Document LJoJ8kbnw1YRYRqR0MYOaGRGq

Reprinted from the Archives of Environmental Health January 1971, Volume 22 Copyright 1971, American Medical Association Occupational Acroosteolysis I. An Epidemiological Study Bertram D. Dinman, MD; Warren A. Cook; Walter M. Whitehxmse, MD; and Harold J. Magnuson, MD, Ann Arbor, Mich; and Theodore Ditcheck, MD, Phoenix, Ariz UCC 100580 61 Occupational Acroosteolysis I. An Epidemiological Study Bertram D. Dininan, MD; Warren A. Cook; Walter M. Whitehouse, MD; and Harold J. Magnuson, MD, Ann Arbor, Mich; and Theodore Ditcheck, MD, Phoenix, Ariz An epidemiological study was performed cover ing 5,011 employees with 21,510 man-years ex perience in various phases of vinyl chloride (VC) and polyvinyl chloride (PVC) manufacturing in 32 plants throughout the United States and Can ada. The total number of definitive cases of acroosteolysis (AOL) was 25; 16 other indi viduals were under suspicion. This condition is clearly associated with the hand cleaning of polymerizers. Workers engaged in oither phases of VC or PVC manufacturing do not appear to be at risk of developing AOL. The importance of Ray naud's phenomenon as a concomitant of AOL is emphasized. Several statistical approaches for rapid medical survey are suggested. Acroosteoly sis appears to be a systemic rather than local disease. Presently, neither the etiological agent nor its portal of entry is known. w ITH THE emergence of a new clinical entity, occupational acroosteolysis (AOL),1 it became a manifest responsibility of the polyvinyl chloride (PVC) industry to define more clearly the character, extent, and pos sible causes of this disorder. Though indi vidual corporations might survey their em ployees, a statistically valid construction of the problem could not be developed without collation of the health experience of all PVC producers. To maintain the privacy of pro- Submitted for publication Feb 9, 1970; accepted March 24. From the Institute of Environmental and Indus trial Health (Dr. Dinman), the Department of Environmental and Industrial Health, School of Public Health (Mr. Cook), the Department of Ra diology, School of Medicine (Dr. Whitehouse), and the School of Public Health (Dr. Magnuson), Uni versity of Michigan, Ann Arbor, and the Depart ment of Radiology, Good Samaritan Hospital, Phoe nix, An/ (Dr. Ditcheck). Reprint requests to Institute of Environmental and Industrial Health, University of Michigan, Ann Arhor 4K1Q4 (Dr. Dinman), prietary information associated with details of manufacture, some disinterested group was needed to evaluate operating processes, as these might relate to the causation of AOL, Accordingly, the Manufacturing Che mists' Association requested the Institute of Industrial Health (Institute of Environmen tal and Industrial Health as of Jan 1, 1970) to conduct an epidemiological survey of the PVC industry in the United States and Canada to meet their responsibility to the workers and the public. The institute was requested to determine the extent of the problem among PVC work ers. By establishing a specific job association with AOL, it was hoped to determine where the risk of incurring AOL arose within the PVC process. It was important to determine whether AOL was associated only with the synthesis stage or if those handling the finished polymer also might be affected. The primacy of this consideration stems from the widespread use of the polymer in the fabrication of a multitude of manufac tured goods that enter general commerce. While the number of workers engaged in the chemical synthesis of PVC is relatively small, there are thousands of small manufac turers engaged in the fabrication of products derived from this versatile plastic. Further, it was hoped that the epidemio logical and industrial hygiene2 investigation might define more precisely how variations in manufacturing processes might relate to variations in the prevalence of AOL among PVC producers. Medical and industrial hy giene studies might furnish some clues to the precise etiological agent(s) as well as suggest preventive measures that might be applied in the industry. In addition, it was planned that a small, Arch Environ Health--Vo! 22, Jan IH7I UCC 100581 62 OCCUPATIONAL ACROOSTEOLYSIS--DINMAN ET AL NAMESoc. Sec. No. (5-13) RACE W N OtherAGESEX M F (14) ~ (15-16) (17) DEPT. NO. (18-19) COMPANY SERVICE DATE Month Year (20-21) (22-23) ARE YOU (RIGHT) (LEFT) HANDED? (24) Do Not Write in This Box Corp.___________ (1-2) Pint.___________ (3-4) TO ANSWER THE FOLLOWING QUESTIONS , PUT A CIRCLE AROUND YOUR ANSWER. FOR EXAMPLE: MONDAY IS THE DAY AFTER SUNDAY (YES) (NO) Have younoticed SUDDEN whiteness. blanching, or paleness of the fingers of your RIGHT HAND LEFT HAND (YES) (YES) (NO) 25 (NO) Have younoticed a sudden redness or blueness of the fingers of your RIGHT HAND LEFT HAND (YES) (YES) (NO) (NO) 26 IF YOU HAVE NOTICED ANY OF THE ABOVE, are these changes brought on by cold (YES) (NO) (DON'T KNOW) 27 HAVE YOUR HANDS OR FINGERS BEEN BOTHERED WITH ANY OF THE FOLLOWING: Tingling or pins-and-needles RIGHT HAND feeling LEFT HAND (YES) (YES) (NO) (NO) 28 Warm feeling in hands or fingers RIGHT HAND LEFT HAND (YES) (YES) (NO) (NO) 29 Cold feeling in hands or fingers RIGHT HAND (YES) (NO) LEFT HAND (YES) (NO) 30 Numbness, fingers fall asleep RIGHT HAND (YES) LEFT HAND (YES) (NO) (NO) 31 Stiffness, fullness, or tightness RIGHT HAND (YES) of fingers LEFT HAND (YES) (NO) 32 (NO) Frequent tenderness, soreness. or RIGHT HAND (YES) pain at fingertips LEFT HAND (YES) (NO) (NO) 33 Arch Environ Health--Vol 22, Jan 1971 OCCUPATIONAL ACROOSTEOLYStS--DINMAN ET AL 63 Have you noticed any of the following changes of the skin of ' your hands, wrists, or arms: Thickening of the skin 111 (YES) (NO) 34 *If yes, is it on (HANDS) (WRISTS) (ARMS) White, raised patches of skin * (YES ) (NO) 35 *If yes, is it on (HANDS) (WRISTS) (ARMS) Whitish pimples -(YES ) (NO) 36 *If yes, is it on (HANDS) ( WRISTS) (ARMS) Shiny skin patches * (YES ) (NO) 37 *If yes, is it on (HANDS) ( WRISTS) (ARMS) Small, raised areas, or lumps less than 1/2 inch around '(YES ) (NO) (NOT including warts) *If yes, is it on (HANDS) ( WRISTS) 38 (ARMS) Have you been bothered by these fingernail troubles: Pits or scars of the fingernails (YES) (NO) Ridges or cracking of the fingernails (YES) (NO) 39 40 Have you used hand-held power tools or air-driven tools at work in the past six years : (DAILY) (ONCE A WEEK) (LESS THAN ONCE A WEEK) (NEVER) 41 i Have you ever broken, fractured, or } crushed a finger? (YES) (NO) If yes, was it the (RIGHT) 42 or (LEFT) hand? Which finger was injured (Thumb) (Index) (Middle) (Ring) (Little) (Thumb) (Index) (Middle) (Ring) (Little) Arch Environ Health--Vol 22, Jan 1971 \}CC 100583 64 OCCUPATIONAL ACROOSTEOLYSIS--DINMAN ET AL selected sample of workers affected with AOL be studied in the clinical research unit of the University of Michigan Medical Center.3 Previously, Rousch and co-workers4 performed an extensive clinical evaluation of a group of such workers. Subsequent ex aminations here were designed to determine whether any new clinical features had de veloped since those prior studies. Polyvinyl Chloride Manufacturing: A Brief Review A more definitive description of the proc ess whereby PVC is synthesized, finished, and compounded is reported elsewhere in this issue.2 A brief review of the operation is appropriate to the premises underlying the epidemiological considerations. Vinyl chloride (VC), together with water, peroxide catalysts, and other additives, are placed in a reactor where polymerization is completed after ten to 16 hours. Unreacted VC is exhausted from the reaction vessel and the suspension or emulsion of the plas tic transferred to equipment for dewatering and drying. Following ventilation of the re actor, one or two workers enter the vessel to remove the residue from its surfaces, using putty knives or chisels. Sixteen plants, in addition to synthesiz ing, also compound PVC while five plants are engaged only in compounding the resin. Methods Survey Techniques.--Nineteen corporations known to be engaged in the polymerization of VC were assessed. The workers who served as the study populations were employed in 32 plants in the United States and Canada. All employees who worked at the time of the survey, or in the past, in any capacity in plants manufacturing VC. PVC, or in compounding such resin were surveyed. In multiple-product plants, all workers having any possible expo sure to the PVC process were studied. This population underwent medical survey, work history questionnaires, and hand x-ray filming. The self-completion medical questionnaire was designed to probe for signs of Raynaud's phenomenon or symptoms associated with pe ripheral vascular insufficiency. In addition, questions were posed to disclose skin or dermal appendage abnormalities, use of power tools, and hand-injury history. X-ray films of both hands were requested from the cooperating plants; specific directions concerning film type, exposure factors, tech nique, and identification were sent to each plant. The x-ray films, when forwarded to the institute, were coded and transmitted without identification to two radiologists for indepen dent interpretation. The findings were recorded on a specifically designed checklist which was then returned for statistical collation. Hand x-ray films that served as controls were drawn from 2,407 men over age 18; this represented the entire male population of the town of Tecumseh, Mich. Work history questionnaires were completed by each individual employee or by the person nel department at each plant or both. These forms recorded work assignments between Jan I, 1961, through Dec 31, 1966; each worker noted the first and last month of assignment for each specific assignment during that period. At each plant, a team of investigators converted local job designations to standard job titles. These are listed in the paper on the industrial hygiene aspects.2 Pilot Run In order to determine the reliability of survey techniques and recording methodo logical patterns, a pilot run was carried out at one large plant. Following this, minor modification of procedures and forms was accomplished. Epidemiological Design.--Acroosteolysis appeared to consist of characteristic x-ray film abnormalities and Raynaud's phenome non. Since the latter would be expected to occur only infrequently in a population not at risk of developing AOL, it could be used as a prime indicator of the existence of AOL. Thus, we hoped to establish an initial dis tinction between two populations, ie, one having AOL and the other not having it. Comparison of prevalence of defects with the Tecumseh control population of 2,407 men suggested that all defects occurred in the PVC population with a greater frequen cy than in the control population (Table 1). However, the significance of this finding was clouded by an unavoidable bias which mini mized the value of the Tecumseh population as a control group, ie, identification marks which could not be readily obliterated were present on these control films. Furthermore, as is apparent from previous discussions of the x-ray film abnormalities, linear radiolucent defects could be confused with old trauma, the absence of a tuft might Arch Environ Health--Vol 22, Jan 1971 OCCUPATIONAL ACROOSTEOLYS1S--DINMAN ET AL 65 follow amputation, or short ened phalanges could he Table 1.--Prevalence of X~Ray Film Delects in Control and PVC Populations consistent with a congenital defect. Defects such as scle Control, PVC Population, N 2,407 N = 5,011 rosis of the shafts, widening of the tufts, and soft-tissue Marginal defect Defect No. No. 1 44 % 1.83 No. 321 % 6.41 clubbing were each suscepti Marginal defect, with residual ble to subjective reader fragment 2 5 0.21 35 0.70 error. Furthermore, each of these defects could arise from other bony disease or injury. Even multiple finger Transverse defect Transverse defect, with distal fragmentation Total resorption of distal portion phalanx 3 4 5 6 0.25 0 2 0.08 40 0.80 21 Q.42 9 0.18 involvement could be con Sclerosis, with recatcification 6 1 0.004 15 0.30 sistent with these etiological Shortening 7 11 0.46 63 1.26 possibilities. Because these problems did exist, signif icant divergencies in x-ray film interpretations were Widening of tuft Widening of shaft Soft-tissue clubbing Total 8 4 0.17 43 0.86 9 1 0.004 7 0.14 1G 2 0.08 40 0.80 55 76 3.16 594 11.87 found (Table 2). Because of these complexities, for the pur reported to occur in conjunction with AOL, poses of first appraisal, marginal defects of those x-ray film abnormalities unrelated to the phalangeal tuft (defect 1) were not con AOL were grouped as defect 11. sidered as indicative of AOL unless both Combining the x-ray film findings and roentgenologists independently concurred as Ravnflnd'sr~several diagnostic criteria were to the presence of this abnormality. The set up (Table 3). The diagnostic criteria for existence of marginal defects with residual category 0 consisted of essentially normal fragment(s) (defect 2), transverse defects workers while category 1 included individu (defect 3), transverse defect with distal frag als having x-ray film abnormalities which by mentation (defect 4), and total resorption of history or radiographic evaluation or both distal portion of phalanx (defect 5) were were found to be consistent with old trauma, considered as possibly diagnostic of the disease other than AOL, congenital or famil bony defect of AOL if found by only one ial defects, etc. The history of persons in radiologist and if other causes for these diagnostic category 3 disclosed cold-induced defects could not be eliminated. Every man finger blanching unassociated with x-ray with any of these five abberations was fol film abnormalities, clearly meeting our crite lowed up subsequent to the field survey. If ria for AOL. Category 4 consisted of persons there was a history of trauma involving the without Raynaud's phenomenon and sugges involved finger or if the appearance of these tive, though not diagnostic, x-ray films for x-ray film lesions was clearly consistent with AOL. Category 5 included those with classi trauma, these individuals were eliminated cal AOL x-ray film defects (2 to 5 and 1, 6, from suspicion. and 7 without and with concurrence, respec Other x-ray film defects of an even less tively) and Raynaud's phenomenon. Cate specific nature are known to be associated gory 6 had typical AOL bone abnormalities with the later stages of AOL but are not in (as diagnosed in category 5) but without themselves considered diagnostic. These x- Raynaud-positive replies to the medical ray film findings included sclerosis of the questionnaire. phalanges (defect 7), widening of the tuft Computational and Statistical Analysis. (defect 8) or of the shaft (defect 9), and --Medical and work questionnaires and x- soft-tissue clubbing (defect 10). Finally, we ray film reports were checked for complete looked with suspicion upon any bony defects ness. They were then coded, verified, involving multiple fingers regardless of punched, and stored in an electronic com whether they were AOL-specific (ie, defects puter. Quality control checking of the punch 2 to 5). cards, using a 25% sample, revealed an er In addition to those defects previously ror of 0.1% among these cards. Arch Environ Health--Vol 22, Jan 1971 66 OCCUPATIONAL ACROOSTEOLYSIS--DINMAN ET AL Tabie 2.--Agreement and Number of Fingers Showing X-Ray Film Detects In 1,302* PVC Workers Marginal defect Defect No. 1 Marginal defect. with residual fragment Transverse defect Transverse defect, with distal fragmentation Total resportion of distal portion phalanx Sclerosis, with recalcification Shortening 2 3 4 5 6 7 Shortening 7 Widening of tuft Widening of shaft Soft-tissue clubbing Totals 8 9 10 No* of Fingers Involved 1 2345678 Agree 30 3 0 1 0 0 0 -- Disagree 237 30 11 4 3 1 1 -- Total 267 33 11 5 3 1 1 -- Agree 6 1 ------------ Disagree 27 1 Total 33 2 Agree 9 0-- 2 0 0---- Disagree 25 1 -- 0 1 1 -- -- Total 34 1 -- 2 1 1 -- -- Agree Disagree Total 1_ 17 -- 18 -- 1 1 ---------- 1 Agree Disagree Total Agree Disagree Total 21 51 72 0 1 0 1 -- 0-- 9 1 2 0-- 1 ---- 9 2 2 1 -- 1---- Agree 6 2 1 0-- 1 -- 1 Disagree 39 6 1 1 -- 1 -- 2 Total 45 8 2 1 -- 2 -- 3 Agree 6 2 1 0-- 1 -- 1 Disagree 39 6 1 1 -- 1 -- 2 Total 45 8 2 1 -- 2 -- 3 Agree 01 ---- 0-- 0 Disagree 35 3 -- -- -- 2 -- 2 Total 35 4 -- -- -- 2 -- 2 Agree Disagree Total 0 0-- 0 0 0---- 3 1-- 1 1 1---- 3 1 -- 1 1 1---- Agree 3 0-- 0 0 0 0-- Disagree 26 1 -- 2 2 1 2 -- Total 29 1 -- 2 2 1 2 -- Agree 57 9 2 4 Disagree 423 45 15 8 Total 480 54 17 12 02 79 7 11 0 3 3 1 4 5 9 10 ---- -- ---- ---- --1 --0 --1 ---- -- ---- ---- 10 11 21 10 11 21 ---- ---- ---- ---- ---- ---- 1-- 1-- 2-- 21 21 42 Negative for Specific % Total Defect Agree 34 287 321 981 10.6 7 28 1,267 20.0 35 12 28 1,262 30.0 40 2 18 1,281 20 9.5 3 6 1,293 33.3 9 2 18 1,287 13.3 20 12 52 1,238 18.9 64 12 52 1,238 18.9 64 1 42 1,259 43 2.3 0 7 1,259 7 2.3 5 35 1,295 40 78 517 595* * The difference between these two figures and those in the defect 11 category, ie, other defects, is not shown here* Frequently, workers performed several different duties during a period spent on one assignment. Where this existed, an estimate was made at each plant of the fraction of time spent in each segment of the job assignment. Since the fraction 1/12 was used for each month a worker spent full time in various jobs during the year, it was convenient to use this fraction of 12ths where his daily or weekly work was divided between two or more jobs. Thus, if a worker spent one half his time performing one spe cific component of his assignment, the value of 6/12 was assigned to our standard job title for that time period on assignment. The remaining 50%, or 6/12, of his time was distributed over one or more applicable job components during the period of assignment. By this method, job experi ence on each of the standard jobs could be computed in terms of months or fractions thereof. The computer was utilized to perform col lation and statistical manipulations of data. Clinical Evaluation of Positive Cases.-- Four employees with active disease were admitted to the Clinical Research Unit of the University of Michigan Medical Center. Extensive clinical, laboratory, and physio logical tests and radiographic, isotopic, nu tritional, and metabolic balance studies were performed. The findings are reported in detail.1 Results General Population Statistics.--The study population consisted of 5,011 workers, of Arch Environ Health--Vol 22, Jan 1971 I I UCC 100586 OCCUPATIONAL ACROOSTEOLYSIS--DINMAN ET AL 67 which 96.4% were men; 95.7% were white, 4.1% Table 3,-- Roentgenographic and Symptomatic Criteria tor Diagnostic Categories 0 to 6 Negro, and 0.2% other. The Hand X-Ray: Specific X-Ray Defect, mean age was 35.8 years, with the median and modal range being 35 to 39 years. These workers were em ployed by 19 corporations at 32 plants in the United States and Canada; of the Raynaud's Diagnostic Phenorn* Category enon 0-- 1 3+ 4-- 5 -I6-- Abnormal, Consistent With AOL * + + Consistent With AOL 1 2 3 4 5 6 7 8 9 10 -- ---------- "-- -- -- -- -- -- -- -- + + t -- -- -- -- + + t -t- + + -1" + + + + + + + + + -1- + -1- + + + 32 plants, all but six were engaged in PVC synthesis. This number of plants rep * Indicates only those lesions that may be seen in AOL (defects 8 to 11 though not diagnostic of AOL. t Single finger involved, no concurrence between readers, t No concurrence between readers. resented 92% of the PVC synthesis industry in this country. The population at Table 4.--Distribution of Diagnostic Categories Among Four Subpoputations With PVC Study Group risk represented a total of 21,510 man years. Population Analysis.--In order to construct several population groups for attri bute comparisons, we ap plied our diagnostic criteria for AOL (ie, x-ray film de fects 2 to 5 plus Raynaud's phenomenon). Using this approach, we found that the clearly defined cases of AOL were limited to seven plants. Total Population Oia nostic Categories 0-1 3* 4t St 65 7 plants with cases of AOL 1,673 1,598 41 9 22 3 3 plants with questionable AOL 910 893 14 3 0 0 17 plants without AOL 1,873 1,851 17 3 0 0 5 plants compounding only 557 554 2 1 0 0 Totals 5,013 4,896 74 16 22 3 * Raynaud's phenomenon only. t Questionable x-ray, no Raynaud's phenomenon. t X-rays characteristic of AOL, with Raynaud's phenomenon. X-rays characteristic of AOL, without Raynaud's phenomenon at time of survey. Using these same diagnostic criteria (Table 3), we detected another group but without Raynaud's phenomenon at time of 21 plants which had all employees in cate of survey). We determined by follow-up that gory 0 and 1, or less than a total of two indi of these three workers, two had manifested viduals in diagnostic category 3 (Raynaud's Raynaud's phenomenon during the 1950's without bone change! or diagnostic category prior to survey period. The third subject 4 ("AOL questionable" x-ray film without was re-x-rayed subsequent to the survey and Raynaud's) or both. To determine whether demonstrated progressive changes character there was a risk of developing AOL among istic of AOL; however, he has never noted those working with the finished polymer, we Raynaud's disorder. In view of this finding extracted for further analysis a third popu that 24 of 25 workers with severe AOL-type lation group, constituted by the five plants x-ray film defects had associated Raynaud's whose employees worked exclusively with phenomenon, we believe it is appropriate to the finished PVC-producing consumer prod link these two abberations in order de ucts. These three groupings left three indi finitively to diagnose AOL. Thus, using as vidual plants in which there remained a our criteria Raynaud's phenomenon or question of AOL's existence. The distribu finger defects qualified as above or both, 25 tion of diagnostic categories among these men were found to have clearly established four populations appears in Table 4. cases of AOL, The reason for requiring the coexistence of x-ray film abnormality and Raynaud's Intrapopulation phenomenon to establish the diagnosis of Differences AOL becomes apparent upon consideration of the three individuals in diagnostic catego Signs and Symptoms.--We determined ry 5 (ie, x-ray films characteristic of AOL the number of workers in each of the four Arch Environ Health--Vol 22, Jan 1971 UCC 100587 68 OCCUPATIONAL ACROOSTEOLYSIS--DJNMAN ET AL Table 5.--Intergroup Prevalence of Raynaud's Phenomenon Not Associated With AOL (Corrected for Male Populations Only) zero symptoms, the Figure) was different in populations Diagnostic Category 3 Alone % Male No. Population Diagnostic Categories 1*2 believed to be free of AOL as contrasted with those probably having cases of AOL (see above). This ob Plants with cases of AOL Plants without cases of AOL Plants with questionable cases of AOL Plants compounding only servation could be used to 41 2.50 1,582 1,623 screen a group of workers at 19 0.82 2,294 unknown risk of AOL de velopment by simply deter mining how many individu 14 1.70 817 831 als reported only a single 0.23 438 symptom. In order to deter mine whether more intensive Table 6.--Abnormal Number ot Cases of Raynaud's Phenomenon Occurring in Various Size PVC-Plant Populations* scrutiny of a group of 25 or more workers at unknown risk of developing AOL is Population Range 25-44 45-82 83-120 121-180 181-220 Abnormal No. of Raynaud's Phenomenon 3 or moref 4 or more 5 or more 6 or more 7 or more warranted, one can use a test based upon the normal ap proximation of the binomial, with _2 . P3 221-280 281-340 341-400 401-460 461-500 8 or more 9 or more 10 or more 11 or more 12 or more * p = o.oi. t Raynaud's, as defined, is cold, white fingers brought on by hand or body chilling or both. ir!> i 3 "3 7i where n is the number of employees questioned, where n is 24 and p is the fraction of population free of symp plant groups having one, two, three, etc, toms. Should the value obtained exceed 1.65, through 26 positive responses to questions further investigation of the population in concerning signs and symptoms. A graphical question should be considered. representation of the frequency with which The foregoing analyses do not relate to respondents noted bilateral hand symptoms any individual symptom sought by our med in those plants where AOL was believed to ical questionnaire. These statistical ap occur and in those free of AOL is provided proaches only considered the absence or in the Figure. presence of any one symptom. While these The one-tailed Kolmogorov-Smirnov (K- findings suggest the possibility of asking a S) two-sample test0 was applied to the two single question to screen a population at distributions. The same analysis was direct risk of developing AOL, is there any one ed to the compounding plant population question that might be used as a more perti AOL-free plant group (see above). The dis nent screening device? tribution of positive signs and symptoms Focusing once more upon Raynaud's phe were found similar in both groups. Accord nomenon, we found, by the use of factor ingly, for the "AOL-free" population in the analysis,0 that this peripheral vascular disor Figure, we have included the employees of der was significantly related to AOL. That the compounding plants. On the basis of is, those questions which appear to be con affirmative responses concerning the pres sistently answered positively as a group con ence of any symptoms, this test indicated stitute a cluster of symptoms or a factor. In that these two distributions did not stem addition, each symptom is associated with from a common population pool (P < 0.001). factors by scores called "loadings," which This K-S test indicates that the frequency weigh the associational strength of the fac of occurrence of the asymptomatic state (ie, tor to the question at hand, ie, the occur- Arch Environ Health--Vol 22, Jan 1971 OCCUPATIONAL ACROOSTEOLYSIS--DINMAN ET AL 69 ti--oI in rH i/i E rn o (N cn 5 CO cn c <C <x> CO > t-* toXo o o> in gz O') (N r-d o Cumulative frequency with which workers reported bilateral hand symptoms in plants with and without cases of AOL. rence of AOL. The most significant of these clusterings of symptoms was found to be, by factor analysis, questions 25, 26, and 27 (questionnaire form), ie, indicators of Ray naud's phenomenon. Accordingly, this con stitutes one reason for examining the preva lence of this particular symptom complex among any group at risk of AOL develop ment. This logically leads to an examination of the relative prevalence of Raynaud's phe nomenon in our various plant-population subgroupings. However, because Raynaud's phenomenon is more frequently associated with women and since that sex represented a relatively small part of the total study population, we first removed the female re spondents from our data pool. Applying analysis, we found that the prevalence of Raynaud's phenomenon was significantly different in the plants believed to be free of AOL (P < 0.01) vs those known to have AOL cases, while no difference was found between the "AOL-free" plants and the compounding locations (Table 5). If indeed the "AOL-free" plants represent a population at no significant risk of developing this disorder, one might infer that Raynaud's phenomenon may occur Arch Environ Health--Vol 22, Jan 1971 Vcc 100589 70 OCCUPATIONAL ACROOSTEOLYSIS--OINMAN ET AL Table 7.--Intergroup Differences Between Four Groups of PVC Plants Question No. 25 26 28 29 30 7 Plants With Cases of AOL, n ss 1,673 VS 17 Plants Without Cases of AOL, n =s 1.873 +* 4* -h + 7 Plants With Cases of AOL, n = 1,673 vs 5 Plants Compounding, n = 557 4-- - -- -- 31 432 33 -134 + 35 36 + 37 + 38 39 -- 40 - -- -I-- 4-- + -1- -- -- * Significant difference, P < 0,01, x2 analysis, t No significant difference, P > 0.01, x* analysis. 17 Plants Without Cases of AOL, n = 1,873 vs 5 Plants Compounding, n 7= 557 --t -- -- + -- -- -- -- -- - 3 Plants With Cases of 7 AOL, n = 910 vs 17 Plants With Cases of AOL, n = 1,673 -- - + -- + 4+ - 4* -- + - -- -- + 3 Plants With Cases Of 7 AOL, n =910 VS 17 Plants Without Cases of AOL, n = 1,873 -- -- -- -- -I- -- -- -- - among the general population in eight of 1,000 men. We are unaware of any epidemio logical data confirming this impression. Extending our screening approach, could one predict the number of cases of Ray naud's phenomenon which might indicate the need for further investigation? Since it appears that the general male population incidence of this circulatory disorder is less than 1% (Table 5), by application of the cumulative probability distribution of the binomial with P = 0.01, one can determine, for various sizes of plant population, the minimum number of positive responses in dicative of a possible AOL problem. Table 6 indicates the number of cases of Raynaud's phenomenon which would only rarely occur by chance (P < 0.01) in various size popu lations. All signs and symptoms believed to be associated with AOL were compared as to their intergroup prevalences (Table 7). There were considerable differences in the prevalence of most signs and symptoms be tween the plants with and without AOL while the compounding plants appeared not to be dissimilar from the population be lieved to be free of AOL. By contrast, the group about which doubt existed as to the occurrence of AOL appeared to be some what more like the AOL-free group than the AOL-affected population. Work History Analysis A detailed review of the 25 clearly estab lished cases of AOL revealed that 18 persons had worked in reactor cleaning at some time during the study; additionally, six others, by follow-up, were found to have worked at this task prior to the study. The one man who had no history of production-line polyclean ing was a laboratory technician. However, detailed review of his work history indicated experience with bench-scale reactor opera tion; this job included hand cleaning of small reactor vessels. Thus, the association of hand cleaning of reactors and AOL dis ease was striking. Because of job progres sion practices, practically all persons suf fering from AOL had worked at other PVC production jobs at one time or another. Since the lowest paying jobs, ie, reactor cleaning and bagging-packing, are given to newcomers, most production workers were given these jobs before progressing upward into the skilled jobs. The fact of this upward progression is underlined by work histories Arch Environ Health---Vol 22. Jan 1971 UCC 100590 OCCUPATIONAL ACROOSTEOLYSIS--DINMAN ET AL 71 Table 8.--Prevalence of AOL by Diagnostic Category and Reactor Cleaning Exposure (Job 17) Prevalence Definitive cases (ie, diagnostic categories 5 and 6) of AOL. 25 Total workers at risk, reactor cleaning, 1,047 Reporting Reactor Cleaning Experience, 1961-1966 Reporting Any Reactor Cleaning Experience, by Follow-up Questionable or no History, Reactor Cleaning 18* 1 case per 1.047 ~ 58 ~ workers 24 l case per 1,047 ~ 41 ~ workers 1 case per 1,047 workers Total man-years, reactor cleaning experience, 1,082 18* 1 case per 1.082 ~ 60 - man years 24 1 case per 1,082 45 "" rnan years 1 case per 1,082 man years Definitive and possible cases (diagnostic categories 3, 5, and 6), 41 Total workers at risk, reactor cleaning, 1,047 (18*+10*) 1 caseper 1.047 -- 37 -- workers Total man-years, reactor cleaning, 1,082 <1B*+10*) 1 case 1,082 39 per manyears Total PVC work force, male Men at risk, 4,751 Total man-years in PVC population, 21,510 28 1 4,751 " 170 " case per male employees in PVC study population 1 768 " case per man-years in PVC study population (28+6t) 1 4,751 TM 140 case per male employees in PVC Study population 34 1 21,510 633 case per man-years in PVC study peculation 7t 1 4,751 679 employees in PVC study population 71 21,510 3,073 man-years in PVC study population * By analysis of survey data. t Based on postsurvey follow-up; si* workers in diagnostic categories 5 and 6 were determined to have had reactor cleaning experience; similar data not available for diagnostic category 3 population. X Assumes that six workers in diagnostic category 3 had had no reactor cleaning experience. of men with AOL; only experience as bag ger-packer or reactor operator approached the frequency of experience of reactor clean ing. Reactor cleaning has a strong association with AOL. The prevalence of AOL in reac tor cleaners was one case per 72 workers at risk (total population at risk, 1,047) among all workers having a history of such job experience. For job categories larger than 500 men, the prevalence rate of AOL by job was greatest for bagger-packers (one case of AOL per 86 workers at risk out of a total population of 860) and reactor operators (one case per 69 at risk out of a total population of 893). On the basis of manmonths at risk, the prevalence of clear-cut AOL was even more strikingly associated with reactor cleaners; this job had a preva lence rate of one case per 721 man-months at risk (12,984 total man-months at risk). Among those jobs reporting 5,000 or more man-months at risk, no other job category approached this prevalence rate. The question might be asked whether plants with AOL differed from plants free of this disorder in regard to the duration of duty tours in reactor cleaning. Analysis did not indicate that men in the former gToup of plants worked longer tours on this high-risk job. In addition to the 25 clear-cut cases of AOL, there remained some question whether individuals demonstrating Raynaud's phe nomenon with nondescript or borderline x-ray films might have represented early or minimal cases of this disorder. Review of the x-ray films and medical questionnaires of the 74 men in diagnostic category 3 (ie, Raynaud's phenomenon) revealed 16 men who might have been manifesting an early phase or minimal degree of AOL. Eight of these men had marginal defects, about whose presence the radiologists could not agree, and two more had widening of the tuft and shaft, with soft-tissue clubbing (one of the latter subsequently developed multi ple osteolytic lesions). The other eight had Arch Environ Health--Val 22, Jan 1971 72 OCCUPATIONAL ACROOSTEOLYSIS--DINMAN ET AL three, five, six, or eight other symptoms of AOL on history but no bony abnormalities. Among these 16 individuals about whom there is suspicion of AOL, we found that ten of them reported experience in reactor cleaning. This reported experience in reactor cleaning was somewhat similar to the work history found among those men with clearcut AOL, ie, 18 men of the 25 cases. It would be of interest to determine the proba bility of fir ling that among 16 men chosen at random fi om the study population, ten or more would report such work experience. Among5,011 individuals surveyed, 1,047, or 21%, reported experience in this specific job. Considering this entire study population, the probability that within a random sample of 16 men we would find by chance ten or more reporting experience in reactor cleaning is only 0.00038 ( ie, 38 chances in 100,000). This determination is based upon application of the cumulative probability distribution, where It is apparent that time alone will tell whether these last eight individuals will de velop AOL, Nevertheless, because we know that at least one individual in this suspi cious group subsequently has developed clear-cut AOL, we believe that 25 cases un derestimates the total number of AOL cases in the population study. Therefore, we have considered that in addition to the 25 clearcut cases, these 16 might also be categorized in the AOL group for statistical purposes, providing a total number of 41. On this basis, the prevalence of cases in the general plant population of AOL is calculated in Table 8. Experience Among Compounding Work ers.--One of the reasons for performing this study was to determine whether fabricators handling PVO polymer risked development of AOL. None of the cases of clear-cut AOL occurred among this group of 1,257 employ ees despite there being 57,657 man-months experience. Among the 557 men working as compounders, there was only one case of Raynaud's phenomenon. The prevalence of Raynaud's phenomenon among compound ers was one case per 6,406 man-months experience while in reactor cleaners, this diagnosis occurred with a prevalence of one case per 721 man-months exposure. Compar ing those plants with cases of AOL with compounding plants by xs analysis, signif icant symptoms (vs) occurred considerably more frequently in the plants with cases (P < 0.01) K than in those compound ing only. In this group of compounders, there were no persons showing x-ray film defects of a questionable nature without Raynaud's phenomenon (diagnostic catego ry 4). In summary, it would appear that manipulation of the finished PVC resin is probably not associated with a risk of con tracting AOL. Comment The disease acroosteolysis appears to be clearly associated with hand cleaning of polymerizers. In those plants using high-pres sure water lances or solvents for cleanup of polymerizers, there were possible cases in only one. There appeared to be a correlation between the extent of degassing prior to entry into the reactor.3 At this time, work practices, rather than any one specific ingredient or combination in the process, appear to determine whether this disease will occur. In view of the low prevalence, the probability of the occurrence of AOL is correspondingly low in a small plant. However, smaller plants without cases at the time of the survey should not consider themselves immune to this problem. Indeed, several plants with relatively small populations do have this disease among their workers. An examination of the intimate details of their work practices might clarify these ambiguities. While plants with cases of AOL provided gloves for their reactor cleaners, we observed in consistent usage. Furthermore, while the standard procedures used to "air out" the reactor before entry are well designed to minimize exposure, these were frequently found, upon worker inquiry, to have been "short-cut." That host characteristics may be determi nant of the pathogenesis of AOL is suggest ed by the low prevalence of this condition, ie, idiosyncratic "sensitization" or "suscep tibility" must be seriously considered. It will be recalled- that a number of inade- :\rrh Eariraa Health -- Val 22. -tan 1921 UCC 100592 OCCUPATIONAL ACROOSTEOLYSIS--DINMAN ET AL 73 quately defined low molecular weight, par tially polymerized materials are believed to reside in the reactor scrapings. In this connection, the sensitizing propensities of dextranpolyvinyl pyrrolidine plasma extend ers should be recalled.7 While it is difficult to see how a high molecular weight (MW 10,000) compound could be absorbed percutaneously, the possibility of respiratory tract absorption and systemic distribution must remain open. In addition, because this ap pears to be a systemic disease, we are more inclined toward the pulmonary portal of entry rather than the transdermal route of absorption. The possibility might be considered that this disease is caused by several agents, each one producing a specific component of the picture. In this connection, there is existent a report of Plesitzer and co-workers8 which notes the occurrence of Raynaud's phenome non resulting from VC exposure. Is it possi ble that the Raynaud's component of AOL stems from residual, unreacted VC associ ated with the PVC remaining on the reactor walls? Contrariwise, the bone changes in animals produced by VC monomer by Viola" did not appear to be associated with vasomotor lability. To answer some of these questions, more data are required. A more complete chemi cal characterization of the residual PVC which remains on the reactor walls is essen tial. Animal exposure to these various possi ble components are desirable. Finally, medical surveillance for symp toms of Raynaud's phenomenon and period ic hand x-ray films appear indicated for the present. It may transpire, as more men and man-years experience accumulate, that those manufacturers now free of AOL among their workers will, as time goes by, see develop ment of some cases. This caution particular ly applies to employees in the PVC synthe sis area. It must, he emphasized that the limitations of any retrospective epidemiolog ical study make impossible any definitive prediction of future consequences of PVC manufacture. Since such studies can only describe events of the past and up to instant of the survey, the recommendations as to work practices and medical surveillance be come especially pertinent for the safety of these workers. This investigation was supported by a grant from the Manufacturing Chemists' Association. A portion of the support for use of computational facilities was made available by grant 1-P07-RR 00417-01 from the Department of Health, Education, and Welfare. Mary E. Patno, PhD, and Leigh S. Whitlock, PhD, aided in the statistical analyses, and Horace J. Dodge, MD, provided the hand x-ray films that served as controls. References 1. Wilson RE, McCormick WE, Tatum CK, et al; Occupational acroosteolysis. JAMA 201:577-581, 1967. 2. Cook WA, Giever PM, Dinman BO, et al: Occupational acroosteolysis: II. An industrial hy giene study. Arch Environ Health 22:74-82, 1971. 3. Dodson VN, Dinman BO, Whitehouse WM, et al: Occupational acroosteolysis: III. A clinical study. Arch Environ Health 22:83-91, 1971. 4. Rousch G Jr: Polycleaner's disease: Clinical aspects. Head before the 22nd annual meeting of the American Academy of Occupational Medicine, Cin cinnati, 1970. 5. Siegel S: Nonparametric Statistics. New York, McGraw-Hill Book Co Tne, 1956, pp 127-136. 6. Harmon HH: Factor analysis, in Mathematical Methods for Digital Computers. New York, John Wiley & Sons Inc, 1959, chap 18. 7. Thompson WL, Walton HP: Elevation of plas ma histamine levels in the dog following adminis tration of muscle relaxants, opiates and macromolecular polymers. J Pharmacol Exp Ther 143:131-136, 1964. 8. Danishevski SL, Egorov MN: Gig Tr Prof Zabol 26:9, 1961. 9. Viola PL: Pathology of vinylchloride, in Pro ceedings of the 16th International Congress on Occupational Health, Tokyo, 1969, to be published. Arch Environ Health--VoJ 22, Jan 1971 Printed and Published in the United States of America Reprinted from the Archives of Environmental Health January 1971, Volume 22 Copyright 1971, American Medical Association Occupational Acroosteolysis II. An Industrial Hygiene Study Warren A. Cook; Paul M. Giever, MPH; Bertram. D. Dinman, MD, ScD; and Harold J, Magnuson, MD, MPH, Ann Arbor, Mich UCC 100594 74 Occupational Acroosteolysis II. An Industrial Hygiene Study Warren A. Cook; Paul M. Giever, MPH; Bertram D. Dinman, MD, ScD; and Harold J, Magnuson, MD, MPH, Ann Arbor, Mich An industrial hygiene survey was conducted in respect to occupational acroosteolysis (AOL) in 32 plants engaged in production and compound ing of polyvinyl chloride. This survey was limited to inspection of facilities, interrogation of tech nical personnel, and statistical analyses. All manufacturing processes and chemicals utilized were explored. A major objective was the iden tification of variables In processes and chemicals that were peculiar to those plants affording known cases of AOL. No such processes or chemicals were found. The survey brought to light a num ber of situations suspected of relationship to the precise cause of AOL, some of which require additional exploration. Most positive of these was the finding that essentially ail of the persons who developed AOL had served as reactor cleaners using manual methods. PoLYVINYL chloride (PVC) has been in production and widespread use for more than 30 years. An extensive account of the chemistry, methods of production, materials utilized, properties, and uses of the PVC resins was recently published in a series of technical articles by Albright.1-5 Uses in clude such diverse products as floor tiles, films for food wrappings, sheeting for show er curtains, phonographic records, and coat ings for cables. As an indication of the magnitude of the PVC industry, the average operating capaci ty in this country for the year 1969 was about 3 billion lb.8 Much PVC is produced in other countries. Even with this extensive production involving thousands of operators, the occurrence of occupational acroosteolysis (AOL) among these workers was not recog nized until about 1961.7 Submitted for publication Feb 9, 1970; accepted April 16, 1970. From the Institute of Environmental and Indus trial Health, University of Michigan, Ann Arbor. Read before the American Industrial Hygiene Conference, Denver, May 15, 1969. Reprint requests to School of Public Health, University of Michigan, Ann Arbor, Mich 48104 (Dr. Dinman). The clinical and epidemiological aspects of this condition are discussed in separate papers in this issue of the archives.8'0 Industrial Hygiene Survey Surveys were made of 32 plants by the authors of this paper, most of the plants being visited by a team of one physician and one industrial hygienist. Twenty-six of these were polymer production plants. Vinyl chloride mon omer production plants and plants engaged in compounding and fabricating of the polymer ized resin were a part of the manufacturing complex at a number of these locations. Five plants conducted compounding and fabricating operations only, and one plant produced only the vinyl chloride monomer. These plants were located throughout the United States and one in Canada. This survey was designed to obtain informa tion on equipment used, procedures followed, materials involved, extent of exposure to these, and names and descriptions of jobs in order to compare practices in the plants where cases of AOL occurred with those in plants experienc ing no cases. The extent of exposures was estimated on the basis of general observation, a knowledge of the physical, chemical, and physiological prop- Arch Environ Health--Vol 22. Jan 1971 Fig 1,--Flow diagram for production of PVC from vinyl chloride monomer. erties of the materials, and discussion with technically informed persons at the plants. By setting up three broad categories of duration and of relative severity of the exposure, semiquantitative estimates were made. Since mate rials responsible for causing AOL might be absorbed either through the respiratory tract or through the skin or both, such estimates were made for each of these avenues of absorption. Production Operations Vinyl Chloride Production.--The two processes for production of vinyl chloride described by Albright12 were found to be in current use. One of these involves the reac tion of hydrogen chloride with acetylene in the presence of a mercuric chloride catalyst. In the other process, 1, 2-dichloroethane is pyrolyzed in the presence of chlorine as a catalyst for the production of vinyl chloride and HC1. In some plants, both methods were used, with the HCl produced in the latter of these processes being used in the former. Polyvinyl Chloride Production.--Poly vinyl chloride can be produced in accord ance with four processes, each with many variations among different plants. The final polymer may have a molecular weight vary ing from 50,000 to 150,000, depending on conditions under which it is produced. The production processes are designated as sus pension, emulsion, bulk (or nonsolvent), and solvent. A typical procedure is present ed in the flow-sheet (Fig 1). The essential steps of these operations are outlined in this paper. For more complete details, reference may be made to the papers by Albright.s-5 In the widely used suspension process, deion ized water is first introduced into a large auto clave, usually glass-lined, followed by a wide variety of additives. These include suspending agents such as polyvinyl alcohol, methyl cellu lose, and gelatin, emulsifying agents such as sulfonated oils, surfactants such as sorbitan monolaurate, and buffers such as sodium bicar bonate. Trichloroethylene may be added to pro duce the lower molecular weight polymers. Cat alysts are added to initiate the polymerization. These are nearly all organic peroxides, with lauroyl peroxide and diisopropylperoxydicarbonate (IPP) being used at the greater number of plants. The reactor is then closed, evacuated to remove oxygen, and vinyl chlo ride, normally a gas, is piped in as a liquid under pressure from the tank farm or the vinyl chloride plant. In some plants, the catalyst is added to a charge tank from which it is intro duced into the closed reactor following addition of the vinyl chloride. The mixture is agitated with rotating propel ler blades, and steam initially is directed into the reactor jacket to raise the temperature to 50 or 55 C. Since the polymerization reaction is exothermic, cooling water must be run into the jacket after the reaction has begun to avoid excessive heating. After ten to 16 hours, the polymerization is essentially complete, and the reactor is brought down to atmospheric pres sure. In most plants, it is placed under vacuum for removal of unpolymerized vinyl chloride monomer and transfer to the monomer recov ery area. The suspension of PVC is then dis charged into a blowdown tank below the reac tor. This tank is placed under vacuum for transfer of remaining vinyl chloride to the recovery area. In some plants, the blowdown tank is eliminated. The general appearance of the charging floor of a |>olymer plant is shown Arch Ennirtm Health--Vol 22, Jan ]!)7l 7G OCCUPATIONAL ACROOSTEOLYSIS--COOK ET AL Fig 2.--Charging floor of PVC production plant showing top portion of reactors. in Fig 2. The reactors extend for ten feet or so Water-soluble initiators such as ammonium down to the floor below. persulfate and hydrogen peroxide are used as The polymer slurry is pumped to blend catalysts. Since the fine polymer particles can tanks, then to centrifuges for removal of much not be readily separated from the emulsion by 1 of the water. The dewatered polymer is passed centrifuging or filtering, a spray dryer is em i through a rotary dryer with heated air, the ployed in this process. temperature of the polymer being kept below The bulk process is similar to suspension 140 F to avoid degradation. The dried granular polymerization but is characterized by the pre polymer is pneumatically conveyed into a cy cipitation of the polymer from the liquid phase, clone separator, with finer particles being re as about 10% of the vinyl chloride becomes moved from the air stream by means of a polymerized. filter-type collector. The polymer particles are Solution polymerization is much the same as sized by vibrating screens, then bagged for suspension polymerization, but no water is in transfer to compounding and fabricating plants troduced, Instead, the polymerization is con or shipped in bulk. In some plants, the polymer ducted in such solvents as cyclohexane, acetone, is stored in bins prior to being shipped. or n-butane. In the emulsion process, the equipment and process is essentially the same as in the suspen sion process up to the drying operations. In order to produce the finer polymer particles characteristic of the emulsion process, a wide variety of emulsifying agents is used. These may be soaps such as ammonium or sodium laurate, surfactants such as salts of naphthalene sulfonic acid, or various synthetic detergents. Copolymer and Terpolymer Production. --In addition to the production of the homopolymer PVC, many of the plants also produce copolymer and terpolymers, includ ing vinyl chloride. The greater percentage of the monomer tends to be the vinyl chloride in all of these. The copolymer vinyl chlor ide-vinyl acetate is in highest production Arrh Environ Health--Va! 22, rlnn l`)71 UCC 100597 OCCUPATIONAL ACROOSTEOLYSIS--COOK ET AL 77 among this group but is a Table 1,--Absolute Pressure-Time Values minor product in compari son to the homopolymer. Other copolymers are vinyl chloride-vinylidene chloride and vinyl chloride-Iauryl- Absolute Pressure-Time Values Below 10 Plant Code No. of No. of Reactor Expected Cleaners AOL Cases Absolute Pressure-Time Values of 10 or Above Plant Code No. of No. Of Reactor Expected Cleaners AOL Cases vinyl ether. An example of a terpolymer is vinyl chloride-n-butyl maleate-dibutyl maleate. EE Plants With AOL Cases 31 0.8 B E F R 216 165 16 39 5-8 4.5 0.4 1.1 Processes, additives, and Plants With Possible AOL Cases CC 32 0.9 catalysts essentially identi Totals 63 1.7 436 11.8 cal to those used for the homopolymer are employed in the production of the co polymers and terpolymers with the addition to the re action mixture of the addi tional monomers. Vinyl Chloride Recovery.-- The unpolymerized vinyl chloride removed from the P Q V w Totals DD Totals 30 22 91 50 193 Plants With No AOL Cases 0.8 D 0.6 H 2.5 L 1.4 S T Z AA 5.3 27 26 28 17 9 17 8 132 Plants Started in Operation in 1965 or 1966 21 0.6 C 54 J 13 21 0.6 67 0.7 0.7 0.8 0.5 0.2 0.5 0.2 3.6 1.5 0.4 1.9 batch following completion of polymerization is conveyed to a vinyl chloride recovery area Table 2.--Classification of Standard Job Titles by the suction pump that Job Title Code Job Title Code places the reactor and the stripping tank under reduced pressure. Here the vinyl chlo ride gas that is removed is placed under pressure to con dense it together with water vapor to a liquid. The water and the vinyl chloride are im miscible, permitting their sep aration. In some plants, an inhibitor, usually phenol, is added to the vinyl chloride Additive makeup (other 1 than catalyst) Bagging and shipping operator 2 Catalyst makeup man 3 Clerical, other office, staff 4 Compounding and fabricating 5 Dryer operator 6 General laborer, custodial 7 Laboratory supervisor S Laboratory technician 9 Maintenance, electrical 10 Maintenance, mechanical 11 Monomer maintenance operator 12 Monomer operator 13 Monomer recovery operator Monomer supervisor Plant guard, security officer Reactor cleaner Reactor cleaner, solvent Reactor operator Slurry blender Supervisor, polymer Tank farm operator Utilities area operator (water treatment, etc) Warehouse operator Transfer operator Other than PVC 14 15 16 17 18 19 20 21 22 23 24 25 30 to prevent spontaneous poly merization during its distillation, the method used for its purification. The recovered vinyl chloride, with the higher boiling phenol remain due on the agitator blades may require hammer and chisel for removal and sometimes even axes. ing in the still, is then pumped to storage for reuse. Reactor Cleaning Operations.--After the polymer has been discharged from the reactor Four plants use water jets with pressures of from 300 to 3,000 Ib/sq in for the cleaning operation. One of these used water jets only in one of two production buildings, as the struc- and the pressure brought to atmospheric, the reactor is aerated. This is accomplished by dropping a flexible duct connected to a portable tional features interfered with movement of the portable high-pressure water unit about the building. Three other plants have been using blower to the bottom of the reactor, with dis solvents for reactor cleaning. One of these charge normally through the open manhole at started to use this method late in 1964, a the top of the reactor into the charging-floor second in 1966, and the third from the begin area. The loose residue is washed out with ning of plant operation in 1966. The reactor is water, after which one or two men enter the filled with the solvent which must be heated reactor to remove the skin of polymer remain and agitated for a prolonged period, as the ing on the reactor surfaces. The usual method PVC is difficulty soluble. is by manual scraping with putty knives. Resi Polyvinyl Chloride Compounding.--Com- Arrh Environ Health--Vol 22, Jan 1971 1 1 i ! ! - v, A ^ 78 OCCUPATIONAL ACROOSTEOLYSIS--COOK ET AL Table 3.--Magnitude of Dosage Expressed as Single Concept at a weekly, monthly, or Seventy/Duration Exposure Parameters! other period. A classification of 26 job Magnitude* of Dosage Slight 1 2 3 4 5 Less Than 1 hr/Week Negligible Moderate Moderate 1-8 hr/Week None Negligible None Negligible More Than 8 hr/Week None None Negligible None None titles for the statistical pur poses of this study was pre pared (Table 1). These job titles corresponded to com parable operations in all the plants and to a common 6 Intermediate 7 8 9 10 11 12 Moderate Appreciable Appreciable Appreciable None Negligible Moderate Moderate Negligible None None Negligible None Negligible Moderate group of types of exposures to production materials. Ev ery job in each of the plants was given its fraction of time related to the job title as defined for the study. De Major 13 14 15 16 17 18 Appreciable Appreciable Moderate Appreciable Appreciable Appreciable None Moderate None Negligible Moderate Appreciable * Expressed in order of increasing dosage received from exposure, which is the summation of the three columns to the right. t Severity of exposure is estimated as none, negligible, moderate, or appreciable for indicated durations. tails of the manner in which this was done are discussed in the accompanying paper on epidemiological aspects of AOL.s Job Progression.--Often the first job assignment of the new employee is that of reactor cleaner. Unless there pounding consists of mixing PVC with a multi are many reactors, this work does not require plicity of plasticizers, antioxidants, pigments, the entire time of the employee. Part of his emulsifiers, stabilizers, and other special com pounds in ribbon blenders and mills. The mix tures are then passed through roll mills or mixers, such as are used in the rubber industry, where the temperature may rise sufficiently to vaporize some of the plasticizer and other vola tile constituents. The compounded polymer is then extruded through rolls or dies to form time may be spent in bagging or shipping finished polymer or in the handling of cata lysts and other additives. As job openings arise and with accretion of seniority, this employee may progress to higher skilled jobs such as reactor, monomer recovery or dryer operator, or any combina sheets, films, or pellets. These are used in tion of these at any one time. fabricating operations. In plants producing other products in ad Job Classification dition to vinyl chloride polymers, employees may transfer between these product areas Many variables were introduced in the either as full time for a period of months or plants surveyed and even in different units dividing their time within a given week or of the same plant in regard to nearly every even day. The latter occurs particularly phase of the operations. Investigation of among maintenance personnel. unit jobs in the plants disclosed a wide Where production of vinyl chloride is part assortment of job titles. In fact, a job title as of the operation, operators may rotate be used by the technical supervisors often dif tween jobs in the production of the mono fered from that for the same job in the mer and those associated with the produc same plant as listed by the personnel de tion of the polymer. partment. The length of time spent at various jobs is Furthermore, a job assignment in one a function of production needs, economic plant might include a different group of unit conditions, plant growth rates, and activity operations than in another plant. Also, the of other product lines. proportion of time devoted to the several unit operations often differed. To compound Occupational Exposures to Materials the difficulties of classification, workers Extent of Exposures.--Information was would frequently rotate through several jobs obtained not only on each of the materials Arch Environ Health--Vol. 22, Jan 1971 UCC 100599 OCCUPATIONAL ACROOSTBOLYSIS--COOK ET AL 79 used or produced but also on the extent of exposure to each. Since this is a function of both duration and severity, each of these factors was estimated. The duration of expo sure was recorded as less than one hour a week, from one to eight hours a week, or more than eight hours a week; the severity as negligible, moderate, or appreciable. Since it was not known whether the materi als might be absorbed through the re spiratory tract, through the skin, or both, the extent of exposure as related to each of these avenues of absorption was recorded. For statistical purposes, it was desirable to combine these two factors of exposure into a single dose value. Although admitted ly the effect of exposure to materials with differing physiological responses would vary somewhat with different combinations of de grees of duration and severity, the 18 possi ble combinations of duration and severity were listed in increasing order of dosage and classified in three groups of dose values des ignated as slight, intermediate, or major. These combinations are presented in Table 2. Materials Involved.--A total of 227 different materials were used or produced in the forma tion of PVC, the copolymer vinyl chloride-vinyl acetate, and other copolymers and terpolymers incorporating vinyl chloride. These materials included monomers, polymers, catalysts, disper sants, inorganic compounds for coalescence reg ulation, emulsifying agents, surfactants, and or ganic solvents (including hydrocarbons, both aliphatic and aromatic, ketones, esters, and chlorinated hydrocarbons). This total of 227 materials does not include the many more used in the compounding and fabricating plants. A computer analysis of exposures to these materials was undertaken, but up to the pres ent time, this has disclosed no constellation of materials introduced into the batch at the plants exhibiting cases of AOL that was not used in plants without cases. However, special consideration was given to certain suspect ma terials and also to certain details of procedure. Vinyl Chloride Exposures.--In evaluating the role of vinyl chloride as a possible in fluence in the causation of AOL, the esti mates of the extent of the exposure of re actor cleaners and reactor operators were reviewed. Since the authors conducted no air analyses, they had to judge the presence of vinyl chloride in the charging floor area principally on the basis of general ventila- Table 4.--Use of Inhibitors With Vinyl Chloride Use of Inhibitors Plants with AOL cases or possible cases Plants without AOL cases In Virgin Vinyl Chloride In Vinyl Chloride Recovery 22 34 tion provided and odor observed. The odor of vinyl chloride could be detected in the immediate vicinity of reactors as they were first opened and also when they were first being vented into the charging room area. Air analyses conducted by others have shown vinyl chloride concentrations in the reactor prior to ventilating to be in the order of 3,000 ppm. The lower limit of detection of vinyl chloride by odor is accept ed as 400 ppm. Except under the foregoing conditions, the odor of vinyl chloride was not observed in the general room air. The reactor cleaner enters the reactor af ter the aeration has reduced the vinyl chlor ide exposure to what is considered satisfac tory limits, normally 15 or 20 minutes. The test is either by sniffing at the manhole opening or by use of a flammable vapor indicator. With the lower explosive limit of vinyl chloride at 4%, it requires 400 ppm for the usual type of flammable vapor indi cator to give a positive reading. Where air analyses have been made in a small number of plants with more sensitive instruments, such as the gas chromatograph, it has been found that the vinyl chloride concentration inside the reactor during scraping operations tends to be below 100 ppm and usually about 50 ppm. With the residue containing some unpolymerized vi nyl chloride, small amounts of this gas are released as scraping is carried on. Air anal yses have shown vinyl chloride concentra tions close to the hand during scraping in the range of 600 to 1,000 ppm. Although this study provided no evidence to suggest that vinyl chloride per se is the etiological agent, the measurement of vinyl chloride concentrations may serve as a use ful index to the adequacy of reactor ventila tion. Inhibitors in Vinyl Chloride,--An inhibi tor, usually phenol, may be added to the Arch Environ Health--Vol 22, Jan 1971 80 OCCUPATIONAL ACROOSTEOLYS1S--COOK ET AL Table 5.--Reactor Cleaning Procedures as a possible factor in the Plant Hand Scraping ............................... - _A--------- - ---------- After Every Less Than Cleaning by Water Jet Cleaning by Solvent causation of AOL. Contact with these inorganic and or ganic peroxides occurs as Code Batch Plants with AOL cases BX X FX KX (through 1964) Every Cycle From 1965 they are handled prior to their introduction into the reactor and also during re actor cleaning if any unre acted catalyst remains in the residue. Twenty-two of the RX U Every 5*12 cycles From 1962 (building A) (building 8) EE X 26 plants used lauroyl perox ide. Eight of these were plants with AOL cases or Plants with possible AOL cases BB Every "several" cycles On glass-lined, intermittently possible AOL cases, and 14 were plants without AOL in 1966, entirely Plants with no AOL cases AX (to September 1964) CX D HX J Every 2Vz cycles L Every 4-8 days NX P Weekly through 1964, every 2 weeks From 1962 From 1965 by December 1966 From September 1964 mon organic peroxide was the IPP, and it, too, was in use in both the positive and the negative plants, in four of the former and in eight of the latter. No catalyst or group of catalysts was used in all positive plants and in no negative rlants. Constituents of Scrapings. --Other suspect materials 1965-1966 QX sX T X Every 1-3 cycles are constituents of scrapings removed from the reactors during manual cleaning. (emulsion) (suspension) V Every 10 cycles 2X AA Every 4 or more cycles DO X GG From start of operations These are believed to include some partially polymerized resins, along with some un reacted catalysts and addi tives, and unpolymerized vinyl chloride. Little specif ic information is currently vinyl chloride to prevent spontaneous poly available concerning these constituents. It has merization in storage. An inhibitor may also been reported that lauroyl peroxide is definite be added to the vinyl chloride recovered ly associated with residual PVC particles, from the reactor. The phenol is removed by whereas little of the IPP remains unreacted treatment with caustic soda prior to its in the polymer.11 transfer to the reactor. The practice over the Further investigation of the scrapings years covered by the study is presented in from the reactor may be expected to furnish Table 3. The use of the inhibitor in the information that may lead to a better under virgin vinyl chloride at two of the positive standing of the factors involved in causation plants and three of the negative plants, with of AOL, A difficulty in such investigation is inhibitor being added in the vinyl chloride that the composition of the scrapings is recovery operation in two positive plants continually changing from the time they are and four negative plants, leads to the con removed from the surfaces of the reactor. It clusion that inhibitors probably are not a is obvious that comparatively small mole factor in AOL causation. cules of partially polymerized vinyl chloride Catalyst Exposures.--Since the catalysts are increasing in size with passage of time, are all active compounds, they were suspect and active constituents such as residual cat- Arrh Environ Health--Vol 22, Jan 1971 OCCUPATIONAL ACROOSTEOLYSIS--COOK ET AL 81 Table 6.--Reactor Cleaning Procedures Method Manual scraping after every reactor cycle Manual scraping after 2-3 cycles Manual scraping after 4-12 cycles Water-jet cleaning Organic-solvent cleaning Positive Plants 7 0 2 2 1 Negative Plants 9 2 4 2 2 alyst and vinyl chloride monomer are de creasing or may no longer be present. Procedures Affecting Exposures of Reactor Cleaners Since all but one of the 25 operators diagnosed as positive for AOL had worked as reactor cleaners and the one exception had cleaned laboratory reactors,8 additional inquiry was made of procedures that might affect the exposures of this group of workers. Reactor Cleaning Procedure.--The number of man hours required to dean the reactors depended not only on the number of reactors but also on the frequency of cleaning. In 16 of the 26 production plants, reactors were manual ly cleaned after every batch. This was true of six of the seven positive plants and one of two plants with possible AOL cases. In only nine of the 17 plants with no AOL cases were reactors manually scraped after every batch. The procedures followed in each of these plants appears in Table 4 and an analysis of the types of procedure by positive and nega tive plants in Table 5. Procedures Following Completion of Poly merization--Since vinyl chloride and other vol atile components of the residue remaining on the reactor surfaces following discharge of the batch might be factors in AOL causation, infor mation was obtained concerning the degree of vacuum applied to the reactor prior to opening it and the length of time that the reactor was under this reduced pressure. The amount of volatile materials remaining in the residue on the reactor surfaces was considered to be roughly proportional to the absolute pressure and inversely proportional to some function of the time. Since the rate of removal of volatile materia] would tend to decrease with time, the logarithm of this factor was used in calculating an absolute pressure-time (AP-T) value. It was hypothesized that the relative amount of vola tile materials remaining in the residue follow ing application of vacuum would approximate a nondimensional value obtained by dividing the absolute pressure in inches of mercury by the logarithm of the time in minutes during which the vacuum was applied. Thus, the lower the AP-T value, the less is the amount of volatile materials remaining in the residue. Another factor affecting removal of vola tile materials is the temperature range over the period that the reactor is under vacuum. The variables in temperature control were so diverse that this factor defied classi fication and could not be incorporated in the parameter under consideration. Actually, variations in degree of vacuum and timing of its application were such that meaningful calculation of an AP-T value could be ac complished only for five of the seven posi tive plants, one of the two with possible cases, and 14 of the 17 negative plants. In the subsequent discussion of these val ues, the three plants that started in opera tion as recently as 1965 and 1966, all with no AOL cases, are not included as it is not known whether sufficient time had elapsed for cases to have developed by the terminal date of the study period. These plants are listed separately in Table 6. The AP-T values ranged from 0.5 to 21. The AP-T values were 10 or above (an entirely arbitrary division line) for four of the five positive plants and for one of the two plants with possible AOL cases. Four of the negative plants gave AP-T values below 10, and seven gave values of 10 or above. Consideration is directed to the number of reactor cleaners and the corresponding num ber of expected AOL cases in the two groups of positive plants as compared with those in the negative plants. The total number of reactor cleaners at risk in the positive or possibly positive plants with an AP-T value of less than 10 is 63; that in such plants with an AP-T value of 10 or more is 436. The relation of these two totals is to be compared with that of the negative plants in which 193 reactor clean ers at risk are in the four plants with an AP-T of less than 10 and 132 are in the seven plants with an AP-T above that value. The number of expected cases of AOL for each plant is also listed in Table 6. These values were obtained by dividing the num ber of reactor cleaners at risk by 37 This Arch Environ Health--Vnl 22. Jan 1971 82 OCCUPATIONAL ACROOSTEOLYSIS--COOK ET AL figure derives from the paper on epidemio logical aspects of AOL8 which reports the prevalence of AOL or possible AOL as one case per 37 reactor cleaners at risk. Of the negative plants with an AP-T val ue of 10 or more, none of those in operation earlier than 1965 had enough reactor clean ers at risk to expect an AOL case, whereas two of the four plants with an AP-T value under 10 have enough workers to expect cases. Cases would be expected from the total population of reactor cleaners in the entire group of the seven negative plants with AP-T values of 10 or more, but the total number would be 3.6, as compared with 5.3 expected cases from the entire group of four negative plants with an AP-T value of less than 10. This contrasts with expected cases in the group of positive plants with an AP-T value of less than 10 as 1.7, compared with 11.8 in such plants with an AP-T value of 10 or more. The proportion of the number of reactor cleaners in the negative plants with an AP-T value of less than 10 to the total number in such plants was 193 to 325. This propor tion for the positive plants was 63 to 499. A Xs test demonstrated that the probability of this difference between the negative and the positive plants occurring by chance is less than 0.001. In order to make this comparison between the positive and negative plants with AP-T values below and above 10, it was necessary to oversimplify a multiplicity of varying de tails of operational procedures from the time the polymerization is completed to the time the cleaner enters the reactor. Accordingly, any interpretation of the findings that re duction of the AP-T value below 10 by reducing the absolute pressure or keeping the reactor at reduced pressure for a longer time or both may results in reduction in development of AOL cases should be ap proached with great caution. Preventive Measures The present knowledge as to AOL is such that categoric statements as to preventive measures are not justified. However, on a somewhat speculative basis, some measures are indicated that may be conducive to AOL prevention: 1. Although some manual removal of resi due from reactor surfaces may be required when solvents or high-pressure water jets are used for cleaning, the more extensive employment of these methods would greatly reduce the number of man hours required for conducting this operation by manual methods. 2. The adequacy of ventilation of reactors prior to entry should be checked with flammable vapor indicator equipped with a scale of greater sensitivity than 1% of the lower explosive limit of vinyl chloride or with other instruments of such sensitivity. 3. Further investigation of the constituen cy of the scrapings from the reactor surfaces should be considered, as such information might disclose the factor responsible for the causation of the disease. References 1. Albright LF: Vinyl chloride processes. Chem Eng 74:123-130, 1967. 2. Albright LF: Manufacture of vinyl chloride. Chem Eng 74:219-226, 1967. 3. Albright LF: Polymerization of vinyl chloride. Chem Eng 74:151-158, 1967. 4. Albright LF*. Vinyl chloride polymerization by suspension processes yields polyvinyl chloride res ins. Chem Eng 74:145-152, 1967. 5. Albright LF: Vinyl chloride polymerization by emulsion, bulk and solution processes. Chem Eng 74:85-92, 1967. 6. PVC producers. Chem Eng News 47:18-19, 1969. 7. "Wilson Till, McCormick WE, Tatum CF, et al: Occupational acroosteolysis: Report of 31 cases. JAMA 201:577-581, 1967. 8- Hillman MIX Cook WA, Whitehouse WM, et al: Occupational acroosteolysis: I. An epidemiologi cal study. Arch Environ Health 22:61-73, 1971. 9. Dodson VN, Dinman BD, Whitehouse WM, et al: Occupational acroosteolysis: III. A clinical study. Arch Environ Health 22:83-91. 1971. Arch Environ Health--Vol 22, Jan 1971 Printed and Published in the United Stales of America Reprinted from the Archives of Environmental Health January 1971, Volume 22 Copyright 1971, American Medical Association Occupational Acroosteolysis III. A Clinical Study Vernon N. Dodson, MD; Bertram D. Dinman, MD; Walter M. Whitehouse, MD; Ahmed N. M. Nasr, MD, PhD; and Harold J. Magnuson, MD, MPH, Ann Arbor, Mich UCC 100604 83 Occupational Acroosteolysis III. A Clinical Study Vernon N. Dodson, MD; Bertram D. Dinman, MD; Walter M. Whitehouse, MD; Ahmed N. M. Nasr, MD, PhD; and Harold J. Magnuson, MD, MPH, Ann Arbor, Mich Four subjects with acroosteolysis were studied clinically. All had osteolytic lesions, especially in the distal phalanges of the hands, and Raynaud's phenomenon. All had worked as polyvinyl chloride reactor-vessel cleaners, with hand scraping being the common mode of operation. Raynaud's phe nomenon anteceded osteolytic lesions. One of the subjects was in negative calcium and phosphorus balance. Plethysmographic abnormalities were present in three. Esophageal motility was normal. Scintiscans of the hands using sodium fluoride labelled with radioactive fluorine (MF) revealed variable uptakes in the fingers which correlated with the radiographic lesions. A wide variety of clinical laboratory parameters were normal. 0 CCUPATIONAL acroosteolysis ( AOL) consists primarily of Raynaud's phenome non and osteolysis of the Hiatal phalanges in .the hands. Suciu et al first described Ray naud's phenomenon in polyvinyl workers in 1963.1 Osteolytic lesions were reported in other cases.2 7 Other associated clinical fea tures such as hepatomegaly, the cutaneous stigmata of scleroderma, hypothroid param eters, and numerous abnormal laboratory tests have been reported.17 Epidemiologically, patients with AOL are associated with the manufacture of vinyl chloride polymers, more specifically with the hand cleaning of the polymerization reactors. This paper reports the results of intensive clinical investigation of four patients with AOL. These four patients were selected from those employees of 32 plants engaged in the manufacture of vinyl chloride poly mers throughout the United States. Methods Each patient was admitted to the Clinical Research Unit, University of Michigan Hospi tal, for a two-week period where detailed medi cal histories were investigated and physical examinations performed. Details of the work histories and environmental exposures were de rived from in-plant epidemiological studies car ried out jointly by the Institute of Environ mental and Industrial Health of the University of Michigan and the Manufacturing Chemists' Association. Table 1 lists the laboratory param eters applied. In addition, calcium and phos phorus balance studies were carried out, using a dietary history to establish a constant cal cium-phosphorus intake to which each patient was accustomed. Daily 24-hour urine specimens were collected, and stool specimens were pooled every three days for calcium and phosphorus determinations. Appropriate correctional com putations for uneaten foods were made. Cate cholamines and 5-hydroxyindole acetic acid were determined on 24-hour urine collections, often partitioned by six-hour units and usually collected on days of provocative cold stresses during plethysmographic determinations. Numerous roentgenographic techniques were introduced, ranging from skeletal survey films to barium swallow radiographs and bone xerog raphy. Scans of the hands and sacral regions, using sodium floride labelled with radioactive fluorine (1SF), were undertaken in two of the patients. Plethysmography of the forearms and hands, using the method of Zweifier et al, was employed.s Esophageal motility studies were made, using Saladin et al's technique.'-1 Results Submitted for publication Feb 9, 1970; accepted March 24. History.--Blanching of the hands was the From the section of Occupational and Environ first sign of AOL. In two of the men, the mental Medicine and the Department of Internal Medicine (Dr. Dodson), and the Institute of Knvic onset was gradual over a few weeks, and in ronmental and Industrial Health (Drs, Dinman and' "two it was abrupt, as shown in Table 2, Nasr), the Department of Environmental and In dustrial Health, School of Public Health (Dr. Mag Cold was the inciting factor in all cases. The nuson). and Ihe Department of Radiology (Dr, blanching was associated- with either tin Whitehouse), University of Michigan, Ann Arbor. gling of the hands or tenderness of the Reprint requests to Institute of Environmental and Industrial Health, University of Michigan. Ann fingertips. Thepolyvinyl work exposure-to~ Arbor. Mich 48104 (Dr. Dinman). first-blanching ranged from one to 23 Arch Environ Health--Vnl fan /?)// Fig 1.--Note transverse defects in distal phalanges of fingers one to four on left hand and fingers one to three on right hand. Marginal defect is present in distal portion of distal phalanx of right fifth finger (case 1). Table 1.--Laboratory Investigation of AOL Patients Biochemistry Sodium Calcium Potassium Phosphorus Carbon dioxide Cholesterol Chloride Glucose (fasting) Bilirubin direct/total Urea nitrogen Creatinine Total serum protein Electrophoresis serum Albumin ol* 1 -2 y Magnesium Lipids Phospholipids Triglycerides Catecholamines Norepinephrine Epinephrine Serum Enzymes Alkaline phosphatase Threonine aldolase Acid phosphatase Malic dehydrogenase Serum glutamic oxaloacetic transaminase Glutamic dehydrogenase Serum glutamic pyruvic transaminase Leucine aminopeptidase Lactic de hydrogenase isogram Aldolase 5 'Nucleotidase Creatine phosphokinase Serine dehydrotase Isocitrate dehydrogenase Immunological Tests Venereal Disease Research Antinuclear factor Laboratory Cryoglobulins LE cell Im mu nophoresis Direct Coomb's IgA Latex-rheumatoid factor lgM Cold agglutinins IgG months. The length of the interval bore no relationship to the mode of onset or the severity of the symptoms. A summary of the habits and dietary history of these patients is detailed in Table 3. Three of the four men were cigarette smokers with smoking histories of nine (case 1), 15 (case 4), and 25 (case 2) pack-years. Half drank coffee (three or nine cups per day). All were casual tea drinkers. Three drank between three and 18 bottles of beer each month. One (case 4) drank 120 cans of beer per month but no whiskey while the others consumed less than four Fig 2.--Hands show further resorption of bone, resulting In shortening of terminal phalanges of all fingers except left fifth and right fourth, latter having developed transverse defect. Right fifth b r UCC 100606 OCCUPATIONAL AC ROOSTEOLYSIS--DODSON ET AL 85 ounces of whiskey per month. Table 2.--Clinical Features of Patients With AOL at Onset Three of the men rarely drank milk, and the other tease 4) averaged two glass es each day. Dietary histor ies indicated calcium intakes between 1,000 mg and 1,300 mg/day except for one pa tient (case 2) whose usual intake was 518 mg/day. This patient's normal dietary in take of phosphorus averaged 2,400 mg/day while the oth ers took in between 1,500 and 1,800 mg/day. None had unusual food cravings or aversions. Clinical Features Blanching, Onset precipitated by cold Associated symptoms; Fingertip tenderness, Tingling of the hands, Joint pain Case 1 Case 2 Sudden Gradual + 4- 4-- Elbows 4Coccyx Low back Joint stiffness Subcutaneous nodules Hands -- Muscle swelling Ekposure-to-first blanch interval 1 mo Intermittent tinnitus + Palpitations with exertion -1- Hands Knees 6 mo + 4- Case 3 Gradual 4- 4-- Hand Knees 4418 mo " 4- Case 4 Sudden 4- _ 41 hip, 1 wrist, 1 knee, hands Low back Sacroiliac -- -- 23 mo 4-- Some details of family his tory are recorded in Table 4. Table 3.--Habits and Dietary Histories of Patients With AOL A firm history of hyperten- Features sionwas present in either the Cigarette smoking (pack-yr) Case 1 9 Case 2 25 Case 3 0 Case 4 15 father or the mother in three of the four patients, and one half also had immediate fam ily members with diabetes mellitus though none per Average coffee consumption (cups/day) Alcohol per month Beer (12-oz unit) Whiskey (oz) Milk (glasses/day) 9 3 1 Rarely 3 13 4 Rarely 0 8 6 Rarely 0 120 8 2 sonally evidenced the dis Calcium intake (mg/day) ease. Notably, the fathers of Phosphorus intake (mg/day) 1,045 1.519 518 1,587 1,327 2,457 1,096 1,797 all four men had heart dis eases. Table 4.--Diseases in Family Member of Patient With AOL The associated symptoms and signs of joint involve ment are listed in Table 5. Intermittent transient ortho static tinnitus was mentioned Family Member Father Case 1 Diabetes mellitus Case 2 Heart disease Case 3 Heart disease, Chronic lung disease Case 4 Congestive heart failure, Hyperlipemia, Hypertension by three of the four patients. This occurred upon rising from sitting or recumbency. Mother Brother Emphysema Hypertension, Obesity Hypertension Obesity The one patient not having tinnitus (case 3) had chronic postnasal drip, nasal stuffiness, and a prior history of child hood bronchial asthma. Physical Examination.--The major physi cal findings were restricted to the skin and joints, as shown in Table 5. Synovial and skin thickening were observed by two or more clinicians independently. Dupuytren's contractures were present bilaterally in one patient (case 1"). The extensor ligaments of the wrist were thickened in another (case 2). The distal phalanges of the hands of one patient (case 3) were shortened. clinical laboratory parameters depicted in Table 1 were found to be essentially normal. All four men vacillated between negative and positive calcium balances day-to-day during the two-week period. Except for one patient (case 2), all averaged a positive balance ranging from 33 to 246 mg/day. This patient managed a deficit of 27 mg/ day, which was primarily a function of renal excretion. All had positive phosphorus balances except this patient, whose phosphoturia always accounted for more than his stool losses. Five-hydroxyindole acetic acid determina Laboratory Parameters.--The numerous tions were performed on 24-hour urine spec- Arch Environ Health--Vol 22, Jan 1971 86 OCCUPATIONAL ACROOSTEOLYSlS--DODSON ET AL Table 5.--Physical Findings on Patients With AOL imens which were sometimes Physical Findings Skin Case 1 Case 2 Case 3 Case 4 partitioned in six-hour incre ments (Table 6). One pa DorsaMhickening hands Blanching Without cold With cold Joints Synorial thickening Incidental + +* + ++ + + ++ PlPf MPJ Knees Oupuytren's contracture Finger R 2 and 3 L 3 and 4 PlPt Thickening of wrist extensors Shorten ing of finger tips tient's third-period value of 2.8 occurred during a cold stress while plethysmogra phy was being done (case 1). Another patient's third value of 1.6 did not increase under similar cold stress (case 2). Plethysmographic studies revealed normal pulse vol umes in all but one patient * Also fingers, f Proximal interphalangeal. X Metacarpophalangeal. (case 2). All patients ex cept one (case 3) showed di minished pulse volumes with cold stress typical of the Raynaud's phenomenon observed in other dis eases. Esophageal motor activity was normal in all subjects as shown by barium fluoro scopy and Saladin's et al motility technique,9 Radiological Findings.--The radiographic findings of the four patients analyzed are summarized in the case reports. Report of Cases Fig 3,*--Erosive and sclerotic changes in sacro-iliac joints bilaterally (case 1). Fig 4.--Erosive lesions of os calcis posterior to bony spur (case 1). Case 1.--X-ray films made elsewhere prior to the patient's hospitalization at the University of Michigan exhibited normal hands on Oct 20, 1966, and, in 1967, showed marginal bone de fects of the right third and left second, third, and fourth distal phalanges and distal fragmen tation of the distal phalanx of the left first digit. On Aug 8, 1967, transverse defects were noted on the distal phalanges of the left fingers one to four and right one to three, with resorp tion of the distal portion of the distal phalanx of the right fifth finger. The appearance of the hands on Oct 13, 1967, when the patient was first seen, is shown in Fig 1. Some reparative changes were noted in comparison with the previous films, with persistent transverse de fects in fingers one to four on the left and one to three on the right, with a marginal defect in the distal portion of the distal phalanx of the right fifth finger. Further check-up examination was obtained on Jan 9, 1969, and Oct 24, 1969, the latter being shown in Fig 2. During this period of time, there has been further resorption of hone, resulting in shorten ing of the terminal phalanges of all fingers except the left fifth, which has remained unin volved, and the right fourth, which has main tained its length but has developed a transverse Arch Environ Health--Vo! 22, Jan 1971 Fig 5,--'Left hand shows marginal defect at tip of distal phalanx of second finger, with transverse defect in distal phalanx of third finger. Right hand shows transverse defects in distal phalanges of second and third fingers and an additional distal marginal defect of second finger (top [case 2j). Scintigram shows uptake of sodium fluoride 18F (bottom). Fig 6.---Sacro-iliac joints show erosive and sclerotic changes, more marked on left (case 2). Arch Environ Health--Vnl 22, Jan 1971 UCC 100609 88 OCCUPATIONAL AC ROOSTEOLYSIS--DODSON ET AL Table 6.--Fiue-hydroxyindole Acetic Acid and Catecholamine Studies in Patients With AOL Case l Case 2 Case 3 Case 4 Urinary Excretion Five^hydroxylndole acetic acid 8am-2pm 2pm-8pm Spm-2am 2am*8am Total (1) Total (2) Total (3) Catecholamine Epinephrine Norepinephrine Urine Creat- Urine Creat- Urine Creat- Urine Creat- Mg Volume inine Mg Volume inine Mg Volume inine Mg Volume inine 0.78 0.92 2.80 0.89 5.39 5.8 4.0 7.6 390 460 730 140 1,720 2,170 1,720 0.36 0.42 0.62 0.12 1.52 1.47 1.46 2,0 265 0.59 2.2 240 0.54 1.2 300 0.69 1.6 210 0.48 7,0 1.015 2.30 23.6 6.1 390 1,208 416 430 2.444 0.30 0.97 0.33 0.34 1.94 defect. The right fifth finger has shown addi tional resorption of the phalangal tip. Softtissue clubbing is noted in the shortened phalanges. Some~sclerosis is noted along the margins adjacent to the transverse defects. This patient also showed extensive erosive and scle rotic changes in the sacro iliac joints, as seen in Pig 3. An erosive lesion was also found in the plantar surface of the os raids, as depicted in Fig 4. Case 2.---This patient had an extensive skele tal survey on Feb 8, 1968. with positive findings limited to the hands and pelvis. As shown in Fig 5, the left hand revealed a marginal defect at the tip of the distal phalanx of the second finger while the third finger showed a transverse defect in the distal pha lanx. At the same time, the right hand present ed transverse lytic defects in the distal pha langes of the second and third fingers and an additional distal marginal defect of the second finger. Figure 5 compares a typical sodium fluo ride ,SF scintiscan of the bands and the corre sponding roentgenograms. A slight increase in sodium fluoride 1SF uptake is seen in the distal phalanges of the left third and right second and third fingers while the uptake is diminished over the distal phalanges of the left fifth and right fourth and fifth fingers. Generally, up takes over the remaining bones of the hands are uniform. The sacro-iliac joints also displayed extensive erosive and sclerotic change, more marked on the left, as shown in Fig 6. Case 3.--Extensive skeletal survey on this Fig 7.--Hands show shortening and widening of distal phalanges of ail fingers, with resultant clubbing of soft tissues (case 3). Arch Environ Health--Vo! 22, Jan 1971 OCCUPATIONAL ACliOOSTEC)LYS/S--DODSON ET AL 89 patient on March 18, 1988, showed shortening and widening of the distal phalanges of all fingers (Fig 7), with resultant clubbing of the soft tissues. These changes were interpreted as the end stages of healing of AOL. The saero-iliac joints showed erosive and sclerotic changes, more marked on the right (Fig 8). Review of previous x-ray films made else where indicated that some sacro-iliac joint changes were present on April 8, 1959, and that the hands had not changed in appearance since October 1965. Case 4.--Review of previous x-ray films made of this patient indicated that no abnor mality of the hands was present on Feb 20, 1967, or Jan 23, 1968. On July 15, 1968, films of the sacro-iliac joints exhibited erosive and scle rotic changes, and the hands revealed marginal defect of the distal phalanges of the left third and the right fourth fingers, slight irregularity of distal phalanx of the right fifth finger, and transverse defects of the distal phalanges of the thumbs bilaterally. Skeletal survey on Oct 20, 1968, noted sepa ration of the tip of the styloid process of the right ulna, marked resorptive and sclerotic changes of the sacro-iliac joints (Fig 9), and some progression of the hand involvement, with transverse defects now involving the distal phalanges of the left digits one to three and Fig 8.--Sacro-iliac joints show erosive and sclerotic changes, more marked on right (case 3). Fig 10.--Hands show transverse defects in distal phalanges of left fingers one to three and right fingers one and four, with an additional marginal defect of left second finger (case 4), Fig 9.--Sacro-iliac joints show erosive and sclerotic changes bilaterally (case 4). Artih Enriron Health--Vuf '2- Jan /V, / UCC 100611 90 OCCUPATIONAL ACROOSTEOLYSIS--DODSON ET AL right digits one and four and an additional marginal defect of the left second digit (Fig 10). Irregularity of the lower pole of the right patella was also noted. Comments Raynaud's phenomenon was the first mSnSTestation noted by these subjects and in the majority of the cases reported. Harris and Adams8 reported one case in which distal digit shortening occurred first, while Wilson et al4 noted this in five of his 31 cases. This suggests that the vascular lesion tends to antecede the bone changes in most individuals. None of these patients exhibited Raynaud's phenomenon of the feet, which confirms Wilson et al's experience. All of these individuals had been "reactor clean ers" ie, they spent some hours each week cleaning the polyvinyl chloride polymeriza tion kettles shortly prior to or at the time of the onset of their disease. No consistent relationship was found between the occur rence of severity of either Raynaud's phe nomenon or the osteolytic lesions and those habit patterns which have been shown to influence vascular reponses, eg, cigarette smoking and the ingestion of alcohol, coffee, or tea. Likewise, neither the bone lesions nor Raynaud's phenomenon seemed to cor relate with the daily consumption of milk or dietary calcium. Dallas and Nordin10 have shown that while dietary deficiencies or cal cium may or may not potentiate osteoporo sis of various types, an adequate calcium intake of 15 mg/kg is protective. Only one patient (case 2) had less than this amount (5.8 mg/kg). Since such dietary histories have been shown to be quite reliable,13 their habitual calcium intakes do not appear to be a consistent causative factor in their cases. Likewise, the calcium-phosphorus balance studies on these men reveal no clear-cut abnormality or pattern. This result might be anticipated since these men have been away from reactor cleaning for several months, and the mass of involved bone is slight compared to the entire skeleton. However, it should be pointed out that a two-week bal ance period is a short observation period though they were placed on calcium and phosphorus intakes comparable to their usu al eating habits to obviate disruption of their "steady state." At first glance, it appeared that there was a rich history of heart disease or hypertension or boffTTmtur parents of these men. Comparison of them with healthy-male chemical workers not involved in polyvinyl chloride manufacture and matched in age suggests a significant correlation by the x2 test, using Yate's correction. Whether such an observation suggests a hereditary frailty to explain the susceptibility of those individ uals developing AOL is speculative. The number of observations is too small to sup port seriously such a contention at this time. The extent and character of other muscu lar skeletal findings such as tendon and synovial thickening confirm the findings of others. We are able to offer no suitable explana tion for the frequent episodes of intermittent orthostatic tinnitus described by three of these men. Suciu et al described liver enlargement, various vague asthenic symptoms, and mani festations of hypothyroidism. Such findings were not present in these men or had they been in the past. The radiological findings in occupational AOL have been reviewed recently on the basis of the epidemiological investigation of the polyvinyl chloride industry. Workers in other countries have described lytic lesions in the distal phalanges of the hands, styloid processes of the ulna and radius, and the sacro-iliac joints.1'3- These four patients demonstrate the en tire spectrum of radiographic hand changes in occupational AOL from the earliest mar ginal erosion of the distal phalanx through the extensive transverse defect, with ulti mate healing resulting in a short and wide phalanx. The fact that all four showed sac ro-iliac involvement suggests that the inci dence of involvement at this site is probably higher than previously suspected. The additional smaller abnormalities found in the ulnar styloid, os calcis, and lower pole of the patella suggest that these areas might well be evaluated in other cases as they appear in order to get a more accur ate estimate of their incidence. The sodium fluoride ,SF scintiscan data suggests that the more advanced osteolytic changes with transverse defects at this time were picking up more fluorine and probably Arch Environ Health--Vol 22. Jan 1971 OCCUPATIONAL ACROOSTEOLYSIS--DODSON ET AL 91 calcium, indicating active remineralization. That the digits with marginal defects picked up less than the normal bones implies some type of uptake block or active demineraliza tion. If the latter is true, it is all the more interesting that different bone lesions ap pear to be mineralizing and demineralizing in the same individual, even the same hand. The etiological agent or agents responsible and the pathogenical mechanisms remain obscure. Medical Prevention.--The interval be tween the beginning of reactor cleaning and the onset of AOL appears highly variable, ie, one to 23 months. Accordingly, the fre quency with which one examines reactor cleaners has no rational basis. By contrast, the occurrence of Raynaud's phenomenon should serve as a sufficient warning for fur ther inquiry. This is not to suggest that prudent medi cal practice does not require periodic exami nation of reactor cleaners and other groups of workers in whom AOL occurs. These groups should be surveyed periodically by use of hand roentgenograms and medical questionnaires. The form the latter might take has been discussed in the epidemiologi cal segment in this series. But we would emphasize that prevention of AOL also re quires the cooperation of supervision. This group should be alerted as to the possible significance of Raynaud's phenomenon in relation to AOL. Upon the discovery of any case of AOL among workers at risk, such individuals should be removed from the job. X-ray films of the hand, back, and knees should be obtained. A medical history questionnaire, similar to those used in the epidemiological survey reported here, should also be pro cured. A medical evaluation for other causes of Raynaud's phenomenon is in order. In view of the unpredictable course of AOL, it would be wise to exclude any individual with_ Raynaud'sphenomenon from~reactpr cleaning. Since seniority practices usually remove workers from reactor cleaning with the passage of time, such excluded workers probably would not return to this particular job. In view of the low prevalence of this disease, we believe that permanent restric tion of these few workers from this job does not pose any undue administrative burden. Until more is known of the natural history of this disease, prudence would dictate the course as here recommended. This investigation was supported by Public Health Service grant 2M01 RR-42 and a grant from the Manufacturing Chemists* Association, References 1. Suciu I, Drejman I, Valeskai: Investigation of the diseases caused by vinyl chloride. Med Intern 15:967-978,1963. 2. Cordier JM Fievez C, LeFevre MJ, et al: Acroosteolysis and skin lesions among workers en gaged in cleaning reactors, Cah Med Travail 4:1419, 1966. 3. Chatelain A, Motilon P: A syndrome of acroos teolysis of occupational origin*. Heretofore not iden tified in France. J Radiol Electr 48:277-280,1967, 4. Wilson RE, McCormick WE, Tatum CF, et al: Occupational acroosteolysis. JAMA 201:577-581, 1967. 5. Harris DK, Adams WGF: Acroosteolysis oc curring in man engaged in the polymerization of vinyl chloride. Brit Med J 3:712-714, 1967. 6. Benoit JP: Acroosteolysis of occupational ori gin, thesis. Lyon, 1968. 7. Hess EV, Schneider H, Roush G: Clinical studies in occupational acroosteolysis: A new dis ease. Read before the 40th meeting of the Central Society for Clinical Research, Chicago, 1967. 8. Zweifler AJ, Cushing G, Conway FJ: The relationship between pulse volume and blood flow in the finger. Angiology 18:591-598, 1967. 9. Saladin TA, French AB, Zarafonitis CJ, et al: Esophageal motor abnormalities in scleroderma and related diseases. Amer J Dig Dis 2:522-535,1966, 10. Dallas I, Nordin BEC; The relation between calcium intake and roentgenologic osteoporosis. Amer J Clin Nutr 11:263-269, 1962. 11. Becker BC, Indik BP, Beeuwkes AM: Dietary Intake Methodologies: A Review, technical report. Ann Arbor. Mich, University of Michigan, i960, pp 55-57. Arch Environ Health--Vol 22, Jan 1971 Printed and Published in the United States of America UCC 100613 Reprinted From The Journal of The American Medical Association August 21, 1967, Vol. 201, pp. 577-581 Copyright 1967, hy American Medtcal Association Occupational Acroosteolysis Report of 31 Cases Rex H Wilson, MD, William E. McCormick, MS, Caroll F. Tatum, MD, and John L, Creech, MD In 31 cases of acroosteolysis of the hands of workmen derma. Some have clubbing of the fingers. Only a few associated with vinyl chloride (CH2CHCI) polymerization of the cases have sought medical attention because processes, the osteolysis was specific to the distal pha of symptomatic complaints; the majority have been langes of hands and was frequently associated with Ray found through x-ray examinations of the hands. naud's symptoms. Affected personnel ranged in age from A summary of the physical findings of the ob 26 to 47. The disorder is believed to have resulted from served cases is shown in Table 1, and a detailed de a combination of physical insult, chemical insult, and scription of the two predominantly common symp personal idiosyncrasy. The specific causes are unknown. toms follows; The prevalence was found to be less than 3% among Raynaud's Phenomenon.--This symptom com employees performing similar work. No cases were found plex has occurred in varying degrees, with one or in workmen using or processing the polymer or manu both hands involved. Generally, its effect is observed facturing commercial resins. as marked discomfort on exposure to cold. None of the cases have exhibited vascular changes of the feet. uring mid-1964, several complaints came to our A unilateral sympathectomy performed on one of D attention of soreness and tenderness of the the most seriously affected cases relieved most of fingertips of workmen in a manufacturing plantthe symptoms on that side. The acroosteolysis was polymerizing vinyl chloride (CH2CHC1). Compre not altered by the sympathectomy. hensive examinations of these individuals revealed Roentgenological Changes.--The most unique nothing of significance except a "Raynaud-like phe characteristic of this syndrome is the unusual roent nomenon" of the hands, and in some cases, definite genological findings. These are described as acroos changes of the distal phalanges were apparent on teolysis and are illustrated in Fig 1. It may be present roentgenograms of the hands. Since then we have in all of the fingers and readily observable, or only in observed additional cases of this unusual syndrome. one finger and barely discernible, as illustrated in It is the purpose of this manuscript to summarize Fig 2. X-ray films of the feet, as well as those of the our observations. long bones and of the skull have been made in some Description of Syndrome of these cases, with no osteolysis found other than in the distal phalanges of the hands. To date we have observed 31 cases of hand dis orders among 3,000 personnel involved in vinyl chloride manufacturing and polymerization. All have been men, between the ages of 26 and 47, We have not observed any ethnic or racial tendency for the syndrome. The great majority have been char acterized by two common factors: symptoms likened to those ascribed to Raynaud's phenomenon and acroosteolysis of the distal phalanges. A few have Acroosteolysis is a rare clinical entity, with only 72 cases of the familial type reported up to 1965, ac cording to Cheney in his excellent review.1 The diagnosis of acroosteolysis requires expert roent genological technique and knowledge. It has been our observations that early cases of acroosteolysis will be missed by the roentgenologist not familiar with this condition. All of the cases on which we are reporting have been confirmed by at least two ex had no symptoms but have the roentgenographic evidence of acroosteolysis. Several of the patients perienced roentgenologists working independently. We established the following roentgenographic have external skin lesions on the dorsal surfaces of the hands and forearms, with a rope-like appear ance resembling changes sometimes seen in sclero- From the Medical Division, B. F. Goodrich Cov Akron, Ohio. Reprint requests to 500 S Main St, Akron, Ohio 44318 (Dr. Wil son). criteria for our diagnoses. These are the result of our having viewed several thousand x-ray films of the hand in our search for the cause of the syn drome. 1. General: Roentgenographically, acroosteolysis, as found in workers who are exposed to vinyl UCC 100614 578 ACROOSTEOLYSIS--WILSON ET AL chloride and polyvinyl chloride manufacturing processes, is somewhat different from that found in familial osteoporosis with acroosteolysis and in fa' milial osteosclerosis with acroosteolysis. In osteopo rosis with acroosteolysis there may be compression fractures in the spine, and basilar impression of the skull, along with destruction of the midpha langes. None of these findings have been seen in these workers. The changes in the distal phalanges are similar in both conditions. Osteosclerosis acroosteolysis observed by Andren et al2 (University Hospital, Malmo, Sweden) in twins showed diffuse sclerosis, with cortical thick ening of the shafts of the long bones and clubbing of the metaphyseal ends. The phalanges of the hands and metacarpals were foreshortened, and the distal phalanges showed acroosteolytic changes. The feet showed the same changes, except that the distal phalanges were not fragmented. There has been no evident destruction in the mid or proximal phalanges of the thumbs or fingers of these individuals and no evidence of a lytic, destruc tive lesion in the feet. We have not observed any loss of calcium salt in the bones of the wrist and remaining bones of the hand and phalanges in any of these individuals, and there has been no evident sclerosis of the wrist and hand bones. 2. Mild stage: The earliest change found in acroosteolysis in these workers has been a loss of the cortex of one or more of the tufts of the distal pha langes, with no destruction of the tuft or shaft of the distal phalanx. The next more advanced stage may be a small, half-moon cut in the cortex of the tuft of one or more Table 1.--Summary of Symptoms and Findings Symptoms Acroosteolysis without Raynaud's symptoms, one hand Acroosteolysis without Raynaud's symptoms, both hands Acroosteolysis with Raynaud's symptoms, one hand Acroosteolysis with Raynaud's symptoms, both hands No, of Cases 4 5 5 17 Total 31 Clubbing of fingers Skin nodules 8 8 2. Mild stage of acroosteolysis with half-moon de fect of distal phalanx of mid finger. 1. Acroosteolysis with involvement of distal phalanges of all fingers. UCC 100615 JAMA, Aug 21, 1967 Vol 201, No S ACROOSTEOLYSIS--WILSON ET AL 579 distal phalanges, or a so-called slice effect along one or more of the tufts. Figure 2 illustrates the small, half-moon cut in the cortex of the distal phalanx of the mid-finger. 3. Advanced stage: A more severe lytic destruc tion may be a complete loss of the tuft and a por tion of the shaft of one or more distal phalanges as illustrated in Fig 1 in which the tuft of the distal phalanx of the right thumb and the tuft of the distal phalanx of the left fifth finger are completely absent. Additionally, there may be in the same hand a portion of the distal rim of the tuft remaining, with loss of the proximal portion of the tuft and a portion of the shaft of the phalanx, which can also be identified in the remaining phalanges in Fig 1, This loss may be of a transverse nature through the shaft and the tuft, or of an oblique type loss of bone structure. 4. Healing stage: In this phase, there is often definite fragmentation of the remaining tuft and of the filled-in area where the previous destruction was noted through the shaft of the distal phalanx. This may go on to a complete bony union or remain as a fibrous union with fragmentation. In Fig 3 and 4, x-ray films of the same individual well demon strate this. This employee had almost complete reunion of the multiple fragments when first seen in November 1965 and again in November 1966 (Fig 3). In the latter, the completely healed shafts and tufts of the distal phalanges are indicated by no residual fragmentation with fibrous union. There is a definite shortening of the shaft and a widening of 3. Above, Acroosteolysis with marked fragmentation of distal phalanges in November 1965. Below, Same individu al in November 1966. 8 JAMA, Aug 21, 1967 Vol 201, No 8 UCC 100616 580 ACR00STE0LYSIS--WILSON ET AL the shaft and tuft, both in the transverse and the anteroposterior diameter. This is most likely due to a combination of constant pressure, and by the nor mal tension of the soft tissues, particularly the ten dons to the distal phalanges. Occupational Aspects of Syndrome Based upon our observations of these 31 cases, it appears as if this syndrome may be of occupational origin and is somehow related to the process of vi nyl chloride polymerization. Its specific cause if not presently known. We are performing extensive re search in an effort to find the cause. Two other pa pers referring to the condition have appeared in the literature. The publication by Suciu et al3 contains no specific information, and merely alludes to some hand problem. That of Cordier et al4 presents case histories with symptoms similar to many of the cases we describe. This syndrome differs from idiopathic and familial acroosteolysis in that only the hands are involved. Vinyl Chloride Polymerization Process Polyvinyl chloride is a widely used synthetic resin. It has been manufactured commercially for more than 30 years and is used in upholstery fabric, floor and wall tile, wire insulation, phonograph records, and many other commonly used commodities. For many of these uses, the resin (CH^CHCD, is mixed with other materials to achieve the desired physical characteristics. The hand syndrome occurs apparent ly only in those people exposed to vinyl chloride or to other chemicals used in the manufacturing process of the resin itself or both. In addition to our exami Table 2.--Age Distribution of Cases Acroosteolysis Without Age Group Raynaud's Symptoms 20-29 1 30-39 4 40-49 4 Acroosteolysis With Raynaud's Symptoms 4 12 6 nations of 3,000 personnel performing vinyl chloride manufacturing and polymerization, we have exam ined more than 1,000 individuals who handle the finished resin or who process it into plastic prodducts. No cases of acroosteolysis have been found in these 1,000 persons. Basically, the manufacture of polyvinyl chloride consists of polymerizing vinyl chloride. The reac tion is accomplished in closed containers (polymerizers) with suitable catalysts and emulsifiers. Copolymers, formed by combining vinyl chloride with other monomers, create variations of the homopolymer. These are commercially produced. Following polymerization, the resin is washed, dried, and sold as a finely divided white powder. The polymerization operations are carried out in closed processes and provide little opportunity for employee exposure. Following the completion of the polymerization reaction, periodic cleaning of the walls and agitator of the polymerizer is necessary. The frequency of this cleaning and its method varies with the type of material used in these vessels and with different manufacturers. The most common practice has been to accomplish the cleaning man ually by using hand scraping techniques, with work ers spending several hours each day on this job assignment. Personnel performing this job are com monly referred to as "polycleaners." UCC 100617 JAMA, Aug 21, 1967 Vol 201, No 8 ACROOSTEOLYSIS--WILSON ET AL 581 Epidemiology We have attempted to study the relationship of job history to the occurrence of the disease. Twen ty-seven of our 31 cases have either been on the "polycleaner" job assignment at the time the syn drome appeared or have had that assignment at some time in the past. This job assignment is the only one showing any positive correlation with the occurrence of the syndrome. Attack Rate.--The syndrome has a low attack rate. Our experience indicates it occurs only in less than 3% of all production employees who at one time have had polycleaning experience. Age Distribution.--The youngest of our cases is 26 years old, and the oldest is 47, with the majority falling in the 30 to 39 age group. The complete dis tribution is shown in Table 2. The appearance of the syndrome among younger employees may be influenced by the fact that the polycleaning job is one of the initial job assign ments into which employees in such plants are hired. Incubation Period.--It this syndrome is related to occupational factors, as we believe, then the time of exposure to these factors should be significant. We have investigated the time spent on polyclean ing and, although accurate job-time assignment in formation is difficult to develop, it appears as if none of these cases has had less than 12 months poly cleaning experience. Comment To our knowledge, this is a unique and, with the exception of references 3 and 4, previously unre ported disorder. The specific cause is presently unknown, although it appears to be related to the manufacture of vinyl chloride and polyvinyl chlo ride. Not only are the x-ray findings of themselves unique, but when accompanied by the symptoms of Raynaud's phenomenon, the syndrome becomes ex tremely specific. As far as we are aware, this has not been an observed response to any toxicant in any of the animal species. We have attempted to arrive at an explanation of its cause, as well as the physio logical mechanism whereby the extreme specificity for the distal phalanges of the hands occurs, but have been unsuccessful. We believe the condition is .the result of three factors, all of which must be present for occurrence: (1) a chemical insult, (2) a physical insult, and (3) a personal idiosyncrasy. The chemical insult could occur from one or more of the monomers, catalysts, and intermediate reaction products existing in polymerizers. A low degree of exposure to these could occur from contact with the solid, slightly moist, residue in the polymerizer or to small quantities of vapor, absorbed either percutaneously or by inhalation. Research studies are in progress in an attempt to verify the chemical in sult factor theory. The physical insult is present in all "polyclean ers" to some degree through the prolonged hand scraping operations as well as the occasional use of hammers to remove the residues. In support of the physical insult factor, we have quite recently ob served the effects of a finger injury to an existing case of this syndrome. Figure 4 (left) shows an x-ray film of the left hand of the first of these cases of acroosteolysis (accompanied by Raynaud's symp toms and skin nodules on the dorsal surfaces of the hands). The bone damage is quite limited. Figure 4 (right) shows the same hand with the roentgeno gram taken four months after a crushing injury to the mid finger, with lysis having resulted to the tuft of the distal phalanx. We believe this activity was stimulated by the trauma. Personal idiosyncrasy appears to be an important factor because of the low incidence of occurrence of the disorder. This is especially significant because, although all polycleaners are subjected to essential ly similar chemical and physical insults, the inci dence of this syndrome is very low, and the explana tion for this can only be made on the basis of person al idiosyncrasy. We suspect pertinent factors here are related to the individual's vascular system, the nerves controlling the blood supply to the fingers, and to the specific type of collagen in the individ ual's hands. We are in the process of investigating these factors. We have observed no serious disability in any of these cases. A few have been partially disabled be cause of hand soreness, to the extent that some restriction in manual activity was necessary. Im provement in the symptoms, as well as in the roent genological findings, has occurred in many cases without adequate explanation. We wish to emphasize that no cases have been found, after extensive search, in individuals either working with the finished polyvinyl chloride or its copolymers, or in processing the polymer into plastic products. In these processes, more exposure to the polymer occurs than does in the manufacturing of the polymers themselves. We presently believe that personnel assigned to polycleaning should be evaluated, prior to assign ment, for any evidence of collagen disease, oste olysis of the hands, or abnormal response of the hands to cold insult. Any evidence of the existence of any of these factors should contraindicate the as signment of an individual to "polycleaning," and thus remove, or at least minimize, the personal idio syncrasy factor. References 1. Cheney, W.D.: Aero-osteolysis, Amer J Roentgen 94:595-617 (July) 1965* 2. Andren, L., et al; Osteopetrosis Acroosteolytica: Syndrome of Osteopetrosis, Acroosteolysis and Open Sutures of Skull, Acta Chir Scand 124:496-507 (Dec) 1962* 3. duciu, 1.; Drejman, I.; and Valaskai, M.: Contribute Studiul Imolnavirilor Produse de Clorura de Vinil, Med Intern 15:967978 (Aug) 1963. 4. Cordier, J.M., et al: Acroosteolyse et Lesides Cutanees Associees Chez Deux Ouvriers, Affecfces au Nettoyage D'AntocIaves, Cah Med Travail 4: (Jan) 1966. JAMA, Aug 21, 1967 Vol 201, No 8 UCC 100618