Document 2qjgQ1B7mvD14n5DD4ooBLD7R

1 c> -- /o-> \J lymphedema.1 In- the case of an abnormal lymphoscintigram, we then recommend a roentgenographic lymphangiogram, as it still is a more definitive test. __ Because of its safety, simplicity, and rewarding yield of diagnostic in formation, we strongly advocate the use of radionuclide lymphangiogra phy as the diagnostic imaging pro cedure of choice in the initial eval uation of patients with unexplained lower-extremity edema. MAJ Frank Vieras. MC, USAF Charles M. Boyd. MD University of Arkansas for Medical Sciences Little Rock 1. Sherman A!, Ter-Pogassian M: Lymph node concen tration of radioactive colloid gold following interstitial injection. Cancer 6:1238-1240, 1953. 2. Sage HH, Gozum BV: Lymphatic scintigrams: A method for studying the functional patterns of lymphat ics and lymph nodes. Cancer 11:200-203, 1958. 3. Hauser W, Arkins HL, Richards P: Lymph node scanning with MTo-sulfur colloid. Radiology 92:1369: 1371, 1969. 4. Fisher DA: Obscure and unusual edema. Pediatrics 37:506-528. 1966. Blood Donors--Paid or Volunteer? To the Editor.-Dr Allen, with charac teristic zeal in the crusade for volun tary blood donation, has missed the point of our article (234:1051,1975) in his letter (235:1327, 1976). The objec tive of our article was to report our experience in changing from a paid to a volunteer blood donor program in compliance with the American Asso ciation of Blood Banks definition of a volunteer donor. Our results empha size the problem with the definition of volunteer vs paid blood donor. Dr Al len apparently believes he can intui tively define paid and volunteer do nor. If the definition were as clear as he implies, there would not have been such a variety of opinion on this sub ject expressed at the March 18, 1976, Pood and Drug Administration public meeting on blood labeling. We have received inquiries con cerning the actual number of cases of posttransfusion hepatitis identified as a result of blood transfusion from our volunteer and paid groups. From two weeks after the beginning to six months after the end of each study period, six cases of clinically overt posttransfusion hepatitis were identi fied from the volunteers, while no cases were identified from the paid group. We have, continued to follow up the donor population receiving three-day passes and have observed a decrease from the hepatitis B surface antigen (HB.Ag) positivity of 0.9% previously reported. From May 1975 through April 1976, five of 2,632 donors (0.2%) were HB.Ag-positive by radio immunoassay. Since the donors still receive three-day passes, this de crease in HB,Ag positivity is unre lated to the three-day pass. Other fac tors, such as repeat donations with the elimination of HB,Ag-positive do nors and possibly fewer drug abusers, must account for this change. Dr Allen's comments do not con tribute significantly, for it appears not to matter whether you call our do nors receiving three-day passes paid or volunteer. Currently, they are a satisfactory donor population. MAJ Paul W. Holley. MC. USA COL Quy C. Glenn, MC, USA Department of Pathology FHzsimons Army Medical Center Denver Rocky Mountain Spotted Fever: Diagnosis To the Editor.--DeShazo et al (235:1353, 1976) have shown that Rickettsia rickettsii infection in mon keys can be confirmed as early as four days after onset of fever. We agree that their data warrant investigation of the monocyte culture technique as a diagnostic tool in human rickettsial disease. If the monocyte culture technique is to be successfully adapted as a useful diagnostic tool for humans with Rocky Mountain spotted fever (RMSF), a major obstacle will have to be overcome. Patients with RMSF often do not seek medical care promptly, and there are sometimes delays in the initiation of appropriate antimicrobial therapy by physicians. In a study of RMSF in the south eastern part of the United States, the mean duration of time from onset of symptoms (defined as the appearance of any one of the following: headache, fever, or skin rash) to the initiation of effective treatment (tetracycline or chloramphenicol) was 8.3 days for eight fatal cases and 5.4 days for 96 nonfatal cases of RMSF (P < 0.05) (J. N. MacCormack, D. J. Sexton, R. N. Philip, unpublished data). There was no significant difference between the mean intervals from onset of symp toms to consultation of a physician for fatal and nonfatal groups: 3.5 days for fatal cases and 2.7 days for nonfatal cases (P < .25). However, the mean interval between consultation of a physician and the initiation of specific antibiotic therapy was 4.8 days for the fatal group and 2.5 days for the nonfatal cases (P < .05). Even a delay of four to six days from onset of symptoms to the initia tion of antimicrobial therapy could be potentially fatal. Indeed, monkey No. 3 died five days after onset of feveronly one day after the appearance of a rash, and only one day after the monocyte culture was first noted to be positive. Similar rapid deaths have oc curred in humans with RMSF.11 Daniel J. Sexton. MD University of Missouri Medical Center Columbia J. N. MacCORuACK. MD. MPH Department of Human Resources Division of Health Services Raleigh. NC 1. Sexton DJ. Banks PM, Weig S, et al: Late appear ance of skin rash and abnormal serum enzymes in Rocky Mountain spotted fever. J Pediatr 87580-582,1975. 2. Parker RR: Symptomatology and certain other as* pects of Rocky Mountain spotted fever, in Moulton FR fed): The Rickettsial Diseases of Man Washington DC, American Association for the Advancement of Science, 1948, pp 139-146. "Meat-wrappers' Asthma" To the Editor.--The renewed"interest in meat-wrappers' respiratory prob lems (235:915; 937; 943,1976) is pleas ing, particularly in the light of three unexpected disappointments in the three years since we wrote our origi nal report on meat-wrappers' asth ma1: (1) the hot-wire technique for cutting polyvinyl chloride (PVC) has not been eliminated; (2) the existence of any association between reversible bronchospastic disease and meat wrap ping is still questioned1; (3) a pro spective study designed to define the incidence, prevalence, and long-term morbidity, if any, of this condition re mains to be done. Although we have been unable to get the support necessary to do a ran domized longitudinal study of pulmo nary function in meat wrappers, I un derstand that such a large-scale study will be undertaken elsewhere. Until this is done, further anecdotes may be helpful to the practicing physician. I know of at least two "disabled" meat wrappers who remain free of symp toms after returning to work using mechanical cutters in lieu of the hot wire. Several other patients who were receiving disability benefits else where for "meat-wrappers' asthma" were found to have normal pulmo nary function at the time of referral to me, suggesting reversibility of the disease process after cessation of' ex posure to PVC pyrolysis products: I agree with Falk and Portnoy that respiratory illness other than asthma should be looked for. Some nonsmok ing, healthy meat wrappers noted an association of smoking meat wrap pers with frequent absenteeism be cause of pneumonia and bronchitis prior to our publication (from per sonal interviews). However, all of our JAMA Sept 6,1976-Vol 236, No. 10 Letters 1117 R&S 131095 f v. patients with objective illness had a reversible bronchospastic pattern of disease, ie, asthma. Asthma in meatwrappers does appear to be a prob lem. This should be distinguished from other conditions occurring in meat wrappers, such as lacrimation and rhinorrhea (which tend to be uni versal with close contact to fumes) and bacterial respiratory infections. Before accepting the conclusion of Andrasch and Bordana that the pricelabel fumes caused their patient's symptoms, I would like some assur ance that their provocation actually simulated real working conditions; if not, did they test those who were not meat wrappers with asthma to rule out the possibility that perhaps any asthmatic would react dramatically to their challenge technique? Lastly, the term "restrictive disease" applied to their patients should be decried. Al though the forced vital capacity was mildly depressed at 72% of predicted value (normal, > 80%), the degree of obstruction suggests that the de crease was due to air trapping and that total lung capacity would be nor mal or "super normal." The term "re strictive" requires that total lung ca pacity be less than 80% of a predicted value, which would be unlikely, to say the least, considering the data pre sented. While awaiting more definitive studies, physicians have an important responsibility to counsel meat wrap pers not to smoke and, whenever pos sible, to support patients' efforts to minimize exposure to pyrolytic prod ucts of PVC. Y.Aeloht.MD Southern Cattfomia Pennanente Group UCLA. Harbor Campus Harbor City 1. Sokol WN, Anlony Y, BeaD GN: Meat-wrapper's asthma: A new syndrome? JAMA 226639-641,1973. 2. Finley T. Mestwrappera' asthms: What is it, tf It ia? Read before the California Thoracic Society Meeting, San Diego, Calif, April 1975. In Reply.--We share Dr Aelony's pleasure with respect to the contin uing interest in the respiratory prob lems of individuals exposed to compo nents of PVC. We wish to emphasize that the term "meat-wrappers' asthma" is somewhat of a misnomer in that the spectrum of exposed indi viduals extends beyond the meatwrapper, and the pulmonary involve ment may occur as acute reactive airway disease or a chronic disorder with both obstructive and restrictive patterns. Wegman* has reported fine nodular changes on the chest x-ray films of three workers of 37 exposed to PVC only. Significant restrictive disease was noted in a larger study of vital capacity and PVC workers by Lilis and his associates.1 Recently, Maccia and his col leagues3 have provided substantia] evidence that phthalic anhydride (a component of some PVC soft wraps, and thermally activated labels) can induce clinical sensitization with high titers of specific IgE. Oddly enough, Kern1 made the truly original obser vation of this phenomenon on clinical grounds in 1939. Phthalic anhydride also has a known irritative effect on the bronchi, and may aggravate ex isting chronic obstructive pulmonary disease in exposed workers.1 As a result of these and many other investigations,* the government has taken steps to protect the exposed worker (29 code of federal regu lations: 1910.1017). In 1975, the Joint Labor Management Committee of the Retail Food Industry awarded a sub stantial grant to Dr Wegman and his associates at the Harvard School of Public Health to conduct a prospec tive five-year study of acute and chronic respiratory disease of work ers in meat departments. Newly de signed automatic wrapping and label ing units with efficient ventilation systems are replacing hot-wire cut ting machines. The work-dmulated-inhalation prov ocation .studies in our case report' simulated with as much accuracy as possible the actual work conditions. The results of a larger series with more specific details of the methodol ogy used will appear in the Journal of AUergy and Clinical Immunology and was recently summarized with picto rial descriptions of procedures.* Sim ilar provocation studies of those who are not meat wrappers with asthma may pose a potential hazard for sensi tization of an uninvolved subject. With regard to the term "restrictive disease" in our patient, we agree that the measurement of total lung capac ity is useful to indicate the presence of restrictive impairment in the set ting of airway obstruction, but this was not performed. However, slow vi tal capacity was performed on numer ous occasions and ranged from 37% to 74% of the predicted value and was similar to the fast vital capacity, sug gesting true restrictive rather than the restrictive effect of air trapping. E. A Bammna, Jr, MO B. H. Amorasch. MD umwntty of Oregon Hsaltti Sciences Center PorSsnd 1. Wegman D: Furtka ranks m polyvinyl ddoride production wwkn Am NY Axmi Sci 246:18-21,1975. X Ufa R, Andenoa H, MSfar A. ct ah Pubnooary changes nnp vinyl-chloride polymerisation worker*. Chest W (supplk296308, 1976. A Maccia CA, Bernatein IL, Emmett EA, et ah In vitro demonstration of specific IgE in phthalic anhydride hypmensitrrity. Am Acs Rap Du 113.-701-704, 1976. A Kern RA: Asthma and allergic rhinitis due to acnaitaaatiw to phthalic anhydride: Report of a case. J Allan 10:164-166.1939. 5 Pope V, Teculnscn D, Staneacu D, et eh Bronchial asthma and asthmatic bronchitis determined by simple chemicals. DU Chest 56:396-401, 1969. 6 Selikoff U, Hammand EC (edsk Toxicity of vinyl chloride polyvinyl chloride, Ann NY Acad Set 2461-322, 1975. 7. Andrasch RH, Berdana E-I: Thermoactivated pricelabel fume intolerance: A -cause of meat-wrappers' asthma. JAMA 235:937,1976. 6 The NU1 Rccvrd. National Institute at Health pplvliestioo No. 11, US Dept of Health, Education, and Wel fare, 1976, ml 38, p 7. Mannitol and Amphotericin B. To the Editor--We wish to comment on the recent report by Rosch et al (235:1995, 1976) that suggests a pro tective effect of mannitol when used in conjunction with the nephrotoxic antifungal agent amphotericin B. Al though the single case report sug gests a role for mannitol, it never theless remains anecdotal. The hazard of extending this type of information to clinical practice must be empha sized. Although a report' from this insti tution likewise suggested a beneficial effect of mannitol in recipients of am photericin B, a controlled study that we are currently concluding indicates no protective role. An identical con clusion was reached by Bullock et al.1 These investigators found no dif ferences in the degree of renal im pairment between groups of patients receiving amphotericin alone, and amphotericin plus mannitol. We therefore urge that purported beneficial results of therapy be sub jected to controlled trial before ac ceptance by the medical community. Temple W. Williams. Jr, MO Juan J. Ouvero, MD Emc L. WESTEftMAN. MD Baylor University College of Medicine Houston Jewry D. Smilack, MD Dalles I. Oiivero JJ, Loeano-Meodes J, Gbafsry EM, et ah Mitigation of amphotericin B nephrotoxicity by mannitoL Bt M*d J1*50-561,1975. X Bullock WE, NnttaD CE, Luke RG, et al: The effi cacy of mannitol in reducing amphotericin B nephro toxicity: A double-blind study, abstract No, 18, in Ab stracts of the IStk JnUncicna Confront* oa Anti microbial Agent* and Chemotherapy. Washington, DC. American Society for Microbiology, 1975- CORRECTION Misspelled Name.--In the original contribution, "Chemoprophylaxis for Patients With Colorectal Can cer," published in the June 28 issue (235:2825-2828, 1976), an error was made in reference 15 (p 2828). The first author of that reference is Ros' wit B, not Burnt, as published. ' 1118 JAMA. Sept 6.1976-Vol 236, No. 10 Letters R&S 131096 R&S 131097 Letters V Letters, if clearly marked, "For Publication, " will be published as space permits and at the discretion of the editor. They should be typewritten double-spaced, with jive orfewer references, should not exceed 500 loords in length, and will be subject to editing. Letters are not acknowledged. of work remains' to be done to make both natural and synthetic materials safer to use. W. R. Sorenson. PhD Continental Oil Company Ponca City, Okie 1. Coroiah HH, Abar EL: Toxicity of pyrolysis prod ucts at vinyl plastic*. Arch Environ Heultk 19:15-21, 1969. Polyvinyl Chloride In Fires. To the Editor.--The article "Polyvinyl Chloride Toxicity in Fires" by Dyer and Esch (235:393-397, 1976) contains a number of factual errors and cer tain misleading inferences. In three places, the combustion of polyvinyl chloride (PVC) is said to produce chlorine and phosgene. In fact, these two materials have never been demonstrated to occur in the py rolysis or combustion of PVC {Jour nal of Polymer Science 8:1887-1890, 1970; Journal of Applied Polymer Sci ence 13:377-391,1969; British Polymer Journal 3:186-193, 1971; Journal of Applied Chemistry 17:366, 1967). The authors point out, correctly, that hy drochloric acid and carbon monoxide are the main toxic combustion prod ucts of PVC but tie these to possible toxic synergy with chlorine or phos gene. It is wrong to raise the specter of toxicants which are not really there. In another instance, vinyl chlo ride is said to be produced. In work by Boettner et al {Journal of Applied Polymer Science 13:377-391, 1969), vi nyl chloride was found in the combus tion products of PVC corresponding to about 0.02% to 0.06% by weight of the original polymer. Carbon monox ide and HC1 were formed at about 45% and 58%, respectively. The vinyl chloride was thus present at a tiny fraction of the primary toxicants and may indeed represent residual vinyl chloride from the original poly merization. The present-day levels of residual vinyl chloride in PVC are .01 or less the levels typical in 1969, when Boettner's work was done. Thus, whether vinyl chloride is a combus tion product needs confirmation. A discussion of the fuel loading in fire situations owing to the presence of plastics carries the statement that .. plastics posses a heat of combus tion 2% times that of other combus tibles." The matter is not so simple. According to the National Bureau of Standards {Modem Plastics 52:81-82, 1975), the heats of combustion of PVC and some common materials, includ ing other plastics, are Edited by John D. Archer. MD, Senior Editor. Cotton PVC Wood Wool Polyester fiber Polyurethane, Polyethylene 7,122 BTU/lb 7,720 BTU/lb 8,825 BTU/lb 8,972 BTU/lb 9,300 BTU/lb 16,000 BTU/lb 20,050 BTU/lb One can conclude from this that, on a pound-for-pound basis, PVC contrib utes less to fuel loading than many other materials in common use. The assignment of a flat 2%-fold factor for plastics over other materials in heat of combustion is, to say the least, misleading. The authors cite work by Cornish and Abar1 on exposure of rats to CO and HC1 from pyrolysis in air or oxygen of pure and formulated PVC. This work pointed out that death from pure PVC pyrolysis was due to CO inhalation unless oxygen was added to the gas stream to keep carboxyhemoglobin blood levels low. It should be. noted that when PVC for mulated for wire insulation or floor tile was pyrolyzed, higher exposure levels were tolerated before death oc curred. Thus, formulated PVC is ap parently less toxic in fires than pure PVC and the former should be the subject of research rather than the latter,, which, is never used commer cially. The commercial impact of PVC is exaggerated. The 8.3 billion kg/yr stated as the annual production in the United States translates to 17.6 bil lion kg/yr in contrast to the 3.7 bil lion kg/yr actually estimated for 1975 for all uses. Roughly 123 billion lb/yr of lumber and plywood were used for construction alone. Upholstered chairs are not stuffed with PVC, as claimed in the article. Chair covering is often PVC-based; the stuffing is usually rubber, urethane, or a nonsynthetic. We want to make it clear that we do not take lightly the potential prob lem that may occur from the combus tion. of articles made from PVC. We stress, however, the need for dis semination of accurate information. It's reasonable to think that society can cope with the true problem, accu rately understood, as well as it has with the toxicity problems from con ventional materials. That means a lot Cardiac Care Units To the Editor.--Frieden and Cooper (235:816,1976) stated that "recent ex periences with monitored cardiac care units have demonstrated that mortal ity can be reduced in the first few days after an acute myocardial in farction." However, critical review of articles on coronary care units fails to confirm a properly designed trial that supports that statement. Adler et al {Br Med Bull 3(h242, 1974), in a re view dealing with research on medi cal care, point out: Mortality within most coronary care units has declined over the years, but there has also been a parallel reduction in mortality in patients suffering from myocardial in farction, who are nursed in conventional wards (MacMillan & Brown, 1971). There is no evidence that hospitals play any part in altering the total picture of die disease. In fact, the first randomized con trolled trial comparing home care vs that in cardiac care units (Mather et al cited by Adler et al Br Med J 3:334, 1971) showed no. striking difference in case fatality ratios between groups treated in the two settings. There may be other impressive ad vantages to grouping patients, equip ment, and trained personnel in one location within the hospital setting. Grouping makes sense. However, re duced case fatality ratios have not been demonstrated. The conviction that cardiac care units are effective has been associated with the view that trials are not only unnecessary but unethical. It is un fortunate that rigorous evaluation was not carried out when units were first introduced, but British experi ence suggests that it is not too late to start. Robert Oseasohn, MD McGill University Montreal In Reply.--It is important to delineate the perspective in which a situation is viewed. It is generally agreed that the majority of cardiac deaths occur early and outside of hospitals. Our report and the reports of the cited authors deal with the subgroup of treated patients. I believe that one should first determin whether or not we are helping these patients. If we JAMA, Sept 27, 1976-Vol 236, No. 13` Letters 1449 R&S 131098 are having a positive effect, then we should try to expand the subgroup in stead of trying to see if treating those within it has a significant effect on the larger group of patients. The statement that "recent experi ences with monitored cardiac care units have demonstrated that mortal ity can be reduced in the first few days after an acute myocardial in farction" was supported by refer ences to Lown et al, Killip and Kimbal, and Day. Their very impressive results would make it difficult to set up a prospective study of the useful ness of cardiac care units unless a subgroup at low risk was identified. After reviewing the article compar ing home vs hospital care by Mather et al, it is difficult to draw any conclu sions except that healthier patients fared better than the sick patients. I believe that the weight of evi dence indicates that careful, vigorous therapy of arrhythmias in specialized, monitored units is lifesaving. As pa tients and physicians are educated to understand that myocardial infarc tion is best treated early after onset, the patient will appear in hospitals sooner, and perhaps the incidence of arrhythmic deaths will be further de creased. Jerome A. Cooper. MD Albert Einstein College of Medicine Bronx, NY Sp ntaneous Rupture of the : Esophagus To the IScfitor.-Campbell and associ ates, in ,their article "Spontaneous Rupture of the Esophagus (Boerhaave Syndrome)" (235:526-528,1976), have clearly indicated the necessity of early diagnosis and treatment of this condition. Because of the sporadic oc currence of this syndrome, it is not unusual to see a patient who has been treated for a number of days with a presumptive diagnosis of other intraabdominal or intrathoracic inflammatory processes. While mis diagnosis is seen at times, mis treatment (in spite of a correct diagnosis) rarely occurs. Correct diag nosis, therefore, should be followed by correct operative technical manage ment of these patients. The technique employed by Camp bell and associates in the operative treatment of their patient (utilizing catgut to close the perforation) is inadequate and may lead to complica tions. Their patient developed an esophageal fistula and eventually re quired additional operations. This type complication occurs infrequent ly1 but usually is preventable if the technique of closure is meticulous. In a previous publication from this insti tution1 we indicated a new technique for management of these cases. Our technique is based on the find ing that the esophageal mucosal tear is always larger than the small open ing seen in the muscular layer. Most surgeons close the rent seen on the muscular layer of the esophagus, ig noring the extent of mucosal layer tear. This, we believe, is the usual reason for fistula formation and post operative leakage. In brief, the tech nique utilizes enlargement of the rup ture in the esophageal muscular layer which is seen at the time of surgery. The opening is enlarged longitudi nally above and below along its long axis. By this maneuver it is possible to see the ends of the esophageal mucosal tear-especially the lower end, which at times extends into esophagogastric junction. Two layers of closure are then accomplished, using 0000 silk cardiovascular sutures. Sutures are taken approximately 3 mm apart and are tied snugly. The muscular layer is closed in the same fashion by interrupted sutures. By this method, we believe the entire rupture is repaired in layers. The me diastinum is then washed thoroughly, as indicated by Campbell and associ ates. With: meticulous postoperative care for management of fluid and electro lytes, all our patients have survived. In our report, this technique was then used in six consecutive cases, and to date we have used it in ten cases. There have been no postoperative esophageal leaks, and no additional surgery was required in any case. We agree that a postoperative esoph- agogram should be done on the fourth or fifth day in all cases prior to initia tion of oral feeding. In those patients in whom esophageal perforation has been present for longer than 48 hours and in whom there is extensive loss of esophageal wall, we suggest the oper ation advocated by Grille and Wil- Irjjift 3 Hooshanq Bolooki. MD Unfverttty of Miami (Fla) L Berne CJ, Shader AE, Doty DB: Treatment of effort rupture of the esaphogiu by epigastric celiotomy. Surg Gynecol Otwtet 129277, 1969. 2, Bolooki H: Spontaneous rupture of the esophagus; Boerfcaave's syndrome. Ann Snrg 174:819*324,1971. 3. Grillo HC, Wilkins EW Jr Esophageal repair follow ing late diagnosis of intra-thoracic perforation. Ana Thorac Surg 20387*399, 1975. To the Editor.--We refer to the publi cation of Campbell and his colleagues on spontaneous rupture of the esoph agus (Boerhaave syndrome). The au thors intended to emphasize the diag nostic approach to this entity. However, we feel that a simple but important diagnostic procedure was omitted: the determination of the pleural fluid pH. The ruptured esoph agus and the repeated vomiting cause the gastric contents to invade the pleural cavity, and a pleural fluid pH of less than 6 is highly suggestive of esophogeal rupture.' Such a procedure resulted in a prompt diagnosis in one of our patients. Dan Aderka. md Bernard Shtamler, MD Harry Suprun, MD Gideon maneus, MD Naharyla Government Hospital Naharyia, Israel 1. Dye RA, Laforet EG; Esophageal rupture; Diag nosis by pleural fluid pH. Chest 66:454-456,1974. Propranolol Blockade of Stress Reaction To the Editor.--Propranolol hydro chloride (Inderal) has the property of blocking catecholamine-induced in creases in heart rate and blood pres sure. As a result of this blockade, the stress response was altered in a pa tient with a duodenal perforation, re sulting in delay in diagnosis and treatment. Report of a Case.--A 51-year-old man with chronic glomerulonephritis and hy pertension who received intermittent he modialysis therapy underwent renal trans plantation on Feb 15, 1975. He received prednisone and azathioprine in standard doses for immunosuppression. Propranolol hydrochloride, 10 mg four times per day, was administered for hypertension from the eighth postoperative day. Other medi cations included methyldopa, 500 mg four times per day; diazepam, 5 mg three times per day; and diphenylhydramine hydro chloride; 50 mg/day. On the 15th postoper ative day, there was a decline in renal function, and the patient was treated for a rejection episode with daily injections of methylprednisolone intravenously: he re ceived L75 gm during the next two days. The prednisone dosage was elevated to 100 mg/day. The patient complained daily of "gas pains," which he attributed to his ant acid medications of magaldrate and alumi num hydroxide, both of which he often re fused to take. The serum creatinine level was 3.1 mg/100 ml on the 16th postopera tive day. On the morning of the 17th postopera tive day, the patient complained more bit terly of abdominal pain. On examination, he had a pulse of 60 beats per minute and blood pressure of 218/95 mm Hg. Bowel sounds were present, and the abdominal examination showed no abnormality. Six hours later the patient's blood pressure was 90 mm Hg systolic, but his pulse re mained slow at 60 beats per minute. Exam ination of the abdomen' at this time re vealed muscle spasm and absent bowel sounds. Free intraperitonea! air was seen on a roentgenogram of the abdomen. At exploratory laparotomy, a superoposterior T450 JAMA, Sept 27, 1976-Vol 236, No. 13 Letters R&S 131099 Angiosarcoma of the Liver in a Rural Population Four Cases Diagnosed in a 29-Month Period Justus J. Fiechtner, MO, Cesar N. Reyes, Jr, MD Angiosarcoma of the liver has recently been publicized because of its association with poiyv-.nyl chloride (FVC) polymerization workers. Four cases of this rare turner have been observed at the Marshfield Clinic within a 29-month period. There are no factories that manufacture PVC products in the immediate area, nor were any of the victims ever involved in such manufacturing work. Thus, It Is possible that other factors may be related to this cluster of the disease. (,JAMA 236:1704-1706, 1976) ANGIOSARCOMA of the liver (ASL), also known as primary hepatic hemangiosarcoma, Kuppfer cell sar coma, malignant hemangioendothelio ma, angiobiastic sarcoma, and malig nant or metastasizing hemangioma, is a rare disease. Alrenga1 estimates that the literature until 1975 contains 165 case reports of the disease, with an autopsy frequency of 6 in 100,000. Heath et al2 state that the expected annual incidence of the tumor is 0.014 in 100,000, or 25 to 30 cases a year, in the entire United States. There were only 8 cases in 40,000 autopsies re ported over a 25-year period from Holland (population of 10 to 14 mil lion).3 However, 14 cases were noted in a population of 5,000 workers in the US PVC polymerization plants in a 15-year period.* The purpose of this communication is to present observations on four cases of ASL evaluated during the 29 months from June 1973 to November 1975 at the Marshfield Clinic-St Jo seph's Hospital in rural central Wis consin (Table). Three of the eases came from within a population base From the Residency Program ot the, Depart ment of Internal Morjicine (Dr Fiechtner) and the Department ot Patnoloqy (Dr Rnves). Marshfield Clinic and MtMic.u Foundation and St Joseph's Hospital. Marshfield, Wis, Reprint requests to Marshfield Clinic, 1000 N Oak Ave. Marshfield, Wl 54449 (Dr Fiechtner). of 130,000 and an 80-km radius. The fourth case came from a community 224 km from this center. The total number of autopsies performed in this time period at the hospital was 391. REPORT OF CASES Case 1.--A 65-year-old man (lumber plant worker, retired) was admitted on Nov 12,1975, with a three-week history of nausea, weakness, anorexia, and slight ab dominal pain. A hemogram prior to admis sion indicated a Coombs-negative hemolyt ic anemia. Physical examination disclosed severe diaphoresis, restlessness, and tachy cardia. His left hemithorax had many moist rales, the abdomen was tender in the right upper quadrant, and the liver was palpated 10 cm below the right costal mar gin. His skin contained ecchymoses, and a bleeding time taken at the bedside was 11 minutes. Soon after the patient was ad mitted his blood pressure decreased precip itously, and he had a cardiopulmonary ar rest. Resuscitation was successful, but it was soon apparent that renal failure had occurred. Laboratory studies were con sistent with a diagnosis of disseminated intravascular coagulopathy and profound hemolytic anemia. A gingival biopsy speci men taken to test for thrombotic throm bocytopenia purpura was unrevealing. The patient was treated wilh heparin stxiium, blood replacement, and large doses of methylprednisolone sodium succinate, A liver scan showed a large lesion in the right iot>e. On Nov 19, 1975, an open liver biopsy of the liver, which hied profusely wr.i" i'-. _ biopsy specimen was taken. A hepatic nv.giogram taken the same day suggested tu mor with hematomas, especially in the right lobe of the liver. The biopsy specimen was interpreted as showing angiosarcoma. The patient's condition deteriorated, and he died Nov 24. At autopsy, there was a large angiosarcoma of the right lobe o: the liver with metastatic implants in the pario tal and visceral peritoneum, the lungs, and the left lobe of the liver. There was aiso evidence of pulmonary edema and acute bronchopneumonia. Case 2.--A 45-year-old male postal em ployee was referred to the Marshlieii Clinic in October 1973 for evaluation ox jaundice following urological surgery. The history included ankle edema for two years and a mild, nonproductive cough. On exam ination at that time, findings included der mal and scleral icterus and a barely pal pable liver. The prothrombin time was 15 seconds (control, 11.5 seconds); total bili rubin level, 7.6 mg/100 ml: indirect bili rubin level, 4.7 mg/100 ml; and there was no evidence of hepatitis-associated antigen (HB.Ag). An intravenous pyelogram showed chronic atrophic pyelonephritis of the right kidney and a large left kid ney. Radioisotope liver-spleen scan showed a small, poorly functioning liver with sple nomegaly. An open liver biopsy provided a pathological diagnosis of peiiosis hepatis. The patient's jaundice cleared and he did well until August 1974. when he was again admitted to St Joseph's Hospital. At that time he had jaundice, fetor hvpaticu-, a harsh grade 4/6 systolic murmur at the cardiac apex, a protuberant alHlomcn wilh obvious ascites, and peripheral edema (3+1. Results of laboratory studies sug gested hemolytic anemia with hypopi.i-hc marrow. The chest roentgenogram showed bilateral pleural etVusioris. The ptj,-nt lapsed into hepatic coma and died Sept 3. Autopsy findings were dilfuse anguea-coma of the liver with Imne marrow metas tasis, fibrinous pericarditis, and a hyjni- 1704 JAMA, Oct 11. 1976-Vol 236, No. 15 Angiosarcoma--Fiechtner & Reyes )1 Case/Age at Death. yr/Sex 1/65/M Length of lllness.t mos 11/7,-11/75 Cases of Angiosarcoma of the Liver" Symptoms, Length and Type 3 weeks: abdominal pain, weakness, nausea 2/46/M 8/74-9/74 3 years, jaundice, abdominal discomfort 3/57/M 7/74-7/74 2 months: peripheral edema, taundice 4/35/M 6/73-6/73 1 month, snouiaor pain and l.nigue 'Seen at the Marshfield Clinic between June 1973 to November 1975. fFrom diagnosis. Diagnosis Method Open liver biopsy Autopsy Open liver bioor-y Autopsy \ .\ Occupation Press operator tor paper company i Postal carrier. 1 bieacher compa.iv worker 1 Rock crusner at roofing paper plant Postal clerk I! __1 R&S 131100 plastic right kidney caused by severe atherosclerosis of the renal artery. Case 3.-A 57-year-old man (rock crusher in a roofing' paoer plant) was referred to the Marshfield Ciimc on Jul> 11. 1974, for evaluation of jaundice. He had peripheral edema for two months, increasing abdomi nal girth, choluria, and achoiia for one month. A liver biopsy at another hospital had been interpreted as showing chronic aggressive hepatitis. Findings at physical examination were icteric selerae and skin, globose abdomen with a grossly enlarged liver, obvious ascites, a palpable spleen, and peripheral edema. Laboratory data in cluded evidence of mild anemia, an alka line phosphatase value of 218 IU (normal, 19 to 91 IU), and absence of HB.Ag. A ra dioisotope liver scan showed evidence of diminished hepatic function and a focal le sion in the right lobe. An open liver biopsyon July 24 was interpreted as showing an giosarcoma. The patient died the following day. Postmortem examination showed a large angiosarcoma disrupting most of the liver, with metastasis to one vertebral body. Case 4.-A 35-year-old male postal clerk was initially seen at the emergency room at St Joseph's Hospital on June 12, 1973, complaining of shoulder pain and fatigue for one month and dark urine and pale stools for ten days. Physical examination showed obvious jaundice, bilateral palmar erythema, distended abdomen with ascites, and a liver palpable 4 cm below the right costal margin and xiphoid process. Labora tory studies showed an elevated -white blood cell count of 31,600/cu mm, bilirubin level of 25.5 mg/100 ml, an alkaline phosphatase value of 985 IU, and absence of HB.Ag. An open liver biopsy showed acute hepatic necrosis. The patient deteri orated rapidly. Exchange transfusions were of no avail, and the patient died on June 15. The autopsy showed widely dis seminated liver angiosarcoma with metastases to the mediastinal lymph nodes and lung. COMMENT Etiology There are several known agents that cause angiosarcoma of the liver. '#..a *-v: W, ...S V-, ' -5'' - ,1 . '7'.4 - * .> >-Vj / -I ; ^' 7'- * v yJ-''T /-- - Vi ' J" Fig 1 .--Cut surface of tumor shows bloodfilled, cavernous spaces and satellite nod ules. Solitary nodule (arrow) in left lobe. These include thorium dioxide, a ra dioactive substance used for angiog raphy in many medical centers as re cently as ten years ago. MacMahon et al1 reported the first such case in 1947. Since that time, approximately 50 ad ditional cases due to ThO. exposure have been reported in the literature. Reviews' -r indicate that a latency pe riod of 22 years exists between injec tion of the material and appearance of clinical manifestations. In addi tion, the biological half-life of ThO- is estimated at 400 years." Arsenic has also been implicated in the disease. Roth," between 1956 and 1959, discovered eight cases of hepatic angiosarcoma among vineyard work ers in the Moselle district of Ger many. He estimated an average in take of 53.7 gm of arsenic over a 12year period in the form of arsenic tri oxide insecticide. Two cases of ASL have also been rejiorted after inges tion of potassium arsenite (Fowler so lution) used to treat psoriasis.""1 One patient took the substance for 17 years (estimated 50 gm) and the other for 15 years (approximately 15 gm). The latency periods in all these cases ranged from 13 to 22 years. Most recently, ASL has been associ ated with PVC polymerization wort ers. Block1- and Creech and Johnson' initially called attention to the rela tionship following the discovery of several cases of the disease in a small population at a manufacturing plant. Since that time, much has been writ ten to substantiate the toxicity of vi nyl chloride gas and imply that the danger may involve large segments of the population with only low-level exposure.u 15 Most authorities, how ever, believe that there is no measur able risk to the general public.'-" As with ThO. and arsenic-induced ASL^_ the latency period is about 20 yeata^^ from first exposure to diagnosis^ None of our patients underwent ThO., arteriography, and only one of the four had any known occupational exposure to a polyvinyl derivative (polyvinyl acetate, not previously im plicated in this disease). Arsenic ex posure has not been documented in any of the cases, but the common thread among them is a past history of farming or farm labor. Arsenical pesticides have been used for many years, but it remains to be seen whether exposure was sufficient to cause disease. It must also be empha sized that a very large proporiion of central Wisconsin residents have lived or worked on farms at some point in their lives. Thus, speculation on cause must certainly await further investigation. Clinical Aspects Our patients had symptoms of ad vanced liver disease, abdominal pain and distension, ankle edema, weak ness, anorexia, and weight loss. Thu duration of symptoms was usuallf brief, except for one patient in whom' three, years elapsed from onset of JAMA, Oct 11. 1976-Vol 236, No. 15 Angiosarcoma--Fiechtner & Reyes 1705 * v;` :* i ,c-. 4 -| Fig 2.--Varying histological oatterns of angiosarcoma of liver (case 1). Upper left. Si nusoidal pattern. Malignant endothelial cells line dilated sinusoids that separate cords of heoatocytes (hematoxylin-eosin. x 125). Upper right, Papillary pattern. Apparently shat tered hepatic cords surrounded by malignant endothelial cells (hematoxylin-eosin, x 31.25). Lower left. Cavernous pattern. Blood-filled spaces lined by malignant cells (hematoxylin-eosin, x 125). Lower right. Solid pattern. Sheets of malignant cells, no in tervening heoatocytes (hematoxylin-eosin, X125). jaundice to death. Physical findings included hepatomegaly, splenomeg aly, dependent edema, and cirrhotic stigmata. Laboratory studies con firmed jaundice and liver disease, often with associated anemia and ab normal bleeding tendencies. The asso ciation of hemolytic anemia with the disease seen in two of our cases has been noted before,and a single re port of concomitant disseminated in travascular coagulopathy exists in the literature.18 Diagnosis was estab lished most efficiently by open liver biopsy. This technique is preferred because of (1) the high risk of bleed ing in these patients and (2) the chance for gross inspection of the ab domen for hemorrhagic ascitic fluid and for characteristic tumor appear ance. The diagnosis should be sus pected before biopsy, however, since two of our patients had initial diag noses of simple peliosis hepatis and acute hepatic necrosis. Pathology The livers were enlarged, with weights ranging from 1,825 gm to 4,500 gm (average, 3,114 gm). Three of the four cases had a dominant mass with numerous smaller, satellite nodules (Fig 1). The fourth case had multiple varying-sized nodules with no single dominant mass. These masses and nodules were character istically described as trabeculated, loculated, or spongy, with varying- sized cystic spaces that were lilled with blood. Scattered, small areas ap peared solid. In one of the cases (case 4) there was extensive necrosis. The histological features we!" al most monotonously similar m aii cases. The predominant picture v, as one of varying-sized. Iilood-containing spaces lined by sintrie or multiple layers of neoplastic endothelial e<_-i!s. Several histologic patterns were ob served. These patterns, in addition to the cavernous pattern, are sinusoidal, papillary, and solid. In case 1. all four patterns were observed (Fig 21. Thomas et al;' described a peculiar in tralobular and capsular fibrosis in tu mor-free portions of livers with an giosarcomas as well as in livers workers without angiosarcoma that had vinyl chloride polymerization ex posure. This was not observed in any of our cases. Unfortunately, none of the autopsy tissues were analyzed for the presence of arsenic. Conclusion We have observed four cases of angiosarcoma of the liver within a 29month period in a rural population, with no documented risk-factor expo sure. The role of an etiological agent in this group of patients must await further investigation. Because of the long latency period of this tumor, we may have observed only the begin ning of a series of such cases. It therefore behooves all clinicians to consider this diagnosis at the earliest possible time. H.G. Skinner, MD.and K. Rentmeester. MPH. Wisconsin Department of Health and Social Ser vices; J. T, Herbert, MD, Center for Disease Con trol; and H. A. Anderson, MD. Mt Sinai School of Medicine, assisted with this communication. 1. Alrenga DP: Primary angiosarcoma of the liver, hit Siinj 110:198-20:1, 1975. 2. Heath CW, Falk H. Creech JL: Character istics of cases of angiosarcoma of the liver among vinyl chloride workers in the United States. Ann NY Arm! .Sri 246:2.11-2:16. 1975. Dalderup LM. Freni SC. Bras G, et al: An giosarcoma of the liver. Ijum I 1:216. ly7(i. 4. Lloyd JW: Angiosarcoma of the liver in vi nyl chloriderpolyvinyl chloride workers../ Ihnip M,d 16:809. 1971. 5. MaeMahon HE. Murphv AS. Bates M: Endo thelial cell sareoma nf the liver following Thorntrust injection. Am J hithiJ 2X:585-611, 1917. 6. da Silva Hurts J, Ahbatt .ID. Cayolla da Motta L. et al: Malignancy and other late edeets following administration of Thorotmsi. f,niW 2:2111-2115. 1965. 7 GrampaG: Radiation injury with imrtictilnr reference to Thomlrast. I'nthm Anm' 6:117-170. 1971. References 8. Hursh JB. Steadman LT. Looney WB. et al: Excretion of thorium and thorium daughters af ter Thorolrast administration. Aftn Radial 47:481-498. 1957. 9. Roth F: 2ur Pathologie derchronschen arsenvengiftung. Zentralb Alb/ 1`athiJ 10U:529-5:|i). 1959. 10. Regelson W. Kim U. Ospina J, et al; Hemangioendothulial sarcoma of liver from chronic arsenic intoxication by Fowler's solution. Grrn-er 21:514-522. 1968. 11. Lander JJ, Stanley ILJ, Sumner HW, et al: Angiosarcoma of the liver assix-oit' d with Fow ler's solution (potassium arsenitel. 'btst rmnUr- 7 68.1582-1586, 1975. 12. Block JB: Angiosarcoma of the liver fol lowing vinyl chloride ex|sisure. JAMA 229:5:i-54, 1971. IT Creech JL Jr, Johnson MX: Angiosarcoma of the liver in the manufacture of |u>lvvin>l chlo ride. J Utxup Mud 16:150-151, 1974. 11. Popper H: The heuristic importance of en vironmental pathology: Lesson- *'rom tI,. vine! chloride program. Anh /*!>2 991.9-71, 1975. 15. Vinyl chloride and cancer, editorial. Mid Asxoc J 112:269-270, 1975 16. Doll R; Discu-sion pan-r of toxicity of vi nyl chIoride-[s>lvvinvl chlorine. .Die AT Arm! Ad 246320-321. 1975.' 17. Alpert LI, BenDch B: Hemangioendothelio ma of the liver associated witn microangiopathic hemolytic anemia: Ke|mrt of four eases. .1./ AM 48:624-1)28, 1970 18. Truell JK, I'e' k SI1.2 R* 4-o5'6l.iin CW. He. manginsarvnma of iho h'1 r mt ncat.-d (,. ,in seminated intrava'C'ii.ir eoag oala.ii. i,'o-/,-,.. ti-ndmfif 65:986-942. i'*.:: 19. Thomas I.H. .......... .. II. H. ti I'll, et a!. Vinyl-chloride-indoee,i :;-.'-r ,! >,... From nine puthir portal hy|>er', *i'i o ii:..ig. . ;.ndronn i lo angiosarcoma.V h-<i` -I M" -92 17-21. 1975. 1706 JAMA. Oct 11. 1976-Vol 236, No. 15 Angiosarcoma--Ficchtner & Reyes R&S 131101