, is aclhc.-·1·e11t t,, the cxtrarcnal part of the pelvis is ,,·ell kno1,. u te every surgeon \\·ho stieks to expose a great.n :1nc:.>. of the tme pelvis for n;asons of lifting the parenc:hyma fl\Y:J.Y from the peh·i:c. 11, I.kw st,rndanl medical dictiollaries. was ttsecl b\' He11le 1 ' in 1872 distinguished tlie hilus reirnlis sinus rcnalis (the hole> or bv the Baslc· Xornina w·ould seem, tJiereforn, to he t.ht" 16 Henle, ,J.: Handbuch der s_vsternatisdtct: Anatomie des lVJem;r:hen. DrnunAclmeig: F. Vimn,:1;:_ nncl Sohn. Hs6H-72. " N arni:h, P. A.: Renal Peh't~ and lhe~er '.\ York. Grune and Stra.tton, 1CJ5!, Ji 15fi.

11s

:392

FRANK HINMAN, JR.

FrG. 9. Relation of capsule to pelvis, according to Fuchs, 6 p. 7. order to accomplish this, the dense connective tissue between capsule and pelvis must be dissected with the scalpel. It is impossible to do it in a blunt manner." "The hilum, though often large in itself, is considerably narrowed by the pelvis and the blood vessels so that the distances which must be bridged by the capsulae are not too great. The various structures that enter the hilum prevent an even and flat formation of a capsular diaphragm. On anatomic section, the formation is rather irregular, but still represents a membrane that seals off the sinus renalis and separates the sinus fat from the fatty tissue of the adipose capsule at the hilum." In summary N arath 18 states, "Leakage of absorbed material from the sinus renalis into the retro peritoneal connective tissue is not easily possible. The capsule with its firm adherence to the parenchyma and the structures of the hilum prevents such an escape." Despite this conclusion, N arath shmvs films (his figures numbered 210 and 220) which we would interpret as showing peripelvic and periureteral distribution of dye, although he labelled them "sinolymphatic absorption." From our injection studies in excised kidneys ,ve determined that once a particulate injected substance reached a point in the intrarenal peripelvic space, it spread widely in that space rather than entering the veins or lymphatic vessels. It would seem reasonable to assume that under some circumstances urine could utilize the same route, in reverse. 1, Ibid, p. 306.

Indeed, Fuchs19 has stated "that if extravasate does not come in contact with a ,,ein, there is propagation towards the hilus and in such cases there are no preformed channels available." The peripelvic connective tissue is relatively loose and infiltrated with fat, so that an anatomic route exists between the rim of the calyx and the renal hilus (fig. 1). The existence of such a route might reasonably be assumed on embryologic grounds since the bud of the wolffian duct pushes its wav into the mesonephric blastema, and in effect, invaginates it around its divisions. In fact, Windholz 20 made use of the pericalyceal fat to visualize the internal structure of the kidney. The observation that a peripelvic route may be taken by urine during retrograde P?elography was first made in 1935 (table 3), when Heckmann21 obtained peripelvic extravasation on a retrograde pyelogram even though he had not passed the catheter into the renal pelvis, and so could not have perforated the collecting structures. Intravenous urograms taken 2 weeks later were normal and no defect was found at autopsy, which was further evidence that actual perforation of the pelvis had not occurred (although the author believed that it had). Others 4 • 19 • 22 have made similar observations. Peripelvic extravasation was first described on an intravenous urogram, made during abdominal compression, by Narath 23 in 1938. His illustration clearly shows extravasation about the pelvis and ureter inside Gerota's fascia. Howeyer, a retrograde pyelogram had been made previously (with the catheter only in the lower ureter) and the kidney was involved in tuberculosis. After removal of the kidney, N arath filled the pelvis with indigo carmine and could detect no leakage. These findings suggest that the extravasation arose from backflow at the fornix rather than from rupture of the pelvis.

1, Kohler R.: Investigations in retrograde pvelograph}'; a roentgenological and clinical study. Acta radio!., suppl. 99, 1953, p. 46. . 20 Windholz F.: The roentgen appearance of the central fat' tissue of the kidney: Its significance in urography. Radiology, 56: 202, 1951. . 21 Heckmann, K.: Ausfliessen des Kontrastmit:· tels in das retroperitoneale Bindgewebe be1 der retrograden Pyelographie. Fortschr. geb. Rontgenstrahlen, 52: 601, 1935. ~ , 22 Puigvert-Gorro, A.: _Tratado de -Crograha Clinica. Barcelona: Ed1tonal Labor, 1944 ... 2, N arath, P. A.: Extrarenal extravasat1on ob~ served in the course of intravenous urography. J. Urol., 39: 65, :1938.

393

PERIPELVIC EXTRAVASATION DURING INTRAVENOUS UROGRAPHY

3. Retrograde pyelograms showing peripelvic extravasation

TABLE

Author

Year

Heckmann 21 . ........ Puigvert2 2 . Narath 4 . Kiihler 19 ...

1935 1944 1951 1953

TABLE

... . . .

Cause

1 1 1 5

"Bulb" pyelogram Retrograde pyelogram Retrograde pyelogram Retrograde pyelogram

.

.

Olsson 12.

Normal Extravasation

Author's Figure Number

3 64 219, 220 166, 17b, 18a

4. Intravenous urograms showing peripelvic extravasation Year

. . . . . . ..

Subsequent IV Pyelogram

of Cases

Number of Cases --- - -

Author

Narath 23 . Renander 24 . Olsson 12 .

Number

1938 1941 1948

1 1 6

1948

3

Ureteral Stone

I

Abdominal Compression, Duration

II

Other

X ? X (all lower ureter)

55, 33, 13 minutes

Author 1 s Figure Number

Treatment

2 1 17b, 18c

N ephrectomy N ephrectomy None

9a, 10a, 10b

None

166, 167 lb 3 1 1, 3 2, 4, 5, 6 2

None None None N ephrectomy Various Various N ephrostomy and drainage Manipulation Indwelling catheter

I

Olsson 25 . Persky and Joelson 26 . Weiner and associates 27 . Fajers and Idbohrn 28 . Forsythe and associates 29 . Forsythe and associates 29 . Hinman ..

1953 1954 1957 1957 1958 I 1958 I 1960

2 1 1 1 3 4 1

X X

Hinman. Hinman (MacDonald).

1960 1960

1 1

X ?

Hinman (Schulte).

1960

1

?

3 4 Ureteral obstruction 5

Hinman.

1960

1

X

6

X ? X

X Ureteral ligation

Pyelostomy and drainage Ureterolithotomy

Since 1938, four more cases of peripelvic extravasation after abdominal compression have been described (table 4). Compression lasted for as long as 55 minutes in one instance. 12 It is probable

that Kitamura's 30 cases are examples of lymphatic backflow. Pathologic obstruction of the ureter, usually by an impacted stone, accounted for 17 more

24 Renander, A. : Another case of spontaneous rupture of the renal pelvis. Acta radiol., 22: 422, 1941. 25 Olsson, 0.: Backflow in excretion urography during renal colic. In: Modern Trends in Diagnostic Radiology. J. W. McLaren, ed. New York: P. B. Roeber, 1953, 2d series, pp. 214-217. 26 Persky, L. and Joelson, J. J.: Spontaneous rupture of renal pelvis secondary to a small ureteral calculus. J. Urol., 72: 141, 1954. 27 Weiner, M. E., Alcorn, "F. S. and Jenkinson,

E. L.: Subcapsular rupture of the kidney during intravenous urography. Radiology, 69: 853, 1957. 28 Fajers, C. M. and Idbohrn, H.: Peri pelvic reflux simulating a tumor of the renal pelvis. Urol. internat., 5: 187, 1957. 29 Forsythe, W. E., Huffman, W. L., Schildt, P. J. and Persky, L.: Spontaneous extravasation during urography. J. Urol., 80: 393, 1958. 3 ° Kitamura, K. and Honda, T.: Statistische Boebachtungen iiber Refluxschatten. Japan. J. Derm. Urol., 36: 537, 1934.

394

FRANK HINMAN, JR.

cases of peripelvic extravasation after intravenous urography (table 4). FACTORS IMPORTANT IN INITIATING AND PERPETUATING BACKFLOW

Kohler 19 found the pressure needed to initiate backflow to be between 80 and 100 mm. Hg, but Ross 31 obtained backflow with pressures as low as 15 mm. Hg. lVIoreover, pressures as high as 100 mm. Hg did not necessarily bring about backflow. In fact, the average pressure during Ross' retrograde studies with backflow was slightly less than in those without backflow. Kiil32 found that the rate of increase in pressure was a more important factor than the absolute nse. That intra.pelvic pressure is reduced during backflow is illustrated clearly by Olsson12 in his figures 5, A and B. Here reduction in the degree of ureteral and pelvic dilatation occurs even though compression of the ureters is maintained. This agrees ,vith experimental observations that once backflow is established, less pressure is required to maintain it. The tone of the renal pelvis probably plays a role, since in all such cases pelvic dilatation is minimal. Observations on the present cases illustrate this point. In each case, although some calyceal overfilling can be seen on the urograms, there is minimal hydrocalicosis. Moreover, some filling of the collecting structures is seen in contrast to the nephrogram without pyelogram usually seen with this degree of obstruction. e can assume that in those patients who show only a nephrogram an adequate route of backflow has not been established and the contrast medium must remain within the renal tubules. Passage of an ureteral catheter immediately produces a pyelogram because it releases the intrapelvic pressure and allows the tubular dye to pass into the renal pelvis. In other cases, in which fairly clear pyelograms are seen but no backflow can be detected, urine probably passes into the lymphatic or venous circulation and proceeds undetected by present radiographic techniques. This "circulation"

,v

31 Ros~, J. A.: One ~housand retrograde pyelograms with manometnc pressure records. Brit. J. Urol., 31: 133, 1959. 32 Kiil, F.: The Function of the Ureter and Renal Pelvis; Pressure Recordings and Radiogr~phic Studies of the Normal and Diseased Urinary Tract of Man. Philadelphia: W. B. Saunders Co., 1957.

. Frn. 10. Intravenous urogram (4¾ hours) with nght ureteral ~bstruction shows persistent pyelogram on opposite side due to contmued backflow of dye from obstructed side. permits the calyces and renal pelvis to fill on the obstructed side, and moreover allows recirculation of the dye so that a pyelogram can still be found on the nonobstructed side hours after the injection (fig. 10). The appearance of extravasated dye during urography performed during ureteral obstruction from calculi occurs rapidly (within 10 minutes) while that induced by concomitant abdominal compression is delayed (13 to 55 minutes). From this it may be conc:luded that pathwavs for backflow become established during tho lm;gcr periods associated with calculous ureteral obstruction so that they appear immediately after injection of the opaque medium. On the other hand if the obstruction and injection are simultan~ous as with abdominal compression, more time is required for the routes to be established and opacifiecl. That the route is not maintained during longstanding obstruction is shown in an oblique way by recent experiments of Guze and Beeson. 33 If cortisone is given during the period of ureteral ligation, the resulting hydronephrosis is much less marked than if no anti-inflammatory agent is used. Our interpretation of these results is that 33 Guze, L. B. and Beeson, P. B.: The effect of cortisone on experimental hydronephrosis following ureteral ligation. J. Urol., 78: 337, 1957.

PERIPELVlC: EXTK-'IVASATI0l\' lllCRlNG INTR:\ VE:\'ffCS \;ROGR:\.PHY

Fil+. Jl, Blocking of perirenal lymplmtics in chronic ttrinary olist.ruciion, from Helmke'"'

om' effect of cortisone i~ to d<:'la.y the perifornical inflammatory reaction and so allow the backflo1,· to continue. Tliis reduces intrapelvic pressure and dt
That. fibro~is in the and prriureteraJ. tissue could result from ''physiologic" urina.ry extra.vasa,tion i~ supported by 8evt:ral obsern1 .. tion~ other than thme, described in the clinical in which fibrosis was found operations for t',al1;uli. Hamperl'l" found material in tbe retroperitoneal tissue at autopsy and described a foreign body reaction to sterile urine. .\lore pertinently, Helmke 3" dcsci-ibccl intt:rstitial infiltration of the pcrirenal tissues in cases of chronic urinary obstruction 11). It is possible that the encl rm ult ol tlwse processes, if cbronieally induced or often re" Hamper!, A. H Viiv11:1dsreaktion ("friimnrnnde-Kropps-reaktion") mot slem och urin. :\:ord. rned., 41: 66, 1949. Helmke, K.: Die Kierenveriindenmgen bei Rarnstarnmg, hesonders iiber die Bildnng von ''Lymphgefiiss-und ,,,,nr,m·n bei c!u-onischer Harnstauung Arch. path. Ann.t., 302; 323, 1938.

peat,ed, would be periJll'h·i,, and p,1riuret112·:i,i fibrosis. kU'.lfMARY ,~i,I> CUNCJ,lJSIOl\S

Peripelvic r·xtravasation, semi b1· i11t.111n,11011s urngraphy aftc,r urderal is ,1.n extreme degree of pdYic bad:flow a, physiologic phenomrnon. The "eclc,rna" fiHid oftl'n Immel in the r•l'n urctcral am! peript
Since completion of the. rn,,,wc;~,:i P,1.wlow"ki (Pawlowski, J. :_\l.: Per-ipelvi,: unne gnmuloma. Am. J. Clin. Path. 34: 64, IH60) ha, dem·ibr,d in greater detail the tissue pcripelvic urinary extravasation. Tbe rwtbor thanh Dr. J. L MacD01mld of Oakland, Cal. am! Dr. ,J. YI.' Schulte or ~hu Francisco, Cal. for pennis,;ion tD r