A prospective analysis of 1250 EDTA anticoagulated blood samples were conducted over duration of two months at the Department of Pathology, ESIC Medical College and PGIMSR, Bangalore, India. Random non-biased sampling technique was employed to select cases from patients attending hospital and for whom the treating physician advised a haematological investigation. The confidentiality of the patient details and ethical standards were maintained through out the study and only the relevant data was recorded in a proforma. This study was not subject to ethical clearance as only the automated analyser derived data was used for the study purpose. Since clinical indications and demographic data were not used for the study, the patients’ confidentiality was not compromised.

Fully automated multichannel haematology analyser, SYSMEX XT 2000i was used for the study. All cases irrespective of age, sex and indication sent for complete haemogram were included in the study. Clinical and demographic details were not included to avoid bias and promote homogeneity and true random sampling in the population. A fixed number of samples were chosen from the daily samples by random selection.

Incomplete platelet parameters, spurious thrombocytopenia and cases with flags for suspected haematological malignancy were excluded from the study. Hence, out of 1250 samples, 1214 samples were selected for the study.

The following RBC parameters – RBC count, HCT, MCV, RDW and haemoglobin and platelet parameters – platelet count, PCT, MPV, PDW were noted. All the present haematological parameters were obtained through the Sysmex autoanalyser and validated as our haematology laboratory undergoes regular external quality control checks as a part of the National Quality Control Programme.

The five RBC and four platelet parameters obtained from each case were statistically analysed for linear dependence using the Pearson Correlation coefficient matrix. Statplus statistical analysis software for Mac was used for computation. High, medium and low correlation was considered when the r value (Pearson Coefficent value) was in the range of - 0.5 to – 1.0 or 0.5 to 1.0; -0.3 to -0.5 or 0.3 to 0.5; -0.1 to -0.3 or 0.1 to 0.3 respectively. Comparable parameters demonstrating direct or inverse relationships were also analysed for statistical significance. A p-value of < 0.05 was taken to be statistically significant.

Pearsons correlation coefficient value for various RBC and platelet parameters in a consolidated form are presented in [Table/Fig-1].

Upon computation of relationship of five RBC parameters with four platelet parameters, the following observations were noted [Table/Fig-2].

A statistically significant direct linear relationship was found between RDW with platlet count and PCT. A statistically significant inverse linear relationship was noticed between the following parameters: i) HCT with platlet count and PCT; ii) MCV with platlet count and PCT; iii) Hb and platlet count. No association of statistical significance was seen between total erythrocyte count and any other platelet indices.

The observations derived from the present study also shows that among RBC parameters, a direct linear significant relationship between RBC count and Hb; and also between RBC count and HCT was seen. An inverse linear significant relationship between RBC count and MCV; and RBC count and RDW was also seen.

Among the platelet parameters, a direct linear significant relationship between platelet count and PCT, and an inverse linear significant relationship between platelet count and MPV; platelet count and PDW was seen. These well known observations validate the present study.

This study was based on the premise of the following two observations; one, undeniable ontogenic relationship between erythrocytes and megakaryocytes which is reflected in their maturation stages and two, RBCs and platelets are analysed in the same detector using hydrodynamic focusing and yield similar parallel comparable parameters.

Erythropoiesis and thrombopoiesis originate from a common bipotent haematopoietic stem cell–Colony Forming Unit –Erythrocyte megakaryocyte (CFU-EMk). This cell divides into divergent pathways and leads to the specific cell lines – the RBCs and the platelets [1]. During the maturation stages of an erythron, the normoblasts lose their nuclei, enter the peripheral circulation to form an immature RBC, known as reticulocyte and eventually develops into a mature RBC. Nucleated forms, named normoblasts are found only in the bone marrow and are seen in the peripheral circulation only in disease states, as in bone marrow damage, stress and potentially serious underlying disorders [6]. Reticulocytes give an estimate of erythropoiesis.

On the other hand, platelets are anucleate cytoplasmic fragments, which are derived from megakaryocytes – a nucleated cell. Platelets, like RBCs also have an immature stage in the circulation, known as reticulated platelet. Newly released platelets are called as reticulated platelets that retain residual RNA, analogous to red cell reticulocytes. Reticulated platelet counts give an estimate of thrombopoiesis and can be used in distinguishing platelet destruction syndromes from hypoplastic platelet production [1]. Nucleated forms of platelets–micromegakaryocytes are seen in the peripheral blood only in disease conditions such as myelodysplastic syndromes and myeloproliferative disorders [2,3].

Thrombopoietin and erythropoeitin belong to the same haematopoietic growth factor subfamily, are majorly produced in the kidney and similarly act by activating the JAK/STAT pathway and Ras signal transduction on their respective precursors [7].

Erythropoietin and thrombopoietin share a high degree of amino acid sequence homology (first 155 amino acids are common). Based on this theory, thrombocytosis in children with iron deficiency anaemia was explained by Bilic E et al., [8]. GATA-1, a transcription factor is expressed in primitive and definite erythroid and megakaryocytic cells and expression of both lineages are dependent on the presence of an intact GATA site [9].

The other noteworthy similarities between RBCs and platelets are that both are anucleate and lack mitochondria, both show mass changes in hyperdestructive and proliferative conditions. All the available haematology analysers determine the erythrocytes and platelet volume using the same aperture and the same dilution [5].

Only a handful of studies have been conducted stating the correlation of RBC and platelet parameters [Table/Fig-3] [10-12]. Wiwanitkit V studied plateletcrit, mean platelet volume and platelet distribution width, and defined their expected values and correlated the same with parallel red cell indices like haematocrit, mean corpuscular volume and red cell distribution width in 215 volunteers [10]. The authors demonstrated a significant correlation between PDW and RDW, whereas no significant correlation between PCT and HCT, MPV and MCV were seen. Saouli Z et al., also performed a study similar to Wiwanitkit V in 303 volunteers [11]. The author’s results were similar to the study done by Wiwanitkit V. Kodikoylu G et al., studied platelet parameters in women with iron deficiency anaemia in 87 cases and concluded that an inverse relationship between PDW and MCV was noted [12].

The direct linear relationship between RDW and platelet count/plateletcrit is observed in inflammatory conditions, for example in slow coronary flow phenomenon, a subclinical inflammatory condition, where both PCT and RDW are independent predictors [13]. Similarly, in inflammatory bowel disease, a significant relationship between RDW and platelet count has been documented [14]. Higher RDW has also been shown to concurrently occur with higher platelet counts in advanced cancer stages, particularly in lung cancer [15].

An inverse linear relationship between MCV and plateletcrit is seen in inflammatory bowel disease [14]. Another important clinical implication of this inverse relationship would be in voluntary plateletpheresis. The AABB demonstrated that higher donor platelet counts reflected a decrease in MCV, thus projecting a higher likelihood of donor iron deficiency in high plateletpheresis yields [16]. An inverse linear relationship between Hb and platelet count is documented in multiple sclerosis [17], inflammatory bowel disease [14] and various other clinical conditions [18].

An important limitation of the study was limitation related to the use of automated analysers. Only average values were quantified. Morphological scatter may be analysed for better correlation. Precision/reproducibility is much lesser for platelets than for red blood cells. However, on the brighter note, one of the highlights of the study was that already known facts showed a linear correlation, which validates the study.

Increasingly, numerous studies are demonstrating a significant association between routine haematological parameters and disease conditions. While most of these studies document similar relationships between RBC and platelet parameters as in the present study, they fail to elucidate in detail the cause for these subtle yet coincidental relationships. The present study demonstrates the inherent ontogenic association at play in these correlations; however, further in-depth evaluation is required to validate these results.

Disclosure: This manuscript was presented as oral paper in 41st Karnataka Chapter – Indian Association of Pathologists and Microbiologists, State level Conference, held at Kempegowda Institute of Medical Sciences, Bangalore. (5th to 7th September 2014).