| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Division of Hematology and Internal Medicine (Drs. Dingli and Tefferi), Division of Pulmonary and Critical Care Medicine (Drs. Utz and Krowka), and Department of Biostatistics and Health Science Research (Dr. Oberg), Mayo Clinic and Mayo Foundation, Rochester, MN.
Correspondence to: Ayalew Tefferi, MD, Division of Hematology, Mayo Clinic, 200 First St SW, Rochester, MN 55905; e-mail: tefferi.ayalew{at}mayo.edu
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: Retrospective chart review of patients who had received diagnoses of both CMD and PH. Patients with a known cause of PH were excluded. The diagnosis of a CMD was based on established criteria. The diagnosis of PH was based on echocardiographic data or right heart catheterization data.
Results: Twenty-six patients satisfied the criteria for both a CMD and PH. Twelve patients had myeloid metaplasia with myelofibrosis (MMM), 5 patients had essential thrombocythemia (ET), 6 patients had polycythemia vera, 2 patients had a myelodysplastic syndrome, and 1 patient had chronic myeloid leukemia. Twenty-two patients (92%) received treatment for their CMDs, which included therapy with hydroxyurea (18 patients), anagrelide (7 patients), and busulfan (3 patients). PH was diagnosed a median of 8 years after recognition of the CMD (range, 0 to 26 years). The median right ventricular systolic pressure (RVsys) was 71 mm Hg (range, 32 to 105 mm Hg). RVsys correlated with the platelet count in patients with MMM (r = 0.30) and ET (r = 0.6) and with the hemoglobin levels in patients with PV (r = 0.77). Treatment of CMD did not seem to affect the severity of the pulmonary artery pressures as measured by serial echocardiography. With a median survival time of 18 months after the diagnosis of PH, the cause of death in the majority of the patients was cardiopulmonary.
Conclusions: The current study suggests a higher than expected incidence of PH in patients with MMM, PV, and ET. Prognosis in such a setting is poor and may not be influenced by aggressive treatment of the underlying hematologic disorder.
Key Words: essential thrombocythemia • myelofibrosis with myeloid metaplasia • myeloproliferative disorders • polycythemia vera • pulmonary hypertension
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The classification of CMDs is based on clinical features and findings on bone marrow histology as well as cytogenetics studies.1 An operational classification of the CMD separates the chronic myeloproliferative disorders (CMPDs) (ie, polycythemia vera [PV], essential thrombocythemia [ET], and myelofibrosis with myeloid metaplasia [MMM]) from chronic myeloid leukemia (CML) as well as from the myelodysplastic syndromes.1 Among the CMPDs, PV and ET have relatively indolent clinical courses that may be interrupted by recurrent thrombohemorrhagic complications, while MMM has the worst prognosis (median survival time, 5 years) and has a higher transformation rate to acute leukemia.1
Isolated pulmonary complications are infrequent in the CMDs. When they occur, they are usually caused either by infection or by thromboembolic events. Pulmonary or pleural extramedullary hemopoiesis is a rare complication that may be associated with MMM.
The possible association of pulmonary hypertension (PH) with CMPD has been suggested by a few case reports2 3 4 and by one cohort of six patients.5 In a retrospective study, we explored the possible associations between various CMDs and PH and reviewed the outcomes of the affected patients.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The initial search yielded 124 charts with both diagnoses. The diagnosis of CMD was based on previously established criteria.1 7 8
PH is defined by a mean pulmonary artery pressure (PAP) of > 25 mm Hg at rest or > 30 mm Hg with exercise.9 The diagnosis of PH was based on measurements obtained from Doppler transthoracic echocardiography (TTE), right heart catheterization, or pulmonary angiography. PH was diagnosed if the patient had an estimated right ventricular systolic pressure (RVsys) of > 35 mm Hg, as measured by TTE, or a mean PAP pressure of > 25 mm Hg during right heart catheterization or pulmonary angiography procedures.9 Patients were excluded from further analysis if any of the following causes of PH was evident: documented pulmonary embolism; moderate-to-severe COPD (ie, FEV1 < 40% of predicted maximum); moderate-to-severe restrictive lung disease (ie, total lung capacity < 70% of predicted maximum); left-sided cardiac failure of any etiology; cyanotic congenital heart disease; connective tissue disorders with pulmonary involvement; carcinoid syndrome; chronic liver disease10 ; sleep apnea syndromes; HIV infection9 ; or use of anorectic medications.
All patients had provided an authorization to use their medical records for research,11 and the study was approved by the Institutional Review Board of the Mayo Clinic.
Reviewing the charts necessary for testing the hypothesis of independence of these two conditions was not feasible. Hence, we calculated the expected number of patients in the United States during the 13-year period of interest. We compared the number of patients seen in the referral population of the Mayo Clinic (a subset of the US population) to this number in search of evidence that there may be an association between the two conditions. The expected number of CMD cases per year was found by multiplying the CMD incidence rate by the average yearly US population from 1987 to 2000.
This count was multiplied by the incidence rate of unexplained PH in order to determine the expected number of cases with both conditions, assuming that they are independent. This number then was multiplied by 13 to get the expected number of cases occurring over the time period of interest.
Spearman’s correlation was calculated using computer software (JMP; SAS Institute; Cary, NC).
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The remaining 26 patients constituted our study population (Table 1 ). Twelve patients had MMM, and of these 4 had postpolycythemic myeloid metaplasia. Six patients had PV, and five patients had ET. One patient had CML with a high platelet count, one female patient had the 5q-syndrome, and one patient had refractory anemia with ring sideroblasts.
|
|
|
-interferon, and
another 2 patients received danazol. One patient received vincristine,
one patient received chlorambucil, and another patient was treated with
cyclophosphamide and 2-chlorodeoxyadenosine. Of the seven patients receiving anagrelide, three had been receiving the drug for an average of 19 months before PH was detected. The other four patients were prescribed the drug either concomitant with the diagnosis of PH or after (two patients in each category). Sixteen patients (61.5%) were receiving antithrombotic therapy when PH was diagnosed, and of these patients 5 were receiving both aspirin and warfarin, 8 were receiving aspirin alone, and 3 were receiving warfarin.
Eighteen patients (69%) had splenomegaly, and 2 patients had portal hypertension. Seven patients (29%) had a splenectomy. Three patients had received a diagnosis of PH before undergoing the procedure (by TTE), while in the other 4 patients, PH was diagnosed a mean of 2.7 years after they underwent the splenectomy.
Severity of PH
All patients had undergone Doppler TTE. The median RVsys at the
time of diagnosis was 71 mm Hg (range, 32 to 105 mm Hg). Twenty-four
patients (92%) had symptoms related to PH, and these led to the
diagnosis. Of these 24 patients, 20 (76.9%) had dyspnea on exertion
and 13 (50%) had edema or signs of right-sided heart failure. One
patient had chest pain, but none had syncope. Testing to rule out
pulmonary embolism was performed in 16 patients (61.5%). Of these, 9
patients (34.6%) underwent ventilation-perfusion lung scanning, 10
(38.5%) underwent ultrafast contrast CT scanning of the chest, 7
(26.9%) underwent Doppler ultrasounds of the lower extremities, and 2
(7.7%) underwent lower-extremity plethysmography. The results of all
these tests for thromboembolism were normal. Five patients underwent a
right heart catheterization, and in these patients the results of
invasive PAP recordings were similar to those obtained by TTE (Table 4
).
|
|
Mortality
Based on death certificate data, 5 patients are alive and 21
patients are dead. The cause of death was available for 18 of the 21
deceased patients. Eight patients (38%) died from congestive heart
failure, four patients (19%) died from pneumonia, two patients died
from sudden cardiac death, two patients died from hemorrhagic strokes,
one patient died from septicemia, and another patient died from
uncontrollable seizures. Based on the patient’s age at the last visit
through July 31, 2000, the median survival time after the diagnosis of
PH was 18 months (range, < 1 to 108 months). One patient died within
a month of the diagnosis. The median age at the time of death was 79.4
years (age range, 48.2 to 92.6 years) No autopsy examinations were
conducted.
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
However, the annual incidence rate of primary PH is only about 0.2 cases per 100,000 population,9 while that of the CMD is of the order of 1 to 8 cases per 100,000 population.13 14 With an average US population between the years 1987 and 2000 of 257 million, we would expect to see 0.04 new cases with both idiopathic PH and a CMD per year in the entire country under the assumption of independence of the two diseases (or 0.53 new cases in 13 years).
The fact that there have been 26 such cases in the referral population of the Mayo Clinic in 13 years provides some indication that these two diseases are related. The supposition is further supported by the substantial difference in the age of onset of PH between that in the current cohort of patients with CMD (median age, 74 years) and the primary PH that usually affects patients in the third or fourth decade.9
In the current study population, 14 of the 26 patients had elevated platelet counts (median platelet count, 593 x 109 cells/L). In patients with MMM, there was a weak correlation between platelet counts and PAP at the time of diagnosis (r = 0.30).
These results are in keeping with the clinical observations of Marvin and Spellberg,2 who reported a fall in PAP with the control of the platelet count in their patient. There is a weak relationship between hemoglobin levels and PAP (r = 0.17) but not WBC count (r = -0.23). In patients with ET, there was a positive relationship between platelet counts and PAP at diagnosis (r = 0.6). In the six patients with PV, hemoglobin levels correlated with PAP (r = 0.77).
Furthermore, the data from the current study suggest that the occurrence of PH in the setting of a CMD may be specific to the CMPD (ie, MMM, ET, or PV). This particular observation raises the potential pathogenetic relevance of thrombocytosis, CMPD-associated hypercoagulable states, and erythroblastosis.
Response to Hematologic Treatment
There is anecdotal evidence of improvement in PAP levels with
control of the CMD.2
4
Eleven of our patients had serial
measurements of PAP over time, and none of them showed a significant
improvement despite good disease control. Thus, once PH develops, it
may become independent of the primary underlying disorder, or there may
be other aspects of the CMD that cause the PH, or these two disorders
may be unrelated.
The prognosis is poor for patients who develop PH. The median time to death was 18 months after the diagnosis. In general, the higher the RVsys, the worse the prognosis. In 10 of the 21 deceased patients, PH seemed to contribute to death, with 8 patients dying from heart failure and 2 patients dying from sudden cardiac death. These are all known complications of PH.
Etiology
The CMD can elevate PAPs by various potential mechanisms. All can
cause a hypermetabolic state resulting in high cardiac output
states that are potentially reversed by a combination of transfusion,
chemotherapy, or, if indicated, splenectomy. Measuring PAP in such
patients without due consideration of their metabolic state can result
in a false diagnosis of PH since PAP may be elevated without an
increase in pulmonary artery resistance.10
Thus, a right
heart study with full hemodynamic evaluation is necessary in this
circumstance. Six of our patients had a high cardiac output state. All
of the patients had MMM, and two of these patients had a right heart
study (patients 2 and 12 in Table 4
). In two other patients, PH
developed after they underwent a splenectomy.
Marvin and Spellberg2 have reported the direct obstruction of pulmonary arteries by circulating megakaryocytes. Two patients also have been reported5 15 16 to have pulmonary parenchymal infiltration by hemopoietic cells and extramedullary hemopoiesis.
Platelets seem to play a central role in the etiology of this process. Most reports describe patients who had elevated platelet counts, and there are reports of patients whose symptoms mirrored the changes in platelet counts.2 Platelet-derived growth factor released from activated platelets is a strong stimulus for smooth muscle hyperplasia,17 and in an animal model of PH the control of the platelet count retards the development of PH.18 Twelve of our patients were receiving aspirin before PH was diagnosed, and this suggests that aspirin may not be effective in preventing the onset of PH in these patients.
Often, patients with CMDs have evidence of chronic disseminated intravascular coagulation19 and this can result in microthrombosis in the pulmonary circulation with PH as a consequence. None of our patients had chronic disseminated intravascular coagulation.
A significant proportion of patients with MMM develop portal hypertension (17% in one series 20 ), and there is a well-known association between portal hypertension and PH.10 21 This may explain the coexistence of the conditions in some patients. Two of our patients had portal hypertension with gastric varices. In patients with PV, thromboembolism and in situ thrombosis occur in about 27% of patients.3
This prothrombotic tendency is due, in part, to the high hematocrit and, in 50% of patients, to elevated platelet counts. These platelets exhibit an increased affinity for fibrinogen and so are functionally prothrombotic.22 Unrecognized, recurrent thrombotic events can result in PH. One of our six patients with PV was receiving warfarin and three patients were receiving aspirin at the time of the diagnosis of PH.
The chemotherapy used in treating these conditions can induce pulmonary parenchymal damage or pulmonary veno-occlusive disease with secondary PH. The most notable of the agents used is busulfan.23 Three of our patients were treated with busulfan. None of them had evidence of pulmonary toxicity on pulmonary function testing.
Anagrelide is being used more often to control platelet counts in patients with CMDs. Edema was reported in about 20% of the first 577 patients studied, 24 and this rate was attributed to the potent vasodilator properties of anagrelide. Fourteen patients (2.4%) developed congestive heart failure, and another patient developed cardiomyopathy. There were no echocardiographic data reported for these patients. In that cohort of 577 patients, pulmonary fibrosis consistent with a drug reaction was reported in one patient, which was proven by biopsy specimen.24 Seven patients in our group have been treated with anagrelide.
Four patients received anagrelide either at the time of diagnosis of PH or after, while the other three patients had taken the drug for a median of 19 months before PH was diagnosed. The small number of patients precludes any statistical analysis of any potential association, but we noted that most of our patients (87%) developed PH without any exposure to the drug.
Hoper et al25 have reported on seven patients who developed PH a mean of 21.3 years after undergoing splenectomies. None of these patients had a CMPD.
Splenectomy, especially in patients with MMM, often results in worsening extramedullary hemopoiesis with the potential for pulmonary infiltration.5 13 In addition, splenectomy in CMD is usually associated with marked elevations in platelet counts and often with a tendency to thromboembolism due to the functionally abnormal platelets. In part, this may explain the marked difference in the interval to development of PH in their group of patients and ours (21.3 vs 2.7 years, respectively). Whether aggressive platelet control after splenectomy in patients with a CMD can prevent the development of PH is not known.
Our study has a number of limitations, mostly because it is retrospective. The main limitation of the study is that pulmonary embolism was not ruled out objectively in all patients, although none had a history of such an event. Doppler TTE can miss pulmonary venous hypertension as a cause of PH, and when this is a possibility it is important for the patient to undergo a right heart catheterization. In addition, no lung tissue was available for testing to establish the presence of pulmonary vascular abnormalities. Our search strategy depended heavily on the proper listing and inclusion of all diagnoses by the physicians caring for these patients.
Thus, we made no attempt to calculate a prevalence of the two conditions together. It is possible that there are many patients with a CMD and unrecognized PH. The incidence of two-disease cases tends to be greater in the large referral population seen at the Mayo Clinic than in the general population. This often leads to a referral bias in studies such as these, causing the incidence or prevalence rates based on our referral population to be too high.26 The 26 cases seen in the referral population of the Mayo Clinic far exceed the 0.53 cases expected in the entire United States in this time frame.
Our data provide evidence for a possible cause and effect relationship between CMPD and PH. The probability of coexistence of the two together by pure chance alone is very small, while plausible mechanisms linking the two exist. The etiology of PH in these patients is probably multifactorial in origin. The development of PH in patients with a CMPD confers a bad prognosis and may result in their death. Larger and prospective studies are required to explore further possible links between these two entities.
|
Footnotes |
|---|
Received for publication January 5, 2001. Accepted for publication April 3, 2001.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. Di Stefano Pulmonary Arterial Hypertension and Chronic Myeloproliferative Disorders Am. J. Respir. Crit. Care Med., September 1, 2006; 174(5): 616 - 616. [Full Text] [PDF]
|
|
|
|
 |
|
 C. Arana-Yi, A. Quintas-Cardama, F. Giles, D. Thomas, A. Carrasco-Yalan, J. Cortes, H. Kantarjian, and S. Verstovsek Advances in the Therapy of Chronic Idiopathic Myelofibrosis Oncologist, September 1, 2006; 11(8): 929 - 943. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Hoffman and M. Xu Is bone marrow fibrosis the real problem? Blood, May 1, 2006; 107(9): 3421 - 3422. [Full Text] [PDF]
|
|
|
|
 |
|
 J.-J. Kiladjian, C. Lamberto, T. Laperche, H. Nunes, H. Mohamed, D. Valeyre, J.-P. Richalet, and P. Fenaux Evidence for Pulmonary Vascular Disease Despite Absence of Overt Pulmonary Arterial Hypertension (PAH) in Myeloproliferative Disorders (MPD) (PV and ET). Blood (ASH Annual Meeting Abstracts), November 16, 2005; 106(11): 4939 - 4939. [Abstract]
|
|
|
|
 |
|
 G. Simonneau, N. Galie, L. J. Rubin, D. Langleben, W. Seeger, G. Domenighetti, S. Gibbs, D. Lebrec, R. Speich, M. Beghetti, et al. Clinical classification of pulmonary hypertension J. Am. Coll. Cardiol., June 16, 2004; 43(12_Suppl_S): 5S - 12S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Tefferi The Forgotten Myeloproliferative Disorder: Myeloid Metaplasia Oncologist, June 1, 2003; 8(3): 225 - 231. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
