Mehran Score for contrast-induced nephropathy after PCI

Clinical background

Contrast-induced nephropathy, now usually termed contrast-associated acute kidney injury, is one of the commonest complications after percutaneous coronary intervention (PCI) and a substantial cause of hospital-acquired acute kidney injury. The risk varies widely between patients, from below 5 per cent in low-risk patients to over 50 per cent in those with several coexisting risk factors. The decision on preventive measures, above all periprocedural hydration and minimisation of contrast volume, must be weighed against the patient's overall risk and the circumstances of the procedure. The Mehran score was developed precisely to combine several known risk factors into a single numerical measure and thereby make risk stratification systematic rather than intuitive.

Calculating the Mehran score

The score is calculated as a weighted sum of eight variables:

$$\text{Mehran} = 5 \cdot H_{\text{hypotension}} + 5 \cdot H_{\text{IABP}} + 5 \cdot H_{\text{heart failure}} + 4 \cdot H_{\text{age}>75} + 3 \cdot H_{\text{anaemia}} + 3 \cdot H_{\text{diabetes}} + \left\lfloor \frac{V_{\text{contrast}}}{100} \right\rfloor + P_{\text{eGFR}}$$

where $H$ is an indicator variable (0 or 1) for each binary factor, $V_{\text{contrast}}$ is the contrast volume in millilitres, and $P_{\text{eGFR}}$ is the renal function score:

$$P_{\text{eGFR}} = \begin{cases} 0 & \text{if eGFR} \geq 60 \ 2 & \text{if eGFR } 40\text{--}59 \ 4 & \text{if eGFR } 20\text{--}39 \ 6 & \text{if eGFR} < 20 \end{cases}$$

in units of mL/min/1.73 m².

The binary variables are defined as follows: hypotension means a systolic blood pressure below 80 mmHg for at least one hour requiring inotropic support or an intra-aortic balloon pump within 24 hours periprocedurally; use of an intra-aortic balloon pump refers to periprocedural insertion; heart failure is defined as NYHA class III/IV or a history of pulmonary oedema; anaemia is defined as a baseline haematocrit below 39 per cent in men and below 36 per cent in women; diabetes mellitus requires an established diagnosis.

The derivation cohort consisted of 8,357 patients who underwent PCI at Columbia University Medical Center in New York during the period up to 2004 [1]. The cohort was randomly split into a development dataset of 5,571 patients and a validation dataset of 2,786 patients. The outcome was contrast-induced nephropathy, defined as an increase in serum creatinine of at least 25 per cent and/or at least 0.5 mg/dL (44 µmol/L) within 48 hours of PCI compared with baseline. Multivariable logistic regression identified eight independent predictors with a p value below 0.0001, and the weighting was based on the respective odds ratios. Patients with acute myocardial infarction were excluded from the derivation cohort, which limits generalisability to primary PCI for STEMI.

Interpretation in practice

The score is divided into four risk bands. In the derivation cohort these corresponded to the following observed incidence of contrast-induced nephropathy [1]:

For low-risk patients (score ≤5) the absolute risk is relatively low but not negligible. Standard hydration with an isotonic crystalloid should be considered, and the contrast volume kept as low as the procedure allows. Nothing further is required beyond routine monitoring of serum creatinine.

At moderate risk (score 6–10), periprocedural hydration should be given systematically, preferably with isotonic saline according to established schedules. The contrast volume should be planned in advance and kept below the maximum allowable, which can be calculated with Cigarroa's formula (5 × body weight in kg / serum creatinine in mg/dL) [2]. Nephrotoxic drugs should be withheld periprocedurally.

At high and very high risk (score ≥11), the incidence of contrast-induced nephropathy is substantial, and the risk of requiring dialysis rises markedly. Here it should be considered whether PCI is the optimal strategy or whether an alternative imaging technique without iodinated contrast can be used. If PCI is necessary, hydration should be intensified, the contrast volume aggressively minimised, and postprocedural monitoring of renal function ensured. Nephrology consultation before the procedure may be warranted.

Validation and performance

In the original validation cohort the score showed moderate discrimination, with a c-statistic of 0.67 [1]. Risk rose exponentially with increasing score, and the relationship was statistically significant (Cochran–Armitage trend test, p < 0.0001).

In a Thai single-centre study of 217 STEMI patients undergoing primary PCI, the Mehran score achieved an AUC of 0.78 (95% CI 0.69–0.87) for predicting contrast-induced nephropathy [3]. The incidence in this cohort was 19.8 per cent, higher than in the derivation cohort and reflecting the different risk profile of acute STEMI patients. The study developed a simpler alternative, the CCIT score, based on only three variables (ejection fraction below 40 per cent, three-vessel disease and IABP use), which achieved an AUC of 0.83, though without statistically significant superiority over the Mehran score (p = 0.082) [3].

An Indian prospective observational study of 55 high-risk patients undergoing elective angiography or PCI confirmed the ability of the score to stratify risk [2]. In this cohort, in which all patients received periprocedural hydration and the contrast volume was kept strictly limited (50–100 mL), the observed CIN incidence was 25.5 per cent. The distribution across risk bands corresponded well with the original cohort: low-risk patients (score ≤5) had an observed CIN incidence of 28.6 per cent, while half the patients in the very high risk group (score ≥16) developed CIN. The high incidence in the low-risk group despite preventive measures illustrates that the score may overestimate the protection at low values in selected high-risk populations [2].

A comparative study of 422 STEMI patients undergoing primary PCI tested six different risk scores against one another [4]. The Mehran score performed well for the narrow CIN definition (rise in serum creatinine ≥0.5 mg/dL) with a c-statistic in the upper part of the range, but less well for the broad definition (≥0.5 mg/dL and/or ≥25 per cent increase), where all the scores had c-statistics between 0.555 and 0.643. The ACEF and AGEF scores had better discrimination and calibration in this population [4].

In 2021 Mehran herself published an updated risk score based on a contemporary cohort of more than 20,000 PCI procedures at Mount Sinai in New York [5]. The new score uses the AKIN definition of acute kidney injury and includes variables such as ejection fraction, periprocedural bleeding and complex PCI anatomy. The c-statistic in the validation cohort was 0.84 for the pre-procedural model and 0.86 when procedural variables were included. The original 2004 Mehran score nonetheless remains the most widely used and most cited in clinical practice.

Limitations

Several limitations should be borne in mind. First, patients with acute myocardial infarction were excluded from the derivation cohort [1], which means that the performance of the score in primary PCI for STEMI is less well supported. External validations in STEMI populations have shown varying results, and simpler scores have performed better in this subgroup in some studies [3, 4].

Second, the score was developed in an era in which both high- and low-osmolar contrast media were used. With modern iso-osmolar contrast media and improved procedural techniques, the absolute risk at a given score may be lower than in the original cohort. The definition of contrast-induced nephropathy has also evolved; the AKIN and KDIGO criteria are now standard, and these capture more cases than the original definition with a fixed creatinine threshold.

Third, the contrast volume is a procedural variable not known until the procedure is planned or under way. The score can therefore be used pre-procedurally with an estimated volume, but the actual volume may differ. This is an inherent weakness, addressed in part in the updated 2021 score, in which a pre-procedural model (Model 1) is separated from a model including procedural variables (Model 2) [5].

Fourth, the score was developed for patients undergoing PCI and should not be applied to other contrast exposure, for example computed tomography, without separate validation. Patients with end-stage renal failure already on dialysis should not be scored; contrast-induced nephropathy is not a relevant outcome for them.

Finally, the score is a tool for risk stratification, not for decision-making in itself. Minimising contrast volume and ensuring periprocedural hydration are the most important modifiable factors whatever the score, and a low score does not justify omitting these measures.

References

1. Mehran R, Aymong ED, Nikolsky E, et al. A simple risk score for prediction of contrast-induced nephropathy after percutaneous coronary intervention: development and initial validation. J Am Coll Cardiol. 2004;44(7):1393–9. PMID: 15464318
2. Gupta H, Singh MM, Sahani KK, et al. Evaluation of Emerging Predictors for Contrast-Induced Nephropathy in High-Risk Patients Undergoing Percutaneous Coronary Intervention. Cureus. 2024;16(7):e64363. PMID: 39130830
3. Koowattanatianchai S, Chantadansuwan T, Kaladee A, et al. Practical Risk Stratification Score for Prediction of Contrast-Induced Nephropathy After Primary Percutaneous Coronary Intervention in Patients With Acute ST-Segment Elevation Myocardial Infarction. Cardiol Res. 2019;10(6):350–7. PMID: 31803333
4. Liu YH, Liu Y, Zhou YL, et al. Comparison of Different Risk Scores for Predicting Contrast Induced Nephropathy and Outcomes After Primary Percutaneous Coronary Intervention in Patients With ST Elevation Myocardial Infarction. Am J Cardiol. 2016;117(12):1896–903. PMID: 27161818
5. Mehran R, Owen R, Chiarito M, et al. A contemporary simple risk score for prediction of contrast-associated acute kidney injury after percutaneous coronary intervention: derivation and validation from an observational registry. Lancet. 2021;398(10315):1974–83. PMID: 34793743