Ultrasound is the primary imaging modality, agreed by all authors because of the lack of ionising radiation, ready availability and low cost. Ultrasound techniques involve identifying renal tract dilation and stones on greyscale imaging, using colour Doppler to look for the ring-down artefact from stones and to evaluate ureteric jets into the bladder and using spectral Doppler to assess the resistance index of blood flow in the interlobar arteries of the kidneys. Ultrasound is limited by the inability to view the whole ureter and by relative insensitivity to small calculi. Transvaginal ultrasound can improve visualisation of the distal ureter. Indeed, finding a dilated ureter below the level of the pelvic brim (where physiological compression is expected to occur) should alert the investigator to the possibility of a lower ureteric stone [19].

Dilation of the pelvicalyceal system alone is non-specific because of the effect of physiological dilation (Fig. 22.4). Left-sided symptoms and dilation have a higher correlation with stone disease because physiological dilation preferentially affects the right side. Renal pelvic diameter of 17 mm or less on an asymptomatic side effectively excludes the presence of stone.

Stones may be seen with ultrasound because of the focus of reflection with a posterior echo shadow (Fig. 22.5). Colour Doppler to show the ring-down artefact behind the stone can increase conspicuity (Fig. 22.6).

Indirect signs of the presence of a ureteral stone are a raised resistance index (RI) and an absent ureteric jet. The kidneys that have become obstructed by a stone have a raised intrarenal pressure that affects the blood flow through the kidney. This does not occur with physiological dilation. An RI over 0.70 or a difference in RI between the two kidneys of greater than 10 % correlates with the presence of a ureteral stone (Fig. 22.7). Limitations on RI are the confounding factors of prior renal disease, treatment with nonsteroidal anti-inflammatory drugs and scanning within 6 h of the onset of obstruction or more than 48 h after obstruction.

Unilateral absence of a ureteral jet into the bladder on the symptomatic side, as shown by colour Doppler, correlates with obstruction (Figs. 22.6 and 22.8). Care must be taken to image the woman in the opposite decubitus position to avoid compression of the ureter by the uterus [20, 21].

Quoted ultrasound sensitivities for stone disease in pregnancy vary widely and is clearly dependent on operator expertise and the thrust of any research study. Rates vary from about a third to nearly nine out of ten cases.

The question arises, what should be the next step when ultrasound has not answered the clinical problem and the patient is still symptomatic despite conservative treatment? The choices lie between CT, diagnostic ureteroscopy or MR scan.

CT is undoubtedly the best test in the non-pregnant patient for the detection of urinary tract stones. It also offers an overview of the rest of the abdomen and pelvis allowing alternative causes of flank pain to be detected. The contraindication to its use in pregnancy is the radiation dose involved. This is a relative contraindication as with low-dose techniques, the dose to the fetus can be reduced to 4 mGy without losing sensitivity for stones. At this low dose, it is arguable that the risk of radiation to cause either deterministic or stochastic effects (e.g. fetal loss, teratogenicity or childhood leukaemia) is so negligible as to be considered nothing. Many authorities still advocate its use in pregnancy as shown by the compendium of guidelines published by Austin and Frush [22]. It is also recognised that CT skills and scanners are more widespread than MR and that radiologists may have far more experience in interpreting CT than MR in the pregnant patient. However, if the alternative of MR is available, then CT should be considered as the last choice given its use of radiation [23]. Repeating CT scans during pregnancy is also frowned upon because of the cumulative radiation dose.

Ureteroscopy, with or without fluoroscopy, is held by many urologists to be the next investigation step after ultrasound. The advantage is the ability to combine diagnosis with treatment [24]. General anaesthesia is used in up to two thirds of cases. It is possible to place a ureteric stent or to perform pneumatic lithotripsy with some authors reporting 100 % stone-free rates after treatment [25, 26]. An element of caution is needed as although the maternal and fetal risks are low, they are not non-existent. The rate of complications increases, the more procedures are done. One series quotes a 4 % rate of preterm delivery [27]; another describes a case of ureteral perforation [28]. In view of this, it is prudent to maximise the non-invasive imaging information prior to embarking on ureteroscopy.

In an ideal world, if ultrasound and conservative management have failed to solve the problem, then MR should be chosen as the second-line imaging test ahead of CT or ureteroscopy. MR has the advantage of not involving ionising radiation. MR without intravenous gadolinium-based contrast agents is considered a safe technique: the American College of Radiology (ACR) advises that pregnant patients can undergo MRI without risks for the fetus. MR imaging at 1.5 T or lower magnetic field strength has been used to evaluate diseases in pregnancy for over 20 years without any documented harmful effects [29]. The first trimester is generally avoided as it is difficult to establish safety because of active organogenesis. Field strengths higher than 1.5 T have not yet been proven to be safe, but as yet in the relatively short time, 3 T has been in use, there have been no adverse events reported. Theoretical risks of acoustic injury or thermal damage are limited by the manufacturers’ inbuilt limits for each pulse sequence.

Gadolinium-based contrast agents cross the placenta and are excreted by fetal kidneys into the amniotic fluid. These are classified as category C drugs by the Food and Drug Administration (FDA), i.e. animal reproduction studies have shown an adverse effect on the fetus. Well-controlled studies in humans are not available at present, but potential benefits may warrant use of the drug in pregnant women despite potential risks. For this reason, gadolinium should only be administered using the smallest dose of the most stable gadolinium agent (macrocyclic) [30] and preferably within the context of a trial.

Calculi do not themselves generate any signal on MR scan. MR needs there to be a signal generating fluid or tissue around the calculus to enable the calculus to be detected as a signal void (Fig. 22.9). MR is able to detect the secondary signs of ureteric obstruction due to stone by being able to reveal the dilated ureter, the oedema in the kidney and the perinephric fluid caused by obstruction. It has been suggested by some authors that MR can be used to define the level of obstruction, and then a limited CT scan focused on that level used to detect the stone, thereby greatly reducing the radiation dose whilst still maintaining sensitivity for stone [31]. There are clearly logistical difficulties with this approach in being able to proceed to CT straight after the MR scan such that the concept has not been adopted by other centres.

There are only a few published series of the use of MR in pregnancy-related renal colic [32, 33]. None have evaluated sensitivity or specificity for renal calculi. Generally, the larger the calculus, the more likely it is to be detected on MR. Thin-section scans increase the conspicuity of smaller stones (Fig. 22.10). However, the main limitation to high-resolution MR imaging in the pregnant patient is fetal movement. As a consequence, sequences able to generate images in a second or two are preferred with longer acquisitions reserved for imaging the level of obstruction.

MR can be poorly tolerated by the pregnant patient because of claustrophobia or because of the need to lie in one position. Women in the third trimester often prefer to lie in a partial decubitus position as lying supine can allow the gravid uterus to compress the cava and cause fainting. A careful explanation of the examination and the use of wider bore scanners help to minimise the non-acceptance rate for MR.

Ideally, a radiologist should start by using coronal thick-slab SSFSE ‘MR urography images’ that provide a ‘road map’ of the urinary system (Fig. 22.11) looking for strictures or calibre changes and potential filling defects using fast imaging such as FISP and water-weighted sequences such as SSFSE and then to focus on these areas of interest using higher-resolution T2W imaging. The technique described in early reports using gadolinium as part of MR excretion urography [34] has not been adopted partly because of the unknown effects of gadolinium on the fetus and partly because it offers no advantage over MR urography using the intrinsic contrast available from urine. Showing delayed excretion on the affected side does not increase detection of stones (Fig. 22.12).