July 2020: Oncology and Transplant ID

Serum Beta-d-Glucan is a poor test for PJP pneumonia in patients with cancer.  COI disclosure that this study comes out of MD Anderson, one of the places I trained as a fellow, and I know Dr. Svalb and Dr. Kontoyiannis in that capacity.  The B-d-glucan assay (aka BDG or Fungitell) identifies a fungal cell wall component shared by many fungi and has proven a sensitive and useful diagnostic test for PJP pneumonia in patients with advanced HIV.  However, patients with advanced HIV develop PJP infections with a high fungal burden, raising the question of whether BDG is diagnostically useful in other hosts and clinical settings.

The authors reviewed patients at their large academic cancer center who 1) did not have HVI, 2) had a PJP qPCR assay performed on either a BAL or bronchial washing for workup of pulmonary infiltrates, and 3) had a serum BDG assay during the same period.  They excluded patients who received empiric treatment for PJP more than 4 days before testing and those who had the qPCR and BDG assays performed more than a week apart.  They used modified Mycoses Study Group criteria to classify PJP as definite, probable, or possible. Patients with a negative qPCR and negative microscopy from BAL or bronchial washings were randomly selected from all patients undergoing such testing as controls.

Over a 5 year period the authors identified 101 patients with both a positive qPCR and BDG testing; these were compared to 74 patients with negative qPCR and microscopy (i.e. controls) who also had BDG testing.  Of these, 22 cases were classified as definite, 51 as probable, and 28 as possible.  Definite/probable cases were more likely to have received corticosteroids than possible cases or controls (70% vs 42%; p=0.012 and 70% vs 32%; p<0001) to have had higher qPCRs than possible cases (median 1300 copies/ml vs 272 copies/ml; p<0.001), and to have had higher serum BDG levels than possible cases or controls (354 pg/mL vs 31 pg/mL; p<0.001 for both comparisons).  In all qPCR positive patients, serum BDG and qPCR values correlated modestly (Spearman’s p=0.38), which improved after excluding patients with other known invasive fungal infections or recent IVIG use (p=0.47); however, the correlation actually weakened when only definite/probably cases of PJP were considered, both before and after controlling for other IFIs and IVIG (p=018 and p=0.29).

The sensitivity of the serum BDG for PJP versus a positive qPCR was poor at thresholds of >80, >200, and >400 pg/mL, and this was only modestly better when PJP positivity was defined as definite/probable PJP with more generous qPCR cutoffs. Even with a qPCR positive defined as >2000 copies/ml and a BDG positive defined as >80 pg/mL, the sensitivity was just 77% with a specificity of 78% (NPV 99% and PPV 11%).  I suppose the cheerful way to look at this is that serum BDG still has a great NPV because the prevalence of PJP in this population is so low (just 3.3%). On the other hand, that’s not much better than doing no testing and just deciding the patient doesn’t have PJP based on the low prevalence, so it doesn’t seem the BDG is actually adding much diagnostic value in this scenario. 32650108

Cutaneous disease following solid organ transplantation affects recipients of all ethnic backgrounds, and some diseases disproportionately affect certain ethnicities.  I started medical school in 2008, and when we were taught dermatology in preclinical lectures it was almost entirely with white skin – even for diseases that are more common in darker skinned folks, like hidradenitis suppurativa, and despite training in a majority-minority city.  In fact, the only images of black and brown skin I recall us being shown were genital lesions in the lectures on sexually transmitted diseases, which I think probably says something about whose bodies our society assigns dignity or considers to have inviolable privacy. 

Anyway, the upshot is that physicians are relatively bad at thinking about skin disease in their non-white patients.  Skin disease – and particularly cutaneous malignancy - is a known complication of solid organ transplantation.  The authors of this study collected data on skin disease in patients attending two post-transplant skin surveillance clinics: one in London, and the other in Philadelphia.  Their goal was to identify what types of skin disease occurred in SOT patients of various ethnicities, and whether some diseases were more prevalent in some populations vs others.  They reviewed records of 1766 patients (64% male); 1024 (58%) were white, 376 (21%) Black, 261 (15%) Asian, 57 (3%) Middle Eastern or Mediterranean, and 48 (3%) Hispanic.  Ninety percent of these transplants were kidney or kidney-pancreas transplants; 4% were liver, 3% lung or heart, and the rest various sequential organ transplants.  Median age at transplant was 47 years and median time since transplant was 9 years, with white and Middle Eastern recipients having had their transplants longer than Blacks and Hispanics.

The authors found at least one skin condition in 76% of patients, and cutaneous malignancy in 26%.  Skin cancer affected white recipients (40%) versus Middle Eastern and Hispanic recipients (12% each) or Black or Asian recipients (7% each).  However, while basal cell carcinomas, squamous cell carcinomas, and melanomas largely occurred in white transplant recipients, Kaposi’s sarcoma more often occurred in black recipients and had a much shorter time from transplant to diagnosis than other types of skin cancer (15mo vs 8.4 years).  Viral diseases were common, with warts affecting 42% of recipients, and were more common among whites; however, cutaneous fungal infection were more common in recipients of Middle Eastern descent, particularly onychomycosis and non-pedis tinea infections.  Finally, 8% of recipients had a non-infectious dermatitis, and this was most common in Black transplant recipients.

The most valuable point I took from this article is that while skin cancer – the most feared cutaneous complication of organ transplant – is more common in white SOT recipients, it is by no means rare in other groups, who also have relatively higher rates of various non-malignant skin diseases. 32659869

Can you give isavuconazonium sulphate capsules via ET tube? Yes!  This study comes to us from Dr. Erin McCreary out of UPMC – she and her colleagues gave isavuconazole by enteral feeding tube to 19 SOT & HSCT recipients for either prophylaxis or treatment of invasive fungal infections.  After a median 7 days of such therapy, the median isavuconazole level was 1.8 ug/mL (range 0.3-5.2), and levels were similar to or greater than those achieved by IV administration in the six patients who also had therapeutic dose monitoring while receiving IV isavuconazole. 32710097