Urolithiasis in pregnancy. I: pathophysiology, fetal considerations and diagnosis
A.D. Joyce, Pyrah Department of Urology, St James' University Hospital, Beckett Street, Leeds LS9 7TF, UK.
‘For the sake of diagnostic precision we should distinguish between the sickness of pregnancy and sickness in pregnancy.’
William Smellie, 1752 .
Urolithiasis in pregnancy represents a major diagnostic and therapeutic challenge to the GP, obstetrician, urologist, radiologist and anaesthetist. Urolithiasis presenting during pregnancy is a cause of major concern, considering the potential adverse effects of radiation exposure, any invasive surgical procedures and anaesthesia on the mother and fetus. The incidence of urolithiasis during pregnancy is 0.026–0.531% [2,3], complicates 1 : 200 to 1 : 2000 pregnancies  and may be a contributing factor in up to 40% of premature births . The incidence of symptomatic stones has been calculated to be the same during pregnancy as in nonpregnant women of childbearing age . Multiparous women seem to be affected more often than primiparae by a ratio of ≈3 : 1 [3,6,7]. However, the incidence in multiparous women is no greater when adjusted for age . Folger  reported that pain resulting from urinary stones is the most common cause of abdominal pain requiring hospitalization during pregnancy. Left- and right-side calculi occur with equal frequency and ureteric stones seem to occur about twice as often as renal calculi. Interestingly, 80–90% of patients present in the second or third trimester of pregnancy, but rarely in the first trimester . The management of urolithiasis during pregnancy is often challenging, requiring close co-operation between urologist, radiologist and obstetrician. Fortunately, with conservative management, 70–80% of symptomatic calculi pass spontaneously with no sequelae .
Effects of pregnancy on the urinary tract
Normal anatomical and physiological changes that occur during pregnancy may predispose to kidney stone formation. Physiological hydronephrosis occurs in 90% of pregnancies, starting at 6–10 weeks of gestation, and generally resolves within 4–6 weeks of parturition . Hydronephrosis is caused by a combination of hormonal and mechanical effects. Progesterone affects the urinary smooth muscle during early pregnancy, causing decreased peristalsis and dilatation of the ureter, above the pelvic brim, with the right side more affected than the left because of compression from the right ovarian vein (engorged) and uterine dextro-rotation. Recent articles suggest that mechanical compression is the main, if not the only, cause for the dilatation. Studies have shown that dilatation is not seen when the ureter does not cross the pelvic brim, as in patients with pelvic kidneys or an ileal conduit [12,13]. Several physiological adaptations occur during pregnancy which may affect renal stone formation. The renal plasma flow increases in pregnancy, leading to a 30–50% increase in GFR, resulting in an increased clearance of creatinine, uric acid and urea. During normal pregnancy there is a physiological state of absorptive hypercalciuria, presumably caused by placental formation of 1,25-dihydroxycholecalciferol and suppressed production of parathyroid hormone. However, the filtered loads of citrate, magnesium and urinary glycosaminoglycans, which inhibit urinary lithogenesis, are increased . Pathological calcium oxalate supersaturation has been reported during pregnancy, but crystalluria is no more common than in nonpregnant woman . Thus, the relative percentage, type and frequency of urinary stones occurring during pregnancy are similar to those in nonpregnant stone-formers .
Fetal considerations of urolithiasis in pregnancy
Risk of radiation exposure
The most important factor complicating the radiological evaluation of stone disease in pregnancy is the risk of radiation exposure to the fetus. The principal effects of irradiation on the mammalian fetus include teratogenesis, carcinogenesis and mutagenesis. These are divided into two categories: (a) those effects that become more severe with increasing dose and for which a threshold is believed to exist, termed ‘non-stochastic’ (threshold), e.g. malformation, growth retardation and cataracts; and (b) those effects where the probability of the effect increases with increasing dose and which have no threshold, termed ‘stochastic’, e.g. cancer and genetic effects.
Much of the information about the effects of radiation in humans originates from the study of survivors of the atomic bomb, who were irradiated with high doses in utero in Nagasaki and Hiroshima . The risk associated with radiation depends critically on the gestational age and the amount of radiation delivered. During early organogenesis, the embryo is very sensitive to the growth-retarding, teratogenic and lethal effects of radiation. During the early fetal period, the fetus has diminished sensitivity to multiple organ teratogenesis but retains CNS sensitivity. During later stages the fetus is not grossly deformed by radiation but can sustain permanent cell depletion of various organs and tissues if the radiation exposure is high enough .
Radiation-induced fetal malformation.
The classical effects of radiation on the developing embryo are gross congenital malformation, intrauterine growth retardation (IUGR) and embryonic death. There is a linear, dose-related association between severe mental retardation and radiation, with the important caveat that most cases followed exposures during weeks 10–17 of gestation . Irradiation of the human fetus from diagnostic exposures of < 50 mGy has not been reported to cause congenital malformations or IUGR . Although the animal and human data support the ‘safe limit’ of 50 mGy, there is the theoretical possibility that functional or biochemical changes may be produced at low levels with a low incidence . Schull and Otake  estimated that there is an ≈0.03 point IQ loss per mGy arising from in utero exposure to ionizing radiation during the most sensitive period of human brain development. In animal experiments, at 10 mGy there was no observable histological effect in the developing brain, and behavioural studies were unable to show neurobehavioural changes at < 200 mGy . In animal models, doses as low as 100–250 mGy induced a variety of potentially relevant cellular effects (induction of apoptosis, p53 and nerve growth factors) in the brain. In general, protracted exposures were less damaging than acute exposures, and the earlier the in utero exposure the greater the effect .
Stewart  suggested that the human embryo was more sensitive to the leukaemogenic effects of radiation and noted that exposure to as little as 10–20 mGy was associated with a slight increase in childhood cancers, by a factor of 1.5–2.0. Lilienfeld  reviewed nine studies, of which six reported a 1.3–1.8-fold increase in the risk of leukaemia after diagnostic radiation exposure in utero. In contrast, Court-Brown et al. examined the relationship between diagnostic X-rays in pregnant patients and childhood cancers, and surprisingly neither study showed a statistically significant increase. At present, some believe that in utero exposure to small amounts of radiation increases the risk of leukaemia and other cancers, whereas others question the contention that the embryo is markedly more sensitive to the leukaemogenic effects of radiation than the child or adult. There is little disagreement with the concept that low doses of radiation present a carcinogenic risk to the embryo and adult, and that there may be different risks with dose at different stages of development .
Genetic effects primarily involve haploid germinal cells. In the general population genetic diseases occur in ≈11% of births and spontaneous mutations account for < 2–3% of genetic disease . The dosage required to double the baseline mutation rate is 500–1000 mGy, far greater than the radiation doses occurring in common radiographic studies . Measuring the genetic effects is difficult because of the high incidence of genetic birth defects inherent in the human population. It appears that radiation is weakly mutagenic and that inherited mutations are rare, especially at low radiation levels .
The radiation dose delivered and its effect on the developing fetus depend upon the equipment, the radiographic technique, the duration of fluoroscopy, number of films and the gestational age. For example, a plain film delivers ≈0.5 mGy to the fetus; Swanson et al. reported that standard IVU exposes the fetus to 3 mGy, whereas limited IVU delivers ≈2 mGy. The National Radiological Protection Board (NRPB) reported mean and maximum fetal doses from the most recent surveys of diagnostic radiology practice (Table 1). The lethal dose to fetal tissue is variable and increases from ≈100 mGy after conception to 500 mGy at the end of the first trimester . In the UK, it is recommended that an investigation resulting in an absorbed dose to the fetus of > 0.5 mGy requires justification . This dose gives a level of risk comparable with that from variations in natural background radiation found in the UK . The possible stochastic effects to the fetus after in utero irradiation are heritable disease and malignancy. The NRPB cites a risk for heritable disease and for excess fatal cancer, up to the age of 15 years, of 0.024/Gy (as 1 in 40 000 per mGy) and at 0.03/Gy (1 in 33 000 per mGy), respectively . During the first month spontaneous abortion is far more common than birth defects, and fortunately, first trimester presentation of urolithiasis is extremely rare. However, clinicians must therefore weigh carefully the risk-benefit ratio of an examination involving radiation during pregnancy during the first trimester.
Table 1. Fetal doses after common diagnostic uro-radiological procedures 
|99mTc kidney scan (DTPA)||1.5||4.0|
|99mTc MAG 3|| ||0.7|
Inhalation anaesthetic agents (nitrous oxide, halothane, cyclopropane) are lipid-soluble and thus easily cross the placenta; animal models have shown that these agents have significant teratogenicity. Exposure to volatile gas agents in the first trimester of pregnancy is estimated to carry a relative risk of 0.5% of a morphogenetic anomaly. Regional techniques of anaesthesia are therefore recommended in the first trimester, and general anaesthesia should be avoided if possible . If an elective procedure is required, it is recommended to temporize and defer intervention until the second trimester, when fetal risks are minimal.
Analgesic agents and antibiotics
Morphine sulphate and meperidine in small doses for episodic pain have had no adverse effect on the fetus, but chronic use of these agents can lead to fetal narcotic addiction, IUGR and premature labour . Compounds containing codeine have been shown to have teratogenic effects when used in the first trimester, but may be used in the second and third trimester for short intervals with little fetal risk . NSAIDs block prostaglandin synthesis and therefore may lead to premature closure of the ductus arteriosus in utero, and should thus be avoided in pregnancy. Patients taking aspirin for analgesia have delayed and prolonged labour as it decreases uterine contractility. Because aspirin decreases platelet aggregation there is an increased risk of bleeding before and after birth. In a neonate exposed to aspirin in utero, platelet dysfunction has been reported 5 days after delivery. No evidence of teratogenicity has been reported for other drugs (e.g. ibuprofen and naproxen), and short courses would be appropriate for up to 48 h, if indicated. Chronic use may lead to oligohydramnios and constriction of the fetal ductus arteriosus. When mild analgesia is needed, acetaminophen should be preferred over aspirin, because it does not prolong the bleeding time in pregnant patients and it has not been shown to be toxic to new-borns. Propoxyphen is an acceptable alternative.
Patients undergoing other than obstetric surgery may require antibiotic treatment and the antibiotics of choice are penicillins and cephalosporin, which have not been associated with any adverse effects; and erythromycin, which is also well tolerated without fetal morbidity, although erythromycin estolate salt compounds should not be used in pregnancy because they can cause cholestatic jaundice in pregnant females . Aminoglycosides, tetracycline, chloramphenicol, fluoroquinolones and sulpha drugs are contraindicated in pregnancy because they have adverse effects on the fetus.
The most common presenting symptoms and signs of urolithiasis are flank pain, gross or microscopic haematuria and urinary infection . The aetiology of loin pain in pregnancy includes both general abdominal conditions and the possible obstetric major complications of pregnancy. Stothers and Lee  reported an incorrect diagnosis of appendicitis, diverticulitis and placental abruption in 28% of patients in whom a stone was subsequently confirmed; this emphasizes the diagnostic difficulties of patients presenting with urinary stone disease in pregnancy. Alder's sign  (palpation eliciting pain disappears with a change in position) may be helpful in differentiating between pain and tenderness of gynaecological and non-gynaecological origin. Microscopic and gross haematuria occur in 75% and 15% of cases, respectively . Patients may present with UTI, irritative LUTS and rarely with pre-eclampsia. Bladder stones are reportedly rare during pregnancy. Cope  reviewed published reports and identified 30 cases of bladder calculi in pregnancy, of which 24 were diagnosed during labour. A bladder stone has even been reported to cause a vesicovaginal fistula during pregnancy .
The most germane questions when presented with a pregnant patient with suspected urolithiasis are how to evaluate the problem, choose the appropriate management and when to intervene surgically. Before the advent of ultrasonography (US) the diagnosis centred on plain radiography of the abdomen, and IVU, but because of the concerns about the radiation risks during pregnancy, US has now become the standard first-line investigation.
Real-time US has revolutionized obstetrics and become the most commonly used imaging method in pregnancy; it has also become the cornerstone of the diagnostic evaluation of suspected renal colic in pregnancy. However, with US it can be difficult to differentiate the physiological dilatation of pregnancy from ureteric obstruction, and it is therefore of limited value in cases of acute obstruction. Stothers and Lee  reported a 34% sensitivity and 86% specificity for US in detecting abnormal findings in the presence of stones. Several measures have been recommended to enhance the performance of US.
Muller-Suur and Tyden  measured the renal pelvic diameter and recommended renography in symptomatic patients with a renal pelvic diameter of > 17 mm. In contrast, Erickson et al. found a maximum pelvic diameter of 27 mm on the right side and 18 mm on the left side during last two trimesters of pregnancy in asymptomatic patients.
Colour Doppler imaging.
MacNeily et al. suggested that the presence of a dilated ureter below the crossing point of the iliac artery is strong evidence of pathological distal ureteric obstruction.
Resistive index (RI).
Doppler US, and more specifically the RI, have been proposed to increase the ability of US to identify urinary tract obstruction. Many authors have proposed that acute or chronic urinary tract obstruction would change the RI in the affected kidney by increasing renal vascular resistance [43,44]. However, other studies have reported RI to be of little use in identifying acute renal colic. Tublin et al. reported a 44% sensitivity and 82% specificity for RI in identifying ureteric obstruction. During acute ureteric obstruction defined changes occur in renal haemodynamics and an important question is whether these changes affect the renal RI. Opdenakker et al. calculated renal RI in 72 patients, referred to the emergency department with acute renal colic, with no known associated renal disease. There was no statistically significant difference within the first 6 h but at 6–48 h the RI in the affected kidney was significantly different from that in the normal kidney. They also reported that after 48 h the sensitivity of RI decreased substantially. Furthermore, Shokeir et al. showed that NSAIDs significantly decreased the RI of the acutely obstructed kidney. Therefore, if precise anatomical and physiological information is needed, the RI cannot be used as a single method for assessing a symptomatic patient with hydronephrosis. Interestingly, in a recent clinical study, Shokier et al. measured the RI and change (ΔRI) in pregnant women with acute unilateral ureteric obstruction caused by a stone, reporting a sensitivity of 45%, a specificity of 91% and an accuracy of 87% for RI. The corresponding values for ΔRI were 95%, 100% and 99%. Despite these efforts, using the RI to detect obstruction remains controversial.
The characterization of ureteric jets may be useful as an ancillary technique. These jets can be visualized by real-time US or colour Doppler US. Deyoe et al. reported that a complete obstruction can be diagnosed with a sensitivity of 100% and a specificity of 91% if there are no ureteric jets detectable on the suspected side of obstruction. In contrast, Burke and Washowich  reported complete unilateral absence of jets in four asymptomatic pregnant patients, and recommended a cautious interpretation of this approach.
Another technological advance that may enhance the diagnosis is the use of transvaginal US. Laing et al. detected 13 distal stones using transvaginal US but only two using the transabdominal approach.
As stated in a regulatory guide , exposure to any level of radiation is assumed to carry some risk. In the absence of scientific certainty about the relationship between low-dose exposure and health effects, any exposure to ionizing radiation may cause undesirable biological effects. The use of radiation for diagnostic studies during pregnancy has been and remains controversial. With proper planning of exposure, the use of tight collimation, low voltages (60–70 kV), a brief exposure, high-speed screen and prone positioning may decrease the radiation dose delivered to the fetus .
Various investigators have suggested that modified or limited IVU can be used to decrease the radiation dose to the fetus, and even be considered safe during the first trimester [4,10]. However, there are data that cause concern with even low doses of radiation exposure to the fetus. Modified IVU has been varyingly defined (Table 2) [2,52–54]. Another limitation of IVU in pregnancy is the difficulty in differentiating delayed excretion of the contrast material associated with physiological dilatation from that associated with obstruction caused by calculus. Furthermore, an enlarged uterus and fetal skeleton may obscure small stones. Although no adverse effects of contrast media on fetal development have been reported, exposure to such agents should be avoided. Ideally, the use of radiographic techniques like modified or limited IVU should be discouraged, as X-rays present inherent risks of ionizing radiation, as does injection with contrast medium to the fetus.
Table 2. The reported modified IVU protocols
|||Two-exposure limited IVU, second film at |
|||Plain film + 20 min, ± delayed films|
|||Plain film + 15 min, if obstruction then 60 min film|
|||Plain film + 1 min + 15 min; faint nephrogram on |
the 1-min film and no excretion on the 15-min
film, delayed films at 120–180 min; dense
nephrogram – further film at 45–60 min
The administration of a radioisotope to a pregnant woman will result in exposure of the fetus to radiation emitted from adjacent maternal organs and from any radioactivity transferred across the placenta. Renography delivers ≈10% of the radiation dose of IVU. The USA Nuclear Regulatory Commission developed a guide to calculate the radiation dose to the embryo/fetus . For 99mTc-labelled radiopharmaceuticals, the absorbed doses is 0.2–1.8 mGy . Importantly, the radioisotope is excreted in urine and the bladder reservoir component will act as a significant source of exposure to the fetus. Therefore, to minimize the radiation risk to the fetus, the patient should be encouraged to maintain a high fluid intake and void as frequently as possible. Renography provides a physiological approach to diagnostic evaluation and its safety has been confirmed . The present authors' protocol is to assess patients with suspected ureteric obstruction in pregnancy with US followed by isotope renography.
MR urography (MRU)
MRU can be used to evaluate the urinary tract without using ionizing radiation and with no administration of contrast medium; this has important considerations for patients in pregnancy. MRI is confirmed as a safe imaging method, with no known teratogenic effects . Roy et al. reported excellent accuracy (sensitivity 100%) using ‘rapid acquisition with relaxation enhancement’ MRU. Using MRU it is possible to differentiate a physiological from pathological ureteric dilatation during pregnancy, but it is an expensive technique and is of limited availability. Thus it should be reserved for special cases when US fails to provide the diagnosis. Recently, Spencer et al. reported the use of gadolinium-enhanced breath-hold gradient-echo MR excretory urography to assess symptomatic hydronephrosis in pregnancy. They compared MR excretory urography with ‘gold standard’ isotope diuretic renography in 11 symptomatic pregnant women, and reported a good correlation between the assessment of excretion from symptomatic kidneys for isotope and MR studies. MRU should be considered the procedure of choice when US fails to establish a diagnosis, and we expect this technique to gain more widespread use.
Other imaging methods
Retrograde pyelography is of limited value because of the risk of sepsis. Several recent studies showed that unenhanced helical CT is a safe, rapid and highly accurate technique for evaluating acute flank pain. However, the radiation dose of CT, and particularly pelvic CT, can be high, thus precluding its routine use during pregnancy.
Urolithiasis during pregnancy, although rare, is a difficult clinical problem, requiring co-operation between obstetrician, urologist and radiologist. Flank pain and haematuria are the common presenting symptoms. The differential diagnosis of flank pain during pregnancy is vast. The major differential diagnosis is between non-urological and urological pathology, for which a thorough medical history and physical examination is helpful. An extremely important part of the evaluation is urine analysis. In addition, US is useful in resolving the diagnostic dilemma at the time of initial assessment. US combined with measurements of renal vascular resistance and ureteric jets appears to be helpful, but the clinical significance of such new approaches remains to be determined. If US fails to detect a calculus in a symptomatic patient with hydronephrosis, isotope renography or MRU is useful in delineating the level and grade of obstruction. If these are not available then insertion of a stent or nephrostomy tube is justified, as it relieves symptoms immediately. Both limited IVU and CT are excellent tools in evaluating ureterohydronephrosis, but there is a potential hazard which demands avoiding radiation if possible.
The recommendations from key organizations that may help clinicians better understand the overall risks from X-rays and other diagnostic imaging methods are:
- • X-ray imaging: NRPB, Royal College of Radiologists, UK: ‘Radiation doses resulting from most diagnostic procedures in an individual pregnancy present no substantial risk of causing fetal death or malformation or impairment of mental development’.
- • National Council on Radiation Protection: ‘Fetal risk is considered to be negligible at 50 mGy when compared to the other risks of pregnancy, and the risk of malformations is significantly increased above control levels at doses > 150 mGy' .
- • American College of Obstetricians and Gynecologists (ACOG): ‘Women should be counselled that X-ray exposure from a single diagnostic procedures does not result in harmful effects. Specifically, exposure to < 50 mGy has not been associated with an increase in fetal anomalies or pregnancy loss’.
- • US: American Institute of Ultrasound in Medicine: ‘Mammalian bioeffects are not seen below a SPTA intensity of 100 mW/cm2’. ACOG: ‘There have been no reports of documented adverse fetal effects for diagnostic ultrasound procedures, including duplex Doppler imaging’.
- • MRI: ACOG and NRPB: ‘Although there is no evidence to suggest that the embryo is sensitive to magnetic and radiofrequency at the intensities encountered in MRI, it might be prudent to exclude pregnant women during the first trimester’[60,61]. Safety Committee of the Society for MRI: ‘MRI is indicated for use in pregnant women if other non-ionizing forms of diagnostic imaging are inadequate, or if the examination provides important information that would otherwise require exposure to ionizing radiation’.
- • Radionuclide scintigraphy: Administration of Radioactive Substances Advisory Committee: ‘Special attention should be given to the optimization of the exposure, taking into account the exposure of the expectant mother and the unborn child’. US Department of Health and Human Services: ‘Consideration should be given to technical modification of the procedure that will minimize fetal radiation exposure’.
The effects of low doses of ionizing radiation, such as those typically involved in diagnostic radiology procedures, are uncertain . A discriminatory approach on the part of the physician with careful weighing of the risk : benefit ratio is warranted.
We thank Dr EJ Will, Department of Nephrology, Dr HC Irving and Dr J Spencer, Department of Radiology, St James's University Hospital, Leeds for their assistance.