Observations on phlebitis

OBSERVATIONS ON PHLEBITIS” EDWARD ALLEN EDWARDS, M.D. BOSTON, MASS. M UCH has been said on the subject of thrombosis and phlebitis, yet for the most part the etiology is still obscure and the treatment. diversified and unsatisfactory. It is necessary to be skeptical of some of our so-called “fundamentals” until more data are secured. The observations here described are of diverse nature, and are not presented in any effort to arrive at conclusive deductions, but rather in the hope that they will arouse a critical viewpoint and lead to further clarification of our problem. THE R6LE OF INFECTION IN THROYBOPHLEBITIS It may be of interest to examine the evidence in regard to the r6le of infection in the production of thrombophlebitis. In this respect phlebitis can quite easily be divided into two major groups, suppurative and nonsuppurative. Suppurative phlebitis is seen in the immediate vicinity of a purulent focus, such as mastoiditis, septic endometritis, or carbuncle. It takes no special argument to establish the presence of infection in this form. The involved vein may not show any more clinical signs of inflammation than a bland phlebitis. The clot within it, however, is infected, and it undergoes purulent softening with a dissemination of bacteria and portions of clot throughout the blood stream. Wherever large enough masses of clot or bacteria block branches of the pulmonary artery, infarcts result, which go on to pulmonary abscess formation. Microscopic examination of the primary (autochthonous) portion of the thrombus shows the vein wall infiltra,ted with an exudate rich in polymorphonuclear leucocytes (Fig. 1). The bacteria are invariably found in this part of the vein and its contained clot. It may be noted that the propagated portion of the clot may be free of bacteria and the vein wall far from the original focus may show an exudate poor in polymorphonuclear leucocytes. In the nonsuppurative category may be classed most of the cases of postoperative phlebitis, especially when it is not in the vicinity of the operative wound ; as well as phlebitis of varicose veins, and finally the so-called “marantic phlebitis” occurring in fevers and in terminal states. At first glance such phlebitis might also seem to be a septic *From the Departments of Surgery. The Boston City Hospital, and Tufts College Medical School. This study was aided by a grant from the Charlton Research Fund. Read before The American Heart Association, section for the study of the peripheral circulation, at Atlantic City, N. J., June 7, 193’7. 428 EDWARDS : OBSERVATIONS ON PHT.EBTTlS 42!, process. The involved vein is apt to be very hot and tender, and the swelling, augmented by phlebitic edema, may be considerable. Nevertheless the evidence for the presence of bacteria in this group is e011troversial and stands in considerable contrast t,o the calear-cut demonstration of bacteria in the septic variety. The bland nature of nonsuppurative phlebitis had seemed well With the advent of bactrrioestablished by the work of Virch0w.l logical examinations the picture was reversed. Ya.quez,’ in ZHYO. recorded the finding of micrococci in what had previously been calletl marantic thrombi, in patients dying of typhoid fever, tuberculosis, and even cancer. There are others who have searched for bacteria in these thromboses, and have also found them with fair frequcucy. Thus, Harris and Longcope found bacteria in thirty-four out of fortyfour cases of phlebitis, most of which were said to have bee11 pet*il,hera1 and of the so-called marantic type. More recently deTakats” has gone a step further. Obtaining positive cultures in a. goodly number of varicaose veins removed at operation, he has formulated the concept of “resting infection” in varicose veins. According to this idea not only is varicose phlebitis an infective process, but the bacteria causin g it may actually lie in varicose veitts in a semidormant condition for long periods of t,ime. All these facts have gradually converted man)- physicians to the belief that all thrombophlebitis represents an itlfection of veins. In the meantime there has been rather strong evidence that nonsuppurative phlebitis is generally a bland process, and is rather due to ptlysical factors such as slowing of the blood stream. AschoP is probabl? the most important contemporary proponent of this theory. This old theory gains modern support in the fact that several investigators have failed to find bacteria in most nonsuppnratire phlebitis. ‘I’husY Barker,” in a review of 166 cases of thrombophlebitis complicating infectious and systemic diseases, found that cdultures were occasionallypositive but usually sterile. Brown,’ in a review of 87 cases of postoperative phlebitis, stated that no bacetrria were foutitl it1 th(JSe that were cultured. The reported results then, ant1 rn~- 1)ersonat experiences, lead me to the caonclusion tbwt bac*tcaria are only sporatlicall) present in these cases. paraSO far as varicose phlebitis goes, my findings are somewhat doxical. While I have never been able to find bacteria. by stain OI culture of the clot or vein wall, I have nevertheless twice come npon gram-positive cocci in surgically excised saphenous veins, wlirre there had been no suspicion of their presence. These patients did not have phlebitis, the wounds did not suppurate, nor was there evidence of any acute infection going on elsewhere in the body. The histology of bland phlebitis differs also from that of the septic variety. There is usually very- little cellular infiltration, and most of 430 THE AMERICAN HEXRT JOURNAL the cells present are of the so-called “chronic inflammatory” or reparative type, that is, lymphocytes, plasma cells, and macrophages.” The infiltration may be located for the most part around the vasa vasorum in the adventitia. Polymorphonuclear leucocytes are uncommon and few in number. This holds true even in those cases where the phlebitis is attended by clinical evidence of acute inflammation with much heat, swelling, and tendernes’s (Fig. 2). There is a form of phlebitis in which it seemed more probable to find bacteria and acute inflammatory cells. I refer to a subacute phlebitis of small veins on the foot which occurs occasionally in patients with varicose veins, especially those which have followed a. deep phlebitis ( ’ ‘ ulcerating phlebitis ’ ’ ) .28 The special features of this phlebitis are the location below either malleolus, the immediate dark pig- mentation of the skin over the vein, and its immediate adhesion to it; and finally the almost irrevocable breaking down of this skin into a shallow, painful, and intractable ulcer. In two such instances I have performed a biopsy of the vein and skin in the pre-ulcerous stage. Bacteriological culture was made at the time of the biopsy, but no bacteria were obtained by either aerobic or anaerobic methods. Nor were bacteria seen in stained sections of the veins and skin. The histology of the tissue was surprisingly mild, an organizing thrombus was present, with very little cellular infiltration. What cells were present were chiefly lymphocytes and plasma cells. *The phlebitis of thromboangiitis obliterans stands out as an exception. EDWARDS : OBSERVATIONS ON 431 PHLEBITIS Evidence of the bland nature of most phlebitis is also to be fomd in its clinical course. Though frequently mentioned, suppuration of a thrombus in bland phlebitis such as that of T-aricose veins, is rare. It is possible that softcAnIndeed, I have never seen a single instance. ing of the clot, through autolysis. has oc*c*asioltwll,v been mistaken for suppuration. Here also it may be nottd that while pulmonary embolism is quite as common as in the septic form, the resultant infarcets do not give rise to pulmonary abscesses. Xoreover, varicose veins wli ic*h are phlebitic map be injected or ligated with safety. In 63 cases trttatc4 by ligat,ion and subsequent injection not a single instance of sepsis ot’ t tic, vein or septicemia has occurred. In the face of the evidence at hand the words of Welch,” spokrltl iI1 we are justified in assigning a far 1909. are still convincing ; “Whilst more prominent place to the agency of micro-organisllls and to prirnar? phlebitis in the etiology of thrombosis t,han, until recent years. has been customary since Virchow ‘s fundamental investigations, rr~lt attempts to refer all thromboses formerly called ‘marantic’ to tlLe direct invasion of micro-organisms, and to phlebitis, go beJ.ond tlrmonstrated facts.” SYMPATHETIC NERYE IRRITATIO’N A~‘~‘O%lPANYING PHLEIIITIS Certain phenomena occur in the course of phlrbitis. which csarl only be explained by a stimulation of the sympa.thrtic nerve fibers of the extremity. The mechanism is not yet known, but it is clear that whether the reactions are due to direct, or to reflex stimulation of the efferent fibers, irritation of a vein is a11 adequate stimulus for these reactions. The simplest, of these reactions involves the Trills themselves. 111 some individuals the puncturing of a vein in the forearm causes an immediate constriction of all the superficial veins of the limb, and it is common experience that a second successful pun&are may be impossib1.e for some minutes. This same phenomenon may be seen in the leg when an irritant solution is injected for the purpose of obliterating varicose veins. Snch a, constriction has been described in phlebitis by Ijeriche,22 and by Ducuing,g who believed it is responsible for much of the pain. There is additional sipnificancr iI1 this, however, as the contraction of the vein slows the ‘blood stream anti fac4itates the prop’ress of the thrombosis. Aside from venous constriction, we may consider a group of reactions which, while originating within the veins, spread to affect other The first. of these is the augmentation of the pilomotor structures. reflex. Ducuiny: JO has remarked that the reflex may be diminished 01 increased. In practice the diminution of this reflex is not easily made out, but its increase is sometimes strikingly apparent. Thus when both the normal and t.he diseased limbs are exposed in the patient’s 432 THE AMERICAN HEART JOURNAL room, the skin of the phlebitic limb may show a pronounced goose flesh appearance, while the normal limb shows it not at all or very slowly. Again, abnormal promptness and intensity of the reflex may be shown by scratching the limb with the back of the nails, when broad streaks of goose flesh appear. The second function of the sympathetic fibers which comes in evidence is that of local sweating. While ordinary generalized sweat- Fig. 3.-Localized sweating distal to somewhat tensely distended finer than shown. varices. The droplets were ing occurs when the body is generally warmed, the local secretion can be seen even when the body is quite cold. This is the same type as that present in some neurological disturbances, in hyperhidrosis of the hands or feet and in those cases of vascular disease characterized by an increase in sympathetic tone. Well recognized examples of the latter are Eaynaud’s disease and thrombo-angiitis obliterans. Localized sweating is frequent in the presence of vein irritation. It is best seen in cases of varicose veins that are large and tense or those EDWARDS : OBSERVATIONS ON 2:u PI-ILEBITIS that are developing rapidly, as in pregnancy? and also in phlebit,is of varicose veins. The sweat may only be a.pparent after the patient has been walking about or standing for some time. Tt appears as millufc or medium-sized droplet,s arranged distal to the varicps (Fig. :I). ‘I’lrt~ droplets may be so small, indeed, as to be apprp(,iatecl only on X~W,V close inspection or by feelin, n moisture 011 touching the leg. Thcw droplets may increase in size, or make their first iII)Dc;lEltt(‘t’, wht~tt tllcb Fig. 4. Fig. Fig. 5. Figs. 4 to 7.-Localized sweating lined by starch iodine. Figs. 4 and the Proximal area lies just below jection. Rig. 5 is a case of varicose 6. F’ig. 7. in varicose veins and varicose phlebitis, out7 show proximal and distal areas. In each case a zone of “phlebitis” induced by sclerosing inulceration. Fig. ti is of uncomplicated varicw vein is pumturecl by a needle. The sweating is perhaps most apparent distal to areas of phlebitis in t.he variwse veins, whether it be spoiltaneous or induced by a sclerosing solution. To render the secretion conspicuous I have employed the starch iodine t,ecJhnique of Victor Minor.ll A glance at the photographs of representative cases will show the fineness of the secretion and its localization into small areas (Figs. 4, 5, 6, 7). That the secretion is not necdrssarily an anpmenta- 434 THE AMERICAN HEART JOURNAL tion of the sweat in areas where it is normally high in amount, is shown by the fact that it is not confined to any one part of the leg or foot, but is called forth in any region distal to vein irritation. Clinically the most important result of the sympathetic irritation is arterial constriction. It is quite commonly known that while, in phlebitis, the limb may be hot directly over the inflamed vein, it may be cold distal to it. Such a lowering of temperature can best be accounted for by a constriction of the arterioles.” Examples of more severe involvement of the arteries have also been mentioned many times. Authors have oft.en spoken of the small pulses in phlebitis. This diminution has been confirmed by the oscillometer in several reported cases, and in some personally observed ones. Finally, phlebitis may occasion a complete spasm of the major artery of the limb. Twenty-odd such cases have been reported, the majority in the French literature, with others in the German and American.13-20 The events in the reported cases resemble those in a personally observed one. CASE REPORT Mr. 0. S., was seen on the Third Medical Service of the Boston City Hospital at the request of Dr. Cadis Phipps. This man, a forty-six-year-oId carpenter, entered the hospital on July 19, 1935, for treatment of a pulmonary embolism resulting from a deep phlebitis of the right leg. Within the next two weeks, the phlebitis ascended to involve the right femoral, then the right iliac vein. On August 6, 1935, a series of events took place which were carefully observed and recorded by Dr. J. C. Druker, the House-Officer on service. I quote from his notes: “August 6, 1935, the patient was seen at 9:00 A.M., and stated that he felt well except for some pain along the medial side of the left thigh. This pain in the left thigh was present for the first time. About 11:30 A.M. the patient was seen again and complained of more pain in this area and there was slight cyanosis of the left lower leg. The pulsation of the femoral artery was present and was of good volume. The patient also complained of some additional pain in the lower right chest (site of previous infarction). Examination of the chest showed no change from what had been heard previously; namely, diminished breath sounds at the extreme lower part of the right lung, posteriorly, and bronchial breathing which was distant above; and also a short terminal inspiratory friction rub anteriorly. “At about 3 :45 P.M. the patient complained that he was unable to feel any sensation in the left leg from the knee down and there was pain above this area. He At this time, the complained also that he was unable to move the foot or toes. leg from the hip down resembled in color an arm which has had a tourniquet applied in the tourniquet test for scarlet fever. The leg was cold, dry, and anesthetic up to the knee, while above, it was hyperesthetic. There were no pulsations made out in the femoral, popliteal, posterior tibial, or dorsalis pedis arteries. “At about 6 :00 P.M. hot water bottles were applied to the leg, and at 7 :30 P.M. the condition of the leg from the knee down had changed considerably. The patient stated that ten minutes after the heat was applied, sensation had returned in the lower leg and in about three-quarters of an hour he was able to move the foot --*AudieF has found in these cases a more rapid absorption rate in the McClureAldrich test. EDWARDS : OBSERVATIONS ON PHI2I3ITIS 4x5 and toes. The leg was pink in color, sensation had returned, and there were goocl pulsations in the femoral, dorsalis pedis, and post,erior tibia1 arteries. The femoral vein was palpable as a tender induration. Two cnlarg~l veins wwc SWII on thc~ Icfi flank which had definitely not been ritiihle hefore. ” ConlnbPnt.-This patient showed temporarily all the signs of an CCelusion of t,he femoral artery; namely, t,lie disappearance of the femoral pulse and the pulses below, the cold white skin, the anesthesis below and the hyperesthexia above, and the loss of muscle power. The diagnosis at the onset of this episode seemed surely to b(> either embolism or thrombosis of the femoral artery-, yet with the application of heat, all the signs wore off and the pulses returJ1rtl. Within a few hours the diagnosis of iliac and femoral phlebitis was definite. The limb subsequently went through the usual sequence that one finds in iliac phlebitis. There was edema from the buttock dower ; there was enlargement of the superficial veins of thigh and abdometr; and there was the tender induration of the femoral vein of the thigh. The edema slowly subsided during the patient’s stay in bed. Later, when the patient was up and about, the edema increased temporaril>* but soon became less again. By the latter part of September, 1935, there was but little edema and the surface veins of the thigh and abdomen were beginning to diminish in size. indicating that the clot, was being canalized. The leg showed not the slightest evidence of arterial impairment. All the pulses were of full volume; an arteriogram showed that there was a normal lumen in both the main and deep femoral arteries; and the oscillometric readings (Pachon) were quite normal. The actual reatlings were as follows: - -__- MEAX IN .____ Right thigh Left thigh Right leg Left leg ARTERIAL MM. OF 120 120 100 I on PKESSI’KE MERCI-RY 0S~‘lLLOMETRIc’ UNITS fi.5 fi 5 5 In cases of embolism or thrombosis of the femoral artery, 011emllst certainl~~ have rare good fortune to observe the spontaneous reappearance of all the pulses of the limb and to find that the objective measurements of the calibers of the vessels are normal and exactly equivalent to the opposite extremity. As a. matter of practice, suc*h an ontcome is but rarely observed, even after a successful embolectom~. One cannot escape the conclusion that this patient had no organic4 obstruction to the arteries at all, but that his arterial occlusion represented a temporary, complete spasm of the femoral artery, incident to the start of the phlebitis in the left iliac or femoral vein. The picture then closely simulates arterial embolism. The femoral artery is the most frequently involved, and the spasm is secondary to 436 THE AMERICAN HEART JOURNAL thrombophlebitis of the femoral or iliac vein. There is a sudden severe pain in the course of the artery, then in the entire leg. The femoral pulse is absent, the limb painful, cold, and paralyzed. After some few hours the attack may wear off spontaneously, or as has rarely happened, actual thrombosis ma,y occur. The arteriospasm may occur after the phlebitis is quite evident, but it usually happens so early that the diagnosis may be difficult to establish. Thus, while the telltale edema and cyanosis of phlebitis may be present, ordinarily the only hint of primary venous involvement is a slight cyanosis of the thigh above the ischemia of the foot and leg, and a8 slight dilatation of the surface veins of t,he thigh and abdomen. In some of the earlier reports it is likely that there were included instances of a true arteritis by extension from pelvic sepsis, or simultaneous phlebitis and arteritis of thrombo-angiitis obliterans type. In many, however, there is no doubt but that there was merely a spasm of the artery.” Several of the reported cases were operated on for the suspected embolism. In all these the artery was found contracted and pulseless but the pulse reappeared immediately after the artery was freed from its sheath. In other patients arteriograms were made and the arteries’ found to be patent. Some of the patients died from pulmonary embolism and autopsy revealed patent arteries. Finally, gangrene occasionally supervened, and examination of the amputat.ed extremities revealed normal arteriesiCHANCES IN THE VENOUS VALVE Apparently the fact that phlebitis damages the venous valve was first appreciated by Homans, 26 although he did not publish demonstrations of this damage. Recently, with Jesse Edwards, I have completed a detailed study of this effect. We made histological examinations of the valves in spontaneous phlebitis in the human and in experimental chemical phlebitis in the dog. This is report,ed elsewhere,2T but it is of interest here to state briefly our findings. When complete thrombosis occurs the forces of organiza,tion and recanalization pay little respect to the cellular nature of the valve, but treat it in the same way as they do the rest of the venous content, that is, the clot. In other words, the valve cusp is immediately subjected to disrupting forces. At first these forces consist of a growth *The exact mechanism for the production of the arteriospasm is not known. Leriche, n. 2? Homans,* and Jennings% have surgically explored the vascular sheath of the iliac and femoral vessels in some cases of phlebitis, and have occasionally found a nonpurulent plastic inflammation involving the adventitia of artery, vein and lymphatics, all with agparently equal virulence. Given such an inflammation, the spasm of the artery is explainable by direct irritation. In most instances, however, the vein is involved by itself and the artery is reflexly contracted. The possible path of such a reflex has been touched upon by Cornil, Mosinger. and Audier,% but has not yet been demonstrated. Since this paper was written, de Takats (Arch. Surg. 34: 939, 193’7) has reported vasospasm originating from phlebitis and other lesions and gives a fine discussion of the probable reflex arc. tThere are several reported cases of gangrene in the course of massive venous thrombosis, in which the arteries were demonstrated to be patent. In these cases, besides the extensive venous obstruction two other factors seem responsible, namely, spasm of the arteries and superimposed infection. EDWARDS : OBSERVATIONS ON PWLEBITIS 437 of capillaries through the base of the cusp and the exudation of reparative cells. But the fibroblasts, as t,hey grow into the clot, cut across the cusp as well and divide it into fragments. At least some recanalization always occurs and the cusp is nom more seriously fragmented by the new blood channels which run through it, and push its fragments apart. The cusp soon completely disappears. Occasionally the valve is open at the moment of thrombosis and its cusps therefore lie close to the vein wall. In such a case its fragmentation may he slight, but the cusp lies imbedded in the organizing thrombns ant1 I Fig. 9. Fig. S. LongiI%-. X.-Destruction of the venous valve by complete thrombophlebitis. tudinal section of dog’s femoral vein at valve site, four months after artificial ablebitis. Recanalization is not yet complete but it has already broken up the tbrombus leaving trabeculae crossing the lumen. In one of these, at C-1, is a small Piece of At C-8 is a long&r ;sJ; elastic tissue, a remnant of the previously existing valve. representing the second cusp imbedded in the new thickened intima. areas in the heavier trabeculae below are nuclei and pigment. Verhoeff’s elastic, tissue plus van Gieson’s stains (X23.5). (From Surg. Gynec. & Obst. 65: 310, 1237.) Fig. 8.-The production of venous stenosis and regurgitation by incomplete thrombophlebitis. Longitudinal section of single cusp from human saphenous vein inThe structure is thickened and volved in spontaneous phlebitis (operative specimen) Here the shortened by the addition of connective tissue, especially on its sinus side. connective tissue effects an adhesion of cusp to vein wall. Compare the thickness of Verhorff‘q the diseased proximal half of the cusps to the normal thin distal half. elastic tissue plus van Gieson’s stains (X22). when recanalization is complete we find it incorporated in the now thickened intima. In any event, complete thrombosis always results in the actual or functional disappearance of the venous valve (Fig. 8). 438 THE AMERICAN HEART JOURNAL Incomplete or parietal thrombosis, as is well known, has a. predileeThe earliest thrombi are seen in the tion for attacking valve sites. valve sinuses, where subsequent organization produces a pad of tissue which narrows the lumen of the vessel, and does not allow complete lateral excursion of the cusp. If the valve is open at the moment of thrombosis, with the cusp lying close to the vein wall, this organization incorporates it and it disappears in the new intima. If the valve is closed and the cusp lies in the center of the lumen, some of it will be bound down by the connective tissue formation in the sinus, and the remainder thickened by a proliferation of connective tissue along This proliferation in the lateral commissure of the its two surfaces. valve may also cause the two cusps to become adherent to each other. The end-result of incomplete thrombosis is, then, a shortened, thickened, adherent cusp of little excursion (Fig. 9). These stubby, rigid valves look altogether different from the veil-like normal ones. The diseased valves resemble the mitral valve which has been involved in a rheumatic endocarditis, and we believe that the term “stenosis and regurgitation” may equally well be applied to both. In the case of complete thrombosis of a main trunk we may see an incomplete thrombosis extending into the mouths of its tributaries. Here the valves which guard these mouths will be involved. The involvement may be of the type above described under the term “incomplete thrombosis ” or there may be a special form of damage. In this latter instance the cusp as it opens into the lumen of the main trunk is caught in the thrombus of the main trunk. Here it may be destroyed if it is entirely embedded in the clot, or it becomes adherent along its free edge if it merely touches the clot. This is the type of involvement which destroys the function of t.he valves in the perforating vessels of the leg when a deep thrombosis occurs. The destruction of the venous valve or the production of a venous stenosis and regurgitation helps to clarify the disability occurring after phlebitis. Reference is of course made to the so-called “postphlebitic” evidences of inefficient return. Difficulty with the venous return exists whether or not the vein has recanalized. If the lumen is adequately restored the valves are absent or crippled. If it is not adequately restored, the blood flows in collateral channels, where, as we have indicated, the valves become incompetent due to the dilatation of these channels. THERAPEUTIC INFERENCES The management of phlebitis is undergoing considerable change. For the author, the essentially bland nature of the phlebitis of varicose veins and the almost absolute rule that both the varices and the phlebitis will recur after the acute process is over, have prompted the immediate treatment of these patients by ligation of the great saphe- EDWARDS : OBSERVATIOKS ON PHLEBITIS 439 Post,nous vein followed by small injections of sclerosinp solution. operatively the patient is treated in the same ambulatory fashion as At the time of the first. any uncomplicated case of varicose veins. observation the thrombosis may make it impossible to determine whether or not the vein was valveless; but one may rest. assured that it will be so when the recanalization has oceurretl. Therefore, whether or not the veins are surely varicosed! they are treated t,he same WH.~~ with the exception of cases of migrating phlebitis. These are treated conservatively because thrombo-angiitis obliterans is likely to be present. No pulmonary embolism or other untoward reaction has bren seen in the 63 cases thus treated.28 The fact that phlebitis produces an irritation of the sympathet,ic fibers of the limb, leads one to inquire how this may be ovrrc~ome wnd whether by so doing the patient will he benefited. Leriche and his associates”“, 3o have performed novocaine infiltration of the Inmbal ganglia with immediate cessation of pain and subsidence of the edema. In t,he chronic postphlebitic edemas this procedure was only of t,emporary value. Other French workers have obtained good results %Jrepeated subcutaneous injection of acetylcholine. The sympathetic: release so obtained by these or any methods is undoubtedly responsible for the improvement noted. The increa.se in arterial flow thta ohtained produces a greater velocity of flow in the veins, increasing their nourishment and diminishing the chances for propagation of the thrombus. If it is also true that venous spasm is a concomitant of phlebitis, its relief by sympathetic release will again increase the venous flow and inhibit the spread of the thrombosis. Apprecia tinfi the presence of sympathetic irrit.ation should make us question the use of cold applications in phlebitis. The use of the ice pack has already fallen into disrepute with many, because it promotes thromboses and devitalizes tissue. Now we have this additional reason for condemning it. Though the above methods of securing vasoclilatation deserve pxtension, I have so far been content to use local warmth in the form (Jf a heated cradle, or as hot compresses, while in the severer cases I have used parenteral foreign protein. Such therapy, as you know, diminishes the coagulability of the blood and lowers the sympathetic tolIe. Striking results have followed the use of intravenous typhoid vaccine, or in sicker patients the intramuscular injectiou of sterile milk produets. Of late the ma,tter of rest and exercise in phlebitis has als(J attracted some attention.31v 14, 32 P reviously, patients with either superficial or deep phlebitis were kept in bed for weeks or montlns in ()rdpt to minimize activity of the limb and thus cut down the chances for pulmonary embolism. The thrombus at the original site (Jf the phlebitis is more or less fixed because of the changes in the rein \vall, antecedent or secondary, but there is frequently a propagatioll of 440 THE AMERICAN HEART JOURNAL, softer more friable clot attached to either end of the original thrombus. This clot is due to the stagnation of the blood and has therefore been called a “stagnation thrombus.” These propagated clots are not necessarily in contact with vein wall all along their course, and no amount of time may suffice for them to become firmly attached. Sooner or later the patient must move his limbs, in bed or out, and if stagnation thrombi have formed they are practically sure to break off and to give embolism. It has been amply demonstrated, in superficial phlebitis at least, that immediate exercise of the limb will prevent the formation of this Deutsch, in 1929,33 and secondary clot from which emboli asise. Jaeger, in 1930,34 reported six hundred, and one hundred cases, respectively, of superficial phlebitis of the leg, treated merely by bandagin,g and forced to remain ambulatory. In these consecutive cases not a single pulmonary embolism resulted. Confronted with a deep phlebitis, it is more difficult to fly in the face of custom and to prescribe exercise. If the phlebitis has lasted for some days with the limb at rest, stagnation clots are already apt to be present. Certainly at this stage very active exercise is dangerous, but it seems logical that there is even more chance for pulmonary embolism if the stagnation is continued, than if regular gentle movement be instituted. If the patient has survived a single pulmonary infarct, he may die from a, second or third. For this reason, whenever feasible, the main venous trunk should be ligated after pulmonary embolism. If the phlebitis is fresh, my feeling is that gentle exercise of the limb should be started at once and be increased daily. The results in several cases thus treated ha,ve been gratifying. Pulmonary embolism is, however, such a serious accident, that a large series of phlebities treated both with and without exercise will have to be compared before this can be laid down as the best method of procedure. REFERENCES 1. Virchow, R.: Zur Geschichte der Thrombose, Wien. med. Wehnsehr. 7: 217, 1857. 2. Vaquez, H.: La Thrombose Caehectique, These, Paris, 1890. 3. Harris and Longcope: Quoted by Welch.8 4. deTakats, G.: “Resting Infection” in Varicose Veins: Its Diagnosis and Treatment, Am. J. Med. SC. 184: 57, 1932. 5. Aschoff, L.: Thrombosis, in Lectures in Pathology, New York, 1924. 6. Barker, N. W.: Thrombophlebitis Complicating Infectious and Systemic Disease, Proc. Staff Meet. Mayo Clin. 11: 513, 1936. 7. Brown, G. E.: Post-operative Phlebitis. A Clinical Study, Arch. Surg. 15: 245, 1927. 8. Welch, W. H.: Thrombosis, in Allbutt and Rolleston’s System of Medicine, Vol. VI, London, 1909. 9. Ducuing, J.: Quelques Signes Neuro-Sympathique des Phlebites Post-operatoires. Leur Valeur dans le Diagnostic Precoce de cette Affection, Prat. m&l. frang. 14: 222, 1933. 10. Ducuing, J.: Le Reflexe Pilo-motor dans le Diagnostic Precoce des Phlebites Post-operatoires, Bull. Sot. d’obst. et de gym%. 22: 255, 1933. EDWARDS : OBSERVATIOA-S ON PIII,EBITIS 441 11. Minor, V.: Ein neues Verfahren zu der klinischen Untersuchung der Schweissabsonderung, Deutsch. Ztschr. f. Nervenh. 101: 302, 1928. 12. Audier, M.: Le Test D’Aldrich et McClure dans le Diagnostic des Phlbbites des Membres, Gaz. med. d. France, p. 18, April 1, 1934. l:?. Homans J . The Operative Treatment of Phlegmasia Alba Dolens, New Engl&d’J. Med. 204: 1,025 1931. I-4. Homans, J.: Thrombophlebitis, in Nelson’s Surgery, Vol. III. 15. Lawen, A.: Arteriospasmus bei akuter massiver Thrombose der V. femoralis, Zentralbl. f. Chir. 56: 1,681, 1934. Thromboses Veineuses. Obliterations 16. Wertheimer, P., and Frieh. P.: hrterielles, et’ Gangrene des’ Membres, Presse med. 43 (1 j : i,OO4, 1935. 17. Audier, M.: La Symptomalogie Arterielle dans le Phlebites des Membres et de leurs Sequelles, Progrbs. med., May 4, 1935, pp. 729T738. 18. Audier, M.: Thromboses Veineuses Aigues simulant L ‘Embolie ArtErielle ties Membres, Paris med. 1: 384, 1936. 19. Gutzeit, R.: Brand durch Venensperre, Miinchen. med. Wchnschr. 83: I ,tit’S. 19::ti Sudden O(lclusion of the Arteries of the 20. McKechnie R. E and Allen E. V.: Extremities, SGg. Gynec. i Obstet. 63: 231, 1936. 21. Leriche, R.: Essai de Traitement Chirurgical des Suites EloignCes des Phlebites du Membre Inferieur, Presse med. 27: 309, 1923. 22. Leriche, R.: Traitement Chirurgical des Suites Eloignees des Phlebites et des Grands Oedemes non Medicaux des Membres Inf&ieurs, Bull. et mcm. Rot. nat de chir 53: 187 1927. 33. Homans, J.: * Phlegmasia Alba Dolens and the Relation of the Lpmphatics to Thrombouhlebitis. A%f. HEART J. 7: 415. 1932. 24. Jennings, Jr E.: Choked Leg, Ann. Surg.‘98: 928, 1933. 25. Cornil, L., Mosinger, P., and Audier, M.: Les Interrelations alrterio-Veineuses Pathologiques des Membres, Presse med. 22: 411, 1937. 26. Homans, J.: The Operative Treatment of Varicose Veins and Ulcers, Based Upon a Classification of these Lesions, Surg. Gyuee. & Obstet. 22: 143, 1916. 27. Edwards, E. A., and Edwards, J. E.: The Effect of Thrombophlehitis on the Venous Valve, Surg. Gvnec. & Obstet. 65: 310. 1937. 28. Edwards, E. A. :’ Phlebitis of Varicose Veins. To be published. 29. Leriche, R., and Kunlin, J.: Traitement Immediat des PhlCbites Post-Operatoires par L’lnfiltration Novocainique du Sympathique Lombaire, Presse m&l. 76: 1,481, 1934. 30. Kunlin, J., and Lucineaco. E.: Resultats du Traitement Immediat ties Phlebitrs Post-Operatoires et Variqueses par L’Infiltration Novocainique du Sym pathique Lombaire. Cinq. Nouvelles Observations. Bull. et mem. Rot. nat. de Whir. 61: 965, 1935. 31. Kilbourne, N. J.: Phlebitis. Its Treatment, California 8; West Med. 45: 176. 1936. 32. Ochsner, A.: Thrombophlebitis. In Dean Lewis’ Practice of Surgery, Vol. XII. 33. Deutsch, H.: Zur Behandlung der akuten lokalizierten Phlebitis der untertn ExtremitPt, Wien. klin. Wchnschr. 42: 1,162, 1929. Zur Behandlung der Thrombose und der Thrombophlebitis, Zent~~~lbl. 34. Jaeger, F.: f. (.‘hir. 57: 1,921, 1930.