H.I.V. — Case-Finding & Diagnosis

The Good News — HIV Infection has become a manageable disease.  Some investigators dare postulate that patients who are fully adherent to their antiretroviral medication may well live normal life spans.  Personally, I find it much easier to care for & control HIV than diabetes, hypertension, & various other chronic conditions.

Not-So-Good News —  New infections keep occurring at similar rates as in the past.  Incidence has increased among 13-29 year-olds, the largest in age range 20-24.  Almost half of all new infections occur among Blacks.  And diagnosis is being delayed; a third of new cases progress to AIDS within a year.

The paradox — HIV specialists are well-versed in subtleties of case finding, but only deal with patients after-the-fact.  This posting will outline this information for primary care practitioners, and for specialists in other fields.

HIV infection evolves through three phases:

• Acute HIV (a.k.a. Primary HIV)
• Long “Asymptomatic” Period
• AIDS

An estimated 50% of transmission occurs during Acute HIV, when enormous amounts of virus circulate prior to sero-conversion (the “window period”).  The “Asymptomatic” Period, in quotes because we’ll discuss symptoms, is when most people will wind up diagnosed.  We won’t address symptomatic AIDS, because patients are so sick that HIV testing is an obvious consideration.

SCREENING

The CDC’s 2006 recommendations are to screen all persons 13-64 y.o. at least once in life.  The US Preventive Service Task Force resisted, but in a new 2013 draft has now signed on.

Screening should also be performed on anyone newly-diagnosed with an STD, active TB, or pregnancy.  Testing is done like for any other condition, no counseling required, no consent needed except opportunity to opt-out.  On occasion for a blood draw, ask, “Have you ever been tested for HIV?”  If not, then announce, “We’ll do one now, OK?”  No objection = didn’t opt out.  [NOTE:  some state laws may vary].

Repeated screening, at least annually, is recommended for certain high-risk people:

• New sex partner (or partner with new partner)
• Injection drug users & their partners
• People exchanging sex for money / drugs
• Partners of HIV-infected persons
• Pregnancy: repeat in 3^rd Trimester in high-risk locales & for women with risk factors

ACUTE HIV  (PRIMARY HIV)

Half of HIV-positive people recall a constellation of symptoms beginning 9 to 30 days from moment of infection, and lasting 1-3 weeks (80% if inquire in depth).  Virtually 100% have fevers.  Other common symptoms include a mononucleosis-like syndrome: Malaise, Sore Throat, sometimes Rash (click for pictures), maybe Arthralgias/Myalgias.  Atypical adenopathy (post. cervical, post. auricular, occipital) is a key clue.  Oral ulcers are highly specific, but fairly rare.

If you ever work-up mono, also “Think 1° HIV.”  A distinct minority present with aseptic meningitis (headache, stiff neck, cranial nerve palsies).  See Primary HIV for summary.

Since we’re in the Window Period, diagnosis is made by PCR.  The viral load is extremely high, often >1,000,000 copies.  Results in the mere tens-of-thousands suggest long-standing infection; a viral load of only a few hundred is likely false-positive [!!!].  The standard serological antibody test for HIV doesn’t sero-convert until:

• 1 month (50% of persons)
• 3 months (95%)
• 6 months (>99%)

The new 4th Generation antibody-antigen test, testing for both HIV Antibody and p24 Antigen [a part of the virus], may be positive around 18 days from infection.  Still, for diagnosing primary HIV, the viral load is most sensitive.

Viral loads are expensive; I only order them if a patient had risk factors in the last month.  Unfortunately, some people are not forthcoming.  But it’s hard to order a test when your patient denies risks.  So I settle for recommending serial antibody testing, but stress how Primary HIV is extremely infectious, & that anyone in the acute phase should abstain from sex.

“ASYMPTOMATIC” LATENT PERIOD

During the uneventful average of 8 years between infection and AIDS, a number of clinical clues may crop up, alerting us to the need to test.  They include risk factors, symptoms, non-AIDS illnesses, incidental findings on physical exam, and subtle laboratory abnormalities.  We’ll take these one by one.

Risk Factors

You know these, but it’d be embarrassing not to mention them here:

• Hx of Multiple Sexual Partners (esp. if unprotected sex)
• Hx of Injection Drug Use
• Hx Blood Products before 1987
• Partner with Hx of Risks

Symptoms

As the immune system becomes compromised, symptoms may develop.  Any of the following warrant an HIV test without further thought:

• Weight Loss
• Recurrent Fevers or Night Sweats
• Chronic Diarrhea

Anecdote (mentioned in a prior post):  In 1995, a 33-year-old woman, typical immigrant from rural Mexico with husband & baby, sought care at a major San Francisco medical center for chronic diarrhea.  Got worked-up for every possible bowel disease, infectious & otherwise, to no avail.  Frustrated, she returned to Mexico — doctors there heard, “Lady from San Francisco, chronic diarrhea!” & nailed the HIV diagnosis immediately.

Illnesses

AIDS-defining Opportunistic Infections don’t tend to occur until the CD4-Lymphocyte count drops below 200 / mm³.  However, certain illnesses (which occur in HIV-Negative persons as well) become more likely with a CD4 count <350.  These are:

• Recurrent Staphylococcal Skin Infections
• Recurrent Candidal Vaginitis
• Recurrent Bacterial Pneumonia / Sinusitis
• Non-Hodgkin Lymphoma  (not CNS)
• Active TB
• Invasive Cervical CA

Encounter any of these illnesses, & automatically order an HIV test.  Do note, however, that many people with “frequent sinusitis” really just have allergies; i.e. clinicians chickened-out & gave antibiotics.

Herpes Zoster (Shingles)  —  Virtually all my HIV patients get this at some point in their lives.  So does half the world, but usually with advanced age.  Whenever I diagnose Zoster in someone <50, I recommend an HIV test “just to be sure.”  It’s invariably negative, but still worth doing.  As opposed to the diseases above, the occurrence of Zoster is not prognostic, and can break out with high CD4 counts.

Incidental Findings on Physical Exam

Most people with HIV maintain some Lymphadenopathy from the time when virus first circulated throughout the body.  Atypical sites (not accounted for by minor mouth & throat infections) include posterior cervical, occipital, axillary, and epitrochlear nodes.  Ignore the inguinal nodes; they’re so common from subclinical foot and local skin infections that they weren’t even included in the first AIDS case-definitions (before we even knew what HIV was).

A word about the Epitrochlear Node, located just above the medial epicondyle of the elbow.  In the absence of chronic hand infections, virtually nobody has palpable epitrochlear adenopathy.  In pre-antibiotic days, a swollen epitrochlear node was considered pathognomonic of syphilis (the STD of the day).  Dermatologists (the syphilis experts) spoke of the “Syphilis Handshake”: at a cocktail party, firmly clutch right hands with a hearty “good to meet you,” cradling their elbow with your left, & furtively palpate.

Find a node, & they’d move on to somebody else for the night.  Today, it’s HIV instead of syphilis.  I check epitrochlear nodes on patients I examine for practically any complaint.

The Mouth — a wonderful place to discover HIV-related signs, all of which occur as CD4 counts decline, but before clinical AIDS.  They include:

** Thrush — White plaques on the palate & buccal mucosa.  In my experience, not particularly on the tongue.  Some thrush is inflammatory: small reddish erosions.

** Hairy Leukoplakia  —  Corrugated vertical white streaks on the sides of the tongue.  Whereas thrush occurs with a variety of immunocompromising conditions (even with steroid inhalers for asthma), hairy leukoplakia is pathognomonic of HIV.  Heaven knows who picked the name.

Ever since I diagnosed an unsuspecting woman in 1998 (index case for husband & 4-y.o. daughter, all still doing fine), whenever I examine the mouth, I say “Stick out your tongue, say ‘Ah,’ now point it to this side, to that side,” while I check.  Just be aware that hairy leukoplakia is perfectly mimicked by occlusal trauma — a pointy molar that scrapes the tongue.

** Kaposi’s Sarcoma (KS) —  No longer AIDS-defining, KS occurs at CD4 levels <350.  In the mouth it appears as an ecchymosis on the palate, maybe palpable.  Doesn’t need treatment, but is only found in HIV.

** Gingivitis (severe)  —  Not too common, but involves horrible gum erosion (far & beyond that occurring from poor hygiene).  Though your immediate thought is “dentist, stat,” be sure to discuss testing.

Click for pictures of Oral Manifestations of HIV.

Rashes — Lots of strange rashes occur as HIV infection advances.  If you need to send someone to Derm for diagnosis, get an HIV test as well.  Some of the more common rashes include:

**  Seborrheic Dermatitis  —  More than simple dandruff, HIV-related Seb Derm involves extensive scaly erythema of the forehead, eyelids, & down the face.

**  Eosinophilic Folliculitis  —  Crops of firm pruritic papules on the face, upper torso, or shoulders.  Is pathognomic of HIV; doesn’t occur in any other circumstance.

**  Kaposi’s Sarcoma  —  Can occur anywhere on the body.  Might be subtle, might be deep, infiltrative, & uncomfortable.  On the penis, can obstruct the urethra.  Only AIDS-defining if visceral (pulmonary or stomach).

Click for pictures of HIV-Associated Rashes.

NOTE — All the above conditions & manifestations (except adenopathy) respond to antiviral therapy.  I’ve had patients with terrible Staph skin abscesses, disfiguring Seb Derm, disabling KS, return to ongoing normal states of health.  The trick is initial diagnosis.

Incidental Findings on Laboratory Tests

At our relatively small primary care clinic we’ve diagnosed a number of HIV patients through pieces of paper.

**  Lymphopenia  —  I picked up a CBC result from an on-call provider, saw the absolute lymphs were low, found the chart & wrote “consider HIV test.”  The next time he came in, another provider noticed, ordered it, then went on vacation.  Lab tech came running a week later with the result, & nobody even knew the guy.

About 50% of lymphocytes have CD4 markers & 50% CD8.  The average CD4 count in uninfected persons is 1,000 / mm³ [very wide range], but as counts decline in HIV, so does the CD4 percent.  I tend to test when a CBC shows Absolute Lymphs <1,000.  So, for example:

• HIV-Neg:  1,000 Lymphs, 50% CD4
  • total CD4 = 500 (lowish, but normal)
• HIV-Pos:  1,000 Lymphs, 15% CD4
  • total CD4 = 150 (AIDS-defining)

There’s no data on sensitivity / specificity of using 1,000 Abs Lymphs as cutoff for HIV testing, but it’s a ballpark.

**  Elevated Serum Globulins  —  Globulins >4 g/dL suggests systemic inflammation, which may be due to a variety of causes.  We’ve diagnosed 2 cases of HIV just by noticing the abnormal lab on a Chemistry result.  High globulins are simply a marker, they’re not prognostic of Immunosuppression.  Clinicians know to order a serum protein electrophoresis (SPEP) for high globulins (r/o multiple myeloma or paraproteinemia), but don’t realize the need for HIV testing too.

However, lots of “Comprehensive Metabolic Panels” don’t include Globulin; you have to subtract the serum Albumin from Total Serum Proteins.  It can be tricky; click the link for Calculating Serum Globulins.

**  Thrombocytopenia  —  Some people develop an idiopathic autoimmune reaction in which HIV antibody attaches to antigens on platelets and destroys the cells.  It can happen at any stage of infection, with high CD4 counts.  Anybody with thrombocytopenia, even if not seriously low, warrants an HIV test.  I routinely order it if platelets are <100,000.

I’ve seen platelet counts drop <10,000 [one patient actually had “zero”!!!], and for some unknown reason they hardly tend to bleed.  Dentists & surgeons aren’t happy, however.  Prednisone boosts platelets temporarily, but treatment of choice is antiretroviral therapy.

**  Low High-Density Lipoprotein (HDL)  —  Anecdotally, lots of my HIV patients have HDL’s in the 20s.  A 1992 study found the mean HDL in HIV was 30 mg/dL, compared to 49 among Hiv-negatives.  I tend to test for HIV when I incidentally find an HDL close to 25; haven’t diagnosed anyone yet.

And that’s all we have to say about Case-Finding persons with Undiagnosed HIV Infection.  It sounds strange to say, “Hope you find some,” but every new diagnosis saves lives & prevents transmission.