The utility of testing erythropoietin level in polycythemia diagnosis

ABSTRACT

Objectives:

Polycythemia vera (PV) is classically thought to be associated with low erythropoietin (EPO) levels. Here, we present a review of the utility of using EPO levels in diagnosing polycythemia.

Methods:

We conducted a systematic literature review of the Medline data through Pubmed and Google Scholar. We included the articles which described confirmed PV associated with elevated EPO level. Our search strategy included the following terms in Pubmed (((polycythemia vera[MeSH Terms]) OR (jak2 protein tyrosine kinase[MeSH Terms])) OR (Myeloproliferative Disorders[MeSH Terms])) AND (Erythropoietin[MeSH Terms]), and ‘polycythemia vera with erythropoietin’ in Google Scholar.

Results:

Our research yielded four cases of PV with elevated EPO levels. The most common symptom was a headache. Thrombotic phenomena happened in a single case in the form of Budd-Chiari syndrome. The mean Hb level was 20.2 gm/dl, and the EPO level was 213 mlU/mL.

Discussion:

Although PV is usually associated with low EPO levels, high levels do not exclude this diagnosis. Workup should include testing for JAK2 mutation and bone marrow biopsy in the presence of suggestive signs and symptoms. Novel biomarkers are also being proposed to aid in the diagnosis.

Conclusion:

Although elevated EPO levels suggest secondary causes of polycythemia, cases where elevated EPO levels were associated with an underlying PV are reported in the literature, and we have summarized a review of them. Workup for polycythemia should include JAK2 mutation testing if signs and symptoms suggest PV even if EPO is elevated.

KEYWORDS:

• Polycythemia vera
• PV
• erythropoietin
• EPO
• myeloproliferative neoplasms
• JAK2 mutation

Introduction

Polycythemia is an increase in red blood cell (RBC) mass, as reflected by the increase in hematocrit (HCT) or hemoglobin (Hb) [1]. It can be classified into secondary polycythemia, in which erythrocytosis is driven by another factor such as hypoxemia, and primary polycythemia, i.e. polycythemia vera (PV), in which erythrocytosis is an autonomous phenomenon [1]. Differentiating between the two disease entities is essential as their treatment is entirely different. Classically, the erythropoietin (EPO) level is used to differentiate between these two entities, as it is frequently high in the former and low in the latter [2]. WHO 2008 and 2016 criteria [2,3] list low EPO level as a minor criterion for PV diagnosis. However, normal or high EPO levels do not reliably rule out PV, especially in the presence of comorbidities and smoking history. EPO level is subject to unpredictable changes in the presence of underlying EPO-producing tumors such as hepatocellular and renal cell carcinoma, as well as in cases of chronic or intermittent hypoxemia like in smoking, chronic obstructive pulmonary disease (COPD) and obstructive sleep apnea (OSA) [1].

The current gold standard for diagnosis of PV is molecular testing for the presence of JAK-2 mutation in addition to bone marrow biopsy that shows hypercellularity with trilineage growth (panmyelosis) [4].

Here, we present a review of the utility of using EPO levels in diagnosing polycythemia and an illustrative case where elevated EPO was associated with an underlying PV diagnosed based on JAK-2 mutation detection and bone marrow findings. We also present a literature review of similar cases.

Material and methods

We have conducted a systematic literature review in the Medline data through Pubmed (years 2008 till 2022) and performed a comprehensive search of Google Scholar data without a duration limit. Additional articles were identified by checking the references of relevant articles, and duplicates were excluded. We included the articles that described confirmed PV associated with elevated erythropoietin levels. Our search strategy included the following terms in Pubmed (((polycythemia vera[MeSH Terms]) OR (jak2 protein tyrosine kinase[MeSH Terms])) OR (Myeloproliferative Disorders[MeSH Terms])) AND (Erythropoietin[MeSH Terms]), and ‘polycythemia vera with erythropoietin’ in Google Scholar.

In total, 1217 records were identified through electronic database searches; 1200 of these records were excluded, and then 17 full-text papers were assessed for eligibility. In total, four papers were included in the review. Two authors (A.I. and A.A.) separately screened the identified articles based on title and abstract. We excluded any PV associated with low EPO levels. We also excluded articles written in languages other than English. Data included demographics, clinical characteristics, risk factors, and molecular studies.

Case presentation

A 47-year-old female with no significant past medical history presented complaining of severe frontal headache radiating to her neck, associated with dizziness and epigastric pain. She also reported left-sided body weakness for one day, which ultimately resolved later.

There was no history of double or blurred vision, dysphagia, speech difficulty, loss of consciousness, shortness of breath, cough, sputum production, wheezes, chest pain, or palpitations. Moreover, there was no history of lower limb swelling, orthopnea, or paroxysmal nocturnal dyspnea. Furthermore, there was no history of dysuria, urgency, or frequency. Additionally, there was no fatigue, abdominal pain, weight loss, or fever. She denied cigarette smoking, alcohol drinking, and drug use.

Her initial vital signs were as follows: Temperature of 36.8 degrees Celsius, blood pressure of 166/82 mmHg, pulse of 76 bpm, respiratory rate of 20 breaths per minute, and oxygen saturation of 98% breathing room air. Physical examination was unremarkable apart from the plethoric face. There was no evidence of central or peripheral cyanosis, added sounds on chest auscultation, heart murmurs, or gallops. Furthermore, no hepatomegaly, splenomegaly, or lymphadenopathy were evident on examination. Jugular venous pulse (JVP) was flat, and lower limbs were without swelling. The neurologic exam did not show any neurologic deficits.

Initial laboratory test results are shown in Table 1.

Table 1. Laboratory results upon admission.

Lab	Value w/Units	Flags	Normal Range
White blood cell (WBC) count	25.4x10^3/uL	High	4.0–10.0
Red blood cell (RBC) count	5.8x10^6/uL	High	3.8–4.8
Hemoglobin (Hb)	17.1gm/dL	High	12.‏0–15.0
Hematocrit	52.‏2%	High	36.0–46.0
Mean cell volume	90.‏2 fL	Normal	83.0–101.0
Platelet count	547x10^3/uL	High	150–400
Basophil count	0.52 x10^3/uL	High	0.02–0.10
Urea	5.4mmol/L	Normal	2.5–7.8
Creatinine	100umol/L	High	44–80
Sodium	140mmol/L	Normal	133–146
Potassium	4.7mmol/L	Normal	3.5–5.3
Bicarbonate	22 mmol/L	Normal	22–29
Alkaline Phosphatase	173 U/L	High	35–104
Alanine transaminase	20U/L	Normal	0–33
Aspartate transaminase	29U/L	Normal	0–32

Peripheral smear showed erythrocytosis, neutrophilic leukocytosis, eosinophilia, basophilia, and moderate thrombocytosis with few giant platelets. A computed tomography scan of the head showed an acute thalamo-capsular infarct with an old pontine infarct. The electrocardiogram (ECG) was unremarkable. Based on the above, the erythropoietin level was sent and found to be elevated at 138.01 mIU/mL (normal range 2.59–18.50 mIU/mL). An abdomen ultrasound showed normal spleen and liver size with no apparent renal mass or other intra-abdominal abnormalities. A whole-body computed tomography scan to rule out malignancy could not be done due to acute kidney injury.

Based on the findings of erythrocytosis, leukocytosis, and acute ischemic stroke in a young, healthy woman, an underlying myeloproliferative neoplasm (MPN) was suspected. A bone marrow biopsy showed myeloproliferative neoplasm with features of PV, which was confirmed later by a positive test for JAK-2 mutation. The patient was treated as a case of PV, started on aspirin and hydroxyurea, and was discharged home. In the hematology clinic weeks later, she reported feeling well and was found to have a decrease in her EPO level down to 40 mIU/mL.

Discussion

Polycythemia refers to an increase in the red blood cell (RBC) mass represented by an increase in hemoglobin (Hb) and hematocrit (HCT) levels. Reference ranges for Hb and HCT for adult men are 16 gm/dl +/− 2 gm/dl and 47% +/− 6%, respectively, and for adult women are 13 gm/dl +/− 2 gm/dl and 40% +/− 6% respectively [1].

Depending on the driving factor, polycythemia is either primary or secondary [1]. Secondary polycythemia is driven by erythropoietin (EPO), which can increase due to hypoxemia, as seen in people living at high altitudes, those with chronic lung diseases, and smoking. It can also be directly released by an underlying EPO-secreting tumor, such as renal cell and hepatocellular carcinoma [1]. On the other hand, PV is a myeloproliferative neoplasm (MPN) in which erythrocytosis is driven by an acquired activating mutation that increases sensitivity to EPO or causes EPO-independent erythroid colony formation [5]. Other diseases that belong to the MPN entity are essential thrombocythemia (ET) and primary myelofibrosis (PMF) [5].

Frequently, polycythemia is an incidental finding on complete blood count (CBC) performed for other purposes. However, some patients may present with symptoms of the underlying disease or symptoms of increased red blood cell mass [6]. Manifestations of PV include fatigue, excessive itching, abdominal pain, weight loss, and fever [7]. Furthermore, the most significant PV complications affecting survival are thrombotic complications such as stroke, venous thromboembolism, and leukemic transformation [7]. The risk for leukemic transformation in patients with PV is about 2–5% over 15 years and is more common in individuals older than 60 [8].

Therefore, history and physical examination play a crucial role in the diagnostic process. They should be focused on eliciting smoking history and signs and symptoms of cardiopulmonary diseases causing chronic or intermittent hypoxemia, as seen in obstructive sleep apnea [1]. On the other hand, abdominal pain, early satiety, fatigue, fever, weight loss, erythromelalgia, and hepatosplenomegaly would be suggestive of PV [1,7]. Traditionally, EPO levels guide the diagnostic process towards primary or secondary polycythemia [1]. However, this strategy has its pitfalls. As mentioned earlier, EPO is released in response to hypoxemia in the context of cardiopulmonary diseases. However, these diseases are prevalent in our population of interest, and multiple pathologic processes can coexist. For example, Kim et al. [9] described a case where secondary polycythemia due to renal artery stenosis was discovered in a patient with an established diagnosis of PV. In that case, the diagnosis of renal artery stenosis was pursued in the context of severe refractory hypertension and severely elevated EPO levels. In the case we presented above, the diagnosis of PV was suspected despite elevated EPO levels due to erythrocytosis, leukocytosis, and acute ischemic stroke in a young, healthy woman.

Of note, primary familial and congenital polycythemia are also associated with low erythropoietin levels [1], which further argues against its utility in polycythemia diagnosis.

Consequently, many experts have suggested avoiding the use of EPO and have called to revise the WHO criteria for PV diagnosis [10,11]. There are a few suggested alternative tests. Janus kinase 2 (JAK2) V617F and non-V617F mutations have a diagnostic sensitivity of 97% and specificity of nearly 100% for PV [12]. Michiels et al. have proposed using a combination of testing for Janus kinase mutations and bone marrow biopsy to increase diagnostic accuracy [12]. Bone marrow biopsy shows hypercellularity with trilineage growth (panmyelosis) in cases of PV [3].

Moreover, new biomarkers are proposed to aid the diagnostic process. Krečak et al. described using several biomarkers to differentiate PV from secondary polycythemia. According to the paper [13], high platelet-to-lymphocyte ratio (PLR) outperformed other studied biomarkers, including neutrophil-to-lymphocyte ratio, total leukocytes, neutrophils, lymphocytes, and platelets. It is, thus, suggested as a cheap and rapidly available diagnostic biomarker. However, these newly proposed biomarkers still need validation on larger datasets.

The WHO 2016 diagnostic criteria [2] for PV include clinical, morphologic, and genetic findings. Diagnosis of PV requires meeting all three major criteria or the first two major criteria and one minor criterion.

Major criteria

• Criterion 1 (clinical)

Hb >16.5 g/dL in men, > 16.0 g/dL in women or

HCT > 49% in men, > 48% in women or

Red cell mass Increased 25% above mean predicted value

• Criterion 2 (morphologic)

BM morphology: hypercellularity for age with trilineage growth (panmyelosis), including prominent erythroid, granulocytic, and megakaryocytic proliferation with pleomorphic, mature MKs (differences in size)

• Criterion 3 (genetic)

Presence of JAK2V617F, or JAK2 exon 12 mutation.

Minor criterion

• Serum Epo level: subnormal

Compared to the 2008 WHO criteria [3], the 2016 criteria upgraded bone marrow morphology to be one of the three major criteria. This reduced the Hb threshold to 16.5 gm/dl from 18.5 gm/dl in men and to 16 gm/dl from 16.5 gm/dl in women. JAK2 mutation remained a major diagnostic criterion, and subnormal EPO level remained a minor criterion as well [2]. Endogenous erythroid colony formation in vitro was discarded as a minor diagnostic criterion. These modifications were applied due to concerns regarding the underdiagnosis of PV based on the old criteria, particularly with the high Hb threshold.

It is worth mentioning that the diagnosis of PV without JAK2 mutation is unlikely. In contrast, its presence does not distinguish PV from ET or PMF, at which point differentiation is possible with bone marrow morphology examination [14].

EPO is a glycoprotein hormone mainly produced by the kidneys [15]. However, it is also produced in small amounts in the liver, spleen, bone marrow, lungs, and brain. The primary regulator of EPO production is the partial pressure of oxygen PaO2 [9]. Multiple case reports and case series suggested an association between PV with elevated EPO and Budd Chiari Syndrome [16]. Thumures et al. described four patients who were diagnosed with Budd Chiari Syndrome with elevated red cell mass and were found to have elevated serum erythropoietin levels. PV was confirmed with JAK2 mutation assay and bone marrow biopsy examination. It was suggested that liver hypoxia with subsequent hepatocyte injury and necrosis caused by Budd Chiari Syndrome leads to EPO level elevation [16].

Accordingly, we assume that patients with PV and elevated EPO have some degree of organ injury that releases EPO into circulation.

Nevertheless, little is known about the significance of elevated EPO levels in patients with PV diagnosis. We have conducted a literature review searching this entity to study the clinical significance and outcomes. Our search yielded 4 cases summarized in Table 2.

Table 2. Summary of literature review findings.

Case number	Age (year)	Sex	Co-morbidities	Hepatomegaly/splenomegaly	Smoking history	Hb (gm/dl)	HCT (%)	Thrombosis	JAK-2 mutation
Index case	47	Female	None	None	None	17	52	None	Yes
1	49	Male	None	Both	None	18	NA	None	Yes
2	64	Male	HTN, HLD, COPD	Splenomegaly	None	20	60	None	Yes
3	58	Female	None	None	None	20	NA	Yes (Bud-chiari)	Yes
4	30	Female	Renal artery stenosis	Splenomegaly	None	22.5	67	None	Yes

The cases were of three males and one female, with a mean age of 50. Two were previously healthy, while the other two had comorbid diseases, as seen in the table. None of them had a smoking history. The most common symptom was headache, followed by erythromelalgia. Thrombotic phenomena happened in a single case in the form of Budd-Chiari syndrome. Hepatosplenomegaly was found in a single case, and splenomegaly was found in two cases. The mean Hb level was 20.2 gm/dl, and the EPO level was 213 mIU/mL. In comparison, our index patient was 47 years old, without comorbidities, including smoking history. As hepatomegaly and splenomegaly were detected in 3 of the 4 cases listed in the table, this can indicate that findings of hepatomegaly or splenomegaly are more suggestive of a diagnosis of PV.

Workup to rule out secondary polycythemia in the abovementioned cases included ruling out hypoxemia and chronic lung pathology by obtaining arterial blood gas (ABG), chest radiography, and pulmonary function tests. Computed tomography of the abdomen was performed to rule out renal and hepatic masses and alpha-fetoprotein to rule out underlying germ-cell tumors. None of the cases discussed how the treatment of PV was modified or how the prognosis changed in the presence of elevated EPO levels.

Our paper highlights the importance of a systematic approach in medicine, where signs and symptoms remain the cornerstone in diagnosis and where laboratory tests should be taken in their appropriate clinical context. Our paper showed that testing EPO levels might be helpful in the appropriate clinical context but should not be used solely to guide the diagnostic process in cases of polycythemia. We did not find clinically significant changes in the prognosis or treatment of PV in the presence of elevated EPO. Consequently, and in line with the recent recommendations of several experts, we recommend testing for JAK2 mutation in polycythemia workup in the proper clinical scenario regardless of the EPO levels.

Several limitations of this study are worth mentioning. The focus of this study was the significance of EPO levels in PV diagnosis. We did not discuss other modalities of diagnosis in depth, including JAK-2 mutation analysis and bone marrow biopsy. Moreover, we did not discuss methods for differentiation between various MPNs and their clinical stages.

Future research should focus on EPO's diagnostic and prognostic significance in polycythemia. It should also focus on studying the utility of novel biomarkers such as PLR and NLR.

Conclusion

Polycythemia vera should be suspected at a Hb level of 16.5 gm/dl or HCT level of 49% in males and a Hb level of 16 gm/dl or HCT level of 48% in females. The diagnostic process includes history and physical examination, JAK2 mutation analysis, and bone marrow biopsy. New biomarkers such as PLR and NLR are being proposed to aid the diagnosis. We also recommend avoiding the use of isolated elevated EPO levels in polycythemia diagnosis as this strategy has many inaccuracies. Future research should focus on the diagnostic and prognostic significance of EPO in polycythemia and on studying the utility of novel biomarkers such as PLR and NLR.

Ethical approval and consent

The patient provided consent to share her case.

Specific author contributions

Conceptualization, Elmustafa Abdalla; Formal analysis, Abdellatif Ismail; Investigation, Abdellatif Ismail, Elmustafa Abdalla, Ashraf Ahmed, Ahmed Alsayed, Methodology, Abdellatif Ismail, Elmustafa Abdalla, Ashraf Ahmed, Ahmed Alsayed; Project administration, Abdellatif Ismail, Ashraf Ahmed and Mohamed Yassin; Supervision, Mohamed Yassin; Writing – original draft, Abdellatif Ismail, Elmustafa Abdalla, Ali Aqel, Abdalla Fadul, Ashraf Ahmed, Ahmed Alsayed, Muzamil Musa; Writing – review & editing, Abdellatif Ismail, Abdellatif Ismail, Elmustafa Abdalla, Ali Aqel, Abdalla Fadul, Ashraf Ahmed, Ahmed Alsayed, Muzamil Musa, Mohamed Yassin. All authors approved of the final submission.

Guarantor of the article: Mohamed Yassin is the guarantor of this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Financial support: If accepted, this research will be funded by Qatar National Library (QNL).