A 35 year old woman with diabetic nephropathy who wants a baby: case progression

Last week (18 September, p 674), we discussed the case of Maria Tomasa who had requested assisted fertilisation despite being advised against pregnancy because of her nephropathy. We discussed her situation within our teams. The opinions were conflicting, but no one had direct experience of a similar case. However, we shared a common view on a patient's right to pursue her pregnancy and on pregnancy being a major determinant of a person's quality of life. Furthermore, we were aware that Tommy had chosen to be under our care because we were sympathetic towards a pregnancy. We therefore did a systematic search to retrieve relevant data.

A search on Medline and Embase in December 2000 (updated in September 2003 for this report) found no information about the specific problem (table). We therefore searched for “in vitro fertilisation” combined with diabetes or nephrotic syndrome. Overall, 221 citations dealt with assisted fertilisation and diabetes, but only nine also included nephrotic syndrome (see bmj.com). There were several examples of in vitro fertilisation in serious conditions, including cystic fibrosis, HIV infection, systemic lupus erythematosus, and thalassaemia major. There were four cases of diabetic women, none with severe nephropathy.

Table 1.

Patient's renal function

Variable	Value
Serum creatinine (μmol/l)	133
Creatinine clearance (ml/min)	57
Proteinuria (g/day)	2.6
Serum albumin (g/l)	38
Total protein (g/l)	68

We still had too few data for a precise quantification of the risks. We therefore substituted “in vitro fertilisation” with generic terms referring to pregnancy. This wider search on maternofetal risks retrieved 382 titles. Of the relevant ones, 27 were primary studies on risks of pregnancy in diabetic patients with “nephropathy” (see bmj.com).

In summary, the reported risks of pregnancy in diabetes vary widely, reflecting different support therapies, definitions, and patient selection. For example, the definition of standard versus intensive insulin treatment has changed over time and varies by area. Congenital malformations range from 2.8% to 9.4%, perinatal mortality from 1% to 4.3%, preterm delivery from 35% to 62%, intrauterine growth retardation from 2% to 4%, and pre-eclampsia from 6% to 42%. Only 11 studies dealt with overt diabetic nephropathy with nephrotic proteinuria; the number of cases was small, and all risks were higher than for women with early diabetic nephropathy.

Assisted fertilisation carries an additional risk of medical complications. The delivery rate ranges from 13.2% to 37.4%. The hormonal conditioning and semisurgical techniques required would probably further increase the risk of medical complications and reduce the success rate. However, evidence concerning its application in patients with different diseases is mostly in the form of case reports; considerable publication bias can be expected since mainly favourable cases are published.

Questions

1. How strong is the evidence for or against assisted conception in this case?

2. How would you help Tommy reach a decision and ensure that she gave informed consent if she decided to have treatment?

3. Is there a moral difference between helping Tommy to conceive and supporting a spontaneous pregnancy?

Please respond through bmj.com, remembering that Tommy is a real patient and that she and her doctors will read your response

There is no evidence to allow quantification of the additional risks of medical complications. Moreover, there is no evidence based recommendation of a risk threshold above which no attempt should be made.