Pregnancy and RA

Well, it’s been a long time since the last post! Sorry about that.

Here’s a topic that came up recently: pregnancy. It’s well known that pregnancy is a particularly dangerous time for women with lupus – flares of disease seem to increase, possibly related to elevated levels of female hormones (eg. estrogen and progesterone). There are a lot of theories about why this happens, and why women are disproportionately affected by lupus. In vitro, estrogen downregulates Fas-mediated apoptosis, and also can inhibit IL-2 production. However, the precise role of these hormones in regulating the immune system remain unclear (Moulton and Tsokos 2012).

What’s funny is that rheumatoid arthritis, another female-predominant disease, gets better with pregnancy, and flares post-partum. The same observations have been made with uveitis (Chan, CC et al. Br J Ophthal 2004). So – why?

Some progesterone-related effects have been invoked. For example, progesterone at pregnancy levels can inhibit Th1 and Th17 responses, skewing towards Th2 and also increasing Treg responses. IgG glycosylation patterns may also be affected. These effects have been observed both in vitro as well as in mouse models of collagen-induced arthritis (reviewed in Hughes, GC 2012).

The SKG mouse spontaneously develops rheumatoid factor-positive inflammatory arthritis, along with other RA-type manifestations such as pneumonitis and rheumatoid nodules. This arthritis remits completely with pregnancy, and is made significantly worse by ovariectomy (Inoue, K et al. 2013). The effects of ovariectomy are ameliorated by exogenous administration of estradiol and, to a lesser extent, progesterone. Treatment with estradiol decreased serum levels of TNFα and IL-6, while progesterone only affected IL-6 levels.

This study did not address what happens to male SKG mice – apparently they do not have as severe arthritis as gonadally intact females. However, castration (orchiectomy) of male SKG mice also worsens the arthritis phenotype, suggesting a protective role for testosterone as well! (Keith RC, et al. 2013). Huh.

Obviously, these observations are rather simplistic. If Th17 responses are important for the pathophysiology of both SLE and RA, why would progesterone be beneficial only for one? Lots of citations in this post, but I don’t feel any closer to understanding this complicated issue. Thoughts appreciated.

1. Moulton, VR and Tsokos, GC. Why do women get lupus? Clin Immunol. 2012 Jul;144(1):53-6.
2. Chan CC, Reed GF, Kim Y, Agrón E, Buggage RR. A correlation of pregnancy term, disease activity, serum female hormones, and cytokines in uveitis. Br J Ophthalmol. 2004 Dec;88(12):1506-9.
3. Hughes GC. Progesterone and autoimmune disease. Autoimmun Rev. 2012 May;11(6-7):A502-14.
4. Inoue K, Inoue E, Imai Y. Female sex hormones ameliorate arthritis in SKG mice.
Biochem Biophys Res Commun. 2013 Apr 11.
5. Keith RC, Sokolove J, Edelman BL, Lahey L, Redente EF, Holers VM, Sakaguchi S, Robinson WH, Riches DW. Testosterone is protective in the sexually dimorphic development of arthritis and lung disease in SKG mice. Arthritis Rheum. 2013 Mar 25.

Alopecia areata

Alopecia areata is associated with other autoimmune diseases, especially autoimmune thyroiditis and vitiligo, and to a lesser extent systemic lupus erythematosus (SLE) and diabetes mellitus. There are many case reports of alopecia areata occurring with other autoimmune conditions as well, including inflammatory bowel disease, psoriasis, and rheumatoid arthritis. As with psoriasis, there are reports of alopecia areata apparently induced by anti-TNF therapy for other conditions. There may be association with atopic disease too, particularly atopic dermatitis.

Pathophysiology seems to be T cell-mediated, though as with all autoimmune conditions the exact pathogenesis is unknown. Healthy hair follicles are normally an immune privileged site – follicular epithelial cells downregulate class I MHC and also secrete inhibitory cytokines that inhibit NK and T cell migration and function. In alopecia areata, this immune privilege is “collapsed”; there is increased MHC class I presentation of follicular autoantigens (although again the exact autoantigens are unknown), and increased NK and autoreactive CD8+ T cell infiltration. Interestingly, a recent study using a mouse model of alopecia in which NK cells were artificially depleted had the opposite result than expected – NK cell deficient mice had worse hair loss, not better. It therefore remains unclear whether the role of NK cells is protective, contributory, or both.

Nail pitting for rheumatologists typically brings to mind psoriasis, but it is also associated with alopecia areata.
In alopecia areata, the pitting is often more diffuse, in a grid-like pattern, so-called “Scotch plaid nails.” Here’s a pic I snapped, but it’s pretty hard to get a good picture of nails –
Nail pitting

Here’s a link to a better photo:

Diagnosing Pain Syndromes: A Pep Talk

by Andy Shulman, rheum school contributor

I’ve been thinking quite a bit about pain syndromes lately. Perhaps you have as well? No doubt, diagnosing a pain amplification syndrome once (or four times!) a day will have that effect. As a newly minted attending pediatric rheumatologist, I would imagine that this recent experience is not unusual. As I work my way through our program’s back log of heavily-triaged “non-urgent” new patient referrals, I’m encountering patients with a dizzying array of symptomatology and the single unifying diagnosis that we all know well. The David Sherry RND patient education handouts are flying off the shelves as fast as I can print them.

As I prepare to knock on the exam room door of a new patient whose records fairly scream the diagnosis, I’ve been refining two pep talks: one for the patients/families and one for me. I’ll do my best to share the latter one with you here.

Diagnosing Pain Syndromes Requires Clinical Acumen

By the time they make their way to our clinics or inpatient floors, patients with pain syndromes have seen many doctors and have been subjected to extensive testing. And yet they remain undiagnosed! Other physicians have heard the story, examined the patient, ordered work-ups and have been unable to make a diagnosis. You are about to do what they lacked the clinical acumen (patience or courage) to do: unravel a medical mystery as challenging as that of a rare autoimmune or inflammatory disease.

The referral packet (look at it, bulging there on the table!) is bloated with third and fourth-order imaging, laboratory and diagnostic procedures done to rule out this-or-that. It takes a great deal of medical knowledge, judgment and time to read through that intimidating and confused pile of reports and interpret so much varied testing accurately, always with the clinical context in mind. Every one of these patients teaches us important lessons in decision analysis, or what I like to call test logic.  Their stories remind us that ordering tests in the inappropriate clinical context with low pre-test index of suspicion has little diagnostic value and the darndest consequences.

I have been trying to bring out this concept for the residents and fellows who have accompanied me on my recent forays into pain syndrome diagnosis. Of course, the ANA and most tests in medicine have lousy performance when sent in the wrong clinical context. The less experienced clinicians who are spending time with you don’t understand this and need to learn it from you. Every physician needs to learn how to select testing based on the context, consider potential results and look ahead to how that information will help lead to a diagnosis and treatment recommendation. Absolutely, this is an opportunity to teach test logic, to reinforce the concepts of sensitivity, specificity, predictive value. These physicians-in-training are extremely fortunate to be able to learn this with you today. The experience of seeing a pain syndrome patient and hearing about their medical testing history might just give life to those abstract concepts and make the lesson stick.

 Diagnosing Pain Syndromes Requires Confidence

Telling a family that other physicians have diagnosed with various frightening diseases that they do not have a terrible disease takes enormous confidence. Since so many possible alternate diagnoses have been considered, the diagnoser of a pain syndrome, in effect, passes judgment over the other diagnoses, dismissing them one by one. The diagnoser puts all of the data into context, provides a radically new explicative pathophysiology and confidently accounts for the etiology of all of the symptoms. Confidence. This is the kind of confidence that comes from Jedi-knight level diagnostic, perceptive and interpersonal skills. Ever notice that our most revered senior colleagues, the people that make us well up with pride to be pediatric rheumatologists when we think of them, are so skilled at diagnosing pain syndromes? Not coincidence.

Diagnosing Pain Syndromes Requires Courage

In the most challenging pain syndrome cases (you’ve been there), many layers of confusion, fear and frustration self-reinforce an alternate reality. When we meet with these families and learn of their stories, we feel the strong pull of that vortex of confusion. We can become disoriented, the horizon can be lost. Our courage is tested.

The easy out, the option chosen by many of our colleagues in other disciplines, would be to send “our tests,” rule out “our diseases” and then pass the patient/family on to the next specialist to continue the cycle. Most doctors do this, but not pediatric rheumatologists. We should be enormously proud that as pediatric rheumatologists we do not engage in this kind of can-kicking. Instead, we speak truth to the confusion, at times, even to outright lies. This requires conviction that it is worth the time and the energy to actually address the patient’s problems in front of you rather than cowardly elect to punt. This requires hope that our recommendations will help people to get better, prevent disrupted lives from becoming ruined ones and stop irresponsible consumption of health care resources. This requires soulful listening, compassion and humor. This requires courage.

Hone skills. Build confidence. Take courage. Go forth and diagnose pain syndromes!

ACR 2012 – Tuesday

I finally attended a pediatric-specific session.

Bernatsky presented some data on the risk of malignancies in patients with pediatric-onset SLE.  She collated information from 10 North American pediatric rheumatology clinics, a total of 1020 SLE patients with an average of 8 follow up years per patient.  I missed the part about where the cancer data came from, but I think this was from cancer registries, and the diagnoses were subsequently verified manually.  She compared the total incidence rate of cancer development over these 8 years to the expected rate calculated from known demographic information controlled for age, sex, geographic location, and calendar year.

There were a total of 14 cancers observed, as compared to 3 expected – SIR 4.7.  Most occurred more than 10 years after the initial SLE diagnosis.  3 were hematologic malignancies (2 non-Hodgkin’s lymphoma, and 1 leukemia), although one was within 1 month of the SLE diagnosis, which makes me question whether she truly had SLE (as opposed to paraneoplastic symptoms and laboratory findings that were more related to her cancer.  The other malignancies were seemingly random – bladder, thyroid, breast, head/neck, and a few unspecified carcinomas.

I am not sure what this overall means.  Certainly it would not be surprising for SLE patients to have higher risk for hematologic malignancies – we know that there is increased white blood cell turnover and increased T cell oligoclonality in SLE.  In addition, some of the genes described in SLE pathogenesis have also been described to play a role in lymphoma/leukemias (eg. STAT signaling proteins).  However, it is less intuitive to understand how risk for other cancers might be increased.  Perhaps it is related to inefficient immune surveillance.  Or maybe it is related to cytotoxic therapies and secondary malignancies.  And I am not sure how knowing about an increased risk of cancer in SLE patients changes how I think about the disease pathogenesis or its treatment.  However, I agree that there is certainly much more to be learned on this topic.

ACR 2012 – Monday

Plenary Session

A couple of years ago, the ACR was all about the GWAS – it seemed like every other presentation was a GWAS describing a risk allele in a certain inflammation-related gene.  What has always been lacking is the translation of these described SNPs with their functional effects.  Robert Plenge from Brigham and Women’s Hospital in Boston described today how his group has addressed this.  First, he discussed a SNP in the CD40 promoter that is associated with RA.  His lab then developed a B cell line containing a luciferase construct with this risk allele-containing promoter, and showed that this promoter had increased activity (eg. higher CD40 expression levels) vs. the non-risk allele.  He then screened >2000 compounds for their effect on luciferase activity in this cell line.  He thus identified multiple  compounds, both known and recognized, that effectively decrease CD40 expression.   This is an interesting model for future drug development.

ACR 2012 – Sunday

ACR Sessions – “Extra-articular Manifestations of RA,” a somewhat misleading title as this session was primarily about pathogenesis…

Ted Mikuls presented data arguing for the role of Porphymonas gingivalis and periodontal disease in RA pathogenesis.  Evidence includes increased risk for RA in patients with periodontitis, and increased risk for periodontitis in patients with RA.  Many other studies cited, including collagen-induced arthritis in mouse models – phenotype is worse if infected with human P. ging

Kevin Deane argued for autoimmunity in RA originating in the lung.  The hypothesis for this is based on observations that lung disease precedes arthritis in some patients, and some patients with lung disease have RA-associated autoantibodies without having synovitis.  In these latter patients, one study found that 3 of 74 pts eventually developed RA.  Interestingly, immunohistochemistry staining of BALT tissue from RA patients show presence of RF and anti-CCP-autoantibodies.  Keane also notes that anti-CCP antibodies can be detected in the sputum (sputa?) of patients with RA.

Patrick Venables further expanded on the role of anti-citrullinated peptide antibodies (ACPA) in RA pathogenesis, covering some of the aspects described above.  One fact that I didn’t know was that the standard anti-CCP assay, while sensitive for >90% of anti-CCP antibodies, does not use a pathogenic antigen for detection.  Rather, pathogenicity of ACPA seems to be largely related to 4 citrullinated proteins:  cit-fibrinogen, cit-vimentin, cit-collagen, and (arguably the most important), cit-enolase.


I also attended the debate on rituximab vs. cyclophosphamide for ANCA-associated vasculitis.  The data was nothing new, but it was interesting to see two experts discuss their interpretation of the data and why they came to their personal preferences.  In general, though, it was not much of a debate as both experts generally agreed on treatment plans.  What did surprise me was the unequivocal preference by both experts for daily cyclophosphamide vs. intermittent IV cyclophosphamide infusions.  We very rarely use PO cyclophosphamide in our pediatric practice, as I think it is generally accepted that the toxicity profile is higher due to the higher cumulative dose (de Groot, et al. 2009).  However, relapse rates also appear to be lower with daily dosing, possibly related again to the higher cumulative dose.

ACR 2012 – Saturday

Live blogging from the ACR!

I am attending the Review Course this year, hoping to catch up on some general knowledge that I’ve missed along the way.  I have difficulty paying attention in lectures, so am taking the advice of an old med school professor who said, “Just learn ONE new thing from each lecture and focus on that.”

So –

1) Ankylosing spondylitis (Desiree van der Heijde) – In patients with AS and axial only manifestations, sulfasalazine showed no effect over placebo (40% response rate for each).  However, in patients with both axial and peripheral disease, there was a small effect on axial symptoms (32% sulfasalazine vs. 21% placebo), and a larger effect on peripheral symptoms (56% vs. 30%).

2) Sclerosing skin conditions (Heidi Jacobe) – One way to differentiate pansclerotic morphea from systemic sclerosis is the distribution/progression of involvement.  Pansclerotic morphea starts centrally and moves peripherally, classically with a well-demarcated ending at wrists/ankles; in contrast, scleroderma exam findings begin distally at the fingertips.  Lots of other good info in this lecture that is too much to capture here.

3) Inflammatory myopathies (Ingrid Lundberg) – There may be a role for IL-1 in the pathogenesis of inflammatory myopathies, and some patients may respond to anakinra treatment.

4) Treatment of SLE (Susan Manzi) – Hydroxychloroquine can have hypoglycemic effects, and there is some evidence that it may reduce the risk of diabetes in patients with RA.

5) Setting the stage for RA (Hani El Gabalawy) – Epitope spreading of anti-CCP antibodies *precedes* the development of clinical symptoms.  This concept (not new) of identifying and following the development of autoimmunity before overt clinical manifestations is fascinating, and of course gets to the fundamental question of whether we can *prevent* these diseases before they happen.

6) Hypermobility syndromes (Elizabeth Russell) – There are many more molecular/genetic associations with hypermobility than I realized, including defects/deficiency of tenascin X, zinc transporters, etc.  Multiple endocrine neoplasia type 2 (RET proto-oncogene) is associated with hypermobility.  Interestingly, the phenotype of Marfan syndrome (fibrillin 1) is thought to be related in part to overactive TGF-beta signaling, and RET is also involved in TGF-beta signaling.