THE HELLP SYNDROME, A PROSPECTIVE STUDY^*

Abstract

Objective.We undertook this study to evaluate the incidence and outcome of HELLP in Irish patients. In addition, duration and trends of the abnormal laboratory results were studied. Study Design.This prospective observational study screened 12068 pregnant women between January 1995 and March 1997. Any pregnant woman with hypertension, proteinuria, thrombocytopenia or anemia was monitored for hemolysis and elevated liver transaminases, from the time of recruitment till six weeks postpartum or resolution. Results.Thirteen of 12068 pregnant women (0.11%) developed HELLP. All had hypertension and 84.6% had proteinuria. Delivery was the only factor found to terminate the syndrome. Acute renal dysfunction was noted in 53.8% but none required dialysis. Laboratory parameters stabilized by the sixth postpartum day. Fetal mortality was 1 out of 14. There were no maternal deaths. Conclusions.HELLP syndrome is a rare but potentially serious complication of pregnancy. Correlation with laboratory data and early intervention are vital in achieving a favorable outcome for both mother and fetus.

• HELLP syndrome
• Ireland
• Renal dysfunction
• Prospective study

INTRODUCTION

Pregnancy induced hypertension (PIH) continues to pose a major medical challenge world wide [[1]]. In some of these women additional complications like pre-eclampsia, fatty liver of pregnancy or the HELLP syndrome can occur. The HELLP syndrome was first defined by Weinstein in 1982 in describing a case of severe pre-eclampsia complicated by Hemolytic anemia, Elevated Liverenzymes and Low Platelets [[2]]. Presentations can be varied [3-4], causing delay in diagnosis and this delay has been shown to be a major predictor of maternal mortality [[5]]. Recent reports show that the HELLP syndrome is the most common cause of acute renal failure in pregnant women [[6]]. Though the etiology is unknown, the underlying pathology is one of thrombotic microangiopathy [7-8].

Varying outcomes associated with this syndrome have been attributed to the different diagnostic criteria used. In an attempt to clarify the situation, Sibai et al proposed standardizing the criteria for diagnosis [[3]]. Questions, however, remain regarding the early detection, adequacy of diagnostic criteria and management of these patients [[4]], [[7]]. In addition there is uncertainty concerning the correlation of morbidity with the severity of thrombotic microangiopathy and the time required for resolution of the thrombotic process once the initiating factors are removed. We undertook this prospective study of the HELLP syndrome, using the diagnostic criteria of Sibai et al, to evaluate the incidence, presentation and outcome (maternal and neonatal) of the HELLP and partial HELLP syndromes in an Irish population. An additional aim was to assess the trends and duration of the abnormal laboratory results.

METHODS

All pregnant women attending the Rotunda Hospital, Dublin, a major obstetric teaching hospital, between January 1995 and March 1997 (n = 12068) were evaluated for hypertension (Blood Pressure > 140/90 mmHg in at least 2 readings more than 4 h apart), proteinuria (excretion more than 300 mg/24 h), thrombocytopenia (platelets less than 100 × 10⁹/L) and/or hemoglobin count of less than 10 g/dL. Those with an abnormality in any of these parameters were referred to an obstetric medicine clinic and a HELLP screen using criteria suggested by Sibai et al [[3]] was undertaken. These tests included the diagnosis of hemolysis by finding schistocytes in peripheral blood smears, increased serum bilirubin (>20 umol/L) and elevated serum LDH levels (> 600 i.u./L). Elevated serum aspartate aminotransferase and alanine aminotransferase were diagnosed at levels of 70 i.u/L or more. Low platelets were defined as counts < 100 × 10⁹/L. Partial HELLP was diagnosed when only one or two of these abnormalities were present.

In addition, renal function and coagulation profiles were measured. Renal dysfunction was defined as serum creatinine more than 80 umol/L. Coagulation parameters more than two standard deviations from the laboratory reference values were taken as prolonged. The severity of the thrombotic process was graded using platelet counts as suggested by Martin et al [[9]] into class 1 (platelet nadir < 50 × 10⁹/L), class 2 (between 50 and 100 × 10⁹/L) and class 3 (>100 × 10⁹/L). Monitoring of blood pressure and proteinuria were continued throughout pregnancy. Pre-eclampsia (PET) in our patients was defined as blood pressure more than 140/90 mm Hg in the third trimester, urinary protein excretion more than 3 gm/24 h with or without edema [[10]]. The modes of presentation as well as, maternal and fetal outcomes were studied. The purpose of clinic evaluation was to also adjust antihypertensive medication and to advise on appropriate timing of induction or cesarean section.

Patients and their babies were then followed up in the obstetric medicine and pediatric clinics till the return of clinical parameters to normal or six weeks postpartum, whichever was later.

RESULTS

1. Incidence

Twelve thousand and sixty eight women were screened, of whom 389 (3.2%) were referred to the specialist clinic for management of hypertension, proteinuria or thrombocytopenia. The majority (387) had hypertension and proteinuria. Thirteen of these women fulfilled the criteria for HELLP/partial HELLP syndrome, an incidence rate of 0.11% of all pregnant women attending the hospital during the study period and 3.3% of those referred to the clinic. Besides these patients, HELLP syndrome was not diagnosed in any other women attending the Rotunda hospital during the study period.

2. Maternal Characteristics of Patients with HELLP

Mean maternal age was 31.7 years (Range 19–43 years). Of these, 10 (76.9%) were primigravidae. One of the multigravidae had a past history of PET but none had suffered from HELLP previously. One patient (7.7%) had a twin pregnancy, the rest being singletons. Eleven patients (84.6%) had normal blood pressure measurements at their early antenatal visits, while the other two were referred in their third trimesters from other centers because of hypertension. Mean gestation at time of diagnosis was 32.8 weeks (range 22–40 weeks).

3. Presentation

Ten patients (76.9%) had edema, 8 (61.5%) had epigastric pain and six (46.2%) tender hepatomegaly. No patient with HELLP syndrome presented with anemia alone. The incidence of clinical and laboratory abnormalities are summarized in Table 1. Mean BP on presentation of 172/101mmHg (range 200/140–135/92). Eleven women had proteinuria, however, only 7 (53.8%) fulfilled the criteria for PET.

Table 1. Incidence and Duration of Clinical & Laboratory Abnormalities

Abnormality	% At Presentation	Overall Incidence (%)	Mean Duration Post-partum in Days (Range)
Hypertension	100	100	23.8 (2–56)
Proteinuria	84.6	84.6	Not estimated
PET	53.8	53.8	–
Hemolysis	53.8	53.8	1^*
Elevated LFT	69.2	84.6	8.4 (3–21)
Low platelets	46.2	84.6	4.3 (2–7)
DIC	15.4	15.4	1
Renal dysfunction	46.2	53.8	5 (2–10)

^*Once DIC settled and LDH levels started to drop, hemolytic parameters were not rechecked.

Of the 4 patients who presented with alanine aminotransferase levels within normal range (2–35 i.u./L) all had a subsequent rise (mean rise of 44 i.u/L) within the first 24 h after presentation. However, only 2 of these reached levels > 70 i.u/L.

Eleven (84.6%) patients overall had thrombocytopenia but low platelet counts were seen on the day of presentation in only half of these. In those with platelet counts < 100 × 10⁹/L at the time of presentation the mean drop in platelet counts over the first 24 h after presentation was 106 × 10⁹/L. In the 2 patients whose platelet count did not reach the diagnostic level, mean drop was 158 × 10⁹/ L.

Seven of thirteen had evidence of hemolysis, two of whom also had an ongoing coagulopathy. The complete HELLP syndrome was present in six (46.2%) patients and ten (76.9%) fulfilled two or more criteria. Using the thrombocytopenic scale proposed by Martin et al, five (38.5%) of our patients were class 1, six (46.2%) were class 2 and two (15.4%) were class 3. Only one patient first developed features of HELLP postpartum. The remaining 12 (92.3%) were diagnosed prenatally.

4. Maternal Outcome

There were no maternal deaths. None had spontaneous vaginal deliveries. Cesarean section was performed in nine (69.2%) and forceps or vacuum assisted delivery in the other four (30.8%) patients. Cesarean section was preferred in patients whose clinical deterioration was rapid. Fetal maturity, fetal distress and failed induction were the additional factors determining mode of delivery. Ten women (76.92%) were delivered within 48 h of presentation. Strict bed rest and control of hypertension arrested the thrombotic microangiopathy in only two patients (15.4%), but in both, recurrence of hypertension and worsening laboratory parameters necessitated delivery within twelve days of initial presentation. Two patients required platelet transfusions and one of these also received red blood cells and fresh frozen plasma. All patients received dexamethasone antepartum, 12 mg on admission and 12mg twelve hours later if not already in labor. The drugs used for blood pressure (BP) control included intravenous hydrallazine in 4 patients (30.76%), oral beta blockers in 6 (46.2%), nifedipine in 11 (84.6%) and alpha methyldopa in 3 (23.1%) cases. The choice of anti-hypertensive had no apparent effect on outcome. Mean duration of hospital stay was 9.5 days (Range 4 to 32 days).

5. Laboratory Trends

The duration of the laboratory abnormalities is summarized in Table 1. The highest level of blood pressure was noted on the day of presentation in ten (76.9%), first postpartum day in two (15.4%) and second postpartum day in one (7.7%) patient. There was no worsening of hypertension in any patient after the second day postpartum. The lowest platelet counts were on the day of presentation or first postpartum day in the majority of patients (77%). No worsening of platelet counts were seen after the fifth day postpartum and platelet counts were above 100 × 10⁹/L in all the patients by the seventh postpartum day.

ALT levels were normal at the time of presentation in four (30.8%) patients. However, the highest ALT levels were noted on the day of presentation in four (30.8%) and on the first postpartum day in five (38.5%). There was no worsening of ALT levels after the sixth day postpartum. AST levels were normal at the time of presentation in two cases. AST levels were highest on the day of presentation in six (46.2%) and first postpartum day in five (38.5%) patients. There was no worsening of AST levels after the sixth postpartum day. Figure 1 depicts these trends.

Figure 1. Recovery in laboratory parameters; x axis – Time in days, y-1 axis – Mean ALT / AST in i.u. /L, y-2 axis – Mean platelet count in (×10⁹/L).

Mean LDH level was 1508.7 i.u./L (range 1081–2152), with the highest level found in all patients at the time of presentation. Impaired renal function was noted in seven (53.8%) patients, six of whom (46.2%) had this finding on presentation. Mean peak serum creatinine level was 118 umol/L. The serum creatinine had returned to within the normal range in all patients by the tenth day postpartum. None required renal replacement therapy.

None of the patients in this series developed ARDS, ruptured liver, or a cerebrovascular accident. DIC was seen in two cases (15.4%), though in an additional eight (61.5%) prolonged prothrombin and activated partial thromboplastin times were noted. None of the latter met the criteria of DIC suggested by Sibai et al [[3]]. There was no worsening of the coagulation profile after the first postpartum day.

6. Fetal Outcome

Fourteen babies were born to thirteen mothers. There was one fetal death (7.1%). This baby was delivered by cesarean section at twenty-five weeks gestation with a birth weight of five hundred and thirty grams, and died less than two hours after birth despite intensive efforts at resuscitation. The average birth weight was 1.7 kg (Range 0.5 kg to 2.7 kg). Mean gestational age was 32.9 weeks (Range 25 to 40 weeks). Using the Gairdner Pearson growth chart [[11]] as reference, intra-uterine growth retardation (IUGR) was noted in ten neonates.

Duration of hospital stay averaged 23.7 days. (Range 5 to 124 days) Neonatal ICU admission was required in 46.2% of the babies with half of these requiring artificial ventilation. Hypoglycemia and hypothermia, related to prematurity were the reasons for ICU admission in the remainder.

Thrombocytopenia defined as <150 × 10⁹/L in the newborn, was present in three (23.1%) and leucopenia defined as <10 × 10⁹/L in one of these (7.7%). None were found to be anemic.

DISCUSSION

The HELLP syndrome continues to be mistaken for other medical conditions with potentially serious consequences [[3]], [[5]]. Though our understanding of the pathogenic mechanisms involved has improved [[8]], [[10]], uncertainty regarding diagnosis, management and prognosis persist [12-13].

We found the HELLP syndrome to be uncommon among Irish women. This observation may be partially due to the low incidence of PET that we have reported previously in an Irish population [[14]]. However, our study also shows that though the majority of those in the HELLP group had proteinuria and all were hypertensive, only half the women suffered from PET. It follows therefore that PET is not a prerequisite for the development of HELLP. Our findings also suggest, however, that hypertension and protein uria can be used to identify patients at increased risk of developing the HELLP syndrome.

Using the thrombocytopenic scale proposed by Martin et al, the majority of our patients had class 1 and 2 disease, suggestive of an aggressive process. Yet, there was zero percent maternal mortality rate, a low rate of DIC, absence of liver rupture and no intracranial bleed. Indeed, the degree of thrombocytopenia did not affect the outcome.

Interestingly, although half our patients developed renal dysfunction, all recovered normal renal function without the need for dialysis. Recent reports show that HELLP syndrome is the most common cause of acute renal failure in pregnancy [[6]]. Our findings suggest that early diagnosis and intervention can prevent the requirement for dialysis in these patients.

Our results, in terms of maternal and fetal survival, compare favorably to those of other series [2-3], [[15]]. This finding suggests that even if adequate control of hypertension is gained, there is little advantage in prolonging the pregnancy once fetal maturity is achieved. Administration of steroids lent no added advantage to the maternal outcome. This is in contrast to the findings of some authors, despite similar dosing schedules [[16]]. When the decision was made to deliver the fetus, Cesarean section was preferred in those patients in whom (a) a trial of induction was considered too great a risk, (b) there was fetal distress or (c) induction of labor had failed.

Sibai et al are to be commended for their attempts to standardize the laboratory definition of the HELLP syndrome. However, we have previously commented on serum levels of liver transaminases (LFTs) being lower in women with uncomplicated pregnancies than in the general population [[17]]. This observation, together with our present finding that a number of women had LFT elevations over twice that of baseline while not reaching the 70 i.u./L level, suggests that this diagnostic criterion could be revised downwards. We also found that there was a mean fall in platelet count of 158 × 10⁹/L in those who did even not reach the 100 × 10⁹/L “thrombocytopenic threshold” suggested by Sibai. A level of 150 × 10⁹/L as suggested by Geary [[18]] might be more realistic. Indeed, acute trends in these laboratory criteria rather than absolute values were very helpful to us in the early detection of the HELLP syndrome. And, since delay in diagnosis continues to result in adverse outcomes [[5]], [[12]], these modified parameters may well translate into earlier pharmacological or obstetric intervention and better patient outcome. It was also noteworthy that 10 of our 13 patients had abnormal coagulation profiles, though only 2 had more than twice the laboratory reference.

In this study, hypertension did not increase after the second postpartum day. No worsening of platelet counts was seen after the fifth postpartum day and all patients had counts >100 × 10⁹/ L by the seventh day. Similarly, there was no deterioration of either ALT or AST levels after the sixth postpartum day and complete normalization was achieved within 21 days. Coagulation abnormalities were common in the perinatal period but did not worsen after the first postpartum day. We found laboratory trends in the postpartum period helpful in confirming the resolution of the pathologic process. Persistence of these abnormalities should alert the physician to the danger of ongoing thrombotic microangiopathy.

Neonates are adversely impacted upon by the need for premature delivery. The most common fetal complication of the HELLP syndrome was IUGR. With our approach, fetal mortality rates compare favorably with others results [12-13].

In conclusion, we have found that awareness of the condition, the subsequent screening of patients followed by intensive monitoring and early intervention when HELLP is diagnosed do translate into improved patient and fetal outcomes.