Document ZYL3wwNnw6pQeZZ4knLRaR0L

) fc// rends.> study. Neoplasms of the central nervous system Epidemiologic considerations Alan K. Percy, M.D., Lila R. Elveback, Ph.D., Haruo Okazaki, M.D., and Leonard T. Kurland, M.D. This paper reports the incidence and sur vivorship for various neoplasms affecting the central nervous system (CNS) in the resident population of Rochester, Minn., for the thirtvfour-year period 1935 through 1968. It is desirable to have such statistics for various populations. Information of this type might disclose trends and population selection of dis ease which could provide relevant clues to etiology. Also, where comparisons can be ex tended to regional or national populations, such incidence data would be of value in plan ning for research and the provision of medical care. Most of the earlier reports dealing with the frequency of neoplasms affecting the central nervous system have been based on the ex perience of individual neurosurgeons, autopsy series, or proportionate rates of hospital ad missions. While providing valuable descrip tions of clinical or pathologic experiences, these data do not necessarily reflect an accu rate picture of the incidence of neoplasms or their frequency by type in the population. Since the early part of this century, the medical practice in Rochester has been cen tered mainly at the Mayo Clinic. For several decades, approximately one-half of the Roch ester population has been seen annually at one of the Mayo Clinic facilities for the diag nosis or treatment of minor or serious acci dents or illnesses or for routine examinations. Diagnoses made by clinic physicians, whether at the clinic, in affiliated hospitals, or at autop- 40 Naurology/Voluma 22/January 1972 sy, have been entered on a master sheet of the patient's record, cross-indexed, sind processed for automated retrieval. In 1958, Kurland1 presented the results of an earlier surveV in which it was noted that primary intracranial tumors represented 1$ of the deaths in the Rochester population during the years 1945 through 1954. That survey, one of the first based on a defined population, yielded incidence rates for neoplasms of the central nervous sjstem that exceeded the rates from all other population studies. The age- f specific incidence rate of primary brain neo plasms was noted to increase dramaticallv with increasing age. However, the population and the number of cases ascertained in Kurljind's survey were admittedlv small, perhaps allow ing for chance fluctuations of some magnitude. The current investigation enlarges the studv period to include all cases available through the enlarged diagnostic index of the Rochester epidemiology program and to explore possible From the Department of Medical Statistics, Epidemiology, arid Population Genetics and the Department of Experi mental and Anatomic Pathology, M.i\o Clinic .md Mayo Foundation, Rochester, Minn. lit. Pvn-y wni the recipient of Teacher-Investigator Award NS-11,0 HI. Tins work was xuptiorlrd in part l>v h-m .im li gr.ittl (JW14231 (mm the National Institute* of Health, Public Health Sen ice. Submitted for publication April 5, 1971; accepted April 12. 1971. Reprint requetts to Section of Publication*. Ma\o Clinic, Roche*ter, Minn. S590I. XIETHO. The and aiu resident vous diagno*.' A resiili .home ester foi of the <i indivulu tate dtusvruptmi . ' All < or mel.i tumors spinal plasms few i,e- the C\ cla'ssifii or prob. logic ot servatu-- All va denth i to the i' diagnos. neurol", slides ol>' f\ the v1 cation. Incuh dale ul the tince I'riu'i.iis i'rtm.irv rttnU.ll \ I nti.ii \ i Mi-t.istal! Total Total 1 M OLiI 6578 t of the i ocessed Milts of (I that 1? of > 'luring . one ion, the he rates i lie age- 111 nVo,lh with mti and .inland's s allowrnitude. e studvthrough ochester possible ctnioJojfy. t Expert* nd Mavo r Award 'int CMk Health April Clinic, NEOPLASMS OF CENTRAL NERVOUS SYSTEM 41 trends during the nearly thirty-five vears under studv. METHODS The review of records of the Mavo Clinic and ancillary resources was confined to those residents of Rochester in whom a central ner vous system neoplasm was suggested on the diagnostic cross-index from 1935 through 1968. A resident was defined as any individual whose home was within the physical limits of Roch ester for at least one year prior to the diagnosis of the disease. Care was taken to exclude any individual who moved to Rochester to facili tate diagnosis or treatment after the onset of jvmptoms. All cases were categorized as being primary or metastatic lesions. The cases of primary tumors were subdivided as neoplasms of the spin^ cord, pituitary, or brain. Primary neo plasms of the nerve roots and meninges and a few neoplasms from adjacent tissues affecting the CNS were also included. All cases were classified as definite (confirmed histologically) or probable (strongly suspected on roentgeno logic or clinical grounds alone or bv direct ob servation of an otherwise inoperable lesion). All case abstracts were reviewed indepen dently to establish proper placement according to the above classification. Almost all clinical diagnoses had been made by Mayo Clinic staff neurologists. One of us (H.0.) reviewed all slides obtained at operation or autopsy to veri fy the diagnosis and provide histologic classifi cation. Incidence rates per 100.000 population by date of diagnosis were calculated for each of the three decades from 1935 through 1964 and for the four years from 1965 through 1968 by using the number of cases in the numerator and the respective decennial census in the denominator. For the period 1965 through 1968, an estimate of the population was based on a special 1965 census. RESULTS During the thirty-four-year period of this study, 297 cases of intracranial and intraspina] neoplasms were identified among the resident population of Rochester. One hundred seventyfour patients had primary neoplasms, including 1 patient with 2 histologically distinct primary neoplasms. The remaining 123 patients had metastatic tumors to the brain or spinal cord, including 4 patients who also had a primary tumor of the brain. Of the 686 cases reviewed, 389 were not included because of nonresidency (131), neoplasm not of the central nervous system (256), or onset prior to 1935 (2). The numbers of primary neoplasms of the brain, pituitary, and spinal cord, as well as the number of metastatic lesions, are presented in Table 1 by sex for each of the study periods. Of the 174 primary neoplasms, 93% were his tologically confirmed and the remainder were diagnosed by direct visualization or by roent genologic evidence. If the pituitary neoplasms, which are often diagnosed on the basis ob clini cal and roentgenologic evidence, are excluded, 95% of the diagnoses of primary neoplasms of the brain and spinal cord were established histologically. Two of the neoplasms lacking histologic confirmation were observed directly at operation but no biopsy could be obtained. In this study, 63 (36%) of the 174 primary central nervous system neoplasms were first TABLE 1 NEOPLASMS OF THE CENTRAL NERVOUS SYSTEM BY SITE, SEX, AND STUDY PERIOD, ROCHESTER, MINN.. 1935-1968 typ* Primary in brain, certain 1935-19*4 MF ---- 9-- 67 Primary in brain, probable 0 0 Pituitary Primary in spinal cord 31 2z Metastatic 48 Total 15 18 Total (M and F) 33 1945-2954 MF 9 23 01 51 o4 9 24 25 53 78 1955-2964 MF 31 36 01 45 20 24 39 61 81 142 2965-2968 MF 9 12 13 20 11 96 22 22 44 1935-1965 MF 55 78 15 14 7 77 46 77 123 174 297 'I ( 42 NEUROLOGY diagnosed at autopsy (Table 2). In 24 (38?) of the 63 cases, there were some neurologic symptoms or signs or both but the diagnosis had been other than neoplasm. In 2 cases, autopsy vvas performed for medicolegal rea sons after automobile accidents. In the re maining 37 cases there were no symptoms. These 63 cases included 36 meningiomas, 12 gliomas, and 15 other histologic types; the ratio of meningiomas to gliomas was 3:1. Among those lesions diagnosed while the pa tient was living, there were 25 meningiomas. 37 gliomas, and 48 others; the ratio of me ningiomas to gliomas was 0.7:1. The ratio of all mlnigiomas (61) to all gliomas (49) in this study was 1.2:1. These ratios contrast with those from neurosurgical and hospital series in which gliomas usually outnumber meningi omas 2:1 or 3:l.2"4 The methods and level of case ascertainment in the present stud>' are be lieved to provide a* more realistic frequency distribution of central nervous system neo plasms by type. Eighty percent of the 174 primary central nervous system neoplasms were intracranial, and 80? of the intracranial lesions were me ningiomas or astrocytomas (Table 3). The 4 intracranial sarcomas represented 2? of the total, in good agreement with the 3? found in a surgical series of 8,000 intracranial neo plasms at the Mayo Clinic.* Chromophobe adenomas were the most common primary tumors involving the pituitary, and schwan nomas were the most common type in the spinal cord. Incidence rates are presented in Table 4. The rates for neoplasms of the pituitarv were 1.6, 2.1, and 2.3 for the first three decades, 1.1 TABLE 2 PRIMARY NEOPLASMS OF THE CENTRAL NERVOUS SYSTEM DIAGNOSED AT AUTOPSY, ROCHESTER, MINN.. 1935-1968 Period Tntttl Xo. of neofdfunu .Vcniifoiim diiwnmtd at (lutO)iiy With nrumli Total tymtUomt or. 1935-1944 1945-1954 1955-1964 1965-1968 1935-1968 21 45 79 29 174 7 14 32 10 63 3 4 14 3 24 TABLE 3 PRIMARY NEOPLASMS* OF THE CENTRAL NERVOUS SYSTEM BY TYPE. ROCHESTER, MINN.. 1935-1988 Tyiw Pfftn* Xumber of totd Spinal cord Schwannoma Ependvmomf Meningioma Chondroblastoma * Melanoma Teratoma 14 6 3 2 1 1* l 1935-1944 1945-1954 1955-1961 1965-1968 1935-1968 * Pituitary Chromophobe adenoma Craniopharyngioma Eosinophilic adenoma Unknown 21 12 0-24 11 4 # 25-44 ,45-64 2 565 4 Intracranial Meningioma Astrocytoma. Schwannoma Other Ependymoma Sarcoma Retinoblastoma* Chordqma* Lipoma Medulloblastoma Epidermoid Hemangiopericytoma Dermoid or paraphyseal evst Unknown 61 49 9 20 4 4 2 2 2 2 1 1 1 1 35 28 5. 12 * Includes selected neoplasms of adjacent tissues affectmi CNS for the last four vears, and 1.9 for the entire thirtv-four-vear period; thev do not differ sig- ` uificantlv. For neoplasms primarv in the spinal cord, the rates were 1.6, 2.1, and 0.5 for the first three decades, 1.1 for the last four vears, and 1.3 for the entire period; these also do not differ significantlv. The rates shown in Table 4 were lower during the first decade than during succeeding periods; this could be due to underdiagnosis. Diagnostic techniques and neurosurgical procedures were still being tie vcloped, and it is unlikely that they were being fully utilized by the population at all ages. Furthermore, autopsy rates, particularly the percentage including central nervous system examinations, were lower in the first period than in later vears. The r. not sign'' Vious tw when tin all diagi plasms |. diagnostthe recu'. tral nei diagnn-iuntil tin autopss for 196" of enrlii". currently group--lb* undo go lesions. 1 four-vear be sntfki. logic ts p. other anal Age-spi 1968 also Significant' results ot Othfcr stu> fifth and subsequen* rates also and spinal toward in* FOLI 6580 neoplasms of central nervocs system 43 NERVOUS 115-1968 Percent of total A 12 TABLE 4 AVERACE ANNUAL INCIDENCE rates per 100,000 POPULATION FOR NEOPLASMS OF THE CENTRAL NERVOUS SYSTEM. ROCHESTER, MINN.. 1933-1968 ^^Jpnrnary in hfnin. ^ So. flufe CJ* ^Jfimary in CSS_, No. Bata Cl* ,__Metastatic to CVS^ So Rata Cl* - Tola! in CVS No. Bate Ct* 1935-1044 1945-1954 1955-1004 1985-1968 1935-1968 11 31 08 25 139 52 3-9 21 11.7 8-16 15 17,1 14-23 7!) 13.4 9-20 29 L2.5 11-15 174 By period 8.5 5-13 15.0 11-21 20,2 16-26 15.5 10-22 15.7 14-18 12 33 6.3 15 123 4.8 2-8 33 11.7 8-17 78 16.1 13-21 142 8.0 4-13 44 11.1 9-13 297 11.3 0-9 27,6 22-35 36.3 31-44 23.6 17-32 26.8 24-30 0-24 2544 45-64 85 By age for entire period 19 3.9 2-6 27 5.5 4-8 3 0.6 0-2 30 6.1 4-9 18 5.6 3-9 22 &9 4-10 17 5.3 3-9 39 12.3 9-17 39 18.9 14-26 53 25.7 20-34 64 31.1 2541 117 56.8 48-69 63 69.0 54-90 72 78.9 63-101 39 42.7 30-59 111 121.6 101-148 *95* confidence mterv.il for the Mte iRfCting ^ entire fer sigi spinal' for the : i ears, eke not Table, an durdue to and ng de being d ages. * Is.' the -\ stem period The rates for the last four-year period are not significantly lower than those in the pre vious two decades; however, it is likely that when this review was made early in 1969, not all diagnoses of central nervous system neo plasms for 1968 had been entered into the diagnostic index. Also, it is likely that with the recognition of svmptoms of a possible cen tral nervous svstem neoplasm, a definitive diagnosis would not be entered on the chart until the patient's operation or demise and autopsv or death certification. Since our rates for 1965 through 1968 are based on the date of earliest neurologic symptoms, that rate is eurrentlv deficient, especially in the metastatic group--these patients would be less likely to undergo operation than those with primary lesions. The deficiency in data entry for the four-vear period, however, is not believed to be sufficiently selective by age, sex, or histo logic tvpe to preclude use of these data in our other analyses. Age-specific incidence rates for 1935 through 1968 also are shown in Table 4.-The incidence of piimarv neoplasms of the brain increased signifieantlv with age, corresponding to the results of the previous Rochester survey.1 Other studies*-* have suggested a peak in the fifth and sixth decades and a decrease in the subsequent decades. Age-specific incidence rates also were computed for primary pituitary and spinal cord neoplasms. There was a trend towaid increase with age in each, although the number of cases was small and the stan dard error so great that no conclusion is war ranted. Incidence rates for metastatic neo plasms showed a similar increase with age-, however, the confidence limits are relatively large in the older age groups. Age-specific incidence rates were calculated for astrocytomas and meningiomas (Table 5). With each type of neoplasm, the age-specific incidence rate increased with age. In terms of sex-specific incidence, a sex difference was ob served with the meningiomas--the female: nfale ratio was 2:1. This is in agreement with pub lished statistics. Previous evidence, particu larly from Sweden7 and Israel, has suggested that astrocytomas occur more frequently in males. In this study the difference in rates (4.7 for males and 4.2 for females) is not sig nificant. TABLE 5 AGE-SPECIFIC INCIDENCE RATES PER 100,000 POPULATION FOR ASTROCYTOMAS AND MENINGIOMAS, ROCHESTER, MINN.. 1933-1988 Aga 0-24 2544 45-64 165 Total Astrocytoma .Vo. Rrtf# Vo, Bata 9 1.8 i 0.2 6 1.9 5 1.6 15 7.3 19 9.2 19 20.8 36 39.4 49 4.4 61 5,5 OLI 6581 i 46 NEUROLOGY 19 (17%) of 109 primary intracranial neoplasms were first discovered at autopsy; this is about one-half of the proportion noted in the present study. The clinical diagnoses in these 19 cases included trauma, toxic encephalopathy, sus pected infection, and behavior disorders. The importance of complete autopsy data in a pop ulation survey of central nervous system dis orders cannot be stressed too strongly. As it appears in most publications, the dis tribution of central nervous system neoplasms by histologic type is the result of a hospital series or a pathology collection in which the special interests or talents of the investigator can be expected to lend some bias to the statistics. Thus, metastatic tumors are infre quently encountered in most surgical series, and certain tumors occur with unusual fre quency, such as pituitary tumors in Cushing's report of his surgical experience. In the series reported by Cushing,2 Olivecrona,'1 and Ziilch,4 the relative occurrences were; gliomas, 43 to 53%; meningiomas, 13 to 19%; neurinomas,* 7 to 9%; pituitary adenomas, 7 to 18%; and metastases, 3 to 4%. Data from Israel," derived from a review of the Tumor Registry for 1961 through 1965, Yielded a somewhat different distribution: gliomas, 27%; meningiomas, 17%; neurofibromas, 12%; and pituitary adenomas, 5%. No statistics were reported for metastases. The Carlisle study*1* revealed 25% gliomas; metastases comprised 45% of all intracranial neoplasms. In the present study the distribu tion of primary neoplasms was; gliomas, 28%; meningiomas, 35%; schwannomas, 5%; and pituitarv adenomas, 7%. Metastatic tumors re presented 41% of the neoplasms. In the present study, the nearly complete histologic confirma tion allows for reasonable certainty of the diag nosis, and the large number of autopsy-dis covered neoplasms suggests more complete case ascertainment. The lack of histologic confirmation in oncthird of the Israeli cases" could explain some of the discrepancy by histologic type as com pared to the current study. However, the cur rent study was a population survey whereas the Israeli study had to rely on case reporting to a central registry, a situation that would favor the detection of malignant or sympto * In these series neurinomas were included to the glioma group. matic lesions over silent or indolent tumors. In addition, the Israeli survey included he mangiomas (2%). Twenty-three cases of he mangioma (vascular malformations) of the cen tral nervous system were identified in the Rochester study, .all but 1 at autopsy. Since these lesions were not regarded as true neo plasms, the cases were not included in the computations. It is difficult to believe that the marked dif ference between the rates in Rochester and in Connecticut (Table 7) reflects a true difference in disease incidence. It seems more reasonable that more complete case ascertainment related to the extensive diagnostic index and .high autopsy rate in Rochester accounts for most, if not all, of the difference noted.' The proportion of metastatic tumors, 41%, in the present .study is considerably higher than the 1 to 8% reported14 in neurosurgical series but lower than the 62.7% reported from chronic disease hospitals providing terminal care for patients with any type of malignancy. In the present study, it is likely that there is more underreporting of metastatic cerebral cancer than of the primary neoplasms of the brain. The metastatic cases were included only when involvement of the central nervous svsteni was specified on the diagnostic master sheet of the patient's record or in the autopsy protocol. In most instances, signs of cerebral dysfunction, focal seizures, or localized paresis had developed; it was not unusual that metas tasis was observed and recorded for the first time at autopsy. It seems leasonable that, if all brain neoplasms were correctly diagnosed, about 50% would be metastatic, as suggested bv Kurland.* In the earlier Rochester study,' there was a preponderance of females with primary neo plasms of the brain and meninges and with metastases. The present results, based on ap preciably larger ninnlicrs, indicate that females still more coinmnnh have meningiomas, but gliomas and most other neoplasms occurred with equal frequency in the two sexes. The present study substantiates one of the more interesting results of the previous Roch ester study*--that the age-specific incidence for primary neoplasms of the central nervous sys tem increases with age. Other studies of tumor incidence by age at diagnosis often failed to account i thus yu-l peaked specificuil ates's \\l continentdards of qnd metl. control lei' major snl tend to a not a con' ing ijitiaeprrespom Dorn an,' tration). Other loma'7 an studied u show tin vancing equivalcu for the < and, then including ability ol all ages Axerage ii` primary age, for H OLI 6582 r -i tumors, loci lienf In' ll io t'Clliii tlie \, Siiice ne neo in the ted difand in (Ference tsonable - related d high n most, n. 41%, higher 'irgieal i d from i. nniniil pliancy is .. oral - of the t d oiilv ms svs* master itopsv viebral I [uresis i *metasthe first* ...t, if all ignosed, iggtsted ie was a n'v neo-' md with l on apfemales nas, but occurred of the is Rochonce for (Ass svjif tumor to NEOPLASMS OF CENTRAL NERVOUS SYSTEM 47 account for the size of the population at risk, thus lielding the impression that incidence peaked at tin* fifth and sixth dorados. This is specifically pointed out liv Doll and associ ates1'1 who compiled age-specific rates for five continents. In those countries where the stan dards of medical practice are relatively high and methods of reporting are reasonably well controlled, the age-specific incidence rates for major subgroups and for all tumors combined tend to increase with age. However, this was not a consistent finding in the subgroup includ ing intracranial neoplasms in which the rates corresponded to the age-distribution found by Dom and Cutler10 in a ten-city study (see illus tration). Other neoplasms, such as multiple mye loma17 and breast cancer,10 have recently been studied in the Rochester population and also show this trend of increasing rates with ad vancing age. The Mayo Clinic provides the equivalent of general as well as specialty edre for the elderlv members of the tfommunitv, and, therefore, there is a greater likelihood of including older patients. This, plus the avail ability of diagnostic facilities for patients of all ages and the high autopsy rate in this com- munitv, may account for the difference be tween age-specific incidence rates in this study and those derived from hospital, surgical, or population snivels elsewhere. SUM M AltY Incidence rates were computed for central nervous system neoplasms for the resident population of Rochester, Minn., for the thirtyfour-year period 1935 through 1968. The rec ords of the Mayo Clinic were the major source of data for the cases; the ancillary rec ords in other medical facilities in and near Rochester could be identified and were avail able through the expanded diagnostic index of the Rochester epidemiology program. There were 174 primary and 123 metastatic CNS neoplasms. Ninety-five percent of the primaribrain and spinal cord neoplasms were histo logically confirmed. During the study period, 36% of the primary CNS neoplasms were first diagnosed at au topsy. Meningiomas accounted for more than one-half of this group. Average annual inci dence rates per 100,000 population were 12.5 for primary brain neoplasms, 1.9 for pituitary neoplasms, 1.3 for primary spinal cord eo- 70 - Rochesier, Minn., ovg annual 1945-54 inertas* Ttn metropolitan areas of US. 1947 / Average annual ineidence rates for primary neoplasms in brain, by age, for Rochester and other areas OLI 6583 48 NEUROLOGY plasms, and 11.1 for metastases to the CNS. No consistent trend over time was observed in the incidence rates for meningiomas, gliomas, or all primary CNS tumors combined. Age-specific incidence rates for primary brain neoplasms increased significantly with age; the rates for pituitary and spinal cord neo plasms showed a trend toward increase with age. No sex predilection could be found for primary brain, pituitary, or spinal cord or metastatic tumors. Meningiomas were found significantly more frequently in females. Five-year survival, determined bv the direct method, was 22% for gliomas^and 59% for meningiomas. REFERENCES 1. Kurland LT. The frequency of intracranial find intraspinal neoplasms in the resident population of Roches* ter, Minnesota, j Neurosufg 15:627-641, 1956 2. Cushing H: Intracnnial Turnon. Springfield III, Charles C Thomas, 1932 t 3. OH\corona H; The cerebellar Angioreticulomas. J NVurosurg 9 317-330, 1932 4. Zulch KJ: Brain Turnon, Their Biology and Pathology. 2nd ed. New York, Springer Publishing Company Inc, 1965. pp 62-69 5. Kernohan JW, Uihlein A: Sarcomas of the brain, Progr Clin Cancer 1.414*437, 1965 6. Earle KM. Richany 5F: Meningiomas; A study of the histology, incidence, and biologic behavior of 243 cases from the Frazier-Crmnt collection of brain tumon. Med Ann DC 36 353-356. 1969 7. Ringertz N. Tomberg B. Sjtistrom X, et al: Cancel Incidence m Sweden 1959. Stockholm Sweden. Na* tional Board of Health. The Cancer Registry, 1962 8- Leibowit% U, Alter M; Tumors of the nervous system: Incidence and population selectivity. (abitr) Neurology {Minneap) 19:292. 1969 9. Schoenberg BS, Christine BW: Neoplasms of the brain and cranial meninges: A study of incidence, epidemio* logical twnds, and survival. fabstr) Neurology (Mia- neap) 20:399, 1970 10. Brewis M, Poskanzer DC, Rolland C. et at: Neuro logical disease in an English city. Acta Neurol Scand ; * 42(Suppl 24)-.1-89, 1966 11. Cohen A, Modan B: Some epidemiologic aspects of neoplastic diseases in Israeli immigrant population. , III. Brain tumours. Cancer 22:1323*1328, 1966 12. Cuomundsson KR: A survey of tumors of the cent*} nervous system in Iceland during the I0*year penod 1954-1963. Acta Neurol Scand 40:538-552, 1970 13- Huntington RW Jr. Cumminfs KL, Moe Tl, et al: Dis- covery of fatal pnmary intracranial neoplasms at md>* colegal autopsies. Cancer 18; 117-127, 1965 14. Wagner JA: Neuropathology of neoplastic1 metastasis to the brain. Southern Med J 48:302-305, 1935 A <>* 15. Doll R,"Payne P, Waterhouse J: Cancer Incidence is Five Continents; A Technical Report, Berlin, SpnngerVerlag, 1966 16. Dorn HF, Cutler $J: Morbidity from cancer in the United States (Public Health Monograph No. 29). * t. i* Washington DC, Government Printing Office, 1955 17. Kyle RA, Nobrcga FT, Kurland LT: Multiple mjtloma in Olmsted County, Minnesota, 1945-1964. Blood 33; * fi r- 739-745. 1969 ? 18. Snyler J, Pa>ne WS, Kurland LT-. Unpublished data. , M i r. n. at' tf r. t. b) iT tu j!.' uy fan Tl e\ tl CtJ* Hi l g,ii` Cilmn or re'-* fe P*`` lev arr OLX 6584