SHORT REPORT

Bone marrow failure in Shwachman–Diamond syndrome does

not select for clonal haematopoiesis of the paroxysmal

nocturnal haemoglobinuria phenotype

Peter Keller,1

Michael R. Debaun,2

Robert J. Rothbaum3

and Monica Bessler1 1Department of Internal

Medicine, Division of Hematology, and Department of Pediatrics, 2Divisions of Pediatric Hematology-Oncology and3Pediatric Gastroenterology, Washington University School of Medicine, St Louis, MO, USA

Received 12 April 2002; accepted for publication 17 June 2002

Summary. Bone marrow failure is believed to be theunderlying condition that drives the expansion of the par-oxysmal nocturnal haemoglobinuria (PNH) clone. Indeed,circulating PNH blood cells have been identified in patientswith acquired aplastic anaemia and with hypoplastic mye-lodysplasia. Whether PNH blood cells are also present inpatients with inherited aplastic anaemia has not beenreported. We screened a large group of patients diagnosed

with Shwachman–Diamond Syndrome (SDS) for PNH bloodcells. None of the patients analysed had detectable circula-ting PNH blood cells, indicating that bone marrow failure inSDS does not select for PNH progenitor cells.

Keywords: bone marrow failure, paroxysmal nocturnalhaemoglobinuria (PNH), Shwachman–Diamond Syndrome(SDS), glycosyl phosphatidylinositol (GPI)-linked proteins.

Paroxysmal nocturnal haemoglobinuria (PNH) is caused bythe clonal expansion of haematopoietic cells that haveacquired a mutation in the X-linked PIGA gene, causing aproportion of blood cells to lack glycosyl phosphatidylinos-itol (GPI)-linked surface proteins (PNH cells). PNH has awell-known association with aplastic anaemia (AA) as allpatients with PNH have signs of bone marrow failure and40–50% of patients with acquired AA have circulating PNHcells. The mechanism allowing the defective clone to expandis not yet understood, but it is believed that the bonemarrow environment prevailing in AA, the PIGA genemutation, provides a growth or survival advantage to PNHprogenitor cells. PNH blood cells have also been found inpatients with myelodysplastic syndrome (MDS), suggestingthat AA and certain forms of MDS share a pathogeneticmechanism that confers an advantage to PNH haemato-poietic progenitor cells (for review see References in Bessleret al, 2001). Less well studied is a possible association ofPNH with inherited forms of bone marrow failure. A singlecase report by Dacie and Gilpin (1944) described a youngman with Fanconi’s familial AA who developed PNH.

Shwachman–Diamond Syndrome (SDS) is a multisys-tem autosomal-recessive disorder characterized by thecombination of bone marrow failure and exocrine pan-creatic dysfunction (Shwachman et al, 1964). Themolecular defect in SDS has not yet been identified. Thedegree of bone marrow failure in SDS patients varies froma mild isolated cytopenia to the full picture of severe AA(Smith et al, 1996). SDS can lead to MDS or acutemyeloid leukaemia (Huijgens et al, 1977). To determinewhether bone marrow failure in SDS provides anenvironment that favours the development of PNH, weexamined the occurrence of circulating PNH blood cellsin 28 children who attended the Second InternationalConference on Shwachman–Diamond Syndrome in StLouis in 1999.

PATIENTS AND METHODS

Families with at least one child with SDS and whoattended the conference were invited to participate in ascientific study to elucidate the genetic epidemiology ofSDS. This provided a unique opportunity to include in thestudy the measurement of PNH cells in peripheral bloodfrom affected children. The local institutional ethicalcommittee approved the study. Written informed consentwas obtained. Every family completed a patient historyquestionnaire. Each child was physically examined and a

Correspondence: Monica Bessler, MD, PhD, Division of Hematology,

Department of Internal Medicine, Washington University School of

Medicine, 660 S. Euclid Ave., Box 8125, St. Louis, MO 63110–

1093, USA. E-mail: mbessler@im.wustl.edu

British Journal of Haematology, 2002, 119, 830–832

830 � 2002 Blackwell Publishing Ltd

complete blood count was obtained. Because of the ratherbroad case delineation, probands were subsequently riskstratified for having SDS. Four classes of risk for SDS weredevised. These were based upon the following three criteria(Ginzberg et al, 1999): bone marrow failure (neutrophils< 1Æ500 · 109/l, haemoglobin < 10 g/dl, or platelets< 150 · 109/l), exocrine pancreatic insufficiency (serumtrypsinogen < 16Æ7 lg/l, abnormal pancreatic stimulationtest, low fat-soluble vitamin levels, or abnormal 72-hfaecal fat study together with a characteristic pancreaticabnormality in an imaging study) and metaphysealdysostosis. The risk classification �SDS� was given whentwo criteria were documented, was considered as �likely�when only one criterion was documented and a secondwas suggested by an answer in the questionnaire, wasconsidered as �possible� when only one of the three criteriawas reported or documented, and was �baseline� whennone of the criteria were fulfilled. The risk was alsoconsidered as �baseline� in the presence of a positive sweatchloride test and height above the 10th percentile.Peripheral red cells and neutrophils were analysed by flowcytometry using monoclonal antibodies for the GPI-linkedantigens CD59 and CD24 (Bessler & Fehr, 1991).

RESULTS AND DISCUSSION

Twenty-eight children (17 boys, 11 girls) with a median ageof 8 years participated in the study (Table I). All studyparticipants were of Caucasian origin. In three patients,bone marrow failure and pancreatic insufficiencies weredocumented, resulting in the diagnosis of SDS. SDS waslikely in 16 children, possible in seven and in the tworemaining patients the risk was unclear because criticalinformation was missing. Two of the 19 children with likelyor diagnosed SDS had undergone bone marrow transplan-tation. The remaining 17 patients had cytopenia of one ormore blood cell lineages; 15 of these were neutropenic. The17 patients with likely or diagnosed SDS who had not beentransplanted presented with a mean haemoglobin of 11Æ7 ±1Æ4 g/dl, a mean cell volume (MCV) of 85Æ9 ± 8Æ5 fl, a meanplatelet count of 174 ± 89 · 109/l and a mean absoluteneutrophil count of 0Æ738 ± 0Æ487 · 109/l. At the time ofthe study, no laboratory tests for pancreatic insufficiencywere obtained. Pancreatic dysfunction was documentedonly in the three patients with diagnosed SDS. In the other16 children, the diagnosis of pancreatic insufficiency wasimplied by a positive answer in the questionnaire. In all

Table I. Patients’ clinical and laboratory characteristics, and likelihood of SDS.

Patient

number Sex Age (years) Hb (g/dl) MCV (fl) Plt (· 109/l) ANC (· 109/l)

Risk classification

of having SDS Clinical

1 F 37 11Æ6 95Æ6 118 0Æ262 Likely

2 M 12 12Æ1 86Æ8 110 0Æ276 Likely

3 M 10 12Æ9 86Æ0 168 0Æ672 Likely

4 M 10 10Æ9 101Æ1 121 0Æ738 Unclear

5 M 3 12Æ1 77Æ4 367 1Æ550 Possible

6 F 4 11Æ8 81Æ4 176 0Æ156 SDS

7 M 1 9Æ9 77Æ0 323 0Æ788 Possible

8 M 3 11Æ3 82Æ1 216 0Æ327 Likely

9 M 9 12Æ5 97Æ4 134 1Æ035 Likely

10 F 3 13Æ7 90Æ4 273 4Æ939 Likely AA, BMT

11 F 6 12Æ6 88Æ8 145 0Æ379 Likely

12 M 2 12Æ5 80Æ5 315 0Æ169 Possible

13 M 8 12Æ8 83Æ3 255 n.a. Possible

14 F 2 11Æ6 79Æ4 385 1334 Likely

15 F 4 11Æ6 84Æ2 130 1Æ655 Likely

16 F 5 12Æ9 85Æ5 224 0Æ488 Likely

17 M 8 13Æ2 89 363 2349 Possible

18 F 7 12Æ4 78Æ1 265 n.a. Possible

19 F 8 11Æ4 84Æ1 198 0Æ901 Likely

20 M n.a. 11Æ7 80Æ9 234 0Æ400 Unclear

21 M 9 12Æ4 92Æ8 72 0Æ619 SDS

22 M 6 11Æ7 72Æ5 325 1Æ209 Likely

23 F 11 11Æ8 94Æ8 105 0Æ816 Likely

24 M 15 13Æ4 87Æ6 217 2Æ182 Likely MDS, BMT

26 M 14 12Æ7 88Æ5 194 1Æ486 Likely

27 M 24 6Æ8 65 20 n.a. Likely MDS

28 M 18 13Æ9 94Æ6 263 2Æ175 Possible MDS, BMT

29 F 4 11Æ3 94Æ8 239 0Æ189 SDS

n.a., data not available; unclear, information missing; Plt, absolute platelet count; ANC, absolute neutrophil count; BMT, bone marrow

transplantation was performed. Sample 25 was a parent and thus was not included.

Short Report 831

� 2002 Blackwell Publishing Ltd, British Journal of Haematology 119: 830–832

patients analysed, the proportion of red cells deficient inCD59 and the proportion of neutrophils with a decreasedexpression of CD59 or CD24 were less than 1%, which wasnot significantly different from the proportion of blood cellsdeficient in GPI-linked proteins identified in 10 normalcontrol individuals. These findings indicated that none ofthe 17 patients with likely or diagnosed SDS had adetectable clone of PNH cells.

This is the first systematic screening for PNH cells inpatients with an inherited form of bone marrow failure.Although the children were only investigated once, therather large number and the different disease stages,including one child whose disease progressed to MDS,suggest that there is no increased incidence of PNH in SDS.PIGA gene mutations occur in normal individuals. How-ever, owing to the lack of a growth advantage, theproportion of PNH cells are usually not detectable (< 1%)(Araten et al, 1999). The occurrence of PNH cells inacquired AA and certain forms of MDS suggests that PNHcells experience a growth or survival advantage in the bonemarrow environment that prevails in these disorders. Bonemarrow failure in the majority of patients with the acquiredform of AA is thought to be immune mediated (Young &Maciejewski, 1997). An autoimmune mechanism has alsobeen implicated in the pathogenesis of certain forms of MDS(Dunn et al, 1999). The cause of bone marrow failure inSDS is less clear. Immunosuppressive therapy seems not toplay a role in the treatment of cytopenia in SDS. A stem celldefect and abnormal bone marrow stroma are thought to beresponsible for the defective haematopoiesis (Dror & Freed-man, 1999). Recent studies suggest that defective prolif-erative properties and an increased tendency to undergoapoptosis may contribute to bone marrow failure in SDS(Dror & Freedman, 2001). Results from our study suggestthat bone marrow failure in SDS does not select for PNHcells, suggesting that a PIGA gene mutation neither protectsSDS bone marrow cells from cell death nor improves theirdefective proliferation. This suggests that bone marrowfailure per se does not select for PNH cells. The expansion ofPNH cells in an aplastic environment is rather due to aspecific mechanism that does not prevail in SDS and that yethas to be defined. In regard to the natural history of SDS,our data suggest that the pathogenesis of bone marrowfailure that prevails in patients who develop PNH and inpatients with SDS might be different. The recent finding thatabout 23% of patients with MDS have PNH cells suggested acommon link between acquired AA and certain forms ofMDS. In SDS, the absence of PNH cells suggests that bonemarrow failure in SDS does not share this common link.

ACKNOWLEDGMENTS

We would like to thank all patients and families thatparticipated in this study. We also thank Shwachman–Diamond Syndrome International and J. M. Mowery, former

director, for their co-operation, the doctors, nurses andsecretaries, especially Cindy Terrill, for their help in dataacquisition, and Philip J. Mason for reading the manuscript.

Supported by the Mallinkrodt Foundation, NationalInstitute of Health CA-89091-2, American Cancer SocietyIRG I#N-36-39, and the Kommission zur Forderungdes akademischen Nachwuchses des Kantons Zurich,Switzerland.

REFERENCES

Araten, D.J., Nafa, K., Pakdeesuwan, K. & Luzzatto, L. (1999)

Clonal populations of hematopoietic cells with paroxysmal noc-

turnal hemoglobinuria genotype and phenotype are present in

normal individuals. Proceedings of the National Academy of Sciences

of the United States of America, 96, 5209–5214.

Bessler, M. & Fehr, J. (1991) Fc III receptors (FcRIII) on granulo-

cytes: a specific and sensitive diagnostic test for paroxysmal

nocturnal hemoglobinuria (PNH). European Journal of Haematol-

ogy, 47, 179–184.

Bessler, M., Schaefer, A. & Keller, P. (2001) Paroxysmal nocturnal

hemoglobinuria: Insights from recent advances in molecular

biology. Transfusion Medicine Reviews, 15, 255–267.

Dacie, J.V. & Gilpin, A. (1944) Refractory anaemia (Fanconi type).

Its incidence in three members of one family, with in one case a

relationship to chronic haemolytic anaemia with nocturnal

haemoglobinuria (Marchiafava–Micheli disease or �nocturnal

haemoglobinuria�). Archives of Disease in Childhood, 19, 155–

162.

Dror, Y. & Freedman, M.H. (1999) Shwachman–Diamond syn-

drome: An inherited preleukemic bone marrow failure disorder

with aberrant hematopoietic progenitors and faulty marrow

microenvironment. Blood, 94, 3048–3054.

Dror, Y. & Freedman, M.H. (2001) Shwachman–Diamond

syndrome marrow cells show abnormally increased apop-

tosis mediated through the Fas pathway. Blood, 97, 3011–

3016.

Dunn, D.E., Tanawattanacharoen, P., Boccuni, P., Nagakura, S.,

Green, S.W., Kirby, M.R., Kumar, M.S., Rosenfeld, S. & Young,

N.S. (1999) Paroxysmal nocturnal hemoglobinuria cells in

patients with bone marrow failure syndromes. Annals of Internal

Medicine, 131, 401–408.

Ginzberg, H., Shin, J., Ellis, L., Morrison, J., Ip, W., Dror, Y.,

Freedman, M., Heitlinger, L.A., Belt, M.A., Corey, M., Rommens,

J.M. & Durie, P.R. (1999) Shwachman syndrome: phenotypic

manifestations of sibling sets and isolated cases in a large patient

cohort are similar. Journal of Pediatrics, 135, 81–88.

Huijgens, P.C., van der Veen, E.A., Meijer, S. & Muntinghe, O.G.

(1977) Syndrome of Shwachman and leukaemia. Scandinavian

Journal of Haematology, 18, 20–24.

Shwachman, H., Diamond, L.K., Oski, F.A. & Khaw, K.T. (1964)

The syndrome of pancreatic insufficiency and bone marrow

dysfunction. Journal of Pediatrics, 65, 645–663.

Smith, O.P., Hann, I.M., Chessells, J.M., Reeves, B.R. & Milla, P.

(1996) Haematological abnormalities in Shwachman–Diamond

syndrome. British Journal of Haematology, 94, 279–284.

Young, N.S. & Maciejewski, J. (1997) The pathophysiology of

acquired aplastic anemia. New England Journal of Medicine, 336,

1365–1372.

832 Short Report

� 2002 Blackwell Publishing Ltd, British Journal of Haematology 119: 830–832