Quantitative observations on iliac bone marrow mast

J. clin. Path., 1975, 28, 947-955

Quantitative observations on iliac bone marrowmast cells in chronic renal failureKATHLEEN M. PEART AND H. A. ELLIS

From the Department ofPathology, University of Newcastle upon Tyne

SYNOPSIS Mast cells have been counted in sections of iliac bone from 61 control subjects atnecropsy. Mast cells were found in all but three, and the range was 0-33 7, median 195 per mm2marrow. The majority (82%) had < 4-99 mast cells per mm2 marrow; in 37-7% there was < 1

mast cell per mm2 marrow.In a group of 45 patients with chronic renal failure there was a significant increase in the numbers

of mast cells (P < 0001) with a range of 096-55-63, median 9-55 per mm2 marrow. Mast cells werecommon in the areas of marrow fibrosis associated with osteitis fibrosa but this was not the solecause of the increase since there was also an excess of mast cells in the non-fibrous parts of themarrow. There was a tendency towards greater numbers of mast cells in those cases with mostmarked osteitis fibrosa in association with the prominent marrow fibrosis, but there was no sig-nificant relationship between mast cell numbers and other features of oesteitis fibrosa such as thenumber of osteoclasts and the amount of woven bone formation. There was no relationship betweenthe numbers of mast cells and the amounts of total bone, osteoid, percentage mineralization ofcancellous bone, or the presence of osteomalacia.

Shortly after we began a long-term study of thehistopathology of azotaemic renal osteodystrophysome eight years ago one of us observed an apparentincrease in the number of marrow mast cells intoluidine blue stained sections of iliac bone frompatients with chronic azotaemia. More recently,Neiman et al (1972), in a letter to the Lancet, ieferredbriefly to their finding of increased numbers ofmastcells in the spleen and bone marrow of patients withuraemia.

There is little quantitative information in theliterature concerning marrow mast cells in man. Inthe past, the finding of such cells has been regardedas abnormal (Fadem, 1951) and it is now recognizedthat mast cells are present in normal bone marrow(Johnstone, 1954; Selye, 1965). The purpose of thepresent paper is to report quantitative observationson the mast cells in histological sections of iliac bonemarrow in 45 patients with chronic renal failure. Forcomparison samples of iliac bone were obtained atnecropsy from 61 cases of sudden death. Numbers ofmarrow mast cells have been considered in relationto the severity of osteitis fibrosa, the presence of

Received for publication 26 June 1975.947

excess osteoid, and osteomalacia, all of which arefeatures of azotaemic renal osteodystrophy.

Material and Methods

BONESamples of iliac bone were examined from45 patientswith chronic renal failure due to a variety of renaldisorders. Iliac bone was obtained by transilialtrephine bone biopsy in 22 patients and at necropsyin 23. Of the patients coming to necropsy, 11 hadterminal peritoneal dialysis, one had a terminalhaemodialysis, and two had been treated by perito-neal dialysis for two months. The biopsy patientswere near end stage renal failure. Three had beentreated by peritoneal dialysis for two to three months,and 12 required dialysis after periods varying froma few days up to three months after bone biopsy.

Control iliac bone samples were obtained atnecropsy from 61 previously physically healthyindividuals after sudden death due to a variety ofcauses such as trauma and poisoning.

Blocks of iliac bone were fixed in neutral 10%formalin, decalcified in formic acid, and doubleembedded using low viscosity nitrocellulose; 5 ,u

copyright. on A

pril 22, 2022 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.28.12.947 on 1 D

ecember 1975. D

ownloaded from

http://jcp.bmj.com/

Kathleen M. Peart and H. A. Ellis

sections were stained by 0-05% toluidine bluesolution at pH 3-8.

MAST CELLS IN MARROW AND FIBROUSTISSUEWe have regarded as marrow tissue mast cells anyconnective tissue cell possessing cytoplasmic granulesthat stain metachromatically. Basophils possesssimilar granules but differ morphologically frommast ce!ls in that they are generally smaller and havea lobulated nucleus. Since human basophilic gianulesare water soluble, metachromasia is lost afteraqueous fixation, and basophils are not liable to beconfused with mast cells in histological sections.Mast cells were counted in the marrow of cancel-

lous bone in a square measuring 0 5 x 0-5 cmmarked out on the cover slip in an area immediatelybeneath the iliac crest in the case of the necropsysamples and in the middle of the biopsy specimens,using x 40 objective and x 10 eyepiece incorporatinga squared graticule. The proportion of total bonewithin the measured area was determined using apoint-counting technique (Ellis and Peart, 1972) andthe number of mast cells per mm2 of marrowwas then calculated. In 24 of the patients with renalfailure the amount of marrow fibrous tissue resultingfrom osteitis fibrosa was also measured using apoint-counting technique (Ellis and Peart, 1973). Inthese patients, mast cells were counted separately infibrous tissue and in mixed fatty and cellular marrowto ascertain whether or not mast cells were concen-trated in the fibrous tissue.

OSTEITIS FIBROSA, EXCESS OSTEOID, ANDOSTEOMALACIAThe severity of osteitis fibrosa was assessed semi-quantitatively on a scale of grades from 0 to 5, andthe number of osteoclasts and amount of wovenbone were determined as previously described(Ellis and Peart, 1973). The amount of osteoid,expressed as a percentage of the measured area, andthe percentage mineralization of cancellous bonewere determined by a point-counting technique usingundecalcified sections (Ellis and Peart, 1972).Osteomalacia was diagnosed on the basis of thepresence of an excess of osteoid with abnormallywide osteoid seams and a reduced or absent calcifi-cation front (Ellis and Peart, 1973; Ellis, 1973).

(P > 0-10) between the means for males (4 70 permm2) and females (2-65 per mm2). Plotting thedistribution of mast cell values iD relation to agerevealed no significant relationship in males orfemales. As the figure shows, there was considei ablevariation in the number of marrow mast cellsbetween individuals, and the distribution is notnormal. Mast cells were found in all but three of the61 samples examined. The highest value (33-76 permm2) was in an 11-year old boy and this in the areaimmediately contiguous with the zone of endochon-dral ossification beneath the iliac crest. A secondcount at a distance of 1-0 cm below this gave a lowervalue of 4-67 mast cells per mm2. In three olderindividuals aged 21 to 22 years, duplicate countsmade at distances of 0 5 cm and 1-5 cm below thecrest revealed less marked differences, the pairedvalues being 3-18 and 7-42, 6-32 and 4-80, and 4-7and 168 mast cells per mm2 of bone marrow,respectively.The majority (82%) of the control subjects had

mast cell values within the range 0 to 4 99 per mm2,and in 37-7% of individuals there was < 1 mast cellper mm2. The six individuals (10%) with > 10 mastcells per mm2 were males aged 11 to 47 years. Themode of death was unrelated to the highei values.Thus one died from asphyxia trapped in a disusedpit, one from multiple injuries, and the others fromvarious poisons including carbon monoxide, barbitu-

60

0

-J

Jam

I-

LU

z)

50

40

30

20

10

0

Results

0 0

a

A A

- AA,& *%AiOA

CONTROLGROUP

a9A

*0 OA*

68O*&A &AO

RENALGROUP

CONTROL SUBJECTSThe arithmetic mean number of marrow mast cellsfor the 38 males and 23 females, varying in age from5 to 93 years, was 3-93 per mm2 with a median of1-95 per mm2. There was no significant difference

Figure Distribution of the numbers of iliac bonemarrow mast cells in 61 control subjects and in 45patients with chronic renalfailure- A and A males,biopsy and necropsy samples; 0 and0 females, biopsyand necropsy samples.

948

copyright. on A


http://jcp.bmj.com

/J C


ecember 1975. D

ownloaded from

http://jcp.bmj.com/

Marrow mast cells in renal failure 949

rates, and aspirin. Other subjects dying from these CHRONIC RENAL FAILUREpoisons had much lower values. Thus four with Total marrow mast cellscarbon monoxide poisoning had 0 to 0-52 mast cells The results of mast cell counts in the iliac boneper mm2. In addition, in a selected group of pre- marrow of the 45 patients with chronic renal failureviously healthy individuals comprising 25 males and are summarized in the figure and giveD in more detail12 females in whom death had been almost instan- in table I. The values in the figure refer to the numberstaneous and predominantly the result of trauma such of mast cells in the whole marrow, including thoseas road accidents, the mean values for males and in mixed fatty and cellular areas and in any fibrousfemales were 3-81 and 2-21 mast cells per mm2, areas resulting from osteitis fibrosa. Although thererespectively, and these did not differ significantly is some overlap with values obtained in the controlfrom those obtained for the control group as a group, in general values were higher and the arith-whole. metic mean was 13-75 per mM2 with a median of

Case No. Age Mast Cells per mm" Mineralization Data Osteitis Fibrosa Dataand Sex

All Marrow Non-Fibrous C M OM OCL WB MF OFGMarrow

234S

6789101112131415161718192021222324252627282930313233343536373839404142434445

ISM16F24F25M27F29M31M32M (2)33F40M (17)42F45F (6)45M45M47F48F49M (18)50M52M53F54M58F23F (27)27F (28)29M (57)35F (15)37M (24)40M (3)42F (31)42M (39)43M (48)45F (44)4SF (5)50M (45)50M (32)SlF (40)53M (41)54F (42)5SF (19)58F (10)59F (43b)60F (I 1)62F (21b)68M (47)70F (37)

9-5244-7431-8933-2914-232-89

22-371-18

19-9826-371-867-87

26-021-32

44-063-25

25-0315-5155-637-8512-1013-011-210-9611-262-62107720-1910-498-882-399.7717-541-625-896-057-62

25-054-04

29-511-579.557-981-392-56

3-012-63

29 5232-9111-692-628-210-93180626-361-466-037-211-32

10-952-82

21 -7013-7443.752-618445-75

11-09

5.47

2-570720-090-150-390-284-020-576-440-310-043.940-490-181-050-111 001-511-811-840-750-480-105.350-152-890080-044650-030-010026220-030-020-040590-180-128-080-360-04264

17 160-01

90-598499-699.397.598988-598-080898-799.982-298-199.395-699.395-093.992-491 896298299.585699692-799-699-886-699.9100-099.979.799.999.999498-299-199-263-698499-673961-0

100-0

0225630-710-240-560-791-550381-490-960-277 702-670201-310-285-003-159-461-580332-160-022490-561-620-300742-640790590-360-670-060480-270-360-960060-960590664-400-350-11

6-8

5-3

50-8

61-6

9-623-310-136-90

7-429-95-13-04-3

33-62-6

21-28-79-6

43-76-616-623-18-3

43-818-3

0

8-614-11-40-5

2-50-96-50-43-30-70-58-0

13-10

4-81-05.7

11-510-43-60-61-50-12-00-91-60-20-72-10-41-00-30-60-11-10-30-72-60

1-41-00-4

13-81-60-1

3421-5

2231-53224413-51

345442-523

2-5232232222-512-522

30

2224-521

Table I Details of the numbers ofmarrow mast cells, mineralization ofbone, and severity of osteltis fibrosa in45 patients with chronic renal failure0, /. measured area comprising osteoid; M, % mineralization of cancellous bone; OM, osteomalacia + or -; OCL, number ofosteoclasts per mm" of cancellous bone and marrow; WB, woven bone O. of cancellous bone trabeculae; MF, /. measured area comprisingfibrous tissue; OFG, osteitis fibrosa gradeFigures in brackets refer to the case numbers in a previous paper (Ellis and Peart, 1973). Present cases 1 to 22 biopsy bone samples;cases 23 to 45 necropsy bone samples.

copyright. on A


http://jcp.bmj.com

/J C


ecember 1975. D

ownloaded from

http://jcp.bmj.com/


9-55 per mm2. Thus in only one individual (2 2%)was there < 10 mast cell per mm2 compared witha corresponding value of 37.7% in the control group.In contrast, the number of mast cells exceeded 5 perMM2 in 68 9% of patients with chronic renal failurecompared with 18-0% of controls. Comparison ofthe control and renal groups by the Mann-Whitneytest revealed a significantly higher mean number ofmarrow mast cells in patients with chronic renalfailure (table II). This was true for renal patientswhether the bone sample had been obtained bytransilial biopsy or at necropsy. The biopsy serieshad significantly more marrow mast cells than didthe necropsy series (p < 0-02).

Mast cells and marrow fibrosisMast cells in the marrow of patients with chronicrenal failure occurred both in the mixed fatty andcellular areas and in the zones of marrow fibrosisconstituting part of the repair reaction to osteitisfibrosa. Mast cells were commonly conspicuouswithin the cellular fibrous tissue occupyingresorptionlacunae and lying alongside cancellous trabeculae.In 24 patients with renal failure (22 bone biopsy and2 necropsy samples) differential counts were made ofthe number of mast cells in the various componentsof the marrow. Very high values were found in thefibrous areas within the range 0 to 536 4 mast cellsper mm2, arithmetic mean 148-8 and median98-27 per mm2. When the number of total marrowmast cells was plotted against the percentage ofmarrow occupied by fibrous tissue there was asignificant linear relationship (P < 0 001) with acorrelation coefficient (r) = 0 474, although there

was a fairly wide scatter of individual values aboutthe calculated regression line.The large number of mast cells within the fibrous

parts of the marrow did not alone account for theincrease in total marrow mast cells in the patientswith chronic renal failure. Thus the differentialcounts in 24 cases also revealed an increased numberof mast cells in the mixed fatty and cellular compon-ent of the marrow with from 0 93 to 43'75 mast cellsper mm2 (arithmetic mean 11 60 and median 7-71per mm2), which was significantly higher than thenumber of mast cells found in the control marrows(P < 0 001).

Number ofmast cells and severity of osteitis fibrosaThat mast cells were common within the fibrouscomponent of the marrow suggested the possibilitythat the increased number of mast cells in azotaemiamight be the result of the osteitis fibrosa due tosecondaiy hyperparathyroidism. Since the osteitisfibrosa grades are non-parametric it was not possibleto calculate any possible linear relationship betweenthe number of marrow mast cells and the severity ofthe osteitis fibrosa. There was a wide range of mastcell numbers within each osteitis fibrosa grade, witha tendency towards most mast cells in the bonesshowing the more severe grades of osteitis fibrosa.Comparison of the mean numbers of mast cells inbone with osteitis fibrosa giades 0 to 2 and grades2X5 to 5 showed more mast cells in the group withsevere osteitis fibrosa (table III) when biopsy andnecropsy specimens were taken together (P < 0-05).There were similar numbers of mast cells in the non-fibrous parts of the marrow in the two osteitis

Mast Cells per mma Marrow Comparisons with ControlsGroup Number (Mann-Whitney Test)

Range Mean Median

Control 61 0-33-76 3 93 1-95Renal (necropsies) 23 0-96-29-51 8-65 7-62 P < 0 001Renal (biopsies) 22 1 18-5563 19 09 14-87 P < 0001All renal 45 096-5563 13-75 955 P < 0001

Table II Marrow mast cell values in 61 controls and 45 patients with chronic renal failure

Group Osteitis Fibrosa Grade Osteitis Fibrosa Grade0-2 2-5-5

Mast Cells per mm' Mast Cells per mm'Number Number

Mean Median Mean Median

Biopsies 10 12-84 7-68 12 24-30' 21-18Necropsies 16 8-02 6-84 7 10*08' 7-98Combined 26 9-88 6-83 19 19-061 15S51

Table III Relationship between number ofmarrow mast cells and severity of osteitis fibrosa'Significant difference between groups with grade 0-2 and 25-5 osteitis fibrosa, P < 0-05'No significant difference between groups grade 0-2 and 25-5 osteitis fibrosa, P > 0-10

950

copyright. on A


http://jcp.bmj.com

/J C


ecember 1975. D

ownloaded from

http://jcp.bmj.com/

Marrow mast cells in renal failure

fibrosa groups in the 24 cases in which differentialcounts were made. Thus the groups osteitis fibrosagrades 0-2 and 2-5-5 had respective means of 11-74and 11 47 mast cells per mm2 of mixed fatty andcellular marrow.

Plotting the numbers of osteoclasts per squaremillim tre of bone against the correspondingnumbers of total marrow mast cells or non-fibrousmarrow mast cellsrevealed arandom scatter of pointswith no suggestion of any significant relationship.The number of marrow mast cells was not related tothe amount of woven bone formation.

Number ofmast cells and state ofbone mineralizationThe presence of osteomalacia was not related to thenumber of marrow mast cells. The values for therange, mean, and median for 10 cases with osteo-malacia were 0-96-29-51, 12-76, and 10 01 mast cellsper mm2, respectively. The corresponding values forthe 35 cases without osteomalacia were, respectively,1-18-55-63, 14-04, and 9 55 mast cells per mm2.Attempts to relate the amounts of osteoid,

expressed as a percentage of a measured area ofbone and marrow, to the numbers of marrow mastcells were unsuccessful in osteomalacic and non-osteomalacic samples. Similarly, there was nosignificant relationship between the percentagemineralization of cancellous bone and the numberof marrow mast cells.

Discussion

The present observations indicate that mast cellsare to be found in histological sections of iliac bonemarrow in the majority of individuals. We haveregarded as tissue mast cells all those mononucleatedcells in the marrow with intracytoplasmic metachro-matic granules. It seems to be generally acceptedthat the solubility characteristics of basophilicleucocytes make them unrecognizable in tissue whenan aqueous or alcoholic fixative has been used(Johnstone, 1954), whereas human mast cell granulesare relatively insoluble in watery fixatives (Johnstone,1954; Kramer and Windrum, 1955; Sagher andEven-Paz, 1967). The solubility of mast cell granulesvaries between different animal species. Mouse, rat,and human mast cell granules are well preserved afterformalin fixation but non-aqueous fixatives arenecessary in the case of lower vertebrates and in theguinea-pig and dog (Benditt and Dunn, 1963). Theshape of mast cells was noted to vary, most beingovoid or rounded but some were elongated aroundthe surface of fat cells or were distinctly spindleshaped. Granules were frequently sufficiently denseand numerous to obscure nuclear detail. Individualmast cells were scattered throughout the marrow

with a tendency to lie alongside cancellous trabec-ulae, around blood vessels or arranged in groupswith wide intervening areas devoid of mast cells.This type of distribution inevitably limits the precisevalue of mast cell counts made in small areas whichmay not necessarily be representative. Any changesobserved in serial bone biopsies from individualpatients must therefore be interpreted with caution.In the present study we are concerned mainly witha comparison of two different groups of subjects sothat any differences encountered between the groupsas a whole may be regarded as meaningful.

In the control necropsy series comprising 61individuals, the distribution of values for thenumbers of marrow mast cells was not a normal oneand we have therefore not quoted the mean andstandard error. The control values were range0-33-76, arithmetic mean 3 93, and median 1-95 mastcells per mm2 marrow. We have no explanation forthe fact that 10% of individuals had mast cell valuesin excess of 10 per mm2 marrow, and these valueswere apparently unrelated to the mode of death.

In the group of 45 patients with chronic renalfailure taken as a whole there were significantly moremairow mast cells (arithmetic mean 13 75 andmedian 9 55 per mm2) than in the control group(P < 0-001). Again the distribution of values wasnot a normal one and comparisons between thevarious groups throughout this study have beenmade using the appropriate Mann-Whitney test(Snedecor and Cochran, 1971). Within the renalgroup there was a significantly higher range of mastcell values for those samples obtained by iliac crestbiopsy as compared to those obtained at necropsy.This raises the possibility that better preservation ofmast cells and hence easier recognition result fromrapid fixation of the biopsy samples, whereas in thecase of the necropsy specimens there would inevit-ably be a delay between death and the subsequentfixation of tissues at the time of the necropsy. Wecannot exclude this possibility, but it does not alterthe significance of the increased numbers of marrowmast cells observed in renal disease since even in thenecropsy specimens significantly more mast cellsoccurred in the renal failure patients than in thecontrols (P < 0-001). Furthermore, we have observedin serial bone biopsies from patients treated by renaltransplantation, where conditions of fixation ofspecimens were identical, a return of marrow mastcells to the control range.The difference between renal biopsy and necropsy

cases may also be attributable in part to the fact thatthere were more severe examples of osteitis fibrosain the biopsy series with more marrow fibrous tissuewhich included frequent mast cells. In the bone ofpatients with chronic renal failure there was com-

951

copyright. on A


http://jcp.bmj.com

/J C


ecember 1975. D

ownloaded from

http://jcp.bmj.com/

5Kathleen M. Peart and H. A. Ellis

monly evidence of osteitis fibrosa due to secondaryhyperparathyroidism. Mast cells were often con-spicuous within the zones of marrow fibrosis in suchcases, and we were able to demonstrate a significantrelationship between the total number of marrowmast cells and the amount of marrow fibrosis. Wehave also observed mast cells in the fibrous tissue oforganizing iliac bone biopsy sites, and it is wellrecognized that mast cells may occur in many sitesofrepair where there is connective tissue proliferationsuch as fracture callus (Selye, 1965; Lindholm et al,1969; Lindholm and Lindholm, 1970). From ourobservations we are not certain that all the cellscounted as mast cells in the marrow fibrous tissueare true mast cells. Some of the cells may have beenfibroblasts which had taken up metachromaticgranules from adjacent disrupted mast cells(Higginbotham et al, 1956).That the mast cells in marrow fibrous tissue did

not alone account for the increased numbers ofmarrow mast cells in renal failure was shown by thefinding of increased numbers of mast cells in themixed fatty and haemopoietic marrow of thesepatients. This suggests that there is someotherfactorin renal failure which is responsible, at least in part,for the increase in marrow mast cells. This conceptis supported by the report that mast cells areincreased in tissues other than the bone marrow inchronic renal failure, such as the spleen (Neimanet al, 1972). These authors put forward thehypothesis that hyperparathyroidism, which isalmost invariably present in chronic renal failure(Gilmour, 1947; Ellis and Peart, 1973), may accountfor tissue mast cell proliferation since there isevidence to suggest in experimental animals thatparathormone stimulates the accumulation of mastcells in bone (Rockoff and Armstrong, 1970;Rasmussen, 1972). This could be the explanation forthe excess of mast cells in the fibrous tissue compon-ent of the osteitis fibrosa and in the remaining non-fibrous marrow. To test this hypothesis we havetried to relate the severity of the osteitis fibrosa,representing the degree of hyperparathyroidism, tothe number of mast cells both in the marrow as awhole and in the non-fibrous marrow. The resultsshowed a tendency towards greater numbers of mastcells in those bone samples with the more severegrades of osteitis fibrosa. When the combined renalbiopsy and necropsy samples were considered, therewas a low order of significant difference (P < 0-05)between the number of total marrow mast cells inthe group with osteitis fibrosa grades 0 to 2 and thatwith grades 2-5 to 5. This difference was almostcertainly associated with the greater amounts offibrous tissue with mast cells in the marrow of thesamples with more severe grades of osteitis fibrosa.

Thus there were similar numbers of mast cells in thenon-fibrous parts of the marrow in the two osteitisfibrosa groupings in 24 individuals where differentialcounts of mast cells in fibrous and non-fibrousmarrow were made. As another measure of hyper-parathyroidism (Ellis and Peart, 1973) we consideredthe number of osteoclasts per square millimetre ofbone in relation to the number of mast cells butthere was no significant relationship. We thereforehave no definite evidence to indicate a relationshipbetween the degree of hyperparathyroidism and thenumber of marrow mast cells. Indeed, what evidenceexists suggests that there may not be any directrelationship. Preliminary observations we have madein serial bone biopsies from patients with chronicrenal failure before and after treatment with calciumcarbonate and aluminium hydroxide or by para-thyroidectomy, and in whom there has been a fallin serum parathormone and a reduction in theseverity of osteitis fibrosa histologically, have shownno corresponding reduction in the excess of mastcells. This is against the hypothesis that there is arelationship between hyperparathyroidism and num-bers of marrow mast cells. A further difficulty in anyattempt to relate the number of marrow mast cellsto the degree of osteitis fibrosa is the duration of thechanges. The number of mast cells may be a functionof the stage of development of the osteitis fibrosaand the age of the marrow fibrous tissue component.By analogy with the changing numbers of mast cellsin the processes of organization and repair elsewhere,such as in the healing of a wound (Wichmann, 1955),one would expect most mast cells to be present duringthe active phase of marrow fibrosis and subsequentlyto be reduced.

Johnstone (1954) observed that the presence andnumber of mast cells in the marrow were not relatedto the iron content of the marrow or to the degreeof erythropoiesis and that anaemia was the onlyfactor common to all the cases in which mast cellswere prominent. The majority of our patients withchronic renal failure had some degree of anaemiabut there was no relationship between the numberof marrow mast cells and the haemoglobin concen-tration of the blood within the group of renalpatients.There are no data in the literature strictly compar-

able with that presented here but Rebel et al (1974)have reported marrow mast cell counts in 21 patientswith renal failure treated by haemodialysis in therange 0 to 600 per mm2, which is within the rangewe have found in renal failure. These authors donot report any control values however. In 11 otherpatients with, for example, osteomalacia or osteo-porosis, the range was 0 1 to 23-0 mast cells per mmn2of marrow. Highest marrow mast cell values were

952

copyright. on A


http://jcp.bmj.com

/J C


ecember 1975. D

ownloaded from

http://jcp.bmj.com/

Marrow mast cells in renalfailure

observed in patients with osteomalacia whether thiswas the result of renal failure or other causes, andthe authors suggested that osteomalacia alone issufficient to bring about an increase in the number ofmarrow mast cells. We were unable to confirm thisrelationship between osteomalacia and numbers ofmarrow mast cells in our patients with chronic renalfailure, and there was no significant relationshipbetween numbers of mast cells and the amount ofosteoid per cent of the measured bone area, or thepercentage mineralization of cancellous bone. Thenature of the bone changes in chronic azotaemiadepends upon the severity and duration of the renaldisease. Hyperparathyroidism occurs at an earlystage, so that in an unselected series of patients withchronic renal failure one can expect about 90% toshow histological evidence of osteitis fibrosa whereasosteomalacia, which develops later, occurs in onlyabout one-third (Ellis and Peart, 1973). In anyunselected series of patients with chronic renalfailure the duration and severity of osteomalacia willvary from case to case. The rate of formation ofosteoid also seems to vary so that some patientsshow relatively little increase in osteoid in serialbiopsies whereas in others it is more rapid, and largeamounts of osteoid accumulate and make a signifi-cant contribution to osteosclerosis. If there is arelationship between the numbers of marrow mastcells and osteomalacia then it might be that mastcells are associated with some particular stage in thedevelopment or persistence of osteomalacia. Theincidence of osteomalacia in haemodialysed patientsincreases with time on dialysis (Ellis and Peart,1971a) so that in a series of dialysed patients, suchas that reported by Rebel et al (1974), some patientswould have recent and active development of osteo-malacia. If this early stage of osteomalacia is associa-ted with increased marrow mast cells, then this couldexplair the discrepancy between the results of theseauthors in haemodialysed patients and in our ownnon-dialysed patients. We are currently accumulatingdata on patients treated by long-term haemodialysiswho have developed osteomalacia and we hope tobe able to elucidate this aspect of the problem in thefuture. Increased numbers of mast cells alongsidebone cancellous trabeculae have been described inrats on a low-calcium and vitamin D deficient diet(Shipley et al, 1922; Urist and McLean, 1957; Casset al, 1958) and the mast cells were reduced innumber after treatment with vitamin D.

Increased numbers of marrow mast cells appearto be associated with experimental resorption andremodelling of bone whatever the cause (Gillman,1958; Lalich, Groupner, and Jolin, 1965; Severson,1969), and it is possible that the presence of excessmast cells in azotaemia is a secondary manifestation

953

of bone resorption and formation. The preciserelationship between mast cells and bone resorptionand formation does not appear to have beenresolved. Local effects of heparin may be responsiblesince it is known that the administration of heparinand similar but synthetic polysaccharides to experi-mental animals is associated with bone resorptionand impaired ossification (Gillman and Naidoo,1957; Ellis, 1965). Osteoporosis has also beenreported following the long-term administration ofheparin to patients (Griffith et al, 1965; Jaffe andWillis, 1965; Schuster et al, 1969). The mechanismsinvolved are unknown although it has been shownthat heparin stimulates collagenase activity (Asherand Nichols, 1965; Sakamoto et al, 1973) and thatheparin is capable of releasing lysozymal enzymeswhich might be involved in bone resorption (Sledgeand Blackburn, 1968). It has been shown (Ellis andPeart, 1970) that heparin increases the resorption ofdiaphyseal bone in cultured mouse limb bones,suggesting that there is a direct effect of heparin onbone. This effect occurred, however, only withhigher concentrations of heparin than are reachedin the blood and tissues of intact animals even afterprolonged administration. For the in vivo changesto be the result of a direct toxic effect on bone therewould have to be a selective accumulation of heparinin bone tissues. The excess marrow mast cells in thepresent cases are, of course, already near to the bone,and there might be localized zones of high heparinconcentration. Others have suggested that the boneresorptive effect of heparin is an indirect one andmediated through hyperparathyroidism resultingfrom the hypocalcaemia following the binding ofserum calcium ions by heparin (Hahnemann, 1965).This hypothesis is not borne out by the observationsthat calcium heparin is equally effective as sodiumheparin in inducing osteoporosis (Ellis, 1965) andthat parathyroidectomy does not prevent develop-ment of heparinoid osteopathy (Ellis and Peart,1971b). There is also evidence from tissue cultureexperiments to suggest that heparin may enhancebone resorption by acting as a parathormone co-factor (Goldhaber, 1965).

In discussing the possible local effects on bone ofmast cells, one has to consider the relatively rarecondition of systemic mastocytosis where there isinfiltration of the skin and other tissue by mast cells(Sagher and Even-Paz, 1967). Bone sections havebeen reported as showing resorption of trabeculaerelated to collections of mast cells, deposition offibrillary lamellar bone, and trabecular thickening,marrow fibrosis, and osteosclerosis (Stark et al, 1956;Jensen and Lasser, 1958; Havard and Scott, 1959;Lerner and Lemer, 1960). In azotaemic renalosteodystrophy osteosclerosis may occur when the

copyright. on A


http://jcp.bmj.com

/J C


ecember 1975. D

ownloaded from

http://jcp.bmj.com/


amounts of osteoid and mineralized woven boneformed exceed the loss of mineralized lamellar andwoven bone due to excess osteoclasis (Ellisand Peart,1973). Woven bone is particularly rich in sulphatedpolysaccharides but we could not find any definiterelationship between the numbers of marrow mastcells and the total amount of bone or the amount ofwoven bone in the present cases.

In conclusion, our results indicate that there is anincrease in the number of mast cells in the marrowof patients with chronic renal failure. Although thereare commonly numerous mast cells within thefibrous parts of the marrow forming part of theosteitis fibrosa reaction, we have been unable toexplain the increased mast cells in renal failure solelyon the basis of hyperparathyroidism. In addition,there does not appear to be any clear relationshipin the present cases between the number of marrowmast cells and the presence or severity of a bonemineralization defect.

We wish to thank Dr. D. R. Appleton of theDepartment of Medical Statistics for statisticaladvice. It is again a pleasure to acknowledge thehelp of our clinical colleagues in the Departmentsof Medicine and Surgery who provided the bonebiopsies. One of us (KMP.) is in receipt of an MRCgrant (no. G973/693/C).

References

Asher, J. D. and Nichols, G., Jr. (1965). Heparin stimulationof bone collagenase activity. Fed. Proc., 24, (Abstract),211.

Benditt, E. P.: Dunn, T. B. (1963). Mast cells and baso-phils; Round Table Discussion. Ann. N. Y. Acad. Sci., 103,452-453.

Cass, R., Marshall, P. B., and Riley, J. F. (1958). 5-Hydroxy-tryptamine and histamine in mast cells of the mouse andrat. J. Physiol. (Lond.), 141, 510-519.

Ellis, H. A. (1965). Effects of the long-term administrationto animals of dextran sulphate. J. Path. Bact., 89, 437-460.

Ellis, H. A. (1973). Bone disease in end-stage renal failure(Letter). Brit. med. J., 3, 232-233.

Ellis, H. A. and Peart, K. M. (1970). The effects of heparinand dextran sulphate on cultured mouse limb bones. Brit.J. exp. Path., 51, 43-52.

Ellis, H. A. and Peart, K. M. (1971a). Renal osteodystrophywith particular reference to the effects of chronic inter-mittent haemodialysis (Abstract). Nephron, 8, 402.

Ellis, H. A. and Peart, K. M. (1971b). Dextran sulphateosteopathy in parathyroidectomised rats. Brit. J. exp.Path., 52, 684-695.

Ellis, H. A. and Peart, K. M. (1972). Quantitative observa-tions on mineralized and non-mineralized bone in the iliaccrest. J. clin. Path., 25, 277-286.

Ellis, H. A. and Peart, K. M. (1973). Azotaemic renalosteodystrophy: a quantitative study on iliac bone. J. clin.Path., 26, 83-101.

Fadem, R. S. (1951). Tissue mast cells in human bone marrow.Blood, 6, 614-630.

Gillman, T. (1958). Mast cell increases after calciferol intoxi-

cation and in experimental odoratism. Acta haenmat.(Basel), 19, 179-186.

Gillman, T. and Naidoo, S. S. (1957). In vitro effects ofheparin and calcium ions on lipaemic serum. Natutre(Lond.), 179, 904-905.

Gilmour, J. R. (1947). The Parathyroid Glands and Skeletonin Reral Discase. Cumberlege, Oxford University Press,London.

Goldhaber, P. (1965). Heparin enhancement of factorsstimulating bone resorption in tissue culture. Science,147, 407-408.

Griffith, G. C., Nichols, G., Jr., Asher, J. D., and Flanagan,B. (1965). Heparin osteoporosis. J. Amer. med. Ass., 193,91-94.

Hahnemann, S. (1965). Heparin and osteoporosis (Letter)Lancet, 2, 855-856.

Havard, C. W. H. and Scott, R. B. (1959). Urticaria pigmcn-tosa with visceral and skeletal lesions. Quart. J. Med.,18,459-470.

Higginbotham, R. D., Dougherty, T. F., and Jee, W. S. S.(1956). Fate of shed mast cell granules. Proc. Soc. exsp.Biol. (N. Y.), 92, 256-261.

Jaffe, M. D. and Willis, P. W., III. (1965). Multiple fracturesassociated with long-term sodium heparin therapy. J. Amer.med. Ass., 193, 158-160.

Jensen, W. N. and Lasser, E. C. (1958). Urticaria pigmentosaassociated with widespread sclerosis of the spongiosa ofbone. Radiologv, 71, 826-832.

Johnstone, J. M. (1954). The appearance and significance oftissue mast cells in human bone marrow. J. clin. Path..7, 275-280.

Kramer, H. and Windrum, G. M. (1955). The metachromaticstaining reaction. J. Histochem. Clvtochem., 3, 227-237.

Lalich, J. J., Groupner, K., and Jolin, J. (1965). The influenceof copper and molybdate salts on the production of bonydeformities in rats. Lab. Invest., 14, 1482-1493.

Lerner, M. R. and Lerner, A. B. (1960). Urticaria pigmentosawith systemic lesions and otosclerosis. Arch. Derm., 81,203-204.

Lindholm, R. V. and Lindholm, T. S. (1970). Mast cells inendosteal and periosteal bone repair. Acta orthop. scand.,41, 129-133.

Lindholm, R., Lindholm, S., Liukko, P., Paasimaki, J.,lsokaanta, S., Rossi, R., Autio, E., and Tamminen,(1969). The mast cell as a component of callus in healingfractures. J. Bone Jt Surg., 51b, 148-155.

Neiman, R. S., Bischel, M. D., and Lukes, R. J. (1972).Uraemia and mast-cell proliferation (Letter). Lanicet, 1,959.

Rasmussen, P. (1972). An experimental study of mast cellsin the rat metaphysis. Calcif. Tiss. Res., 9, 325-330.

Rebel, A., Riberi, P., Bregeon, P., and Malkani, K. (1974).Mastocytes et remaniements du tissu osseux. Path. et Biol.,22, 213-220.

Rockoff, S. D. and Armstrong, J. D., Jr.,(1970).Parathyroidhormone as a stimulus to mast cell accumulation in bone.Calcif. Tiss. Res., 5, 49-55.

Sagher, F. and Even-Paz, Z. (1967). Mastocytosis and theMast Cell. Karger, Basle and Year Book MedicalPublishers, Chicago.

Sakamoto, S., Goldhaber, P., and Glimcher, M. J. (1973).Mouse bone collagenase: the effect of heparin on theamount of enzyme released in tissue culture and on theactivity of the enzyme. Calcif. Tiss. Res., 12, 247-258.

Schuster, J., Meier-Ruge, W., and Egli, F. (1969). ZurPathologie der Osteopathie nach Heparinbehandlung.Dtsch. med. Wschr., 94, 2334-2338.

Selye, H. (1965). The Mast Cells. Butterworths, WashingtonD.C., and London.

Severson, A. R. (1969). Mast cells in areas of experimental

954

copyright. on A


http://jcp.bmj.com

/J C


ecember 1975. D

ownloaded from

http://jcp.bmj.com/

Marrow mast cells in renal failure

bone resorption and remodelling. Brit. J. exp. Path., 50,17-21.

Shipley, P. G., Park, E. A., McCollum, E. V., and Simmonds,N. (1922). Is there more than one kind of rickets? Amer. J.Dis. Child., 23, 91-106.

Sledge, C. B. and Blackburn, W. W. (1968). Heparin, dextranand dextran sulfate: effect on lysosomes from embryoniccartilage. Calcif. Tiss. Res., 2, Suppl. (Abstract), 65-65a.

Snedecor, G. W. and Cochran, W. G. (1971). StatisticalMethods, 6th edition. p. 130. Iowa State University Press,Ames, Iowa.

955

Stark, E., Van Buskirk, F. W., and Daly, J. F. (1956).Radiologic and pathologic bone changes associated withurticaria pigmentosa. Report of a case. Arch. Path., 62,143-148.

Urist, M. R. and McLean, F. C. (1957). Accumulation ofmast cells in endosteum of bones of calcium-deficient rats.Arch. Path., 63, 239-251.

Wichmann, B. E. (1955). The mast cell count during theprocess of wound healing: an experimental investigationon rats. Acta path. microbiol. scand., suppl. 108, 1-35.

copyright. on A


http://jcp.bmj.com

/J C


ecember 1975. D

ownloaded from

http://jcp.bmj.com/

Documents

Quantitative observations on iliac bone marrow mast