Splenectomy:Indications and Results in Hematologic Disorders *

WILLIAM R. SANDUSKY, M.D., BYRD S. LEAVELL, M.D., BURTON I. BENJAMIN, M.D.

From the Departments of Surgery and Internal Medicine, University of Virginia Medical Center,Charlottesville, Virginia

THIS is a report concerning the indica-tions for and the results of splenectomyperformed for hematologic disorders. Dur-ing the period January 1, 1946 throughDecember 31, 1962, 254 splenectomies wereperformed at the University of VirginiaMedical Center. Of these 160 were carriedout for non-hematologic conditions, andinclude splenectomy for trauma, as part ofan en bloc resection, incident to the per-formance of splenorenal venous anasto-mosis, for the purpose of facilitating someother surgical procedure and, in isolatedinstances, for tumor, abscess, aneurysm ofthe splenic artery, or for diagnosis. Therewere 94 splenectomies for hematologic dis-orders. These form the basis for this report.

Indications for SplenectomyAs will be noted in the subsequent dis-

cussion, there are several hematologic con-ditions in which opinion concerning theindications for and the timing of splenec-tomy are in dispute. During the periodunder consideration the indications whichwere employed by us have, in many in-stances, remained essentially constant; inothers significant changes in practice havebeen made to conform with new evidenceconcerning therapy, new concepts concern-ing disease mechanisms, as well as to takeadvantage of experience gained during theperiod. Within this framework our indica-tions for splenectomy in heinatologic con-ditions may be stated as follows:

* Presented before the Southern Surgical Asso-ciation, Hot Springs, Virginia, December 10-12,1963.

695

An established diagnosis of hereditary sphero-cytosis was considered sufficient reason for sple-nectomy unless some contraindication to operationexisted. Even if the hemolytic process was com-pensated, or nearly so, splenectomy was advisedbecause of the tendency to cholelithiasis that ex-ists in this disease, to avoid the later complicationsof gallstones, and to avoid the consequences ofan aplastic crisis that sometimes follows an in-fection.

In those patients classified as having hemo-lytic anemia other than hereditary spherocytosis,splenectomy was performed when satisfactory im-provement did not follow the use of corticosteroidpreparations, or when transfusions were requiredor over a prolonged period, or when studiesshowed selective splenic sequestration of Cr'tagged red blood cells.

Splenectomy was carried out on patients withidiopathic thrombocytopenic purpura who did notrespond within six weeks to therapy with adreno-corticosteroids, and on those who have relapsedafter the drug was discontinued.

Splenectomy was performed in those patientswith myeloproliferative disorders who could notbe managed with corticosteroids, testosterone, bu-sulfan, or transfusions, when repeated hemorrhages,frequent infections, recurrent anemia or dis-comfort associated with the abdominal mass ledto invalidism, or when there was selective seques-tration of Cr" tagged erythrocytes in the spleen.

In the group who were diagnosed as aplasticanemia splenectomy was done in the moderatelyill patient who had developed splenomegaly asso-ciated with hyperhemolysis, when it was hopedthat operation might mitigate this aspect of theillness; it was also performed in some seriously illpatients as a last resort measure when the needfor transfuision was almllost continuouis.

All of the splenectomiiized patients consideredin the grooip entitled congestive splenomegaly withhyp)er.sp)leni.sn lhad Laemiiec's eirrlhosis and, inaddition, one or more of the components of hyper-splenism, namely, splenomegaly, anemia, leuko-penia or thrombocytopenia.

SANDUSKY, LEAVELL AND BENJAMIN Annals of SurgeryMay 1964

TABLE 1. Hereditary Spherocytosis

Before Splenectomy After Splenectomy**

Hemato- Reticulo- Hemato- Reticulo-crit cytes crit cytes

Case Patient Sex, Age* % % % % Result Followv up

1. S. M. F 8 29 8 40 2 Excellent 13 yr.2. D. L. S. F 1 mo. 12 12 37 2.8 Excellent 13 yr.3. M. K. F 15 26 - - Excellent 9 yr.4. J. E. W. F 8 21 12 40 0.2 Excellent 4 yr.5. M. A. C. M 54 25 4.7 38 - Excellent 4 yr.6. E. A. C. MI 29 25 6.1 46 - Excellent 3 yr.7. F. C. F 28 32 1.7 32*** - Excellent 21 yr.

94. M. B. T. F 38 13 8 41 1.5 Excellent 21 r.8. G. E. D. F 43 28 15 47 - Excellent 2 yr.9. E. E. F. M 54 6**** 14 15**** - Excellent 1 yr.10. I. K. A. M 2 17 30 39 - Excellent 7 mo.11. P. J. K. F 4 6.2**** 10 13.2**** - Excellent 4 mo.12. L. P. R. F 13 28 2.3 47 0.2 Excellent 2 mo.13. D. D. F 8 8.6**** - 13**** - Excellent 2 mo.14. K. E. N. M 10 wvk. 21 2.6 36 0.2 Excellent 2 mo.15. A. R. D. F 25 30 7.6 41 - - Lost to follow up16. M. L. D. M 6 33 - - - - Lost to follow up17. M. L. V. M 9 15 18 - - - Lost to follow up18. W. XV. S. M 65 30 6.8 - - Operative

death

* In this Table and in those to follow the ternm Age refers to the patient's age at time of splenectomy.** In this Table and in those to follow the term after splenectomy refers to the most recent laboratory data available.

* Hematocrit: 39%70 two months after splenectomy (see text).** Ilemoglobin Gni. Ct,

In the patients classified as having granulo-matous diseases, splenectomy was performed be-cause of the association of splenomegaly withevidence of hypersplenism. Often a definitive diag-nosis was not disclosed by the preoperative diag-nostic studies.

Individuals with lymphomatous diseases whohad splenectomy did not have the correct diag-nosis established before operation. The indicationsfor operation were the presence of an enlargedspleen associated with evidence of hypersplenismor gastro-intestinal bleeding, or the presence of a

left upper quadrant mass that was causing con-

siderable distress. Essentially the same considera-tions led to splenectomy in the group of miscel-laneous hematologic conditions.

Results of splenectomy performed on 94patients for hematologic conditions are pre-

sented in Tables 1 through 9, and in thetext that follows.

Hereditary Spherocytosis. In this group

of 19 patients, 15 have been observed longenough to justify classification as excellentresults subsequent to splenectomy. Thirteenhave maintained normal lhematocrit or hle-mo1globinl levels, despite the contitiled pro-

duction of abnormal erythrocytes. Onepatient (Case 7) showed an increase in

hematocrit from 32, prior to splenectomy,to 39, two months later. She subsequentlyhas developed a blood loss anemia asso-

ciated with menorrhagia. The only opera-

tive death occurred in the oldest patient(Case 18), a 65-year-old man, who was

a very poor-risk patient, and who expired24 hours after operation due to intra-abdominal bleeding. Three patients havebeen lost to follow-up.

Hemolytic Anemia Other than Heredi-tary Spherocytosis. The cases in this group

are divided into autoimmune and non-

autoimmune hemolytic anemias, predomi-nately on the basis of the Coombs test.Both subdivisions include a variety ofunderlying disorders which were responsi-ble for, or related to, the hemolytic anemia.The nature of the primary disease was an

important factor in determining the post-operative result. Of the five patients withatutoimmtine hemolytic anemias, tvo (Case19, 20) hlaving been followed for tlhree and

eiglht years, respectively, atre classified itsgreatly improved because of normal hema-

tocrit findings, although they continue to

696

s* *

_ 0-. 0.0 00

U>

>Cs

C 0 0

eso Ln eq oo oo

)

0 0

- SV*-* *- S0. 0. Co

fi

tr) o ~ "0I 0sIo

I- et i

o0 I 00 n O 000 - i0 e O

- \0do-4 -0 1-)\ \0-f-NI ONf- 0if)O N u1

o IT) - - C-I 0 ON C-I t- f I ) - en 0f--i c- CN f-I - Cf cfN - fN f-- r r f

U L0 0 "0 "0

> > >0 0 0E-_o Eo 0.0.

.b5oE*t *= ._-n U) ;D >D

cl,

I--,C

f"i f-i f-4 i -

C~ - e l

r, I C- C1 r'l 0Z

0 "0:> 0.:> >:> Q Q:00>

._ ._ *- *_.

o

Qi .

Q3;

If- o 0 e .4 t-CIA fI ("I e e) e e

697

0o

0

SU

S

Cd

- 0*0

*4 cd m 0. m

._t

Se =

ASo.dZ 2b

cfU) U2CflU)..........""0; ,<Q

0 r

Volume 159Number 5 SPLENECTOMY

0

0

>1 = b

0

r.

< Oi

n II I IqIo oI0 ~~00C.,)

C-,

>1

0

(10

0

6.1cd .. QE b,(V

Xi

10

dI

C14 \0 C.l tn 0 "t in -4 0 Cs "o C., m4 t- IC Ul) \0 M tn -4 1- M -4 M C14

4 .. ;;.4 .. .. .. ;.-- .. ;.- .. ..

mI.-I

1.141-1 tn(i -"

;4 C/)- ,1.(i ,4 .;

F4 0-

SANDUSKY, LEAVELL AND BENJAMIN Annals of SurgeryAMay 1964

TABLE 3. Thrombocytophlenic Purpura

Platelets per cu. mm.

Response to Before AfterSteroids Pre- Splen- Splen-

Case Patient Sex Age operatively ectomy ectomy Result Follow up

34. H. F. F 24 No therapy 30,000 200,009 Excellent 15 yr.35. M. F. F 29 No therapy 5,000 475,000 Excellent 8 yr.36. L. H. F 56 Unchanged 28,000 124,000 Excellent 8 yr.37. B. H. M 17 No therapy 15,000 594,000 Excellent 5 yr.38. S. L. B. MI 8 Unchanged 6,000 350,000 Excellent 4 yr.39. A. S. M 9 Unchanged 10,000 290,000 Excellent 1 yr. 4 mo.40. P. C. F 34 Unchanged 15,000 230,000 Excellent 1 yr. 3 mo.41. E. H. F 42 Slightly improved 23,000 500,000 Excellent I yr.42. L. A. H. M 41 Slightly improved 12,000 251,000 Excellent I yr.43. N. L. A. F 16 Unchanged 8,000 - Excellent 10 yr.44. R. W. B. M 8 Transient im- 18,000 - Excellent 8 yr.

provement45. H. H. M 43 No therapy * 123,000 Greatly improved 3 yr.46. R. H. M 35 Unchanged 20,000 40,000 Moderately improved 3i yr.47. J. H. F 10 Unchanged 20,000 48,000 Moderately improved 2i yr.48. E. S. H. F 62 No therapy 7,000 53,000 Moderately improved 2 yr. 4 mo.49. S. S. M 44 Unchanged 48,000 55,000 Unchanged** 11 yr.50. G. T. E. M 36 Moderately im- 20,000 84,000 Unchanged"* 31 yr.

proved51. J. F. M. F 33 Slightly improved 36,000 520,000 Lost to follow up 8 mo.

* Studies done but records of laboratory data lost.** On steroids.

receive steroids. The remaining three pa-

tients died between six weeks and threemonths, respectively, following operation.Case 22 had an underlying reticulum cellsarcoma which probably contributed to thegram-negative septicemia, resulting in hisdeath one month after operation. Case 23had an associated thrombocytopenia witha platelet count of 37,000, and had beensuspected of having thrombotic thrombo-cytopenic purpura; he died three monthsafter operation, but autopsy findings didnot confirm this diagnosis. Case 21, whoseunderlying disorder was chronic lymphaticleukemia, died six weeks following opera-

tion of an intra-abdominal staphylococcalabscess and pneumonia. This was the onlysurgical complication in these groups.The remainder of the patients with he-

molytic anemias (the non-autoimmunegroup) showed somewhat better results.Two (Case 24, 25) are classified as excel-lent results, having showed no clinical or

laboratory evidence of hemolysis during theeight months and two years of post sple-nectomy observation. Four are classified as

moderately improved, inasmuch as the rateof hemolysis has decreased; two patients

are unchanged, and one patient (Case 33)died three months postoperatively of theprimary condition, Hodgkin's disease. Case26 had associated ulcerations of the legsand osteomyelitis. Case 29 had associatedthrombocytopenia. Case 30 has associatedrheumatoid arthritis and requires steroidsto control the hemolytic anemia. A young

girl (Case 31), diagnosed as having con-

genital non-spherocytic anemia, underwenta splenectomy in another hospital; shefailed to improve and was subsequentlyoperated upon by us at which time an

accessory spleen, 1.5 x 1.5 x 0.8 cm. insize, and a calculus gallbladder were re-

moved. Although her strength is sustained,she remains anemic and icteric and is clas-sified as unimproved. Case 32 has an un-

diagnosed disorder with cyclic episodes ofhemolysis, and Case 33 had underlyingHodgkin's disease.

Idiopathic Thrombocytopenic Purpura.Nine of the 18 patients with idiopathicthrombocytopenic purpura have shown no

clinical or laboratory evidence of diseasefollowing splenectomy and are classified as

excellent results. Of these, five have beenfollowed for four years or more, and four

698

SPLENECTOMY

have been followed for at least one year.

Two patients (Case 43, 44) are classifiedsimilarly, since they have been free of any

bleeding tendency and without therapy forat least eight years, although platelet countsare not available. Four patients, havingbeen followed for two to three years, are

classified improved on the basis of de-creased bleeding tendency and a higherplatelet count. Two patients are unchangedand require steroids for control. One pa-

tient was lost to follow up after eightmonths, at which time his platelet countwas normal. In two patients classified as

excellent result (Case 35) and improved(Case 45), respectively, there is uncer-

tainty concerning the part splenectomyplayed in the improvement, because remis-sion did not occur in these cases until sixmonths and two years, respectively, afteroperation. The only complications encoun-

tered in this group were two instances ofpostoperative thrombophlebitis (Case 36,45).

Myeloproliferative Disorders. Of theseven patients with myeloproliferative dis-orders, there were but two which showedeven moderate improvement after splenec-tomy. Two died within two weeks of opera-

tion, two died approximately two years

after splenectomy of their primary dis-order, and one has been lost to follow up.

In four cases, the definitive diagnosis was

not established prior to the time of sple-nectomy.Case 52, with known polycythemia vera,

developed persistent gastro-intestinal bleed-ing; cavernous transformation of the splenicvein was demonstarted by splenoporto-graphy; therefore splenectomy was per-

formed. There has been no further bleed-ing. Case 53, with known polycythemiavera and myelofibrosis, required splenec-tomy because of a hemolytic process. Threedays following operation she developedpyelonephritis, complicated by E. coli sep-ticemia; she responded favorably to anti-biotic therapy. Splenectomy has alteredfavorably the hemolysis. Case 54 presented

with hepatosplenomegaly and mild pan-

cytopenia. He was unimproved by splenec-tomy, and bone marrow examination ninemonths postoperatively revealed changesof myelofibrosis. Case 55 presented withgastro-intestinal bleeding and was found tohave carcinoma of the stomach, spleno-megaly, and the hematological changes ofchronic myelogenous leukemia. Case 56presented with splenomegaly and pancyto-penia and had 11 per cent myelocytes inthe peripheral blood. Following operation30 per cent blast forms appeared in theperipheral blood and he expired two weekspostoperatively with pneumonia. Case 57was an undiagnosed problem of spleno-megaly and pancytopenia; she died 16 dayspostoperatively, having developed a staphy-lococcal wound abscess and pneumonia.Autopsy findings revealed changes com-

patable with myeloid metaplasia. Case 58had previously known polycythemia vera

and myelofibrosis, and developed anemiaand persistent left upper quadrant pain re-

quiring splenectomy. She has been lost tofollow up.

Aplastic Anemia. All eight patients inthis group presented with pancytopenia.Three patients (Cases 27, 59 and 62) hadassociated hemolytic components, whichconstituted the indication for splenectomy.In this group of eight there was one excel-lent result (Case 59), a patient who hasmaintained normal blood counts for fouryears and whose marrow has reverted fromhypocellular to normocellular. One patientwas unchanged, one died two weeks fol-lowing operation from the primary disease;two went on to develop the changes ofacute myelogenous leukemia. One (Case27) presented with associated pancytopeniaand, following splenectomy, the leukocyteand platelet counts returned to normal andthe rate of hemolysis decreased. This im-provement was maintained for six years

only to be followed by worsening of anemiaand death. The primary disease remainsobscure. One case was lost to follow up.Case 89 developed hemolytic anemia dur-

Volume 159Number 5 699

SANDUSKY, LEAVELL AND BENJAMIN Annals of SurgeryMlay 1964

TABLE 4.

Before Splenectomy

Indication Hemato- WBC Plate- Myelo-for Opera- crit per lets per cytes

Case Patient Sex Age tion Final Diagnosis Cu. mm. cu. mm. Marrow %

52. S. D. M 46 GI bleeding Polycytlhemia 27 11,000 28,000 Normoblastic 0vera hyperplasia

53. V. H. F 59 Hemolytic Polycythemia 28 7,400 500,000 Hypocellu- 4process vera with lar, non-

myelofibrosis diagnostic

54. D. R. F. M 56 Spleno- Myelofibrosis 35 3,900 92,000 Hypocellu- 0megaly lar, non-

diagnostic

55. E. R. D. M 68 Spleno- Chronic myelo- 30 22,000 - * 6megaly, genous leu-gastric kemia, &mass gastric car-

cinoma

56. J. W. K. M 40 Anemia, Chronic myelo- 27 12,000 86,000 Hypocellth- 11spleno- genous leu- lar, non-megaly kemia &/or diagnostic

myeloid meta-plasia

57. R. S. F 63 Pancyto- Chronic myelo- 32 3,100 33,000 Normoblastic, 0penia genous leu- nondiag-

kemia &/or nosticmyeloid meta-plasia

58. W. M. F 69 Sympto- Polycythemia 28 11,000 293,000 Normoblastic, 1matic vera with myeloid hy-spleno- myelofibrosis perplasiamegaly

* Probable chronic myelogenous leukemia or leukemoid reaction.

ing the course of his congenital hypoplasticanemia; he died five months postoperativelyof hepatic hemosiderosis. Cases 27, 59 and62 have been reported in detail by Moblerand Leavell.12Congestive Splenomegaly with Hyper-

splenism. Six patients with Laennec's cir-rhosis of the liver and congestive spleno-megaly also exhibited some degree ofhypersplenism and underwent splenectomyfor this reason. Two of these six patientswere followed and are classified as excel-lent results, in that their blood counts have

returned toward normal; two have beenlost to follow uip; and two died followingoperation. One of these (Case 67) had

Laennec's cirrhlosis of the liver, portal hy-

pertension, esophageal varices ad hIyper-splenism. A portacaval shunt was consid-ered to be the procedure of choice, butbecause of a clotting defect splenectomy

was performed initially. Following thisthere was a significant rise in platelets, butthere was delayed clotting and poor clotretraction. Three weeks later a side-to-sideportacaval shunt was done, but unfortu-nately the patient died on the eighth post-operative day when a retroperitoneal hema-

toma ruptured into the left thorax. Theother patient (Case 68) died in the post-operative period of uncontrolled bleeding.He was a 55-year-old man with Laennec'scirrhosis of the liver, ascites, anemia, leuko-penia, thrombocytopenia and splenomegaly.He was not considered a good candidatefor portacaval shlint, hNt it was thoughtthat splenectovwoldyl)vd l)elbeficial incorrectinlg thle Iypersplenisnii and ill reduc-

inig tile portal bloodlHow. tIhiis w,as per-

formed, but in the postoperative period lhedeveloped uncontrollable hemorrhage and,in spite of re-operation (at which time no

700

Volume 159Number 5

M1Wyeloproliferative Disorders

SPLENECTOMY

After SplenectomyWeight

Hemato- WBC Plate- Myelo- ofcrit per lets per cvtes Spleen% cli. mm. Cu. mm. Marrow 70 Gm. Result Follow up

53 13,000 1,000,000 - 0 1170 Improved, no 3 yr.GI bleeding

31 18,000 500,000 Normoblastic 4 1975 Moderate im- Death 2 yr. afterhyperplasia, provement operationmyeloid hy-perplasia

18 21,000 - Myelofibrosis 8 1020 Death 2 yr. afteroperation

- - - 12 637 Death 18 mo.after operation

- - -- 2157 Died 2 wk. post-operativelyblastic crisis& pneumonia

- - - - - 1782 Postoperativedeath; woindabscess &pneumonia

- - - - - 870 Lost to follow up

specific bleeding point was determined)and transfusion with massive quantities ofblood, he died.Granulomatous Diseases. Of the five pa-

tients with granulomatous disease, threeare classified as moderately improved on

the basis of relief of symptoms or improvedblood counts, or both. One patient was un-

changed, and one patient (Case 74) diedfive days after operation of disseminatedhistoplasmosis. Three patients were foundto have nonspecific granuloma in theirspleens resembling sarcoidosis, but have

not developed further evidence of the dis-ease; one patient does have sarcoidosis.Lymphomatous Diseases. Iii this grotip

of seven patienits, there were fourl WvittlHodgkin's disease, two xvitli lyinplosar-coma and one with chronic lymphocyticleukemia. In all but two cases the correctdiagnosis was not establisbed until after

operation. The results in this group are

uniformly poor, except for one patient(Case 76) with lymphosarcoma, who hasbeen followed for one and a half years

since splenectomy, and who has shown im-proved blood counts. Another (Case 77),with chronic lymphocytic leukemia, has

been followed for three years withoutchange in blood counts. The remaining pa-

tients died of the primary disease withinone month to two years following operation.

Miscellaneous Hematological Condi-tions. The majority of the patients in thisgroup were undiagnosed at the time ofsplenectomy and the operation was per-formed b)ecdalse of hN7persplenisiii. Ulnfor-tullately, tile tw!o patienits withl lepato-splenomegaly anid pancytopeniia (Case 90,91) classified unsatisfactorily as idliopathiichypersplenism were lost to follow up. Hadthey been observed for a longer period, it is

701

SANDUSKY, LEAVELL AND BENJAMIN Annals of SurgeryMay 1964

'CS40 L0 Q

o _cd

0

on = E

*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

-b ano .0 .< Ca >.., > 00- ;'30. le 0 - Z

o; O '0 .

0 \0 00

O

0 00

U)~~~~~~~~~~~~~~~m C)s

o~ o o o

o V 0 oo

O r~~~~~~~~~~~-

o o -

o 0 0

o) 0 0 0

_ XooC)

CC'S N 00

Cl) C N-

_;- r,CZ

la at :O.R .~ 0 0

0

8

0

8Cl)

CS

CCC

t-

8lr

t-

m

Ce4

0

ul

0

0

It~

1°~~~~~IO cn

O, '~ :-

-~ 0

0

0

_ 0I 8_

0N ~

a E ad E

0CS '-.C a.) a...oCS

\0~~~~~~~~~~~( CZ0 =C0 u Z r N

5o S:

*~ SC * -C C 0 0*

co d co _dmSoS !) z

0000 00 C07

00~~ ~ ~~~'

ii>

CsCad-CZ CSZ

'0 0'

0 0 0~

zzz

oJoeat

_

702

3C-

0Ia4

~cV

>1

0

nC,)-

0

C,)

0

an

CC)

I)ItzS

'CC~4

H

0

Cd

CS

0

CS

_Eicd

o..

CS

0

CS

CS

CZI

0

~CC

CSQ.

SPLENECTOMY

possible that the underlying primary diseasemight have been discovered. An additionalpatient (Case 83) might be similarly classi-fied, although his bone marrow was sugges-

tive of reticulum cell sarcoma. Case 84 hadsplenectomy for anemia and thrombocyto-penia and was later shown to have Hemo-globin C disease; he died six years later ofcardiac failure unrelated to his hemato-logic disorder. Case 85 had proven Letterer-Siwe's disease and underwent splenectomybecause of a massive spleen which inter-fered with normal movement; there was as-

sociated anemia and thrombocytopenia.Case 86 developed a hemolytic componentin the course of an unusual anemia which,since splenectomy, has been shown to bea familial iron loading disorder; he hassubsequently died of a cause unrelated tothis disorder. Case 87, with massive spleno-megaly and thrombocytopenia, was shownto have Gaucher's disease on examinationof the spleen. Case 88 underwent splenec-tomy in hope of improving his idiopathicerythrocytic hypoplasia. Case 92 had mi-gratory polyarthralgias for seven years andskin ulcerations for six months prior tosplenectomy. The spleen revealed patho-logical changes of periarteritis nodosa. Shehas been on steroids since splenectomy andhas maintained normal blood counts. Case93, an eight-year-old Negro girl with sicklecell anemia, splenomegaly, hemolysis andpancytopenia, has been reported in detailby Shotton, Crockett and Leavell.13 Fol-lowing splenectomy there was marked ob-jective and subjective improvement as de-termined by a careful observation over a

ten-year period.Mortality Following Splenectomy. Each

death occurring within the first six weeksafter splenectomy has been discussed underthe appropriate grouping. Among the 37patients with hereditary spherocytosis andidiopathic thrombocytopenic purpura therewas one death, giving an operative mor-

tality rate of 2.7 per cent. Among the re-

maining 57 patients, four died within six

i0

00i

IC,.

C,)

ICA

::

-

0

>

0

CI)

Q

0

4)

6

.S~

CC-

C.)

m

E

IL=l O

EL.

Xx

;14

C)

z0

'-ZCZu1-1z

0 ~

-o

O4

Fo ¢:4-

> >

Z

d C_

00

00

SrI

Cl

S

o .-

Ce0

o0

0 .

o CI)

0-. if*)

O

'C

0 c

8) 8 8 8 8e le r .4

e) dq O c C

ll 0-I Cli C)0-

ot 0t 0I 00

CD .EnoJ

as0

.;

£

U)

._L

I1).

C,)u

0

C.)

Ce

0'4-j

V)* CVl*

O U0O

0I

;L - o

I)0-S.

vz S.

o

),

6-o

0

4-0

u

1')

I')

uz .EnU

0

.J,=

ZI..(L-

Ce

0

uz

C')

cd ._

o>,

;o

~c \0

Volume 159Number 5 703

6ZCo;.- 0-

SAND)USKY, LEAVELL A

a

0C14 Li.~~

I

4

._

C)

oE oE o.E YCS 0 C0 C

C0E )0 0c

80CN

o 0

o o-

It m

%0 s- of '-U

CSd CS 'C C

CS CS Cd Cd cn cd

5) bO cB BC0d 0 e0 0 .001. E S. °. E S.S

oo " bO Ch HM bi

U 0d 0 U 0S . C .C00 0 0% '-*'~~E

C If) CU)

- 4' - r

.Cl)

cU0

0z

od

06H

SS

C/)bo

Cd

%.5

0

S

u.z3 E

C)c

0

Cl)

E0

CSo

0E-0

0z

0_'a000otOC._-

C)

C)Cl

0oCS

C)o

Q

NI) BENJAMIN

-CS'

0C)

_ea

CS0E

w >E

4 .1C)0

c^

o0 E

e0

4

8

ul:

80\lf-^

ICSIb00C'

Annals of SurgeryMay 1964

C)

CSC o oo-

CSrCSY 4Y o

- 00

.so.0 o CS>- -l 1

°q Id

0 o

*

6 WstU) ein -

tr en oo

C- o -_ef) _-

CC

C C

E E CS

u

If)

Ur)

8

lu:'

0 U0~~E En

CSE

>~~~~>Col E e °; cQo ~ c @ ;

:Y cd 1e

>,CS c, CSCS Cd

- 8 0 0°- o E-Y

If) e o eo c O s

m* so N e rsto

to o1i ooo

&.; i

u) tr

*_ C-l

00 0

0

E.0

Q)0

704

0

044

E0

L

Cl)1(n)v

¢.00

C.)

t3Co)

r

EH

S0

CLC)

0

4)14

'E

6

K..

CS

'S

CS g

0

CS

0

Un.

000

53-

0

C)C)W

0

C)CJ)CS

SPLENECTOMY

U) U)

C

0 r

00

1--z

c

0

0

4-i

U1)

7r >0

0

E l- .-

00

Ul)

0~8 8xo o

e 00 0) 14eC1 CIe 't

o o

C.) %0

-4 '0 ei 4 0%t U.'Cm C1 44 C..C 44 -4 e'

U1)

U)adcY

vS._

v._

u4(4 00

U)

U)

U)U)

a.00

C1)

00

CC

eCd

C0 C.)C

CCU

0u -ocu 0

0 0n U)

'0CdC

~~~ H ~~~~~ 00~~C

0%

.

o;

2E\0

I

r0

oo

If)

00

- 0 00to) -4 It

0% 0%N

Volume 159Number 5 705

0sZ

C14U.-0

U-4

o)

C)rllo

00'

0

8~o00C1)

C.,

0

0

0

.)cr.Iz

--H

c

3=

Li

E E

C C

cnB

0

C)

._

0n

o

1.-

*0_

C)

C)

UU

0%

0

olc(1f00l

m

00

0*

e 0m m

0

0U)O._

E .U az

Cd O E

o) , cd0 00

C. 0C.)

P H

CU

CU'

0.

a.

CUv

U)

-

4C

EcU

C.)

.

-o

0LCeO

U)

* * ** *

*

00

706 SANDUSKY, LEAVELL AND BENJAMIN

weeks after splenectomy of causes judgeddue to the primary disorder; while therewere five who died in the postoperativeperiod as the result of splenectomy; thus,in this grouip the surgical mortality is 8.7per cent.

Discussion

It is generally agreed that splenectomy isthe treatment of choice in hereditaryspherocytosis. Although spherocytes andincreased fragility remain, anemia andjaundice almost invariably are relieved.Unless there is a contra-indication to opera-tion, an established diagnosis of hereditaryspherocytosis is sufficient reason for sple-nectomy. There are, however, those whowould avoid, if possible, performing sple-nectomy during infancy, on the basis ofa presumed increase in susceptibility ofthe splenectomized infant to infection.9

In those hemolytic anemias other thanhereditary spherocytosis the response tosplenectomy is unpredictable and thechance of a lasting remission is no betterthan fifty per cent. For this reason, it isa general practice to employ splenectomyin this group of hemolytic anemias onlywhen there has been failure of responseto treatment with the adrenocorticoster-oids 10 or when there is selective splenicsequestration of Cr51 tagged erythrocytes.

Idiopathic Thormocytopenic Purpura.Subsequent to 1916, when the medicalstudent, Kaznelson, persuaded his professorof surgery to perform the first splenectomyfor thrombocytopenia purpura,1 this opera-tion has occupied a significant place in themanagement of this disorder. Modern meth-ods of platelet transfusion and the introduc-tion of the adrenocorticosteroids have allbut eliminated the requirement for emer-gency splenectomy. Moreover, the employ-ment of the latter agents has altered theselection of cases for splenectomy.Although the unpredictable course of the

disorder makes it difficult to evaluate vari-ous forms of treatment, it is interesting to

Annals of SurgeryMay 1964

note the effectiveness of splenectomy aloneas reported by Doan, Bouroncle and Wise-man.4 In 1960 these authors were able toreport a good response without recurrencein 85 per cent of 167 patients with idio-pathic thrombocvtopenic purptura treatedby splenectomy alone. On the basis of anextensive experience over a 28-year period,these authors conclude that splenectomyremains the treatment of choice for throm-bocytopenic purpura, both primary andsecondary, except in selected patients. Cor-ticosteroids are reserved for children andyoung adults with thrombocytopenic pur-pura secondary to transitory infections, forthose in whom operation is contraindicated,and for splenectomy failures.At the University of Virginia we have

approached the management of idiopathicthrombocytopenic purpura somewhat dif-ferently. When first seen these patients arestarted on Prednisone, 40 to 60 mg. daily';those that do not respond to this therapywithin six weeks, and those who have re-lapsed after the drug is discontinued, aretreated by splenectomy. It is to be notedthat on this program 60 per cent of thosefollowed have had complete remission. Hadsplenectomy been resorted to initially, tothe exclusion of other forms of therapy,undoubtedly it would have been used insome patients in whom either spontaneousor drug-induced remission might have oc-curred. Moreover, by employing cortico-steroid therapy initially, it is possible thatthe more refractory cases remain and theresults from splenectomy may appearpoorer. In this connectin, it is of interestthat when Leavell and Thorup 11 reviewedthe experience at the University of Virginia,it was apparent that failure of hormonaltherapy did not prejudice the results ofsplenectomy, nor did the failure to respondto splenectomy affect the future responseto ACTH or adrenocorticosteroids.

I In the series there are a few exceptions inwhich Prednisone therapy varied from this sched-ule in some detail.

SPLENECTONiY

Myeloproliferative Disorders. Utntil re-

cently the presence of myeloproliferativedisease of the spleen was considered almostuniversallyv a contraindication to splenec-tomv. It was reasoned that in this disorderextramedullarv blood formation occurredin the spleen and liver as a compensatoryreaction, and that removal of the spleeneliminated an important site of blood for-mation. It was believed that splenectomyunder these circumstances would be fol-lowved bv harmful or even fatal conse-

quences. Nonetheless, occasionally sple-nectomy had been done in mveloprolifera-tive disease, and bv 1953 Green, Conlev,Ashburn and Peters(; wvere able to collect29 reported cases, five of w\hich were theirown. Thev use the term agnogenic mlyeloidmetaplasia for this condition, and conclude

that in no case did splenectomy have a

deleterious effect; one patient showed he-matologic improvement. Moreover, theirreview of the literature failed to confirmthe belief that splenectomy leads to disas-trous effects because of removal of a largeblood-forming area; most of the reportedcases showed no improvement, but therewere occasionally patients, particularlythose with severe thrombocytopenic or he-

molvtic anemia, who did derive benefitfrom splenectomy. It has not been our prac-

tice to employ splenectomy in those pa-

tients w\Nho could be managed adequatelyby other forms of treatment, such as corti-costeroids, testosterone, busulfan or trans-fusions. Howvever, when repeated hemor-

rhages, frequent infections, recurrent ane-

mia, or the discomfort of splenomegaly ledto invalidism, or when there is selectivesplenic sequestration of Cr51 tagged redblood cells, splenectomy has been em-

ployed.Aplastic Anemia. There is considerable

diversity of opinion relative to the employ-ment and usefulness of splenectomy inaplastic or hypoplastic anemias. The ra-

tonale for its use lies in the hope of stop-ping the production of auto-antibodies,

imiiproving:, 1bon1e imarrow7 ftunietioni itand pro-

longing the survival of red cells wvhen a

hemolytic element is present."' Moore-'believes it should be performed only ifthere is evidence of a hemolytic componentto the anemia, or a response to cortico-steroid therapy. Others wvould employsplenectomy when other forms of therapyhave failed and still others believe it isof no value.Our indications for splenectomy in aplas-

tic anemia, in the moderately ill patient,has been splenomegaly with associatedhvperhemolysis, and in the seriously illpatient, as a means of last resort when theneed for transfusion has been almost con-

tinuous, or in either group when survivalstudies indicate increased destruction in thespleen.During the period 1930 to 1956, Heaton,

Crosby and Cohen found 35 documentedcase histories of splenectomy for hypo-

plastic or aplastic anemia. Of these, 20appeared to have been improved by, sple-nectomy. To the reported cases these au-

thors add 12 of their own; six were appar-

ently benefited by operation; three were

not; and three died.MIohler and Leavell 12 in an analysis of

50 cases of aplastic anemia seen at theUniversity of Virginia Medical Center be-tween 1933 and 1956, reported the use ofsplenectomy in five; there was one remis-sion, two improvements and two failures.These five cases are not in those collectedfrom the literature by Heaton, Crosby andCohen. Scott, Cartwright and Wintrobe 13

have employed splenectomy in 15 patientswith aplastic anemia, five of whom showedsubstantial reduction in transfusion require-ment following operation. More recentlyDuckett5 has reported the results of sple-nectomv in 19 cases of aplastic and hypo-

plastic anemia. There were two excellentand four good results.Hypersplenism is the term used to de-

note a syndrome characterized by: 1) sple-nomegaly; 2) anemia, leukopenia, throm-

X'oltime 159NuLnmber 5 707

SANDUSKY, LEAVELL AND BENJAMIIN

bocytopenia, singly or in combination; 3)a marrow that has normal or increased cel-lularity; and 4) correction of the blood pic-ture by splenectomy.

This syndrome occurs in a variety of dis-orders that are associated with spleno-megaly. The most common of these may

be grouped as follows:

1. Congestive splenomegaly2. Chronic infection (especially tuberculosis,

brucellosis, funguis, syphilis, malaria, bacterial en-

docarditis)3. Diseases of the lymphoma grouip (especially

lymphosarcoma, giant follicle lymphomna, Hodg-kin's disease)

4. Other diseases of uncertain etiology (rheni-matoid arthritis, collagen diseases, sarcoidosis)

5. Splenomegaly of undetermined cause (?idiopathic)

In nearly every case the diagnosis can beestablished before splenectomy, or at leastat the time of the operation. However, inthe occasional patient the cause of thesplenomegaly remains obscure, even aftercareful histologic study of the spleen andother biopsy material. It is these patientswho are classified, somewhat unsatisfac-torily, as idiopathic or primary.At present the mechanism responsible for

the reduction of the circulating cellularelements in many patients with spleno-megaly is not understood. The actual se-

questration and destruction of the cells or

platelets in the spleen have been consid-ered the important factors by some authors.3The presence of a marrow inhibiting fac-tor, elaborated in the spleen and trans-ported in the plasma, also has been sug-gested.) The role of antibodies, particularlyleukocyte antibodies, has not been clari-fied despite considerable study; autoim-mune hemolytic anemia and iodiopathicthrombocytopenic purpura, syndromes in

which the role of antibodies has been more

firmly established, usually are not includedin hypersplenism.The decision regarding the advisability

of splenectomy in this syndrome can be

made only after careful consideration of theindividual patient. The first objective is therecognition of the disease responsible forthe splenomegaly by means of appropriatediagnostic procedures. In addition to a

study of the peripheral blood and bonemarrow, these may include radiographicstudy of the chest, esophagus and stomach,blood cultures, skin tests, agglutinationtests, liver function tests, serum electro-phoresis, and biopsy of the liver or avail-able lymph nodes. Recognition of theprimary disease is important because sple-nectomy relieves only one aspect of theunderlying disorder and rarely effects a

cure. If the underlying disease is one thatis amenable to more definitive treatment,

splenectomy may not be necessary becausethe hypersplenism syndrome may disappearafter such therapy. When the nature andcourse of the underlying disease are suchthat an early fatal outcome can be pre-

dicted, splenectomy is rarely a justifiableprocedure. If the nature of the underlyingprocess cannot be diagnosed, or if it isidentified as one for which no satisfactorytherapy is available, splenectomy is indi-cated if the presence of anemia, neutro-penia, or thrombocytopenia is disabling or

hazardous to the patient. In such circum-stances gratifying results occur often, btitnot invariably.

Susceptibility to Infection after Sple-nectomy. Recently there have appeared a

number of reports concerning increasedsusceptibility of the splenectomized indi-vidual to subsequent infection; there are

equally as many reports which refute this.Our observations in this regard have not

been as complete as they would have beenhad we followed each patient from thetime of splenectomy with this point inmind, however, from a retrospective stand-

point, the only serious post splenectomyinfections occurred in those individualswhose primary disorder was one predis-posing to infection.

708 Annals (f SturgeryMay 1964

olumne 159 SPLENECTOMY 709INumber 570Summary

We have presented the experience of auniversity medical center with splenectomyfor hematologic disorders. In some patientssplenectomy was performed for well-established indications, namely, hereditaryspherocytosis and idiopathic thrombocyto-penic purpura, with the expectation andthe achievement of excellent hematologicresults and a 2.7 per cent mortality. In otherpatients, those with hemolytic anemia,aplastic anemia, myeloproliferative disor-ders and secondary hypersplenism, it wasrecognized that, at best, the outcome wasproblematical. While the over-all results inthe latter groups are far from excellent, andthis includes surgical mortality of 8.7 percent, individual instances of significant im-provement have been most rewarding.

Bibliography1. Allen, J. G., in Harkins, H. N., C. A. Moyer,

J. E. Rhoads and J. G. Allen: Surgery: Prin-ciples and Practice. Philadelphia, J. B. Lip-pincott Company, 1961, 801 pp.

2. Dameshek, W.: Hypersplenism. Bull. N. Y.Acad. Med., 31:113, 1955.

3. Doan, C. A. and C. S. Wright: Primary Con-genital and Secondary Acquired Splenic Pan-hematopenia. Blood, 1:10, 1946.

4. Doan, C. A., B. A. Bouroncle and B. K.Wiseman: Idiopathic and Secondary Throm-

bocytopenic Purpura: Clinical Study andEvaluation of 381 Cases over a Period of28 Years. Ann. Int. Med., 53:861, 1960.

5. Duckett, J. W.: Splenectomy in Treatment ofSecondary Hypersplenism. Ann. Surg., 157:737, 1963.

6. Green, T. W., C. L. Conley, L. L. Ashburn andH. R. Peters: Splenectomy for MyeloidMetaplasia of the Spleen. New Engl. J.Med., 248:211, 1953.

7. Heaton, L. D., W. H. Crosby and A. Cohen:Splenectomy in the Treatment of Hypo-plasia of the Bone Marrow with a Report ofTwelve Cases. Ann. Surg., 146:637, 1957.

8. Jacobson, L. O., C. V. Moore and W. H.Crosby: Panels in Therapy IV. The Thera-peutic Management of a Case PresentingSplenomegaly, Pancytopenia, and a Hypo-cellular Marrow. Blood, 10:753, 1955.

9. Leavell, B. S. and 0. A. Thorup, Jr.: Funda-mentals of Clinical Hematology. Philadel-phia, W. B. Saunders Company, 1960, 161PP.

10. Ibid., p. 196.11. Ibid., p. 267.12. Mohler, D. N. and B. S. Leavell: Aplastic

Anaemia: An Analysis of 50 Cases. Ann.Int. Med., 49:326, 1958.

13. Scott, J. L., G. E. Cartwright and M. M.Wintrobe: Acquired Aplastic Anemia: anAnalysis of Thirty-nine Cases and Review ofthe Pertinent Literature. Medicine, 38:119,1959.

14. Shotton, D., C. L. Crockett, Jr. and B. S.Leavell: Splenectomy in Sickle Cell Anae-mia: Report of a Case and Review of theLiterature. Blood, 6:365, 1951.

DISCUSSION

DR. S. W. MOORE (New York): There areseveral points that I would like to mention. One,we have heard about finding the ruptured spleenwhen doing a thoracotomy. I woud like to call toyour attention that sometimes it is very safe todo a thoracotomy and take out the spleen. Onepatient had two previous attempts at splenectomy,both unsuccessful; we opened the chest, spentfour hours, and finally were able to remove hisspleen; so it is very useful to have an additionalapproach.

Another patient with collagen disease had beenon cortisone, developed fluid in the chest, andcame in with all the symptoms and signs of

blood and hemorrhage in the abdomen. We finallydecided to explore this patient and to the chagrinof everyone, found a ruptured spleen as the re-sult of aspiration of the chest.

Spontaneous rupture of the spleen, I think,is very doubtful, and if we look further, we willusually find that there is a history of trauma.

Abdominal exploration is very necessary. Attimes it does cause trouble. We have had a recentdeath following abdominal exploration in whichthe spleen was ruptured and not recognized, andI must admit that I have taken out the spleentwice fairly recently, because of a small tearwhile doing an exploration and trying to mobilizeviscera.