Comparison of radiologic and gross examination for detection of cancer in defleshed skeletons.код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 96:357-363 (1995) Comparison of Radiologic and Gross Examination for Detection of Cancer in Defleshed Skeletons BRUCE M. ROTHSCHILD AND CHRISTINE ROTHSCHILD Arthritis Center of Northeast Ohio, Youngstown, Ohio 44512; Department of Medicine, Northeastern Ohio Uniuersities College of Medicine, Rootstown, Ohio 44272; Department of Biomedical Engineering, University of Akron, Akron, Ohio 44242; The Carnegie Museum, Pittsburgh, Pennsylvania 15213; and Kansas Museum of Natural History, Lawrence, Kansas 66045 KEY WORDS X-ray, Skeletal pathology, Cancer, Metastasis ABSTRACT The reliability of visual examination of defleshed bones was assessed for detection of postcranial metastatic disease in individuals known to have had cancer. This was compared with standard clinical radiologic techniques. The skeletons of 128 diagnosed cancer patients from an early 20th century autopsied skeletal collection (Hamann-Todd Collection) were examined. Radiologic examination detected evidence of metastatic disease in 33 individuals, compared to 11 by visual examination of the postcranial skeletons. Four of these cases were detected by both techniques. Blastic lesions were most commonly overlooked on visual examination, because they were localized to trabecular (internal bone) structures. The ilium was the most commonly affected bone, with lytic or blastic lesions detected in 30 of 33 individuals. While the proximal femur was affected in only nine individuals, x-ray of the proximal femur and ilium detected all individuals with postcranial evidence of metastatic disease. Skeletal distribution of metastases provides no clue to the location of origin or histologic subtype of the cancer. Survey of archeological human remains for metastatic cancer requires radiologic examination. Such skeletal surveys should x-ray at least the ilia and femora. 0 1995 Wiley-Liss, Inc. Because bone is secondary only to lung and liver as a site of metastasis (Alvarez, 1948; Abrams et al., 1950; Wierman et al., 19541, study of skeletal populations should be a reasonable approach to determining the presence of cancer in antiquity. The relationship of malignancy to environmental conditions and occupations allows the anthropologist another tool with which to further characterize prehistoric living conditions. Presence and extent of cancer provides insights to support networks (familial and community-based). Metastatic cancer has, however, only sporadically been recognized in archeologic human remains (Allison et al., 1980; Grupe, 1988; Manchester, 1983). Such reports were premised on initial visual recognition of pathology. Systematic archeologic human radiology does 0 1995 WILEY-LISS. INC not appear to have been pursued. Given the critical role of skeletal x-rays in clinical evaluation for presence of metastatic disease, it seemed appropriate to radiologically assess a well-defined skeletal population with cancer. Metastatic cancer typically appears as multiple holes (lytic areas) or areas of increased density (blastic areas) of quite variable size (Huvos, 1991; Kim and Barry, 1964; Resnick and Niwayama, 1988). Although metastatic disease has been reported in archeological human remains Received February 21,1994; accepted October 19,1994. Address reprint requests to Bruce M. Rothschild, Arthritis Center of Northeast Ohio, 5701 Market Street, Youngstown,OH 44512. 358 B.M.ROTHSCHILD AND C. ROTHSCHILD (Ortner and Putschar, 1985; Steinbock, 1976; Tkocz and Bierring, 1984), findings have been predicated upon visual recognition (usually with subsequent radiologic confirmation), The sensitivity of visual examination of defleshed bones for detection of postcranial metastatic disease has not been previously assessed. The current study was initiated t o determine the relative sensitivity of visual and radiologic examination for detection of metastatic cancer. Given the expense of radiologic evaluation, it was further desired to identify which bones must be x-rayed (to assure that all individuals in the population with metastatic disease would be identified). The skeletons of diagnosed cancer patients from an early 20th century, autopsied skeletal collection (Rothschild and Woods, 1991) were therefore examined for evidence of visible and radiologic skeletal alterations. METHODS Defleshed, postcranial skeletal remains of 129 autopsied individuals with documented cancer were examined. These were identified from the Hamann-Todd Collection (Cleveland Museum of Natural History) compiled by T. Wingate Todd between 1913 and 1933. The skeletons (484 women, 2422 men) were derived from cadavers of unclaimed status which, after autopsy (including histologic study), were subjected to defleshing (with sodium hydroxide to remove soft tissues). The medical condition and cause of death records attended most of the individuals. The Hamann-Todd Collection is the world's largest and best preserved collection of human skeletons for which a demographic background is known (Rothschild and Woods, 1991; Rothschild and Martin, 1993). The representatives of this collection has been documented in multiple studies (Mensforth and Latimer, 1989; Rothschild and Woods, 1991; Rothschild et al., 1990; Shaibani et al., 1991). Age and sex noted in the medical records were confirmed, as described previously (Rothschild et al., 1990). The skeletal remains were examined visually t o identify all occurrences of osseous alterations, to specify the types of bony alterations a t each occurrence, and to map the distribution of occur- rences in each skeleton. Lesions were distinguished from artifact, as previously described (Rothschild et al., 1990). In the event of disagreement as to whether a lesion represented erosion or artifact, for the purpose of this study it was treated as artifact. Radiographs and fluoroscopy of the entire skeleton of each individual were obtained with the bones arranged and oriented approximately as they would be in living patients. Fluoroscopy of all postcranial skeletons was performed using the Xi Scan 1000 fluoroscopy/digital storage system (X-Tech Inc., Randolph, NJ). All radiologic evaluation was performed in a blinded manner, without knowledge of the presence or absence of visible evidence of metastatic disease in any other portion of the skeleton. One individual placedmanipulated the bone(s), while the other solely viewed the resulting image. The Chi-square statistic was utilized to assess if there was a significant difference in recognition of metastatic osseous disease by visual examination and radiologic techniques. The representativeness of the individuals evaluated was also investigated by comparison of frequency and distribution with that reported in the literature (Chi square, P < 0.05). RESULTS Demographics One hundred twenty-nine individuals (4.3%)in the Hamann-Todd Collection had diagnosed cancer. Ninety-eight men (4.0%of the total Hamann-Todd Collection human male component) and 30 women (6.2%of the female component) had diagnosed cancer, for a female predominance of 1.51. The ratio was calculated on the basis of percentage of women affected (from the 484 at risk [number of female skeletons represented]) versus the percentage of men affected (from the 2,422 at risk). The average ages of affected men and women were also indistinguishable (61.0 and 60.3, respectively). Radiologic-visualcomparisons Radiologic examination revealed cancer in 33 individuals, compared to 11 individuals by visual examination. The detection rate of radiologic examination was three times that of visual examination (Chi-square, RADIOLOGIC DETECTION OF METASTATIC CANCER Fig. 1. Lytic metastatic ilia and rib lesions. A Lytic iliac lesions with un-sharpborders (Hamann-Todd Collection No. 2645). B Lytic iliac lesions with more sharply defined borders (Hamann-Todd Collection No. 788). C: Lytic rib lesions (Hamann-Todd Collection No. 788). 359 B.M. ROTHSCHILD AND C. ROTHSCHILD 360 Fig. 2. Blastic metastatic ilia and rib lesions. A: Pure radiodense (increased blockage of x-ray beam, producing dark image on x-ray), blastic rib lesion (Hamann-Todd Collection No. 1670). B: Radiodense, blastic iliac metastases with associated lytic involvement (Hamann-ToddCollection No. 217). TABLE I . Prevalence (number (percent7) of grossly visible and radiologically detected mefastatic disease in Hamann-Todd Collection cancer individuals Tumor Clinically diaenosed Stomach Lung Colon hostate Breast Esophagus Pancreas Urethrabladder Uteruslcervix Undetermined I Not 32 8 7 3 8 5 5 6 10 35 All 1 (3) 3 (38) 0 1 (33) 1 (13) 0 0 1 (33) 1 (10) 3 (9) Visual detection Only' Total 0 3 0 0 1 0 0 1 1 7 9 (28) 3 (38) 2 (29) 1 (33) 2 (25) 3 (60) 2 (40) 0 0 12 (34) Radiologic detection Blastic Lvtic 4 5 0 1 3 0 0 1 2 2 1 2 0 1 0 0 4 - Total detection 1 0 8 recognized radiologically. P < 0.001). Radiologically, the lesions in 23 individuals were lytic (Fig. 1)in nature, while 12 were blastic, occasionally with mixed patterns (Fig. 2). Lesions were irregular in outline (withoutdiscernable pattern). Metastatic cancer was detected in four individuals by both techniques (visible examination and x-ray).The frequency, percentage of cancer occurrences by site of cancer origin, and type of radiologic reaction are delineated in Table 1. The site of origin or histologic subtype of the cancer was not identified in 26% of individuals with metastatic bone disease in the Hamann-Todd Collection. The distribution of skeletal involvement is listed in Table 2. Vertebral lesions were less frequently detected radiologicallythan visually. No relationship was noted between distribution of metastatic disease and the location of the initial cancer or its histologic RADIOLOGIC DETECTION OF METASTATIC CANCER TABLE 2. Distribution of visible and radiologic skeletal metastases in the Humann-Todd Collection and clinical literature Site Vertebrae Pelvis Ribs Femur Tibia Scapula Humerus Radius Clavicle X-ray Number Percent 4 30 5 9 1 2 3 1 1 Visible (%) Literature' 12 91 15 27 3 45 64 21 69 6 9 18 18 3 3 0 0 36 9 41 25 25 3 6 10 0 0 361 nation, because most appear localized to trabecular structures. Visual and radiologic examination are complementary. We suggest, therefore, that study of cancer in the skeletal record should use both approaches. Because ilia or proximal femur were concommittently affected (radiologically) in all individuals with any evidence of metastatic disease (radiologically), it would seem that at a minimum, any metastatic disease survey should include their x-rays. Representatives of findings Radiologic evidence of metastatic bone disease in 26% of the Hamann-Todd Collecsubtype. Ninety-one percent of recognized tion falls well within the 7-27% range reradiologic metastases were identified on ported by Copeland (1964). The detection x-ray of the ilia. Femoral metastases were frequencies by combined visible and radioradiologically detected in four individuals logic examination were also indistinguishwho had no evidence of metastatic disease able for the various tumors (Chi square, on pelvic x-rays. All individuals with radio- P > 0.05) from those reported from clinical logically detected metastases had involve- and autopsy studies (Clain, 1965; Tubianament of the ilia and/or femora. Hulin, 1991). The pattern of distribution of metastatic disease (Table 2) also proved inDISCUSSION distinguishable from that reported in the litOnly 33 of 128 Hamann-Todd Collection erature (Clain, 1965). Thus, the representaindividuals with diagnosed cancer had ra- tiveness of the Hamann-Todd Collection diologically recognizable skeletal alter- (Mensforth and Latimer, 1989; Rothschild ations. The detection rate of radiologic ex- and Woods, 1991; Rothschild et al., 1990; amination was three times that of visual Shaibani et al., 1991)is further documented. Two apparent discrepancies between clinexamination (Chi square, P < 0.001). Although Ortner and Putschar (1985) sup- ical and radiologic detection of cancer in deported visual inspection for detection of fleshed bones must be addressed. Clinical presence of cancer in osseous remains, their localization of vertebral metastases is freconjecture was not based on systematic skel- quently based on isotope bone scans, rather etal radiology. Their radiologic evaluation than x-rays. Bone scans are more sensitive was apparently limited to individuals (per- for detection of vertebral metastases (Resnick haps even limited just t o visually affected and Niwayama, 1988). Increased radiologic bones of individuals) with visual evidence of detection in the defleshed pelvis is probably cancer. The current study confirms their no- related to lack of interference by soft tissues tation that some visually detected lesions (as in the intact individual). Subtle changes were not seen on x-ray. Because there must Seen on x-ray of defleshed bones are o h n not be a 3050% increase or decrease of bone visualized in the intact specimen (Rothschild density before it can be recognized radiolog- et al., 1990). ically (Resnick and Niwayama, 1988; RothsRelationship of radiologic finding to child and Martin, 1993), it is not surprising variety of cancer that all visually detected lesions cannot be detected in this matter. No relationship was noted between distriHowever, if visual inspection is the only bution of metastatic disease and the location technique utilized, two thirds of metastatic of origin or histologic subtype of the cancer. disease would have been missed in the Ha- This corroborated the observations of Wilmann-Todd Collection. Metastatic blastic le- son and Calhoun (1981) and Morgan et al. sions especially are missed by visual exami- (1990) that the skeletal distribution of sites 'Chin. 1965. 362 B.M. ROTHSCHILD AND C. ROTHSCHILD of metastasis was independent of tumor histologic type or tissue of origin. The skeletal distribution of metastases provides no clue to the site of origin or histologic subtype of the cancer. archeologic populations. As this can be accomplished using one 14 by 17-inch x-ray plate to obtain a single anterior-posterior view, this should meet the need for a costeffective approach. Differential diagnosis ACKNOWLEDGMENTS We acknowledge the kindness of Drs. Bruce Latimer and Lyman Jellema and the Cleveland Museum of Natural History for cogent comments and for logistical and technical support in examining The HamannTodd Collection, and David Vargo of %-Scan for providing the fluoroscopy equipment. Distinguishing the lytic lesions of cancer from those of multiple myeloma (plasma cell dyscrasia), lymphomas, and hematologic malignancies is another subject for future investigation. The lesions of multiple myeloma are typically considered to produce sharply defined lytic or (rarely) blastic lesions, all comparable in size (Resnick and Niwayama, 1988). The lesions of metastatic cancer are generally thought to blend with surrounding bone, with great variability of size of individual metastases in a single individual (Resnick and Niwayama, 1988). However, notation in diagnosed multiple myeloma of lesions of variable size (unpublished observations) and of “punched out” (sharply defined) lesions in diagnosed metastatic cancer preclude its use to distinguish between the two phenomena. Presence of lytic lesions in defleshed skeletons allows consideration of metastatic cancer, but hematologic malignancies, multiple myeloma, lymphomatous disorders, granulomatous infections (e.g., fungal, mycobacterial [such as tuberculosisl, brucellosis), and even treponemal disease must also be considered (Resnick and Niwayama, 1988; Rothschild and Martin, 1993; Rothschild and Turnbull, 1987). The latter are typically associated with characteristic patterns of periosteal reaction, which should allow easy discrimination (Rothschild and Rothschild, 1994). Implications Radiologic examination must accompany visual examination of skeletons if the epidemiology of metastatic disease is to be reliably determined in archeologic human remains. The expense of total skeleton radiologic surveys must be overcome for routine radiologic examination of the entire skeleton of archeologic human remains to become practical. X-ray of ilia and femora (Table 2) would seem the minimum required to detect presence of metastatic disease in LITERATURE CITED Abrams HL, Spiro R, and Goldstein N (1950) Metastases in carcinomas: Analysis of 1000 autopsied cases. Cancer 3: 74-85. Allison MJ, Gerszten E, Munizaga J, and Santoro C (1980) Metastatic tumor of bone in a Tiahuanaco female. Bull. N. Y. Acad. Sci. 56:581-587. Alvarez GH (1948) Clinica del cancer de pulmon. Rev. Assoc. Med. Argent. 62:690-694. Clain A (1965) Secondary malignant disease of bone. Brit. J . Cancer 19:15-29. Copeland MM (1964) Treatment of metastases to bone from primary tumors in other sites. Tumors Soft Somatic Tissues Bone 8502-516. Grupe G (1988) Metastasizing carcinoma in a Medieval skeleton: Differential diagnosis and etiology. Am. J . Phys. Anthropol. 75r369-374. Huvos AG (1991) Bone Tumors: Diagnosis, Treatment, and Prognosis, 2nd ed. Philadelphia: Saunders. Kim SK, and Barry WF Jr (1964)Bone Island. AJR Am. J. Roentgenol. 92:1301-1306. Lawrence W Jr (1986) Gastric Cancer. CA 36:216-236. Manchester K (1983) Secondary cancer in an AngloSaxon female. J. Arch. Sci. 10:475-482. Mensforth RP, and Latimer BM (1989) Hamann-Todd Collection aging studies: Osteoporosis fracture-syndrome. Am. J. Phys. Anthropol. 80:461479. Morgan JW, Adcock LA, and Donohue RE (1990) Distribution of skeletal metastases in prostatic and lung cancer. Urology 36:31-34. Ortner DJ, and Putschar WG (1985) Identification of Pathological Conditions in Human Skeletal Remains. Washington: Smithsonian Institution Press. Resnick D, and Niwayama G (1988) Diagnosis of Bone and Joint Disorders, 2nd ed. Philadelphia: Saunders. Rothschild BM, and Martin L (1993) Paleopathology: Disease in the Fossil Record. London: CRC Press. Rothschild BM, and Rothschild C (1994) Syphilis, Yaws and Bejel: Definitive differential diagnosis by population criteria. Am. J . Phys. Anthropol. 94:174. Rothschild BM, and Turnbull W (1987) Treponemal infection in a Pleistocene bear. Nature 329:61-62. Rothschild BM, and Woods RJ (1991) Spondyloarthropathy: Erosive arthritis in representative defleshed bones. Am. J . Phys. Anthropol. 85:125134. RADIOLOGIC DETECTION OF METASTATIC CANCER Rothschild BM, Woods RJ,and Ortel W (1990)Rheumatoid arthritis in the buff: Erosive arthritis in representative defleshed bones. Am. J. Phys. Anthmpol. 82: 441-449. Shaibani A, Workman R, and Rothschild B (1991)Enthesitis as a disease manifestation. Am. J. Phys. Anthropol. (Suppl.) 12:159-160. Steinbock RT (1976)Paleopathological Diagnosis and Interpretation. Springfield: Thomas, pp. 385-397. Tkocz I, and Biemng F (1984)A Medieval case of metastasizing carcinoma with multiple osteosclerotic bone lesions. Am. J. Phys. Anthropol. 65373-380. 363 Tubiana-Hulin M (1991)Incidence, prevalence and distribution of bone metastases. Bone 12 (Suppl. I):S9s10. Wierman WH, Clagett OT, and McDonald JR (1954) Articular manifestations in pulmonary diseases. An analysis of their occurrence in 1024 cases in which pulmonary resection was performed. J.A.M.A. 155: 1459-1463. Wilson MA,Calhoun Fw (1981)The distribution of skeletal metastases in breast and pulmonary cancer. J. Nuel. Med. 22.594497.